Você está na página 1de 7

Int J Gynecol Cancer 2003, 13, 633639

Anemia before and during concurrent


chemoradiotherapy in patients with cervical
carcinoma: Effect on progression-free survival
A. OBERMAIR*,, R. CHEUK, K. HORWOOD, M. NEUDORFER, M. JANDA, G. GIANNIS,
J.L. NICKLIN*, L.C. PERRIN* & A.J. CRANDON*
*Queensland Center for Gynaecological Cancer, Royal Womens Hospital; Queensland Radium Institute; and Division of
Oncology, Royal Brisbane Hospital; Queensland University of Technology, School of Public Health, Brisbane, Australia;
and University Hospital Vienna, Department of Gynecology and Obstetrics, Vienna, Austria

Abstract. Obermair A, Cheuk R, Horwood K, Neudorfer M, Janda M,


Nicklin JL, Perrin LC, Crandon AJ. Anemia before and during
concurrent chemoradiotherapy in patients with cervical carcinoma: Effect
on progression-free survival. Int J Gynecol Cancer 2003;13:633639.

To determine the impact of anemia before and during chemoradiation in


patients with cervical cancer, we collected data on hemoglobin (Hb)
levels before and during treatment from 60 unselected patients with
cervical carcinoma. All patients had FIGO stage IB to IVA disease and
were treated with concurrent chemoradiation for the aim of cure.
Patients with an Hb value below or equal to the lower 25th quartile
were considered anemic. Progression-free survival (PFS) was evaluated
by univariate and multivariate analyses. After a median follow-up of
26.3 months, 20 patients developed disease progression. The lowest Hb
during chemoradiation (nadir Hb), the stage of disease, and parametrial
involvement were correlated significantly with PFS. On multivariate
analysis, the nadir Hb (relative risk [RR] 0.29) and tumor stage (RR
3.4) remained the only prognostically relevant factors predicting PFS.
At 60 months the PFS was 39.1% for anemic patients and 48.0% for
nonanemic patients (P < 0.0002). In patients undergoing chemoradiation
for cervical carcinoma, a low nadir Hb is highly predictive of shortened
PFS, whereas the Hb before treatment is prognostically not significant.

KEYWORDS: anemia, chemoradiotherapy, cervical cancer, prognostic factors,


treatment.

Tumour-associated anemia is a common finding in cer- impaired prognosis for anemic compared to nonanemic
vical cancer patients in the US(1). More than one-third of cervical cancer patients who undergo radiotherapy. It has
the patients have a hemoglobin level of 120 g/l at pre- been argued that anemia might be a surrogate marker for
sentation(1). The causes for tumor-associated anemia are tissue hypoxia and thus mediate resistance to radiother-
not fully understood(2,3). A number of studies suggest an apy rather than anemia representing a biologically more
aggressive tumor(1,46).
Until 1999, radiotherapy was the standard of care
for patients with locally advanced cervical cancer.
Address correspondence and reprint requests to: Andreas Therefore, studies which examined the impact of anemia
Obermair, MD, Queensland Center for Gynaecological
Cancer, Royal Womens Hospital, Brisbane, Australia. Email:
were based on patients who were treated with
andreas_obermair@health.qld.gov.au radiotherapy as a single modality treatment. Recently,

# 2003 IGCS
634 A. Obermair et al.

five prospective randomized trials reported a reduc- apy. Fourteen patients (23.3%) had a pelvic/aortic
tion of 24% to 51% for the risk of death for patients lymph node debulking prior to chemoradiation. Radical
treated with concurrent chemoradiation when com- pelvic radiotherapy was given with the aim of cure in all
pared with conventional radical radiotherapy(711). patients. The pelvis was treated by external beam radio-
As a consequence, chemoradiation has replaced therapy with a linear accelerator using minimum
radiotherapy as the current standard of treatment photon-beam energy of 4 MV within a standard four-
for locally advanced cervical carcinoma(12). field box. Patients with stage IB, IIA, and IIB cervical
Until recently, no data were available about the carcinoma received a dose of 4045 Gy at the ICRU
significance of anemia in patients undergoing concur- (International Commission of Radiation Units) reference
rent chemoradiation for cervical carcinoma. In our point external beam radiotherapy given in 2528
previous study the nadir hemoglobin rather than the fractions over 55 weeks. Patients with stage III and
hemoglobin at presentation was highly predictive of IVA cervical carcinoma received a dose of 52.5 Gy as
the response to treatment in patients with carcinoma external beam radiotherapy given in 30 fractions over 6
of the cervix who received radical chemoradiation(13). weeks. External beam radiotherapy was followed by
Because of the short follow-up, no data regarding the intracavitary low-dose rate brachytherapy up to 2
prognostic effect of anemia could be provided. There- weeks after completion of external beam radiotherapy.
fore, in the present study we aimed to describe a possible Patients with stage IB, IIA, and IIB cervical carcinoma
relationship between hemoglobin levels before and received 50 Gy to point A in one or two insertions.
during treatment and the prognosis in patients under- Patients with stage III and IVA cervical carcinoma
going chemoradiation for cervical cancer. received 25 Gy to point A in one insertion. Due to poor
compliance with treatment, one patient did not receive
brachytherapy.
Patients and methods Chemotherapy consisted of cisplatin alone (n 54),
cisplatin combined with vinblastine (n 1), cisplatin
Patients
combined with 5-fluorouracil (n 3), or 5-fluorouracil
Sixty two patients with histologically proven cervical as single agent (n 2). Chemotherapy started at the
carcinoma underwent definitive concurrent chemoradia- commencement of radiotherapy (day 1) and was
tion at the Queensland Radium Institute in Brisbane, given every 7 days thereafter. One liter of normal
Australia between January 1996 and March 2001. We saline was given both before and after cisplatin. Pre-
excluded two patients because they underwent a comple- medication included ondansetron and dexametha-
tion hysterectomy after radiotherapy. For the aim of this sone. The median dose of cisplatin was 40 mg per
analysis, patients had to have: histologically proven square meter of body-surface area per week (range,
cervical carcinoma, stage IB to IVA disease; radical radio- 3050 mg/m2). Generally the total dose of cisplatin
therapy with concurrent chemotherapy; and no previous did not to exceed 70 mg per week. Two patients had
radical or standard hysterectomy. Patients with stage IVB single agent 5-fluorouracil at a dose of 800 mg/m2 per
disease were not included into this series because the aim day intravenously (iv) continuous for 96 h and in
of treatment was palliation rather than cure in those another two patients 5-fluorouracil was given at a
patients. Histologic grading was based on the proportion dose of 800 mg/m2 per day i.v. continuous for 96 h
of the tumor that was undergoing keratinization with the combined with 80 mg/m2 of cisplatin on day 1. This
formation of squamous pearls and the number of mito- schedule was repeated after 4 weeks. One patient had
sis(14). Tumors were graded as well differentiated (grade cisplatin 40 mg/m2 combined with vinblastine at a
1), moderately differentiated (grade 2), or poorly differ- dose of 4 mg/m2. The median number of cycles was
entiated (grade 3). The stage of disease was determined five (range, four to six). Treatment with cisplatin was
clinically according to the 1995 FIGO staging system(15). withheld if neutrophil count was less than 1.0109
per liter or if the platelet count was less than 80109
per liter. During treatment patients were seen weekly
Treatment
and a full blood count and serum electrolytes, liver
Before commencement of chemoradiotherapy, all function tests, and renal function tests were taken.
patients underwent a CT scan of the pelvis and the abdo-
men and an X-ray of the chest. Patients with bulky pelvic
Determination of anemia
lymph nodes had a extraperitoneal lymph node debulk-
ing prior to chemoradiation. All patients received radical A full blood count was determined by venous punc-
radiotherapy to the pelvis and concurrent chemother- ture before commencing radiotherapy and then every
# 2003 IGCS, International Journal of Gynecological Cancer 13, 633639
Anemia before and during concurrent chemotherapy in cervical cancer patients 635

7 days thereafter. We recorded the hemoglobin before (25th) quartile was used as the cut-of level for both
treatment and the lowest hemoglobin level which was the initial and the nadir hemoglobin levels. Patients
observed throughout chemoradiation (nadir Hb). with values below or equal this value were considered
Hemoglobin values are given in gram per liter102 anemic. Univariate and multivariate Cox models
(g/dl). Seventeen patients (28.3%) received at least were calculated in order to identify independent
one blood transfusion during chemoradiation and prognostic factors for PFS. We did not include the
the median number of blood transfusions was four histologic type as a covariate in the Cox model
(range, two to nine) with a median trigger Hb of because only five of 62 patients had a histologic type
8.6 g/dL (range, 6.3 g/dl to 10.6 g/dl). Two, four, other than squamous cell carcinoma (SCC). Univariate
eight, and three patients received their first blood analysis of PFS was performed as outlined by Kaplan
transfusion at chemotherapy cycle numbers 1, 2, 3, and Meier(22) and survival curves were compared using
or 4, respectively. There was no uniform policy on the log-rank test(23). P-values are the result of two-sided
blood transfusion and no patient received erythropoietin tests and a P-value <0.05 was considered to indicate a
prior or during chemoradiation. Patients were considered statistically significant difference. Statistical analysis
smokers if they regularly smoked more than five was done using SPSS (Version 9.0, SPSS, INC, Chicago,
cigarettes per day at the time of diagnosis. IL).

Statistical methods Results


The primary statistical endpoint was progression-free Analysis is based on the data of 60 patients with a
survival (PFS), which was defined as the period from mean age of 48.3 years (range, 2977 years). The
the commencement of chemoradiation until the date distribution of patients to stages IB, IIA, IIB, III, and
of first recurrence. Data on patients who were tumor- IVA was 13, 3, 24, 18, and 2 patients, respectively. The
free were censored at the last follow-up visit. By using median tumor size (25th quartile, 75th quartile) was
the ANOVA test, mean hemoglobin levels were 5 cm (4 cm, 6 cm). Data on the grade of differentiation
compared among various clinical and histopathologic were available in 53 patients but not available in
parameters (patients characteristics). Hemoglobin seven patients. Further patients characteristics are
values were entered in the ANOVA test as a continuous given in Table 1.
variable. Hemoglobin values were dichotomized. In Forty-seven patients (75.3%) had a complete clinical
the absence of an established cut-off level the lower response to chemoradiation, 10 patients (16.7%) had

Table 1. Patients characteristics


Hb presentation Nadir Hb
n (mean  SD) P-value (mean  SD) P-value
Stage
IB, IIA 16 13.5  1.5 10.8  1.9
IIB 24 12.8  1.2 10.9  1.2
III, IVA 20 12.1  1.2 0.007 10.6  1.8 0.287
Histologic type
SCC 56 12.7  1.4 10.6  1.6
Adenocarcinoma, adenosquamous carcinoma 4 13.2  0.9 0.479 10.7  2.3 0.895
Grading
Well/moderately 28 12.3  1.5 10.8  1.8
Poorly 25 12.8  1.1 0.183 10.4  1.3 0.437
Parametrial involvement
None 14 13.4  1.6 10.7  1.8
Unilateral 27 12.6  1.3 10.7  1.4
Bilateral 19 12.3  1.2 0.068 10.3  1.7 0.695
Lymph nodes
Negative 46 12.7  1.3 10.4  1.7
Positive 14 12.7  1.8 0.919 11.1  1.2 0.221
Tumor size
5 cm 34 13.0  1.5 11.0  1.4
>5 cm 26 12.3  1.1 0.064 10.0  1.6 0.008

# 2003 IGCS, International Journal of Gynecological Cancer 13, 633639


636 A. Obermair et al.

a partial response, and two patients (3.3%) did not Nine of 14 patients who were anemic before the
respond to treatment. In one patient we were commencement of treatment experienced anemia dur-
unable to determine the response to chemoradiation. ing treatment, whereas five patients remained non-
The mean PFS for patients who recurred was 15.0 anemic. Forty-five patients were considered not
months (25% to 75%, 6.022.6). After a median anemic before treatment and only seven of those
follow-up of 26.3 months (25% to 75%, 18.326.3 patients developed anemia during chemoradiotherapy
months), 20 patients developed recurrent disease. (Chi-square 12.82, P 0.001).
The site of the first recurrence was the pelvis in 13
patients, six patients developed metastases at distant
Association of hemoglobin levels with PFS
sites (supraclavicular lymph nodes, n 1; lung, n 2;
umbilicus, n 1), and one patient developed a central Univariate Cox models revealed that stage, nadir Hb,
local recurrence. At the time of statistical analysis, 34 and parametrial involvement were determinants of
patients were alive without any evidence of disease, PFS, whereas the Hb at presentation, tumor size, histo-
nine patients were alive with disease, 12 patients had logic grade, lymph node involvement, and cigarette
died of disease, and three patients had died of causes smoking were not (Table 2). A multivariate Cox model
other than cervical cancer (biliary sepsis due to a including the covariates stage, nadir Hb, and parame-
calculus in the biliary duct, n 1; small bowel trial involvement demonstrated an independent prog-
obstruction due to radiation changes, n 1). Two nostic effect on PFS for stage and nadir Hb but not for
patients were lost for follow-up after 10 and 21 parametrial involvement (Table 3).
months, respectively. The overall survival probability At 60 months the PFS was zero and 66.6% for
was 78.7% at 5 years for all patients. patients who started chemotherapy with an initial Hb
of 11.6 and who were therefore considered anemic
and for nonanemic patients, respectively (log-rank
P 0.147) (Fig. 1A). At 60 months the PFS was 39.1%
Association of hemoglobin levels with clinical and and 48.0% for patients who had a nadir Hb of 9.4 g/dl
histopathologic features or less during chemoradiation compared to patients
with a lowest recordable Hb of >9.4 g/dl throughout
For all patients the mean Hb (SD) at presentation
chemoradiation (Log-rank P 0.005) (Fig. 1B).
and the mean nadir Hb (SD) were 12.8 g/dl (1.4 g/
Within the follow-up period, local control could be
dl) and 10.6 g/dl (1.6 g/dl), respectively. Patients
maintained in 45 patients (75.0%). The probability of
with lower tumor stage presented with higher Hb
local control at 60 months was 41.9% and 81.3% for
values before chemoradiation (P 0.007) compared
patients who had a nadir Hb of 9.4 g/dl or less during
to patients with higher tumor stages. Patients with
chemoradiation compared to patients with a lowest
tumors measuring larger than 5 cm had lower nadir
recordable Hb of >9.4 g/dl throughout chemoradia-
Hb values throughout treatment (P 0.013) than
tion (Log-rank P 0.002).
patients with smaller tumors. No other statistically
significant association of the initial Hb or nadir Hb
with stage, tumor size, parametrial or lymph node Discussion
involvement, histologic grading, or histologic type Disease progression was significantly more likely in
was found (Table 1). patients who became anemic during chemoradiation

Table 2. Univariate Cox models with regard to disease-free survival in 60 patients with concurrent chemoradiation for cervical
carcinomaa
Covariates OR 95% CI P-value
Stage (IB, IIA vs. IIIB vs. III, IVA) 3.4 1.67.0 0.001
Hb nadir 0.3 0.10.7 0.009
Size (5 cm vs. >5 cm) 1.6 0.63.9 0.299
Parametrial involvement (nil vs. unilateral vs. bilateral) 1.9 1.03.7 0.037
Hb at presentation (continuous) 0.5 0.21.3 0.155
Grading (g1,g2 vs. g3) 0.9 0.81.0 0.323
Lymph nodes (negative vs. positive) 0.7 0.22.4 0.637
Cigarette smoking (yes vs. no) 1.01 0.42.5 0.967
a
Abbreviations: OR, odds ratio; 95% CI, 95% Confidence interval;

# 2003 IGCS, International Journal of Gynecological Cancer 13, 633639


Anemia before and during concurrent chemotherapy in cervical cancer patients 637

Table 3. Multivariate Cox model (nonstepwise) with regard to disease-free survival in 60 patients with concurrent chemoradiation
for cervical carcinomaa
Covariates OR 95% CI P-value
Stage (IB, IIA vs. IIB vs. III, IVA) 3.1 1.37.2 0.007
Hb nadir (continuous) 0.4 0.151.0 0.05
Parametrial involvement (none vs. unilateral vs. bilateral) 0.9 0.44.9 0.556
a
Abbreviations: OR, odds ratio; CI, confidence interval

compared to patients who maintained normal Hb the patients had a nadir Hb of 9.4 g/dl and experi-
levels. In a multivariate analysis, the nadir Hb and enced a PFS of only 39% at 5 years. In contrast,
the stage of the disease were the only prognostic patients who maintained a Hb of >9.4 g/dl had a
factors predictive of tumor progression. A quarter of 48% chance of being disease-free at the same time
(Fig. 1B). Interestingly, as with previous studies,
anemia at presentation was not prognostically signifi-
A cant (Fig. 1A).
Recently, the impact of anemia and blood transfu-
sion on the survival of 605 patients with cervical
cancer has been investigated at seven centers in
Canada(1). A low average weekly nadir Hb had a
negative effect on disease-free and overall survival.
The 5-year survival was 74% for patients who main-
tained a normal Hb (120 g/l) throughout radiother-
apy and only 45% for patients who dropped their Hb
below 110 g/l. In contrast, the Hb at presentation and
blood transfusion were prognostically irrelevant(1).
Similar findings were reported in another study of
386 patients treated with radiation therapy for locally
advanced cervical carcinoma(6). Multivariate analysis
demonstrated that a Hb <10 g/dl before radiation
therapy was not prognostically relevant. In contrast,
B
patients with at least one Hb value below 10 g/dl
during therapy had an 80% increased risk of local
regional failure compared with patients with all
their values above 10 g/dl(6).
Three possible explanations exist why anemia
appears to be associated with a worse outcome: 1)
The presence of anemia might indicate tumors
which are biologically more aggressive. However,
previous findings(1,6,13) and the findings from this
study indicate that the Hb values recorded before
treatment are prognostically irrelevant, whereas the
Hb values which were recorded during or even at
the end of radiotherapy are prognostically important;
2) Anemia might be a surrogate marker of tumor
hypoxia. This is known to mediate resistance to radio-
Fig. 1. Kaplan-Meier estimates of progression-free survival (PFS) in therapy(1821); 3) A mechanism different from anemia
60 patients who underwent chemoradiation for cervical carcinoma.
The bold lines represent anemic patients and the dotted lines
might not only mediate resistance to therapy but also
represent nonanemic patients. Patients who had an Hb below the induce anemia as a consequence.
25th quartile were considered anemic. A) Patients were In contrast to histopathologic factors, anemia is
dichotomized according to the Hb at the time of commencing potentially reversible. Various approaches have been
treatment (11.6 g/dl vs. >11.6 g/dl, Log-rank P 0.147). B)
Patients were dichotomized according to the lowest Hb during
evaluated to overcome the effects of hypoxia. Carbo-
chemoradiation (9.4 g/dl vs. >9.4 g/dl, Log-rank P 0.0057). gen, a mixture of carbon dioxide and oxygen, reduces

# 2003 IGCS, International Journal of Gynecological Cancer 13, 633639


638 A. Obermair et al.

the hypoxic cell fraction and increases tumor oxygen- adjuvant therapy after radical surgery in high-risk
ation(22,23). Blood transfusions and human recombin- early-stage cancer of the cervix. J Clin Oncol 2000;18:
160613.
ant Erythropoietin (rh-EPO) are the most frequently
8 Keys HM, Bundy BN, Stehman FB et al. Cisplatin, radia-
used methods to correct anemia. For the correction of tion, and adjunct hysterectomy compared with radiation
chronic anemia, blood transfusions are not optimal and adjuvant hysterectomy for bulky stage IB cervical
because the effect is only short lasting. Therefore, carcinoma. New Eng J Med 1999;340:115461.
multiple transfusions may be required throughout a 9 Morris M, Eifel PJ, Lu J et al. Pelvic radiation with con-
treatment program that lasts 5 to 6 weeks. In addition, current chemotherapy compared with pelvic and para-
aortic radiation for high-risk cervical cancer. New Eng J
there are small but definitive risks of transfusion Med 1999;340:113743.
reactions and viral infection and the availability of 10 Whitney CW, Sause W, Bundy BN et al. Randomized
blood products is limited(24,25). As an alternative, comparison of fluorouracil plus cisplatin vesus
research has concentrated on rh-EPO as its effect is hydroxyurea as an adjunct to radiation therapy in stage
long-lasting and it is synthesized recombinantly. IIB-IVA carcinoma of the cervix with negative para-aortic
lymph nodes: a Gynecologic Oncology Group and
In summary, we have demonstrated that anemia dur- Southwest Oncology Group study. J Clin Oncol 1999;17:
ing treatment identifies a group of patients who are at an 133948.
unacceptably high risk for treatment failure. The likeli- 11 Rose PG, Bundy BN, Watkins EB et al. Concurrent
hood of patients with a nadir HB  9.4 g/dl during cisplatin-based radiotherapy and chemotherapy for locally
chemoradiation being tumor-free at 30 months is only advanced cervical cancer. New Eng J Med 1999;340:
114453.
39%. This is similar to the results for patients receiving
12 Thomas GM. Concurrent Chemotherapy and radiation
palliative radiotherapy for stage IV disease(26). for locally advanced cervical cancer: the new standard of
As with every retrospective study, the authors care. Semin Radiat Oncol 2000;10:4450.
acknowledge the limitations of this analysis. While 13 Obermair A, Cheuk R, Horwood K et al. Impact of
an association of a low Hb during chemoradiotherapy hemoglobin levels before and during concurrent
chemo-radiotherapy on the response of treatment in
with treatment outcome is likely, a treatment recom-
patients with cervical carcinoma: Preliminary results.
mendation cannot be drawn. In the absence of Cancer 2001;92:9038.
explanations as to why some patients become anemic 14 Broders AC. Squamous-cell epithelioma of the lip: a
and others maintain normal Hb values throughout study of five hundred and thirty-seven cases. JAMA
chemoradiation, future research should investigate 1920;74:65664.
causes and pathogenetic mechanisms of anemia. 15 Creasman WT. New gynecologic cancer staging. Gynecol
Oncol 1995;58:1578.
Clinical trials currently evaluate whether rh-EPO not 16 Kaplan EL, Meier P. Non-parametric estimation from
only increases the blood Hb level but also increases incomplete observations. J Am Statist Assoc 1985;53:
the response to chemoradiation in cervical carcinoma 45781.
patients. 17 Mantel N. Evaluation of survival data and two new rank
order statistics arising in its consideration. Cancer
Chemother Rep 1966;50:16370.
References 18 Hoeckel M, Vaupel P. Tumor hypoxia. Definitions and
current clinical, biologic, and molecular aspects. J Nat
1 Grogan M, Thomas GM, Melamed I et al. The importance Cancer Inst 2001;93:26676.
of hemoglobin levels during radiotherapy for carcinoma 19 Nordsmark M, Overgaard M, Overgaard J. Pretreatment
of the cervix. Cancer 1999;86:152836. oxygenation predicts radiation response in advanced
2 Spivak JL. Cancer-related anemia. its causes and charac- squamous cell carcinoma of the head and neck. Radiother
teristics. Semin Oncol 1994;21:38. Oncol 1996;41:3140.
3 Mercadante S, Gebbia V, Marazzo A, Filosto S. Anaemia 20 Hockel M, Schlenger K, Hockel S, Vaupel P. Hypoxic
in cancer. pathophysiology and treatment. Cancer Treat cervical cancers with low apoptotic index are highly
Rev 2000;26:30311. aggressive. Cancer Res 1999;59:45258.
4 Bush RS. The significance of anemia in clinical radiation 21 Henke M, Bechtold C, Momm F, Dorr W, Guttenberger R.
therapy. Int J Radiat Oncol Phys 1986;12:204750. Blood hemoglobin level may affect radiosensitivity-
5 Dische S. Radiotherapy and anaemia-the clinical experi- preliminary results on acutely reacting normal tissues.
ence. Radiother Oncol 1991;20(Suppl.):3540. Int J Radiat Oncol Phys 2000;48:33945.
6 Girinski T, Pejovic-Lenfant MH, Bourhis J et al. Prognos- 22 Partridge SE, Aquino-Parsons C, Luo C, Green A,
tic value of hemoglobin concentrations and blood Olive PL. A pilot study comparing intratumoural
transfusions in advanced carcinoma of the cervix treated oxygenation using the comet assay following 2.5% and
by radiation therapy: results of a retrospective study of 5% carbogen and 100% oxygen. Int J Radiat Oncol Biol
386 patients. Int J Radiat Oncol Phys 1989;16:3742. Phys 2001; 49:57580.
7 Peters WA, Liu PY, Barrett RJ 2nd et al. Concurrent 23 Marinez JC, Villar A, Cabezon MA et al. Hyperfraction-
chemotherapy and pelvic radiation therapy alone as ated chemoradiation with carbogen breathing, with or

# 2003 IGCS, International Journal of Gynecological Cancer 13, 633639


Anemia before and during concurrent chemotherapy in cervical cancer patients 639

without erythropoietin: a stepwise developed treatment April, 2001 .http://www.health.gov.au/hsdd/bodt/


schedule for advanced head-and-neck cancer. Int J Radiat index.htm.
Oncol Biol Phys 2001;50:4753. 26 Benedet J, Odicino F, Maisonneuve P et al. Carcinoma of
24 American Association of Blood Banks. Facts about blood the cervix uteri. J Epidemiol Biostat 1998;3:534.
and blood banking. Update 2001. http://www.aabb.
org/All_About_Blood/FAQs/aabb_faqs.htm#donor.
25 Blood and Organ Donation Taskforce. Department of
Health and Aged Care. Fact Sheet no. 1. Updated 6 Accepted for publication May 30, 2003.

# 2003 IGCS, International Journal of Gynecological Cancer 13, 633639