Gynecologic tumors in Thai women with breast cancer: a cross-sectional surveillance and ten-year cohort study

Breast cancer is one of the most common cancers affecting Thai women and women worldwide ( 1 , 2 ). Breast cancer and gynecologic tumors share some risk factors, e.g., gene mutation and personal characteristics. Therefore, women with breast cancer are presumed to have a higher prevalence of gynecologic tumors than women without breast cancer ( 3 ). The prevalence of endometrial abnormalities varies from 3.3% to 36% in women with breast cancer ( 4 - 6 ), whereas, in the general population, it is lower than 1% ( 7 ). In addition, the estimated risk of primary ovarian cancer is doubled for breast cancer survivors ( 3 ). Despite the high theoretical risk of gynecologic tumors in women with breast cancer, there was unexpectedly sparse information regarding this issue. Most relevant reports focus only on the risk of endometrial abnormalities in women with breast cancer taking tamoxifen. Tamoxifen users have up to 7.5 times greater risk of endometrial cancer than non-users. In absolute terms, this corresponds to an estimated 20 additional endometrial cancers per 1,000 women over 10 years of tamoxifen exposure ( 6 , 8 , 9 ). In addition, previous studies have shown the correlation between tamoxifen usage and the risk of various benign gynecologic tumors, including endometrial polyp, leiomyoma, and ovarian cyst ( 10 ). Therefore, some authorities have recommended pelvic examination, transvaginal ultrasonography, and sonohysterography to monitor current tamoxifen users ( 10 - 12 ). On the contrary, the American College of Obstetricians and Gynecologists (ACOG) [2014] stated that routine endometrial surveillance (e.g., transvaginal ultrasonography and/or endometrial biopsy) in asymptomatic tamoxifen users has not been shown to improve early detection and is not recommended, as it may lead to more invasive and costly diagnostic procedures ( 13 ). They recommended that all postmenopausal women with breast cancer be screened for pre-existing gynecologic conditions before tamoxifen initiation ( 13 ). However, this recommendation has not been applied by most health care providers, as it was found that up to 50% of women with breast cancer had never had gynecologic screening ( 14 ). This shortcoming is also present in our institute. To achieve this recommendation to practice, pertinent caregivers must recognize the magnitude of the risk of gynecologic tumors and the importance of regular gynecologic surveillance in this specific group of patients. The present study aimed to survey the prevalence of concurrent gynecologic tumors in Thai women with recently diagnosed breast cancer. In addition, we determined the incidence of de novo gynecologic tumors arising during a ten-year follow-up period and the risk of genital synchronous and metachronous cancers. We present this article in accordance with the STROBE reporting checklist (available at https://gs.amegroups.com/article/view/10.21037/gs-2025-1-422/rc ).

The present study comprised two phases: the initial surveillance phase and the routine follow-up phase. The initial phase was a cross-sectional study conducted at the Gynecologic Endocrinology Unit, Department of Obstetrics and Gynecology, and the Division of Head-Neck and Breast Surgery, Department of Surgery, Faculty of Medicine Siriraj Hospital, Mahidol University between September 2007 and January 2009. The initial phase aimed to survey the prevalence of concurrent gynecologic tumors in Thai women with breast cancer. Thereafter, the cohort underwent a routine annual gynecologic examination. The follow-up data up to February 2019 were retrospectively analyzed. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The initial cross-sectional study and the analysis of ten-year follow-up data were approved by the Institutional Review Board of Siriraj Hospital. Written consent was obtained from all participants who agreed to participate in the study. Participants were Thai women with recently diagnosed breast cancer who were at least 18 years of age, were not pregnant, had never received medications for breast cancer, did not take sex hormones within three months before enrollment, and had intact uterus and ovaries. After informed consent was obtained, each patient underwent the surveillance process for gynecologic tumors. All patients were interviewed for the following information: demographic data, menstrual and obstetric history, medical and surgical history, and family history of cancers. They then underwent pelvic examination, cervical cytology testing, and pelvic ultrasonography. Some patients needed further investigative or operative procedures as indicated. Breast cancer was managed according to the standard protocol of the Division of Head-Neck and Breast Surgery, Department of Surgery, including surgery with or without chemotherapy, radiotherapy, hormone therapy, and targeted therapy. Any abnormal findings in the gynecologic examination were managed according to the Department of Obstetrics and Gynecology protocol, including expectant management, endometrial sampling, fractional curettage, hysteroscopy, and laparotomy. After the initial surveillance, the patients who retained genital organs were appointed for annual gynecologic examination. Data of newly arising gynecologic conditions, including gynecologic tumors, cervical cancerous/pre-cancerous tumors, and metachronous genital cancers, were collected from the case records of our clinic and the Hospital Information System (iSOFT, IBA Health Group Co., Ltd., Sydney, Australia). The initial diagnosis of gynecologic tumors was made using ultrasonography. The ultrasonographic diagnosis included endometrial polyp, uterine leiomyoma, adenomyosis, and adnexal mass. In the patients who underwent an operative procedure, histopathology results were used for the final diagnoses. Genital cancer in the present study was considered the second primary cancer if it was not metastatic breast cancer. It was categorized into: (I) synchronous genital cancer ; cancer of reproductive organs detected within six months of the diagnosis of breast cancer; and (II) metachronous genital cancer ; cancer detected later than six months of the diagnosis of breast cancer. Pelvic ultrasonography was performed using an ultrasound machine, GE Voluson ® 730 Expert Diamond (USA), equipped with a 7.5-MHz vaginal probe. The ultrasound probe was inserted into the vagina. If the transvaginal examination was impossible, the probe was inserted into the rectum instead. In premenopausal women, all ultrasonography-related procedures were performed after the menstrual period finished and not later than menstrual cycle day 10. The endometrial ultrasonogram with a well-defined trilaminar pattern or a thickness of less than 5 mm was considered normal endometrium; therefore, no further endometrial surveillance was needed. Saline infusion sonohysterography was indicated for the case of a suspicious endometrial lesion, e.g., the ultrasonogram showing irregular lining, focal lesion, or heteroechoic endometrium ( 15 ). The sample size was calculated using the formula for a descriptive study. When the estimated prevalence of the leiomyoma (p) =20% ( 10 ) and that of endometrial abnormalities (p) =18.6% ( 8 ), a total of 269 women would be required for alpha =0.05 and absolute precision error (d) =4.6%. Statistical analysis was performed using Stata statistical software, version 14 (StataCorp LLC, College Station, TX, USA). As appropriate, data were presented in mean ± standard deviation (SD), number (n), and percent. The prevalence of concurrent gynecologic tumors and synchronous genital cancers were presented as a percentage and 95% confidence intervals (CIs). The incidence of new gynecologic tumors and metachronous gynecologic cancers during the 10-year follow-up period were present at an incident rate per 100 person-years and 95% CI. The data were analyzed using the Fisher’s exact test for categorical data and the t -test or the Mann-Whitney U test for continuous data. Regression analysis was performed to compare risk factors between groups. Statistical significance was considered when a P value was less than 0.05.

Initially, 329 patients with breast cancer were invited, and 48 declined. Finally, 281 women were enrolled. The baseline characteristics are shown in Table 1 . They were 50.4±10.6 years of age, and 43.1% were postmenopausal. Their average body mass index (BMI) was 23.6±3.6 kg/m 2 , and approximately one-quarter were overweight. Approximately one-third (32.7%) were nulliparous. Menstrual history demonstrated early menarche (≤10 years old) and late menarche (≥16 years old) in 0.3% and 13.0% of the patients, respectively. Overall, 16.0% of patients reported a family history of breast cancer; however, almost half (49.1%) reported a family history of associated gynecologic malignancies, including breast, endometrial, uterine, or ovarian cancer. Users of hormonal contraceptives comprised 34.5% of all women, and those of postmenopausal hormonal therapy comprised 10.7% of postmenopausal ones (4.6% of all patients). Most of the patients had estrogen receptor-positive breast cancer (64.4%), stage I or II breast cancer (78.5%), and invasive ductal carcinoma (80.9%). Data are presented as mean ± standard deviation or n (%). † , four cases could not recall age at menarche. ‡ , associated cancers included breast cancer, endometrial cancer, uterine cancer, ovarian cancer, and prostate cancer. § , stage of breast cancer was unknown in one case whose operation was performed in another hospital. ¶ , pathological reports and estrogen receptor status were not available in three cases whose operations were performed in other hospitals. BMI, body mass index. The prevalence of gynecologic tumors at the initial screening is shown in Table 2 . The overall prevalence of gynecologic tumors was 40.6% (95% CI: 34.8–46.6%) which included the prevalence of uterine leiomyoma (25.3%, 95% CI: 20.3–30.8%), adenomyosis (1.8%, 95% CI: 0.6–4.1%), adnexal mass (8.9%, 95% CI: 5.9–12.9%), and endometrial polyp (15.9%, 95% CI: 11.8–20.7%). The overall prevalence of gynecologic tumors was significantly higher in the premenopausal than in the postmenopausal group. However, only the prevalence of endometrial polyp had a statistically significant difference between the two groups. The proportion of pathologic adnexal mass was comparable between the postmenopausal women (33.3%) and the premenopausal ones (31.3%). The prevalence of abnormal cervical cytology was 2.5%. There were three patients with high-grade precancerous lesions [two cases of high-grade squamous intraepithelial lesion (HSIL) and one case of atypical squamous cells—cannot exclude HSIL (ASC-H)] and one with invasive squamous cell carcinoma; these comprised 57% (4/7 cases) of overall abnormal cervical cytology. A total of 125 patients underwent gynecologic procedures, including endometrial biopsy (n=98), fractional curettage (n=8), hysteroscopy (n=8), and laparotomy (n=11). Two cases had a second primary cancer, including an ovarian serous cystadenocarcinoma and cervical cancer. Thus, the prevalence of genital synchronous cancers was 0.7% (95% CI: 0.3–1.7%). Data are presented as n (%). † , data were compared between pre- and post-menopausal groups using Chi-squared test or Fisher’s exact test. ‡ , the patient who had more than one type of tumor was counted only once. ASC-H, atypical squamous cells—cannot exclude HSIL; ASC-US, atypical squamous cells of undetermined significance; HSIL, high grade squamous intraepithelial lesion; LSIL, low grade squamous intraepithelial lesion. Demographic and clinical parameters of 114 patients with gynecologic tumors and 167 patients with normal gynecologic findings are shown in Table 3 . The parameters that were significantly different between the two groups included age at menarche (13.7±1.6 vs. 14.2±1.6, P=0.01), previous hormonal contraceptive use (31% vs. 66%, P=0.04), and endometrial thickness (7.1±10.1 vs. 4.6±4.0, P<0.001). The estimated risk of having any concurrent gynecologic tumor in patients with breast cancer was higher in the patients who had never been exposed to hormonal contraceptives, had an earlier age at menarche, or had thicker endometrium ( Table 4 ). Data are presented as mean ± standard deviation or n (%) and were analyzed using t-test or Mann-Whitney U test for continuous data, and Chi-squared or Fisher’s exact test for categorical data. BMI, body mass index; NA, not applicable. Data were analyzed using logistic regression analysis. † , adjusted for exposure to hormonal contraceptive, endometrial thickness; ‡ , adjusted for age at menarche, endometrial thickness; § , adjusted for age at menarche, exposure to hormonal contraceptive. CI, confidence interval; OR, odds ratio. Of all 281 patients, 167 had at least one visit for gynecologic examination 10 years after the initial surveillance. There were 30 de novo gynecologic tumors in 29 patients, providing an incident rate of 3.1 per 100 person-years (95% CI: 2.1–4.4). The number of patients who remained in the annual gynecologic examination cohort and the cumulative incidence of newly arising gynecologic tumors is shown in Figure 1 . Clinical characteristics of patients with and without new gynecologic tumors were compared and presented in Table 4 . The estimated risk of having any de novo gynecologic tumor was significantly lower in the patients who had an older age at menarche [risk ratio (RR) 0.75, 95% CI: 0.58 to 0.98, data not shown]. Three patients had metachronous genital cancers, including an ovarian serous cystadenocarcinoma, endometrial cancer, and cervical cancer, providing an incident rate of 1.4 per 100 person-years (95% CI: 0.3–4.2). Clinical characteristics of survivors with and without newly arising gynecologic tumors are provided in Table S1 . Cumulative events of newly-arising gynecological tumors and/or abnormal cervical cytology in breast cancer survivors and the number of patients each year. ASC-US, atypical squamous cells of undetermined significance; HSIL, high grade squamous intraepithelial lesion; LSIL, low grade squamous intraepithelial lesion.

The overall prevalence of concurrent gynecologic tumors at the initial screening in our study (40.6%, 95% CI: 34.8–46.6%) was much higher than those in the general population (7–20%) ( 8 , 10 , 16 - 19 ) and postmenopausal Thai women with breast cancer (1.5–5.4%) ( 20 , 21 ). The higher prevalence in our study could be due to the difference in race, age, menopausal status, and diagnostic measures. In the present study, we included Thai women with breast cancer in all age groups regardless of their menopausal status or the presence of gynecologic symptoms. However, the cumulative incidence and the incident rate of de novo gynecologic tumors arising within 10 years after the initial screening in our study were comparable with the report from other studies ( 10 , 14 ). As most gynecologic tumors are hormone-dependent, a higher overall prevalence of the tumors is expected in premenopausal women ( 10 ). In the present study, the endometrial polyp was the only tumor with a significantly higher prevalence in premenopausal women than in postmenopausal women. Other gynecologic tumors might originate during the premenopausal period, persist to the postmenopausal period, and be stimulated during tamoxifen therapy. This hypothesis seemed accurate in adnexal mass, as postmenopausal women had a slightly lower prevalence than premenopausal women. Still, they had a significantly higher proportion of pathologic tumors. In premenopausal women, some adnexal masses were functional cysts that could spontaneously resolve, whereas the nonfunctional ones persisted or grew with time. We expected that postmenopausal women would have a lower prevalence of leiomyoma, but our study showed this not to be statistically significant. Some types of asymptomatic leiomyoma might be less responsive to sex hormones ( 22 ). Diagnostic modality is an essential factor influencing the prevalence of documented tumors. In 2004, Sinawat et al. used transvaginal ultrasonography to screen for gynecologic tumors in 66 postmenopausal Thai women with breast cancer and found an overall prevalence of 7.6% ( 20 ). In 2005, Chalas et al. analyzed data of 4,138 women at high risk for breast cancer in the Breast Cancer Prevention Trial (BCPT). They reported that benign gynecologic conditions had an overall prevalence of 5.16% ( 10 ), but the report did not mention the diagnostic tool. We used transvaginal or transrectal ultrasonography as the diagnostic measure in the present study. Moreover, we used sonohysterography to investigate lesions in the endometrial cavity. Transvaginal ultrasonography combined with sonohysterography has higher sensitivity, specificity, positive predictive value, and negative predictive value for the diagnosis of submucous myoma and focal endometrial lesions, compared with transvaginal ultrasonography alone ( 23 , 24 ). This difference in imaging approach may have contributed to the higher detection of endometrial lesions in our cohort, compared with a previous study ( 20 ). Our data demonstrated that risk factors for concurrent gynecologic tumors in women with breast cancer were earlier age at menarche, never using hormonal contraception, and a thick endometrium ultrasonogram measured via the transvaginal or transrectal route. In addition, the estimated risk of having de novo gynecologic tumors also increased in the patients who had an earlier age at menarche. Further study is warranted to prove the impact of these factors and identify the appropriated cutoff value of ultrasonographic endometrial thickness and age at menarche. Most of the gynecologic tumors found in the present survey were benign. However, the second primary genital cancer was not uncommon. The prevalence and incidence rate of second primary genital cancers of 0.7% (95% CI: 0.3–1.7) and 3.1 per 100 person-years (95% CI: 2.1–4.4) in the present study were consistent with previous reports ( 25 - 27 ). The findings highlighted the need for the surveillance of genital tumors in patients with breast cancer, both at the initial diagnosis and the following period. In our study, hormone receptor subtype and endocrine therapy details (e.g., tamoxifen) were not consistently recorded for all patients during the baseline period [2007–2009], which precluded a robust therapy-stratified analysis; therefore, we did not attempt to attribute gynecologic findings to endocrine therapy exposure in this cohort. Importantly, most concurrent gynecologic findings detected in this study were benign and may be asymptomatic; however, we consider them clinically relevant because they inform symptom-based follow-up and targeted evaluation in breast cancer survivors. Our study is intended to describe the surveillance yield of standard gynecologic assessment and pelvic ultrasonography rather than to imply that all detected lesions require immediate intervention. Further diagnostic procedures or treatments were undertaken only when clinically indicated. In the present study, the 2.5% prevalence of abnormal cervical cytology concurrently found at breast cancer diagnosis was slightly lower than the 3.7% prevalence in a 2009 report of general Thai women ( 28 ). The discrepancy might represent the difference in sexual behavior relating to age groups. The population in the present study was older than those in the 2009 report (average age of 50.4±10.6 vs. 42.66±12.21 years). Although the prevalence of high-grade and malignant lesions in the present study (1.8%) was comparable to the 2009 report, the proportion of these lesions in the abnormal cytology specimens was slightly higher in the present study than in the 2009 report (57.1% vs. 50.3%). Our findings regarding the prevalence and incidence of gynecologic tumors, second primary genital cancers, and abnormal cervical cytology indicated the significance of gynecologic surveillance in Thai women with breast cancer. The surveillance should begin with a routine gynecologic examination and cervical cytology testing. Application of transvaginal (or transrectal) ultrasonography with sonohysterography may benefit breast cancer women with a higher risk of gynecologic tumors, i.e., postmenopausal age, earlier age at menarche, and non-exposure to hormonal contraception. All women with breast cancer were encouraged to undergo regular gynecologic surveillance. However, further studies are mandatory to determine the cost-effectiveness of the gynecologic surveillance program in Thai women with breast cancer. A limitation of the present study was the diagnostic modality of tumors, which was made using only ultrasonography. In most cases, the gold standard histopathology was not available for the final diagnosis. Since the ultrasonographic diagnosis suggested benign diseases in most cases, performing an invasive procedure to obtain tissue diagnosis was unnecessary. Nonetheless, the technique used in the present study has a diagnostic accuracy as high as 90% ( 24 ). Therefore, we believe that the prevalence of gynecologic tumors in the present study was very close to the actual one.

Gynecologic tumors are prevalent in Thai women with breast cancer. The overall prevalence of concurrent gynecologic tumors is 40.6% (114/281), and the overall incidence of de novo tumors during the first 10 years is 3.1 per 100 person-years (95% CI: 2.1–4.4). The prevalence of synchronous genital cancers is 0.7% (95% CI: 0.3–1.7%), and the incidence of metachronous gynecologic cancers during the first 10 years is 1.4 per 100 person-years (95% CI: 0.3–4.2). Therefore, it is recommended that Thai women with breast cancer have gynecologic surveillance, especially those necessitating adjuvant therapies that might deteriorate the pre-existing gynecologic tumors.

The article’s supplementary files as