| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Genitourinary Imaging |
1 From the Departments of Radiology (L.E.H., D.M.L., A.M.B., D.L.S.), Epidemiology and Biostatistics (W.S.), and Surgery (M.C.), Memorial Sloan-Kettering Cancer Center, 1275 York Ave, New York, NY 10021. Received August 11, 1999; revision requested September 27; revision received October 12; accepted October 26. Address correspondence to L.E.H.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: A database search yielded 54 patients with breast cancer and with adnexal masses at US and histopathologic examinations. Clinical, US, and histopathologic findings were correlated.
RESULTS: Forty (74%) patients had benign adnexal masses, and 14 (26%) had malignant masses; three patients had both benign and malignant ovarian masses. Seven patients had primary ovarian cancer, and seven had breast metastases to the ovary. All breast metastases to the ovary were bilateral solid masses at histopathologic examination and occurred in women with stage IV breast carcinoma at the time of US. Eleven ovaries with breast metastases were solid at US. The remaining three ovaries with breast metastases had cystic components at US because of hemorrhage or coexistent benign ovarian cysts. Four of seven patients with primary ovarian carcinoma had bilateral ovarian tumors, and seven of 11 ovarian carcinomas were predominantly cystic at US. No patient with primary ovarian carcinoma had stage IV breast cancer.
CONCLUSION: In this small series, half the ovarian malignancies in patients with breast cancer were primary ovarian carcinomas and half were breast metastases to the ovary. Breast metastases to the ovary most frequently are bilateral solid masses at US and are associated with stage IV disease at the time of US.
Index terms: Breast neoplasms, metastases, 00.32 • Ovary, neoplasms, 852.3111, 852.3112, 852.3117, 852.3119, 852.3171, 852.3172, 852.318, 852.32, 852.339 • Ovary, US, 852.12984, 852.12989
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Adnexal masses detected in patients with breast cancer are of particular concern because these patients have an increased risk of both primary and secondary ovarian neoplasms. The estimated risk of developing ovarian cancer is approximately double for all patients with breast cancer (2). In the subset of patients with a hereditary predisposition to breast cancer because of BRCA mutations, the risk of developing ovarian cancer is even greater: 15%–65% for patients with BRCA1 heterozygotes and 10%–20% for patients with BRCA2 mutations (3). In addition, the ovary frequently is a site of metastatic spread from primary breast carcinoma. Ovarian metastases are reported at autopsy in 50% of patients with breast cancer and at oophorectomy for metastatic disease in 25% of patients with breast cancer (4,5).
Investigators in prior studies have used US to screen for adnexal disease in high-risk patients, who include women with a breast cancer (6–8), but, to our knowledge, no investigators have previously addressed the relatively frequent clinical problem of the patient with breast cancer who has an adnexal mass found at US. We performed this study to correlate the US, clinical, and histopathologic findings in patients with breast cancer who underwent surgery because of adnexal masses that were evident at US.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The mean patient age was 54 years (age range, 29–84 years). Twenty-six (48%) patients were premenopausal, and 28 (52%) were postmenopausal. Four patients had undergone prior unilateral oophorectomy for benign abnormalities. Breast cancer histopathologic diagnoses were as follows: ductal carcinoma in situ (n = 8), infiltrating ductal carcinoma (n = 33), infiltrating lobular carcinoma (n = 6), medullary carcinoma (n = 3), tubular carcinoma (n = 1), and unknown type (n = 3).
US Examination
Indications for US included suspected mass (n = 42), pain (n = 5), screening (n = 4), increased abdominal girth (n = 2), and bleeding (n = 1). The mean time between breast cancer diagnosis and US was 3.5 years (range, 1 month to 15 years). The mean time between US and surgery was 3.4 weeks (range, 3 days to 24 weeks).
All US was performed by using 5.0–7.5-MHz endovaginal, 2.5–4.0-MHz sector, and/or 2.5–3.5-MHz curved-array transducers with a model XP machine (Acuson, Mountain View, Calif). Thirty-eight (70%) patients underwent both endovaginal and transabdominal pelvic US, eight (15%) underwent endovaginal US, and eight (15%) underwent transabdominal US. Twenty-three ovarian masses in 20 (37%) patients were investigated with duplex Doppler US, and resistive indexes and pulsatility indexes were calculated.
Image Analysis
US findings were determined with retrospective analysis of the images. US images were reviewed independently by two experienced radiologists (L.E.H. and D.M.L.), who had no knowledge of the histopathologic findings. The radiologists classified the adnexal masses according to morphology into six predetermined categories: nonvisualized ovary, normal ovary, cyst, complex cyst, mixed cystic and solid ovarian mass, and solid enlarged ovary.
Cysts were defined as anechoic or as having low-level echoes and/or septa of 3 mm or thinner. The size criterion for cysts in premenopausal women was a diameter of greater than 25 mm. Complex cysts had mural nodules smaller than 3 mm or had septa thicker than 3 mm. Mixed cystic and solid masses contained solid areas greater than or equal to 1 cm or contained papillary projections greater than 3 mm. Solid enlarged ovaries exceeded the normal volume for the patient's age and menstrual status. Ovarian volume was calculated by using the formula for an ellipse (length x width x height x 0.523). Normal premenopausal ovarian size was defined as ovarian volume less than 18 cm3 (9). In postmenopausal ovaries, 8 cm3 (10) was the upper limit of normal volume, with decreasing volume for older age, according to standards reported by Andolf et al (11). Ovarian volume more than twice that of the contralateral side was also considered abnormal. Discrepancies in the classification of ovarian morphology occurred for five ovaries in five patients and were resolved with consensus.
The resistive index and pulsatility index findings were recorded for the 23 masses that were investigated with Doppler US. A resistive index less than or equal to 0.4 or a pulsatility index less than 1.0 was considered abnormal. The presence of ascites also was recorded.
Clinical Findings
One of three radiologists (L.E.H., D.M.L., or A.M.B.) reviewed medical records to determine the histopathologic findings in the original primary breast cancer and to determine the stage of breast cancer at the time of US. Breast cancer stage was classified in accordance with the recommended stage grouping of the American Joint Committee on Cancer (12). The histopathologic findings in the ovarian masses were determined with review of the histopathologic reports.
Statistical Analysis
All statistical analyses were performed (W.S.) with SAS software release 6.12 (SAS Institute, Cary, NC). The histopathologic findings were correlated with the US findings, which included ovarian mass morphology, size, location, resistive index, pulsatility index, and the presence of ascites. Ovarian histopathologic findings were correlated with breast cancer histopathologic findings and with breast cancer stage at the time of US. In addition, ovarian histopathology was compared in pre- and postmenopausal women. The two-tailed Fischer exact test or Kruskal-Wallis test was performed, and a P value of less than .05 was considered to indicate a significant difference.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The histopathologic diagnoses in the adnexal masses are shown in Table 1. In 40 patients, there were 47 benign ovarian masses; simple cyst, serous cystadenoma, endometrioma, and corpus luteum were the most frequent benign diagnoses. There were bilateral masses in seven patients per Table 2, five of whom had bilateral masses with identical histopathologic findings, which included bilateral fibromas, endometriomas, simple cysts, serous cystadenomas, and corpora lutea. The other two patients had masses with different histopathologic findings. One of these patients had a right ovarian simple follicular cyst and a left ovarian hemorrhagic cyst, and the other patient had a right ovarian teratoma and a left ovarian fibrothecoma. In addition, in three patients, three benign ovarian masses (a fibrothecoma, a serous cyst, and a hemorrhagic cyst) occurred in ovaries that also had malignant disease.
|
|
The seven patients with primary ovarian cancers had eight serous cystadenocarcinomas, two mixed serous and endometrioid cancers, and one clear cell carcinoma. Four of the patients had bilateral ovarian tumors. Four patients had stage I primary ovarian carcinoma, and three patients had stage III ovarian carcinoma.
US Findings versus Ovarian Histopathologic Findings
The US morphologies of the adnexal masses are compared with histopathologic diagnoses in Table 2. Benign masses had variable US morphologies. Primary ovarian carcinomas also had variable appearances, but seven (64%) of 11 masses were complex cysts or were predominantly cystic masses with mural nodules or septa (Fig 1). In contrast, 11 (79%) of 14 breast metastases to the ovary were solid at US (Fig 2). Patients with such metastases included an 84-year-old woman with a 53-cm3 right ovary and with a 5.2-cm3 left ovary that was also enlarged for her age. Both ovaries were infiltrated diffusely with metastatic breast carcinoma. Three ovaries with breast metastases appeared cystic at US; two had coexistent benign ovarian cysts that included a fibrothecoma and a serous cyst (Fig 3). Another patient with bilateral breast metastases to the ovary had a cystic left ovary at US from hemorrhage within a lobular breast cancer metastasis; a benign hemorrhagic cyst in the right ovary was occult at US. In patients with breast cancer, there was a significant correlation between solid adnexal masses at US (solid vs simple cyst, complex cyst, and mixed cystic and solid masses) and breast metastases to the ovary (breast metastases vs benign histopathologic findings and primary ovarian cancer) (P < .001).
|
|
|
|
|
Ascites was identified in one (3%) of 40 patients with benign ovarian histopathologic findings (the three patients with coexistent benign and malignant masses in the same ovary were excluded) and in seven (50%) of 14 patients with malignant ovarian masses. Ascites was correlated with ovarian malignancy (P < .001). The frequency of ascites was similar for metastatic breast carcinoma to the ovary and for primary ovarian carcinoma. Four (57%) of seven patients with breast metastases to the ovary and three (43%) of seven patients with primary ovarian carcinoma had ascites.
Doppler US indexes were available for 23 masses in 20 patients. Seven (54%) of 13 malignant masses and four (40%) of 10 benign masses had abnormal pulsatility indexes. An abnormal resistive index was present in only one (8%) of 13 malignant ovarian masses and was present in no benign masses. Neither the resistive index nor the pulsatility index was correlated with ovarian histopathologic findings.
Clinical Findings
There was no difference in the incidence of ovarian malignancy in premenopausalversus postmenopausal women with breast cancer. Seven of 26 premenopausal women and seven of 28 postmenopausal women had malignant ovarian tumors (P = .87).
Table 3 compares ovarian histopathologic findings with the stage of breast cancer at the time of US. Three (6%) of 54 patients had an unknown breast cancer stage according to their medical records. Five (13%) of 38 patients with benign ovarian masses had stage IV breast cancer. No patient with primary ovarian cancer had stage IV breast cancer. In contrast, all patients with breast cancer metastases to the ovary had known stage IV breast cancer. Compared with the group of patients with benign ovarian masses or with primary ovarian cancer, the group of patients with breast cancer metastases to the ovary more frequently had known stage IV breast cancer at the time of US (P < .001).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Approximately half the benign ovarian lesions were neoplasms such as serous cystadenoma, mucinous cystadenoma, or teratoma, but simple cysts also were relatively frequent. It is of interest that simple cysts were of sufficient concern to the gynecologist that surgery was performed, but we found that all simple cysts were benign, even in a group of patients with breast cancer who were at high risk for malignancy. Our findings support findings in other articles (13–15), in which the reliability of US morphology for the differentiation of benign from malignant ovarian lesions is documented.
It is our policy to perform follow-up US in 6 weeks or after the next menses when adnexal masses are detected at US to eliminate physiologic cysts. This is particularly important, since the frequency of diagnosis of breast cancer is increasing in younger women (1), as illustrated in the findings of our study, in which half the women were premenopausal. It is interesting that two postmenopausal women also had corpora lutea. These women were 44 and 53 years old; both had been postmenopausal since chemotherapy 4 and 5 years previously. Premature menopause secondary to chemotherapy is a well-recognized complication of breast cancer treatment (16), but women with premature menopause, in particular those in a younger age group, may have some residual ovarian function with physiologic ovarian cysts.
In the group with malignant ovarian tumors, half the patients had primary ovarian carcinoma and half had breast metastases to the ovary. Curtin et al (5) found that 60 (50%) of 121 patients with breast cancer and with a preoperative diagnosis of pelvic mass who underwent oophorectomy had malignant disease, and that the ratio of primary ovarian carcinoma to ovarian metastases from the breast was three to one. Results of long-term follow-up of 644 patients with early-stage breast cancer have shown that nonmammary malignancies occurred in 86 (13%) and that ovarian carcinoma is the most common primary malignancy after treatment for breast cancer (17). The frequent association of breast and ovarian carcinomas has been attributed to genetic risk factors. Patients with mutations in the BRCA1 gene on chromosome arm 17q and in the BRCA2 gene on chromosome arm 13q have an autosomal-dominant predisposition to breast and ovarian carcinoma. These patients have a 66%–85% risk of developing breast cancer by the age of 70 years; the risk of ovarian carcinoma is estimated at 15%–65% for patients with BRCA1 and at 10%–20% for patients with BRCA2 mutations (3,18).
We observed that all breast cancer metastases to the ovary were bilateral at histopathologic examination and that most were solid at US. Breast cancer metastasizes to the ovary through the hematogenous route, and ovarian involvement most commonly is nodular or infiltrating (4). The hematogenous spread differs from that in Krukenberg tumors and metastases from gastrointestinal tumors that involve the ovary secondarily through the peritoneal route or from gynecologic malignancies that may invade the ovary directly from contiguous tumor involvement (19). The metastatic pattern of breast carcinoma differs in accordance with the histopathologic findings (4,20). Infiltrating lobular carcinoma more frequently involves the ovaries, the gastrointestinal tract, and the peritoneum (4). Our findings also showed a high percentage of ovarian involvement in patients with infiltrating lobular carcinoma. Although only 11% of patients in our study had infiltrating lobular carcinoma, 43% of the ovarian metastases were from infiltrating lobular carcinoma.
It has been postulated that metastases to the ovary are solid when small and become more cystic as they increase in size (21–23). Megibow et al (24) found that breast cancer metastases to the ovary have a variable appearance at computed tomography, with the cystic pattern predominating. In contrast, we found that a majority of breast cancer metastases to the ovary were solid at US, even when the metastases were large. Only one breast metastasis to the ovary had a cystic component that was secondary to hemorrhage within the tumor. The remaining two ovaries with breast metastases that appeared cystic at US had coexistent benign cysts. Because breast cancer metastases are solid, ovarian enlargement may be the only US abnormality, and metastases in normal-size ovaries may be occult at US. The US appearance of breast metastases to the ovary may be mimicked by benign ovarian fibromas that produce either unilateral or bilateral enlargement of solid ovaries.
The stage of disease should be considered when evaluating a patient with an adnexal mass evident at US and with a personal history of breast cancer. We found that all breast cancer metastases to the ovary occurred in patients with known stage IV advanced breast cancer at the time of US. In this small series of patients with adnexal masses detected at US, ovarian metastases were present in 58% of all patients with stage IV breast cancer. Our results are consistent with those in other articles in which an association between advanced breast cancer and ovarian metastases is shown (5,25). Curtin et al (5) observed that ovarian metastases occurred in five (25%) of 20 women who underwent prophylactic oophorectomy for the hormonal ablation of stage IV breast cancer.
Stage IV breast cancer and solid, bilateral, adnexal masses were the best predictors of ovarian metastases from breast carcinoma. Differentiation of primary ovarian carcinoma from ovarian metastases may be useful in clinical practice. Patients with ovarian carcinoma require extensive staging laparotomy; their disease potentially is curable if the ovarian cancer is stage I, as shown in four of our seven patients with primary ovarian cancer. In contrast, patients with breast metastases to the ovary may be treated with less invasive surgery such as laparoscopy and with medical treatment of the metastatic disease.
The retrospective design, the small sample size, and the inherent selection bias were limitations of this study. Since only patients with surgical correlation at a tertiary care cancer center were included, there may have been an overrepresentation of patients with complex masses, with large masses, or with malignant disease. Also, since we selected only patients with surgical correlation, we did not attempt to estimate the relative frequency of adnexal disease in patients with breast cancer. Such information can be obtained from only the results of large prospective screening studies. Some ovarian screening studies have been performed in patients who are considered to be at high risk for ovarian disease because of a family history of ovarian carcinoma (6–8), but to our knowledge few researchers have investigated the incidence of ovarian disease in patients with a personal history of breast cancer (8). Weiner et al (8) screened 600 patients with breast cancer prospectively with endovaginal US and found 14 patients with persistent abnormal screening examination findings, 11 of whom consented to surgery. Seven of these patients with breast cancer and with US-histopathologic correlation had benign ovarian abnormalities, three had primary ovarian carcinoma, and one had metastases to the ovaries. In that study (8), the frequency of primary ovarian carcinoma in patients with breast cancer who underwent endovaginal US screening was one (0.5%) in 200 patients.
In conclusion, a majority of patients with breast cancer who undergo surgery for adnexal masses that are evident at US have benign diagnoses, and approximately one-third have malignant disease. In this small series, half the malignant adnexal masses were primary ovarian carcinomas, and half were breast metastases to the ovary. Breast metastases to the ovary occur most frequently in patients with stage IV breast carcinoma at the time of US and manifest as bilateral, solid, adnexal masses at US.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
3 Current addresses: Department of Radiology, Brigham and Women's Hospital, Brookline, Mass.
Author contributions: Guarantor of integrity of entire study, L.E.H.; study concepts and design, L.E.H.; definition of intellectual content, L.E.H.; literature research, L.E.H.; clinical studies, L.E.H., A.M.B., D.M.L., D.L.S.; data acquisition, L.E.H., D.L.S., D.M.L., A.M.B.; data analysis, L.E.H., W.S.; statistical analysis, W.S.; manuscript preparation, L.E.H.; manuscript editing, L.E.H., A.M.B., W.S.; manuscript review, L.E.H., A.M.B., M.C., W.S.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  W.-C. Chang, M. D. Meux, B. M. Yeh, A. Qayyum, B. N. Joe, L.-m. Chen, and F. V. Coakley CT and MRI of Adnexal Masses in Patients with Primary Nonovarian Malignancy. Am. J. Roentgenol., April 1, 2006; 186(4): 1039 - 1045. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Simpkins, M. Zahurak, D. Armstrong, F. Grumbine, and R. Bristow Ovarian Malignancy in Breast Cancer Patients With an Adnexal Mass Obstet. Gynecol., March 1, 2005; 105(3): 507 - 513. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Sonmezer and K. Oktay Fertility preservation in female patients Hum. Reprod. Update, May 1, 2004; 10(3): 251 - 266. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Oktay and M. Sonmezer Ovarian tissue banking for cancer patients: Fertility preservation, not just ovarian cryopreservation Hum. Reprod., March 1, 2004; 19(3): 477 - 480. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. L. Alcazar, M. J. Galan, C. Ceamanos, and M. Garcia-Manero Transvaginal Gray Scale and Color Doppler Sonography in Primary Ovarian Cancer and Metastatic Tumors to the Ovary J. Ultrasound Med., March 1, 2003; 22(3): 243 - 247. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. L. Brown, K. H. Zou, C. M. C. Tempany, M. C. Frates, S. G. Silverman, B. J. McNeil, and A. B. Kurtz Primary versus Secondary Ovarian Malignancy: Imaging Findings of Adnexal Masses in the Radiology Diagnostic Oncology Group Study Radiology, April 1, 2001; 219(1): 213 - 218. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
