Escolar Documentos
Profissional Documentos
Cultura Documentos
The purpose of this study was to evaluate the association between postmenopausal hormone therapy (HT) and the risk for
breast cancer in recently postmenopausal Finnish women. All Finnish women with first invasive breast cancer diagnosed
between the ages of 50 and 62 years during 1995–2007 (n 5 9,956) were identified from the Finnish Cancer Registry. For
each case, 3 controls of the same age were retrieved from the Finnish Population Register. The cases and controls were
linked to the national medical reimbursement register to assess the use of HT. The odds ratios (ORs) and 95% confidence
intervals (CIs) for breast cancer were calculated with conditional logistic regression analysis, adjusting for parity, age at the
first birth and health care district. Estradiol-only therapy (991 users with breast cancer, n) or oral progestagen (n 5 138) was
not accompanied by an increased risk. Estradiol-progestagen therapy (EPT) (n 5 1,731) was associated with an elevated risk
in the whole series (OR 1.36; 95% CI 1.27–1.46). The risk became detectable in less than 3 years of use. Continuous EPT use
tended to be associated with a higher risk for breast cancer than the sequential EPT use. The use of tibolone (n 5 80) (1.36;
1.15–1.96), a levonorgestrel-releasing intrauterine system (LNG–IUS) alone (n 5 154) (1.45; 1.97–1.77) or as a complement
to estradiol (n 5 137) (2.15; 1.72–2.68) was also associated with an increased risk. The association between HT use and the
risk for breast cancer shows a large variation between various forms of HT, and also the use of LNG-IUS may carry a risk.
Abundant data from the United States and many European that in the general female population. EPT use carried a
countries imply that the prolonged use of postmenopausal higher risk for breast cancer than ET, a result that resembles
hormone therapy (HT) is associated with an elevated risk for to a large extent to those obtained in many other countries. 1
breast cancer.1 National differences exist in the pattern of HT Our cohort could not control confounding factors, and the
use,2 and thus, data from one given country may not hold role of tibolone or the progestagen-releasing intrauterine sys-
for another. Our research team has conducted 2 nationwide tem (IUS) as possible risks for breast cancer could not be an-
studies on the risk for breast cancer in women using either alyzed. Therefore, we designed a case-control study to com-
estradiol-only therapy (ET)3 or estradiol-progestagen therapy
Epidemiology
pare the use of HT between patients with breast cancer and
(EPT) in Finland.4 In these studies, we used cohort analyses, women without breast cancer in Finland.
which compare the risk for breast cancer in the HT users to
Hormone therapy exposure Table 1. Type and dose of progestogens and estradiol in fixed
The history of HT use by the cases and controls was traced commercial products in Finland
from the national Medical Reimbursement Registry. This Dose of Dose of
registry contains data on the purchases of systemic HT since progestagen estradiol
Regimen (mg) (mg)
1994 nationwide. Its coverage is 100%, because in Finland
systemic HT, available only with a doctor’s prescription, is Sequential estradiol-progestagen therapy, type of progestagen
partly reimbursed by the national health care system. Only Oral
some vaginal estrogens carrying no risk for breast cancer3 are Norethisterone acetate 1 1 or 2
available without prescription. Only women who had pur- Medroxyprogesterone acetate 10 2
chased systemic HT regimens for at least 6 months of use Medroxyprogesterone acetate 1
20 2
were considered as users in this study.
Levonorgestrel 0.25 2
In Finland, the only systemic estrogen as a part of HT is
Dydrogesterone 10 1 or 2
estradiol, which is available both for oral and for transdermal
use. According to a national policy, only hysterectomized 20 2
women should use ET, whereas nonhysterectomized women Trimegestone 0.5 2
must complement estradiol with progestagen (EPT). The Transdermal
mode of the EPT regimen refers here to the mode of the Norethisterone acetate 0.17 or 0.25 0.05
progestagen use, which can be sequential or continuous. The
Levonorgestrel 0.01 0.05
sequential EPT regimen includes progestagen courses of 10–
0.02 0.065
14 days once a month or every 3rd month (long cycle) in
fixed preparations. A long cycle sequential regimen can also Continuous estradiol-progestagen therapy, type of progestagen
be formed individually when a user complements oral or Oral
transdermal estradiol with oral progestagen courses of 10–14 Norethisterone acetate 0.5 1
days with 2–3 months intervals. The continuous EPT 1 or 0.7 2
regimen includes progestagen every day. Progestagen can be
Medroxyprogesterone acetate 2.5 or 5 1
given either orally or transdermally both in sequential and
5 2
in continuous EPT. One popular option to give progestagen
for endometrial protection in Finland is the levonorgestrel- Dydrogesterone 5 1
releasing (20 lg/d) intrauterine system (LNG-IUS), Drospirenone 2 1
which can be fitted to women using estradiol orally or Transdermal
transdermally. Norethisterone acetate 0.17 or 0.25 0.05
If a woman had used 2 or more different progestogens or
1
regimens,the progestagen/regimen she had used for more Long cycle, progestagen for 14 days every 3rd month.
than 50% of the total exposure time determined the category
to which this woman was included in the analysis. The most mg/d, lynestrenol 5 mg/d, progesterone 200–300 mg/d and
megestrol acetate 10 mg/d.
Epidemiology
proportions.6 All statistical analyses were performed with Table 2. Distribution of selected characteristics of the cases and
STATA software, release 9.2. controls
Cases Controls
N Percent N Percent
Results Total 9,956 100 29,868 100
Breast cancer cases were more often nulliparous (p < 0.001) Year of birth
or had 1–2 children in the ages >25 years than controls (p < 1945–1946 2,774 27.9 8,322 27.9
0.0001) (Table 2). The proportion of women from southern
1947–1948 2,471 24.8 7,413 24.8
Finland was higher among cases than among controls (p <
0.0001). In the final model, parity, age at first birth and 1949–1950 1,865 18.7 5,595 18.7
health care district were all significant variables. The OR was 1951–1952 1,329 13.4 3,987 13.4
especially low (0.48; 0.36–0.64) for women with 5 children 1953–1954 939 9.4 2,817 9.4
and an age of younger than 25 years at first birth as com- 1955–1957 578 5.8 1,734 5.8
pared with nulliparous women. The ORs for breast cancer Age at inclusion
were highest in central Finland and southern Finland (1.16;
50–52 2,805 28.2 8,427 28.2
1.07–1.27 and 1.16; 1.08–1.26, respectively), whereas northern
Finland had the lowest breast cancer incidence as reference 53–55 3,178 31.9 9,521 31.9
category. 56–58 2,282 22.9 6,851 22.9
Cases had used HT more often (45%) than controls 59–61 1,456 14.6 4,364 14.6
(39.9%) (p < 0.0001) (Table 2). Almost 40% of the HT users 62–63 235 2.4 705 2.4
had used EPT and more than 20% had used ET. The use of
Age at first birth and parity
EPT was more common in cases than in controls (p <
Nulliparous 1,752 17.6 4,515 15.1
0.001), whereas among controls ET use was more common
(p < 0.01). Progestagen-only, tibolone and mixed therapy 1–2 children, first 2,160 21.6 7,000 23.4
were equally often used by the cases and controls. Around birth ≤25 years
14% of all HT users had LNG-IUS. 1–2 children, first 4,018 40.0 10,918 36.6
birth >25 years
The use of ET or oral progestagen-only therapy was not
associated with an increased risk for breast cancer (Table 3), 3þ children, first 895 9.0 3,640 12.2
whereas the use of tibolone, LNG-IUS alone and mixed ther- birth ≤25 years
apy use were significantly associated with an elevated risk for 3þchildren, first 1,131 11.4 3,795 12.7
birth >25 years
breast cancer (Table 3).
Sequential EPT (once a month) was accompanied with an Health care district at the time of diagnosis
elevated risk already in less than 3 years of use, and the risk Northern 1,046 10.5 3,680 12.3
was highest (OR 1.70; 1.43–2.036) after 5 years of use (Table Eastern 1,484 14.9 4,841 16.2
4). The ORs for long cycle EPT (49% MPA-containing prod- Western 1,330 13.4 4,156 13.9
Epidemiology
ucts) were similar in less than 5 years of use, but did not Middle Finland 2,329 23.4 6,737 22.6
reach significance in any duration category (Table 4). Contin-
Southern 3,681 36.9 10,201 34.2
uous use of progestagen as a part of EPT was significantly
associated with an elevated risk for breast cancer for any ex- Unknown 86 0.9 253 0.8
posure time category and this risk increased along with an Hormone therapy
increasing duration of exposure. The combined use of oral or No 5,473 55.0 17,956 60.1
transdermal estradiol plus LNG-IUS was accompanied with Yes1 4,483 45.0 11,912 39.9
more than 2-fold elevated risk for breast cancer after 3 years Estradiol-only therapy 991 10.0 3,300 11.0
of exposure (Table 4).
Estradiol-progestagen 1,731 17.4 4,243 14.2
Sequential NETA with estradiol was accompanied with an therapy
increased risk for breast cancer in less than 3 years of use
Progestagen-only therapy 138 1.4 476 1.6
(Table 5). The risk for breast cancer in users of sequential 2
dydrogesterone with estradiol did not reach the significance LNG-IUS 329 3.3 708 2.4
in any duration category. A sequential MPA-containing regi- Mixed therapy3 927 9.3 2,534 8.5
men showed an increased risk for breast cancer in the use of Estradiol plus LNG-IUS 287 2.8 473 1.6
from 3 years on. Tibolone 80 0.8 178 0.6
Continuous NETA with estradiol showed an increased
1
risk for breast cancer in all exposure categories, and the risk At least 6 months hormone therapy. 2Levonorgestrel releasing
intrauterine system. 3Mixture of estradiol-only, progestagen-only,
increased with an increasing duration of use (Table 6). Con- estradiol-progestagen therapy, or tibolone.
tinuous estradiol-MPA use was associated with an increased
Table 3. Relative risk of invasive breast cancer among Table 5. Relative risk of invasive breast cancer among
postmenopausal women using hormone therapy postmenopausal women using sequential1 estradiol-progestagen
therapy by type of progestagen, and by duration
Therapy Cases Controls OR1 95% CI p
2 Cases Controls OR2 95% CI p
No user 5,473 17,956 1.00 (Reference)
NETA (years)3
Estradiol-only therapy 991 3,300 1.01 0.93–1.09 0.88
Progestagen-only 138 476 0.97 0.80–1.17 0.73 <3 285 748 1.26 1.09–1.46 0.001
therapy 3<5 136 343 1.34 1.09–1.64 0.005
LNG-IUS 3
329 708 1.53 1.33–1.75 0.001 ≤5 127 248 1.75 1.41–2.19 0.001
Estradiol-progestagen 1,731 4,243 1.36 1.27–1.46 0.001 MPA (years)4
therapy <3 223 724 1.02 0.87–1.20 0.78
Estradiol plus LNG-IUS 287 473 2.07 1.78–2.41 0.001 3<5 113 263 1.45 1.16–1.82 0.001
Mixed therapy4 927 2,534 1.22 1.12–1.33 0.001
≤5 70 156 1.63 0.22–2.18 0.001
Tibolone 80 178 1.36 1.15–1.96 0.003
LNG (years)5
1
Adjusted with age, parity, age at first birth and health care district. <3 63 143 1.44 1.07–1.95 0.02
2
Had bought HT never or for less than 6 months. 3Levonorgestrel
3<5 30 61 1.70 1.09–2.65 0.02
releasing intrauterine system. 4Mixture of estradiol-only, progestagen-
only, estradiol-progestagen therapy, or tibolone. ≤5 8 27 1.04 0.47–2.31 0.92
Dydrogesterone (years)
Table 4. Relative risk of invasive breast cancer among <3 109 309 1.16 0.93–1.45 0.19
postmenopausal women using estradiol-progestagen therapy by 3<5 36 87 1.37 0.92–2.03 0.12
duration and mode of progestagen
≤5 18 39 1.53 0.87–2.69 0.14
Cases Controls OR1 95% CI p
6
Other progestogens (years)
Sequential EPT (years)
<3 26 74 1.17 0.74–1.84 0.50
<3 619 1,752 1.17 1.06–1.29 0.002
3<5 17 57 0.99 0.58–1.71 0.98
3<5 287 700 1.39 1.20–1.61 0.001
≤5 19 52 1.29 0.76–2.19 0.35
≤5 202 414 1.70 1.43–2.03 0.001
1
Long cycle EPT (years)2 Includes long cycle progestagen and progestagen once a month.
2
Reference category is never users of HT. Odds ratio adjusted with age,
<3 83 236 1.15 0.89–1.48 0.29 parity, age at first birth and health care district. 3Norethisterone
3<5 41 98 1.38 0.95–2.00 0.08 acetate. 4Medroxyprogesterone acetate. 5Levonorgestrel. 6Includes
sequential and continuous administration of progesterone, lynesterol
≤5 35 93 1.34 0.90–2.00 0.15 or megestrol acetate, trimegestone, or drospirenone (70% sequential
Continuous combined EPT (years) users).
≤5 128 207 2.16 1.71–2.71 0.001 use was associated with an elevated risk for breast cancer,
Continuous estradiol plus LNG-IUS (years)3 and the continuous EPT showed higher risks than sequential
<3 117 245 1.65 1.32–2.06 0.001 EPT. Also, the use of tibolone was associated with a signifi-
3<5 75 110 2.38 1.76–3.21 0.001 cant risk elevation. A most surprising finding was that the
sole use of LNG-IUS was associated with a significantly ele-
≤5 95 117 2.79 2.12–3.67 0.001
vated risk for breast cancer and that this risk tended to be
1
Reference category is never users of HT. OR adjusted for age, parity, further elevated by estradiol replacement in these women.
age at first birth and health care district. 2Estradiol with oral
One of the strengths of our study is the exact information
progestagen courses of 10–14 days with 2–3 months intervals.
3
Levonorgestrel releasing intrauterine system. of the HT purchases. It is very likely that the women truly
used the HT preparations they had bought, because HT is
risk for breast cancer after 5 years of use, and also the use of only partly reimbursed. We were also able to collect data on
LNG-IUS with estradiol carried a risk (Table 6). the women wearing LNG-IUS alone or in combination with
estradiol; these data are novel. Another strength is the whole-
population character of the study; all Finnish women, regard-
Discussion less of their place of residence, were included. The women
Because of the national differences in the pattern of HT living in big cities and leading the most urban type of life are
use,2,7 the impact of HT use for breast cancer should be stud- characterized by a higher risk for breast cancer.8–10 This was
ied in each country where HT is being used. In this large also seen in our study; southern (city health care districts)
national case-control study, we found no risk elevation for health care districts showed the highest risk for breast cancer.
Table 6. Relative risk of invasive breast cancer among do not complain about hot flushes. Four percent of the cases
postmenopausal women using continuous estradiol-progestagen had been exposed to such therapy for short periods.
therapy by type of progestagen, and by duration
Although a progestagen component of EPT may be a stron-
Cases Controls OR1 95% CI p ger determinant for the risk of breast cancer, short oral pro-
2
NETA (years) gestagen therapy (without estrogen) use was not accompanied
<3 155 359 1.44 1.19–1.75 0.001 by an increased risk for breast cancer. This is a reassuring
3<5 110 193 1.97 1.55–2.50 0.001 message for clinicians, and basically our presentdata are in
≤5 102 167 2.13 1.65–2.74 0.001 line with previous finding in Sweden.14
MPA (years)3 The use of conjugated equine estrogen only was not asso-
ciated with an elevated risk for breast cancer in the WHI
<3 27 81 1.16 0.75–1.80 0.50 study,15 whereas in our previous cohort study carried out
3<5 15 44 1.11 0.61–2.00 0.73 with standardized incidence ratio analysis, the use of estradiol
≤5 20 21 3.36 1.80–6.28 0.001 for more than 5 years was accompanied by an elevated risk
LNG-IUS (years)4 for breast cancer (1.44; 1.29–1.59).3 In this study, we did not
<3 118 247 1.65 1.32–2.07 0.001 find an elevated risk for breast cancer among ET users. This
discordant finding is possibly due to the difference in age
3<5 75 110 2.39 1.77–3.22 0.001 and exposure duration; the previous study included women
≤5 95 117 2.78 2.11–3.67 0.001 50 to 85þ years with more than 10 years of ET use, whereas
Dydrogesterone (years) in this study all ET users between 50 and 62 years had less
<3 13 34 1.31 0.69–2.49 0.41 than 12 years of ET exposure. In the United Kingdom, cur-
3<5 8 14 1.86 0.78–4.46 0.16 rent use of estrogen-only therapy (including also other than
estradiol) was also accompanied by an elevated risk of breast
≤5 1 4 0.90 0.10–8.18 0.92
cancer for the use of less than 5 years.16
1
Reference category is never users of HT. Odds ratio adjusted with age, We have presented data based on the cohort study that the
parity, age at first birth and health care district. 2Norethisterone
acetate. 3Medroxyprogesterone acetate. 4Levonorgestrel releasing use of EPT is associated with an increased risk for breast can-
intrauterine system. cer also in Finland, and this risk elevation was seen after 3
years of EPT use.4 This finding was consistent with the present
case-control study. In line with previous data,2,12,17–20 continu-
Moreover, we were able to control the age at the first birth ous EPT carried a higher risk than sequential EPT, although
and the parity, which are important risk factors for breast the difference was not significant. It is noteworthy that the
cancer. We could not control other confounders such as age control of some important confounders in this study did not
at menarche, age at menopause, body weight or socioeco- modify the general conclusion on the impact of EPT for the
nomic status. However, socioeconomic differences among risk of breast cancer or in the study population, in which EPT
postmenopausal HT users and nonusers declined in Finland was generally taken younger. Taking into account some im-
by the 1990s and are virtually nonexistent among women portant confounders, our present study, however, demon-
younger than 55 years.11 Some women may have used HT strates that the use of EPT is associated with a risk elevation
Epidemiology
before the age of 50 years. However, such a possible use for breast cancer already within the first 3 years of use.
should not modify the risk for breast cancer significantly. There are sparse data on the possible differences between
Moreover, the late age at menopause is associated with various progestogens in regard to the breast cancer risk. An
increased risk for breast cancer12 and these women start the English study16 reported comparable risks for users of EPT
use of HT later. Therefore, both these factors may contribute regimens containing MPA, NETA and LNG. Our cohort
to the speculation that our risk estimates related to short use study in women with longer use of EPT revealed a higher
of EPT may be slightly too high. risk for breast cancer among NETA than in MPA users after
We could not analyze whether the cases and controls had 5 years of exposure,4 and a comparable finding has also been
undergone a mammography with equal frequency. However, reported from Germany.20 This study on EPT users in
the Finnish health care system provides mammography women aged 50–62 years and with shorter exposure failed to
screenings free of charge every 2nd year, the coverage of demonstrate any difference between NETA and MPA-con-
which is 95% among 50- to 59-year-old women.13 This taining EPT regimens. In agreement with an earlier cohort
should reduce the chance of diagnosis bias in our study. study by our team and also with data from France, 7 EPT-
Finally, our conclusions related to HT use of more than 5 containing dydrogesterone was not associated with a signifi-
years cannot be very strong, because the entire time window cantly elevated risk for breast cancer in the present case-con-
of the study was only 12 years. trolled study. However, a relatively small sample size may
According to Finnish guidelines in management of meno- limit firm conclusions of the breast safety of dydrogesterone.
pause, oral progestagen-only therapy should be used for the The impact of tibolone on the risk for breast cancer is
control of cycle irregularities in perimenopausal women who controversial; this issue could not be addressed in our
previous study.4 In this study, we report an association rise in breast cancer incidence. Moreover, a family risk for
between tibolone and breast cancer, which in line with some breast cancer may have led to the insertion of the LNG-IUS,
previous studies.16,21 Likewise in a recent large placebo-con- alone or as a part of the HT, because the ‘‘low’’ release of
trolled trial of approximately 3,100 women with a history of LNG into the blood may have been assumed to be neutral to-
breast cancer, the use of tibolone (2.5mg/day) was associated ward the breasts. Unfortunately, we were unable to collect
with a 48% increase in the risk of new cancer or recurrence data on weight, cycle disturbances or family history of breast
within the first 4 years of use.22 The dose issue can be impor- cancer in our study. We must also admit that the accurate
tant; another randomized trial on osteoporotic elderly women exposure to LNG-IUS could not be set. Generally taken, LNG-
demonstrated that a smaller dose of tibolone (1.25 mg/d) led IUS users are satisfied with this intrauterine system,27,28 and
to a significantly reduced risk for breast cancer.23 therefore we are confident that the LNG-IUS users had worn
The contraceptive use of LNG-IUS has no effect on the this system for 5 years, as approximated in our analy- sis.
risk for breast cancer in women of fertile age.24 Therefore, However, a chance of prescription bias of LNG-IUS may have
the elevated risk observed in this study for breast cancer in contributed to the excess risk for breast cancer in our study.
women wearing a LNG-IUS, alone or as a complement to es- In summary, our case-control study showed that the use
tradiol, was unexpected. There are many explanations for this of estradiolalone or progestagen alone does not relate to a
finding. First, it could be possible that LNG, entering first into rise in the risk for breast cancer. The use of EPT was associ-
the blood and then also into the breast tissue in women with an ated with an elevatedrisk for breast cancer in less than 5
IUS, can cause cancerous changes in breast cells and/ or, more years of use, and the risk elevation was higher for continuous
likely, can promote the growth of the existing unde- tectable EPT rather than for sequentialEPT use. The use of tibolone,
cancer.25 This explanation sounds unlikely, because the levels LNG-IUS, alone or as a complement to estradiol, was accom-
of LNG in blood in postmenopausal women with LNG-IUS panied with an elevated risk for breast cancer.
are very low.26 Second, we must admit a chance of selection
bias; women characterized with an inherently increased risk Acknowledgements
for breast cancer could have been fitted with an IUS. These We are grateful to Dr. Timo Klaukka (deceased May 21, 2009) in the Social
characteristics may include hyperestrogenic states, such as Insurance Institution of Finland for the impact in the collection of the study
population.
obesity, or higher age at spontaneous meno- pause, both of
which are known to be accompanied with a
References
1. Collins JA, Blake JM, Crosignani PG. 9. Pukkala E, Weiderpass E. Time trends in 15. Anderson GL, Limacher M, Assaf AR,
Breast cancer risk with postmenopausal socio-economic differences in incidence Bassford T, Beresford SA, Black H, Bonds
hormone treatment. Hum Reprod Update rates of cancers of the breast and female D, Brunner R, Brzyski R, Caan B,
2005;111:545–60. genital organs (Finland, 1971–1995). Int J Chlebowski R, Curb D, et al. Effects of
2. Lee SA, Ross RK, Pike MC. An overview of Cancer 1999;81:56–61. conjugated equine estrogen in
menopausal oestrogen-progestin hormone 10. Pukkala E, Patama T, Engholm G, postmenopausal women with hysterectomy:
therapy and breast cancer risk. Br J Cancer Ólafsdóttir GH, Bray F, Talbäck M, the Women’s Health Initiative randomized
Epidemiology
20. Flesch-Janys D, Slanger T, Mutschelknauss LIBERATE Study Group. Lancet Oncol 26. Raudaskoski T, Tapanainen J, Tomas E,
E, Kropp S, Obi N, Vettorazzi E, Braendle 2009;10:135–46. Luotola H, Pekonen F, Ronni-Sivula H,
W, Bastert G, Hentschel S, Berger J, 23. Cummings SR, Ettinger B, Delmas PD, Timonen H, Riphagen F, Laatikainen T.
Chang-Claude J. Risk of different Kenemans P, Stathopoulos V, Verweij P, Intrauterine 10 microg and 20 microg
histological types of postmenopausal breast Mol-Arts M, Kloosterboer L, Mosca L, levonorgestrel systems in postmenopausal
cancer by type and regimen of menopausal Christiansen C, Bilezikian J, Kerzberg EM, women receiving oral oestrogen
hormone therapy. Int J Cancer 2008;123: et al. The effects of tibolone in older replacement therapy: clinical, endometrial
933–41. postmenopausal women. N Engl J Med and metabolic response. BJOG 2002;109:
21. Stahlberg C, Pedersen AT, Lynge E, 2008;359:697–708. 136–44.
Andersen ZJ, Keiding N, Hundrup YA, 24. Backman T, Rauramo I, Jaakkola K, Inki P, 27. Backman T, Huhtala S, Luoto R,
Obel EB, Ottesen B. Increased risk of Vaahtera K, Launonen A, Koskenvuo M. Tuominen J, Rauramo I, Koskenvuo M.
breast cancer following different regimens Use of levonorgestrel-releasing intrauterine Advance information improves user
of hormone replacement therapy frequently system and breast cancer. Obstet Gynecol satisfaction with the levonorgestrel
used in Europe. Int J Cancer 2004;109: 2005;106:813–17. intrauterine system. Obstet Gynecol 2002;
721–27. 25. Horwitz KB, Sartorius CA. Progestins in 99:608–13.
22. Kenemans P, Bundred NJ, Foidart JM, hormone replacement therapies reactivate 28. Jensen JT, Nelson AL, Costales AC. Subject
Kubista E, Von Schoultz B, Sismondi P, cancer stem cells in women with and clinician experience with the
Vassipoulou-Sellin R, Yip CH, Egberts J, preexisting breast cancers: a hypothesis. J levonorgestrel-releasing intrauterine system.
Mol-Arts M, Mulder R, Van Os S, et al. Clin Endocrinol Metab 2008;93:3295–98. Contraception 2008;77:22–2.
Epidemiology