Você está na página 1de 94

New Answers for

Cancer
Cutting-edge drugs and
research are helping
solve the puzzle
Cancer
Resource
Guide
Where to
turn for help
Stem Cells
The real culprits?
The Cancer
Genome
Mapping out
an attack plan
© 20 08 SCIENTIFIC AMERIC AN, INC.
letter from the editor
New Answers for Cancer is published

In This Issue
by the staff of Scientific American,
with project management by:

EDITOR IN CHIEF: John Rennie


E XECUTIVE EDITOR: Mariette DiChristina
is sue EDITOR: Dawn Stover
ART DIREC TOR: Patricia Nemoto
P H O T O G R A P H Y E D I T O R S : Emily Harrison,
Smitha Alampur n e a r ly 4 0 y e a r s since we de-
PRODUC TION EDITOR: Richard Hunt
clared war on cancer, how goes the
COPY DIREC TOR: Maria-Christina Keller
cOPY CHIEF: Daniel C. Schlenoff campaign against this intractable and
A s sis tant cOPY CHIEF: John Matson ancient adversary? As you will learn in
COPY AND RESE ARCH: Rachel Dvoskin,
Aaron Fagan, Aaron Shattuck, Kenneth Silber, this special edition, our enemy intelli-
Kevin Singer, Michelle Wright gence has improved over the years, en-
EDITORIAL ADMINIS TR ATOR: Avonelle Wing abling us to get a better bead on where
SENIOR SECRETARY: Maya Harty
the trouble begins. And we have devel-
A SSOCIATE PUBLISHER, PRODUC TION:
oped stronger weapons, to more pre-
William Sherman
MANUFAC TURING MANAGER: Janet Cermak cisely pursue and annihilate diseased
ADVERTISING PRODUC TION MANAGER: tissue.
Carl Cherebin
PREPRESS AND QUALIT Y MANAGER:
Silvia De Santis Finding Enemy Forces. Cancer’s ori-
PRODUC TION MANAGER: Christina Hippeli
CUS TOM PUBLISHING MANAGER:
gins are multifaceted, a combination
Madelyn Keyes-Milch of an individual’s genetic factors and
A SSOCIATE PUBLISHER, CIRCUL ATION: influences from the surrounding environment and his or her personal history and
Simon Aronin lifestyle. Even stem cells — which, in other contexts, offer promise for the treat-
CIRCUL ATION director: Christian Dorbandt
rene wal s manager: Karen Singer ment of a variety of ailments — could be to blame. To learn more, turn to page 40
FULFILLMENT AND DIS TRIBUTION MANAGER: for “Stem Cells: The Real Culprits in Cancer?” by Michael F. Clarke and Michael
Rosa Davis
W. Becker. And in “Mapping the Cancer Genome,” on page 22, Francis S. Collins
VICE PRESIDENT AND PUBLISHER:
Bruce Brandfon and Anna D. Barker explain how such a tool will help us chart a course across the
SALES DE VELOPMENT MANAGER: David Tirpack landscape of human malignancies.
SALES REPRESENTATIVES: Jeffrey Crennan,
Stephen Dudley, Stan Schmidt
Destroying the Targets. While scientists are grappling to gain a better under-
A SSOCIATE PUBLISHER, S TR ATEGIC PL ANNING:
Laura Salant standing of cancer’s complex beginnings, they also have improved ways of stalling
PROMOTION MANAGER: Diane Schube the advance of the disease. In “Taming Vessels to Treat Cancer,” on page 64, for
RESE ARCH MANAGER: Aida Dadurian
PROMOTION DESIGN MANAGER: Nancy Mongelli instance, Rakesh K. Jain describes how calming the chaos in tumors’ blood ves-
Vice president, Finance, and
sels could facilitate attacking them. Francisco J. Esteva and Gabriel N. Horto-
GENER AL MANAGER: Michael Florek bagyi explain how we are also “Gaining Ground on Breast Cancer” with tar-
BUSINESS MANAGER: Marie Maher
MANAGER, ADVERTISING ACCOUNTING AND COORDINATION:
geted therapies, beginning on page 88.
Constance Holmes
DIREC TOR, SPECIAL PROJEC TS: Hope in the Trenches. Patients with cancer, empowered by expanding informa-
Barth David Schwartz tional resources and the changing, more open attitudes of doctors, are living
DIREC TOR, special media:
Jeremy A. Abbate longer and better than ever today, as Lisa Stein writes in “Living with Cancer”;
SALES REPRESENTATIVEs, online:
see page 6. Although medicine clearly has much work to do, with advances oc-
k enn b r o w n M o n d o l i t h i c S t u d i o s

Gary Bronson, Thomas Nolan, Taylor West curring rapidly, we are on the path to managing this chronic disease.
DIREC TOR, ANCILL ARY PRODUC TS:
Diane McGarvey
PERMISSIONS MANAGER: Linda Hertz

CHAIRMAN: Brian Napack


PRESIDENT: Steven Yee Mariette DiChristina
VICE PRESIDENT AND MANAGING DIREC TOR, INTERNATIONAL: Executive Editor
Dean Sanderson
VICE PRESIDENT: Frances Newburg
Scientific American
CHAIRMAN EMERITUS: John J. Hanley editors@SciAm.com

w w w. S c i A m . c o m SCIENTIFIC A MERIC A N 1
© 20 08 SCIENTIFIC AMERIC AN, INC.
Contents Volume 18, Number 3, 2008

1 Letter from the Editor

6 Living with Cancer


by Lisa Stein
Keep up your spirits and tap available resources
to make the disease manageable.
Plus: Cancer Resource Guide

NEW UNDERSTANDING
OF A DREADED DISEASE
14 Evolved for Cancer?
by Carl Zimmer
Natural selection lacks the power to erase cancer
from our species and, some scientists argue, may
even have provided tools that help tumors grow.

22 Mapping the Cancer Genome


by Francis S. Collins and Anna D. Barker
Pinpointing the genes involved in cancer will help
chart a new course across the complex landscape
of human malignancies.

30 Untangling the Roots of Cancer


by W. Wayt Gibbs
Recent evidence challenges long-held theories 6
of how cells turn malignant—and suggests new
ways to stop tumors before they spread.

40 Stem Cells: The Real Culprits


in Cancer?
by Michael F. Clarke and Michael W. Becker 14
A dark side of stem cells—their potential to turn
malignant—is at the root of a handful of cancers
and may be the cause of many more. Eliminating
the disease could depend on tracking down and
destroying these elusive killer cells.

48 A Malignant Flame
by Gary Stix
Understanding chronic inflammation, which
contributes to heart disease, Alzheimer’s and
a variety of other ailments, may be a key
to unlocking the mysteries of cancer.

56 The Long Arm of the Immune System


by Jacques Banchereau
Dendritic cells catch invaders and tell the immune
system when and how to respond. Vaccines
depend on them, and scientists are even employ­
ing the cells to stir up immunity against cancer.

Cover image by Kenn Brown, Mondolithic Studios.


Articles in this special edition are updated from previous issues of Scientific American.

2 S C I E N T I F I C A M E R I C A N N e w ans w ers f o r C ancer


© 20 08 SCIENTIFIC AMERIC AN, INC.
Contents
NEW WEAPONS AGAINST A POTENT ENEMY
64 Taming Vessels to Treat Cancer
by Rakesh K. Jain
Restoring order to the chaotic blood vessels inside a 88
tumor opens a window of opportunity for attacking it.
Surprisingly, drugs meant to destroy vasculature can
make the repairs and may help reverse conditions that
lead to cardiovascular disease and blindness.

72 Tumor-Busting Viruses
by Dirk M. Nettelbeck, Ronald D. Alvarez and
David T. Curiel
A technique called virotherapy harnesses viruses,
those banes of humankind, to stop another
scourge—cancer.

80 New Light on Medicine


by Nick Lane
Pigments that turn caustic on exposure to light can fight
cancer, blindness and heart disease. Their light-induced
toxicity may also help explain the origin of vampire tales.

88 Gaining Ground on Breast Cancer20


by Francisco J. Esteva and
Gabriel N. Hortobagyi
The newest targeted therapies are helping doctors to ­tailor
increasingly effective treatments to individual patients.

64

Scientific American Special (ISSN 1936-1513), Volume 18, Number 3, 2008, published by Scientific American, Inc., 415 Madison Avenue, New York, NY 10017-1111. Copyright © 2008
by Scientific American, Inc. All rights reserved. No part of this issue may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying
and recording for public or private use, or by any information storage or retrieval system, without the prior written permission of the publisher. Canadian BN No. 127387652RT;
QST No. Q1015332537. To purchase additional quantities: U.S., $10.95 each; elsewhere, $13.95 each. Send payment to Scientific American, Dept. can2008, 415 Madison
Avenue, New York, NY 10017-1111. Inquiries: fax 212-355-0408 or telephone 212-451-8442. Printed in U.S.A.

4 S C I E N T I F I C A M E R I C A N N e w ans w ers f o r C ancer


© 20 08 SCIENTIFIC AMERIC AN, INC.
Living
with
Cancer
Keep up your spirits and tap
available resources to make the
disease manageable
By Lisa Stein

I
t was February 2003, and
Kris Carr, a photographer
and actress, was on a roll.
Overview The bubbly, green-eyed
stunner was in high demand.
■  Rather than surrendering to
She was considered “the Julia Rob-
despair and impersonal medical
treatments, growing numbers of erts of advertising” (at least accord-
cancer patients are empowering ing to her agent), thanks to her suc-
themselves with information and cess in two popular Bud Light com-
control over their therapies. The mercials that aired during the Super
trend is finding acceptance in Bowl. She also had some impressive
mainstream medicine and
theater and film credits, among
helping people with cancer lead
healthier lives. them a role in Arthur Miller’s Mr.
■ T he experiences of author and Peter’s Connections, in which she
filmmaker Kris Carr, who was performed (in the buff, no less)
diagnosed with a rare, incurable alongside actor Peter Falk.
malignancy, illustrate how
successfully one can manage Like many of her hip young com- Festival, where a film she had appeared
cancer as a chronic disease. peers, Carr, then 31, routinely burned in premiered, and she was dragging.
■ T he following resource guides the candle at both ends. She existed on Time to detox, cleanse her body and
l isa c o cciar d i ( p h o t o g r a p h)

offer tips on developing a energy bars, fast food and coffee soul, exercise and eat right for a spell.
strategy for managing the downed between nonstop auditions She swore off drinking for a month and
illness, asking the right and takes. Every so often her frenetic took a vigorous Jivamukti-style yoga
questions of physicians and lifestyle would catch up with her as it class to kick-start her new get-healthy-
getting the right professional did now: she had just returned home to quick scheme.
and personal support. New York City after “partying like a “The following morning I woke up
rock star” at Florida’s Sarasota Film feeling like I was hit by a truck,” Carr

6 S C I E N T I F I C A M E R I C A N ne w ans w ers f o r C ancer


© 20 08 SCIENTIFIC AMERIC AN, INC.
“ I don’t think anyone has a better life than me,”
says cancer patient Kris Carr.

You Have
Cancer:
Now What?
Diagnosis: cancer. Your head is
spinning, and you feel like the wind
has been sucked out of you. In a split
second, life as you knew it is gone.
“Getting diagnosed throws your entire
universe into a free fall,” Carr writes in
her 2007 book Crazy Sexy Cancer Tips.
“There’s no sugarcoating it: cancer is
a devastating blow, one that takes
time to process.”
The first things you should do (after
taking a deep breath and trying
to chill):
F ind the best doctor for your
disease: Be willing to travel and
always get second, third and even
fourth opinions to make sure that
you’re getting the best treatment.
Design a healing plan: Pull together
a team of Western physicians as well
as integrative doctors (to teach you
how to build up your immunity and
spiritual grit) to create the best get-
healthy recipe. Ask family and
friends to chip in and scour the
Internet and bookshelves for
information. “If you want to heal, you
have to take initiative, have a voice
and use it,” Carr says.
Focus on lifestyle changes: “The
only thing that you can control is
what you eat, what you drink and
how you move,” Carr says. She
recommends exploring healthy
diets, exercise and alternative
therapies such as massage, yoga
and meditation to boost and
maintain your physical and
emotional well-being.
says. Every muscle ached. She dis- least of Carr’s problems. Her pain had Create a support system: “Nobody
ser g i y timas h o v i S t o c k p h o t o (a l l i c o n s)

missed her sore body as a sign that she worsened, and it was now accompa- understands you quite like another
was more out of shape than she had nied by shortness of breath and severe cancer survivor,” Carr says. “There is
thought and, as usual, slipped into abdominal cramping. She made an ap- incredible strength in that.”
tight jeans, slathered on a mask of pointment to see her doctor the follow- Live! “Don’t wait for permission to
makeup and headed to an audition: a ing day. live. Just because you have cancer
commercial for a diet shake. (She didn’t Gallbladder trouble, the physician does not mean that your life is over,’’
get it: too fat, says the slender onetime surmised after a quick examination. Carr insists. “Start living. It’s that
model.) Recommended treatment: yank the simple.”
By evening, stiff muscles were the pear-shaped organ that, when healthy,

w w w. S c i A m . c o m SCIENTIFIC A MERIC A N 7
© 20 08 SCIENTIFIC AMERIC AN, INC.
helps the liver flush fats from the body It wasn’t. blissfully ignorant of the potential ram-
but, when faulty, causes excruciating “When they did the ultrasound, ifications. “I didn’t know,” she says,
pain. He gave Carr a prescription for they found the ‘lesions.’ They could see “that lesions meant tumors.”
painkillers and sent her for an ultra- there were spots all over my liver— so A battery of tests over the next few
sound to confirm that her gallbladder many that it looked like Swiss cheese,” days revealed that Carr was suffering
was indeed the culprit. Carr says. She was concerned but still from epithelioid hemangioendothelio-
ma (EHE), a vascular cancer in the lin-

Questions to Ask ing of the blood vessels in her liver and


lungs so rare that only 0.01 percent of
Studies show that cancer (and other) the cancer population has it. Around
patients who arm themselves with 200 to 300 cases are diagnosed nation-
information typically fare better and experience wide every year. The cause: unknown.
fewer side effects than those who simply follow The cancer was stage IV— incurable
doctors’ orders, no questions asked. Being informed and inoperable, the doctor said. “Some
gives them some control over their disease—and
people say it could have come on like a
that feeling of empowerment plays a role in the
meteor shower,” Carr says; others sus-
healing process. No. 1 rule: do not be cowed by your
pect the tumors had been developing
doctor. Ask him or her to explain anything and
her whole life.
everything you don’t understand. Prepare
questions in advance of appointments (to reduce EHE is typically a slow-moving
stress and the odds of forgetting any)—and bring a notebook to jot down answers and cancer. There are studies under way but
other important info. Below are some questions you should ask: currently no cures or definitive treat-
ments. The doctor recommended a
“watch and wait” approach. That is,
q What causes this type of cancer?
that they take their cues from the tu-
q What are the risk factors? If it’s genetic, are other family members at risk?
mors— monitor them for two months to
q What lifestyle changes (diet, exercise, rest) do you recommend? gauge whether they were holding steady
q What are my treatment options? or moving slowly or swiftly. They were
qAre there activities that should be avoided because they might trigger or quiet for now, “indolent” in cancer-
exacerbate symptoms? speak, and the hope was they would
qWhat happens if new symptoms crop up or existing ones worsen? stay that way.
qWhat medical tests or procedures are necessary? How often? It was February 14. “Happy Valen-
qWhat stage is my cancer? What does that mean?
qWhat is my overall prognosis or chance of recovery?
qWhat are the average survival and cure rates?
qCould my disease go into remission?
“How could I live
qWhat is the recommended treatment?
qHow often will I have to undergo treatment— and for how long? tine’s Day. You have cancer,” Carr
qWhat are the potential side effects? wrote in her journal that night.
qWhat are the benefits versus the risks of each treatment option?
qAre there alternative therapies? What are they?
Why Me?
“ i f e lt l i k e I was punched in the
qWhat are the expected results of treatment? stomach by God,” she recalls. “Cancer
qIs the treatment painful? If so, is there a way to make it more bearable? is such a frightening word. How could
qHow long is the recovery? Will it require a hospital stay? this be happening to me? Cancer hap-
qWhen can I resume my normal activity (if it’s been curtailed)? pened to other people. I was young and
qHas my cancer spread? If so, how does this change treatment decisions? vibrant. I was the Bud Girl, for Christ’s
qAm I eligible for any clinical trials? sake. I felt like I was staring down the
qWhat happens if my disease progresses while I’m in a clinical trial? barrel of a gun, waiting to find out how
many bullets were inside.”
nic h o l as m o n u i S t o c k p h o t o

qWho foots the bills if I participate in a clinical trial?


qWhere can I find emotional, psychological and spiritual support? There were 24 — to be exact— litter-
qWhom should I call with questions or concerns after office hours? ing her liver and lungs.
Carr pressed the doctor on her op-
qMay I contact you or a nurse if I have questions or more symptoms? (If the
answer is “no,” find another doctor.) tions. “Just try and live a normal life,”
he told her.
With two dozen time bombs ticking

8 S C I E N T I F I C A M E R I C A N ne w ans w ers f o r C ancer


© 20 08 SCIENTIFIC AMERIC AN, INC.
inside her? “How the hell could I do Becoming a “Healing Junkie” and sprouts. She formed a “posse” with
that? How could I live with cancer c a r r h i t t h e b o ok s and the In- other young women with cancer. She
without thinking of dying every day?” ternet. (“I tell people I have a Ph.D. explored alternative therapies, includ-
she wondered. from Google University,” she says, ing massage and meditation, and even
Well, he offered, she could try to laughing.) She traded in fast food for a spent time in a Zen monastery. And she
strengthen her immune system through vegan diet and swapped martinis for a began the empowering process of docu-
diet and lifestyle changes. green brew of cucumbers, kale, celery menting and filming her journey— ev-
“He did not know it, but in that mo-
ment he planted the seeds for personal
After doctors
revolution,” Carr says. “I was not going found tumors in
to kick back and wait for the unknown. her liver (below
I was going to dive in and become a left), Carr hit the
full-time healing junkie.” books to learn
more about her
She set about trying to find out ev-
disorder.
erything she possibly could about can-
cer. She sought second, third and fourth
opinions. “If I had listened to one of the
first doctors I talked to, I would have
ended up sliced, fried and hauling
around not one but three organs that
didn’t belong to me,” she says.

with cancer without thinking of dying every day?”


Your Odds of Beating Cancer
Success in the battle against cancer is often measured which collects survival data from state registries covering about 26
in terms of the “five-year relative survival rate.” That rate percent of the U.S. population.
c o u rtes y o f k ris carr (u l t r a s o u n d ) ; brian fassett (C a r r)

is the number of patients who are still alive five years after being Survival rates have increased dramatically over the years,
diagnosed, relative to the number thanks to earlier detection and
who would be expected to survive if Five-Year Survival Rates (percent) better treatments. The five-year
they had not come down with the Prostate 99 relative survival rate for patients
disease. Five years might not seem Melanoma (skin) 91 diagnosed with any type of cancer
like a lot, but it is, considering that Breast 89 in 1975 was 50 percent; the rate
67 is the median age for diagnosis. Endometrium 83 jumped to 67 percent in 2000.
Below is a sampling of five-year Urinary bladder 80 Bear in mind that survival
relative survival rates for common Kidney 67 rates vary widely depending on the
types of cancer diagnosed between Non-Hodgkin’s lymphoma 65 type of cancer and the patient’s
1996 and 2004. These rates are Colon and rectum 64 age, gender, general health,
calculated by the National Cancer Ovary 46 lifestyle and ethnicity. You can
Institute’s Surveillance, Epidemiology Lung and bronchus 15 find more detailed statistics at
and End Results (SEER) program, Pancreas 5 http://seer.cancer.gov

w w w. S c i A m . c o m SCIENTIFIC A MERIC A N 9
© 20 08 SCIENTIFIC AMERIC AN, INC.
erything and everyone she met, from the
Getting Support: physicians to the gurus to the quacks.
(Beware of quick fixes, she warns: “If
Tips, Tools and Tenderness anyone offers guarantees — run!”)
You’ve just been diagnosed with cancer. Now She conducted her search for an on-
what? First and foremost, do not try to handle cologist as though she were CEO of a
this on your own. Allow family and friends to help, company that she dubbed Save My Ass
and find others in your situation to lean on. Technologies, Inc., treating prospective
Online resources: doctors as though they were job appli-
■ www.crazysexycancer.com: Carr’s Web site. Have cants. “If it was the perfect fit: fine,”
questions? Want to dish? You can visit her online she says. “If not: next!” She nixed some
community, www.crazysexylife.com. of the candidates for their poor bedside
■ http://berniesiegelmd.com and www.ecap-online. manner (“There should be mutual re-
org: These sites of physician Bernie Siegel, author of spect”), others because of their pro-
Love, Medicine & Miracles and Peace, Love & Healing
(both from Harper Paperbacks, 1990), offer info and posed treatment plans. Among the dis-
tools based on the science of mind-body-spirit medicine. missed: the one who recommended a
■ www.cancercare.org: Need a professional cancer assistant? Try the next best thing. This
triple organ transplant (her liver and
site is designed to help patients navigate their way through cancer— answering questions, both lungs). “Some doctors are still
finding help or just “listening” when they need to vent. caught up in the old model of nuke it
■ http://nccam.nih.gov: The National Center for Complementary and Alternative Medicine of and cut it out— and sometimes it is re-
the National Institutes of Health provides information here on alternative and ally not necessary.... In my case it was
complementary therapies, discoveries and clinical trials. not the protocol,” Carr says. “Do you
■ http://hippocrateshealthinstitute.com: Site of the Hippocrates Health Institute, a world- want them to be stabbing at you if
renowned healing center in Florida.
they’re taking that stab in the dark? It’s
■ www.mercola.com: An alternative medicine and education site.
important to make sure you’re in the
■ www.heardsupport.org: This site is specifically geared toward patients with right hands. They can help you, or they
hemangioendothelioma, the rare cancer that Carr has.
can kill you. It’s that simple.”
■ www.livestrong.org: Site of seven-time Tour de France winner and cancer survivor Lance
The more physicians she inter-
Armstrong.
viewed, the more she came to realize
■ www.ulmanfund.org: Provides support programs and resources for patients and their
that “half the time they don’t have the
families. Also helpful: a downloadable book penned by founders Doug and Diana Ulman.
■ www.thechinastudy.com: The China Study, by T. Colin Campbell, probes the relationship
between diet and cancer and other diseases.
■ www.cancer.gov: This site of the National Cancer Institute is a comprehensive source of
state-of-the-art treatments and clinical trials (including a database of open trials).
“We are still
■ www.imtooyoungforthis.org: An invaluable source of support and research for survivors in
their 20s and 30s and their families. answers,” but it is the ones willing to
■ www.cancersurvivorsunite.org: Camps and support programs for young adults with cancer.
admit that fact who hold the most
■ www.youngcancerspouses.org: A site designed to connect couples dealing with the ups
promise of finding them. Enter the doc-
and downs of cancer. tor she “hired”: George Demetri, direc-
■ www.cancerconsultants.com: Contains detailed, consumer-friendly information on the
tor of the Center for Sarcoma and Bone
latest treatment developments. Oncology at the Dana-Farber Cancer
■ www.americancancersociety.com: This American Cancer Society site provides basic Institute in Boston, who, unlike many
information, alternative therapies, ways to manage the disease, and support programs . of the other “job applicants,” not only
■ www.oncolink.com: This University of Pennsylvania site offers key cancer info and has the medical credentials but, she
pointers. says, is also “kind and compassionate”
■ www.cancerguide.org: A how-to on researching your disease, searching for clinical trials, and welcomes his patients’ input.
and finding out about the latest traditional and alternative therapies.
■ www.cancer.net: American Society of Clinical Oncology site provides oncologist-approved Keeping Tumors at Bay
information to help patients make informed decisions about their health care. c a r r s ay s Demetri believes that she
■ www.gildasclub.org: Named for Saturday Night Live comedian Gilda Radner, who died of
manc h an / Gett y I ma g es

can live her “whole life” with the dis-


ovarian cancer, this site provides a support network for patients and their families. ease but that it may have to be treated
■ www.thewellnesscommunity.org: The Wellness Community provides support and with drugs at some point. “We don’t
education for cancer patients and caretakers—and hooks them up with others going
know. There is currently no cure,” she
through the same thing. It provides info on local wellness communities and even offers
a virtual wellness community in Spanish. notes, “but there’s no doubt in my mind
that any new information, drugs, and

10 SCIENTIFIC A MERIC A N ne w ans w ers f o r C ancer


© 20 08 SCIENTIFIC AMERIC AN, INC.
Medical Resources
Finding a doctor who specializes in cancer care and choosing a treatment
facility are essential steps in any patient’s recovery program. One good place
to start is with the 63 cancer centers that the National Cancer Institute recognizes for
“scientific excellence and the capability to integrate a diversity of research
approaches” (http://cancercenters.cancer.gov/cancer_centers). You can also check
whether the American College of Surgeons’ Commission on Cancer (www.facs.org/­
cancerprogram) approves of a given program. Some of the things to look for in a cancer
center include a low mortality index, a high ratio of nurses to patients and
opportunities to participate in clinical trials. For more tips, see www.cancer.gov/
cancertopics/factsheet/Therapy/doctor-facility. Here is a selection of some of the
most respected cancer treatment centers around the country:

Dana-Farber Cancer Institute Children’s Hospital of Philadelphia Mayo Clinic


Boston 215-590-1000 Rochester, Minn.
866-408-DFCI www.chop.edu/consumer/jsp/division/ (facilities also in Arizona and Florida)
www.dfci.harvard.edu service.jsp?id=26696 507-284-2511
www.mayoclinic.org/cancer-treatment
Memorial Sloan-Kettering Cancer Center Duke Comprehensive Cancer Center
New York City Durham, N.C. University of Washington Medical Center
212-639-2000 888-ASK-DUKE Seattle
www.mskcc.org www.cancer.duke.edu 206-598-4100
http://uwmedicine.washington.edu/
Sidney Kimmel Comprehensive Cancer University of Texas M. D. Anderson PatientCare/MedicalSpecialties/
Center at Johns Hopkins Cancer Center SpecialtyCare/UWMEDICALCENTER/Cancer
Baltimore Houston
410-502-1033 877-MDA-6789 UCLA Medical Center
www.hopkinshospital.org/health_info/ www.mdanderson.org Los Angeles
Cancer/index.html 800-UCLA-MD1
University of Chicago Medical Center www.uclahealth.org
888-UCH-0200
www.uchospitals.edu/specialties/cancer

empowered. I have cancer, but I’m dealing with it.”


treatment is going to come out of
this place [Dana-Farber]. I’m in the
right place to be monitored.”
Four years after turning the
camera on herself, Carr turned her
healing journey into a documenta-
ry called Crazy Sexy Cancer, which
h ein z Lin k e i S t o c k p h o t o (l a b) ; Lisa c o cciar d i (C a r r)

TLC bought in the fall of 2006.


Last year it had its world premiere
at the South by Southwest Film Fes-
tival in Austin, Tex.
“I’m not saying that cancer is
sexy,” she stresses. “What I’m say-
ing is that we are still empowered.
We are still alive and whole. I might
have cancer, but I’m dealing with it
and I’m still all that. The most im-
portant thing is to have a voice and
use it.”
Carr is among a growing num-

w w w. S c i A m . c o m SCIENTIFIC A MERIC A N 11
© 20 08 SCIENTIFIC AMERIC AN, INC.
every week. “We want to be the bridge,

The 411 on Health Insurance one of many bridges, between Western


and alternative medicine,” she says.
Worried that your health insurance won’t When first diagnosed, Carr viewed
cover your treatment? Wondering if you’re cancer as a freight train to death; now
entitled to disability benefits? These Web she views it as a “catalyst” for change.
sites may help: She changed her lifestyle, met a new
community of women and ditched act-
■ www.healthinsuranceinfo.net ing for writing, something she never be-
■ www.patientadvocate.org
lieved she could do. Last year she wrote
and published Crazy Sexy Cancer Tips
■ www.patient.cancerconsultants.com
(Globe Pequot Press), a book chock-full
■ http://cancerguide.org/disability.html of practical advice on everything from
■ www.thedisabilityexpert.com doctor shopping to diet to how to keep
your wits about you when diagnosed
■ www.ssa.gov/applyfordisability
with the Big “C” (or any other disease,
for that matter). She wrote a companion
ber of people living and thriving with lives and their immune systems [in- book, Crazy Sexy Cancer Survivor:
cancer, thanks to medical advances as tact],” she says. “Plus, there is so much More Rebellion and Fire for Your Heal-
well as a progressive philosophy in on- that we as patients can do to help our ing Journey, due out in September— and
cology that recognizes past mistakes of bodies regain health.” is set to pen a diet and lifestyle manual
overtreatment and welcomes alterna- Carr is currently developing a non- to be published next year.
tive medicine as a partner in the healing profit organization that will work with Perhaps most important, she says,
process. The new approach, she says, top oncologists on studies and research cancer led her to her “soul mate.” She
shatters the stigma that cancer is either using data from the more than 1,000 recruited Brian Fassett to help her film,
a death sentence or something that has members of her online community edit and produce her documentary. Dur-
to be eradicated — and opens the door (www.crazysexylife.com) and the ing the project, they fell in love — and
to treatments designed to keep tumors 5,000 to 10,000 people who visit her Fassett and Carr (who, when first diag-
in check, which could buy time while Web site (www.crazysexycancer.com) nosed, thought she would never date
new therapies are developed. “Many
amazing new treatments are targeting
tumors and leaving patients with their “Once I was able to change my focus,
How to Stay Healthy
Patients undergoing treatment can shore up their
physical (and emotional) reserves by eating well, exercising
and cutting stress (which impairs the immune system). The
American Institute for Cancer Research, which funds studies on
the role of food and exercise in cancer prevention and treatment,
recommends a diet that’s at least two-thirds vegetables, fruit,
whole grains and beans. Below is a roundup of research related to
■ A pair of 2006 studies showed that regular exercise reduced by
pame l a m o o re ( f o l d e r s) , jenn y h i l l (r u n n e r) an d

staying healthy:
up to 61 percent the odds of death in colorectal cancer
■ A study of 22,000 healthy Greeks showed their “Mediterranean patients. The findings held even in patients who did not start
diet,” rich in vegetables, whole grains, olive oil, fruit and fish, exercising until after diagnosis.
v ict o r me l nici u c ( f r u i t) i S t o c k p h o t o

reduced their risk of dying from cancer by at least 25 percent. ■ A 2005 study showed that 92 percent of nearly 3,000 women with
Other studies have found that nutrients in dark, leafy greens breast cancer who walked or did other exercise three to five hours
may inhibit the growth of tumor cells in breast, skin, lung and weekly were still alive 10 years after their diagnosis, compared
stomach cancers and that green tea may thwart cancer with 86 percent of those who exercised less than an hour a week.
development in colon, liver, breast and prostate cells. (A ■ A 30-year review of the scientific literature, published in 2004,
leading theory: flavonoids in tea and carotenoids in leafy suggested that cancer patients who feel helpless or who
greens, which act as antioxidants, may protect against cancer suppress negative emotions may be at greater risk of having
by rooting out free radicals.) their cancer spread than those who play a role in their healing.

12 SCIENTIFIC A MERIC A N ne w ans w ers f o r C ancer


© 20 08 SCIENTIFIC AMERIC AN, INC.
Looking
Ahead: Start
a Family?

Today Carr says she


is “healthier than
before I was diagnosed.”

Does a cancer diagnosis spell the end


of your dreams to have a family? In a
word—no. Note to readers: check your
options before undertaking
treatments that may cause infertility.
In the event that you cannot become

desperation led to inspiration.”


pregnant, there is always surrogacy
and adoption. Despite what you’ve
heard, it is possible to adopt if you’ve
again, let alone marry) got hitched in the ‘watch and live’ approach. I’m not wait- had cancer. The key: pick an agency
fall of 2006. “It was one of the happiest ing, putting my life on hold. I’m living and country that are open to working
with cancer survivors.
days of my life,” she says. “We vowed to my life, just with the knowledge that
be fellow adventurers. We thought it cancer is in my body. For more, check out:
would be way too melodramatic to say “I think that life is just too sweet to ■ w ww.fertilehope.org: This site
‘till death do us part.’ This was a day be bitter. Once I was able to change my provides unvarnished facts about
that cancer just was not a part of.” They focus, desperation led to inspiration. I fertility risks associated with cancer
are now considering having kids. (“Will made so many changes, and I thought: treatment as well as fertility-
B rian F assett (C a r r) ; pa v e l l o se v s k y i S t o c k p h o t o ( f a m i l y)

the hormones wake the sleeping dragon? This is an awesome life. I mean, hon- preservation and parenthood
We don’t know,” she says, “but I refuse estly, I don’t think anyone has a better alternatives before, during and after
to live my life in fear.”) And they have life than me. How can you live with the treatment. It outlines the success
started their own production company, knowledge of cancer? I might not ever rates, costs and time requirements
Red House Pictures. be able to get rid of it, but I can’t let that for a variety of fertility procedures
So how is the 36-year-old Carr to- ruin my life.... I think: Just go for it. and also addresses other
day, more than five years since her life- Life is a terminal condition. We’re all possibilities, including egg and sperm
donation, surrogacy and adoption.
altering diagnosis? “I am happy and, I going to die. Cancer patients just have
think, healthier than I was before I was more information, but we all, in some ■ w ww.pregnantwithcancer.org: This

diagnosed.” Her last scan in February ways, wait for permission to live.” Web site links newly pregnant
showed the tumors are stable. cancer patients with others with a
Looking back on her healing jour- Lisa Stein is news editor for similar cancer who have already
ney, she muses: “The doctors told me to Scientific American’s Web site, been there, done that.
‘watch and wait.’ What I prefer is the www.SciAm.com.

w w w. S c i A m . c o m SCIENTIFIC A MERIC A N 13
© 20 08 SCIENTIFIC AMERIC AN, INC.
r ?
nce
C a
or
d f ZIM
M ER

v e ARL
l Y C
B

v o
E
CREDIT

14 SCIENTIFIC A MERIC A N ne w ans w ers f o r C ancer


Natural selection lacks the power
to erase cancer from our species and,
some scientists argue, may even have
provided tools that help tumors grow

Charles Dar win, 1881


Natural selection is not natural perfection.  
Living creatures have evolved some remarkably complex ad-
aptations, but we are still very vulnerable to disease. Among
the most tragic of those ills— and perhaps most enigmatic— is
cancer. A cancerous tumor is exquisitely well adapted for
survival in its own grotesque way. Its cells continue to divide
long after ordinary cells would stop. They destroy surround-
ing tissues to make room for themselves, and they trick the
body into supplying them with energy to grow even larger.
But the tumors that afflict us are not foreign parasites that
ric h ard mi l ner arc h i v es

have acquired sophisticated strategies for attacking our bod-


ies. They are made of our own cells, turned against us. Nor
is cancer some bizarre rarity: a woman in the U.S. has a 39
percent chance of being diagnosed with some type of cancer
in her lifetime. A man has a 45 percent chance.
CREDIT

w w w. S c i A m . c o m SCIENTIFIC A MERIC A N 15
These facts make cancer a grim yet fasci- great evolutionary success that came with ac-
nating puzzle for evolutionary biologists. If quiring a body. But bodies also present a pro-
natural selection is powerful enough to pro- found risk. Whenever a cell inside a body
duce complex adaptations, from the eye to the divides, its DNA has a small chance of ac-
immune system, why has it been unable to quiring a cancer-causing mutation. “Every
wipe out cancer? The answer, these investiga- time a cell divides, it’s going to be at risk of
tors argue, lies in the evolutionary process developing into cancer,” Campisi says.
itself. Natural selection has favored certain Rare mutations, for instance, may cause
defenses against cancer but cannot eliminate a cell to lose restraint and begin to multiply
it altogether. Ironically, natural selection may uncontrollably. Other mutations can add to
Natural even inadvertently provide some of the tools the problem: They may allow deranged cells
selection that cancer cells can use to grow.
The study of cancer evolution is still in its
to invade surrounding tissues and spread
through the body. Or they may allow tumor
has favored infancy, with much debate about the mecha- cells to evade the immune system or attract
nisms involved and much testing of hypoth- blood vessels that can supply fresh oxygen.
certain eses left to carry out. Some medical research- Cancer, in other words, re-creates within
defenses ers remain skeptical that the work will affect
the way they fight the disease. Evolutionary
our own bodies the evolutionary process that
enables animals to adapt to their environ-
against cancer biologists agree that they are not about to ment. At the level of organisms, natural se-
discover a cure for cancer, but they argue lection operates when genetic mutations
but cannot that understanding cancer’s history could re- cause some organisms to have more repro-
eliminate it veal clues that would otherwise remain hid-
den. “Obviously, we always have that in the
ductive success than others; the mutations
get “selected” in the sense that they persist
altogether. back of our minds in everything we do,” says and become more common in future genera-
Judith Campisi of Lawrence Berkeley Na- tions. In cancer, cells play the role of organ-
tional Laboratory. isms. Cancer-causing changes to DNA cause
some cells to reproduce more effectively than
The Dawn of Cancer ordinary ones. And even within a single tu-
at i t s ro o t, cancer is a disease of multi- mor, more adapted cells may outcompete less
cellularity. Our single-celled ancestors repro- successful ones. “It’s like Darwinian evolu-
duced by dividing in two. After animals tion, except that it happens within one or-
emerged, about 700 million years ago, the gan,” explains Natalia Komarova of the Uni-
cells inside their bodies continued to repro- versity of California, Irvine.
duce by dividing, using the molecular ma-
chinery they inherited from their progeni- Limits to Defenses
tors. The cells also began to specialize as they a lt houg h ou r b odi e s may be vulner-
divided, forming different tissues. The com- able to cancer, they also have many ways to
plex, multicellular bodies animals have to- halt it. These strategies probably resulted
day were made possible by the emergence of from natural selection, because mutations
new genes that could control how cells divid- that made our ancestors less likely to die of
ed— such as by stopping the cells’ reproduc- cancer in their prime could have raised their
tion once an organ reached its adult size. The reproductive success. But given the many
millions of animal species are evidence of the millions of people who get cancer every year,
it is obvious that these defenses have not
Overview/Cancer Evolution eradicated the disease. By studying the evolu-
tion of these defenses, biologists are trying to
n  Natural selection has only a limited ability to prevent cancer. It has understand why they fall short.
provided some defenses, but these tend to delay the disease until late Tumor suppressor proteins are among the
in life rather than eliminating it entirely. most effective defenses against cancer. Stud-
n In addition, evolutionary forces have apparently favored some genes ies suggest that some of these proteins pre-
that can contribute to cancer’s development or aggressiveness. vent cancer by monitoring how a cell repro-
n A n understanding of cancer’s evolutionary history—and how individual duces. If the cell multiplies in an abnormal
tumors evolve in the body—could suggest fresh angles of attack on way, the proteins induce it to die or to slip
the disorder. into senescence, a kind of early retirement.
The cell survives, but it can no longer divide.

16 SCIENTIFIC A MERIC A N ne w ans w ers f o r C ancer


Tumor suppressor proteins play a vital role in ity to propagate their information through
our survival, but scientists have recently dis- the generations.”
covered something strange about them: in
some respects, we would be better off with-
Anticancer proteins such as p16 may fa-
vor the young over the old. When p16 pushes
“Every time
out them. a cell into senescence, the cell does not just a cell divides,
Norman E. Sharpless of the University of stop multiplying. It also begins producing an
North Carolina at Chapel Hill genetically odd balance of proteins. Among the proteins it’s going to
engineered mice to study the effect of one of
these proteins, called p16 (or, more properly,
it makes is vascular endothelial growth fac-
tor (VEGF), which triggers the growth of
be at risk
p16-Ink4a). He and his colleagues created a more blood vessels. VEGF fosters the growth of developing
line of mice that lacked a functional gene for of tumors by supplying them with extra nu-
p16 and thus could not produce the protein. trients. In young people, p16’s main effect into cancer.”
In September 2006 the group published three
studies on the mice. As expected, the animals
may be to suppress cancerous cells. But over
time, it may create a growing population of
—Judith Campisi
were more prone to cancer, which could arise senescent cells, which could make people
when they were only a year old. more vulnerable to cancer in old age.
But losing the p16 gene had an upside.
When the mice got old, their cells still behaved
as if they were young. In one experiment, the
scientists studied older mice, some of which
had working p16 genes and some of which did
not. They destroyed insulin-producing cells in
the pancreases of the animals. The normal ro-
dents could no longer produce insulin and de-
veloped fatal diabetes. But the ones without
the p16 protein developed only mild diabetes
and survived. The progenitors of their insulin-
producing cells could still multiply quickly,
and they repopulated the pancreas with new
cells. The scientists found similar results when
they examined cells in the blood and brains of
the mice: p16 protected them against cancer
but also made them old.
These results support a hypothesis Cam­
pisi has developed over the past few years.
Natural selection favors anticancer proteins
such as p16, but only in moderation. If these
Cell in the final stage of division
proteins become too aggressive, they can cre-
ate their own threats to health by making
bodies age too quickly. “It’s still a working Another way to delay cancer is to set up
hypothesis,” Campisi admits, “but the data several lines of defense. Studies on colon can-
p a u l andre w s U n i v e r s i t y o f D u n d e e / P h o t o R e s e a r c h e r s , I n c .

are looking stronger and stronger.” cer, for example, show that cells in the colon
must acquire mutations to several genes be-
Delaying the Inevitable fore they turn cancerous. These defense lines
a d e f e n s e ag a i n s t c a n c e r does not do not prevent people from getting colon
have to eradicate the disease completely to cancer— in fact, it is the third most common
be favored by natural selection. If it can just form of the disease. But the need for multiple
delay tumors until old age, it may allow peo- mutations to occur in a cell may reduce the
ple to have more children, on average, than chances that colon cancer will arise in young
others who lack the defense. It may seem cru- individuals. The average age of people diag-
el for evolution to stick old people with can- nosed with colon cancer is 70.
cer, but as Jarle Breivik of the University of Not all cancers strike the old, of course.
Oslo points out, “natural selection does not Most victims of a cancer of the retina called
favor genes because they let us live long and retinoblastoma, for example, are children.
happy lives. They are selected for their abil- But Leonard Nunney of the University of

w w w. S c i A m . c o m SCIENTIFIC A MERIC A N 17
Making Tools for Tumors
r e c e n t r e se a rc h sug g e s t s that nat-
ural selection may have altered genes in ways
that make cancer cells more dangerous. Evo-
lutionary biologists discovered this disturb-
ing possibility as they searched for the chang-
es that have made us uniquely human. After
our ancestors diverged from other apes about
six million years ago, they experienced natu-
ral selection as they adapted to a new way of
life as a toolmaking, savanna-walking hom-
inid. Scientists can distinguish between genes
that have not changed significantly since the
origin of hominids and those that have under-
gone major alteration as a result of selection
pressures. It turns out that among the genes
that have changed most dramatically are
some that play important roles in cancer.
Scientists suspect that the adaptive advan-
tages brought by these genes outweigh the
Resin cast of blood vessels in a tumor harm they may cause. One of these highly
evolved cancer genes makes a protein called
California, Riverside, argues that evolution fatty acid synthase (FAS). Normal cells use
is responsible for that difference between the the protein encoded by this gene to make
two cancers. He points out that colon cells some of their fatty acids, which are used for
have many more opportunities for acquiring many functions, such as building membranes
The ability to dangerous mutations than retinal cells do. and storing energy. In tumors, however, can-
stimulate new The colon is a large organ made of many
cells, which continue replicating throughout
cer cells produce FAS protein at a much high-
er rate. The protein is so important to them
blood vessel a person’s life as old cells slough off and new that blocking the activity of the gene can kill
ones take their place. That risk puts a big cancer cells. By comparing the sequence of
formation evolutionary premium on defenses that can the FAS gene in humans and other mammals,
serves a tumor prevent colon cells from turning cancerous.
The retina, on the other hand, is “the
Mary J. O’Connell of Dublin City University
and James McInerney of the National Univer-
just as it does smallest bit of tissue you can imagine,” as sity of Ireland found that the gene has under-
Nunney puts it. That small set of retinal cells gone strong natural selection in humans.
a placenta. also stops multiplying by the time a child turns “This gene has really changed in our lineage,”
five. With fewer cell divisions occurring, the McInerney says.
retina has far fewer opportunities to turn can- McInerney cannot say what FAS does dif-
cerous. As a result, retinoblastoma is extreme- ferently in humans, but he is intrigued by a
ly rare, striking only four people in a million. hypothesis put forward by the late psychia-
Because the risk is so much lower, Nunney trist David Horrobin in the 1990s. Horrobin
argues, natural selection cannot drive the argued that the dramatic increase in the size
C l o u ds Hi l l I ma g in g L T D . P h o t o R e s e a r c h e r s , I n c .

spread of new defenses against retinoblasto- and power of the human brain was made pos-
ma. A defense against cancer in the retina sible by the advent of new kinds of fatty acids.
would make very little difference to the aver- Neurons need fatty acids to build membranes
age reproductive success of a population. and make connections. “One of the things
that might allow a larger brain size was our
CARL ZIMMER writes frequently about evolution for the New York ability to synthesize fats,” McInerney specu-
the author

Times, National Geographic and other publications. He is the au- lates. But with that new ability may have
thor of six books, including Parasite Rex and Soul Made Flesh. His come a new tool that cancer cells could bor-
latest book is Microcosm: E. coli and the New Science of Life row for their own ends. Cancer cells may, for
(Pantheon, 2008). His blog, The Loom (www.scienceblogs.com/ example, use FAS as an extra source of energy.
loom), is a winner of Scientific American’s Science and Technol- Many fast-evolving cancer genes normal-
ogy Web Awards. ly produce proteins in tissues involved in re-

18 SCIENTIFIC A MERIC A N ne w ans w ers f o r C ancer


production — in the placenta, for instance. ate specifically in such cells are also among
Bernard J. Crespi of Simon Fraser University the fastest evolving in the human genome. A
in British Columbia and Kyle Summers of gene that allows a progenitor sperm cell to
East Carolina University argue that these divide faster than other cells will become
genes are part of an evolutionary struggle be- more common in a man’s population of
tween children and their mothers. sperm. That means it will be more likely to
Natural selection favors genes that allow get into a fertilized egg and be passed down
children to draw as much nourishment from to future generations.
their mothers as possible. A fetus produces Unfortunately for us, genes that make for
the placenta, which grows aggressively into fast-breeding sperm cells can make for fast-
the mother’s tissue and extracts nutrients.
That demand puts the fetus in conflict with
its mother. Natural selection also favors genes Evolution of
that allow mothers to give birth to healthy
children. If a mother sacrifices too much in a Cancer-Causing Virus
the pregnancy of one child, she may be less The American Cancer Society estimates that 17 percent of all cancer cases —
likely to have healthy children afterward. So more than 1.8 million a year— are caused by viruses and other infectious
mothers produce compounds that slow down agents. Scientists are studying the evolution of these cancer-causing
the flow of nutrients into the fetus. pathogens to find hints for fighting them. One such pathogen is the human
Each time mothers evolve new strategies papillomavirus, responsible for most of the half a million cases of cervical cancer
to restrain their fetuses, natural selection fa- diagnosed annually. The virus can cause host cells to divide long after normal
vors mutations that allow the fetuses to over- cells would stop and also prevents them from repairing mutations to their DNA.
come those strategies. “It’s a restrained con- Scientists have reconstructed some of the virus’s evolutionary history by
flict. There’s a tug-of-war about how much sequencing and comparing the genomes of hundreds of different types of
the fetus is going to take from the mother,” viruses. Papillomaviruses, which form a large family, are found in most
Crespi says. vertebrates, in whom they typically engender only warts and other benign
Genes that allow cells to build a better growths. Yet when Homo sapiens first emerged — about 200,000 years ago in
placenta, Crespi and Summers argue, can get Africa — our ancestors already carried a number of
hijacked by cancer cells — turned on when strains that could infect our species and no other
they would normally be silent. The ability to animal, and these included cancer-causing types.
stimulate new blood vessel formation and ag- After about 100,000 years, H. sapiens expanded
out of Africa to other continents, bringing the viruses
gressive growth serves a tumor just as it does
with them. As human populations became isolated
a placenta. “It’s something naturally liable to
from one another, their papillomaviruses did as well.
be co-opted by cancer cell lineages,” Sum-
Consequently, the genealogy of human papilloma­
mers says. “It sets up the opportunity for mu-
viruses reflects human genealogy. The oldest
tations to create tools for cancer cells to use lineage of the viruses is most common in living
HUMAN PAPILLOMAVIRUS
to take over the body.” is shown in a computer- Africans, for example. Native Americans descended
Yet even though activation of these usu- generated rendering. from Asians, and their viruses share that kinship.
ally quiet genes may make cancers more po- This coevolution may be medically relevant,
Linda m . S tannard U n i v e r s i t y o f C a p e To w n / P h o t o R e s e a r c h e r s , I n c .

tent, natural selection may still have favored because the viruses appear to have adapted to their hosts. In August 2006
them because they helped fetuses grow. “You scientists published a report in the Journal of the National Cancer Institute on
may get selection for a gene variant that helps the persistence of various virus types in different ethnic groups. A woman who
the fetus get a little more from mom,” Crespi becomes infected by a virus having an ancient association with her ethnic group
says. “But then, when that kid is 60, it might will carry the virus for a longer time than if she were infected by another type.
increase the odds of cancer by a few percent. Scientists are also investigating how certain benign papillomaviruses
It’s still going to be selected for because of the evolved to cause cancer. Their discoveries will become all the more important
strong positive early effects.” as vaccines are introduced against the viruses. The FDA has approved a
Sperm are another kind of cell that mul- vaccine against the most dangerous human papillomavirus strain, known as
tiplies rapidly. But whereas placental cells H16. But evolutionary studies indicate that on rare occasions, human
proliferate for a few months, sperm-making papillomavirus types have traded genes involved in triggering cancer. The
cells function for a lifetime. “For decades, global HIV epidemic might raise the risk of this gene swapping. As HIV weakens
human males are producing an enormous a person’s immune system, more types of human papillomaviruses can invade
amount of sperm all the time,” says Andrew and coexist. This mingling could conceivably give rise to a new cancer-causing
strain for which today’s vaccines would be less protective.  — C.Z.
Simpson of the Ludwig Institute for Cancer
Research in New York City. Genes that oper-

w w w. S c i A m . c o m SCIENTIFIC A MERIC A N 19
breeding cancer cells. Normally, nonsperm biology may elucidate one of the most daunt-
cells prevent these genes from making pro- ing challenges faced by oncologists: the emer-
teins. “These are genes that need to be firmly gence of drug-resistant tumors.
silenced, because they are dangerous genes,” Chemotherapy drugs often lose their ef-
Simpson says. It appears that in cancer cells, fectiveness against cancer cells. The process
mutations can unlock these sperm genes, al- has many parallels to the evolution of resis-
lowing the cells to multiply quickly. tance to antiviral drugs in HIV. Mutations
that allow cancer cells to survive exposure to
How vs. Why chemotherapy drugs enable the tumor cells
e vol u t ion a ry biol o g i s t s hope that to outcompete more vulnerable cells. Under-
their research can help in the fight against standing the evolution of HIV and other
cancer. In addition to clarifying why evolu- pathogens has helped scientists to come up
tion has not eradicated cancer, evolutionary with new strategies for avoiding resistance.
Now scientists are investigating how under-
standing the evolution within tumors could
The Surprising History lead to better ways of using chemotherapy.

of a Dog Cancer The concepts evolutionary biologists have


been exploring are relatively new for most
A canine cancer called Sticker’s sarcoma can be transmitted both through sex cancer biologists. Some are reacting with
and by licking or touching a tumor. Once established in a new host, it can great enthusiasm. Simpson believes, for in-
produce tumors that grow to the size of grapefruits before gradually stance, that deciphering the rapid evolution
disappearing. Many scientists once thought that the disease, like cervical of sperm-related genes could help in the fight
cancer, was spread by viruses. Now they know that the cancer cells themselves against tumors that borrow them. “I think
move from dog to dog and have been spreading this way for centuries. it’s absolutely crucial to understand exactly
A team of scientists from University College London and the University why there is such strong selection on these
of Chicago analyzed the genes of Sticker’s sarcoma cells collected from dogs genes,” Simpson says. “Understanding that
around the world. They found that the will give us a real insight into cancer.”
tumors are much more genetically Bert Vogelstein of the Howard Hughes
similar to one another than they are to Medical Institute also finds it useful to view
the dogs in which they grew. Additional cancer through an evolutionary lens. “Think-
research confirmed that the tumors ing about cancer in evolutionary terms jibes
belong to a single lineage of cancer cells. perfectly with the views of cancer molecular
“It represents the evolution of a geneticists,” he says. “In one sense, cancer is
cancer cell into a successful parasite of a side effect of evolution.”
worldwide distribution,” the scientists But Vogelstein is not yet persuaded by the
wrote in 2006 in the journal Cell.
significance of fast-evolving cancer genes:
Investigators have identified only a
“One has to be a little cautious. The first
few other possible examples of parasitic
question I would ask is, Are they looking at
some HUSKy might have started the cancer. Tasmanian devils, for example,
spread of Sticker’s sarcoma cells the whole genome in a wholly unbiased
can spread a facial tumor by biting one
among dogs hundreds of years ago. another. Why aren’t there more parasitic way?” McInerney acknowledges that such
cancers? Organ transplantation may systematic studies have not yet been conduct-
offer a clue. One of the biggest dangers in organ transplantation is rejection, ed, but the early results have prompted him
in which a patient’s immune system violently attacks the organ. All verte­ and other scientists to begin them.
brates reject grafts of foreign tissue with this kind of ferocity. It is possible Some cancer specialists are leery of the
that this rejection response evolved hundreds of millions of years ago as a entire approach. Christopher Benz of the
defense against parasitic cancers. Buck Institute for Age Research in Novato,
Sticker’s sarcoma appears to have evolved its way around this ancient Calif., says that no insights from evolution
defense. The cells in the tumor make very few of the surface proteins that should be accepted until they are put to an
danie l b o sc h u n g z e f a / C o r b i s

vertebrates use to distinguish self from nonself — allowing them to evade experimental test the way any other hypoth-
an all-out attack from the dog’s immune system. Instead the immune system esis would be. “Call me skeptical,” he says.
erodes the tumor slowly over the course of several months, and individual Crespi is familiar with this skepticism,
cancer cells can survive even after the tumor is gone. Rather than being and he thinks that it may emerge from the
just an ordinary cancer that dies with its host, it has become a cancer that different kinds of questions evolutionary bi-
can live for centuries.  — C.Z. ologists and cancer biologists ask. “The peo-
ple working on cancer are working on the

20 S C I E N T I F I C A M E R I C A N ne w ans w ers f o r C ancer


Ultimately,
study of the
evolution
of cancer may
Large brain tumor,
highlighted in blue
reveal why
eradicating
how question, and the evolution people are selecting against cancer defenses. “We have the disease
working on the why,” he says. changed their life histories by selecting on
Perhaps by asking different questions, their timing of reproduction,” Crespi says. has proved
evolutionary biologists will be able to con-
tribute to some of the debates among cancer
Ultimately, the study of the evolution of
cancer may reveal why eradicating the dis-
so difficult.
biologists. One long-standing argument fo- ease has proved so difficult. “There is no real
cuses on whether mice are good models for solution to the problem,” Breivik says. “Can-
cancer in humans. Some evolutionary biolo- cer is a fundamental consequence of the way
gists argue that they are not, because of their we are made. We are temporary colonies
separate history. Rodents inherited the same made by our genes to propagate them to the
set of genes as we did from our common an- next generation. The ultimate solution to
cestor some 100 million years ago, but then cancer is that we would have to start repro-
sim o n fraser rnc , N e w c a s t l e u p o n Ty n e / P h o t o R e s e a r c h e r s , I n c .

many of those genes underwent more change ducing ourselves in a different way.”
in the two lineages. Cancer-related genes such
as FAS may have experienced intense evolu- more to explore
tionary change in humans in just the past few Evolutionary Biology of Cancer. Bernard J. Crespi and Kyle Summers
million years, making them significantly dif- in Trends in Ecology and Evolution, Vol. 20, pages 545–552; 2005.
ferent from their counterparts in mice. Human Papillomavirus Type 16 and 18 Variants: Race-Related
Distribution and Persistence. L. F. Xi, N. B. Kiviat, A. Hildesheim,
Mice may also be a poor choice for a can- D. A. Galloway, C. M. Wheeler, J. Ho and L. A. Koutsky in Journal of the
cer model because of the way they reproduce. National Cancer Institute, Vol. 15, pages 1045–1052; August 2, 2006.
Scientists have bred lab mice to produce more Clonal Origin and Evolution of a Transmissible Cancer. Claudio
pups at a faster rate than their wild cousins. Murgia, Jonathan K. Pritchard, Su Yeon Kim, Ariberto Fassati and
Such manipulation may have altered the evo- Robin A. Weiss in Cell, Vol. 126, pages 477–487; August 11, 2006.
lutionary trade-off faced by mice, so that Ageing: Balancing Regeneration and Cancer. Christian M.
Beausejour and Judith Campisi in Nature, Vol. 443, pages 404–405;
they are rewarded for investing energy into September 28, 2006.
growing quickly and reproducing young. At Positive Selection in the Evolution of Cancer. Bernard J. Crespi and
the same time, this artificial selection may be Kyle Summers in Biological Reviews, Vol. 81, pages 407–424; 2006.

w w w. S c i A m . c o m SCIENTIFIC A MERIC A N 21
Mapping the
Cancer
Genome
Pinpointing the genes involved in
cancer will help chart a new course
across the complex landscape
of human malignancies

By Francis S. Collins and Anna D. Barker

‘‘I
f we wish to learn more about cancer, we must ute. As the U.S. population ages, this rate is expected to
now concentrate on the cellular genome.” ­Nobel rise significantly in the years ahead unless investigators
laureate Renato Dulbecco penned those words find ways to accelerate the identification of new vulnera-
more than 20 years ago in one of the earliest bilities within cancerous cells and develop novel strategies
public calls for what would become the Human for attacking those targets.
Genome Project. “We are at a turning point,” Dulbecco, a Still, however noble the intent, it takes more than a
pioneering cancer researcher, declared in 1986 in the jour- desire to ease human suffering to justify a research enter-
nal Science. Discoveries in preceding years had made clear prise of this magnitude. When applied to the 50 most com-
that much of the deranged behavior of cancer cells stemmed mon types of cancer, this effort could ultimately prove to
from damage to their genes and alterations in their func- be the equivalent of more than 10,000 Human Genome
tioning. “We have two options,” he wrote. “Either try to Projects in terms of the sheer volume of DNA to be se-
discover the genes important in malignancy by a piece- quenced. The dream must therefore be matched with an
meal approach, or … sequence the whole genome.” ambitious but realistic assessment of the emerging scienti­
Dulbecco and others in the scientific community fic opportunities for waging a smarter war against cancer.
grasped that sequencing the human genome, though a
monumental achievement itself, would mark just the first A Disease of Genes
step of the quest to fully understand the biology of cancer. t h e ide a t h at a lt e r at ions to the cellular genome
With the complete sequence of nucleotide bases in normal lie at the heart of all forms of cancer is not new. Since the
human DNA in hand, scientists would then need to clas- first identification in 1981 of a cancer-promoting version
sify the wide array of human genes according to their func- of a human gene, known as an oncogene, scientists have
tion—which in turn could reveal their roles in cancer. Over increasingly come to understand that cancer is caused
the span of two decades Dulbecco’s vision has moved from primarily by mutations in specific genes. The damage can
pipe dream to reality. Less than three years after the Hu- be incurred through exposure to toxins or radiation, by
man Genome Project’s completion, the National Institutes faulty DNA repair processes or by errors that occur when
of Health officially launched the pilot stage of an effort to DNA is copied prior to cell division. In relatively rare
create a comprehensive catalogue of the genomic changes cases, a cancer-predisposing mutation is carried within a
involved in cancer: The Cancer Genome Atlas (TCGA). gene variant inherited from one’s ancestors.
The main reason to pursue this next ambitious ven- Whatever their origin, these mutations disrupt bio-
ture in large-scale biology with great urgency is cancer’s logical pathways in ways that result in the uncontrolled
s l im fi l ms

terrible toll on humankind. Every day more than 1,500 cell replication, or growth, that is characteristic of cancer
CREDIT

Americans die from cancer— about one person every min- as well as other hallmarks of malignancy, such as the abil-

22 S C I E N T I F I C A M E R I C A N n e w a n s w ers f o r C a n cer
© 20 08 SCIENTIFIC AMERIC AN, INC.
CREDIT

w w w. S c i A m . c o m SCIENTIFIC A MERIC A N 23
© 20 08 SCIENTIFIC AMERIC AN, INC.
Many Pathways to Malignancy
Gene malfunctions underlie the ability of cancer cells to escape normal constraints on a cell’s behavior. Because genes
give rise to proteins that serve as cellular building blocks, signals and regulators of other genes, a mutation that
disables one gene, or causes it to be overactive, can have multiple deranging effects on the cell (below).
Nevertheless, cells usually need to accrete several cancer-promoting, or oncogenic, mutations in separate genes
to acquire the hallmark properties of malignancy (box at right). Identifying all the genes whose alteration can
produce those traits should one day reveal which mutations are the true drivers of specific types of cancer— even
of a specific patient’s malignancy — and therefore the most effective ways to intervene in the disease.
Gene variants mutated
EXTERNAL in multiple cancers
GROWTH EGFR SOS Ras
FACTORS

Gene mutated in
many solid tumors pI3K
▲ Complex circuitry Gene family members
The extraordinarily complex mutated or duplicated RAF MEK ERK
molecular interactions in a in many lung and
human cell can be viewed as breast cancers AKT
a network of parallel and
intersecting pathways. A
simplified depiction (right) B-RAF gene mutated in RSK CELL
of just one such pathway that Cell most melanomas GROWTH
promotes cell proliferation Suicide
begins with a family of
epidermal growth factor
receptors (EGFR) in the cell ▲ Oncogenic mutations
membrane. Their stimulation In a significant portion of lung and breast tumors, members of the EGFR gene family are mutated or
by growth factors outside the duplicated, which boosts the number or function of the receptors they encode, overstimulating this growth
cell transmits signals to pathway. Damage to downstream genes can have similar results. Changes in the B-RAF gene, seen in some
additional proteins and genes, 70 percent of melanomas, also promote hyperactive cell proliferation. Versions of the RAS gene are mutated
ultimately prompting the cell in many cancer types, which can affect cell growth as well as intersecting pathways — for example,
to “grow” by dividing. interfering with a suicide program that normally destroys damaged cells.

ity to invade neighboring tissues and to spread to sites through- changes, called the catalogue of somatic mutations in cancer,
out the body. Some mutations may disable genes that nor- or COSMIC, is maintained by Michael Stratton’s group at

t h e C e n ter f o r C a n cer R esearc h at M assac h u setts Ge n era l H o s p ita l a n d D a p h n e W. B e l l


mally protect against abnormal cell behavior, whereas others the Wellcome Trust Sanger Institute in Cambridge, England.
increase the activity of disruptive genes. Most cells must ac- But no one imagines that it is the complete list.

o f t h e NHG R I C a n cer Ge n etics B ra n c h f o r t h eir ad v ice o n t h e p at h w ay i l l u strati o n


L u c y R eadi n g - I k k a n da ; A c k n o w l ed g me n t: s p ecia l T h a n k s t o J effre y S ett l ema n o f
quire at least several of these alterations before they become So does it make sense to continue exploring the genomic
transformed into cancer cells— a process that can take years. basis of cancer at cottage-industry scale when we now pos-
Over the past two decades many individual research sess the means to vastly increase the scope and speed of dis-
groups have used groundbreaking molecular biology tech- covery? In recent years a number of ideas, tools and tech-
niques to search for mutations in genes that are likely candi- nologies have emerged and, more important, converged in a
dates for wreaking havoc on normal patterns of cell growth manner that has convinced many leading minds in the cancer
and behavior. By early 2007 such small-scale approaches had and molecular biology communities that it is time for a sys-
identified 350 cancer-related genes and yielded many signifi- tematic, collaborative and comprehensive exploration of the
cant insights into this diabolical disease. A database of these genomics of cancer.
The Human Genome Project laid a solid foundation for
Overview/Cancer Connections TCGA by creating a standardized reference sequence of the
three billion DNA base pairs in the genome of normal human
n  Changes in the structure or activity of genes underlie tissues. Now another initiative is needed to compare the DNA
the malignant behavior of cancer cells. sequences and other physical traits of the genomes of normal
n Identification of genes involved in certain cancers is cells with those of cancerous cells, to identify the major ge-
already advancing diagnosis and treatment. netic changes that drive the hallmark features of cancer [see
n T he Cancer Genome Atlas is a monumental initiative to box on opposite page]. The importance of international part-
eventually identify all the genetic alterations in nerships in large-scale biology to pool resources and speed
different forms of cancer so that gene changes driving scientific discovery was also demonstrated by the Human Ge-
the disease can be targeted directly. nome Project, and TCGA is exploring similar collaborations.
Finally, the Human Genome Project spurred significant

24 S C I E N T I F I C A M E R I C A N n e w a n s w ers f o r C a n cer
© 20 08 SCIENTIFIC AMERIC AN, INC.
sequencing 20 genes in 378 cancer samples. But the investiga-
tors hit pay dirt a year later when they found that a gene
called B-Raf was mutated in about 70 percent of the malig-
Hallmarks of Cancer nant melanoma cases they examined. A variety of researchers
The six abnormal capabilities listed below together give tumors their swiftly set their sights on this potential new therapeutic tar-
lethal power to overrun their native tissue and spread through the body.
get in the most deadly form of skin cancer. They tested mul-
Self-sufficiency in growth signaling tiple approaches — from classic chemical drugs to small inter-
Cancer cells amplify external growth cues or generate their own. fering ribonucleic acids — in cell lines and in mice, to see if
these interventions could block or reduce the activity of B-
Insensitivity to antigrowth signals
Cancer cells become deaf to quiescence cues from surrounding tissue. Raf or inhibit a protein called MEK that is overproduced as
a result of B-Raf mutations. Today the most promising of
Evasion of cell suicide these therapies are being tested in clinical trials.
Mechanisms that should trigger or carry out a self-destruct program
Other research groups have already zeroed in on genetic
in damaged cells are disabled or overridden.
mutations linked to certain types of breast cancer, colon can-
Limitless replicative potential cer, leukemia, lymphoma and additional cancers to develop
Cancer cells evade intrinsic limits on the number of times a normal molecular diagnostics, as well as prognostic tests that can
cell can divide.
point to an agent in the current arsenal of chemotherapies to
Sustained blood vessel growth which a particular patient is most likely to respond. Cancer
Tumors emit signals promoting the development of new genomics has also helped to directly shape the development
blood vessels to deliver oxygen and nutrients. and use of some of the newest treatments.
Invasiveness and motility The drug Gleevec, for example, was designed to inhibit
Cancer cells defy multiple signals and an enzyme produced by a mutant fused version of two genes,
Dividing lung
forces that hold a cell in place and prevent cancer cells called Bcr-Abl, that causes chronic myelogenous leukemia.
it from traveling to — and thriving Gleevec is proving dramatically effective against that disease
in — other tissues.
and showing value in the treatment of more genetically com-
Adapted from “The Hallmarks of Cancer,” by
Douglas Hanahan and Rober t A . Weinberg, plex malignancies, such as gastrointestinal stromal tumor
in C ell, Vol. 100; Januar y 7, 2000.
and several other relatively rare cancers that involve similar
enzymes. Herceptin, an agent that targets a cellular signal–
receiving protein called HER2, is successful against breast
cancers with an abnormal multiplication of the Her2 gene
advances in the technologies used to sequence and analyze that causes overproduction of the receptor protein.
genomes. At the start of that project in 1990, for example, Strategies for selecting treatments based on specific gene
the cost of DNA sequencing was more than $10 per “finished” mutations in a patient’s cancer are also being tested in studies
nucleotide base. Today the cost is less than a penny per base of the drugs Iressa and Tarceva for lung cancer, as well as
and is expected to drop further with the emergence of innova- Avastin for lung, colon and other cancers. The performance
tive sequencing methods. Because of these and other techno- of these new gene-based diagnostics, prognostics and thera-
logical developments, the large-scale approach embodied in peutics is certainly good news, although the list of such inter-
TCGA— unthinkable even a few years ago — has emerged as ventions remains far shorter than it would be if researchers
perhaps the most efficient and cost-effective way to identify in academia and the private sector had ready access to the
the wide array of genomic factors involved in cancer. entire atlas of genomic changes that occur in cancer.
Recent studies illustrate both the potential power of large-
Proofs of Concept scale genomics applied to the discovery of cancer genes and
p i l e s of da t a are, of course, not worth much without the tremendous undertaking a comprehensive cancer genome
evidence that comprehensive knowledge of cancer’s molecu- atlas will be. In 2006 Johns Hopkins University researchers
S TE V E GS CHMEIS SNER Photo Researchers, Inc.

lar origins can actually make a difference in the care of peo- sequenced about 13,000 genes in tumor tissues taken from 11
ple. Several recent developments have provided proofs of con-
cept that identifying specific genetic changes in cancer cells FRANCIS S. COLLINS and ANNA D. BARKER are leaders of the
the authors

can indeed point to better ways to diagnose, treat and prevent Cancer Genome Atlas initiative in their positions as, respec-
the disease. They offer encouraging glimpses of what is to tively, director of the National Human Genome Research Insti-
come and also demonstrate why the steps toward those re- tute and deputy director for Advanced Technologies and Stra-
wards are complex, time-consuming and expensive. tegic Partnerships of the National Cancer Institute. Collins led
In 2001, when the Wellcome Trust Sanger Institute began the Human Genome Project to its completion of the human DNA
its own effort to use genomic technologies to explore cancer, sequence, and Barker has headed drug development and bio-
the project’s immediate goal was to optimize robotics and technology research efforts in the public and private sectors,
information management systems in test runs that involved with a particular focus on fighting cancer.

w w w. S c i A m . c o m SCIENTIFIC A MERIC A N 25
© 20 08 SCIENTIFIC AMERIC AN, INC.
Genes and Cancer
A connection between genetic abnormalities and the aberrant colorectal cancer patients and 11 breast cancer patients. They
features of cancer cells was first suggested more than 100 reported finding potentially significant mutations in nearly
years ago by German biologist Theodor Boveri and others. But 200 different genes, most of which had not been previously
over the past few decades evidence that gene alterations associated with those two cancers. A follow-up analysis by the
directly cause the deranged behavior of cancer cells began Broad Institute in Cambridge, Mass., raised statistical con-
accumulating. Calls arose by 1986 to sequence the normal cerns about those conclusions, however, indicating that much
human genome to study malignant gene changes comprehen-
larger numbers of samples are likely required to distinguish
sively. The Human Genome Project was completed in 2003. The
Cancer Genome Atlas project recently started cataloguing the true cancer-causing mutations, or drivers, from background
gene mutations found in three types of human cancer. mutations, or passengers, in these and other cancers.
A 2007 study of lung cancer by the National Human Ge-

S T E V E G S C H M E I S S N E R P h o t o R e s e a r c h e r s , I n c . (t u m o r c e l l ) ; M A R K J . W I N T E R P h o t o R e s e a r c h e r s , I n c . (G l e e v e c m o l e c u l e) ; C E C I L H . F OX P h o t o R e s e a r c h e r s , I n c . (B i o s p e c i m e n C o r e R e s o u r c e) ;
Theodor Boveri


1890–1914 nome Research Institute’s Tumor Sequencing Project took
advantage of a larger sample set, getting much closer to the

aff y metri x (C a n c e r G e n o m e C h a r a c t e r i z a t i o n C e n t e r s) ; N A T I ON A L HU M A N G E NO M E R E S E A R C H I N S T I T U T E (G e n o m e S e q u e n c i n g C e n t e r s) ; © cdc / p h i l C o r b i s (D a t a C o o r d i n a t i n g C e n t e r )


Studies of abnormal chromosome distribution
during cell division suggest a role in malignancy. size we are aiming for in TCGA. That analysis of 371 lung
adenocarcinoma samples, plus 242 matched samples of nor-
1950s–1960s mal tissue, uncovered a critical gene alteration not previous-
Multiple discoveries reveal that tumor viruses can cause cancer
by injecting their genes into cells. ly linked to any form of cancer.
Among the major challenges encountered by researchers
1960

P h o t o R esearc h ers , I n c . (B o v e r i ) ; D E P A R T M E N T O F C L I N I C A L C Y T OG E N E T I C S , A D D E N B R OOK E ’ S HO S P I T A L P h o t o R e s e a r c h e r s , I n c . (P h i l a d e l p h i a c h r o m o s o m e) ;


sequencing cancer cell genomes is the difficulty of distin-
First genetic defect associated with a specific
guishing meaningless mutations in the tumor samples from
cancer—an abnormality known as the Philadelphia
chromosome—is discovered in chronic those that are cancer-related. Somewhat surprisingly, early
myelogenous leukemia (CML) cells. sequencing studies have also found very little overlap among
the genetic mutations present in different types of cancer and
1976 even substantial variation in the pattern of genetic mutations
Scientists discover that src, a nonviral gene found in animal cells, among tumor samples from patients with the same type of
can cause cancer.
cancer. Such findings underscore the idea that many different
1979 possible combinations of mutations can transform a normal
P53, later found to be the most frequently mutated gene in human cell into a cancer cell. Therefore, even among patients with
cancer, is discovered. cancers of the same body organ or tissue, the genetic profile
of each individual’s tumor can differ greatly.
1981 To grasp the full scope of what TCGA hopes to achieve,
H-RAS is the first human oncogene (a gene whose alteration is
one must consider the complexities identified in such early
cancer-promoting) to be discovered.
efforts and imagine extending the work to more than 100
1983 types of cancer. It is enough to give even veterans of the Hu-
Altered methylation of DNA, suspected to affect gene activation, man Genome Project and seasoned cancer biologists pause.
is found in cancer cells. Yet TCGA participants and other scientific pioneers from
around the world are forging ahead, because we are con-
1986 vinced that amid the intricacies of the cancer genome may lie
Renato Dulbecco, writing in Science, calls for sequencing the human the greatest promise for saving the lives of patients.
genome to advance cancer research.
Recognizing the immensity of this challenge and leveraging
1986 knowledge gained from TCGA and earlier efforts in England,
U.S. Department of Energy considers sequencing researchers from Asia, Australia, Europe and North America
the human genome to further study of radiation effects. recently announced plans to form the International Cancer
Genome Consortium. The goals of the consortium, in which
1986 the U.S. will play a central role, will be to produce high-qual-
First tumor suppressor gene, RB1, is identified. ity genomic data on a number of cancers and to make such data
rapidly and freely available to researchers around the globe.
1987 Although investigators will probably take many years to
Fused gene BCR-ABL in the Philadelphia chromosome
complete a comprehensive catalogue of all the genomic muta-
is found to cause CML.
tions that cause normal cells to become malignant, findings
1990 with the potential to revolutionize cancer treatment are like-
Model of multistep tumor genesis clarifies ly to appear well before this compendium is finished, as the
the role of accumulated gene changes in proofs of concept have shown. As each new type of cancer is
cellular transformation to malignancy. studied and added to TCGA, investigators will gain another
rich new set of genomic targets and profiles that can be used
1990
Human Genome Project begins. to develop more tailored therapies.

© 20 08 SCIENTIFIC AMERIC AN, INC.


Gleevec model


1993
Preclinical testing starts on drug that would
Compiling a Colossal Atlas become Gleevec, the first therapy developed to
a p h a se d -i n st r at e g y that proved successful at the be- target a known gene-based cause of a cancer.
ginning of the Human Genome Project was to test protocols
and technology before scaling up to full DNA sequence “pro- 1999
Gene-activity profiles are first shown to distinguish between
duction.” Similarly, TCGA is beginning with a pilot project to cancer types and to predict chemotherapy response.
develop and test the scientific framework needed to ultimately
map all the genomic abnormalities involved in cancer. 2001
In 2006 the National Cancer Institute and National Hu- Gleevec earns FDA approval.
man Genome Research Institute selected the scientific teams
and facilities that are participating in this pilot project, along 2002
with the cancer types they have since begun examining. Over Wellcome Trust Sanger Institute tumor genome survey discovers
a mutation in B-RAF gene common to 70 percent of melanomas.
the next few years these two institutes are devoting $100 mil-
lion to compiling an atlas of genomic changes in three tumor
2003
types: glioblastomas of the brain, lung cancer and ovarian can- Human Genome Project is completed.
cer. These particular cancers were chosen for several reasons,
including their value in gauging the feasibility of scaling up this 2005
project to a much larger number of cancer types. Indeed, only The Cancer Genome Atlas (TCGA) pilot project is
announced by the National Institutes of Health.
if this pilot phase achieves its goals will the NIH move forward
with a full-fledged project to develop a complete cancer atlas. 2006
The three malignancies selected for the pilot collectively TCGA names pilot project participants and three
account for more than 210,000 cancer cases in the U.S. every cancer types for sequencing and genetic analysis.
year and caused an estimated 191,000 deaths in this country
in 2006 alone. Moreover, tumor specimen collections meeting 2007–2010
TCGA will collect and analyze tumor samples obtained from designated
the project’s strict scientific, technical and ethical requirements biorepository institutions treating patients with cancer. The project’s
exist for these cancer types. In September 2006 our institutes four primary components—a Biospecimen Core Resource, seven Cancer
announced the selection of several biorepositories to provide Genome Characterization Centers, three Genome Sequencing Centers
and a Data Coordinating Center—will cooperate to test methods
both tumor and matched samples of normal tissue from the and technologies and to generate and manage data that will be made
same patients. These repositories, along with other qualifying available to the wider research community.
academic institutions and private hospitals, are delivering ma-
terials to a central Biospecimen Core Resource, one of four How Will It Work?
major structural components in TCGA’s pilot project.
Cancer Genome Characterization Centers, Genome Se- Patient samples
from biorepositories
quencing Centers and a Data Coordinating Center constitute
the project’s other three main elements [see illustration at
right], and all these groups collaborate and exchange data
openly. Specifically, the seven Cancer Genome Characteriza-
tion Centers are using a variety of technologies to examine the
activity levels of genes within tumor samples and to uncover
and catalogue so-called large-scale genomic changes that con- Biospecimen Core
Resource
tribute to the development and progression of cancer. Such
alterations include chromosome rearrangements, changes in
gene copy numbers and epigenetic changes, which are chemi-
Technology
cal modifications of the DNA strand that can turn gene activ- Cancer Genome Development
ity on or off without actually altering the DNA sequence. Characterization Genome Sequencing
Centers Centers
Genes and other chromosomal areas of interest identified
by the Cancer Genome Characterization Centers are targets
for sequencing by the three Genome Sequencing Centers. In
addition, families of genes suspected to be involved in cancer,
such as those encoding enzymes involved in cell-cycle control
known as tyrosine kinases and phosphatases, are being se-
quenced to identify genetic mutations or other small-scale Data Coordinating
Center
changes in their DNA code. At present, we estimate that some
2,000 genes— in each of perhaps 1,500 tumor samples—will be Research
sequenced during this pilot project. The exact numbers will, of Databases
course, depend on the samples obtained and what is ­discovered

w w w. S c i A m . c o m SCIENTIFIC A MERIC A N 27
© 2008 SCIENTIFIC AMERIC AN, INC.
about them by the Cancer Genome Characterization Centers. always harbor some nonmalignant cells, which can dilute the
Both the sequencing and genome characterization groups, sample. If the tumor DNA to be sequenced is too heteroge-
many of which were participants in the Human Genome neous, some important mutations may be missed.
Project, can expect to encounter a far greater level of com- The extent of these and other challenges should be known
plexity than that in the DNA of normal cells. Once cells be- soon, as researchers begin to examine the extraordinary sets
come cancerous, they are prone to an even greater rate of of multidimensional data being generated by TCGA. Though
mutation as their self-control and repair mechanisms fail. still in its early phases, the atlas team’s comprehensive analy-
The genomic makeup of individual cells can therefore vary sis of glioblastoma has already identified at least one new
dramatically within a single tumor, and the integrated teams gene that may prove to be important in understanding the
will need to develop robust methods for efficiently distin- cause of this deadly form of brain cancer.
guishing the “signal” of a potentially biologically significant Following the lead of the Human Genome Project and oth-
mutation from the “noise” of the high background rate of er recent medical genomics efforts, scientists are making all
mutations seen in many tumors. Furthermore, tumors almost these data swiftly and freely available to the worldwide re-

From Genome to Cancer— Why the Time Is Right


By Renato Dulbecco
When in 1986 I suggested a new project directed at identifying cancers (such as the acute phase of myeloid leukemia or the
all human genes, one of my overriding goals was to find those metastatic phase of other cancers), however, the participation
genes involved in cancer development— a feat I hoped would of many other genes is required. Most of them are still unknown.
lead to new tools for cancer research and, ultimately, to new An exception is the recently observed phenomenon of
therapies. That original human genome project has now been oncogene addiction in certain tumor cells: despite the presence
carried out and has demonstrated its usefulness for the of numerous mutations to the cellular genome, turning off the
discovery of genes involved in many diseases, including cancer. activity of one so-called oncogene causes the cells to commit
Moreover, the genome sequencing effort has been extended to suicide via a mechanism known as apoptosis. But how generally
other organisms — from bacteria to chimpanzees — and is this phenomenon occurs is also unknown. To approach these
showing the unity of life by revealing how many genes questions, it will be necessary to have a complete
distant species share. Cancer catalogue of the structural and functional alterations
In the course of this work, new technologies results from of genes and other cellular comp­onents that cause
have also provided a much more detailed the stepwise the loss of regulation in cancer cells. This process, in
understanding of the complicated processes by loss of cellular turn, will require a complete determination of their
which genes give rise to a variety of functional self-control. connections into networks by computational
molecules. An important outcome of this research is means — a task for the future.
the realization that genes do not act alone but are On the way to this goal, however, many other
participants in extensive networks of activity within cells. Any unanswered questions can be explored by the research
change in the functioning of one gene can therefore be community. A possible role for stem cells in cancer, for
accompanied by changes in the workings of multiple genes and example, is supported by similarities in the behavior of stem
proteins involved in the cells’ self-maintenance. cells and cancer cells: both have an unlimited ability to divide;
The complexity of this system in normal both are very sensitive to the cellular environment, or niche,
cells is evident in what we already know in which they grow; and many of the genes known to be active
about cancer— that it results from the in stem cells are also activated in cancer cells.
stepwise loss of such cellular self-control, The advent of genomics has provided welcome insight into
which becomes more and more complete as the mechanisms by which normal cells become cancerous, but
the disease progresses. That progression is our picture is still incomplete. The time has come to obtain a
sa l k i n stit u te f o r bi o l o g ica l st u dies

caused only in part by physical alterations, or truly comprehensive catalogue of the genes involved in cancer,
mutations, in specific genes; mostly it is the bringing to bear all the power of the new tools of genomics and
result of consequent changes in the molecular biology to the problem. The Cancer Genome Atlas
activity of many other genes involved project aims to do just that.
in cell regulation. Single genes may
therefore be responsible in the Renato Dulbecco is president emeritus of the Salk Institute
initiation of cancer and so potential for Biological Studies and co-recipient of the 1975 Nobel Prize in
therapeutic targets. To reach the Physiology or Medicine for discoveries related to the interaction
more advanced stages of these of tumor viruses and the genetic material of the cell.

28 S C I E N T I F I C A M E R I C A N n e w a n s w ers f o r C a n cer
© 20 08 SCIENTIFIC AMERIC AN, INC.
search community. To further enhance its usefulness to both
basic and clinical researchers and, ultimately, health care pro-
Targeting Gene Changes in Cancer
fessionals, TCGA is linking its sequence data and genome TCGA pilot project teams will examine the DNA of some
analyses with information about observable characteristics of 1,500 tumor samples from patients with cancers of the lung,
the patient donors and their clinical outcomes. Developing the ovaries or brain (glioblastoma), looking for genetic changes.
bioinformatic tools to gather, integrate and analyze those mas- Approximately 2,000 suspect genes in each sample will be
sive amounts of data, while safeguarding the confidentiality of sequenced to identify specific mutations. The list of target
genes will be tailored to each cancer type and largely deter­
patient information, is therefore another hurdle that must be
mined by what the Cancer Genome Characterization Centers
cleared to turn our vision into reality.
find in the samples, although candidates will also be drawn
from categories of genes already associated with cancer.
Uncharted Territory
t h e roa d a h e a d is fraught with scientific, technological
and policy challenges— some of which are known and others
as yet unknown. Among the uncertainties to be resolved: Will
new sequencing technologies deliver on their early promise in
time to make this effort economically feasible? How quickly
can we improve and expand our toolbox for systematically
detecting epigenetic changes and other large-scale genomic From left to right: Glioblastoma, lung cancer, ovarian cancer
alterations involved in cancer, especially those associated with
Gene Categories Examples
metastasis? How can we harness the power of computational
biology to create data portals that prove useful to basic biolo- Genes identified by TCGA • In some brain tumor cell lines, a
gists, clinical researchers and, eventually, health care profes- Cancer Genome Character­ gene encoding the intracellular
sionals on the front lines? How can we balance intellectual- ization Centers as having protein NF-kB is much more active
aberrant structure or than it is in normal brain tissue
property rights in a way that promotes both basic research
activity in a significant
and the development of therapies? The list goes on. number of tumor samples
To avoid raising false expectations, we also must be clear
C N R I P h o t o R e s e a r c h e r s , I n c . ( g l i o b l a s t o m a a n d o v a r i a n c a n c e r ) ; J ON A T H A N A S H T ON P h o t o R e s e a r c h e r s , I n c . (l u n g c a n c e r)

about the questions this project will not attempt to answer. Well-known oncogenes • Growth factor receptor genes:
Although it will serve as a resource for a broad range of bio- (genes whose overactivity HER2 (breast and lung cancers),
or alteration is EGFR (lung and colon cancers)
logical exploration, TCGA is only a foundation for the future
cancer-promoting) • S ignaling protein genes:
of cancer research and certainly not the entire house. And we BCR-ABL (chronic myelogenous
face the sobering issue of time — something that is in short leukemia), RAS (many cancers),
supply for many cancer patients and their families. As we B-RAF (skin cancers)
survey the considerable empty spaces that exist in our current • Regulators of cell death:
map of genomic knowledge about cancer, the prospect of fill- BCL-3 (lymphoma)
ing those gaps is both exhilarating and daunting. Scientists
Well-known tumor • C ontrollers of cell division:
and the public need to know up front that this unprecedented suppressors RB1 (retinoblastoma)
foray into molecular cartography is going to take years of (genes that protect • DNA repairers: HNPCC (colon cancer,
hard work and creative problem solving by thousands of re- cells from malignant endometrial cancer)
searchers from many different disciplines. transformation, unless • P romoters of programmed cell
disabled by mutation) suicide: p53 (lung, colon, breast and
Where all this work will lead can only be dimly glimpsed
brain tumors)
today. In this sense, our position is similar to that of the early
19th-century explorers Meriwether Lewis and William Clark. Genes related to known • T he oncogenes HER2 and EGFR are
As they ventured up the Missouri River into the largely un- oncogenes and tumor part of the epidermal growth factor
charted Northwest Territory in 1804, their orders from Pres- suppressor genes by receptor signaling pathway, which
ident Thomas Jefferson were to “take observations of latitude similarity or pathway contains at least half a dozen other
membership genes suspected of playing key
and longitude at all remarkable points.... Your observations
roles in cancer development
are to be taken with great pains and accuracy; to be entered and progression
distinctly and intelligibly for others, as well as yourself.”
Although Lewis and Clark did not find the much-longed-
for water route across the continent, their detailed maps more to explore
proved valuable to their fledgling nation in myriad ways that The New Era in Cancer Research. Harold Varmus in Science, Vol. 312,
Jefferson could never have imagined. For the sake of all those pages 1162–1165; May 26, 2006.
whose lives have and will be touched by cancer, we can only Characterizing the Cancer Genome in Lung Adenocarcinoma. Barbara
hope our 21st-century expedition into cancer biology exceeds A. Weir et al. in Nature, Vol. 450, pages 893–898; December 6, 2007.
even Renato Dulbecco’s grandest dreams. The Cancer Genome Atlas: http://cancergenome.nih.gov

w w w. S c i A m . c o m SCIENTIFIC A MERIC A N 29
© 20 08 SCIENTIFIC AMERIC AN, INC.
CREDIT

30 S C I E N T I F I C A M E R I C A N ne w ans w ers f o r C ancer


© 20 08 SCIENTIFIC AMERIC AN, INC.
Roots of Cancer
Recent evidence challenges long-held theories of how cells turn malignant—
and suggests new ways to stop tumors before they spread By W. Wayt Gibbs

What causes cancer?


­ obacco smoke, most people would say. Probably
T somes, each containing 1,000 or more genes, are
too much alcohol, sunshine or grilled meat; infec- often lost or duplicated. Pieces of chromosomes are
tion with cervical papillomaviruses; asbestos. All frequently scrambled, truncated or fused together.
have strong links to cancer, certainly. But they can- Chemical additions to the DNA, or to the histone
not be root causes. Much of the population is ex- proteins around which it coils, somehow silence im-
posed to these carcinogens, yet only a tiny minority portant genes, but in a reversible process quite dif-
suffers dangerous tumors as a consequence. ferent from mutation. And scans of the genomes of
A cause, by definition, leads invariably to its ef- malignant cells within tumors have found that they
fect. The immediate cause of cancer must be some typically harbor myriad rare mutations rather than
combination of insults and accidents that induces a handful of common genetic alterations.
normal cells in a healthy human body to turn ma- The accumulating evidence has spawned new
lignant, growing like weeds and sprouting in un- hypotheses that compete with the standard dogma
natural places. to explain what changes come first and which aber-
At this level, the cause of cancer is not entirely a rations matter most in the decadelong transforma-
mystery. In fact, a decade ago many geneticists were tion of a cell and its descendants from well-behaved
confident that science was homing in on a final an- tissue to invasive tumor. The challengers dispute
swer: cancer is the result of cumulative mutations that the dominant view of the disease as the product of
alter specific locations in a cell’s DNA and thus a defined genetic state. They argue that it is more
change the particular proteins encoded by cancer- useful to think of cancer as the consequence of a
related genes at those spots. The mutations affect two chaotic process, a combination of Murphy’s Law
kinds of cancer genes. The first are called tumor sup- and Darwin’s Law: anything that can go wrong
pressors. They normally restrain cells’ ability to di- will, and in a competitive environment, the most
vide, and mutations permanently disable the genes. prolific variants will dominate.
CAREFULLY
The second variety, known as oncogenes, stimulate Despite that shared underlying principle, the
CHOREOGRAPHED growth — in other words, cell division. Mutations new theories make different predictions about what
dance of lock oncogenes into an active state. Some researchers kind of treatments will work best. Some suggest
chromosomes still take it as axiomatic that such growth-promoting that many cancers could be prevented altogether by
j eff j o h ns o n H y b r i d M e d i c a l A n i m a t i o n

occurs during cell changes to a small number of cancer genes are the better screening, changes in diet, and new drugs —
division. Missteps
that mangle chromo­ initial event and root cause of every human cancer. or even by old drugs, such as aspirin. Other theories
somes or that send For the past few years, however, prominent on- cast doubt on that hope.
the wrong number cologists have increasingly challenged that theory.
to each daughter No one questions that cancer is ultimately a disease Marks of Malignancy
cell may be critical of the DNA. But as biologists trace tumors to their a wor k a bl e t h e ory of cancer has to explain
events early in the
development of roots, they have discovered many other abnormali- both why it is predominantly a disease of old age
cancer, according ties at work inside the nuclei of cells that, though not and why we do not all die from it. A 70-year-old is
to new theories. yet cancerous, are headed that way. Whole chromo- roughly 100 times as likely to be diagnosed with a

w w w. S c i A m . c o m SCIENTIFIC A MERIC A N 31
© 20 08 SCIENTIFIC AMERIC AN, INC.
malignancy as a 19-year-old is. Yet For example, cancer cells continue so those rare few that survive must re-
most people make it to old age without dividing in situations in which normal produce indefinitely if the tumor is to
getting cancer. cells would quietly wait for a special grow. The survivors do so in part by ma-
Biologists estimate that more than chemical signal — say, from an injured nipulating their telomeres, gene-free
10 million billion cells must cooperate neighbor. Somehow they counterfeit complexes of DNA and protein that pro-
to keep a human being healthy over the these progrowth messages. Conversely, tect the ends of each chromosome.
course of an 80-year life span. If any tumor cells must ignore “stop dividing” Tumors that develop these five facul-
one of those myriad cells could give rise commands that are sent out by the ad- ties are trouble, but they are probably not
to a tumor, why is it that fewer than jacent tissues they squeeze and by their deadly. It is a sixth property, the ability
half the population will ever contract a own internal aging mechanisms. to invade nearby tissue and then metas-
cancer serious enough to catch a doc- All cancerous cells have serious prob- tasize to distant parts of the body, that
tor’s attention? lems of some sort with their DNA, and gives cancer its lethal character. Local
One explanation is that a cell must as they double again and again, many invasions can usually be removed surgi-

“If you look at most solid tumors in adults, it looks like


someone SET OFF A BOMB in the nucleus.”
— William C. Hahn, Dana-Farber Cancer Institute
acquire several extraordinary skills to cells in the resulting colony end up far cally. But nine of every 10 deaths from
be malignant. “Five or six different reg- from the blood vessels that supply oxy- the disease are the result of metastases.
ulatory systems must be perturbed in gen and nutrients. Such stresses trigger Only an elite few cells in a tumor
order for a normal cell to grow as a can- autodestruct mechanisms in healthy seem to acquire this ability to detach
cer,” asserts Robert A. Weinberg of the cells. Tumor cells find some way to avoid from the initial mass, float through the
Whitehead Institute at the Massachu- this kind of suicide. Then they have to circulation and start a new colony in a
setts Institute of Technology. In a 2002 persuade nearby blood vessels to build different organ from the one that gave
review paper, he and William C. Hahn the infrastructure they need to thrive. birth to them. Unfortunately, by the
of the Dana-Farber Cancer Institute in A fifth derangement that almost all time they are discovered, many cancers
Boston argued that all life-threatening cancers acquire is immortality. A culture have already metastasized— including,
cancers manifest at least six such de- of normal human cells stops dividing af- in the U.S., 76 percent of lung cancers,
ranged abilities. (Although Weinberg is ter 50 to 70 generations. That is more 55 percent of colorectal cancers, and 37
one of the founding proponents of the than enough doublings to sustain a per- percent of breast cancers. By then the
standard paradigm, even those who son through even a century of healthy prognosis is frequently grim.
challenge that theory tend to agree with life. But the great majority of cells in tu-
this view.) mors quickly die of their genetic defects, The Order of Disorder
d o c t or s c ou l d c at c h incipient
Overview/How Cancer Arises tumors sooner if scientists could trace
the steps that cells take down the road to
■ Cancer is a genetic disease. Alterations to the DNA inside cells can derange cancer after the initial assault to their
cells in ways that disable normal restraints on growth, allowing a rare few to DNA by a carcinogen or some random
grow anywhere and to continue dividing indefinitely. biochemical mishap. Researchers broad-
■ F or many years, most cancer researchers focused on mutations to a ly agree on the traits of the diseased cells
relatively small set of cancer-related genes, suspecting that disruptions to that emerge from the journey. It is the
these genes are the decisive events in the transformation of healthy cells to propelling force and the order of each
malignant tumors. milestone that are under active debate.
■ New evidence, however, has challenged this view, leading to new theories of The dominant paradigm for 30 years
the origin of cancer. One such theory hypothesizes that a breakdown in DNA has been that tumors grow in spurts of
duplication or repair leads to many thousands of random mutations in cells. mutation and expansion. Genetic dam-
Another suggests that damage to a few “master” genes mangles the age to a cell deletes or disrupts a tumor
chromosomes, which then become dangerous. A third challenger proposes suppressor gene— RB, p53, BRCA2 and
that abnormal numbers of chromosomes in a cell may be the first milestone APC are among the best known—there-
on the road to cancer. by suppressing proteins that normally
ensure the integrity of the genome and

32 SCIENTIFIC A MERIC A N ne w ans w ers f o r C ancer


© 20 08 SCIENTIFIC AMERIC AN, INC.
Six DIABOLICAL derangements OF CANCER
1. G rowth even in the
absence of normal
“go” signals
Most normal cells wait for an
external message before divid­
ing. Cancer cells (right) often
counterfeit their own progrowth
messages.

2. G rowth despite “stop”


commands issued by
neighboring cells
As the tumor (yellow) expands, it
squeezes adjacent tissue, which
sends out chemical messages that
would normally bring cell division to
a halt. Malignant cells ignore
the commands.

3. Evasion of built-in
FRANK LYNC H QualTek Molecular Laboratories; ANDREJS LIEPINS P h o t o R e s e a r c h e r s , I n c . ; CNRI/photo researchers, inc.; photo researchers, inc.

autodestruct
mechanisms 5. Effective immortality
In healthy cells, genetic damage Healthy cells can divide no more
above a critical level usually than 70 times. Malignant cells need
activates a suicide program. more than that to make tumors. So
Cancerous cells (magenta) by­ they work around systems—such
pass this mechanism, although as the telomeres (yellow) at the
Clockwise from top right: CHRIS JONES Corbis; PETER L ANSDORP University of British Columbia; photo researchers, inc.;

agents of the immune system end of chromosomes (blue)—that


(orange) can sometimes suc­ enforce the reproductive limit.
cessfully order the cancer cells
to self-destruct. 6. Power to invade other
tissues and spread to
other organs
4. Ability to stimulate blood Cancers usually become life-threat­
vessel construction ening only after they somehow
Tumors need oxygen and nutrients
disable the cellular circuitry that
to survive. They obtain them by
confines them to a specific part
co-opting nearby blood vessels
of the particular organ in which
to form new branches (brown
they arose. New growths (orange
streaks) that run throughout the
and yellow) appear and eventually
growing mass.
interfere with vital systems.

the process of cell division. Alternative- gene looses another shackle, initiating form any cell. Or so the theory goes.
ly, a mutation may increase the activity another burst of growth. The mutant-gene paradigm gained
of an oncogene — such as HER2/NEU, Cells normally have two copies of almost universal acceptance because it
c-fos or c-erbB3 — whose proteins then every chromosome — one from the explained very well what scientists saw
stimulate the cell to reproduce. mother, the other from the father— and in their experiments on genetically engi-
Changes to cancer genes free the cell thus two copies, or alleles, of every gene. neered mice and human cell cultures.
of one or more normal restraints, allow- (In males, the single X and Y chromo- But new technologies now allow re-
ing it to outbreed its neighbors. The cell somes are notable exceptions.) A muta- searchers to survey large fractions of the
passes abnormalities in its DNA se- tion to just one allele is enough to acti- genomes of cancerous and precancerous
quence on to its descendants, which be- vate an oncogene permanently. But it cells taken directly from people. Many
come a kind of clone army that grows takes two hits to knock out both alleles recent observations seem to contradict
to the limits of its capacity. Eventually of a tumor suppressor gene. Four to 10 the idea that mutations to a few specific
another random mutation to a cancer mutations in the right genes can trans- genes lie at the root of all cancers.

w w w. S c i A m . c o m SCIENTIFIC A MERIC A N 33
© 20 08 SCIENTIFIC AMERIC AN, INC.
MALIGNANT MUTATIONS: A PARTIAL Map
WNT
Well over 1,000 genes have been found to be frequently mutated in
Frizzled one kind of cancer or another. According to the standard paradigm, the
Disheveled
proteins normally produced by these cancer-related genes, which include
tumor suppressor genes (red circles) and oncogenes (green circles), are
GSK3ß organized into complex biochemical circuits that control the reproduction
and survival of cells. Mutations that cause parts of the circuitry to fail
APC
(crosses) or become hyperactive (arrows) prompt cells to multiply into
tumors. But the sheer number of cancer genes — only a tiny fraction of
ß-catenin which are shown below — have frustrated attempts to deduce which ones
TGF-ß are necessary and sufficient to cause the disease.
TCF

TGF-ßR
MYC
SMADs
INK4B RB E2Fs Cell divides

TERT
INK4A CYCD-CDK4
LT Cell becomes immortal

CYCE-CDK2
ARF MDM2 p53 WAF1 ALT
Protein Cell autodestructs
dosages BAX
change
still under
MAPK
Cell alters metabolism construction
MEK
and behavior Maps of genes that control ...
RAF ... formation of new blood
RSK vessels in tumors
RAS PTEN PP2A eIF4E ... metastasis of cancer cells to
distant parts of the body
GRB2-SOS ... subversion of neighboring cells
so that they aid the tumor
TOR ST ... destabilization of
Growth RTK PI3K AKT the chromosomes
factor
... evasion of destruction
G protein by the immune system
Cytokine GPCR A d a p t e d f r o m “A S u b w a y M a p o f C a n c e r P a t h w a y s ,” b y W i l l i a m C . H a h n a n d R o b e r t A . W e i n b e r g i n
N a t u r e R e v i e w s C a n c e r, M a y 20 02; a v a i l a b l e a t w w w. n a t u r e . c o m /n r c /p o s t e r s /s u b p a t h w a y s

Unexplained Phenomena
© COPYRIGHT 20 02 MACMILL AN MAGAZINES LTD., REPRINTED BY PERMISSION
common human cancers, just a small human cells (and from one another) that
i n 2 0 03, for example, a team led by fraction of the cells in a tumor are re- they might fairly be called new species.
Michael F. Clarke, then at the University sponsible for its growth and metasta- In 2006 Bert Vogelstein of Johns
of Michigan at Ann Arbor and now at sis — and so for the illness and death of Hopkins University and a large team of
Stanford University, reported that it had the patient. The discovery of such cells, collaborators used new genetic se-
identified distinguishing marks for a which some began calling “cancer stem quencing technologies to examine more
rare subset of cells within human breast cells,” posed a problem for the mutant- than 13,000 genes in samples taken
cancers that can form new tumors. As gene theory of cancer. If mutations, from 11 breast tumors and 11 colorec-
few as 100 cells of this type quickly which are copied from a cell to its prog- tal cancers. They found noninherited
spawned disease when injected into eny, give tumor cells their deranged mutations in 1,149 (nearly 9 percent) of
mice lacking an immune system. Tens of powers, then shouldn’t all clones in the those genes. Even more startling than
thousands of other cells, harvested from army be equally powerful? the sheer number of genes mutated was
the same nine breast malignancies but In fact, most tumors are not masses the lack of consistency from one tumor
lacking the telltale marks, failed to do of identical clones. On the contrary, clos- to the next— or even from one cell to
so. These were the first tumor-initiating er examination has revealed amazing another within the same tumor. Not a
cells ever isolated for solid tumors, and genetic diversity among their cells, some single gene was mutated in more than 5
later studies confirmed that for many of which are so different from normal percent of the tumors.

34 SCIENTIFIC A MERIC A N ne w ans w ers f o r C ancer


© 20 08 SCIENTIFIC AMERIC AN, INC.
Moreover, some of the most com- gested that aneuploid cells might cause be converging as they bend to accom-
monly altered cancer genes have oddly the disease. But scientists could find no modate new experimental results.
inconsistent effects. Vogelstein’s group recurrent pattern to the chromosomal The modified form of the standard
has also reported that the much studied chaos —indeed, the genome of a typical dogma revives an idea proposed in 1974
oncogenes c-fos and c-erbB3 are curi- cancer cell is not merely aneuploid but by Lawrence A. Loeb, now at the Uni-
ously less active in tumors than they are unstable as well, changing every few versity of Washington. He and other ge-
in nearby normal tissues. The tumor generations. So Boveri’s idea was neticists have estimated that, on aver-
suppressor gene RB was shown to be dropped as the search for oncogenes age, random mutation will affect just
hyperactive — not disabled — in some started to bear fruit. The aneuploidy one gene in any given cell over the
colon cancers, and, perversely, it ap- and massive genomic instability inside course of a lifetime. Something— a car-
pears to protect those tumors from tumor cells were dismissed as side ef- cinogen, reactive oxidants, or perhaps a
their autodestruct mechanisms. fects of cancer, not prerequisites. malfunction in the DNA duplication
The “two hit” hypothesis— that both But the oncogene/tumor suppressor and repair machinery of the cell — must
alleles of a tumor suppressor gene must gene hypothesis has also failed, despite dramatically accelerate the mutation
be deactivated— has also been upended three decades of effort, to identify a par- rate, Loeb argues.
by the discovery of a phenomenon called ticular set of gene mutations that occurs For many years, he suggested that
haploinsufficiency. In some cancers, it in every instance of any of the most “early during the genesis of cancer there
turns out, tumor suppressors are not common and deadly kinds of human are enormous numbers of random mu-
mutated at all. Their output is simply cancer. So now, Hahn says, “the ques- tations — 10,000 to 100,000 per cell,”
reduced, and that seems to be enough to tion is which comes first: mutations or but he had little evidence to support the
push cells toward malignancy. This ef- aneuploidy?” idea. In 2006, however, technology ad-
fect has now been seen for more than a There are at least four competing vanced to the point that Loeb was able
dozen tumor suppressor genes, and in- answers. The newest and most contro- to test his hypothesis by comparing the
vestigators expect to find many more versial theory— that cancers often arise rate at which a noncoding portion of
like them. Searching for the mere pres- from the very stem cells that give healthy the p53 gene suffered small mutations
ence or absence of a gene’s protein is too organs their growth and healing abili- in both normal and malignant human
simplistic. Dos­age matters. ties— is discussed elsewhere in this issue cells. Cancerous cells, his test revealed,
[see “Stem Cells: The Real Culprits in harbored anywhere from 65 to 475 mu-
Beyond Mutation Cancer?” by Michael F. Clarke and Mi- tations per 100 million nucleotides,
r e s e a r c h e r s a r e n o w looking chael W. Becker, on page 40]. So we will whereas normal cells had four or fewer.
more closely at other phenomena, be- focus on the three more established That seems to be a far stretch from the
sides errors in a gene’s DNA sequence, ideas. Let us call them the modified 100,000-fold increase in mutation rate
that can dramatically alter the dosage of dogma, the early instability theory and that Loeb had anticipated, but it is
a protein in a cell. The loss or gain of a the all-aneu­ploidy theory. Encourag- nonetheless an important discovery that
chromosome (or part of one) containing ingly, all four of these theories seem to demands explanation.
the gene can do this. So can tweaks to
ABERRANT CHROMOSOMES in a cancer cell can
regulatory factors that control how of- alter the dosage of thousands of genes at once. A
ten a gene is translated into protein, as healthy cell (below) contains one pair of each of
well as epigenetic phenomena that alter the 22 kinds of chromosomes (distinct colors),
gene activity by reversible means. All plus two sex chromosomes. In a malignant cell
HESED M. PADILL A-NA SH AND THOMA S ried National Cancer Institute

(right), some chromosomes contain arms of


these changes are nearly ubiquitous in
different types (multicolored, at left edge).
established cancers. Others are missing limbs (royal blue) or are
“If you look at most solid tumors in present in the wrong number (lime green).
adults, it looks like someone set off a
bomb in the nucleus,” Hahn says. “In
most cells, there are big pieces of
chromo­somes hooked together and du-
plications or losses of whole chromo-
somes.” Cancer biologists call such
cells aneuploid.
Almost a century ago German bi-
ologist Theodor Boveri noticed the
strange imbalance in cancer cells be-
tween the numbers of maternal versus
paternal chromosomes. He even sug-

w w w. S c i A m . c o m SCIENTIFIC A MERIC A N 35
© 20 08 SCIENTIFIC AMERIC AN, INC.
THE GENESIS OF CANCER: FOUR THEORIES
For decades, the most widely accepted view of how cancer begins re­cently, three alternative theories have gained currency. One mod­
has been that mutations to a handful of special genes eliminate i­fies the standard paradigm by postulating a dramatic increase in
tumor suppressor proteins and activate oncoproteins. More the accumulation of random mutations throughout the genomes of
TUMOR SUPPRESSOR
GENES 2 Mutations in tumor suppressor
genes cause growth-inhibiting
proteins encoded by the genes to
APC p53
STANDARD DOGMA

disappear, allowing the cell


RB
1 Carcinogens, to survive and continue
dividing when it
such as ultra-
violet sunlight should not.
and tobacco, 3 At the same time, mutations to
directly alter the DNA oncogenes cause oncoproteins
BRAF sequence of cancer- to become hyperactive,
related genes. prompting the cell to grow in
c-fos situations in which it
c-erbB3 normally would not.
ONCOGENES

1 Something disables one


or more genes needed to
accurately synthesize or
MODIFIED DOGMA

repair the DNA.


DNA repair gene
2 As the cell divides,
random mutations
are introduced and
go unrepaired,
accumulating by the
tens of thousands.
Eventually the cancer-
related genes are hit.

1 Something silences one or more The dosage of genes in the cell changes
“master” genes that are as chromosome pieces are added
required for cell division. or deleted.
EARLY INSTABILITY

2 As the chromosomes are


duplicated, mistakes occur.
Some daughter cells get the
wrong number of chromosomes or
chromosomes with missing arms or
extra segments. The aberrations
get worse with each generation.

1 A mistake during cell division 2 The misplaced or truncated


produces aneuploid cells. chromosomes change the relative
amounts of thousands of genes. Teams
ALL-ANEUPLOIDY

of enzymes that normally cooperate


to copy or fix DNA begin to fail. Most
aneuploid cells die as a result.
CREDIT

36 S C I E N T I F I C A M E R I C A N ne w ans w ers f o r C ancer


© 20 08 SCIENTIFIC AMERIC AN, INC.
Loeb’s modified dogma thus may add
a prologue to the long-accepted life his-
precancerous cells. Two others focus on the role of aneuploidy: large-scale aber­rations in the
tory of cancer. But the most important
chromosomes. Aneuploidy could lead to genomic instability early on and later mutate known
plot points in that story would still re-
cancer genes. Or it may form tumors through an almost infinite variety of genetic changes.
main the same: mutations to genes that
serve to increase the reproductive success
4 The excess of oncoproteins and of cells. Mangled and ever changing
lack of tumor suppressor proteins lead
mutant cells to reproduce excessively. chromosomes are, in this narrative, mere
fortuitous by-products.

Unstable from the Outset


vo g e l s t e i n of Johns Hopkins and his
former co-worker Christoph Lengauer,
5 After many rounds of now at Novartis, proposed an alternative
mutation and expansion, theory in which chromosomal instability
one cell in the mass of can occur early on. The genetic flux then
mutants breaks free of all
combines forces with natural selection to
restrictions on its growth.
The colony invades adjacent produce a benign growth that may later
tissue in the host organ. be converted to an invasive malignancy
3 As in the standard and life-threatening metastases.
view, the elimination In their hypothesis, there are several
of tumor suppressor “master” genes whose function is critical
proteins and the for a cell to reproduce correctly. If just
activation of
one of these genes is disabled, either epi-
oncoproteins
short-circuit the genetically or by mutation, the cell stum-
autodestruct bles each time it attempts the carefully
mechanisms of the choreographed dance of cell division,
cell so that it cannot muddling some of the chromosomes into
commit suicide.
an aneuploid state. One result is to in-
crease 100,000-fold the rate at which
cells randomly lose one of the two alleles
3 In time, the dosage of tumor of their genes. For a tumor suppressor
suppressor proteins drops below
a critical threshold . . . gene, a lost allele may effectively put the
gene out of commission, either because
. . . and extra copies the remaining copy is already mutated or
of oncogenes can 6 In the most advanced
raise the dosage stages of its evolution, because of the haploinsufficiency effect.
of oncoproteins to the cancer leaks cells Lengauer and Vogelstein still assume that
dangerous levels. into the bloodstream.
These metastatic cells some cancer genes must be altered before
form new colonies at a malignancy can erupt.
distant sites throughout the Some observations do support the ear-
body, ultimately interfering
with life-critical functions. ly instability theory. In 2000 Lengauer’s
laboratory examined colon adenomas —
benign polyps that occasionally turn ma-
3 But a few survive lignant— and observed that more than 90
and produce progeny percent had extra or missing pieces of at
that are also least one chromosome. More than half
aneuploid, though
in ways different 5 Evolving over years or had lost the long arm of chromosome 5,
from the decades, the cells gradually home to the APC tumor suppressor gene,
parent cells. acquire the ability to
invade neighboring tissue long implicated in the formation of colon
of different types. cancer. Other researchers have discovered
similarly aberrant chromosomes in pre-
4 Eventually one or more cells acquire a mix of aberrant cancerous growths taken from the stom-
chromosomes that conveys one or more of the superpowers ach, esophagus and breast.
of cancer. The cells multiply into a precancerous tumor. The early instability theory still has
some loose ends, however. How can cells

w w w. S c i A m . c o m SCIENTIFIC A MERIC A N 37
© 20 08 SCIENTIFIC AMERIC AN, INC.
with shifty chromosomes outcompete bonus. Asbestos fibers, Duesberg notes, BRANCHING POINTS IN The Evol uti
their stable counterparts? Under normal can physically disrupt the process.
conditions, they probably do not, sug- Most aneuploid cells are stillborn or Theodor
gests immunologist Jarle Breivik of the growth-retarded. But in the rare survi- Boveri
University of Oslo. But in a “war zone,” vor, he suggests, the dosage of thou- 1927 Muller
observes that
where a carcinogen or other stressor is sands of genes is altered. That corrupts radiation
continually inflicting damage to cells, teams of enzymes that synthesize and mutates cells
normal cells stop dividing until they maintain DNA. Breaks appear in the
have completed repairs to their DNA. double helix, destabilizing the genome
Genetically unstable cells get that way further. “The more aneuploid the cell is, 1951 Muller
proposes
because their DNA repair systems are the more unstable it is, and the more theory that
already broken. So they simply ignore likely it will produce new combinations multiple
the damage, keep on proliferating, and of chromosomes that will allow it to mutations
turn a cell
thus pull ahead, Breivik hypothesizes. grow anywhere,” Duesberg explains. malignant
But what jumbles the chromosomes Unlike the three other theories, the
in the first place? No genes have yet all-aneuploidy hypothesis predicts that
been conclusively identified as master the emergence and progress of a tumor 1914 Boveri suggests that
genes, although several strong suspects are more closely connected to the as- aberrant chromosomes
have surfaced. Beth A. A. Weaver of the sortment of chromosomes in its cells may cause cancer
University of Wisconsin–Madison and than to the mutations in the genes on
her collaborators may have uncovered a those chromosomes. Some observa-
clue in their studies of a gene called tions do seem to corroborate the idea.
CENP-E. The protein produced from Thomas Ried, chief of cancer ge- 1915 1920 1925 1930 1935 1
this gene is one of several that are cru- nomics at the National Cancer Institute,
cial for guiding the chromosomes as has obtained supporting evidence in hu- be expanded to incorporate the still
they replicate and separate during cell mans from his investigation of aneu- mysterious role of epigenetic phenome-
division. At a meeting of the Society for ploidy in cervical and colorectal cancers, na, which may be pivotal.
Chromosomal Cancer Research in which he presented at the Oakland con- It is important to determine which
Oakland, Calif., this year, Weaver re- ference. “Unequivocally, there are re- of the ideas is more right than the oth-
ported that in mice embryos missing current patterns of genomic imbalanc- ers, because they each make different
one of their two copies of CENP-E (and es,” Ried avers. “Every single case of predictions about the kinds of therapy
thus making only half the normal [nonhereditary] colorectal cancer, for that will succeed best against the most
amount of CENP-E protein), chromo- example, has gains of chromosomes 7, 8, common and lethal cancers. In the
somes quickly went askew. Within a 13 or 20 — or a loss of 18. In cervical can- standard view, tumors are in effect ad-
couple of weeks, more than half the em- cer, aneuploidy of chromosome 3 hap- dicted to the proteins produced by on-
bryonic cells were aneuploid. And when pens very early, and those cells seem to cogenes and are poisoned by tumor
these animals grew up, they developed have a selective advantage.” Ried finds suppressor proteins. Medicines should
many more spleen and lung cancers the average number of abnormal chro- therefore be designed to break the ad-
than did control mice with normal mosomes increasing gradually from 0.2 diction or supply the poison. Indeed,
CENP-E genes. in a normal cell to 12 in the cells of met- this strategy is exploited by some newer
astatic colon tumors. drugs, such as Gleevec (for rare forms
Aneuploidy All the Way Down “So I actually think Duesberg is of leukemia and stomach cancer) and
on t h e o t h e r h a n d, maybe cells right that aneuploidy can be the first Herceptin (for one variety of advanced
can become malignant even before any genetic aberration in cancer cells,” Ried breast cancer).
master genes, oncogenes or tumor sup- says. “But he also argues that no gene But all existing therapies, including
pressor genes are mutated. Peter Dues- mutations are required. This is simply Gleevec and Herceptin, fail in some pa-
c h risty krames (t w o p r e c e d i n g p a g e s)

berg of the University of California, not true.” tients because their tumors evolve into
Berkeley, has put forth a third theory: a resistant strain. Loeb fears that there
nearly all cancer cells are aneuploid (leu- Stopping Cancer at Its Roots may be no easy way around that prob-
kemia being one exception) because they neither the standard dogma nor any lem; his latest work, after all, suggests
start that way. Lots of things can inter- of the new theories that challenge it can that a tumor of any significant size har-
fere with a dividing cell so that one of its fully untangle the knotted roots of the bors up to a trillion random mutations.
daughter cells is cheated of its normal 100-odd diseases we call cancer and ex- “If I am right, then within any given
complement of 46 chromosomes and plain them as variations of a single prin- tumor, there will be cells with random
the other daughter is endowed with a ciple. And all the theories will need to mutations that protect them from any

38 S C I E N T I F I C A M E R I C A N ne w ans w ers f o r C ancer


© 20 08 SCIENTIFIC AMERIC AN, INC.
l ution of cancer theory
1971 Alfred G. Knudson 1997 Christoph Lengauer, 2002 Thomas Ried 2008 The number of
Hermann J. explains different vogelstein and co­workers identifies recurrent identified cancer genes,
Muller rates of inherited and demonstrate dramatic patterns of now well over 1,000,
spontaneous retinal increase in gain and loss aneuploidy in human continues to grow rapidly
cancer with the of chromosomes in colon cervical and colon
hypothesis that two tumor cells and propose cancer
“hits,” or damaging that chromosomal
mutations, are needed instability is a critical
to disable both alleles early event that leads
of the RB gene and that to the mutation of
one mutation can be Bert oncogenes and tumor
inherited Vogelstein suppressor genes
1990 vogelstein and Peter Duesberg
1974 Lawrence A. Eric R. Fearon publish a model of
Loeb argues that sequential gene mutations that
random mutations 1999 Duesberg and collaborators
lead to colon cancer publish detailed theory of how
must accumulate
much faster than is 1986 Robert A. weinberg and aneuploidy may be sufficient to cause
1960 Discovery that an normal inside cells colleagues isolate RB, the first cancer itself, even without mutations to
exchange of DNA between that become tumor suppressor gene any particular set of genes
chromosomes 9 and 22 malignant
leads to chronic
myelogenous leukemia

1940 1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005

treatment you can conceive,” Loeb aneuploid, cells that were highly aneu- why, doctors have discovered that a dai-
says. “So the best you can hope for is ploid actually suppressed tumor forma- ly baby aspirin seems to prevent colon
to delay the tumor’s growth. You are tion. It may be possible to design drugs adenomas and to lower the risk of one
not going to cure it.” that kill cancer cells by exacerbating kind of breast cancer in some adults.
For the elderly— who, after all, are their aneuploidy. The effect is small. But it is a step from
the main victims of cancer— a sufficient One day science will produce a de- chemotherapy toward a better alterna-
delay may be as good as a cure. And finitive answer to the question of what tive: chemoprevention.
even better than slowing the growth of causes cancer. It will probably be a very
From left to right: photo researchers, inc. (Boveri); HULTON-DEUTSCH COLLECTION/ CORBIS (Muller);

a tumor would be to delay its formation complicated answer, and it may force us W. Wayt Gibbs, a contributing editor
in the first place. If Lengauer, Weaver to shift our hope from drugs that cure at Scientific American, is executive
and others succeed in identifying mas- the disease to medicines that prevent it. editor of Intellectual Ventures in
ter genes, then it should also be possible Even without a clear understanding of Bellevue, Wash.
to make drugs that protect or restore
their function. The Johns Hopkins more to explore
group has already licensed some of its Aneuploidy Precedes and Segregates with Chemical Carcinogenesis. Peter Duesberg, Ruhong Li,
cell lines to the pharmaceutical indus- David Rasnick, Charlotte Rausch, Andreas Willer, Alwin Kraemer, George Yerganian and Ruediger
Hehlmann in Cancer Genetics and Cytogenetics, Vol. 119, No. 2, pages 83–93; June 2000.
try to use in drug screening.
ALEX WONG Getty Images (Vogelstein); MAR C GELLER ( Duesberg)

Chromosome Segregation and Cancer: Cutting through the Mystery. Prasad V. Jallepalli and
Screening of a different kind may be Christoph Lengauer in Nature Reviews Cancer, Vol. 1, No. 2, pages 109–117; November 2001.
the best approach if the all-aneuploidy Rules for Making Human Tumor Cells. William C. Hahn and Robert A. Weinberg in New England
theory is correct. There are no known Journal of Medicine, Vol. 347, No. 20, pages 1593–1603; November 14, 2002.
means of selectively killing cells with ab- Multiple Mutations and Cancer. Lawrence A. Loeb, Keith R. Loeb and Jon P. Anderson in
normal chromosomes. But a biopsy that Proceedings of the National Academy of Sciences USA, Vol. 100, No. 3, pages 776–781;
February 4, 2003.
turns up a surfeit of aneuploid cells
Human Cancers Express a Mutator Phenotype. Jason H. Bielas, Keith R. Loeb, Brian P. Rubin,
might warrant careful monitoring or Lawrence D. True and Lawrence A. Loeb in Proceedings of the National Academy of Sciences USA,
even preventive surgery in certain cases. Vol. 103, No. 48, pages 18238–18242; November 28, 2006.
Duesberg suggests that foods, drugs and Aneuploidy Acts Both Oncogenically and as a Tumor Suppressor. Beth A. A. Weaver et al.
chemicals should be tested to identify in Cancer Cell, Vol. 11, No. 1, pages 25–36; January 2007.
compounds that cause aneuploidy. And Chromosomal Chaos and Cancer. Peter Duesberg in Scientific American, Vol. 296, No. 5,
pages 52–59; May 2007.
Weaver found that although CENP-E-
Comparative Lesion Sequencing Provides Insights into Tumor Evolution. Siân Jones et al. in
deficient mouse cells were highly tum- Proceedings of the National Academy of Sciences USA, Vol. 105, No. 11, pages 4283–4288;
origenic if the cells were moderately March 18, 2008.

w w w. S c i A m . c o m SCIENTIFIC A MERIC A N 39
© 20 08 SCIENTIFIC AMERIC AN, INC.
40 SCIENTIFIC A MERIC A N
© 20 08 SCIENTIFIC AMERIC AN, INC.
ne w ans w ers f o r C ancer
Stem Cells:
The Real Culprits in Cancer?
By Michael F. Clarke and Michael W. Becker

A dark side of stem cells — their potential to turn malignant— is at the root of a handful of
cancers and may be the cause of many more. Eliminating the disease could depend on
tracking down and destroying these elusive killer cells

A fter more than 30 years of declared age of tumor cells have the power to produce
war on cancer, a few important vic- new cancerous tissue and that targeting
tories can be claimed, such as 85 per- these specific cells for destruction may be a
cent survival rates for some childhood can- far more effective way to eliminate the dis-
cers whose diagnoses once represented a ease. Because they are the engines driving
death sentence. In other malignancies, new the growth of new cancer cells and are very
drugs are able to at least hold the disease at probably the origin of the malignancy itself,
bay, making it a condition with which a pa- these cells are called cancer stem cells. But
tient can live. In 2001, for example, Gleevec they are also quite literally believed to have
was approved for the treatment of chronic once been normal stem cells or their imma-
myelogenous leukemia (CML). The drug has ture offspring that have undergone a malig-
been a huge clinical success, and many pa- nant transformation.
tients are now in remission following treat- This idea — that a small population of
ment with Gleevec. But evidence strongly malignant stem cells can cause cancer— is
suggests that these patients are not truly far from new. Stem cell research is consid-
cured, because a reservoir of malignant cells ered to have begun in earnest with studies
responsible for maintaining the disease has during the 1950s and 1960s of solid tumors
not been eradicated. and blood malignancies. Many basic prin-
Conventional wisdom has long held that ciples of healthy tissue genesis and develop-
any tumor cell remaining in the body could ment were revealed by these observations
potentially reignite the disease. Current treat- of what happens when the normal process-
ments therefore focus on killing the greatest es derail.
number of cancer cells. Successes with this Today the study of stem cells is shedding
approach are still very much hit-or-miss, light on cancer research. Scientists have
however, and for patients with advanced cas- filled in considerable detail over the past 50
es of the most common solid-tumor malig- years about mechanisms regulating the be-
ONE AMONG THOUSANDS of tumor
nancies, the prognosis remains poor. havior of normal stem cells and the cellular
cells may be a cancer stem cell Moreover, in CML and a few other can- progeny to which they give rise. These fresh
responsible for driving the disease. cers it is now clear that only a tiny percent- insights, in turn, have led to the discovery of

w w w. S c i A m . c o m SCIENTIFIC A MERIC A N 41
© 20 08 SCIENTIFIC AMERIC AN, INC.
similar hierarchies among cancer cells within a tumor, pro- responsible for specific tasks, ranging from defending against
viding strong support for the theory that rogue stemlike cells infection to carrying oxygen to tissues [see box on opposite
are at the root of many cancers. Successfully targeting these page]. By the time a cell reaches that final functional stage, it
cancer stem cells for eradication therefore requires a better has lost all ability to proliferate or to alter its destiny and is
understanding of how a good stem cell could go bad in the said to be terminally differentiated.
first place. The stem cells themselves meanwhile remain undifferenti-
ated, a state they maintain through their unique capacity for
Orderly Conduct self-renewal: to begin producing new tissues, a stem cell di-
t h e h u m a n b ody is a highly compartmentalized system vides in two, but only one of the resulting daughter cells might
made up of discrete organs and tissues, each performing a proceed down a path toward increasing specificity. The oth-
function essential to maintaining life. Individual cells that er daughter may instead retain the stem cell identity. Num-
make up these tissues are often short-lived, however. The skin bers in the overall stem cell pool can thus remain constant,
covering your body today is not really the same skin that you whereas the proliferation of intermediate progenitors allows
had a month ago, because its populations of specific he-
surface cells have all since matopoietic cell types to ex-
sloughed off and been re- Stem cells’ power to pand rapidly in response to
placed. The lining of the gut
turns over every couple of self-renew already exempts changing needs.
The capacity of stem cells
weeks, and the life span of
the platelets that help to clot them from the rules. to re- create themselves
through self-renewal is their
blood is about 10 days. most important defining
The mechanism that maintains a constant population of property. It gives them alone the potential for unlimited life
working cells in such tissues is consistent throughout the span and future proliferation. In contrast, progenitors have
body and, indeed, is highly conserved among all complex some ability to renew themselves during proliferation, but
species. It centers on small pools of long-lived stem cells that they are restricted by an internal counting mechanism to a
serve as factories for replenishing supplies of functional cells. finite number of cell divisions. With increasing differentia-
This manufacturing process follows tightly regulated and or- tion, the ability of the progenitors’ offspring to multiply de-
ganized steps wherein each generation of a stem cell’s off- clines steadily.
spring becomes increasingly specialized. The practical significance of these distinctions can be ob-
This system is perhaps best exemplified by the hematopoi- served when hematopoietic stem cells or their descendants
etic family of blood and immune cells. All the functional cells are transplanted. After the bone marrow of a mouse is irradi-
found in the blood and lymph arise from a single common ated to destroy the native hematopoietic system, progenitor
parent known as the hematopoietic stem cell (HSC), which cells delivered into the marrow environment can proliferate
resides in bone marrow. The HSC pool represents less than and restore hematopoiesis temporarily, but after four to eight
0.01 percent of bone marrow cells in adults, yet each of these weeks those cells will die out. A single transplanted he-
rare cells gives rise to a larger, intermediately differentiated matopoietic stem cell, on the other hand, can restore the en-
population of progenitor cells. Those in turn divide and dif- tire blood system for the lifetime of the animal.
ferentiate further through several stages into mature cells The hematopoietic system’s organization has been well
understood for more than 30 years, but similar cellular hi-
Overview/Cancer Stem Cells erarchies have recently been identified in other human tis-
sues, including brain, breast, prostate, large and small intes- I l l ustrati o n b y k enn br o w n M o n d o l i t h i c S t u d i o s ( p a g e 4 0)

n  C ancer cells are often perceived as all having the same tines, and skin. Principles of regulated stem cell behavior are
potential to proliferate and expand the disease, but in also shared across these tissues, including specific mecha-
many types of cancer only a small subset of tumor cells nisms for controlling stem cell numbers and for directing
has that power. decisions about the fates of individual cells. Several genes
n T he tumor-generating cells share key traits with stem and the cascades of events triggered by their activity— known
cells, including an unlimited life span and the ability to as genetic pathways — play key roles in dictating stem cells’
generate a diverse range of other cell types, and are fate and function, for example. Among these are signaling
therefore considered cancer stem cells. pathways headed by the Bmi-1, Notch, Sonic hedgehog and
n T hese malignant progenitors are believed to spring Wnt genes. Yet most of these genes were first identified not
from regulatory failures in damaged stem cells or their by scientists studying stem cells but by cancer researchers,
immediate offspring. because their pathways are also involved in the development
n C ancer treatments must target cancer stem cells to of malignancies.
eradicate the disease. Many such similarities between stem cells and cancer cells
have been noted. The classical definition of malignancy itself

42 S C I E N T I F I C A M E R I C A N ne w ans w ers f o r C ancer


© 20 08 SCIENTIFIC AMERIC AN, INC.
hierarchy in blood-forming cells
Stem cells in the blood-forming, or hematopoietic, system and another short-lived daughter termed a multipotent
illustrate principles governing the activity of stem cells in progenitor cell (MPP). The MPP, in turn, divides to produce
other tissues as well. A small population of hematopoietic progenitors committed to generating cells in the myeloid
stem cells (HSCs) in the bone marrow is the source of most (blood) or lymphoid (immune) lineages. As the descendants
of the different blood and immune cell types that circulate of progenitors become increasingly specialized, they
in the human body. HSCs reside in an environmental niche, experience a programmed decline in their ability to
surrounded by stromal cells that provide important regulatory proliferate until they stop dividing and are said to be
signals to the stem cell. When new blood or immune cells are terminally differentiated. Only the stem cell retains unlimited
needed, an HSC divides to produce one daughter cell that proliferative potential through its ability to renew itself
remains in the niche and retains the long-term HSC identity indefinitely by dividing without differentiating.

Macrophage

Bone marrow
Neutrophil
Granulocytic/
monocytic Basophil
precursor
Eosinophil

Myeloid
progenitor Platelets

Megakaryocyte
Stromal cell
Hematopoietic
stem cell Megakaryocytic/
erythrocytic Red blood cell
Multipotent precursor
progenitor cell Erythroblast

Plasma
Niche cell
B cell
precursor
T cell
Lymphoid
progenitor
Natural
T cell killer cell
precursor

Pluripotent stem cell Committed progenitors Terminally differentiated cells

Self-renewal Degree of differentiation


Proliferation potential

includes cancer cells’ apparent capacity to survive and multi- them from the rules limiting life span and proliferation for
ply indefinitely, their ability to invade neighboring tissues and most cells. Stem cells’ ability to differentiate into a broad range
to migrate (metastasize) to distant sites in the body. In effect, of cell types allows them to form all the different elements of
the usual constraints that tightly control cellular prolifera- an organ or tissue system. A hallmark of tumors, too, is the
T ami T o l p a

tion and identity seem to have been lifted from cancer cells. heterogeneity of cell types they contain, as though the tumor
Normal stem cells’ power to self-renew already exempts were a very disorderly version of a whole organ. Hematopoi-

w w w. S c i A m . c o m SCIENTIFIC A MERIC A N 43
© 20 08 SCIENTIFIC AMERIC AN, INC.
etic stem cells have been shown to migrate to distant parts of oncogenic scars, a stem cell’s enormous proliferative capac-
the body in response to injury signals, as have cancer cells. ity makes it an ideal target for malignancy. Because nature
In healthy stem cells, strict genetic regulation keeps their so strictly regulates self-renewal, a cell population already
potential for unlimited growth and diversification in check. possessing that ability would need fewer additional muta-
Remove those control mechanisms, and the result would be tions for malignant transformation than would cells lacking
some­thing that sounds very much like malignancy. These that capacity.
commonalities, along with growing experimental evidence, With these considera­tions in mind, several possible paths
suggest that failures in stem cell regulation are how many to malignancy become apparent. In one model, mutations
cancers get started, how they perpetuate themselves, and pos- occur in the stem cells themselves, and their resulting loss of
sibly how malignancies can spread. control over self-renewal decisions produces a pool of stem
cells predisposed to malignancy. Subsequent additional on-
Achilles’ Heel cogenic events that trigger proliferation of the malignant cells
t h e p r e se n c e of s t e m c e l l s in certain tissues, espe- into a tumor might happen in the stem cells or in their descen-
cially those with high cell turnover such as the gut and the dants, the committed progenitor cell population. A second
skin, seems to be an overly complicated and inefficient system model holds that oncogenic mutations initially occur in stem
for replacing damaged or old cells. Would it not appear to cells but that the final steps in transformation to cancer hap-
make more sense for an organ- pen only in the committed pro-
ism if every cell could simply genitors. This scenario would
proliferate as needed to supply Several possible paths require the progenitors’ lost
replacements for its injured
neighbors? On the surface, to malignancy self-renewal capacity to be
somehow reactivated.
perhaps— but that would make
every cell in the body a poten-
become apparent. Current evidence supports
both models in different can-
tial cancer cell. cers. And at least one example
Malignancies are believed to arise when an accumulation exists of both processes playing a role in different stages of the
of “oncogenic” changes to key genes within a cell leads to the same disease. Chronic myelogenous leukemia is a cancer of
abnormal growth and transformation of that cell. Gene mu- the white blood cells caused by the inappropriate fusion of
tations typically happen through a direct insult, such as the two genes. Insertion of the resulting fused gene will transform
cell being exposed to radiation or chemicals, or simply a normal hematopoietic stem cell into a leukemia stem cell.
through random error when the gene is improperly copied Untreated, CML invariably progresses to an acute form
before cell division. Because the rare stem cells are the only known as CML blast crisis. Catriona Jamieson and Irving
long-lived cells in the organs where most cancers develop, Weissman, both then at the Stanford University School of
they represent a much smaller potential reservoir for cumula- Medicine, demonstrated that in patients who progressed to
tive genetic damage that could eventually lead to cancer. Un- CML blast crisis, the specific additional genetic events respon-
fortunately, because stem cells are so long-lived, they also sible for this more virulent version of the disease had con-
become the most likely repository for such damage. ferred the ability to self-renew on certain progenitor cells.
Indeed, stem cells’ longevity would explain why many
cancers develop decades after tissues are subjected to radia- Steady Pursuit
tion — the initial injury may be only the first in a series of ov e r t h e pa s t de c a de , evidence that stem cells could
mutations required to transform a healthy cell into a malig- become malignant and that only certain cancer cells shared a
nant one. In addition to accumulating and preserving these variety of traits with stem cells strengthened the idea that the
driving force underlying tumor growth might be a subpopula-
MICHAEL F. CLARKE and MICHAEL W. BECKER worked together tion of stemlike cancer cells. The theory has a longer history,
the authors

in Clarke’s laboratory at the University of Michigan at Ann Arbor, but in the past the technology to prove it was lacking.
where breast tumor stem cells were first isolated in 2003. By the 1960s a few scientists were already beginning to
Clarke is now associate director, as well as professor of cancer note that groups of cells within the same tumor differed in
biology and of medicine, at the Stanford Institute for Stem Cell their ability to produce new tumor tissue. In 1971 C. H. Park
Biology and Regenerative Medicine. He continues to work on and his colleagues at the University of Toronto showed that
identifying cancer stem cells and the mechanisms by which within a culture of cells taken from an original, or “primary,”
they, as well as normal stem cells, regenerate. Becker is as- myeloma (a cancer affecting plasma cells in bone marrow),
sistant professor of medicine in the hematology and oncology the cells displayed significant differences in their ability to
division of the University of Rochester Medical Center. Becker’s proliferate. At the time, Park’s group could not interpret this
research focus is characterizing leukemic stem cells, and his phenomenon decisively, because at least two explanations
clinical work centers on peripheral blood and bone marrow were possible: all the cells might have had the ability to mul-
transplantation. tiply in culture but by chance only some of them did, or else

44 S C I E N T I F I C A M E R I C A N ne w ans w ers f o r C ancer


© 20 08 SCIENTIFIC AMERIC AN, INC.
possible paths to cancer
The existence of cancer stem cells that drive tumor growth programming in concert with signals they receive from their
has been established in several blood cancers and a handful environmental niche. Changes in the way cancer stem cells
of solid-tumor types, but how these malignant stem cells carrying oncogenic gene mutations respond to niche signaling
arise is still uncertain. Like a normal stem cell, a cancer may therefore play an important role in the cells’ final
stem cell has the ability to self-renew by dividing without transition to malignancy (a, b and c). Alternatively, mutations
differentiating and can therefore potentially give rise to an in the stem cell might be preserved in its immature
unlimited number of the abnormal differentiated cells that descendants, the progenitor cells, which subsequently
make up the bulk of a tumor. Those progeny have a finite life undergo further mutations that reactivate the self-renewal
span and are not themselves tumorigenic — that is, they capacity normally possessed only by stem cells (d).
cannot generate new cancer cells. The behavior of normal Evidence of all these possibilities has been observed in
stem cells is tightly controlled by their own genetic different cancers.

Cancer Abnormal Abnormal Alternative


Niche cell
stem cell progenitor cell differentiated cell niche cell

a b

Tumor

Niche

Expanded niche. Cancer stem cells with oncogenic Alternative niche. Oncogenic mutations in cancer stem
mutations are held in check by healthy niche signals until a further cells include changes that enable the cells to adapt to a new niche.
alteration in the cancer stem cells or in the niche causes the niche The cancer stem cells can expand their own numbers and prolif­er­
to expand. The larger niche allows the malignant stem cells ation and possibly invade neighboring tissues or metastasize
to increase their own population and consequently the number to distant locations in the body.
of abnormal cells they generate.

c d Self-renewing
malignant progenitor
Mutated
stem cell

Niche independence. Mutation renders stem cells that Self-renewal mutation. Progenitor cells predisposed to
are already predisposed to malignancy independent of niche malignancy by oncogenic mutations inherited from their parent stem
signaling, lifting all normal environmental controls on the cancer cell undergo further mutation that restores the ability to self-renew.
tami t o l p a

stem cells’ self-renewal and proliferation. These progenitors thereby gain unlimited life span and tumorigenic
capacity and become cancer stem cells.

w w w. S c i A m . c o m
© 20 08 SCIENTIFIC AMERIC AN, INC.
SCIENTIFIC A MERIC A N 45
Cornering Cancer Stem Cells and this new mouse model, Dick began in 1994 to publish a
series of seminal reports identifying cancer stem cells in leuke-
Techniques for sorting live cancer cells and for determining mia. In 2003 Richard Jones of Johns Hopkins University iden-
whether they possess the ability to self-renew have led tified a cancer stem cell population in multiple myeloma.
to the positive identification of cancer stem cells within Earlier the same year our own laboratory group at the
larger cancer cell populations. In the cancers listed below, University of Michigan at Ann Arbor had published the first
the malignant stem cells have been demonstrated capable
evidence of cancer stem cells in solid tumors. By transplanting
of self-renewal and able to regenerate the entire mixture
sorted populations of cells from human breast tumors into
of cell types found in the original tumor. Those properties
mice, we were able to confirm that not all human breast can-
mean that a small number of cancer stem cells could have
cer cells have the same capacity to generate new tumor tissue.
given rise to the whole tumor, could continually replenish
its much larger cell population — the majority of which is Only one subpopulation of the cells was able to re-create the
nontumorigenic— and could reconstitute the original cancer original tumor in the new environment. We then compared
even if most or all of the tumor were destroyed. Eradicating the phenotype, or physical traits, of those new tumors with
the disease would therefore require treatments to target that of the patient samples and found that the profile of the
the cancer stem cells successfully. new tumors recapitulated the original. This finding indicated
that the transplanted tumorigenic cells could both self-renew
Cancer Type (year cancer stem cells identified) and give rise to all the different cell populations present in the
Acute myeloid leukemia (1994) original tumor, including the nontumorigenic cells.
Acute lymphoblastic leukemia (1997) Our study attested to the presence of a hierarchy of cells
Chronic myelogenous leukemia (1999) within a breast cancer similar to those identified in blood
Breast (2003) malignancies. Since then, the investigation of cancer stem cell
Multiple myeloma (2003) biology has exploded, as labs across the world continue to
find similar subpopulations of tumorigenic cells in other
Brain (2004)
forms of cancer. In 2004, for example, the laboratory of Peter
Prostate (2005)
Dirks of the University of Toronto identified cells from pri-
Lung (2005) mary human central nervous system tumors with the capac-
Colorectal (2006) ity to regenerate the entire tumor in mice. Since then, the
Pancreatic (2007) phenotypes of cancer stem cell populations for colorectal
Melanoma (2008) cancer, lung cancer, melanoma, prostate cancer and pancre-
Ovarian (2008) atic cancer have been reported.
A related area of recent intensive investigation is also pro-
a hierarchy of cells was present in the tumor and cancer stem viding support for the cancer stem cell model. The signaling
cells were giving rise to cells that were nontumorigenic, or environment, or niche, in which tumors reside appears to
incapable of proliferation. strongly influence the initiation and maintenance of malig-
Philip J. Fialkow of the University of Washington had al- nancy. Studies of normal body cells as well as of stem cells
ready demonstrated in 1967 that the stem cell model was prob- have already established the essential role of signals emanat-
ably the correct one for leukemia. Using a cell-surface protein ing from surrounding tissue and the supportive extracellular
marker called G-6-PD, which can identify a cell’s lineage, Fi- matrix in sustaining a given cell’s identity and in directing
alkow showed that in some women with leukemia, both the its behavior. Normal cells removed from their usual context
tumorigenic cells as well as their more differentiated nontum- in the body and placed in a dish have a tendency to lose some
origenic progeny had all arisen from the same parent cell. of their differentiated functional characteristics, for exam-
These early studies were critical in the development of the ple. Stem cells, in contrast, must be cultured on a medium
stem cell model for cancer, but they were still limited by re- that provides signals telling them to remain undifferentiated,
searchers’ inability to isolate and examine different cell pop- or they will quickly begin proliferating and differentiating—
ulations within a tumor. A key event in stem cell biology, seemingly as though that is their default programmed
therefore, was the commercial availability, beginning in the ­behavior, and only the niche signals hold it in check.
1970s, of an instrument called a flow cytometer, which can In the body, stem cell niches are literal enclaves surround-
automatically sort different living cell populations based on ed by specific cell types, such as stromal cells that form con-
the unique surface markers they bear. nective tissue in the bone marrow. With a few exceptions,
A second crucial event in the evolution of cancer stem cell stem cells always remain in their niche and are sometimes
studies was the advent during the 1990s of conclusive tests for physically attached to it by adhesion molecules. Progenitor
self-renewal. Assays to establish self-renewal in human cells cells, on the other hand, move away from the niche, often
did not exist until Weissman of Stanford and John E. Dick of under escort by guardian cells, as they become increasingly
the University of Toronto developed methods that allowed nor- differentiated.
mal human stem cells to grow in mice. Using flow cytometry The importance of niche signaling in maintaining stem

46 S C I E N T I F I C A M E R I C A N ne w ans w ers f o r C ancer


© 20 08 SCIENTIFIC AMERIC AN, INC.
cells’ undifferentiated state and in keeping them quiescent Stem cells cannot be identified based solely on their ap-
until they are called on to produce new cells suggests that pearance, however, so developing a better understanding of
these local environmental signals could exert similar regula- the unique properties of cancer stem cells will first require
tory control over cancer stem cells. Intriguing experiments improved techniques for isolating and studying these rare
have shown, for example, that when transplanted into a new cells. Once we learn their distinguishing characteristics, we
niche, stem cells predisposed to malignancy because of onco- can use that information to target cancer stem cells with tai-
genic mutations will nonetheless fail to produce a tumor. Con- lored treatments. If scientists were to discover the mutation or
versely, normal stem cells transplanted into a tissue environ- environmental cue responsible for conferring the ability to
ment that has been previously damaged by radiation do give self-renew on a particular type of cancer stem cell, for in-
rise to tumors. stance, that would be an obvious target for disabling those
Many of the same genetic pathways identified with signal- tumorigenic cells.
ing between stem cells and their niche have been associated Encouraging examples of this strategy’s promise have been
with cancer, which also suggests a role for the niche in the final demonstrated by Craig T. Jordan and Monica L. Guzman of
transition to malignancy. For example, if malignant stem cells the University of Rochester. In 2002 they identified unique
were being held in check by molecular features of malig-
the niche but the niche was
somehow altered and ex- Destroy the engine nant stem cells believed to
cause acute myeloid leukemia
panded, the malignant stem
cell pool would have room to
driving the disease, leaving (AML) and showed that the
cancer stem cells could be
grow as well. Another possi-
bility is that certain oncogen-
nontumorigenic cells preferentially targeted by spe-
cific drugs. In 2005 they re-
ic mutations within cancer
stem cells could permit them
to die off. ported their discovery that a
compound derived from the
to adapt to a different niche, again letting them increase their feverfew plant induces AML stem cells to commit suicide while
numbers and expand their territory. Still a third alternative is leaving normal stem cells unaffected.
that mutations might allow the cancer stem cells to become Some research groups are hoping to train immune cells or
independent of niche signals altogether, lifting environmental antibodies to recognize and go after cancer stem cells. Tobias
controls on both self-renewal and proliferation. Schatton, Markus H. Frank and their co-workers at Children’s
Hospital Boston showed that an antibody could inhibit the
Closing In growth of melanoma stem cells. Yet another idea under inves-
t h e i m p l ic at ion s of a stem cell model of cancer for the tigation is that drugs could be developed to force cancer stem
way we understand as well as treat malignancies are clear and cells to differentiate, which should take away their ability to
dramatic. Current therapies take aim against all tumor cells, self-renew.
but our studies and others have shown that only a minor frac- Most important is that cancer investigators are now on
tion of cancer cells have the ability to reconstitute and per- the suspects’ trail. With a combination of approaches, aimed
petuate the malignancy. If traditional therapies shrink a tu- at both targeting genetic pathways unique to the mainte-
mor but miss these cells, the cancer is likely to return. Shideng nance of cancer stem cells and disrupting the cross talk be-
Bao, Jeremy N. Rich and their colleagues at Duke University tween tumor cells and their environment, we hope to be able
showed that this was indeed the case for glioblastoma, where soon to find and arrest the real culprits in cancer. 
the stem cells were resistant to radiation. Treatments that
specifically target the cancer stem cells could destroy the en- more to explore
gine driving the disease, leaving any remaining nontumori- The Reversal of Tumor Growth. Armin C. Braun in Scientific American,
genic cells to eventually die off on their own. Vol. 213, No. 5, pages 75–83; November 1965.
Circumstantial evidence supporting this approach already The Proteus Effect: Stem Cells and Their Promise for Medicine.
Ann B. Parson. Joseph Henry Press, 2004.
exists in medical practice. Following chemotherapy for tes-
Context, Tissue Plasticity, and Cancer: Are Tumor Stem Cells Also
ticular cancer, for example, a patient’s tumor is examined to
Regulated by the Microenvironment? Mina J. Bissell and Mark A.
assess the effects of treatment. If the tumor contains only LaBarge in Cancer Cell, Vol. 7, pages 17–23; January 2005.
mature cells, the cancer usually does not recur and no further Glioma Stem Cells Promote Radioresistance by Preferential
treatment is necessary. But if a large number of immature- Activation of the DNA Damage Response. S. Bao et al. in Nature,
looking— that is, not fully differentiated— cells are present in Vol. 444, pages 756–760; December 7, 2006.
the tumor sample, the cancer is likely to return, and standard The Prognostic Role of a Gene Signature from Tumorigenic Breast-
Cancer Cells. R. Liu, X. Wang, G. Y. Chen, P. Dalerba, A. Gurney, T. Hoey,
protocol calls for further chemotherapy. Whether those im-
G. Sherlock, J. Lewicki, K. Shedden and M. F. Clarke in New England
mature cells are recent offspring that indicate the presence of Journal of Medicine, Vol. 356, No. 3, pages 217–226; January 18, 2007.
cancer stem cells remains to be proved, but their association Identification of Cells Initiating Human Melanomas. T. Schatton et al.
with the disease prognosis is compelling. in Nature, Vol. 451, pages 345–349; January 17, 2008.

w w w. S c i A m . c o m SCIENTIFIC A MERIC A N 47
© 20 08 SCIENTIFIC AMERIC AN, INC.
Understanding chronic inflammation, which
contributes to heart disease, Alzheimer’s
and a variety of other ailments, may be

FLA
a key to unlocking the mysteries of cancer
by Gary Stix

A Malignant
M
ore than 500 million years ago a set nate immune system, basically thinking of it
of specialized enzymes and proteins as a crew of biochemical bouncers that pum-
evolved to defend our primitive mel anything able to penetrate the tiniest
ancestors against assaults from the outside opening in a living being’s skin or shell. They
world. If a microbe breached the shell of some lavished their attention, instead, on the more
Cambrian-era fauna, the members of this ear- advanced adaptive immune system, which
ly-vintage immune system would stage a sav- can marshal antibodies and other weaponry
age but coordinated attack on these interlop- that identify and then target an intruder with
ers— punching holes in cell walls, spitting out a specificity lacking in the untamed innate
chemical toxins, or simply swallowing and system.
digesting the enemy whole. Once the invaders In the past 15 years, innate immunity has
were dispatched, the immune battalion would come into its own. Inflammation, its hall-
start to heal damaged cells, or if the attacked mark characteristic, has gained recognition
cells were too badly damaged it would put as an underlying contributor to virtually ev-
them to rest. ery chronic disease — a list that, besides obvi-
This inflammatory immune response ous culprits such as rheumatoid arthritis and
worked so well that many aspects of it have Crohn’s disease, includes diabetes and depres-
been preserved during the protracted aeons sion, along with major killers such as heart
of evolution. We know this to be true because disease and stroke [see “Common Cause,” on
studies have found that we share many of the page 53]. The possibility of a link with a
same immune genes as the lowly fruit fly— third major killer— cancer— has received in-
and vertebrates and invertebrates diverged tensive scrutiny in this decade. “The connec-
from a common ancestor in excess of half a tion between inflammation and cancer has
billion years ago. moved to center stage in the research arena,”
Immunology researchers have long paid notes Robert A. Weinberg of the Massachu-
relatively little attention to this thuggish in- setts Institute of Technology’s Whitehead
Institute for Biomedical Research, who has
Overview/Hijacked Healing highlighted the changing emphasis in a revi-
sion of his leading textbook, The Biology of
n  Until recently, cancer researchers had focused primarily on genetic Cancer (Garland Science, 2006).
changes as the underlying cause of the disease. This transformation recognizes that the
n In this decade investigators have come to realize that the developing immune inflammatory state serves as a key
tumor can commandeer the immune system’s inflammatory mediator of the middle stages of tumor de-
component— normally part of the wound-healing process — to foster velopment. Cancer begins with a series of
carcinogenesis. genetic changes that prompt a group of cells
n A new generation of anti-inflammatory drugs may join traditional to overreplicate and then invade surround-
chemotherapies, which in combination could keep solid tumors or ing tissue, the point at which true malignan-
premalignancies localized. cy begins. Eventually some tumor cells may
break off and establish new growths (metas-

48 SCIENTIFIC A MERIC A N ne w ans w ers f o r C ancer


© 20 08 SCIENTIFIC AMERIC AN, INC.
ME tases) at distant sites. That much has
been understood for a long time. But can-
cer biologists and immunologists have be-
gun to realize that the progression from dis-
eased tissue to full-blown invasive cancer
often requires cells that normally participate
in healing cuts and scrapes to be diverted to
the environs of the premalignant tissue,
where they are hijacked to become co-con-
spirators that aid and abet carcinogenesis. As
some researchers have described the malig-
nant state: genetic damage is the match that
lights the fire, and inflammation is the fuel
that feeds it.
In this rewriting of the textbooks, a tu-
mor is not just a clump of aberrant cells;
it also includes a support system, a tumor
microenvironment, which encompasses a
multitude of varying immune cell types and
crisscrossing chemical signals, along with a
network of blood vessels. The tumor assumes
the status of an outlaw organ that exists not
to pump blood or rid the body of tox-
ins but to serve only its own ends.
This new view implies that root-
ing out every last cancer cell in the body might
not be necessary. Anti-inflammatory cancer
therapy instead would prevent premalignant
cells from turning fully cancerous or would
impede an existing tumor from spreading to
j eff j o h ns o n H y b r i d M e d i c a l A n i m a t i o n

distant sites in the body. Cancer sufferers


might then be able to survive, in the same way
that new drugs have let HIV patients live lon-
ger. “I don’t think a cure is necessarily the
goal. It doesn’t need to be,” comments Lisa M.
Coussens, a cancer biologist at the University
TUMOR DE V ELOPMENT progresses in some cancers through the
of California, San Francisco. “If you can effects of what cancer biologists have labeled a “smoldering”
manage the disease and live your natural life inflammation, in which the tumor recruits immune cells that linger
span, that’s a huge win.” in its surroundings and within the malignant mass.

w w w. S c i A m . c o m
© 2008 SCIENTIFIC AMERIC AN, INC.
Multiple Lines of Defense man pathologist Rud­olf Virchow noted the
c om p r e h e n sion of the link between in- presence of so-called lymphoreticular infil-
flammation and cancer requires knowing how trate (white blood cells) in malignant tissue.
the body reacts to invaders— and how normal As early as 1978 Alberto Mantovani of Hu-
healing is then subverted into promoting can- manitas Clinical Institute and the University
cer when the inflammatory state lasts too long. of Milan had observed that innate immune
THE PLAYERS After you step on a nail, the bacteria that in- cells tend to congregate around some tumors.
The immune system consists of vade the sole of your foot receive a welcome Cancer biologist Harold F. Dvorak of Har-
innate cells, which form a first line from an array of proteins and white blood cells vard Medical School remarked in 1986 that
of defense against pathogens, that resemble rejects from central casting for tumors are “wounds that do not heal.” The
and members of the adaptive
the movie Creepshow 2. Just one example: status quo, though, lay elsewhere. Even a de-
system, which targets invaders
with greater specificity. Some 20 complement proteins, so called be- cade ago many biologists still hewed to the
cause they complement other bodily defense idea that the immune system serves not only
innatE mechanisms, chemically spritz pathogens until to eliminate pathogens but to ferret out cells
Macrophage the invaders explode into a big protoplasmic that are the abnormal precursors of cancer.
This immune defender mess. While the complement system slimes the But a closer look at the microenvironment
engulfs and consumes area, an assemblage known in immunology surrounding tumors found the unexpected.
pathogen invaders.
textbooks as professional phagocytes— literal-
Mast cell
ly “expert eating cells”— goes to work. Hunting Pigeons
This cell releases histamine
Lacking table manners, these Pac-Man- i n t h e l a t e 19 9 0 s Frances Balkwill of
and other chemicals that like macrophages and neutrophils proceed to the Institute of Cancer at Queen Mary, Uni-
promote inflammation. engulf and consume the uninvited guests. versity of London, had been doing research
Other members of the attack brigade include on a cytokine (a hormonelike immune signal-
Granulocyte natural killer cells, mast cells and eosino- ing molecule) known as tumor necrosis factor
Three cell types with tiny phils. Healing represents more than launch- (TNF), which was named for its ability to kill
granules in their interior—
ing an offensive against invaders. Blood cancer cells when administered directly into
the neutrophil,
eosinophil and platelets involved with clotting migrate to the a tumor at high levels. But when TNF lingers
basophil— par­ break in the skin from an inner layer infused as a chronic, low-level presence in the tumor,
ticipate in the with blood vessels. Enzymes direct the repair it acts very differently. Balkwill’s lab turned
inflammatory response. of the extracellular matrix, the protein-based off the TNF gene in mice so that the rodents
mortar in which the cells are immobilized. A could not produce the protein: to the re-
Dendritic cell scab forms, the skin grows back and the searchers’ surprise, the mice did not contract
It presents antigens —
fragments of protein or other
whole process of inflammation ends. Some- tumors. “That really put us as the cat among
molecules from pathogens or times, though, inflammation does not stop. pigeons,” she recalls. “All the people who
cancer cells — to adaptive immune Any tissue (not just skin) that is chronically were working on TNF as an anticancer agent
cells, inducing the cells to attack inflamed because of the persistent presence were horrified. This cytokine they thought
bearers of the displayed antigens. of pathogens, toxins or genetic damage helps was a treatment for cancer was actually work-
to spur illness, from heart disease to cancer. ing as an endogenous tumor promoter.”
Natural killer Beyond this first layer of defense, verte- The ready availability of knockout mice,
cell
This cell destroys the body’s
brates are equipped with additional weap- in which the effects of selectively switching
own cells that have become infected onry. The adaptive system learns an invader’s off genes could be tested, helped to highlight
with pathogens; it also goes after specific molecular signature and then uses it the cancer-inflammation link. Coussens and
cancer cells. as a target for killing. Among the protago- her U.C.S.F. colleagues Douglas Hanahan
nists are B cells, which produce antibody and Zena Werb reported in 1999 that mice
ADAPTIVE molecules able to neutralize pathogens or engineered with activated cancer genes but
B cell mark them for destruction, and T cells, which without mast cells (another type of innate im-
Antigens stimulate this
cell to divide and produce
prompt infected cells to kill themselves or se- mune cell) developed premalignant tissue that
antibodies that neutralize invaders crete chemicals that direct the activities of did not progress to full malignancy. In 2001
or tag them for killing. other immune players. Jeffrey W. Pollard and his co-workers at the
In recent years a body of evidence has ac- Albert Einstein College of Medicine described
T cell cumulated to show that chronic inflamma- mice that were genetically engineered to be
A killer T cell destroys an tion can play an important role in the pro- susceptible to breast cancer tumors but that
infected cell in which
gression of some types of tumors from a pre- produced precancerous tissue that did not
it detects the presence of
antigens. Other T cells—such as malignant state to full-blown disease. A link turn fully malignant unless it enlisted the as-
helper and regulatory types— between cancer and inflammation has long sistance of macrophages.
coordinate the immune response. been suspected. In 1863 the prominent Ger- The altered picture does not completely

50 SCIENTIFIC A MERIC A N ne w ans w ers f o r C ancer


© 20 08 SCIENTIFIC AMERIC AN, INC.
overturn the old one. In fact, it reveals that but not all— and its link to blood-borne can- CANCER BASICS
the immune system functions as a double- cers is not well characterized.
A developing malignancy
edged sword. The network of molecules and When looking for culprits, researchers proceeds in stages—a process
cells, second in complexity only to the brain, have often focused their microscopes on that may take years, even
remains a paradox: sometimes it promotes macrophages, which occupy a meaningful decades, to fully evolve.
cancer; other times it hinders disease. Some spot among the white blood cells in the tu-
types of innate immune cells, such as natural mor microenvironment. The macrophages INITIATION
killer cells, can actually protect against tu- are capable of killing tumor cells or sending Hereditary mutations or exposure to
chemicals or radioactivity results in
mor growth. Others may nurture a malig- out an alarm to T cells of the adaptive im- genetic changes in one or more cells.
nancy only when the microenvironment is mune system that something is amiss. But
“polarized” into an inflammatory state; work by Pollard and other researchers has PROMOTION
when not, they may blot it out. Inflammation, detailed how macrophages are “reeducated” Cells in premalignant tissue be­g in
moreover, produces tumors in many organs by cancer cells to do their bidding. They be- to proliferate, often in the
presence of an inflammatory
stimulus. Their appearance
cancer hijacks wound healing becomes increasingly abnormal.
Innate immunity responds to an insult, such as a puncture wound, with a cellular and chemical
arsenal. Cancer biologists have recently begun to understand how chronically inflamed PROGRESSION
Tumor cells begin to invade
premalignant tissue uses many of the same biochemical players to promote cancer development.
surrounding tissue and to spread
to the blood and lymph nodes,
at which point full malignancy
Wound
develops. Metastases may
Skin Clot establish themselves at
distant sites.

Bacterium Epithelial cells


Fibrin Antimicrobial toxins
Platelet
Neutrophil
normal wound healing Mast cell
Cells of the innate immune system,
including macro­phages, neutrophils
and mast cells, converge on bacteria,
either consuming them or spraying
them with toxins. Some of the innate Macrophage
Cytokines
cells emit chemical signals (cytokines)
to coordinate the defensive actions.
Meanwhile blood platelets and a
protein called fibrin form a clot. Blood vessel

WOUND HEALING RUN AMOK Tumor cells


Chronically inflamed precancerous
tissue commandeers many of the
same innate cells and signals
present in the acute immune
response to help the tumor grow and
develop a blood supply. Ultimately
the tumor starts to invade
surrounding tissue, the point at
which it becomes fully malignant.

Macrophage

Cytokines
Epithelial cells Neutrophil
j en c h ristiansen

Mast cell

Blood vessel

w w w. S c i A m . c o m SCIENTIFIC A MERIC A N 51
© 20 08 SCIENTIFIC AMERIC AN, INC.
the immune paradox
Two arms of the immune
system — the innate and E Activated ●
1 A dendritic or other cell (the exact one
has not been identified experimentally)
NS E
B cell
the adaptive — are “displays” a piece of the tumor to adaptive
TI V B cell

PO
P

ES
exquisitely well adapted T and B cells.

A
GR

AD
for fighting pathogens,

T IN
2 In some instances,
but their role in combating

T U M O R- S U P P O R
T cells and B cell–
cancer is decidedly more produced antibodies
paradoxical. The innate identify and destroy
system furnishes an Antibodies
tumor cells, as they
initial inflammatory would a pathogen.
Antigen
response to a microbial Tumor
insult by attacking any T cell
invading pathogen Inflammation
indiscriminately, whereas response
TUMOR B cell
adaptive immunity TUMOR RETREATS
furnishes a delayed IN N A GROWS
TE Attacked
response that homes in tumor cell
on a particular pathogen.
In cancer, both systems

E
may sometimes attack

P ON S
tumor cells. But a tumor ●
3 B
 ut when B cells receive
certain yet to be delineated

RE S
protects itself by signals, they produce

IN G
recruiting the innate antibodies that activate the TI

VE
AP

HT
system to enhance innate immune system to AD

IG
F
its development. help tumor cells to survive
Antibodies
O R-
and flourish. TUM

come factories for cytokines and growth fac- In 2004 Yinon Ben-Neriah and Eli Pikar-
Genetic damage tors that nurture tumor development. sky of the Hebrew University of Jerusalem and
Turning the macrophages into traitors be- their colleagues reported that mice engineered
is the match gins when tumor cells send out help signals to develop hepatitis (which can cause liver can-
that lights that attract cells that become macrophages cer) contracted precancerous lesions that did
once they reach the tumors. Inside the tu- not progress to full malignancy when NF-kB
the fire of mors, proliferating cells grow so quickly that was curtailed through a genetic alteration or
malignancy, and they begin to die for lack of oxygen. A com- when the proinflammatory TNF signaling
bination of hypoxia and messages from the molecule was shut off. In the latter group, a
inflammation is tumor cells initiates a process whereby the neutralizing antibody blocked TNF and pre-
the fuel that newly arrived macrophages assume their vented it from binding to a receptor on the pre-
bad-boy identity as tumor promoters. Cancer malignant liver cells; loss of the receptor pre-
feeds the flames. biologists give the name tumor-associated vented the TNF from triggering a molecular
macrophages to these mutineers that congre- cascade that turns on the NF-kB master switch.
gate in and around the tumor. Blocking NF-kB prompted the precancerous
Biologists have now been able to follow liver cells to initiate apoptosis, or programmed
the inflammation link down to the level of in- cell death. In a related finding that year, Mi-
dividual signaling molecules, providing hard- chael Karin and his collaborators at the Univer­
er evidence for a connection to carcinogenesis. sity of California, San Diego, found that inhib-
For example, nuclear factor-kappa B (NF-kB) iting NF-k B in mice engineered to develop
is a complex of proteins that acts as a master colitis, which can lead to colon cancer, also
switch for turning inflammation genes on and promoted apoptosis. Shutting down the path-
for controlling cell death. As biological path- way in inflammatory cells, such as macrophag-
ways go, NF-k B’s is world-famous, having es, deterred tumor development as well.
been discovered and patented for use in drug So far the clearest evidence of a link be-
j en c h ristiansen

development by scientific stars that include tween cancer and inflammation is the data
Nobelists David Baltimore and Phillip A. demonstrating that inflammation encourages
Sharp and having subsequently become the the conversion of precancerous tissue to full
object of multimillion-dollar patent litigation. malignancy for many cancers. But the biolog-

52 SCIENTIFIC A MERIC A N ne w ans w ers f o r C ancer


© 20 08 SCIENTIFIC AMERIC AN, INC.
ical response may also be involved in initiating cancer. The research suggests that a cytokine
the disease and in advancing metastasis. In- produced by inflammatory cells near a pros-
fections with Helicobacter pylori bacteria in- tate tumor induces cancer cells to decrease
duce inflammation that greatly increases the production of a protein that blocks metasta-
risk of gastric cancer, and the hepatitis C virus sis. This result, Karin notes, may explain the
can bring on liver cancer, to name just two puzzling observation that cutting into tumors,
malignancies. Pathogens may also generate such as for a prostate biopsy, sometimes
free radicals, which can damage DNA. But seems to encourage metastasis. If he is correct,
although inflammation may be involved from the inflammation generated by the interven-
the outset, few studies have shown yet that an tion could be at fault. Around the same time,
inflammatory condition actually alters DNA Pollard’s group reported in Cancer Research
to provide the initiating spark. on a study in mice that observed that mac-
The case for a role in metastasis is stron- rophages accompany breast tumor cells in
ger— and recent studies lend credence to this their migration toward blood vessels that
hypothesis. Karin’s group reported in the then transport them to remote sites.
April 5, 2007, Nature that inflammation, not The innate immune system has received
genetic changes in cancer cells, spurs metas- the most attention in explorations of how in- COMMON CAUSE
tasis in mice engineered to acquire prostate flammation might cause cancer. As with in- Chronic inflammation contributes
to many diseases, not just cancer.

Immune System as Cancer Fighter HEART DISEASE

T
Macrophages, stars of innate
umors waylay the immune system to concept of the cancer vaccine has garnered its immunity, are key players. They
promote their own growth and survival. share of critics. “When you give cancer ingest “bad” cholesterol (low-
But the opposite also holds. The vaccines, sometimes you raise T cells against density lipoproteins), and then the
antibodies and killer T cells of the adaptive the vaccines, but they’re just not powerful cells are encased in a fibrous cap
immune system can, at times, target and enough to keep the cancer from growing,” notes that forms arterial plaque, which
destroy cancer cells. Drug companies and Steven A. Rosenberg, chief of surgery at the can break off and create a clot
that blocks an artery, leading
scientists have tried to turn this knowledge National Cancer Institute. Rosenberg and his to a heart attack.
into new therapies, with mixed results. co-workers have pursued a different approach,
Among the most successful new adoptive cell transfer, in which T cells that DIABETES
biotechnology drugs are monoclonal target tumors are selected from white blood When exposed to the metabolic
antibodies — identical antibodies that are cells removed from the body. They are then stress that occurs from being obese,
capable of attacking a cancer antigen, a multiplied in the laboratory and reinfused into both innate immune cells and fat
molecular fragment found on the surface patients whose immune systems have been cells (adipocytes) manufacture
of a cancer cell. Monoclonals are generally chemically suppressed. In a 2005 study in the signaling molecules called
cytokines, such as tumor necrosis
a “passive” immunotherapy because they Journal of Clinical Oncology, about half of 35
factor. These molecules interfere
are produced in cell culture or in mice and melanoma patients saw their metastatic with the normal function of insulin
injected into patients instead of relying tumors regress at least partially. and can lead to diabetes.
on the patient’s own immune system to In recent years, the picture for cancer
produce antibodies. vaccines has brightened somewhat. Early ALZHEIMER’S DISEASE
In contrast, cancer vaccines — the object positive reports were presented at the Microglial cells, the neural
of a frustrating decades-long quest— are American Association for Cancer Research in equivalent of macrophages,
“active” therapies. A patient receives an April 2007 for vaccines for breast, prostate, express cytokines and cell-
injection of an antigen, usually along with and head and neck cancers. But more bad damaging free radicals, while
interacting with the beta-amyloid
another helper molecule, an adjuvant, that news came the next month: the Food and Drug
proteins that build up in the
precipitates an immune response. Administration delayed approval of what plaques that are characteristic
Cancer antigens are more difficult to would have been the first U.S. therapeutic of the disease. The resulting
identify than those for pathogens because cancer vaccine, formulated by Seattle-based inflammation can damage neurons.
cancer cells are mutant forms of the body’s Dendreon for prostate cancer.
own cells. The adaptive immune system Marshaling the patient’s own immune DEPRESSION AND
often does not see them as foreign — and a system to fight cancer may still be possible. SCHIZOPHRENIA
tumor can trick the body into turning off any But meeting that goal may well depend on High levels of inflammatory
immune response that does arise. deepening the growing understanding of how molecules — interleukin-6 and
C-reactive protein — have been
After hundreds of previous trials, and the the immune system serves as a two-edged
found in depressed patients.
absence of virtually any evidence from them sword that can either foster or block cancer Some evidence even suggests
that vaccines cause tumors to regress, the progression.  — G.S. that elevated cytokines correlate
with schizophrenia.

w w w. S c i A m . c o m SCIENTIFIC A MERIC A N 53
© 20 08 SCIENTIFIC AMERIC AN, INC.
The Big Eaters cells, a subclass of adaptive immune cells re-
sponsible for quieting other T cells.
Macrophages, also known as big eaters, are inflammatory cells that Meanwhile Coussens and her colleagues at
coordinate many critical steps in cancer development, from nurturing the U.C.S.F. found in a 2005 study, published in
tumor to helping its cells spread to distant sites. Cancer Cell, that the removal of antibody-mak-
ing B cells from mice engineered to be prone to
Help to bring a Release proteins skin cancer prevented the tissue changes and
Tumor cells
blood supply to that suppress the angiogenesis that are prerequisites for disease
oxygen-starved adaptive immune progression. In their normal role as pathogen
Macrophage parts of tumors response
fighters, B cell–produced antibodies circulate
through the bloodstream and mark viruses
and bacteria for destruction by innate immune
cells. In response to a signal from precancer-
ous tissue, however, the antibodies induce the
innate system to collaborate in cancer develop-
Membrane
Thwarted T cell ment. An open research question is how this
process starts. One possibility suggests that a
Tumor cell
cancer cell may send a message to innate im-
Break down mune cells, perhaps dendritic cells, that then
the membrane activate B cells. Signaling may involve toll-like
surrounding
the tumor Growth factor receptors, which have emerged as prominent
intermediaries in innate immune messaging.

Macrophage
Cancer Blockers
Release a growth factor that t h e r e c o g n i t i o n that can-
Blood vessel guides cancer cells to blood cer is more like an organ than
vessels, enabling their spread Tumor just a clump of cells with DNA
throughout the body necrosis Proteins
factor that stop mutations in cell nuclei may
cell death also explain why some of the
A MASTER SWITCH previous approaches to che-

Tumor cell
Macrophages produce inflammatory compounds, Receptor motherapy have met with
such as tumor necrosis factor, which can activate a limited success. “People have
gene switch, a complex of proteins called nuclear Nucleus
Activated NF-�B taken cells and then trans-
factor-kappa B (NF-B), in a tumor cell. NF-B enters the formed them in culture and stuck
cell nucleus and turns on production of proteins that stop them into animals,” Pollard says.
cell death and promote inflammation and cell proliferation. DNA
“They grow as little balls. They do cer-
tain things there. But they are not complex
nate immunity, the adaptive immune sys- tissues, whereas a naturally occurring tumor
tem— the T cells and antibodies produced by is a very complex tissue.”
B cells that target specific molecules on invad- Instead of just killing cancer cells— the goal
ing cells — contributes to pathology or may of current drug therapies and radiation— new
also fight against it. For decades, immuno- approaches may supplement existing drugs by
therapies designed to enhance T cell respons- slowing inflammation. Without the involve-
es against cancer have been explored, though ment of macrophages and other innate cells,
often with disappointing results [see box on the premalignant tissue would remain in check.
preceding page]. Cancer could, in essence, become a chronic dis­
Furthermore, an emerging picture has be- ease akin to rheumatoid arthritis, another in-
gun to reveal an intricate cross talk between flammatory condition. “Keep in mind almost
innate and adaptive immune cells that may no one dies of primary cancer,” says Raymond
participate in the promotion of malignant DuBois, provost of the University of Texas M.
disease. Researchers working on cancer vac- D. Anderson Cancer Center and a researcher
cines may need to take account of these inter- of anti-inflammatory agents for cancer. “A pa-
j en c h ristiansen

actions in designing their treatments if they tient almost always dies from a metastasis.”
are ever to prove effective. One study showed A pharmaceutical against chronic inflam-
that ovarian tumors produce a signaling mation represents a more alluring proposition
molecule that serves to attract regulatory T than massacring malignant cells (and, un-

54 S C I E N T I F I C A M E R I C A N ne w ans w ers f o r C ancer


© 20 08 SCIENTIFIC AMERIC AN, INC.
avoidably, healthy ones), a consequence of ex- livery scheme that takes advantage of the natu-
isting chemotherapies. Taken alone, such an
agent might be benign enough to use every day
ral attraction of macrophages to the oxygen-
starved areas in tumors. They have engineered
Instead of
as a preventive for high-risk patients. Epide- macrophages to deliver a therapeutic virus to killing cancer
miological and clinical studies have shown
some promise for the use of nonsteroidal anti-
hypoxic tumor regions, which respond poorly
to conventional treatments such as chemother-
cells— the goal of
inflammatory drugs (NSAIDs) such as aspirin apy and radiation because of an insufficient current drug
to stave off the onset of some solid tumors. In-
vestigations continue on more selective block-
blood supply. Once the macrophages arrive in
a tumor (grown in culture so far), each one
therapies and
ing of the production of prostaglandins, the ­releases thousands of copies of the virus, which radiation— new
regulatory molecules that are curtailed by
NSAIDs. In particular, drugs that inhibit pro-
then infect the cancer cells, after which a pro-
tein in those cells activates the therapeutic gene
chemotherapies
duction of prostaglandin E2 may curb inflam- in each virus. This action then directs synthe- may supplement
mation and tumor growth, while avoiding the
cardiovascular side effects of drugs such as
sis of a cell-killing toxin. “The macrophage is
migrating into a site and doing what we want
existing drugs
Vioxx and the gastrointestinal problems of the it to do rather than driving tumor development by turning down
earlier class of NSAIDs. The anti-inflammato-
ry effects of the ubiquitous statins used to low-
in a normal way,” Lewis says.
The exact outlines of an anti-inflammato-
inflammation.
er cholesterol are also being contemplated. ry strategy against cancer have yet to be elu-
Some treatment options already exist. cidated. Tweaking immune cells that form a
The drug Avastin inhibits production of the defensive barrier against pathogens bears its
angiogenesis-promoting VEGF, although own risks. “It’s a very complicated issue,”
oncologists must contend with other mole- DuBois notes. “If you magically shut
cules in the tumor microenvironment that down the immune system, you will have
promote blood vessel growth. Drugs devel- problems with opportunistic infections,
oped for more familiar inflammatory diseas- just like with AIDS.” Use of TNF blockers
es may also fight cancer— and these medi- in other inflammatory disorders has been
cines might be combined into HIV-like drug linked to tuberculosis and other infections,
cocktails that also include angiogenesis in- even potentially lymphoma. Moreover, inhib-
hibitors and cell-killing agents. iting the NF-kB pathway can paradoxically
Inhibitors of TNF have received approval promote cancer in some instances. Constrain-
for treatment of rheumatoid arthritis, Crohn’s ing NF-kB can at times lead to tissue damage
disease and other disorders and are now in and a process of abnormal regeneration of
clinical trials for both solid tumors and blood that tissue that can foster cancer.
cancers. The drug Rituxan, a monoclonal an- Still, it seems likely that a new generation
tibody that represses B cells in rheumatoid ar- of anti-inflammatory agents will join the che-
thritis and B cell lymphoma, might prevent the motherapeutic arsenal. Chronic diseases— and
inflammatory response that fuels formation their underlying inflammatory conditions —
of solid tumors. Other cytokines and related are hallmarks of an aging population. “We’re
molecules (IL-6, IL-8 and CCL2, among oth- all a little bit overinflamed,” Pollard observes.
ers) are also potential targets, as is NF-kB. Treating the smoldering embers that surround
Some existing compounds, including the tumor rather than just mutant cells could
NSAIDs and even one found in the spice tur- make cancer a disease we can live with. 
meric, exert at least some of their effects by
inhibiting NF-kB. But major pharmaceutical Gary Stix is senior writer for
laboratories are investigating highly selective Scientific American.
inhibitors of this molecular linchpin, many
of them targeted at the enzymes (such as I-kB more to explore
kinase) that regulate NF-kB activity. Smoldering and Polarized Inflammation in the Initiation and Promotion
of Malignant Disease. Frances Balkwill, Kellie A. Charles and Alberto
A Chemical Trojan Mantovani in Cancer Cell, Vol. 7, No. 3, pages 211–217; March 2005.
o n e g r o u p is contemplating a radically Distinct Role of Macrophages in Different Tumor Microenvironments.
Claire E. Lewis and Jeffrey W. Pollard in Cancer Research, Vol. 66, No. 2,
ambitious treatment, a molecular Trojan horse pages 605–612; January 15, 2006.
of sorts. Claire E. Lewis and Munitta Mutha- Paradoxical Roles of the Immune System during Cancer Development.
na of the University of Sheffield in England Karin de Visser, Alexandra Eichten and Lisa M. Coussens in Nature
and their colleagues have designed a drug de- Reviews Cancer, Vol. 6, No. 1, pages 24–37; January 2006.

w w w. S c i A m . c o m SCIENTIFIC A MERIC A N 55
© 20 08 SCIENTIFIC AMERIC AN, INC.
The
Long
Dendritic cells Arm
of the
catch invaders and tell
the immune system when
and how to respond. Vaccines
depend on them, and scientists are
even employing the cells to stir up
immunity against cancer
By Jacques Banchereau

56 S C I E N T I F I C A M E R I C A N
© 20 08 SCIENTIFIC AMERIC AN, INC.
ne w ans w ers f o r C ancer
Immune System

w w w. S c i A m . c o m
© 20 08 SCIENTIFIC AMERIC AN, INC.
SCIENTIFIC A MERIC A N 57
T hey lie buried— their long, tentaclelike arms out­stretched— in all the tissues of our bodies that
interact with the environment. In the lining of our nose and lungs, lest we inhale the influ­
enza virus in a crowded subway car. In our gastrointestinal tract, to alert our immune system
if we swallow a dose of salmonella bacteria. And most important, in our skin, where they lie in wait as stealthy
sentinels should microbes breach the leathery fortress of our epidermis.
They are dendritic cells, a class of
white blood cells that encompasses
some of the least understood but most
fascinating actors in the immune sys­
tem. Over the past decade, researchers
have begun to unravel the mysteries of
how dendritic cells educate the im­
mune system about what belongs in the
body and what is foreign and poten­
munization by “vaccinating” cancer
patients with dendritic cells loaded
with bits of their own tumors to acti­
vate their immune system against their
cancer. Dendritic cells are also respon­
sible for the phenomenon of immune
tolerance, the process through which
the immune system learns not to attack
other components of the body.
Rare and Precious
d e n d r i t i c c e l l s are relatively
scarce: they constitute only 0.2 percent
of white blood cells in the blood and are
present in even smaller proportions in
tissues such as the skin. In part because
of their rarity, their true function eluded
scientists for nearly a century after they
were first identified in 1868 by German
tially dangerous. Intriguingly, they But dendritic cells can have a dark side. ana­t­o­m­ist Paul Langerhans, who mis­
have found that dendritic cells initiate The human immunodeficiency virus took them for nerve endings in the skin.
and control the overall immune re­ (HIV) hitches a ride inside dendritic cells In 1973 Ralph M. Steinman of the
sponse. For instance, the cells are cru­ to travel to lymph nodes, where it infects Rockefeller University rediscovered the
cial for establishing immunological and wipes out helper T cells, causing cells in mouse spleens and recognized
“memory,” which is the basis of all vac­ AIDS. And those cells that become active that they are part of the immune system.
cines. Indeed, physicians, including at the wrong time might give rise to auto­ The cells were unusually potent in stim­
those at a number of biotechnology immune disorders such as lupus. In these ulating immunity in experimental ani­

Illustration by jeff johnson H y b r i d M e d i c a l A n i m a t i o n (preceding pages)


companies, are taking advantage of cases, shutting down the activity of den­ mals. He renamed the cells “dendritic”
the role that dendritic cells play in im­ dritic cells could lead to new therapies. because of their spiky arms, or den­
drites, although the subset of dendritic
Overview/Dendritic Cells cells that occur in the epidermis layer of
the skin are still commonly called Lang­
■ Dendritic cells—named for their long arms, or dendrites—exist in many erhans cells. This groundbreaking re­
tissues, particularly the skin and mucous membranes. They reel in search laid the foundation for all of the
invaders, chop them into pieces called antigens and display the antigens progress that we are seeing in dendritic
on their surfaces. cell science today. In 2007 Steinman re­
■ A ntigen-bearing dendritic cells travel to lymph nodes or the spleen, where ceived the Albert Lasker Basic Medical
they interact with other cells of the immune system—including B cells, which Research Award for his pioneering work
make antibodies, and killer T cells, which attack microbes and infected cells. in rediscovering and characterizing
■ C ancer vaccines composed of dendritic cells bearing tumor antigens are dendritic cells.
now in clinical trials involving humans. Scientists are also hoping to For almost 20 years after the cells’
turn off the activity of dendritic cells to combat autoimmune diseases rediscovery, researchers had to go
such as lupus. through a painstakingly slow process to
isolate them from fresh tissue for study.

58 S C I E N T I F I C A M E R I C A N ne w ans w ers f o r C ancer


© 20 08 SCIENTIFIC AMERIC AN, INC.
But in 1992, when I was at the Schering- by the rest of the immune system [see
Plough Laboratory for Immunology Re­ box on next two pages]. The cells use
search in Dardilly, France, my co-work­ pitchfork-shaped molecules termed the
ers and I devised methods for growing major histocompatibility complex (MHC)
large amounts of human dendritic cells to display the antigens on their surfaces.
from bone marrow stem cells in culture The antigens fit between the tines of the
dishes in the laboratory. At roughly the MHC, which comes in two types, class
same time, Steinman— in collaboration I and class II. The two types vary in
with Kayo Inaba of Kyoto University in shape and in the ways they acquire their
Japan and her colleagues — reported that antigen cargo while residing inside cells.
he had invented a technique for cultur­ Dendritic cells are very efficient at
ing dendritic cells from mice. capturing and presenting antigens: they
In 1994 researchers led by Antonio can pick up antigens that occur in only
Lanzavecchia, now at the Institute for minute concentrations. As they process
Research in Biomedicine in Bell­in­zona, antigens for presentation, they travel to
Switzerland, and Gerold Schuler, now the spleen through the blood or to lymph
at the University of Erlangen-Nurem­ nodes through a clear fluid known as
berg in Germany, found a way to grow lymph. Once at their destinations, the
the cells from white blood cells called cells complete their maturation and
monocytes. Scientists now know that present their antigen-laden MHC mol­
monocytes can be prompted to become ecules to naive helper T cells, those that
either dendritic cells, which turn the im­ have never encountered antigens before.
mune system on and off, or macrophag­ Dendritic cells are the only cells that can
es, cells that crawl through the body educate naive helper T cells to recognize
scavenging dead cells and microbes. an antigen as foreign or dangerous. This
jacques banchereau (top and top middle); olivia gresser Marseilles-Luminy Center of Immunology INSERM/CNRS (bottom middle);

The ability to culture dendritic cells unique ability appears to derive from co-
offered scientists the opportunity to in­ stimulatory molecules on their surfaces
vestigate them in depth for the first that can bind to corresponding recep­
time. Some of the initial discoveries ex­ tors on the T cells.
panded the tenuous understanding of Once educated, the helper T cells go
how dendritic cells function. on to prompt so-called B cells to pro­
There are several subsets of dendrit­ duce antibodies that bind to and inacti­
ic cells, which arise from precursors vate the antigen. The dendritic cells and
that circulate in the blood and then take helper cells also activate killer T cells,
up residence in immature form in the which can destroy cells infected by mi­
william e. bowers, ste ve hale y and richard hunt University of South Carolina (bottom)

skin, mucous membranes, and organs crobes. Some of the cells that have been
such as the lungs and spleen. Immature educated by dendritic cells become
dendritic cells are endowed with a “memory” cells that remain in the body
wealth of mechanisms for capturing in­ for years — perhaps decades — to combat
vading microbes: they reel in invaders the invader in case it ever returns.
using suction cup–like receptors on Whether the body responds with
their surfaces, they take microscopic antibodies or killer cells seems to be de­
sips of the fluid surrounding them, and termined in part by which subset of
they suck in viruses or bacteria by en­ dendritic cell conveys the message and
gulfing them in sacks known as vacu­ which of two types of immune-stimu­
oles. Yong-Jun Liu, a former colleague lating substances, called cytokines, the
of mine from Schering-Plough who is dendritic cells prompt the helper T cells
SPIKY ARMS are common to mature dendritic
now at the University of Texas M. D. to make. In the case of parasites or cells from humans (top and top middle), mice
Anderson Cancer Center, has found some bacterial invaders, type 2 cyto­ (bottom middle) and rats (bottom). The rat
that some immature dendritic cells can kines are best because they arm the im­ dendritic cell is interacting with what is probably
also zap viruses immediately by secret­ mune system with antibodies; type 1 a helper T cell. Through such interactions,
dendritic cells teach the immune system what it
ing a substance called interferon-alpha. cytokines are better at mustering killer
should attack. Cells matured in the laboratory,
Once they devour foreign objects, cells to attack cells infected by other such as the one at the top middle, are being used
the immature cells chop them into frag­ kinds of bacteria or by viruses. in cancer vaccines.
ments (antigens) that can be recognized If a dendritic cell prompts the wrong

w w w. S c i A m . c o m SCIENTIFIC A MERIC A N 59
© 20 08 SCIENTIFIC AMERIC AN, INC.
DENDRITIC CELLS AND INFECTION
Present in the lungs, skin, gut and lymph nodes, dendritic cells
orchestrate the immune response against invaders (here, Bacterium
bacteria entering a cut in the skin).

Bacteria enter
cut in the skin.

Skin
Lung Lymph node

Gut Dendritic cell

Epidermis

Dermis

Dendritic cells bind to helper T cells, killer T cells and — perhaps —


B cells. The binding prompts the helper T cells to make substances
called cytokines that stimulate killer T cells and cause B cells to
begin making antibodies. The antibodies and killer T cells migrate
to the cut to fight the infection. Memory cells persist in case the
body becomes infected again.
Type 2 cytokine

Type 1 cytokine

Helper
T cell

Killer
T cell T cell
receptors Adhesion protein
Antigens
Memory
T cell
Co-stimulatory
molecule
MHC class II

MHC class I
CREDIT

60 SCIENTIFIC A MERIC A N ne w ans w ers f o r C ancer


© 20 08 SCIENTIFIC AMERIC AN, INC.
Dendritic cells ingest bacteria and chop them up
into bits called antigens. As they exit infected
tissues, they mature and display the antigens using
molecules called MHC class I and class II.

Immature dendritic cell

Antibody Antigen

MHC class I
Antigen
After traveling to the
lymph nodes in a fluid MHC class II
called lymph, dendritic
cells activate other cells
of the immune system
that are capable of
recognizing the antigens
they carry. The activation
readies the immune cells
to fight invaders bearing
the antigens.

Memory
B cell

Lymph node
B cell

Unknown signal
terese w ins l o w

Mature dendritic cell


CREDIT

w w w. S c i A m . c o m SCIENTIFIC A MERIC A N 61
© 20 08 SCIENTIFIC AMERIC AN, INC.
type of cytokine, the body can mount tients and grown in the laboratory to­ ever, in more than 1,000 vaccinated pa­
the wrong offense. Generating the ap­ gether with tumor antigens. During tients.) Conversely, the tumors might
propriate kind of immune response can this process, the dendritic cells pick up mutate to “escape” the immune on­
be a matter of life or death: when ex­ the antigens, chop them up and present slaught engendered by a dendritic cell
posed to the bacterium that causes lepro­ them on their surfaces. When injected vaccine. Tumor cells could accomplish
sy, people who mount a type 1 response back into the patients, the antigen- this evasion by no longer making the an­
develop a mild, tuberculoid form of the loaded dendritic cells are expected to tigens the vaccine was designed to stim­
disease, whereas those who have a type ramp up patients’ immune response ulate the immune system against. This
2 ­response can end up with the poten­ against their own tumors. problem is not unique to dendritic cells,
tially fatal lepromatous form. Various researchers—including our though: the same phenomenon can oc­
own group as well as scientists at several cur with traditional cancer therapies.
Cancer Killers biotechnology companies — are testing In addition, tailoring a dendritic
ac t i va t i n g n a i v e helper T cells is this approach against cancers as diverse cell vaccine to fight a particular pa­
the basis of vaccines for everything as melanoma, B cell lymphoma, and tu­ tient’s tumors might not be economi­
from pneumonia to tetanus to influ­ mors of the prostate and colon. There cally feasible. But many scientists are
enza. Scientists are now turning the have been glimmers of success. In Sep­ working to circumvent the costly and
new knowledge of the role that dendrit­ tember 2001, for instance, my co-work­ time-consuming steps of isolating cells
ic cells play in immunity against mi­ ers and I, in collaboration with Stein­ from patients and manipulating them
crobes and their toxins into a strategy man’s group, reported that 16 of 18 pa­ in the laboratory for reinjection.
to fight cancer. tients with advanced melanoma to whom One approach involves prompting
Cancer cells are abnormal and as we gave injections of dendritic cells load­ dendritic cell precursors already present
such are thought to generate molecules ed with melanoma antigens showed in a person’s body to divide and start
that healthy cells do not. If researchers signs in laboratory tests of an enhanced orchestrating an immune response
could devise drugs or vaccines that ex­ immune response to their cancer. What against their tumors. While at Immunex
clusively targeted those aberrant mol­ is more, tumor growth was slowed in the in Seattle, David H. Lynch, now at Bain­
ecules, they could combat cancer more nine patients who mounted responses bridge Biopharma Consulting in Bain­
effectively while leaving normal cells against more than two of the antigens. bridge Island, Wash., and his co-work­
and tissues alone — thereby eliminating Scientists are now working to refine ers discovered a cytokine that causes
some of the pernicious side effects of the approach and test it on larger num­ mice to make more dendritic cells, which
chemo­therapy and radiation, such as bers of patients. So far cancer vaccines eventually induce the animals to reject
hair loss, nausea and weakening of the based on dendritic cells have been tested grafted tumors. Other scientists, includ­
immune system caused by destruction only in patients with advanced cancer. ing Drew M. Pardoll of Johns Hopkins
of the bone marrow. Although researchers believe that patients University, have observed that tumor
Antigens that occur only on cancer­ with earlier-stage cancers may ­respond cells that have been genetically engi­
ous cells have been hard to find, but re­ better to the therapy— their immune sys­ neered to secrete large amounts of cy­
searchers have succeeded in isolating tems have not yet tried and failed to erad­ tokines that activate dendritic cells have
several of them, most notably from the icate their tumor— several potential prob­ the most potential as cancer vaccines.
skin cancer melanoma. In the early lems must first be considered. Another approach, pioneered by
1990s Thierry Boon of the Ludwig Can­ Some researchers fear that such vac­ Steinman and his Rockefeller colleague
cer Institute in Brussels, Steven A. cines might induce patients’ immune Michel C. Nussenzweig, is to selectively
Rosenberg of the National Cancer Insti­ systems to attack healthy tissue by mis­ target antigens by coupling them to
tute and their colleagues independently take. For instance, vitiligo —white patch­ monoclonal antibodies that bind to re­
identified mel­a­noma-specific antigens es on the skin caused by the destruction ceptors on the surface of dendritic cells.
that are currently being targeted in a va­ of normal pigment-producing melano­ These receptors need to allow internal­
riety of clinical trials involving humans. cytes — has been observed in melanoma ization of the antigens and their pro­
Such trials generally employ vac­ patients who have received the earliest cessing for presentation on both MHC
cines made of dendritic cell precursors antimelanoma vaccines. (No major ad­ class I and class II antigens. Several such
that have been isolated from cancer pa­ verse events have been reported, how­ molecules are now under intense scru­
tiny. Studies in mice have shown that
JACQUES BANCHEREAU has directed the Baylor Institute for Immunology Research in targeting the antigens in the absence of
the author

Dallas since 1996. The institute aims to manipulate the human immune system to treat dendritic cell activation results in toler­
cancer as well as infectious and autoimmune diseases. Before 1996 Banchereau led ance induction. In contrast, delivering
the Schering-Plough Laboratory for Immunology Research in Dardilly, France. He ob- the antigen together with dendritic cell
tained his Ph.D. in biochemistry from the University of Paris and holds many patents on activators induces immunity, which can
immunological techniques. be protective.

62 SCIENTIFIC A MERIC A N ne w ans w ers f o r C ancer


© 20 08 SCIENTIFIC AMERIC AN, INC.
Shutting Immunity Down ents from rejecting their new tissues.
i n t h e m e a n t i m e , other scientists A new treatment for AIDS might
are looking at ways to turn off the ac­ also rest on a better understanding of
tivity of dendritic cells in instances dendritic cells. In 2000 Carl G. Figdor
where they exacerbate disease instead and Yvette van Kooyk, then at the Uni­
of fighting it. Usually, in a phenomenon versity Medical Center St. Radboud in
known as central tolerance, an organ in Nijmegen, the Netherlands, identified a
the chest called the thymus gets rid of subset of dendritic cells that makes
young T cells that happen to recognize DC-SIGN, a molecule that can bind to
the body’s own components as foreign the outer coat of HIV. These cells pick
before they have a chance to circulate. up HIV as they regularly prowl the mu­
Some inevitably slip through, however, cous membranes and deep tissues.
so the body has a backup mechanism When they travel to the lymph nodes,
for restraining their activity. they unwittingly deliver the virus to a
But this mechanism, termed periph­ large concentration of T cells. Drugs
eral tolerance, appears to be broken in that block the interaction between DC-
patients with autoimmune disorders SIGN and HIV might slow the progres­
such as rheumatoid arthritis, type 1 sion of AIDS.
diabetes and systemic lupus erythema­ Other infectious diseases — includ­
tosus. In 2001 my colleagues and I re­ ing malaria, measles and cytomegalovi­
ported that dendritic cells from the rus — also manipulate dendritic cells for
blood of people with lupus are unnatu­ their own ends. Red blood cells that
rally active. Cells from these patients have been infected by malaria parasites,
release high amounts of interferon-­ for instance, bind to dendritic cells and
alpha, an immune-stimulating protein prevent them from maturing and alert­
that causes precursors to grow into ing the immune system to the presence
mature dendritic cells while still in the of the invaders. Several groups of re­
bloodstream. The mature cells then in­ searchers are now devising approaches
gest DNA, which is present in unusual to prevent such microbes from hijacking
amounts in the blood of people with dendritic cells; some are even seeking to
jacques banchereau (top, top middle and bottom middle ); matthe w l. albert Rockefeller University (bottom)

lupus, and that in turn causes the indi­ use supercharged dendritic cells to fight
vidual’s immune system to generate an­ the infections.
tibodies against his or her own DNA. As we learn more about the mole­
These antibodies result in the life- cules that control dendritic cells, we
threatening complications of lupus will find ways to harness their thera­
when they lodge in the kidneys or the peutic potential. The increasing num­
walls of blood vessels. ber of scientists and pharmaceutical
Accordingly, we propose that block­ corporations working on dendritic cells
ing interferon-alpha might lead to a portends that we will soon be able to
therapy for lupus by preventing den­ maximize the biological power of these
dritic cell activation. A similar strategy cells to treat and prevent the diseases
might prevent organ transplant recipi­ that plague humankind.

more to explore
Dendritic Cells and the Control of Immunity. Jacques Banchereau and Ralph M. Steinman in
Nature, Vol. 392, pages 245–252; March 19, 1998.
Dendritic Cells as Vectors for Therapy. Jacques Banchereau, Beatrice Schuler-Thurner,
A. Karolina Palucka and Gerold Schuler in Cell, Vol. 106, No. 3, pages 271–274; August 10, 2001.
IMMATURE dendritic cells can be stained to
Taking Dendritic Cells into Medicine. Ralph M. Steinman and Jacques Banchereau in Nature, show up green in breast cancer tissue (top) or
Vol. 449, pages 419–426; September 27, 2007. red in normal skin (top middle). As the cells
Dendritic Cell Subsets in Health and Diseases. H. Ueno, E. Klechevsky, R. Morita, C. Aspord, T. Cao, mature, they make proteins that allow them to
T. Matsui, T. Di Pucchio, J. Connolly, J. W. Fay, V. Pascual, A. K. Palucka and J. Banchereau in stick to one another (bottom middle). They also
Immunological Reviews, Vol. 219, No. 1, pages 118–142; October 2007. produce forklike receptors (green dots,
Background information on the immune system and on experimental cancer therapies such bottom), which they use to show bits of
as those using dendritic cells can be found on the American Cancer Society’s Web site: invaders to other immune cells.
www.cancer.org

w w w. S c i A m . c o m
© 20 08 SCIENTIFIC AMERIC AN, INC.
SCIENTIFIC A MERIC A N 63
Taming Vessels
to Treat Cancer
W
hile still a graduate student in 1974, I
Restoring order to the chaotic blood vessels had a chance to see malignant tumors
from a most unusual perspective. I was
inside a tumor opens a window of opportunity working at the National Cancer Institute in the
for attacking it. Surprisingly, drugs meant laboratory of the late Pietro M. Gullino, who had
developed an innovative experimental setup for
to destroy vasculature can make the repairs studying cancer biology— a tumor mass that was
connected to the circulatory system of a rat by just
and may help reverse conditions that lead a single artery and a single vein. As a chemical engi-
neer, I decided to use this opportunity to measure
to cardio­vas­c­u­lar disease and blindness  how much of a drug injected into the animal would
flow to the tumor and back out again. Amazingly,
By Rakesh K. Jain most of the substance injected into the rat never
entered the tumor. To make matters worse, the
small amount that did reach the mass was distrib-
uted unevenly, with some areas accumulating hard-
ly any drug at all.
My immediate concern was that even if a small
fraction of the cancer cells in a human tumor did not

64 S C I E N T I F I C A M E R I C A N ne w ans w ers f o r C ancer


© 20 08 SCIENTIFIC AMERIC AN, INC.
receive an adequate dose of whatever anticancer storing blood vessels is also opening doors to treat- ABNORMAL blood
drug was being applied, those cells could survive — ing other vascular disorders, such as macular degen- vessels (left) add to
causing the tumor to grow back sooner or later. Per- eration, the leading cause of blindness in the U.S. the havoc inside a
tumor and prevent
haps the engineer in me was also drawn to trying to treatments from
understand and solve the apparent infrastructure Tortuous Road reaching cancer
problem inside tumors that posed a major obstacle t h e jou r n e y t h at l e d to our recent success- cells. Normalizing
to the delivery of cancer therapies. es began in earnest a few years after I completed those vessels
Over the subsequent decades my colleagues and my doctoral studies. Determined to find out why (above) makes
them into functional
I have investigated what makes the vasculature with- drugs do not penetrate tumors uniformly, my col- weapons that can
in tumors abnormal and how these disordered blood leagues and I started by monitoring every step of be turned against
vessels not only stymie traditional cancer treatments the process in rodents. Using a variety of tech- the tumor.
but also contribute directly to some of the malignant niques, we observed the progress of drugs as they
properties of solid cancers. Building on these in- entered the tiny blood vessels of a tumor, crossed
sights, we developed approaches to normalizing tu- the vessel walls into the surrounding tissue, en-
mor blood vessels and tested them successfully in tered into cancer cells and eventually exited the
mice. In the process, we also discovered a seeming mass. Together with my students and collabora-
paradox— a class of drugs designed to destroy the tors, we developed methods for tracking mole-
blood vessels of tumors actually acts to repair them,
creating a window of opportunity to attack the can-
cer most effectively.
Overview/Controlling Chaos
In recent years we have finally been able to start n Abnormal and dysfunctional blood vessels are a hallmark of solid tumors,
testing this idea in cancer patients, and the excite- one that contributes directly to malignant properties of a cancer as well
ment in our lab was overwhelming when we saw the as preventing treatments from reaching and attacking tumor cells.
first clinical evidence of tumors shrinking in re- n Normalizing tumor vessels allows cancer therapies to penetrate the mass

sponse to vascular normalization, just as we had and to function more effectively.


anticipated. Much more work remains before we n Unexpectedly, drugs originally designed to destroy tumor blood vessels
k eit h k asn o t

can perfect this therapeutic approach and gauge its act to repair them for a time, opening a new avenue for cancer treatment
usefulness in patients with different types of malig- as well as restoration of abnormal vasculature in other diseases.
nancy. But what we have already learned about re-

w w w. S c i A m . c o m SCIENTIFIC A MERIC A N 65
© 20 08 SCIENTIFIC AMERIC AN, INC.
cules, such as oxygen, within blood vessels and culature and gained insight into the consequences
tissues. Eventually we could even watch as genes of the abnormalities. The picture that emerged was
turned on and off inside cells. grim: the very first thing we realized was that tu-
Early on it was apparent that the vessels within mor blood vessels are not just disorganized in their
tumors bear little resemblance to normal ones. appearance but highly aberrant in every aspect of
Healthy tissues are fed by straight vessels that their structure and function. We found that blood
branch predictably into successively smaller capil- flows quite briskly in some vessels within a tumor,
laries and microvessels, creating a pervasive net- whereas it is static in others. In a given vessel, blood
work for delivering oxygen and nutrients to cells. may travel in one direction for a while and then
Tumors, which stimulate the growth of new vascu- reverse direction. These flow patterns alone create
lature of their own, tend to generate a tangle of ves- a major obstacle to uniform drug delivery. More-
sels. These connect to one another randomly, with over, some parts of the vessel walls are overly leaky
some oversize branches, many extraneous imma- and others are unusually tight, which means that
ture microvessels and areas of a tumor that will drugs and other molecules that managed to pene-
lack vessels altogether. trate the vasculature would be distributed into the
Over the course of many years we managed to surrounding tumor tissue unevenly.
delineate the processes that govern the movement When we began investigating the causes of this
of fluids, drugs and cells within this tortuous vas- nonuniform porousness, we discovered that in

Abnormal Vessels Make Trouble


Malformed vasculature inside a
tumor turns a bad situation worse High
VESSEL FUNCTION interstitial
(boxes). Flaws in the organization n O versize pores in vessel walls leak fluid fluid
and functioning of blood vessels into interstitial areas (between cells, pressure
create barriers that prevent vessels and other structures)
n High
 interstitial fluid pressure blocks
therapies from reaching tumor cells transport of drugs out of vessels
and foster an environment where to tumor tissue
Impaired Abnormal
those treatments are less effective. transport blood vessel
These un­natural internal conditions of drugs
also contribute to malignant
properties of the cancer itself. Immature
microvessel

VESSEL ORGANIZATION
n Oversize diameter and chaotic
layout create irregular blood flow
n Absent or immature vessels

make some tumor regions


impenetrable
Leaky oversize pore

Tumor Endothelial cell

FLUID BUILDUP
n Tumor tissue

TUMOR MICROENVIRONMENT swells, causing


n Dysfunctional vessels produce conditions of Swelling painful symptoms
low oxygen (hypoxia) and high acidity n F luid pressure
Fluid drives tumor-
Radiation and certain chemo­ther­a­pies that
n
and cells
require oxygen to kill tumor cells are ineffective escape generated proteins
n Immune cells that might attack cancer cells and cells toward
cannot function in acidic, hypoxic environments Lymphatic vessel healthy tissues
Hypoxia causes changes in gene activity and into lymphatic
k eit h k asn o t

that promote tumor cell migration toward vessels, increasing


healthy tissues risk of metastasis
Healthy tissue

66 SCIENTIFIC A MERIC A N ne w ans w ers f o r C ancer


© 20 08 SCIENTIFIC AMERIC AN, INC.
some tumors the pores in blood vessel walls
could be as large as one or two microns in di- BLOOD VESSELS in a normal
mouse muscle’s capillary
ameter, which is more than 100 times the size
bed (left) and inside a mouse
of pores in healthy vessels. As a result, these tumor (right) differ distinctly.
vessels are unable to maintain normal pres- The tumor vessels branch
sure gradients across their walls. Fluid pres- erratically, vary in diameter
sure inside healthy blood vessels is typically along their lengths and are
generally oversize — all
much higher than in the surrounding tissue.
features that contribute to
Because tumor vessels are so porous, escaping irregular blood flow.
fluid raises the outside — or interstitial— pres-
sure until it nearly equals that inside the blood
vessels. mately improve cancer treatment. To accomplish
This unnatural pressure gradient is not just an such a reversal, we first had to gain a better under-
impediment to the ability of drugs to reach tumor standing of what makes tumor vessels abnormal
cells; the accumulation of interstitial fluid produces and keeps them that way.
swelling in and around tumor tissues. In patients
with brain cancers, where tissue expansion is lim- Restoring Balance
ited by the skull, that swelling becomes a severe, w e b e g a n t o l o o k at the molecular factors
often life-threatening problem in itself. In those involved in normal blood vessel formation, known
with other types of cancer, the exuded fluid can as angiogenesis, including the single most potent
also accumulate in body cavities. Wherever it goes, one, vascular endothelial growth factor (VEGF).
the fluid oozing from a tumor carries with it tumor First discovered and named vascular permeability
cells as well as various tumor-generated proteins factor by my Harvard University colleague Harold
that promote the growth of new blood and lym- Dvorak, VEGF promotes the survival and prolif-
phatic vessels in the surrounding normal tissue and eration of endothelial cells, which form the inner
lymph nodes— which can then serve as conduits for lining of blood vessels. In excess, it also makes ves-
the metastatic spread of the cancer cells to other sels leaky—hence its original name. In normal tis-
parts of the body. sues, however, the collective action of VEGF and
Beyond the difficulty of delivering drugs other growth-stimulating molecules like it is coun-
through chaotic tumor vasculature and the danger- terbalanced by the actions of natural antiangiogen-
ous fluid buildup caused by leaky vascular walls, esis molecules, such as thrombospondin, that in-
the abnormalities of tumor vessels create a highly hibit blood vessel growth.
unnatural microenvironment inside a tumor as Whether healthy or diseased, tissues that need
I N NATURE MEDICINE, V O L . 9 ; 2 0 0 3 , re p rinted w it h p ermissi o n o f M acmillan Publis h ers LT D .

well. Because many areas of a tumor lack vascula- new blood vessels increase their production of an-
ture and existing vessels are unable to deliver suf- giogenesis stimulators or reduce their production of
ficient oxygen to surrounding tissues, a general inhibitors, or do both, tipping the balance in favor
state of hypoxia (low oxygen) and high acidity pre- of angiogenesis. In healthy processes such as wound
F R O M “ M o lecular R e g ulati o n o f V essel M aturati o n , ” B Y R a k es h K . J ain ,

vails in the tumor. Hypoxia in turn makes tumor healing, a balance between growth and inhibitory
cells more aggressive and prone to metastasis. In factors is eventually reinstated once the new vessels
addition, the body’s immune cells, which might are established. But in tumors and a number of oth-
help fight a tumor, are hampered by acidity and er chronic diseases, an imbalance persists — and
cannot function in low oxygen. Nor can radiation blood vessels grow increasingly ­abnormal.
treatments and a subset of chemotherapy drugs
that depend on chemical processes that require ox- Rakesh K. Jain is Andrew Werk Cook Professor of Tumor Biology and director
the author

ygen to kill cancer cells. of the Edwin L. Steele Laboratory for Tumor Biology in the radiation oncology
Thus, what began as an inquiry into seemingly department of Massachusetts General Hospital and Harvard Medical School.
simple aberrations in the flow of drugs inside tu- His research incorporates biology, imaging, engineering and mathematics in
mors revealed that the abnormalities in tumor the study of blood and lymphatic vessels and their tissue environment, as well
blood vessels are obstacles to treatment in even as the adaptation of basic findings to patient treatment. He would especially
more ways than I had initially imagined. By 1994 like to acknowledge the National Cancer Institute for continuous support of
our observations were beginning to suggest to my his work since 1980 and more than 200 graduate students, postdoctoral fel-
research collaborators and me that if we knew how lows and collaborators worldwide who have shared his journey into the world
to repair the structure and function of tumor-asso- of solid tumors. Jain also serves as an adviser to several pharmaceutical and
ciated blood vessels, we would have a chance to biotechnology companies and is a member of both the National Academy of
normalize the tumor microenvironment and ulti- Engineering and the Institute of Medicine.

w w w. S c i A m . c o m SCIENTIFIC A MERIC A N 67
© 20 08 SCIENTIFIC AMERIC AN, INC.
Tipping the Balance toward NormalcY
Restoring tumor vasculature to a more normal
state optimizes conditions for anticancer Normal vasculature Abnormal TUMOR VESSELS
therapies to reach tumor cells and to work more Balance between progrowth and Excessive amounts of proangiogen­
antigrowth factor signaling esis factors, primarily VEGF, cause
effectively. Many abnormalities arise in tumor produces an organized network, overgrowth, yielding haphazardly
vessels because factors that stimulate vessel with large vessels regularly organized, oversize vessels and
growth, or angiogenesis, overwhelm growth branching into ever smaller ones. many dysfunctional microvessels.
inhibitors that usually check vessel proliferation.
Antiangiogenesis drugs that suppress the primary
progrowth factor, VEGF, tip the balance back
toward normal vessel formation and maintenance.
Normalization does not last indefinitely, because
the drugs may ultimately destroy most vessels or
a tumor may become resistant to them.
q healthy vessel Growth and Maintenance
Endothelial cells form blood vessels in response to
signals from molecules that promote and inhibit growth.
Pericyte cells straddle vessels, and a basement
membrane surrounds them; both provide support.

Endothelial cell Anti Pro Anti Pro


Pericyte
Basement
membrane Because VEGF is ceptors on the cell surface. Other antibodies bound
abundant in most to the VEGF receptors themselves, preventing the
solid tumors, I growth factor from making contact. Remarkably,
suspec ted t hat both forms of VEGF inhibition caused some of the
VEGF finding a way to immature and inefficient blood vessels characteris-
Receptor
mop up the excess tic of many tumors to be pruned away and induced
VEGF or inter­fere the remaining vessels to remodel themselves so that

B Y L A N C E l . M U N N I N “ N O R M A L I Z I N G T U M O R V A S C U L A T U R E W I T H A N T I - A N G I OG E N I C T H E R A P Y: A N E W P A R A D I G M
with the growth they began to resemble normal vasculature. Those

K E I T H K A S N O T (i l l u s t r a t i o n s) ; jen c h ristiansen (s e e s a w s) ; A D A P T E D F R O M O R I G I N A L I L L U S T R A T I O N S

F O R C O M B I N A T I O N T H E R A P Y, ” B Y R A K E S H K . J A I N , I N N AT U R E M E D I C I N E , V O L . 7, N O . 9 ; S E P T E M B E R 2 0 0 1
Antigrowth signals it generates normalized blood vessels were less leaky, less di-
factors could restore bal- lated and less tortuous. We could also detect func-
ance and cause tumor tional improvements in the tumors, including low-
Progrowth vas­cula­ture to revert to er interstitial fluid pressure, higher oxygenation
factors
a more normal state. Al- and improved penetration of drugs.
ternatively, increasing the As excited as we were by these results and by the
concentration of angiogenesis- fact that they were later reproduced in animals by
inhibiting factors could have the same nor- other researchers, we still could not know whether
malizing effect on the blood vessels. I also theo- the same responses would occur in cancer patients.
rized that vessels treated in either way would not And many researchers were understandably skepti-
remain normal forever— they would be destroyed if cal of our approach. By the late 1990s, when I first
the inhibitors were potent enough or would become proposed the idea of tumor vessel normalization
abnormal again if the tumors developed the ability publicly, scientists in academia and industry had
to make different stimulators, such as basic fibro- been working on making drugs to destroy blood
blast growth factor (bFGF), which can mimic many vessels. Their pursuit was based on the hypothesis
of the effects of VEGF. The only way to find out was put forward in 1971 by the late Judah Folkman, my
to try angiogenesis inhibitors on tumors and see former Harvard colleague, that tumor growth
what happened. could be halted by starving the mass using antian-
In 1995 antibody-based drugs that could neu- giogenic drugs. Indeed, the drug Avastin, approved
tralize the effects of VEGF were already in develop- by the Food and Drug Administration for use in
ment, so we were able to use these to test our ap- cancer treatment in 2004, is a VEGF-neutralizing
proach in mice. Certain of the antibodies attached antibody originally developed as such an antian-
directly to VEGF, hindering its ability to send a giogenesis agent.
growth signal to endothelial cells by binding to re- In laboratory testing and clinical trials, Avastin

68 SCIENTIFIC A MERIC A N ne w ans w ers f o r C ancer


© 20 08 SCIENTIFIC AMERIC AN, INC.
duced as well. And a form of programmed cell
death known as apoptosis, characteristic of oxy-
Normalized vessels Normalization
gen and nutrient deprivation, increased among
Inhibiting progrowth signals from Window Closed tumor cells that no longer had access to the pruned
VEGF prunes immature vessels and A potent enough vasculature.
restores the remaining ones to antiangiogenesis
a more normal arrangement, size Anti Pro Surprisingly, however, there was no concur-
drug could prune
and functioning. so much vascula­ rent decrease in a sign of overall energy use by
ture that addi­ the tumor, its uptake of a glucose analogue, as
tion­al therapies might be expected if the tumor were only being
would no longer
pene­trate the starved. Instead it seemed that the remaining
tumor (top). tumor vessels had become more efficient in sup-
Or, other growth porting the energy needs of the surviving cancer
factors might
surge to com­ cells. Furthermore, the rate of proliferation of
pensate for VEGF cancer cells increased in some tumors, reflect-
inhibition, mak­ ing their access to better-functioning vessels
ing the tumor
resistant to and a more normal tissue microenvironment.
antiangiogenesis Although increased proliferation is usually not
therapy and desirable when it comes to cancer cells, that
causing a return
to abnormal state would make them more sensitive to che-
vasculature. motherapy drugs, which generally target divid-
Anti Pro Anti Pro ing cells.
Together these results provided a first
glimpse of how a drug such as Avastin works in
has been shown to destroy blood vessels in animal patients and thereby revealed why it might improve
p o lymer cast F R O M “ Ge o metric resistance and micr o vascular net w o r k arc h itecture o f h uman c o l o rectal carcin o ma , ”

and human tumors, although when used alone it the outcome of radiation or chemotherapy for a
does not increase overall survival in cancer pa- time. As the drug blocks the effects of VEGF, some
tients. In a pivotal clinical trial that led to its ap- tumor vasculature is pruned away immediately, but
proval, however, Avastin did increase the survival the vessels that remain become less abnormal. In
of patients with advanced colorectal cancer but addition to improving the overall tumor microen-
B Y J . R . L ess , M . C . P o sner , T. C . S k ala k , N . W o lmar k A N D R . K . J ain , I N M I C R O C I R C U L AT I O N , V O L . 4 , N O . 1 ; 1 9 9 7

only when it was used in conjunction with stan- vironment, those normalized vessels also make the
dard chemotherapy. That positive outcome seemed surviving cells more vulnerable to the treatments
quite paradoxical at the time because, in principle, that they can now deliver more efficiently. Restor-
a drug that was designed and deployed to destroy
blood vessels should reduce the effectiveness of
TUMOR VESSELS cast in
chemotherapy, which requires functioning blood
polymer from a surgically
vessels to reach tumor cells. Some published stud- excised colon cancer form
ies have in fact shown that antiangiogenic agents a dense crystalline mass
can hinder radiation and chemotherapy. So how representing a chaotic
could these apparently contradictory findings be profusion of microvessels.
In contrast, large gaps
reconciled?
indicate areas of the tumor
Our group had the chance to find out by closely that lacked any blood supply.
examining the structure and function of blood ves-
sels in the tumors of rectal cancer patients receiving
Avastin and combined chemotherapy and radiation
in a 2002 clinical trial supported by the National
Cancer Institute and led by Christopher Willett,
now at Duke University Medical Center. Very
quickly, we saw that the changes to tumor vascula-
ture in those patients were not limited to simple
vessel destruction.
Two weeks after a single injection of Avastin,
blood flow within the tumors did drop by 30 to 50
percent in six consecutive patients. The density of mi-
crovessels, the overall number of blood vessels and the
interstitial fluid pressure in the tumors were all re-

w w w. S c i A m . c o m SCIENTIFIC A MERIC A N 69
© 20 08 SCIENTIFIC AMERIC AN, INC.
ing normal function to tumor vessels thereby cre- signed to block the main VEGF receptor used by
ates a period during which treatment with a variety endothelial cells and saw signs of vessel normaliza-
of cancer therapies should be maximally effective. tion begin after one day. During the normalization
window—which lasted only about five to six days—
Window of Opportunity tumor oxygenation increased and radiation therapy
t o t ru ly b e n e f i t from this new insight into yielded the best therapeutic outcome. Other teams
the way that antiangiogenic therapy can work with working with laboratory animals have subsequent-
radiation or chemotherapies, an oncologist would ly reported similar observations.
need to know when a patient’s tumor vessels first Enough evidence supported this model, in fact,
begin to normalize and how long they remain that that we were able to test it in another National Can-
way. My research group returned to experimenting cer Institute clinical trial, completed in late 2006.
with mice to better characterize this period we came Led by my Massachusetts General Hospital col-
to call the “normalization window.” We treated leagues Tracy Batchelor and Gregory Sorensen, the
brain tumors in the animals with an antibody de- trial included 30 patients whose brain tumors, known
as glioblastomas, had regrown despite aggressive sur-
gery, radiation and chemotherapy. These patients
Vessel Repair: Beyond Cancer had a life expectancy of less than six months.

H
They received a daily oral dose of Recentin, an
undreds of millions of people around the world suffer from
experimental drug that potently inhibits the three
noncancerous conditions that involve abnormal vasculature.
primary cellular receptors for VEGF. Using ad-
Modifying blood vessel growth and function might become a key
component of the therapeutic arsenal for those diseases as well, so drugs that vanced imaging techniques, we were able to look for
normalize blood vessels have the potential to vastly impact human health. effects in their tumors and saw them almost imme-
Among the most widespread of problems in this category, for example, is diately [see illustration on opposite page]. The signs
atherosclerosis, an artery disease whose features include an accumulation of vascular normalization included reduced vessel
of fatty plaques within the inner walls of blood vessels. Inside such plaques, diameter and leakiness, which continued for at least
inflammation-inducing blood cells and other detritus accumulate, gradually 28 days, with some normalized characteristics per-
enlarging the lesion. New blood vessels sprout within this growing mass to sisting for the entire four-month duration of the
feed it, much like a tumor. These new vessels also share many abnormal study. Moreover, as anticipated by our original mod-
features with tumor vessels, such as leakiness and disorganization. In el, the normalization was accompanied by a rapid
principle, therefore, applying antiangiogenic agents should normalize decrease in swelling in and around the tumor, an ef-
intraplaque vessels, stabilizing the lesions, halting their expansion and fect that continued as long as the patients took the
reducing their potential for rupture. Recentin. Because the side effects of VEGF inhibi-
Eye diseases such as diabetic retinopathy and the so-called wet form of tion can be severe, however, some patients asked for
age-related macular degeneration (AMD) are also characterized by vascular a break from the treatment during the trial, which
abnormalities similar to those seen in tumors. A hallmark of wet AMD is, in allowed us to observe the tumor vessels becoming
fact, the leakiness of blood vessels in the retina at the back of the eye. As abnormal again when Recentin was discontinued
a result, blood oozes into surrounding tissue, causing partial or total and renormalized when the drug was resumed.
vision loss. More than nine million Americans are currently affected. Not These results were the first to define how long
surprisingly, the greatest progress outside the realm of cancer treatment in the period of vascular normalization can last in hu-
using antiangiogenesis to repair vascular abnormalities has been in wet
mans and have led to a much larger ongoing clinical
AMD. Two drugs, Lucentis and Macugen — both inhibitors of VEGF — have
trial involving 300 patients to further define the
already been approved for treating the condition and most likely work by
role that Recentin, with and without chemothera-
normalizing the leaky vessels.
py, might play in the treatment of glioblastoma. We
These same normalization principles may also be useful for controlling
conditions that cause fluid buildup (edema) and for tissue engineering and are also studying a number of anti­angiogenic drugs
regenerative medicine, which require the creation and maintenance of in combination with traditional therapies in newly
normally functioning vasculature.  — R.K.J. diagnosed and recurrent tumor cases in more types
of cancer.
WET AMD affects the macula, a
At the same time, we are also investigating ways
k an g z h an g and james g ilman

region of the retina at the back


of the eye that appears slightly of expanding the window of normalization so that
darkened in a healthy human survival improvements can be extended from months
eye (left) but can sprout excess to years. Any potential strategy to repair vessels
blood vessels leading to must recognize that VEGF blockade alone may not
macular degeneration. The
abnormal vessels leak blood always be sufficient to achieve or sustain normaliza-
(right) into surrounding tissue, tion, because tumors can substitute other growth
obscuring vision. factors to get around the loss of VEGF signaling. As

70 SCIENTIFIC A MERIC A N ne w ans w ers f o r C ancer


© 20 08 SCIENTIFIC AMERIC AN, INC.
Before treatment After treatment
1 day 27 days 111 days
BR AIN TUMOR response to
TUMOR SIZE
tumors grow larger, for instance, antiangiogenesis therapy
they tend to make a diverse array reflects some effects of
vascular normalization.
of proangiogenic molecules in ad-
MRI images of a patient’s
dition to VEGF, so their vessels brain one day before
may gradually lose responsiveness beginning treatment and
to a treatment such as Avastin. at subsequent intervals
In rectal cancer patients, for show a shrinking white area
corresponding to the SWELLING
instance, our group discovered
malignant tumor’s location
that blood levels of VEGF and (top row). Another set
PlGF (placental growth factor), a of views highlights
related molecule, actually in- accumulated fluid in the
creased after VEGF was mopped same area receding
(middle row). As tumor-
up with Avastin, suggesting that
generated edema
the tumor or other tissues began dimin­ishes, reduced WHITE MATTER
manufacturing more of those fac- compression of brain tissue
tors in response. And in recurrent is visible in the rebounding
glioblastoma patients, blood lev- of fibrous white matter
tracts, which are
els of multiple proangiogenic mol-
F r o m “A Z D 2 17 1 , a Pan - V E G F R ece p t o r T yr o sine Kinase I n h ibit o r , N o rmalizes T um o r V asculature and A lleviates E dema in Gli o blast o ma Patients , ”

color-coded here by brain


ecules rose as tumors escaped the region (bottom row).
Recentin treatment.
This diversification of pro-
growth signals illustrates that the challenge for the longer valid. If a drug is sufficiently potent and spe-
oncologist will be to formulate cocktails of agents cific to destroy enough tumor vasculature to starve
specifically tailored to the molecular profile of each the ­entire tumor and save a patient’s life, then that
patient’s primary and metastatic tumors and to would be a happy outcome for everyone. But the
changes in those profiles that will likely occur over ability to use these drugs for vascular repair as well
time. It is worth noting, however, that the available makes them valuable tools for attacking tumors in
tools for promoting vascular normalization are not more than one way. In the longer term, this research
by T. B atc h el o r et A l . , in C A N C E R C E L L , v o l . 11 ; J anuary 2 0 0 7, © 2 0 0 7 R e p rinted w it h p ermissi o n O F E lsevier

limited to drugs targeting VEGF or other growth can also benefit the many millions of people around
factors directly. We have shown in mice, for exam- the world suffering from other diseases caused by
ple, that the drug Herceptin— an antibody that tar- abnormal vasculature, such as age-related macular
gets a tumor cell-surface protein called HER2 and degeneration and atherosclerosis [see box on oppo-
that is given to about a quarter of women with breast site page].
cancer— can mimic the responses produced by an More than 30 years ago, when I first set out to
antiangiogenic cocktail and normalize tumor ves- understand the tortuous and dysfunctional blood
sels. Herceptin indirectly lowers cellular manufac- vessels of tumors, I never imagined where that road
ture of several proangiogenic molecules while in- would lead. Nor could I have pictured a day when a
creasing the cells’ production of the antiangiogenic patient with a disease of abnormal blood vessels
thrombospondin-1. could walk into a clinic, have various biomarkers
In addition to identifying new and existing med- measured, then receive a tailored regimen of nor-
ications that can foster vascular normalization, it malizing drugs to repair those vessels. But now that
will be important to find minimally invasive and day looks closer than ever before. 
affordable ways for doctors to monitor the normal-
ization process, to best exploit it when delivering more to explore
treatments. To that end, my colleagues and I have Normalization of the Tumor Vasculature: An Emerging Concept in Anti-Angiogenic
been working to identify so-called biomarkers: Therapy. Rakesh K. Jain in Science, Vol. 307, pages 58–62; January 2005.
readily identifiable signs that reflect what is happen- Lessons from Phase III Clinical Trials of Anti-VEGF Therapy for Cancer. Rakesh K.
ing inside the tumor and thereby reveal the onset Jain, Dan G. Duda, Jeffrey W. Clark and Jay S. Loeffler in Nature Clinical Practice
Oncology, Vol. 3, No. 1, pages 24–40; January 2006.
and duration of the normalization window in indi-
Angiogenesis in Brain Tumors. Rakesh K. Jain et al. in Nature Reviews
vidual patients. Such markers might include, for ex- Neuroscience, Vol. 8, pages 610–622; August 2007.
ample, proteins in the bloodstream or in urine
Antiangiogenic Therapy for Normalization of Atherosclerotic Plaque Vasculature:
whose levels rise or fall during this time window. A Potential Strategy for Plaque Stabilization. Rakesh K. Jain et al. in Nature Clinical
Finding that antiangiogenesis drugs can nor- Practice Cardiovascular Medicine, Vol. 4, No. 9, pages 491–503; September 2007.
malize vasculature should not suggest that the orig- For more articles by Rakesh K. Jain and animations explaining vessel normalization,
inal purpose for which they were developed is no go to http://steele.mgh.harvard.edu

w w w. S c i A m . c o m SCIENTIFIC A MERIC A N 71
© 20 08 SCIENTIFIC AMERIC AN, INC.
TUMOR-bustING
A technique called virotherapy harnesses
viruses, those banes of humankind, to stop
another scourge— cancer
By Dirk M. Nettelbeck, Ronald D. Alvarez
and David T. Curiel

V
iruses are some of the most insidious creations in nature. They travel light:
equipped with just their genetic material packed tightly inside a crystalline
case of protein, they latch onto cells, insert their genes, and co-opt the cells’
energy-producing, gene-copying and protein-making machinery, using them to
make thousands of copies of themselves. Once formed, the new viruses percolate to
the cell surface, pinch off inside minuscule bubbles of cell membrane
and drift away, or else they continue reproducing until the cell finally
bursts. In any case, they go on to infect and destroy other cells, result-
ing in diseases from AIDS to the common cold.
Different viruses cause different diseases in part because each virus enters a cell
by first attaching to a specific suction cup–like receptor on its surface. Liver
cells display one kind of receptor used by one family of viruses, whereas nerve
cells display another receptor used by a different viral family, so each type of
virus infects a particular variety of cell. Cancer researchers have envied this
selectivity for years: if they could only target cancer therapies to tumor cells
and avoid damaging normal ones, they might be able to eliminate many of the nox-
ious side effects of cancer treatment.
Some scientists, including ourselves, are now genetically engineering a range of
viruses that act as search-and-destroy missiles: selectively infecting and killing can-
cer cells while leaving healthy ones alone. This new strategy, called virotherapy, has
shown promise in animal tests, and clinical trials involving human patients are now
under way. Researchers are evaluating virotherapy alone, as well as viruses “armed”
with therapeutic genes that enable them to administer traditional chemotherapies
solely to tumor cells. They are also developing methods to label viruses, or the cells
infected by these viruses, to track the movement of the viral agents in patients.
One of the first inklings that viruses could be useful in combating cancer came in
CREDIT

1912, when an Italian gynecologist observed the regression of cervical cancer in

72 SCIENTIFIC A MERIC A N ne w ans w ers f o r C ancer


© 20 08 SCIENTIFIC AMERIC AN, INC.
ADENOVIRUSES explode from
a cancer cell that has been
selectively infected in order
to kill it. The viruses can
CREDIT

spread to and wipe out other


tumor cells.

w w w. S c i A m . c o m SCIENTIFIC A MERIC A N 73
© 20 08 SCIENTIFIC AMERIC AN, INC.
a woman who was inoculated with a rabies vaccine made TARGETING MELANOMA
from a live, crippled form of the rabies virus. Physicians first
injected viruses into cancer patients intentionally in the late The skin cancer melanoma can be highly lethal unless detected
1940s, but only a handful appeared to benefit. Twenty years early; it arises from pigment cells in the skin or eye, called
later scientists found that a virus that causes the veterinary melanocytes, by uncontrolled growth and spread. Scientists
disorder Newcastle disease shows a preference for infecting are using the virotherapy approach to selectively kill melanoma
cells while leaving healthy cells alone. One technique for
tumor cells and began to try to enhance that tendency by
studying melanoma involves combining melanoma cells
growing the viruses for generations in human cancer cells in
laboratory culture dishes. Although critics countered that
such viruses could be exerting only an indirect effect against
cancer by generally activating an individual’s immune system
and making it more likely to detect and kill cancer cells, re-
ports continued to pop up in the medical literature linking
viral infection and cancer remission. In the early 1970s and
1980s two groups of physicians described patients whose
lymphomas shrank after they came down with measles.
The modern concept of virotherapy began in the 1990s,
when researchers genetically modified viruses to selectively
replicate in, and kill, tumor cells. Teams led by Frank McCor- Adenoviruses are distinct from the types of viruses usu-
mick of ONYX Pharmaceuticals in Richmond, Calif., and ally used in gene therapy to treat inherited disorders. Gene
Daniel R. Henderson of Calydon in Sunnyvale, Calif., inde- therapy traditionally employs retroviruses to splice a func-
pendently published reports showing they could target viro- tioning copy of a gene permanently into the body of a patient
therapy to human cancer cells grafted into mice, thereby elim- in whom that gene has ceased to work properly. Unlike retro-
inating the human tumors. (ONYX is no longer developing viruses, however, adenoviruses do not integrate their DNA
therapeutic viruses, and Calydon has been acquired by Cell into the genes of cells they infect; the genes they ferry into a
Genesys in South San Francisco, Calif.) Both groups used de- cell usually work only for a while and then are lost. Scientists
rivatives of an adenovirus, a cause of the common cold that has have investigated adeno­viruses extensively in gene therapy ap-
been intensively explored for virotherapy. (Other viruses under proaches to treat cancer, in which replication-defective de-
study include herpes simplex, measles, parvovirus, vaccinia, rivatives of the viruses are exploited as vectors for transferring
reo­virus and the avian Newcastle disease virus.) Adeno­virus is into cells genes that, for example, make cancer cells more sus-
appealing in part because researchers understand its biology ceptible than normal ones to chemotherapy. In general, tests
very well after years of trying to cure colds and of using the involving adenovirus vectors have been safe, but regrettably a
virus in molecular biology and gene therapy research. It con- volunteer died in 1999 after receiving an infusion of adenovi-
sists of a 20-sided protein case, or capsid, filled with DNA and ruses as part of a clinical trial to test a potential gene therapy
equipped with 12 protein “arms.” These protrusions have for a genetic liver disorder [see box on page 78].
evolved over millennia to latch onto a cellular receptor whose Gene therapists have been working to tailor adenoviruses
normal function is to help cells adhere to one another. and other viral vectors, or gene-delivery systems, to improve
their safety and reduce the chances that such a tragedy might
Overview/Anticancer Viruses occur again. It is perhaps even more essential for researchers
such as ourselves, who are investigating virotherapy, to develop
■  V irotherapy is a strategy to treat cancer by selectively safer, more targeted vectors, because virotherapy by definition
infecting and killing tumor cells. Researchers are aims to kill the cells the viruses infect and thereby produce a new
testing various approaches to target viruses to cancer generation of infectious viruses, not just insert a therapeutic gene
illustration by terese winslow (preceding pages)

cells, leaving normal cells untouched. into them. Replicating in the wrong cells could be dangerous.
■ T he viruses used in virotherapy can kill tumor cells

either by bursting them open or by combining virus Homing In on the Target


replication with the delivery of genes that make the v i ro t h e r a p is t s a r e de v isi ng two main strategies to
cells more susceptible to traditional chemotherapies. make sure their missiles hit their objectives accurately with no
■ T he same types of viruses used in virotherapy can also collateral damage. In the first approach, termed transduction-
be labeled with fluorescent tags or devices for al targeting, researchers are attempting to adapt the viruses
detection by nuclear medicine. Once delivered into the so that they preferentially enter, or transduce, cancer cells.
body, they home in on cancer cells. In the future, The second method, called transcriptional targeting, involves
physicians might be able to use this imaging technique altering the viruses so that their genes can be active, or tran-
to detect the presence of tiny tumors. scribed, only in tumors [see box on pages 76 and 77].
Transductional targeting is particularly necessary because

74 SCIENTIFIC A MERIC A N ne w ans w ers f o r C ancer


© 20 08 SCIENTIFIC AMERIC AN, INC.
(dark dots in micrograph below left) with normal skin cells called keratino­cytes appear red; viral proteins in cells show up green. The
keratinocytes and collagen to make cancer-bearing artificial skin center micrograph was made using viruses that were not
that can be grown in laboratory culture dishes. One of us specifically targeted to melanomas. The viruses were able to grow
(Nettelbeck) and colleagues have devised an adenovirus that can in healthy cells, making those cells look yellow. In contrast, the
specifically reproduce in melanoma cells. In the center and right targeted virus (below right) did not replicate in healthy cells, so
micrographs of fluorescently labeled artificial skin (below), healthy none of the cells are yellow.  — D.M.N., R.D.A. and D.T.C.

viruses have not evolved to infect and kill tumor cells. Unfor- must make much more of the pigment melanin than do liver
tunately, adenoviruses bind more efficiently to the variety of cells, which have little use for the protein. Accordingly, the pro-
normal tissues in the human body than they do to most tumor moter for the key enzyme for making melanin gets turned on
cells. We can reverse this pattern using specially generated in melanocytes but generally is off in most other body tissues.
adapter molecules made of antibodies that snap onto the arms In the deadly skin cancer melanoma, the gene encoding this
of the virus like sockets on a socket wrench. By attaching care- enzyme is frequently fully functional, making the tumors ap-
fully chosen antibodies or other molecules that selectively bind pear black. We, and others, have engineered adenoviruses that
only to a specific protein found on tumor cells, we can render have a promoter for the enzyme adjacent to genes that are es-
adenoviruses unable to infect any cells but cancerous ones. sential for the viruses’ ability to replicate. Although these vi-
Once the antibody-bearing virus latches onto a targeted cell, ruses might enter normal cells, such as liver cells, they can re-
the hapless cell engulfs it in a membrane sac and pulls it inside. produce only inside melanocytes, which contain the special
As the sac disintegrates, the viral capsid travels to a pore in the combination of proteins needed to turn on the promoter.
cell’s nucleus and injects its own DNA. Soon the viral DNA Researchers are currently tailoring adenoviruses with a va-
directs the cell to make copies of the viral DNA, synthesize viral riety of promoters that limit their activity to particular organs
proteins and combine the two into thousands of new adenovi- or tissues. In liver cancers, for example, the promoter for the
ruses. When the cell is full to capacity, the virus prompts the cell gene α-fetoprotein — which is normally shut down after fetal
to burst, releasing the new viruses to spread to other cells. development— becomes reactivated. Adenoviruses containing
The viruses can also be engineered more directly. In this that same promoter hold promise for eradicating liver tumors.
regard, Curiel’s group at the University of Alabama’s Gene Scientists led by Jonathan W. Simons, now at the Prostate Can-
Therapy Center has designed adenoviruses that bind to cellular
proteins called integrins. These molecules help cells stick to the DIRK M. NETTELBECK heads a research group focusing on viro-
the authors
and Louise T. Chow University of Alabama at Birmingham (micrographs)

network of connective tissue, called the extracellular matrix, therapy at the German Cancer Research Center and at Heidel-
dirk m. nettelbeck and N. S anjib Banerjee, Thomas R. Broker

that organizes the cells into cohesive tissues. Although integ- berg University Hospital. He was a molecular biologist and post-
rins are also made by healthy cells, cancer cells produce them doctoral fellow of the German Research Association at the Uni-
in abundance as they become metastatic and begin to squeeze versity of Alabama at Birmingham (U.A.B.) from 2000 to 2003.
through tissue layers and travel throughout the body. The Uni- He received his Ph.D. in 2000 from Philipps University in Mar-
versity of Alabama research group has had encouraging results burg, Germany. RONALD D. ALVAREZ is professor in and director
using the engineered viruses in mice bearing human ovarian of the division of gynecologic oncology at U.A.B., and he directs
cancers. The viruses homed in on the ovarian tumor cells and the Women’s Cancer Program of the U.A.B. Comprehensive Can-
killed them, ridding the treated animals of the disease. cer Center. He received his M.D. from Louisiana State University
Transcriptional targeting generally takes advantage of ge- in 1983. He has extensive experience in translational gene ther-
netic switches (promoters) that dictate how often a given gene apy research for ovarian cancer. DAVID T. CURIEL is director of
is functional (gives rise to the protein it encodes) in a particular the division of human gene therapy at the Gene Therapy Center
type of cell. Although each body cell contains the same ency- at U.A.B. Curiel, who holds an M.D. and a Ph.D., is Jeanne and
clopedia of genetic information, some cells use different chap- Anne Griffin Chair for Women’s Cancer Research at U.A.B. and a
ters of the encyclopedia more often than others to fulfill their professor of gene therapy at the Free University of Amsterdam.
specialized tasks. Skin cells called melanocytes, for instance, He has an equity interest in Birmingham-based VectorLogics.

w w w. S c i A m . c o m SCIENTIFIC A MERIC A N 75
© 20 08 SCIENTIFIC AMERIC AN, INC.
zapping cancer cells with viruses
Two main strategies are being explored for virotherapy, which is the technique of using reproducing viruses to kill tumors.
In the first method, dubbed transductional targeting (below), scientists are attempting to engineer viruses such as
adenoviruses — which normally cause respiratory infections — to selectively infect and destroy only cells that have turned
cancerous. They are attaching adapter molecules onto the viral outer coat proteins or directly modifying these proteins
to try to prevent the viruses from entering normal cells and to instead prompt them to home in on tumor cells.

NORMAL virotherapy with transductional targeting


ADENOVIRUS

Cell bursts,
and virus
Viral infects and
outer kills other
coat cancer cells
proteins

Adapter molecule
on engineered
adenovirus

Receptor
Normal made only by
adenovirus tumor cells
receptor

No infection Targeted virus


or cell killing takes over
cancer cell,
making so many
copies of itself
that it kills the
infected cell

Normal cancer
cell cell

Viral dna

cell dna cell dna


terese winslow

76 S C I E N T I F I C A M E R I C A N ne w ans w ers f o r C ancer


© 20 08 SCIENTIFIC AMERIC AN, INC.
The second approach (below) involves placing a snippet of DNA called a tumor-specific promoter next to one of the
adenovirus’s essential genes. The promoter acts as an “on” switch that permits the gene to function only in cancer cells.
The engineered viruses can enter normal cells, but they cannot reproduce and kill them. Once they enter cancer cells, however,
the tumor-specific promoter lets them make thousands of copies of themselves and ultimately burst the cancer cells. They
can then spread to — and destroy — other tumors.  — D.M.N., R.D.A. and D.T.C.

virotherapy with transcriptional targeting

Engineered adenovirus
with tumor-specific Cell bursts,
promoter linked to and virus
essential virus gene infects and
kills other
cancer cells

Promoter

Infection occurs,
but normal cell does
not have switch to
turn on viral gene;
virus cannot
replicate or kill cell

Normal cancer
cell cell

Viral dna Tumor-


specific
promoter

Cancer cell has


switch to turn on
cell dna viral replication
genes

w w w. S c i A m . c o m SCIENTIFIC A MERIC A N 77
© 20 08 SCIENTIFIC AMERIC AN, INC.
cer Foundation, have tested the approach in men whose pros- A further strategy for targeting virotherapy makes the most
tate cancer recurred following treatment with radiation. The of one of cancer’s hallmarks: the ability of tumor cells to divide
researchers used adenoviruses that had been engineered by again and again in an uncontrolled manner. Healthy cells
Cell Genesys to contain the promoter for prostate-specific an- make proteins that serve as natural brakes on cell division —
tigen, a protein made in abundance by prostate tumors. They notably the retinoblastoma (Rb) protein. As cells turn cancer-
administered the virotherapy to 20 men who received varying ous, however, the genes that code for one or the other of these
doses of the adenoviruses. In 2001 Simons and his colleagues proteins become mutated or otherwise inactivated. Certain
reported that none of the men experienced serious side effects viruses, including adenoviruses, interfere with the braking
and that the tumors of the five men who received the highest mechanisms of a normal cell by making proteins that stick to
doses of the virotherapy shrank by at least 50 percent. and inactivate Rb. They do this because they can replicate only
in cells that are preparing to divide.
Other Strategies Several research groups and biotechnology companies have
v i ro t h e r a p i s t s might end up combining the transduc- engineered adenoviruses that fail to make the Rb blockers.
tional and transcriptional targeting strategies to ensure that the Normal cells, which make these blockers, will stall the replica-
viruses kill only tumor cells and not normal ones. ­Adeno­viruses tion of these viruses by putting the brakes on cell division. But
engineered to contain the promoter for the enzyme that makes these viruses will replicate in cells in which the Rb protein is
melanin, for instance, can also replicate in normal ­melan­ocytes, already disabled — cancer cells — and kill them. Curiel and Al-
so on their own they might cause spots of depigmentation. And varez are currently conducting a clinical trial of this approach
adenoviruses that are designed to bind to receptors on the sur- using a virus that is also integrin-targeted, as described above.
faces of tumor cells can still invade a small proportion of healthy The first phase I study began in early 2007. In this trial, escalat-
cells. But viruses altered to have several fail-safe mechanisms ing dosages of a modified adenovirus are injected daily for
would be expected to be less likely to harm normal cells. There three days into the abdomens of patients with recurrent ovar-
are no results at present, however, to demonstrate that a com- ian or other gynecological cancers. The objectives are to de-
bination of approaches makes viruses more targeted. termine the maximum dose that can be tolerated, the toxicity,

But Is It Safe? adenoviruses had spread to his spleen,


lymph nodes and bone marrow, and an
Many approaches to virotherapy use
adenoviruses, which caused a death examination of his records revealed that
in a gene therapy clinical trial his liver function was probably too
impaired for him to be a volunteer in the
In September 1999 18-year-old Jesse trial. A number of scientists have also
Gelsinger died after receiving an infusion of suggested that he might have mounted
adenoviruses into his liver. He had a mild such an extreme immune reaction
form of an inherited liver disease called because he had previously been infected
ornithine transcarbamylase deficiency with a naturally occurring adenovirus.
(OTCD) and was participating in a clinical Since Gelsinger’s death, gene
trial of a new gene therapy to use therapists and virotherapists alike have
adenoviruses to ferry a corrected copy of focused on refining adenoviruses to make
the gene encoding OTCD into his liver cells. them safer. But researchers are still unsure
Unfortunately, four days after an infusion Jesse Gelsinger, who died in 1999 after why Gelsinger reacted so violently to the
of the viruses, he died of acute respiratory receiving an infusion of adenoviruses, adenoviral infusions: a second patient
distress syndrome and multiple organ in a family photograph.
participating in the same clinical trial
failure, apparently caused by an over­ tolerated a similar dose of the viruses. And
whelming immune reaction to the large dose of adenoviruses hundreds of other people worldwide have been treated so far
he had been administered as part of the trial. with adenoviruses, including replication-competent
Although Gelsinger’s death was part of a gene therapy trial, adenoviruses, with no serious side effects.
the tragedy also has ramifications for the young field of A National Institutes of Health report generated in the
virotherapy. Gene therapy uses crippled versions of viruses aftermath of Gelsinger’s demise recommends that all
such as adenoviruses to introduce a new gene into cells; participants in such clinical trials be monitored closely for
virotherapy employs actively replicating viruses (which may or toxic reactions before and after the infusion of therapeutic
may not contain added genes), including adenoviruses, to kill viruses. It also stipulates that volunteers be screened for any
mbr / ne w sc o m

specific types of cells. predisposing conditions that would increase their sensitivity
Gelsinger’s autopsy showed that the engineered for the viruses.  — D.M.N., R.D.A. and D.T.C.

78 SCIENTIFIC A MERIC A N ne w ans w ers f o r C ancer


© 20 08 SCIENTIFIC AMERIC AN, INC.
Selected Companies Involved in Virotherapy
Company Headquarters Virus Diseases Viral Modifications Clinical Trial Status
BioVex Abingdon, England Herpes simplex Breast cancer and Carries the gene for Phase I/II
virus (HSV) melanoma granulocyte-macrophage
colony stimulating factor, an
immune system stimulant
Cell Genesys South San Francisco, Adenovirus Prostate cancer Targeted to prostate cancer Phase I/II
Calif. cells using prostate-specific
promoters
Crusade Glasgow, Scotland HSV Glioma (brain cancer), Has a gene deletion that Phase II for glioma and
Laboratories head and neck cancer, restricts it to actively head and neck cancer;
melanoma dividing cells such as cancers phase I for melanoma
MediGene Martinsried, Germany HSV Glioma and colon Harbors two gene deletions Phase II for glioma;
cancer that has that prevent it from phase I for colon cancer
spread (metastasized) reproducing in normal cells metastases
Oncolytics Calgary, Alberta Reovirus Prostate cancer Able to replicate only in Phase II for prostate
Biotech and glioma cancer cells bearing the cancer; phase I/II
activated oncogene RAS for glioma
VectorLogics Birmingham, Ala. Adenovirus Ovarian cancer Expanded tropism for cancer Phase I/II
cells and reduced replication
in normal cells
Note: Phase I tests are designed to evaluate safety in small numbers of patients.
Phases II and III are intended to determine the appropriate dose and efficacy, respectively.

the immunological response, and any antitumor effects of the techniques. The fluorescent protein might work best for can-
adenovirus. To date, nine patients have been treated with dos- cers that are accessible by an endoscope, such as cancers of the
ages of up to 10 billion virus particles a day. No toxicity has larynx. Physicians could peer into the endoscope and see ex-
been detected, and more patients are joining the trial. actly where the viruses — and, therefore, cancer cells — are by
Researchers are also arming therapeutic viruses with looking for fluorescence. So far the approach has worked best
genes that make the cells they infect uniquely susceptible to with viruses that do not kill cells, however. Nevertheless, we
chemo­therapy. The technique involves splicing into the vi- are convinced that such sophisticated imaging technologies
ruses genes that encode enzymes that turn nontoxic precur- will enable scientists to draw more meaningful conclusions
sors, or “prodrugs,” into noxious chemotherapies. In one from future clinical trials of virotherapy.
example, which was reported in 2002, André Lieber of the In 1995 gene therapy pioneer W. French Anderson of the
University of Washington and his co-workers designed ade- University of Southern California School of Medicine predict-
noviruses to carry genes encoding the enzymes capable of ed in Scientific American that “by 2000 ... early versions of
converting innocuous prodrugs into the anticancer com- injectable vectors that target specific cells will be in clinical
pounds camptothecin and 5-fluoro­uracil. The scientists en- trials.” These trials indeed began on schedule, as well as some
gineered the viruses so that they could make the enzymes he could not have envisioned then. We envision a substantial
only in actively dividing cells, such as cancer cells. When they role for viruses — that is, therapeutic viruses — in 21st-century
injected the viruses and the prodrugs into mice bearing im- medicine.
planted human colon or cervical cancer cells, they found that
the viruses reproduced and spread in the tumors. more to explore
Such “smart” virotherapies are the vanguard of the fu- Gene Therapy: Designer Promoters for Tumour Targeting. D. M.
Nettelbeck, V. Jérôme and R. Müller in Trends in Genetics, Vol. 16, No. 4,
ture. But physicians will also need to track the activity of pages 174–181; April 2000.
virotherapies in a patient’s body to best assess how well the Replicative Adenoviruses for Cancer Therapy. R. Alemany, C. Balagué
strategies are working and to refine them further. Virothera- and D. T. Curiel in Nature Biotechnology, Vol. 18, No. 7, pages 723–727;
pists are now teaming with radiologists to establish novel July 2000.
Recent Progress in the Battle between Oncolytic Viruses and
imaging technologies to easily measure how effectively a giv- Tumours. K. A. Parato, D. Senger, P.A.J. Forsyth and J. C. Bell in Nature
en virotherapy is replicating. Reviews Cancer, Vol. 5, No. 12, pages 965–976; December 2005.
The imaging strategies involve inserting a gene that gov- Engineering Targeted Viral Vectors for Gene Therapy. R. Waehler,
erns the production of a tracer molecule into a virus or virus- S. J. Russel and D. T. Curiel in Nature Reviews Genetics, Vol. 8, No. 8,
pages 573–587; August 2007.
infected cell. The tracer can be either a fluorescent protein that
Gene therapy clinical trials database of Journal of Gene Medicine:
can be observed directly or one that binds to, or activates and www.wiley.com/legacy/wileychi/genmed/clinical/
traps, the radionuclides used in standard radiological imaging American Society of Gene Therapy: www.asgt.org

w w w. S c i A m . c o m SCIENTIFIC A MERIC A N 79
© 20 08 SCIENTIFIC AMERIC AN, INC.
Pigments that
turn caustic on
exposure to light
can fight cancer,
blindness and heart

New disease. Their light-


induced toxicity

Light
may also help
explain the origin
of vampire tales

on Medicine
Stories of vampires date back thousands of years.
Our modern concept stems from Bram Stoker’s quirky classic Dracula and Hol-
lywood’s Bela Lugosi— the romantic, sexually charged, bloodsucking outcast with
a fatal susceptibility to sunlight and an abhorrence of garlic and crosses. In con-
trast, vampires of folklore cut a pathetic figure and were also known as the undead.
In searching for some underlying truth in vampire stories, researchers have specu-
lated that the tales may have been inspired by real people who suffered from a rare
blood disease, porphyria. And in seeking treatments for this disorder, scientists
have stumbled on a new way to attack other, more common serious ills.
Porphyria is actually a collection of related diseases in which pigments called
porphyrins accumulate in the skin, bones and teeth. Many porphyrins are benign
in the dark but are transformed by sunlight into caustic, flesh-eating toxins. Without
treatment, the worst forms of the disease (such as congenital erythropoietic porphy-
ria) can be grotesque, ultimately exacting the kind of hideous disfigurement one
might expect of the undead. The victims’ ears and nose get eaten away. Their lips
and gums erode to reveal red, fanglike teeth. Their skin acquires a patchwork of
scars, dense pigmentation and deathly pale hues, reflecting underlying anemia. Be- LIGHT-ACTIVATED DRUGS
used in photodynamic therapy
cause anemia can be treated with blood transfusions, some historians speculate that
could treat diseases of the eye,
cancers such as those of

By Nick Lane
CREDIT

the esophagus, and coronary


artery disease.

80 SCIENTIFIC A MERIC A N ne w ans w ers f o r C ancer


© 20 08 SCIENTIFIC AMERIC AN, INC.
CREDIT

w
www w.
w. SS cc ii AA m
m .. cc oo m
m S C I E N T I FSI CC IAEMNET RI FI CI CA ANMREERPIOCRA TNS 81
81
© 20 08 SCIENTIFIC AMERIC AN, INC.
in the dark ages people with porphyria might have tried drink- generates energy by transferring electrons to oxygen in a crit-
ing blood as a folk remedy. Whatever the truth of this claim, ical step of cellular respiration).
those with congenital erythropoietic porphyria would certain- Porphyria arises because of a flaw in the body’s heme-
ly have learned not to venture outside during the day. They making machinery. The body produces heme and other por-
might have learned to avoid garlic, too, for some chemicals in phyrins in a series of eight coordinated stages, each catalyzed
garlic are thought to exacerbate the symptoms of the disease, by a separate enzyme. Iron is added at the end to make heme.
turning a mild attack into an agonizing reaction. In porphyria, one of the steps does not occur, leading to a
While struggling to find a cure for porphyria, scientists backlog of the intermediate compounds produced earlier in
came to realize that porphyrins could be not just a problem but the sequence. The body has not evolved to dispose of these
a tool for medicine. If a porphyrin is injected into diseased tis- intermediates efficiently, so it dumps them, often in the skin.
sue, such as a cancerous tumor, it can be activated by light to The intermediates do not damage the skin directly, but many
destroy that tissue. The procedure is known as photodynamic of them cause trouble indirectly. Metal-free porphyrins (as
therapy, or PDT, and has grown from an improbable treatment well as metalloporphyrins containing metals that do not in-
for cancer in the 1970s to a sophisticated and effective weapon teract with the porphyrin ring) can become excited when they
against a diverse array of malignancies today and, most re- absorb light at certain wavelengths; their electrons jump into
cently, for macular degeneration and pathological myopia, higher-energy orbitals. The molecules can then transmit their
common causes of adult blindness. Ongoing research includes excitation to other molecules having the right kind of bonds,
pioneering treatments for coronary artery disease, AIDS, au- especially oxygen, to produce reactive singlet oxygen and
toimmune diseases, transplantation rejection and leukemia. other highly reactive and destructive molecules known as free
radicals. Metal-free porphyrins, in other words, are not the
Molecular Mechanisms agents, but rather the brokers, of destruction. They catalyze
t h e su b s ta n c e s at the heart of porphyria and photody- the production of toxic forms of oxygen.
namic therapy are among the oldest and most important of all Photosensitive reactions are not necessarily harmful.
biological molecules, because they orchestrate the two most Their beneficial effects have been known since ancient times.
critical energy-generating processes of life: photosynthesis In particular, some seeds and fruits contain photosensitive
and oxygen respiration. Porphyrins make up a large family of chemicals (photosensitizers) called psoralens, which indirect-
closely related compounds, a colorful set of evolutionary vari- ly led scientists to experiment with porphyrins. Psoralens
ations on a theme. All porphyrins have in common a flat ring have been used to treat skin conditions in Egypt and India for
(composed of carbon and nitrogen) with a central hole, which several thousand years. They were first incorporated into
provides space for a metal ion such as iron or magnesium to modern medicine by Egyptian dermatologist Abdel Monem
bind to it. When aligned correctly in the grip of the porphyrin El Mofty of Cairo University some 60 years ago, when he
rings, these metal atoms catalyze the most fundamental ener- began treating patients with vitiligo (a disease that leaves ir-
gy-generating processes in biology. Chlorophyll, the plant regular patches of skin without pigment) and, later, those
pigment that absorbs the energy of sunlight in photosynthe- with psoriasis using purified psoralens and sunlight. When
sis, is a porphyrin, as is heme, which is at the heart of the activated by light, psoralens react with DNA in proliferating
oxygen-transporter protein hemoglobin and of many en- cells to kill them.
zymes vital for life, including cytochrome oxidase (which Two American dermatologists, the late Aaron B. Lerner
of Yale University and the late Thomas B. Fitzpatrick of Har-
Overview/Light Therapy vard University, were struck by the potential of psoralens. In
the 1960s they showed that psoralens are activated by ultra-
■ I n photodynamic therapy, light-activated chemicals violet (UVA) rays, and the researchers later refined psoralen
called porphyrins are used to destroy fast-growing cells therapy using an ultraviolet lamp similar to those used in
and tissue. Doctors could apply the treatment to a solaria today. Their method became known as PUVA (short
variety of ailments, including age-related macular for psoralen with UVA) and is now one of the most effective
degeneration, tumors and atherosclerotic plaques. treatments for psoriasis and other skin conditions.
photograph by sam ogden (preceding page)

■ A few porphyrin drugs are on the market, and several

others are undergoing human trials. A Way to Kill Cancer Cells?


■ Researchers got the idea for photodynamic therapy i n t h e e a r ly 19 7 0 s the success of PUVA impressed
from their knowledge of the rare disease porphyria, in Thomas J. Dougherty of the Roswell Park Cancer Institute
which porphyrins accumulate in the skin and certain in Buffalo, N.Y., leading him to wonder if a variant of it could
organs. Unless the disease is managed, victims of the be effective against cancer. Activated psoralens can kill rogue
severest type of porphyria can become disfigured, cells to settle inflammation, but in comparison with porphy-
leading some researchers to speculate that they may rins they are not potent photosensitizers. If psoralens could
have inspired medieval vampire legends. kill individual cells, could porphyrins perhaps devour whole
tumors? His idea was the beginning of true photodynamic

82 SCIENTIFIC A MERIC A N ne w ans w ers f o r C ancer


© 20 08 SCIENTIFIC AMERIC AN, INC.
Chlorophyll tail

Chlorophyll

No metal
Iron

Magnesium

Basic Porphyrin Ring Heme

therapy, in which photosensitizers catalyze the production of PORPHYRINS all have in common a flat ring, mainly composed of carbon
oxygen free radicals. The concept was built on pioneering and nitrogen, and a central hole where a metal ion can sit. The basic ring
( far left) becomes caustic when exposed to light; molecules useful for
work from German physicians Oscar Raab, then a medical photodynamic therapy also share this trait. Nontoxic examples include
student, and Hermann von Tappeiner, his professor. Around heme (a component of the oxygen transporter hemoglobin) and the
the start of the 20th century, Raab and von Tappeiner showed chlorophyll that converts light to energy in plants.
that acridine, when activated by light, reacts with oxygen to
kill protozoa such as paramecia. Von Tappeiner had even them reported burns and skin rashes in the weeks after PDT.
gone on to treat patients with skin cancer using eosin, a pho- Potency was another issue. The early porphyrin prepara-
tosensitive component of coal tar, and white light. Dougherty tions were mixtures, and they were seldom strong enough to
realized, however, that these early forms of photodynamic kill the entire tumor. Some porphyrins are not efficient at pass-
therapy lacked the raw power of porphyrins. He also drew ing energy to oxygen; others are activated only by light that
on two other medically useful properties of porphyrins dis- cannot penetrate more than a few millimeters into the tumor.
covered in the mid-20th century: porphyrins accumulate se- Some biological pigments normally present in tissues, such as
lectively in cancer cells and are activated by red light, which hemoglobin and melanin, also absorb light and in doing so can
penetrates more deeply into biological tissues than do short- prevent a porphyrin from being activated. Even the porphyrin
er wavelengths, such as white light or UVA. itself can cause this problem if it accumulates to such high
Dougherty injected a mixture of porphyrins into the blood- levels that it absorbs all the light in the superficial layers of the
stream of mice with mammary tumors. He then waited a few tumor, thus preventing penetration into the deeper layers.
days for the porphyrins to build up in the tumors before shin- Many of these difficulties could not be resolved without the
ing red light on them. His early setup was primitive, passing help of specialists from other disciplines. Chemists were need-
light from an old slide projector through a 35-millimeter slide ed to create new, synthetic porphyrins, ones that had greater
colored red. His results were nonetheless spectacular. The light selectivity for tumors and greater potency and that would be
activated the porphyrins within the tumor, which transferred activated by wavelengths of light able to reach farther into
their energy to oxygen in cells to damage the surrounding tis- tissues and tumors. (For each porphyrin, light activation and
sues. In almost every case, the tumors blackened and died after absorption occur only at particular wavelengths, so the trick
the light treatment. There were no signs of recurrence. is to design a porphyrin that has its absorption maximum at
Dougherty and his colleagues published their data in 1975 a wavelength that penetrates into biological tissues.) Physicists
in the Journal of the National Cancer Institute, with the brave were needed to design sources that could produce light of
title “Photoradiation Therapy II: Cure of Animal Tumors with particular wavelengths to activate the new porphyrins or that
Hematoporphyrin and Light.” Over the next few years they could be attached to fine endoscopes and catheters or even
refined their technique by using a low-power laser to focus red implanted in tissues. Pharmacologists were needed to devise
light onto the tumors. They went on to treat more than 100 ways of reducing the time that porphyrins spent circulating in
patients in this way, including people with cancers of the breast, the bloodstream, thereby restricting photosensitive side ef-
lung, prostate and skin. Their outcomes were gratifying, with fects. Finally, clinicians were needed to design trials that could
a “complete or partial response” in 111 of 113 tumors.
Sadly, though, cancer is not so easily beaten. As more phy- Nick Lane studied biochemistry at Imperial College, University
the author

sicians started trying their hand with PDT, some serious draw- of London. His doctoral research, at the Royal Free Hospital, con-
backs began to emerge. The affinity of porphyrins for tumors centrated on oxygen free radicals and metabolic function in or-
turned out to be a bit of an illusion— porphyrins are taken up gan transplants. Lane is an honorary reader at University Col-
by any rapidly proliferating tissue, including the skin, leading lege London and a regular contributor to Nature. His books in-
slim films

to photosensitivity. Although Dougherty’s original patients clude Oxygen: The Molecule That Made the World and Power, Sex,
were no doubt careful to avoid the sun, nearly 40 percent of Suicide: Mitochondria and the Meaning of Life.

w w w. S c i A m . c o m SCIENTIFIC A MERIC A N 83
© 20 08 SCIENTIFIC AMERIC AN, INC.
How Photodynamic Therapy Works
Doctors who administer photodynamic
therapy deliver photosensitive
... AT THE MOLECULAR LEVEL

1 3
chemicals called porphyrins
intravenously. These chemicals then A porphyrin absorbs light, Singlet oxygen reacts with
becoming activated. other substances in cells
collect in rapidly proliferating cells and, Singlet to produce destructive oxygen free
when exposed to light, initiate a cascade oxygen radicals; then cells die.
of molecular reactions that can destroy
Porphyrin
those cells and the tissues they in cells
compose. Some targets for the therapy Light
include abnormal blood vessels in the
retinas of people with age-related Oxygen
molecule (02)
macular degeneration (the leading
cause of adult blindness), cancerous
tumors, and atherosclerotic plaques
in coronary arteries.

2 The activated porphyrin


passes this light energy to
oxygen molecules, converting them Activated
porphyrin Cell dying from
to singlet oxygen. oxidative damage

... IN THE EYE


1 To treat macular degeneration, a
porphyrin (green) is injected into a
patient’s arm. It takes just 15 minutes for the
Fast-growing vascular tissue Photoreceptors porphyrin to accumulate in abnormal blood
Retinal pigment of the retina
epithelium vessels under the macula, the central part of
the retina responsible for color vision.
Normal blood vessels Area damaged
by disease

2 A red laser light activates the porphyrin,


which leads to the destruction of the
vascular tissue.
Red laser beam

3 After therapy halts damage to the


retina, the treated vascular tissue is
reabsorbed by the body, and the overlying
photoreceptors may settle back into
place. Because vessel growth could recur,
the patient may require several
additional treatments.

Damaged vascular tissue Normal blood vessels


being reabsorbed
CREDIT

84 SCIENTIFIC A MERIC A N ne w ans w ers f o r C ancer


© 20 08 SCIENTIFIC AMERIC AN, INC.
prove an effect and determine the best treatment regimens.
The ideal drug would be not only potent and highly selective
for tumors but also broken down quickly into harmless com-
pounds and excreted from the body. The first commercial prep-
. . . DEEP IN THE BODY aration, porfimer sodium (Photofrin), was approved by the U.S.
Even long wavelengths of visible light cannot penetrate Food and Drug Administration for the treatment of various
very far into tissue, so photodynamic therapies for diseased cancers. Although it has been helpful against certain cancers
tissue deep within the body require an internal light- (including esophageal, bladder, head and neck, and skin can-
delivery system. cers and some stages of lung cancer), it has not been the break-
through that had been hoped for and cannot yet be considered
1 Here, in an
experimental
therapy, an optical
Atherosclerotic
plaque containing
a porphyrin (green)
an integral part of cancer therapy. Surprisingly, though, the
first photosensitizing drug to fulfill most of the stringent crite-
fiber has been ria for potency and efficacy without causing photosensitivity,
threaded into an Artery verte­porfin (Visudyne), was approved in April 2000 by the FDA
artery in which a not to treat cancer at all but to prevent blindness. As the theo-
porphyrin has
ries converged with reality, researchers came to realize that
accumulated in
atherosclerotic PDT can do far more than destroy tumors.
plaques.
Battling Blindness
on e t h i ng i t could do, for instance, was combat age-re-
lated macular degeneration (AMD), the most common cause
of legal blindness in our maturing Western population. Most
Light people who acquire AMD have a benign form and do not lose
source
their sight, but about a tenth have a much more aggressive
type called wet AMD. In this case, abnormal, leaking blood

2 The fiber
produces red
light, activating the
vessels, like miniature knots of varicose veins, grow under-
neath the retina and ultimately damage the sharp central vi-
porphyrin. sion required for reading and driving. As the disease pro-
gresses, central vision is obliterated, making it impossible to
recognize people’s faces or the details of objects.
Most attempts to hinder this grimly inexorable process
have failed. Dietary antioxidants may be able to delay the
onset of the disorder but have little effect on the progression
of established disease. Until recently, the only treatment
proved to slow the progression of wet AMD was a technique
called laser photocoagulation. The procedure involves apply-
ing a thermal laser to the blood vessels to fuse them and thus
halt their growth. Unfortunately, the laser also burns the nor-
mal retina and so destroys a small region to prevent later loss
of vision in the rest of the eye. Whether this is worth it de-

3 Over the course


of a few days,
the porphyrin
pends on the area of the retina that needs to be treated. For
most people diagnosed with wet AMD, the area is located
destroys unwanted below the critical central part of vision or is already too large
plaques. to benefit from laser coagulation.
Against this depressing backdrop, researchers at Harvard
and at the biotechnology firm QLT in Vancouver, B.C., rea-
soned that PDT might halt the growth of these blood vessels
and delay or even prevent blindness. If porphyrins could accu-
mulate in any rapidly proliferating tissue — the very problem in
H y b rid M edica l A nimati o n

cancer— then perhaps they could also accumulate in the blood


vessels growing under the retina. Verteporfin, a novel synthet-
ic porphyrin, seemed promising because it had a good track
record in preclinical animal studies at QLT and at the Univer-
sity of British Columbia in the late 1980s and early 1990s.
CREDIT

Verteporfin accumulates in abnormal retinal vessels re-

w w w. S c i A m . c o m SCIENTIFIC A MERIC A N 85
© 20 08 SCIENTIFIC AMERIC AN, INC.
400 nm 500–550 nm 630–650 nm Coronary angioplasty is a minimally invasive procedure
for treating arteries affected by atherosclerosis. It uses a tiny
0.5 Epidermis balloon to open arteries, so that atherosclerotic plaques do
not occlude the entire vessel. Photoangioplasty could sidestep
Papillary many of the problems of conventional angioplasty, notably
1.0 dermis
the restenosis (renarrowing) of treated arteries. The proce-
Depth of Light Penetration (millimeters)

dure involves injecting a porphyrin into the bloodstream,


1.5
waiting for it to build up in the damaged arterial walls and
Porphyrin then illuminating the artery from the inside, using a tiny light
2.0 activation
to 1–2 cm source attached to the end of a catheter. The light activates
Reticular the porphyrins in the plaques, destroying the abnormal tis-
dermis Percentage of
2.5 light penetration sues while sparing the normal walls of the artery.
(width of cone) An even more ingenious method for treating deep cancers
3.0 is the use of self-lighting nanoparticles. Wei Chen of the Uni-
Subcutaneous versity of Texas at Arlington recently developed a method of
fatty tissue
attaching porphyrins to “scintillation luminescence nanopar-
3.5
ticles.” The nanoparticles generate visible light when exposed
EACH WAVELENGTH of light reaches a different depth in tissues, and to x-rays during radiotherapy, and this in turn activates the
any given porphyrin absorbs light at specific wavelengths. A porphyrin porphyrin. The method effectively combines radiotherapy
activated by deeper-penetrating light might be best for treating an
with PDT, optimizing both methods and allowing the treat-
internal tumor. In contrast to porphyrins, the psoralens used in PUVA
treatments for psoriasis are activated by near-ultraviolet light (400 ment of large or deeply buried tumors with relatively low-
nanometers), which barely penetrates the skin. dose radiotherapy. But how will the bulky nanoparticles be
targeted at tumor cells? In January, Chen and his colleagues
markably quickly: within 15 minutes of injection into an arm reported preliminary findings suggesting that the entire com-
vein. When activated by red laser light, verteporfin seals off plex can be attached to folic acid, which is then taken up by
the vessels, sparing the overlying retina. Any blood vessels folate receptors on cancer cells.
that grow back can be nipped in the bud by further treat- An alternative approach for targeting porphyrins or even
ments. Two major clinical trials, headed by Neil M. Bressler nanoparticle complexes at tumors involves attaching them to
of the Wilmer Eye Institute at Johns Hopkins University, con- antibody fragments that recognize cancer cells. Mahendra
firmed that PDT can be given six or seven times over a three- Deonarain and his colleagues at Imperial College London,
year period without damaging a healthy retina. For people who are pioneers of this method, have managed to attach
with the most aggressive form of AMD (with mostly “classic” more than 10 porphyrin molecules to a single antibody frag-
lesions), verte­porfin halved the risk of moderate or serious ment without destroying its ability to target cancer cells. The
vision loss over a two-year period. The effect is sustained over group used antibody fragments rather than whole antibodies
at least three or four years: patients who are not treated lose to reduce the size of the complex and enable it to be cleared
as much vision in three months as those treated with verte- from the body more quickly. The team’s initial findings, re-
porfin lose in three years. ported in March, showed the complexes accumulating in tu-
Some participants in the trials gained little benefit from mors at 10 times their concentration in blood and 50 times
PDT. Their disease may have already progressed too far. A their concentration in muscle.
reanalysis of clinical data presented by Bressler in April 2002 Accumulation of porphyrins in active and proliferating
at the International Congress of Ophthalmology in Sydney cells raises the possibility of treating other conditions in which
showed that smaller lesions respond much better to treatment abnormal cell activation or proliferation plays a role — among
than older, larger ones, implying that early detection and them, infectious diseases. Attempts to treat infections with
treatment may optimize the benefits of PDT. the pigments had long been frustrated by a limited effect on
gram-negative bacteria, which have a complex cell wall that
Other Treatment Avenues obstructs the uptake of porphyrins into these organisms. One
t h e s u c c e s s of ophthalmic PDT has inspired research solution, developed by Michael R. Hamblin and his colleagues
activity in other fields but also reveals the drawbacks of the at Harvard, involved attaching a polymer— usually polylysine,
treatment. In particular, even red light penetrates no more a repetitive chain of the amino acid lysine — to the porphyrin.
Hybrid Medical Animation

than a few centimeters into biological tissues [see illustration The polymer disrupts the lipid structure of the bacterial cell
above]. This limitation threatens the utility of PDT in inter- wall, enabling the porphyrins to gain entry to the cell. Once
nal medicine — its significance might seem to be skin deep. inside, they can be activated by light to kill the bacteria. In
There are ways of turning PDT inward, however. One inge- studies of animals with oral infections and infected wounds,
nious idea is called photoangioplasty, which is now being the altered porphyrin showed potent antimicrobial activity
tested to treat coronary artery disease. against a broad spectrum of gram-negative and gram-positive

86 S C I E N T I F I C A M E R I C A N ne w ans w ers f o r C ancer


© 20 08 SCIENTIFIC AMERIC AN, INC.
Photodynamic Therapies
The light-activated drugs listed below are a sampling of those on the market or in development.
DRUG TARGET MAKER STATUS
LEVULAN Acne and actinic keratosis DUSA PHARMACEUTICALS On the market for actinic keratosis;
(5-aminolevulinic acid) (precancerous skin disorder), Toronto phase II trials (relatively small studies in
Barrett’s esophagus humans) have begun for acne
(precancerous condition)

Photofrin Cancers of the esophagus and AXCAN SCANDIPHARM On the market for esophageal cancer and
(porfimer sodium) lung, high-grade dysplasia from Birmingham, Ala. nonsmall cell lung cancer; FDA-approved
Barrett’s esophagus for high-grade dysplasia

Visudyne Age-related macular QLT and NOVARTIS On the market


(verteporfin) degeneration, pathological OPHTHALMICS
myopia and ocular Vancouver, B.C., and
histoplasmosis (eye disorders) Duluth, Ga.

Metvix Actinic keratosis, basal cell skin PHOTOCURE FDA-approved for actinic keratosis
(methylaminolevulinic acid) cancer and squamous cell skin Oslo, Norway and basal cell skin cancer
cancer

LS11 Colorectal cancer, hepatocellular Light Sciences Oncology In phase III trials for colorectal cancer
(talaporfin sodium) carcinoma, and glioma Bellevue, Wash. with recurrent liver metastases and for
hepatocellular carcinoma; phase II trial
for glioma completed in February

TOOKAD Prostate cancer Steba Biotech In phase II/III trials for prostate cancer
(vascular-targeted PDT) Toussus-le-Noble, France after radiation therapy

bacteria. As antibiotic resistance becomes more intractable, es, components of our own body erroneously activate immune
targeted antimicrobial PDT could become a useful weapon in cells. These activated clones then proliferate in an effort to
the medical arsenal. destroy the perceived threat— say, the myelin sheath in mul-
Several other, related photodynamic methods hinge on tiple sclerosis or the collagen in rheumatoid arthritis. When
the finding that activated immune cells take up greater organs are implanted, activated immune cells may multiply to
amounts of photosensitizing drugs than do quiescent im- reject the foreign tissue — the transplanted organ or even the
mune cells and red blood cells, sparing the quiet cells from body tissues of the new host, in the case of bone marrow trans-
irreversible damage. In most infections, nobody would wish plants. In leukemia, immune cells and their precursors in the
to destroy activated immune cells: they are, after all, respon- bone marrow produce large numbers of nonfunctional cells.
sible for the body’s riposte to the infection. In these cases, In each instance, PDT could potentially eliminate the unwant-
targeting immune cells would be equivalent to “friendly fire” ed immune cells, while preserving the quiescent cells, to main-
and would give the infection free rein to pillage the body. tain a normal immune response to infection. As in HIV infec-
In AIDS, however, the reverse is true. The AIDS virus, tion, the procedure might work either extracorporeally or
HIV, infects the immune cells themselves. Targeting infected transdermally. Much of this research is in late-stage preclini-
immune cells would then be more like eliminating double cal or early clinical trials. For all the cleverness in exploring
agents. In the laboratory, HIV-infected immune cells take up possible medical applications, though, we can only hope that
porphyrins, thereby becoming vulnerable to light treatment. more extensive clinical studies will bear fruit.
In patients, the light could be applied either by withdrawing
blood, illuminating it and transfusing it back into the body more to explore
(extracorporeal phototherapy) or by shining red light onto the The Colours of Life: An Introduction to the Chemistry of Porphyrins
skin, in what is called transdermal phototherapy. In the trans- and Related Compounds. L. R. Milgrom. Oxford University Press, 1997.
Lethal Weapon. P. Moore in New Scientist, Vol. 158, No. 2130,
dermal approach, light would eliminate activated immune cells pages 40–43; April 18, 1998.
in the circulation as they passed through the skin. Whether the Investigation of Water-Soluble X-ray Luminescence Nanoparticles
technique will be potent enough to eliminate diseased immune for Photodynamic Activation. Y. Liu, W. Chen, S. Wang and A. G. Joly in
cells in HIV-infected patients remains an open question. Applied Physics Letters, Vol. 92, Article 043901; January 29, 2008.
Autoimmune diseases, rejection of organ transplants, and Targeted Photodynamic Therapy with Multiply-Loaded Recombinant
Antibody Fragments. M. Bhatti, G. Yahioglu, L. R. Milgrom, M. Garcia-
leukemias are also all linked by the common thread of acti- Maya, K. A. Chester and M. P. Deonarain in International Journal of
vated and proliferating immune cells. In autoimmune diseas- Cancer, Vol. 122, No. 5, pages 1155–1163; March 2008.

w w w. S c i A m . c o m SCIENTIFIC A MERIC A N 87
© 2008 SCIENTIFIC AMERIC AN, INC.
Gaining Ground on
Breast Cancer The newest targeted therapies are helping doctors to tailor increasingly
effective treatments to individual patients
by Francisco J. Esteva and Gabriel N. Hortobagyi

B
reast cancer is the most commonly diagnosed tein that trastuzumab was designed to attack, called
malignancy among women and, after lung HER2, promotes aggressive tumor growth. Before
cancer, the second leading cause of cancer-re- trastuzumab, diagnosis with a tumor that overpro-
lated deaths in North America. Yet unlike the sur- duces HER2 was dreaded news for patients. Now
vival rate for individuals diagnosed with lung cancer, it can be one of the tumor types with the best prog-
the rate for women diagnosed with breast cancer has nosis, because doctors have an increasing number
been rising dramatically over the past decade — to of effective weapons against HER2.
the point where breast cancer could soon lose its The next decade promises to be an exciting and
ranking as the second-greatest cancer killer. Noth- productive time in the field of molecular-targeted
ing would delight clinicians like us more. cancer therapy: additional drugs currently being
This improvement in overall outlook for wom- tested in people and animals are making it possible
en diagnosed with breast cancer is attributable in to go after an increasing variety of molecular tumor
part to earlier detection, which results from great- features that play a critical role in the initiation and
er awareness of, and access to, regular breast survival of malignancies and in the cancers’ pro-
screening. But breast cancer patients are also ben- gression to increasingly threatening stages. Along
efiting from accelerated research that has led to a with improvements in older therapies and support-
much better understanding of the disease and a ive care, this newer generation of drugs gives doc-
wider variety of treatment choices that doctors can tors more options for customizing treatment to
mix and match to tailor therapy for a particular cope with a tumor’s particular suite of molecular
patient. In just the past decade, it has even become characteristics and reflects our growing realization
possible to target drugs to specific molecules with- that breast cancer is not a single disease.
in tumors that help to drive the disease.
Breast cancer was, in fact, the first type of solid- Evolving Treatment Approaches
tumor cancer to be treated with this molecular-tar- a lt h ou g h t h e p ro sp e c t of tailoring treat-
geting therapeutic approach, when the drug trastu- ment to the molecular features of individual tumors
zumab (Herceptin) was approved in 1998. The pro- is incredibly encouraging, prior advances are also
contributing to the declining mortality rate for wom-
Overview/Targeted Treatment en diagnosed with breast cancer. Improved screening
techniques, for instance, are definitely helping to
Breast cancer survival rates have been steadily climbing in North
■  catch and confirm more cases at an earlier stage,
America and Europe, thanks to increased early detection and novel which is a boon, because breast cancer is highly cur-
treatment options. able if detected early. Newer imaging methods in-
■ Many new treatments target specific molecules on tumor cells, clude digital mammography (which produces a
allowing doctors to tailor medication to an individual patient’s clearer picture than screen-film mammography), ul-
tumor profile. trasound and magnetic resonance imaging (MRI).
■ B reast cancer was the first solid-tumor type for which molecular- In addition, surgical approaches to tumor exci-
j ames p o rt o

targeted therapy became available, and the success of the sion have changed over the past 20 years from radical
approach promises further dramatic advances. tissue removal in women whose tumor appears con-
fined to a small part of the breast to breast-conserv-

88 S C I E N T I F I C A M E R I C A N ne w ans w ers f o r C ancer


© 20 08 SCIENTIFIC AMERIC AN, INC.
CREDIT

w w w. S c i A m . c o m SCIENTIFIC A MERIC A N 89
© 20 08 SCIENTIFIC AMERIC AN, INC.
treatment MILESTONES
fast facts ing therapy. More focused radiation is also less dam-
aging to normal tissues such as those of the heart and Doctors started 1880s–1890s
Inherited mutations
lungs. These changes have made treatment less de- aggressively treating Hormonal
in the BRCA1 gene can
multiply lifetime structive, with equally successful results. breast cancer in the cancer
breast cancer risk by Besides these refinements in the detection and 19th century, with connection
the first mastectomy suggested when
10 times, but only in local management of breast tumors, the use of sys-
done in 1882. But physicians report
the past year have George T.
temic therapies as supplementary, or adjuvant, treat- significant
researchers discov- insights into mech­ regression of Beatson
ered why. BRCA1 is ments has become more sophisticated thanks to the anisms driving the breast cancer tumors
involved in DNA repair, availability of new drugs, improvements in their de- disease that would following ovary removal or
so its malfunction livery, and management of side effects. Such treat- lead to in­creas­ingly onset of menopause.
makes errors in other ments aim to eradicate any malignant cells not elim- 1896: First ovary removal
targeted therapies
cancer-promoting performed as a breast cancer
inated by surgery or radiation. The approach is often began with discov­ treatment by Beatson.
genes more likely.
warranted because even tumors that are tiny and ap- eries in the 1950s.
Following a 2002 parently self-contained can already have quietly
report that hormone spawned microscopic metastases, undetectable tu- to supportive care that increases their quality of life.
replacement therapy mors at distant sites in the body. By attacking these Another mainstay of breast cancer treatment, at
(HRT) increased invisible tumors, adjuvant chemotherapy can pro- least for patients with tumors that are determined
breast cancer risk
long disease-free intervals and overall survival rates. to be dependent on estrogen or progesterone, is en-
in postmenopausal
women, HRT use fell. Adjuvant chemotherapy is also improving the docrine therapy. Indeed, hormonal manipulations
The next year there odds for those with more advanced tumors. In the to treat breast cancer date as far back as the 1890s,
was a dramatic drop 1970s our clinical group and others began develop- when doctors observed tumors regressing after they
in the incidence of ing multidisciplinary treatment programs for pa- had removed the ovaries of premenopausal women

M ar k j . w inter P h o t o R e s e a r c h e r s , I n c . (e s t r o g e n r e c e p t o r) ; science m u se u m (t a m o x i f e n) ; p. mara z z i P h o t o R e s e a r c h e r s , I n c . (H e r c e p t i n)


both invasive (7.3 per-
tients with so-called locally advanced breast cancer, with advanced breast disease. In 1966 researchers
cent) and noninvasive
(5.5 percent) breast which has invaded neighboring tissue. Such patients identified hormone receptors— molecules that bind
cancers in the U.S. are often not diagnosed until their cancer is incur- to specific hormones— in various tissues, including
able with surgery alone. Our approach is to treat that of the breast. Subsequent studies showed that
them with preoperative, or neoadjuvant, chemother- a significant number of invasive breast cancers— as

an g e l a r o w l in g s A P P h o t o (r a c e r s) ; Lisa A p fe l bac h er (a l l p i l l s) ; time l ine : w e l l c o me l ibrar y, l o nd o n (B e a t s o n) ;


apy to shrink their tumors to operable size, after many as 75 percent— contain estrogen receptors or
which surgery is performed, followed by additional progesterone receptors, or both, causing these mol-
chemotherapy and radiation therapy. Using this ecules to quickly become therapeutic targets.
sandwich approach over the past three decades, spe- The antiestrogen drug tamoxifen was first ap-
cialized teams of doctors, nurses and other health proved in the U.S. in 1977 to treat advanced breast
professionals have greatly enhanced the cure rate for cancer in postmenopausal women. The drug mole-
these patients. Even those whose breast tumor has cule binds to the estrogen receptor, preventing estro-
already metastasized to other organs now have ac- gen from doing so. Tamoxifen has since proved ef-
cess to novel therapies that prolong their survival and fective for patients with localized breast tumors that
display estrogen or progesterone receptors and as a
preventive therapy in healthy women who are at
high risk for breast cancer. Meanwhile newer drugs
that inhibit the aromatase enzyme, suppressing nat-
ural estrogen manufacture in the body, have proved
superior to tamoxifen in postmenopausal women.
In a sense, then, estrogen and progesterone re-
ceptors became the first molecular features of tu-
mors that could be directly targeted by drugs, al-
though an important distinction should be noted
between these targets and newer ones identified in
the past decade. The sex steroid receptors promote
cell proliferation, or growth, in healthy tissues as
well as in tumors, so suppressing their ability to
transmit growth signals does help to check tumor

R AISING AWARENESS of the importance of early detection, as


well as raising funding for research, has paid off in notable
declines in breast cancer mortality in the developed world.

90 SCIENTIFIC A MERIC A N ne w ans w ers f o r C ancer


© 20 08 SCIENTIFIC AMERIC AN, INC.
1950s–1960s 1970s–1980s 1990s–2008
1951: Estrogen and testosterone 1976: Cancer- 1994: BRCA1 gene, known to increase susceptibility
found to drive the growth of promoting to breast cancer, is isolated.
breast and prostate cancers, “oncogenes” Tamoxifen, a 1998: Trastuzumab, the first molecular-targeted cancer
respectively. first discovered selective estrogen therapy, approved in U.S. for use in breast cancer.
1958: Cancer researchers identify in mammals. receptor modulator
2007: Lapatinib, an inhibitor of growth signaling,
additional “growth factor” proteins 1976: Clinical trials begin to show that lumpectomy approved in U.S. for use in breast cancer.
that help tumors thrive. with radiation can be as effective as mastectomy. 2008: Bevacizumab, an inhibitor of blood
1966: Estrogen 1977: Hormone-blocking drug tamoxifen approved vessel formation in tumors, approved in U.S.
receptor identified. in U.S. for treatment of breast cancers sensitive to for use in breast cancer.
estrogen or progesterone.
Estrogen Trastuzumab, also known by the
receptor protein 1988: Trial results show that preoperative chemotherapy trade name Herceptin
structure shrinks tumors, making less invasive surgeries possible.

enlargement. And changes to the receptors’ shape Targeting HER2


or function may sometimes contribute to the gen- i n t h e e a r ly 198 0 s the gene that gives rise to
eral malignant characteristics of tumor cells. But HER2 was first discovered in mutated form in rat
the gene encoding the estrogen receptor is rarely neural tumors by investigators at the Massachu-
mutated in breast cancer, which means that it is not setts Institute of Technology, who named that on-
a true cancer-causing gene. cogene Neu. Soon researchers realized that the gene
Perhaps the most important realization in can- was a mammalian version of one previously identi-
cer research since the era when sex hormone recep- fied in viruses called ERBB, so Neu also came to be
tors were discovered is that particular genes, when known as ERBB2. This gene was not done accumu-
they become mutated, can cause a normal cell to lating names, however. When scientists identified
turn cancerous. Such genes, once they do mutate, the protein encoded by ERBB2, they realized that
are referred to as oncogenes, and they are believed it was closely related to a cell-membrane protein
to be responsible both for initiating the transfor- called epidermal growth factor receptor (EGFR).
mation of a normal cell into a cancerous one and Thus, when they finally isolated the human version
for driving tumor growth. That is why breast can- of the ERBB2 gene, they named it human epider-
cer (like all cancers) is described today as funda- mal growth factor receptor 2 (HER2).
mentally a disease of genes. An oncogenic muta- As it turns out, the entire EGFR family of pro-
tion, such as a small change in the DNA nucleotide teins has proved important to tumor cell growth in
sequence of a gene, might disable a protective gene a variety of cancers. When activated by specific mol-
or boost the activity of a tumor-promoting one. In ecules that bind to them (their ligands), such recep-
some cases, entire genes are deleted or duplicated tors transmit a proliferation signal to the cell by ini-
[see sidebar on page 93]. tiating a cascade of internal molecular interactions—
Tumors can now be classified according to the spurring activity by genes whose encoded proteins
genes that are overactive or suppressed in their cells regulate the activity of still more “downstream”
and according to the resulting changes in the man- genes. Shortly after the HER2 gene was discovered,
ufacture and function of proteins encoded by those scientists noted that it was frequently duplicated in
genes. The damaged genes can vary from tumor to breast cancer cells and that having multiple copies
tumor, and this heterogeneity at the genetic level of the gene was associated with a poor prognosis.
explains why breast cancers in individual patients Laboratory studies confirmed that adding cop-
might behave differently. Some cancers have lim- ies of the HER2 gene to a normal cell could trans-
ited invasiveness and metastatic potential, for in-
stance, whereas others spread quickly to distant Francisco J. Esteva is director of the Breast Cancer Translational Research
the authors

organs. Knowing the molecular profile of a pa- Laboratory at the University of Texas M. D. Anderson Cancer Center, where
tient’s tumor should permit a doctor to focus on Gabriel N. Hortobagyi is director of the Multidisciplinary Breast Cancer
inhibiting the mechanisms driving that particular Research Program. Esteva, an associate professor in the departments of
tumor, one day choosing from an arsenal of drugs breast medical oncology and molecular and cellular oncology, focuses on
a set that will interfere with the specific molecules bridging the gap between basic research and patient treatment. Hortobagyi,
involved in the initiation, growth and spread of the a recent president of the American Society of Clinical Oncology, is also chair
cancer. The success of trastuzumab and other of, and a professor in, the department of breast medical oncology. His re-
HER2-targeted therapies illustrates the potential search spans all aspects of breast cancer biology, management and thera-
of this approach in combating breast cancer. peutics and has earned numerous awards.

w w w. S c i A m . c o m SCIENTIFIC A MERIC A N 91
© 20 08 SCIENTIFIC AMERIC AN, INC.
form it into a cancer cell— a hallmark ability of on- act as growth-stimulating factors, and inhibiting
cogenes. Because 20 percent of breast cancer tu- them directly is another way of squelching EGFR-
mors overproduce the HER2 protein, it became a mediated growth signaling in cells. That is why
therapeutic target for drug researchers. Genentech pharmaceutical companies are avidly pursuing the
scientists created trastuzumab in the late 1980s by clinical development of such drugs. Lapatinib (Ty­
manufacturing so-called monoclonal antibodies kerb) is a dual EGFR/HER2 tyro­sine kinase inhib-
that bind to the HER2 receptor, preventing it from itor that has shown remarkable laboratory results,
being activated. In clinical trials, it was found that leading to growth arrest and cell suicide in breast
trastuzumab could lengthen the survival of patients cancer cell lines that overproduce HER2.
with both early-stage and metastatic breast cancer. One way to improve the effectiveness of HER2-
The success of trastuzumab has led to the devel- targeted therapy is, therefore, to combine a drug such
opment of similar antibody-based therapies, such as trastuzumab with a tyrosine kinase inhibitor such
as pertuzumab, which binds to HER2 at a different as lapatinib. In breast cancer cell lines, that combina-
site than trastuzumab and has the added effect of tion produces greater synergistic growth inhibition
preventing the receptor from interacting with other and higher rates of cell suicide. Even in cell lines that
members of its family in the cell membrane, such as have developed resistance to trastuzumab after long-
EGFR and HER3. Blocking such interactions re- term treatment, lapatinib has proved just as effec-
duces growth signaling along the intracellular path- tive at inducing cell suicide. A recent large (phase
ways of molecular communication downstream of III) clinical trial among patients with HER2-over-
these receptors. Pertuzumab can even disrupt cer- producing metastatic breast cancer, whose disease
tain types of HER2 activation in tumor cells that had become resistant to trastuzumab, demonstrat-
have become resistant to trastuzumab. Moreover, ed that lapatinib plus capecitabine chemotherapy

j en c h ristiansen
we have shown that combining trastuzumab with doubled the median time to progression as com-
pertuzumab can boost the rate of cell death in pared with capecitabine alone. On the basis of these
breast cancer cells overproducing HER2. results, in 2007 the U.S. Food and Drug Adminis-
Still another method of wielding antibodies tration approved the use of lapatinib combined with
against the HER2 recep-
tor is to attach a potent tumor Pathways
toxin to them, which the
antibodies then transport Excessive activity of certain proteins and genes in
ll
into the cancer cell. After breast cancer cells can switch on a series of ce
c er
molecular interactions, or pathways, that Growth an
the toxin-antibody pair factor tc
encourage the cells’ proliferation (growth) e as
is internalized by a cell, Br
and survival. Among these proteins are
the toxin detaches and various cell-surface receptors, such as IGF-1R
kills the cell. This ap- HER2 (a member of the epidermal
proach has been success- growth factor receptor, or EGFR,
ful in other types of can- family) and the insulinlike growth Signaling
HER2
cer, such as acute myeloid factor 1 receptor (IGF-1R). pathway
leukemia, and clinical The On Switch EGFR
trials are under way in When stimulated by a
Growth
patients with metastatic growth factor, an individual factor
receptor protein joins, or
breast cancer to deter- dimerizes, with a neighbor.
mine the safety and effi- The pairing causes an Dimerized
cacy of such trastuzumab- enzyme called a tyrosine pair
kinase to add a phosphate
­based conjugates. group to both units, Phosphorylation
To send a growth sig- setting off a signal to the
nal into a cell, the intra­ cell nucleus. Those
receptors are thus Tyrosine kinase
cellular region of the important targets for
EGFR family of proteins antitumor therapies
Estrogen
must first be acted on by (box at far right). Estrogen receptor
tyrosine kinases, enzymes Cell membrane Like other growth factor receptors, estrogen
that chemically modify a receptor proteins dimerize when activated
segment known as the ty- by estrogen. The receptor pair then acts
directly on DNA (right) to switch on genes
rosine kinase domain. Ty-
CREDIT

involved in cell growth and survival.


rosine kinases can thus

92 S C I E N T I F I C A M E R I C A N ne w ans w ers f o r C ancer


© 20 08 SCIENTIFIC AMERIC AN, INC.
capecitabine for treating metastatic disease. Clini- to commit suicide. We have even seen
cal trials to determine lapatinib’s value as an adju- cells lose or disable the extracellular
vant treatment in a wider variety of circumstances binding site that trastuzumab attaches
are ongoing, as are trials of several other tyrosine to on the HER2 receptor.
kinase inhibitors that target HER2 and EGFR. In light of these observations, iden-
Finding alternative means of interfering with tifying additional molecular targets to
the same growth pathways is important because, as attack in cells that overproduce HER2,
can happen with trastuzumab, cancer cells often do as well as targets in the other 80 percent
eventually find ways to evade individual drugs. of tumors that do not have HER2 muta-
Also under way is research into how and why cancer tions, is a high research priority.
cells develop resistance to trastuzumab, so that in-
vestigators can use those insights as a guide to de- Expanding the Arsenal her2 amplification
A gene encoding the HER2
signing more effective combinations or new agents a m o n g t h e m o s t promising new
growth factor receptor is
for patients whose tumors overproduce HER2. targets for breast cancer therapy is the tagged with red fluorescence
In studies of cell cultures and animals, for in- IGF-1 receptor as well as the growth in breast cancer cells (top). In
stance, our laboratory has discovered that cancer hormone molecules that activate it, HER2-positive cancer cells
cells employ many different mechanisms to survive IGF-1 and IGF-2. High levels of IGF-1 (bottom), the gene is
duplicated many times over,
in the presence of trastuzumab, including increas- in the bloodstream have been linked
leading to overproduction of
c o u rtes y o f francisc o j . este v a

ing their production of other growth factor recep- with increased risk of breast cancer, and HER2 proteins that causes
tors, either from the EGFR/HER family or from many laboratory and clinical studies the cells to receive excessive
other families, such as the insulinlike growth factor have implicated its receptor in the devel- growth signals.
1 (IGF-1) receptor. The cells may also lose or inac- opment, maintenance and progression
tivate the tumor suppressor gene PTEN. This gene of multiple cancer types. Signaling by
normally blocks a survival pathway involving the the IGF-1 receptor regulates a variety of cellular pro-
enzyme phosphotidylinositol 3-kinase (PI3K), which cesses, including growth, motility and protection
allows damaged cells to ignore signals telling them from cell suicide. In fact, such signals have been

Attacking Tumor Proteins


Nucleus Drugs designed to inhibit growth factor
receptors on tumor cells can act in a variety
of ways, including inducing the receptors
to deliver toxins into the cells.

The Effects Growth Drug


The signals activate specific genes, such as RAS, which give rise to proteins that factor
act on additional genes. This series of gene-protein interactions, depicted in Attack
abbreviated form here, leads to cell growth and the suppression of mechanisms growth
that would usually cause an abnormal cell to commit suicide. Mutations in any factor
gene along such signaling pathways can produce similar results, making those
genes and their encoded proteins therapeutic targets as well.
Block growth
Gene factor binding site
activation Cell
RAS raf mek mapk growth

Inhibit
Cell dimerization
pI3k akt mTOR survival

Block
tyrosine
DNA kinase
activity
Toxin
Deliver toxin
into cell
CREDIT

w w w. S c i A m . c o m SCIENTIFIC A MERIC A N 93
© 20 08 SCIENTIFIC AMERIC AN, INC.
shown to protect tumor cells from the effects of che- various other growth factor receptors — including
motherapy and radiation therapy. Conversely, in- estrogen, HER2 and additional EGFRs — is a key
hibiting IGF-1 receptor activity during radiation mechanism for the growth and survival of breast
therapy or chemotherapy has been found to enhance cancer. This codependence and communication be-
tumor cell suicide rates in animal studies. tween different intracellular pathways is thought to
In addition to exploring IGF-1 receptor inhibi- play an important role in drug resistance. Our re-
tion as a direct therapeutic tool for breast cancer, search group has shown, for example, that blocking
scientists are evaluating ways to apply it toward pre- the IGF-1 receptor with a monoclonal antibody re-
venting or reversing resistance to other treatments, stores the sensitivity of resistant cells to trastuzu­
such as endocrine therapies, trastuzumab and lapa- mab and disrupts the interaction between the IGF-1
tinib. Cross talk between the IGF-1 receptor and the and HER2 receptors. Suppressing the IGF-1 recep-
tor also kills the resistant cells. Furthermore,
targeted therapies lapatinib appears to have inhibitory effects
on IGF-1 signaling in the trastuzumab-resis-
A growing list of drugs Target drug
designed to inhibit tant cells, suggesting that its ability to limit
Estrogen/ ● A nastrozole
specific tumor pro- tumor cell proliferation may result not only
progesterone ● L etrozole
teins are approved receptor proteins from its anti-EGFR/HER2 activities but also
●E  xemestane
to treat breast cancer from direct IGF-1 receptor inhibition.
● T amoxifen
patients (bold) or are The tangle of signaling pathways leading
undergoing clinical ● F ulvestrant
from the receptors we have been describing
trials. HER2 receptor ● T rastuzumab to the cellular processes that actually cause a
protein ● Pertuzumab cell to divide or to resist suicide despite DNA
Drug type ● L apatinib damage is highly complex. But scientists are
●N  euVax finding that key genes along those pathways
●A romatase
inhibitor ●d  HER2 are also frequently mutated or dysregulated
(blocks an enzyme ● MV
 F-HER-2 in tumor cells. Among the best characterized
involved in estrogen ●E  1A (gene therapy) examples is the PI3K gene, whose encoded
and progesterone protein chemically modifies another protein
synthesis) IGF-1 receptor protein ● IMC
 -A12
● C P-751, 871
known as AKT, which in turn modifies a
●M onoclonal
complex called the mammalian target of ra-
antibody ● AM
 G 479
(impedes activation ●h  7C10
pamycin (mTOR). This PI3K/AKT/mTOR
of cellular receptors) ● OSI
 -906 pathway plays a critical role in the body’s use
of glucose for energy and other important
●K inase inhibitor PI3K/AKT/mTOR cell ●B  GT226
(inhibits signaling physiological processes in normal cells, but it
survival pathway ●B  EZ235A
by cellular receptors) is pathologically overactivated in cancer cells,
● RAD 001
● Vaccine


prolonging their survival. Because the path-
●R  apamycin
(stimulates way’s effects are ubiquitous in the body, deliv-
production of VEGF receptor protein ● B  evacizumab ering drugs that inhibit it could disrupt healthy
antibodies specific (involved ● S unitinib cells as well as cancerous ones— a drawback
to tumor proteins; in forming tumor
can be composed ● V atalinib that has so far limited the use of such agents.
blood vessels)
of cells or peptide ●P  azopanib Several mTOR inhibitors are nonetheless
molecules) ● A ZD2171 being tested in clinical trials, both as single
●O ther ● AM
 G706 agents and combined with other therapies. At
(includes direct ● AM
 G386 the moment, studies using the mTOR-sup-
inhibitors of other ●P  TC299 pressing antibiotic rapamycin, along with an
molecules or
gene therapy to inhibitor of the IGF-1 receptor, suggest that
Other targets ●D  asatinib (SRC inhibitor)
alter cellular protein such combinations yield additive antitumor
● T HERATOPE
manufacture) effects as compared with single agents.
●D  endritic cell vaccines
Another approach showing great prom-
●P  53 peptide vaccine
ise is combining direct antitumor agents with
● ALT
 801 (p53 inhibitor)
compounds that target elements in a tumor’s
● A d5CMV-p53 (gene therapy)
environment. Cancers secrete a variety of
● A nti-p53 T cell reinfusion
growth factors to attract the endothelial cells
● A ZD2281 (PARP protein inhibitor)
that build new blood vessels in a process
●B  SI-201 (PARP inhibitor)
called angiogenesis. Overproduction of the

94 SCIENTIFIC A MERIC A N ne w ans w ers f o r C ancer


© 20 08 SCIENTIFIC AMERIC AN, INC.
Global Power
T argeted therapies will be most powerful, in
principle, when they are used together in
combinations tailored to the tumor features
continuing to turn up new molecular targets that
both reveal more about the underlying mechanisms
of cancer and provide potential leads for drug de-
driving an individual patient’s cancer. Clinical trials to velopment. Terumi Kohwi-Shigematsu of Law-
test specific drug combinations provide critical rence Berkeley National Laboratory and her col-
information about which treatments work most effectively
leagues announced one such discovery earlier this
on different tumor profiles and reveal unexpected
year. They identified a gene called SATB1 as the “mas-
interactions between drugs. But trials take time, often
years, to enroll a sufficient number of participants to ter regulator” of activity for more than 1,000 genes
generate statistically significant results. That is why involved in breast cancer metastasis. Kohwi-Shige-
multinational research consortia based in Europe and the matsu showed that the influence of the SATB1 protein
U.S. are pooling resources to conduct a 50-country trial, encoded by the gene is both necessary and sufficient
the Adjuvant Lapatinib and/or Trastuzumab Treatment for breast cancer cells to become metastatic, which
Optimization Study (ALTTO), which has just begun makes it an appealing therapeutic candidate.
recruiting in the U.S. Progress in the molecular targeting of breast can-
Some 1,500 testing sites will treat patients with early cer and individualized therapy will generally rely on
(stage I or II) breast cancers that overproduce the HER2 the continuing development of profiling tools to de-
protein, giving them chemotherapy and either trastuzu­
termine whether a patient’s tumor overproduces pro-
mab or lapatinib alone, or one of those drugs followed by
teins such as HER2, SATB1 and others that might
the other, or both drugs together. The trial will provide the
first side-by-side comparison of these HER2-targeted be direct drug targets. In addition, genetic testing
treatments that work by different mechanisms. can help characterize a tumor’s overall gene activity
With a goal of including as many as 8,000 women on six patterns— a potential signature of a good or poor
continents, ALTTO has the potential to quickly generate prognosis. Still other tests already available or near-
results that can then be applied to patients everywhere. ing approval can help profile the patient herself to
Moreover, this global data-sharing model can highlight establish whether she has genetic variations that
differences in treatment responses or toxicity among might make her body process a medication more
different ethnic groups, a phenomenon observed with slowly than average — a situation that can be prob-
certain types of chemotherapy because of genetic lematic with a drug such as tamoxifen that depends
variations that affect the way the drugs are metabolized
on the body to convert it to active form.
by patients’ bodies. Having such information about the
Meanwhile further clinical trials of various
newer targeted therapies will help doctors to further
personalize treatment, tailoring it to both the tumor drug combinations are needed to validate the ef-
and the patient.  —The Editors fectiveness of multipronged attacks. A 50-country
trial has recently begun recruitment in the U.S., for
example, to test lapatinib and trastuzumab alone
most important of these, vascular endothelial and in combination with each other and with tra-
growth factor (VEGF), is thought to make tumors ditional chemotherapies [see box on this page].
more dangerous, and high levels correlate with Such a large trial exemplifies the considerable
worse survival rates in human invasive breast can- resources and attention focused on breast cancer re-
cers. Genentech’s bevacizumab (Avastin) is a mono- search, in recognition of its importance as a global
clonal antibody directed against VEGF that was health threat. Doctors’ ability to profile a tumor and
first approved for use in colon cancer in 2004. In tailor treatment to fight it with a growing arsenal of
more recent clinical trials among patients with heav- weapons is already making a difference in the sur-
ily treated metastatic breast cancer, bevacizumab vival rates of patients, and the coming decade prom-
alone had limited activity, but certain patients who ises even more dramatic progress.
received it in combination with capecitabine chemo-
therapy showed improved responses. In another more to explore
study, HER2-negative metastatic breast cancer Molecular Oncology of Breast Cancer. Edited by Jeffrey S. Ross and Gabriel N.
progressed more slowly in patients who received Hortobagyi. Jones and Bartlett Publishers, 2005.
paclitaxel chemotherapy with bevacizumab than it Long-Term Cardiac Tolerability of Trastuzumab in Metastatic Breast Cancer:
did in patients who received paclitaxel alone. Based The M.D. Anderson Cancer Center Experience. Valentina Guarneri et al. in Journal
of Clinical Oncology, Vol. 24, No. 25, pages 4107–4115; September 1, 2006.
on such results, bevacizumab was recently approved
Trastuzumab: Triumphs and Tribulations. Rita Nahta and Francisco J. Esteva in
for use in breast cancer patients, and other VEGF Oncogene, Vol. 26, No. 25, pages 3637–3643; May 28, 2007.
inhibitors are also in development, such as Pfizer’s Advances in the Treatment of Breast Cancer. Stacy Moulder and Gabriel N. Hortobagyi
sunitinib (Sutent), a tyrosine kinase inhibitor tar- in Clinical Pharmacology & Therapeutics, Vol. 83, No. 1, pages 26–36; January 2008.
geted against the VEGF receptor. Breast Cancer. Second edition. Edited by Kelly K. Hunt, Geoffrey L. Robb, Eric A.
At the same time, very basic biology research is Strom and Naoto T. Ueno. Springer, 2008.

96 SCIENTIFIC A MERIC A N ne w ans w ers f o r C ancer


© 20 08 SCIENTIFIC AMERIC AN, INC.

Você também pode gostar