Ovarian pathology in risk-reducing salpingo-oophorectomies from women with BRCA mutations, emphasizing the differential diagnosis of occult primary and metastatic carcinoma.

摘要

Risk-reducing salpingo-oophorectomy (RRSO) is an effective prophylactic procedure for women with mutations in BRCA1 or BRCA2 genes, both of which confer an increased lifetime risk for ovarian, tubal, peritoneal, and breast cancer. In addition to lowering this risk, RRSO also offers the opportunity to detect occult early-stage fallopian tube or ovarian carcinoma. The differential diagnosis of occult tubal/ovarian cancer includes a spectrum of benign tubal and ovarian alterations and also occult metastatic breast cancer, although only rare cases of the latter have been reported in RRSO. Neoadjuvant breast cancer chemotherapy may contribute to diagnostic difficulty due to treatment-induced cytologic alterations. With the aim of elucidating features which may help with differential diagnosis, this study reports the incidence and pathologic features of benign ovarian alterations, benign ovarian tumors, and occult primary and metastatic malignancies in prophylactic oophorectomies from 108 women with a BRCA mutation and from 35 women with other strong risk factors for hereditary breast/ovarian carcinoma. We direct particular emphasis on morphologic features of primary ovarian lesions that may mimic occult metastatic breast cancer. We also evaluate histologic alterations due to neoadjuvant breast cancer chemotherapy in the ovary and fallopian tube of patients who received such treatment immediately preceding RRSO. Comparison is made to ovarian metastases of breast cancer in our hospital-based population of breast cancer patients, none of whom underwent RRSO. Overall, 69% of RRSO patients had a personal history of breast cancer. Neoadjuvant breast cancer chemotherapy was administered in 15%. Occult primary carcinoma occurred in 7 (6.5%) BRCA patients (5 in fallopian tube, 1 in fallopian tube and ovary, 1 in ovary). Ovarian metastasis of breast cancer occurred in 1 (1%) BRCA patient undergoing RRSO and in up to a similar proportion (0.8%) of the hospital-based population of breast cancer patients. The metastasis in the RRSO patient was clinically occult, unilateral, 0.2 cm, and demonstrated mild atypia without mitoses. Abundant foamy, vacuolated cytoplasm due to neoadjuvant chemotherapy exposure was notable. In contrast, ovarian metastases in the non-RRSO population were all clinically detected, bilateral, large, and exhibited well-developed malignant cytologic features. None of the normal cell types in the ovary or tube demonstrated any cytologic alterations in RRSO patients who received neoadjuvant chemotherapy. The main morphologic mimics of metastasis with superimposed chemotherapy-induced alterations in RRSO were stromal hyperthecosis (n=8), nodular hyperthecosis (n=2), adrenal rests (n=3), hilus cell nodules (n=43), and hilus cell hyperplasia (n=4). Occult primary ovarian carcinoma was reliably distinguished from ovarian metastases of breast cancer by WT-1+, p53+, mammaglobin-, GCDPF-immunoprofile. These results demonstrate that evaluation of RRSO specimens requires awareness of a spectrum of ovarian lesions which may mimic occult primary or metastatic carcinoma; awareness of the masquerading effects of neoadjuvant chemotherapy; and awareness of the potential morphologic differences between occult metastatic breast cancer in RRSO and non-RRSO specimens.