Progestagens Use Before Menopause and Breast Cancer Risk According to Histology and Hormone Receptors

Breast cancer is not a single entity. In particular, tumors differ in their histologic type and hormone receptor status. Histologic type and hormone receptor status may further define groups of tumors that are etiologically distinct, as suggested by the fact that associations between a number of risk factors and breast cancer differ according to these characteristics (1-4).

Previous studies showed that combined menopausal hormone therapy (MHT; estrogen plus progestin) increases the risk of breast cancer (5) but also suggest that the risk differs according to the dose and/or to the type of progestagen used in the combination (5, 6). It has been suggested that progestagens may increase breast cell proliferation as well as breast cancer cell proliferation (7), but the mechanisms by which progestagens operate in breast carcinogenesis remain largely unknown. In an earlier report, we showed a potential increase in breast cancer risk associated with the use of oral progestagens prescribed alone (i.e., without estrogen) before menopause (8), a treatment often prescribed in France for perimenopausal disorders, contraception, and gynecologic disorders such as breast pain, uterine, or ovarian pathologies and irregular menstruations (9).

At present, the French E3N cohort is the only known large cohort with the ability to study the effects of progestagens used alone before menopause on breast cancer risk. In addition to results of other studies on progestagens used in postmenopause, our results may help to better understand the mechanisms of progestagens on breast carcinogenesis.

Two main arguments can be put forward to support the hypothesis that progestagens are differently associated with the risk of different types of breast cancer: (a) progestagens act through a hormonal mechanism and may therefore differently effect the risk of cancer with different hormone receptor profiles; (b) combined MHT has been described as being more strongly associated with lobular than ductal breast cancers (10).

To provide more informative data on the potential effect of progestagens on the risk of different subtypes of breast cancer, we studied the association between the use of progestagens in premenopause and after the age of 40 years and the risk of breast cancer defined by histologic and hormone receptor status among women from the French E3N cohort.

E3N is a French prospective cohort set up in 1990 to investigate cancer risk factors in women. A total of 98,995 women ages 40 to 64 years belonging to the Mutuelle Générale de l'Education Nationale, a French health insurance scheme primarily covering teachers, and residing in France, agreed to volunteer by filling in the first questionnaire and a consent form. Since June 1990, participants have been asked at ∼2-y intervals to complete self-administered questionnaires requesting information on various exposures and medical diagnoses.

Information on lifetime use of hormonal treatments, including progestagens, was first recorded in the January 1992 questionnaire. To facilitate accurate recall, a booklet presenting an extensive list and color photographs of hormonal treatments marketed in France was mailed to all study participants. For each episode of treatment (defined as the continuous use of the same hormonal brand) brand name, age at initiation of use, and duration were recorded. Information on hormonal treatment use was updated in each of the subsequent questionnaires. Information on the dose and number of treatment days in the cycle was not requested.

In the present study, the exposure of interest was the use of oral progestagens prescribed alone (i.e., not combined with an estrogen) before menopause but after the age of 40 years. “Mini pills” did not contribute to this exposure and because they were only occasionally used in our study population, were classified with estrogen-progestagen oral contraceptives (OC).

We defined the date of menopause as the date preceding 12 consecutive mo of amenorrhea (if the last menstrual period occurred before any MHT use; and if amenorrhea was not due to hysterectomy) or the date of bilateral oophorectomy; or, in decreasing order of priority, the self-reported date of menopause, the date that MHT use began, the date when menopausal symptoms began, or an imputed date corresponding to age 47 y if menopause was due to oophorectomy and to age 51 y otherwise (which were the median ages for surgical and natural menopause in the cohort, respectively).

Cases were identified from self-reports of participants: all questionnaires asked them whether any cancer had been diagnosed, requesting the address of their physicians and permission to contact them to obtain the pathology reports. The pathology reports were obtained for 96.6% of the incident cases identified. Information on histologic type and estrogen receptor (ER) and progesterone receptor (PR) status was extracted from these reports, and invasive breast cancer cases were classified accordingly as ductal, lobular, or other (which included mucinous, comedocarcinoma, tubular, medullar, papillar, and mixed ductal/lobular) for histologic type, and ER+/PR+, ER+/PR−, ER−/PR+, ER−/PR−, or missing for joint ER/PR status. We excluded cases whose information on histology (n = 87) or joint hormone receptor status (n = 522) was missing from the corresponding analyses.

Follow-up started at the date of return of the second questionnaire (sent out in January 1992). It continued until the return of the follow-up questionnaire sent out in June 1993, January 1995, April 1997, June 2000, or July 2002, whichever was answered last. Person-years accrued until that date or until diagnosis of any cancer, whichever occurred first.

Women who had a prevalent cancer other than basal cell carcinoma before inclusion (n = 8,031) were excluded, as well as women who had never menstruated (n = 25), those who did not answer the second questionnaire (n = 7,673) or for whom the second questionnaire was the last they had responded to (n = 3,169), and those who did not report either their date at the start of progestagen treatment or duration in Q2 (n = 5,815) or in a subsequent questionnaire (n = 183). To focus the study on intake of progestagens during the perimenopausal period, the analysis was further restricted to women who had never used a progestagen before the age of 40 years and who reached menopause after the age of 40 y. This left us with 67,057 women for the analysis, accruing 636,188 person-years, with an average age at start of follow-up of 51.4 y (SD, 6.7), a mean follow-up time of 9.5 y (SD, 2.0), and a rate of lost of follow up of 8.5% (a women is lost of follow-up if the date of end of follow-up is neither date of diagnosis of any cancer nor July 2002).

Relative risks for breast cancer were estimated using Cox proportional hazards models according to a competing risk method (11). Age was used as the time scale. Known risk factors for breast cancer were included in the model, as well as confounding variables if they improved model fit by the P value of <0.3 criterion; we adjusted for body mass index before and after menopause (time dependent), menopausal status (time dependent), age at menopause, parity, age at full-term pregnancy (FTP), age at menarche, familial history of breast cancer in first-degree relatives, personal history of biopsy-defined benign breast disease (time dependent), personal history of benign uterine or ovarian disease (time dependent), use of OC (time dependent), use of post-MHT (time dependent) and previous mammography (time dependent), period of time (time dependent), further stratified on year of birth; cutoff are indicated in the footnotes of the tables. Imputation to the median was used for adjustment factors with 5% or less of missing values; for those with >5% of missing value, we created a “missing” category. Each variable reported on a specific questionnaire was used to prospectively categorize women for the period that followed (i.e., we considered the actualized information reported in questionnaire i and earlier for the follow-up period between questionnaire i and questionnaire j where j is the next questionnaire the women completed, j> = i+1). In particular, progestagen use was included in the model as a time-dependent variable. The reference group in each model consisted of women who indicated that they had not used yet any progestagen alone.

Hazard ratios (HR) are given with 95% confidence intervals (CI). Methods for estimating and testing the model variables were obtained by likelihood methods and Wald test. All analyses were done using SAS system, version 9.1.

The main characteristics of the 67,057 women included in the analysis are shown in Table 1 according to use of progestagen treatment alone use. At the end of follow-up, ever users had later menopause and more frequently had a personal history of benign disease of the breast, uterus, or ovary than never users. Ever use of OC (estrogen-progestagen and mini pills) or of MHT, and mammographic follow-up were more frequent in ever users of progestagens, and young generations were more likely to have used progestagens than older women. During follow-up, 2,264 cases of primary invasive breast cancer (first cancer) were identified among the 67,057 women in the cohort.

Pathology reports were obtained for 96.6% of cases. Histologic type was reported for 96.2% of cases (n = 2,177) and joint ER/PRstatus for 76.9% of cases (n = 1,742). Joint distribution is showed in Table 2.

At the end of follow up, 57,587 women were menopausal. Among them, 52,019 had a natural menopause, 4,237 were menopausal due to a bilateral oophorectomy and for the remaining 1,331 menopausal women, and we did not know whether menopause was natural or surgical. Among menopausal women, 5,305 had an unknown age at menopause, and their age at menopause was therefore imputed (4,452 for naturally menopausal, 758 for unknown type of menopause, and 95 for surgical menopause).

We found a differential (P = 0.01) association between ever-use of progestagens before menopause and the risk of ductal and lobular breast cancer: compared with never-users, the risk of lobular breast cancer was significantly increased (HR, 1.40; 95% CI, 1.12-1.76); there was no significant association with the risk of ductal breast cancer (HR, 0.97; 95% CI, 0.87-1.08; Table 3).

We further stratified our analysis according to duration and recency of use. As in the previous study (8), we used two classes of duration: less then 4.5 years (corresponding to the three first quartiles) and 4.5 years or more (corresponding to the fourth quartile). Current use of a treatment for <4.5 years was not significantly associated with ductal carcinoma risk (HR, 1.01; 95% CI, 0.82-1.25); current use of a treatment for 4.5 years or more was significantly (P = 0.048) associated with ductal carcinoma risk (HR, 1.49; 95% CI, 1.00-2.22); and past use was not significantly associated with ductal carcinoma risk (HR, 0.94; 95% CI, 0.84-1.06). In contrast, the increase in lobular carcinoma risk was apparent in current users (HR, 1.51; 95% CI, 1.02-2.24) and in past users (HR, 1.38; 95% CI, 1.08-1.75; Table 3).

When progestagen use was considered as ever/never, we did not find any differential association according to joint ER/PR status (Table 4). When stratifying according to recency of use and duration, current use of progestagens was significantly associated with ER+/PR− carcinomas (HR, 1.53; 95% CI, 1.01-2.32) and current use of progestagens for 4.5 years or more was significantly associated with ER+/PR+ carcinomas (HR, 1.68; 95% CI, 1.05-2.68; Table 4).

We studied the association between ever-use of progestagens and the risks of joint histology and receptor-defined breast cancers (data not shown). We observed a significantly differential association between use of progestagens before menopause and the risk of ductal or lobular ER+/PR+ breast cancer for both current and past use (P for homogeneity = 0.012). There was no significant differential association between use of progestagens before menopause and the risk of ductal or lobular ER+/PR−, ER−/PR+ and ER−/PR− breast cancer (number of cases is very low).

Risks associated with different types of progestagens (progesterone, retroprogesterone, medrogestone, demegestone, cyproterone acetate, medroxyprogesterone acetate, nomegestrol acetate, chlormadinone acetate, ethynodiol, norethisterone acetate, lynestrenol, and promegestone) did not vary significantly (number of cases in subcategories is very low) among the subgroups of breast cancers defined by histologic type (lobular, ductal, and other) or hormone receptor status (ER+/PR+, ER+/PR−, ER−/PR+, and ER−/PR−; data not shown).

In the present analysis, we found an increased risk of lobular carcinoma associated with premenopausal use of progestagens, whatever the recency of use. Moreover, we found that current use of progestagens was associated with an increase in the risk of ER+/PR− carcinoma. For ductal and ER+/PR+ breast carcinomas, which represent the majority of breast cancer cases, we found an increased risk associated with current use of progestagens that lasted longer than 4.5 years.

We found no study in the literature on the relationship between the use of progestagens alone and the risk of breast cancer according to histology and hormone receptor status. Associations between MHT or OC and breast cancer risk have been shown to differ according to hormone receptor status and histologic type.

Regarding histologic type, it has been reported that combined MHT use was more strongly related to lobular than to ductal carcinomas (10, 12-14). Furthermore, one study reported that combined MHT (estradiol plus medroxyprogesterone acetate) preferentially stimulated lobule cells rather than ducts cells in a normal breast (7). Although results on OC are scarcer, a number of studies have shown an association with the use of OC that was limited to lobular breast cancers (15, 16). Thus, our result indicating that progestagen use before menopause is associated with the risk of lobular breast cancer is consistent with these findings.

There exist biological arguments to explain a greater increase in risk of lobular carcinoma with premenopausal progesterone use than in ductal carcinoma. Studies on PR–knocked-out mouse suggest that progesterone induces lobulo-alveolar development, whereas estradiol stimulates ductal elongation and PR expression (17). It has been shown in humans that lobular tumors are more frequently ER and PR positive compared with breast tumors of other histologic types, suggesting that they are more hormone dependent (13, 18). Indeed, we found an increased risk of lobular ER+/PR+ carcinomas associated with current use of premenopausal progestagens.

It has also been suggested that use of combined MHT was more likely to be associated with ER+/PR+ tumors than other subtypes (4, 19-21). In contrast, the use of OC was more strongly associated with ER− tumor risk than ER+ tumor risk in one case-control study (22), and with ER−/PR− tumors than ER+/PR+ tumors in two others (23, 24). However, one case-control study found that this differential association could be reversed with menopausal status (25). These results on combined MHT and OC are discordant and our results do not fit with any of them. In the present study, we indeed found that current premenopausal use of progestagens was associated with both ER+/PR+ and ER+/PR− breast carcinoma risk, and we could not exclude the possibility that it also affected the risk of ER− tumors due to the low number of ER− cases.

Several mechanisms may explain our results reporting a higher increase in ER+/PR− carcinomas. First, it is well-known that progestagens induce PR down-regulation in breast cancer cells in vitro (26), but this down-regulation is transient and disappears 48 hours after progestagen withdrawal (27). Second, the phenotype ER+/PR− may be due to overexpression of human epidermal growth factor receptor 2 (28). However, to the best of our knowledge, human epidermal growth factor receptor 2 expression has not been studied after progestagen use, and most studies failed to find significantly more frequent human epidermal growth factor receptor 2 overexpression in breast cancers diagnosed in MHT users versus MHT nonusers (29-31), except for one study based on very small numbers (32). Third, the absence of PR may indicate high insulin-like growth factor, epidermal growth factor and heregulin activities, which down-regulate PR independently of ER status (28). As progestagens up-regulate insulin-like growth factor and epidermal growth factor receptors (33), they may increase the potency of growth factors and therefore preferentially affect the risk of ER+/PR− tumors.

The associations observed on current use of premenopausal progestagens cannot be explained by the effect of MHT on breast cancer risk because MHT is used after menopause. Furthermore, a sensitivity analysis censuring women at age of start of MHT yielded similar estimates of the risk associated with past use of progestagens alone before menopause.

Among the strengths of our study are its large population, a 2-year update on hormone exposure, and its ability to adjust for numerous confounding factors, especially the exclusion of women who were exposed to progestins before the age of 40 years. In particular, we adjusted our analyses on previous mammograms as a time-dependent variable so as to limit a potential detection bias.

Limitations described previously (8) include the absence of information on reasons for prescribing progestagens, their dose, and duration per cycle. Although the reasons for prescribing progestagens were not recorded, a potential “prescription” bias is unlikely because we adjusted for the variables “personal history of benign breast disease” and “personal history of benign uterine or ovarian disease”, and because the effect of progestagens on breast cancer risk did not vary significantly according to a personal history of benign breast disease or to a personal history of benign uterine or ovarian disease.

Assessment of hormone receptors and histology of tumors, as determined from various laboratories and techniques, may vary over time and across laboratories. However, the resulting misclassification should not be associated with progestagen use. Moreover this potential bias had been accounted for by adjusting for the date of diagnosis (before 1994/1994-1996/1997 or later).

In conclusion, our study suggests that the use of oral premenopausal progestagens (without estrogens) by women over 40 years may preferentially increase the risk of lobular rather than ductal carcinoma, and differentially affect the risk of breast cancer according to hormone receptor status. Our results are noteworthy but have yet to be confirmed or refuted by other studies.

No potential conflicts of interest were disclosed.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

We thank all the participants for providing data and to the practitioners for providing pathology reports, R. Chaït for obtaining pathology reports, and to M. Fangon, M. Niravong, and L. Hoang for managing the data.