Simpkins, Fiona MD*; Zahurak, Marianna MS†; Armstrong, Deborah MD†; Grumbine, Frances MD‡; Bristow, Robert MD†
From the *National Cancer Institute, Bethesda, Maryland; †The Kelly Gynecologic Oncology Service, Department of Gynecology and Obstetrics, The Johns Hopkins Medical Institutions, Baltimore, Maryland; and ‡Division of Gynecologic Oncology, Department of Obstetrics and Gynecology, Greater Baltimore Medical Center, Baltimore, Maryland.
Reprints are not available. Address correspondence to: Fiona Simpkins, MD, National Cancer Institute, 10 Center Drive, Bldg. 10, B1-B40, Bethesda, Maryland 20892; e-mail: email@example.com.
Received June 21, 2004. Received in revised form October 10, 2004. Accepted October 27, 2004.
Breast cancer is the most common cancer in American women. In the United States 182,000 new invasive cancers are diagnosed and 43,000 women die yearly.1–3 The gynecologist, as a primary care physician for many women, plays a leading role in the diagnosis of breast cancer.4 Given the increase in incidence of breast cancer, combined with earlier diagnosis and improved therapies, more women are breast cancer survivors.5 Breast cancer treatments can result in hormonal, reproductive, and sexual changes that will increase the need for gynecologic care. Breast cancer survivors are a growing population that will be encountered frequently in the general gynecologist's practice. It is estimated that up to 22% of the patients seen by the general obstetrician–gynecologist at the Johns Hopkins Outpatient Clinic have a history of breast cancer.
An adnexal mass in a breast cancer patient is of particular concern because of the increased risk for malignant involvement of the ovary. A prior history of breast cancer is a risk factor for the subsequent development of ovarian cancer. The estimated risk of developing ovarian cancer is increased approximately 2-fold for patients with a history of breast cancer.6 Ovarian cancer is the most common second malignancy diagnosed after breast cancer.7 In addition to the increased risk of primary ovarian malignancy, the ovary is a common site of breast cancer metastasis. The lobular histologic subtype of breast cancer has a particular propensity for ovarian and abdominal spread.8
More than 480,000 oophorectomies are performed each year in the United States, making them the 11th most common surgeries performed.9 Generalist gynecologists perform the majority of these surgeries and therefore diagnose the majority of all ovarian cancers.10 The result is that the recommended complete surgical staging is often not done with the patient's first surgery.11 According to The American College of Obstetricians and Gynecologists in their recently published Committee Opinion, “Role of the Generalist in the Early Detection of Ovarian Cancer,” states that pre- and postmenopausal women with a adnexal mass with at least one of these indicators—elevated CA-125, ascites, evidence of metastasis, or a family history of breast or ovarian cancer—should be referred to a gynecologic oncologist.12 Should these recommendations also apply to the breast cancer patient with an adnexal mass? We performed a PubMed search using the dates 1969 to present and the keywords “breast cancer” and “ovarian/pelvic/adnexal mass.” A search of MEDLINE and PubMed using the keywords “breast cancer” and “ovarian/pelvic/adnexal mass” was performed from 1969 to 2004, and no study was found to guide the gynecologist in management of breast cancer patients with an adnexal mass in the absence of features suggestive of malignancy. The objectives of this study were to estimate the rate of ovarian malignancy in breast cancer patients with an adnexal mass and to identify variables predictive of malignancy.
MATERIALS AND METHODS
We conducted a computer search using International Classification of Diseases coding and discharge diagnoses using “ovarian cyst,” “ovarian neoplasm,” and “ovarian cancer” along with “history of breast cancer” as search titles. In addition, the pathology computer data base was searched using similar titles to ultimately represent a comprehensive patient population. Data base review from 1990 to 2002 was performed at John Hopkins Hospital, Johns Hopkins Bayview Medical Center, and Greater Baltimore Medical Center. This study was approved by both the Johns Hopkins Medicine Institutional Review Board and the Greater Baltimore Medical Center Institutional Review Board. All patients with a diagnosis of breast cancer (regardless of menopausal status) with an adnexal mass who underwent oophorectomy were included. Adnexal masses could be incidental or symptomatic. Anyone with preoperative evidence of disseminated malignancy was excluded. Exclusion criteria included ascites, pleural effusions, bowel obstruction, omental caking, or evidence of metastatic breast disease at the time of adnexal mass diagnosis. Women who underwent a prophylactic oophorectomy, a therapeutic oophorectomy as hormonal treatment for breast cancer, or an incidental oophorectomy were also excluded. All ovarian specimens were read by pathologists specializing in gynecologic pathology. Ovarian pathology was classified as benign, primary ovarian carcinoma, or metastatic breast carcinoma.
Factors associated with malignant ovarian neoplasms were selected based on cross tabulations and logistic regression modeling. Cross tabulations were analyzed using χ2 or Fisher exact tests where appropriate. Logistic regression models13 were used to estimate odds ratios, confidence intervals and P values for these factors. These computations were performed using the Statistical Analysis System 5 (SAS Institute, Cary, NC). LogXact software (Cytel Software Corporation, Cambridge, MA) was used when small sample size required an exact method for estimating the odds ratio. Confidence intervals (CIs) are at the 95% level and P values are 2-tailed.
This study included 129 women aged between 30 and 89 years. Patient demographics are represented in Table 1. Most patients had localized breast cancer at diagnosis (123 of 129 were stage II or less). Most patients in our study population had ductal histopathology (85%) compared with lobular (15%). Breast cancer pathology was invasive lobular or ductal carcinoma or both in 79% of patients; the remaining were in situ. Of the 75 patients for whom this information was available, 88% (n = 66) had hormone-receptor–positive breast cancer. Forty-nine patients (38%) had a family history of breast carcinoma, whereas 9 (7%) had a family history of ovarian cancer. Twenty-two patients (17%) had a breast carcinoma recurrence before the diagnosis of adnexal mass. The majority of patients (64%) were aged older than 50 years at the time of oophorectomy. Interval from breast carcinoma diagnosis to oophorectomy was a median of 3 years. A simple cystectomy or oophorectomy was performed in 32% of cases, total abdominal hysterectomy and bilateral salpingo-oophorectomy alone in 56% of cases and total abdominal hysterectomy with bilateral salpingo-oophorectomy and ovarian cancer staging procedure in the remaining 12% of cases.
The adnexal mass was about equally diagnosed incidentally or symptomatically. Incidental diagnosis frequently occurred when breast cancer patients on tamoxifen were followed up with pelvic ultrasonography to evaluate endometrial effects of tamoxifen. Patients in the symptomatic group included women with complaints of abdominal pain, pelvic pressure, bloating, or vaginal bleeding. Serum CA 125 values were available for only a limited number of patients; however, the value was normal in 73% of these cases. Ultrasound characteristics showed complex and simple adnexal masses with equal frequency (50% each). In addition, adnexal mass size was of equal frequency for masses 5 cm or less in size and more than 5 cm in size (50% each). Adnexal masses were unilateral in 83% of cases and bilateral in only 17% of cases. Approximately one half of the study patients were taking tamoxifen at the time of oophorectomy.
Adnexal masses were found to be benign in 113 of 129 cases, or 88%, (95% CI 80.6–92.7%), and malignant in 16 cases, or 12% (95% CI 7.3–19.4%). Table 2 shows the breakdown of ovarian histopathology results for the benign and malignant ovarian masses. Benign adnexal masses were primarily serous cystadenomas, functional cysts, and endometriomas. Of malignant adnexal masses, 11% were primary ovarian carcinoma, and 1.5% were metastatic breast cancer. Thus the ratio of primary ovarian carcinoma to metastatic breast cancer was approximately 7:1. Of interest in the primary ovarian cancer group, 7 cases were low malignant potential tumors, and 7 cases were invasive ovarian epithelial tumors. In the epithelial group, 5 of 7 were early stage ovarian malignancies (stages I and II). The 2 remaining epithelial cancers were stage III at the time of surgery. One of these 2 patients with stage III epithelial ovarian cancer had a known deleterious BRCA1 mutation. BRCA status was known for only 4 patients, and all but 1 patient had no BRCA mutation. The 2 adnexal masses with metastatic breast carcinoma were due to infiltrating ductal carcinoma. The malignancy rate for symptomatic adnexal masses was 15% (11/69) compared with 8% (5/59) for asymptomatic adnexal masses. The proportions malignant across hospital centers were not significantly different: 15.9% (95% CI 8.2–26.7%) at John Hopkins Hospital, 8.0% (95% CI 1.0–26.0%) at Johns Hopkins Bayview Medical Center, and 8.6% (95% CI 1.8–23.1%) at Greater Baltimore Medical Center.
In addition to ovarian malignancy, endometrial cancer was found in 2 cases. However, these patients both had benign adnexal pathology. The first patient presented with pelvic pressure, and ultrasonography showed a 10-cm adnexal mass and an endometrial stripe of 2.6 cm. This patient was diagnosed with a mucinous cystadenoma of the ovary and malignant mixed mesodermal sarcoma of the endometrium (stage IC by Federation Internationale Gynecologica Obstetrica for endometrial cancer). The patient was receiving tamoxifen at the time of diagnosis. The second patient with endometrial cancer was symptomatic with vaginal bleeding. Ultrasonography showed a 3-cm simple adnexal mass. Endometrial biopsy showed complex atypical hyperplasia. Final pathology showed ovarian serous cystadenoma and endometrial adenocarcinoma with no myometrial invasion. Table 3 shows demographic information for the patients found to have primary ovarian carcinoma or metastatic breast carcinoma in the oophorectomy specimen (endometrial cancer cases not included).
Univariate logistic regression estimates for all factors are given in Table 4. Reference group for breast cancer stage was zero. The relative risk of malignancy was increased for patients whose breast cancer was more advanced stage, although this was not statistically significant for any stage group. Stage II patients were 1.5 times more likely (95% CI 0.34–6.18) to have a malignant neoplasm compared with the stage 0 group, whereas the stage III/IV patients were 3 times as likely (95% CI 0.37–24.3) to be malignant compared with the same reference group.
A complex mass characterization was the strongest risk factor associated with this outcome. Of the 64 masses characterized as complex, 16 (25%) of them were malignant. None of the simple cysts were malignant. Complex masses were 29.2 (95% CI 4.77 to +infinity) times more likely to be malignant compared with the simple masses (P < .001). Estrogen-receptor–negative breast cancer significantly increased the probability of malignancy (OR 12.4, 95% CI 2.4–65.1; P = .003). Adnexal mass size and CA 125 status were also significant risk factors for malignancy. An elevated CA 125 level increased the risk of malignancy 6.3-fold compared with patients with normal CA 125 (95% CI 1.3–31.5; P = .02). Adnexal mass size greater than 5 also significantly increased the risk of malignancy (OR 4.6, 95% CI 1.2–17.3; P = .02). Patients aged older than 50 years had an increased risk for malignant masses (OR 2.7, 95% CI 0.7–9.9) but this was not significant (P = .14). Tamoxifen therapy was associated with a nonsignificant decrease in the risk of malignancy (OR 0.4, 95% CI 0.12–1.5; P = .17).
In this study group, a complex adnexal mass was found in 42% of benign cases, and 100% of the malignant cases. Size more than 5 cm was found in 45% of benign cases and 80% of the malignant masses. Elevated CA 125 was found in 24% of the benign adnexal masses and 63% of the malignant ones. When combining the 2 variables, complex mass and size more than 5cm, 22% were benign and 86% were malignant. The 3 major identified variables of complex mass, size more than 5 cm, and elevated CA 125 were found in 0.03% of benign cases and in 50% of malignant cases. Of note, CA 125 values were attainable for only 55 patients. Furthermore, because of the small number of cancer cases found, a multivariate logistic regression analysis was not possible.
Adnexal masses are not uncommon in women with a history of breast cancer. The challenge is to determine which features are suggestive of malignancy and which suggest a high likelihood of a benign pathology. The results of this study indicate that breast cancer patients diagnosed with an adnexal mass most likely have a benign ovarian cyst; 88% of adnexal masses were benign, and 12% were malignant. Risk of malignancy in patients with asymptomatic adnexal masses was 8% compared with 15% in the symptomatic patient subset. This is consistent with findings by Goff et al14 that patients with ovarian cancer have more abdominal, pelvic, and urinary complaints than those patients with benign pelvic masses.
In the group with a malignant ovarian neoplasm, primary ovarian carcinoma was more common than metastatic breast carcinoma, occurring in a ratio of 7:1. A complex mass by ultrasonography, size more then 5 cm, and an elevated CA 125 were 3 variables found to be statistically significant predictors of malignancy in breast cancer patients with an adnexal mass (P < .05). We conclude that when any of these 3 high-risk features are present, the patient should be referred to a gynecologic oncologist.
The present study has some limitations. Because of the small number of cancer cases found, a multivariate logistic regression analysis was not possible. In addition, data were not available for variables such as serum CA 125 (n = 55) and breast cancer hormone receptor status (n = 75) on many patients. Although all ovarian pathology specimens were read by pathologists specialized in gynecology, it was beyond the scope of this investigation to perform a retrospective review of breast cancer pathology specimens. Finally, the patient population identified in this study is heavily weighted to hormone-receptor–positive breast cancer patients. This may be due to the increased use of pelvic imaging in women taking tamoxifen. Thus, the finding in this study that women with hormone-receptor–negative breast cancer are at higher risk for malignant involvement of an adnexal mass must be viewed with caution.
We are not aware of any study to date that provides clear guidance to the gynecologist in management of breast cancer patients with an adnexal mass. Earlier diagnosis of breast cancer and better overall prognosis mean that there are a growing number of breast cancer survivors. The general gynecologist will be increasingly involved in the care and management of this growing patient population. The American College of Obstetrics and Gynecology recommends that a patient with an adnexal mass who has a first-degree relative with breast or ovarian cancer should be referred to the gynecologic oncologist for management.12 It is notable that only 5 of the 16 patients with malignant adnexal masses met this criterion. There are currently no definitive recommendations for management of the breast cancer patient with an adnexal mass.
Three prior studies have evaluated breast cancer patients with an adnexal mass. In the first study, Curtin et al5 found a 50% malignancy rate in a study arm of 121 breast cancer patients with a preoperative diagnosis of an adnexal mass. In this study, the ratio of primary ovarian cancer to metastatic breast cancer was 3:1. However, 14 patients (12%) had ascites identified preoperatively, suggesting malignancy. Thus, the rate of malignancy may have been overrepresented. In the second study, Hann et al15 found a malignancy rate of 26% in 54 breast cancer patients with an adnexal mass. Of the malignant masses, 50% were primary ovarian carcinoma. In this study, 8 patients (15%) had ascites, again suggesting a diagnosis of disseminated malignancy preoperatively. Finally, Quan et al16 demonstrated that the majority of stage IV breast cancer patients with a pelvic mass have metastatic breast cancer (68%), and only a smaller percentage of patients have primary ovarian malignancy (10%).
The current study is distinct from the above studies in that it addresses the adnexal mass in breast cancer patients in the absence of other evidence of disseminated abdominal malignancy. The health care needs of the breast cancer patient are complex, and decision making requires a multidisciplinary approach with consideration of multiple factors. Based on the results of this study, we created an algorithm to assist the clinician in management of the adnexal mass in the breast cancer patient (Fig. 1). When a breast cancer patient without other evidence of disseminated malignancy is diagnosed with an adnexal mass, CA 125 should be measured, and transvaginal ultrasonogram characteristics should be evaluated. A patient with a simple, 5 cm or smaller adnexal mass and a normal serum CA 125 has a low risk of malignancy. However, patients with an adnexal mass with complex architecture, size more than 5 cm, or an elevated CA 125 have a high risk of malignancy and should be referred to a gynecologic oncologist.
1. American Cancer Society. Cancer facts & figures—1999. Atlanta (GA): American Cancer Society; 1999.
2. Parker SL, Tong T, Bolden S, Wingo PA. Cancer statistics. CA Cancer J Clin 1997;47:5–27.
4. Breast diseases. In: DiSaia, PJ, Creasman, WT. Clinical gynecologic oncology. 6th ed. St. Louis (MO): Mosby; 2002. p. 385–422.
5. Curtin J, Barakat R, Hoskins W. Ovarian disease in women with breast cancer. Obstet Gynecol 1994;84:449–52.
6. Lingeman CH. Etiology of cancer of the human ovary: a review. J Natl Cancer Inst 1974;53:1603–18.
7. Rosen PP, Groshen S, Kinne DW, Hellman S. Nonmammary malignant neoplasms in patients with stage I and stage II breast carcinoma. Am J Clin Oncol 1989;12:169–74.
8. Lamovec J, Bracko M. Metastatic pattern of infiltrating lobular carcinoma of the breast: an autopsy study. J Surg Oncol 1991;48:28–33.
9. Rutkow I. Surgical operations in the United States. Arch Surg 1997;132:983–90.
10. McGowan L. Patterns of care in carcinoma of the ovary. Cancer 1993;71:628–33.
11. Munoz KA, Harlan LC, Trimble EL. Patterns of care for women with ovarian cancer in the United States. J Clin Oncol 1997;15:3408–15.
12. The role of the generalist obstetrician-gynecologist in the early detection of ovarian cancer. ACOG Committee Opinion No 280. American College of Obstetricians and Gynecologists. Obstet Gynecol 2002;100:1413–6.
13. Cox DR. The analysis of binary data. London (UK): Methuen; 1970.
14. Goff BA, Mandel LS, Melancon CH, Muntz HG. Frequency of symptoms of ovarian cancer in women presenting to primary care clinics. JAMA 2004;291:2705–12.
15. Hann LE, Lui DM, Shi W, Bach AM, Selland DL, Castiel M. Adnexal masses in women with breast cancer: US findings with clinical and histopathologic correlation. Radiology 2000;216:242–7.
16. Quan ML, Fey J, Eitan R, Abu-Rustum NR, Barakat R, Borgen P, et al. Role of laparoscopy in the evaluation of the adnexa in patients with stage IV breast cancer. Gynecol Oncol 2004;92:327–30.