Comparison of Prophylactic Oophorectomy Specimens from Carriers and Noncarriers of a BRCA1 or BRCA2 Mutation

Abstract & commentary

Source: Stratton JF, et al. J Natl Cancer Inst 1999; 91:626-628.

Synopsis: Many of the histologic "abnormalities" described in "normal" ovaries are, in fact, variations of the normal and are not associated with the development of cancer.

The natural history of ovarian cancer is not well understood, and, to date, there is conflicting evidence as to whether there is a demonstrable precursor lesion. Some women at high risk of developing ovarian cancer because of their family history elect to have a prophylactic oophorectomy. To determine whether a recognizable premalignant lesion could be defined in familial ovarian carcinogenesis, Stratton and colleagues reviewed ovarian tissue specimens from women whose ovaries were removed prophylactically before gene testing became available, and who were tested subsequently for BRCA1 or BRCA2 gene mutations. They analyzed ovarian tissue specimens from 37 women. The specimens were examined for the presence of the following four features: inclusion cysts, clefts and fissures, ovarian epithelial metaplasia, and the presence of papillae on the ovarian surface epithelium. The specimens were also examined closely for the presence of dysplasia and occult neoplasia. Furthermore, the occurrence of endometriosis and benign ovarian tumors was documented in these women. The protein truncation test, nonradioactive single-stranded conformation polymorphism analysis, and heteroduplex analysis, followed by DNA sequencing, were used to identify BRCA1 or BRCA2 mutations in either blood samples or ovarian tissue specimens. Eleven women had inherited a mutated BRCA1 or BRCA2 gene and 26 women had not. There was no difference between these groups for any of the features studied. Stratton et al conclude that their data suggest many of the histologic "abnormalities" described in "normal" ovaries are, in fact, variations of the normal and are not associated with the development of cancer.

Comment by David M. Gershenson, MD

One of the major goals of investigators working in the area of the pathogenesis of ovarian cancer has been to define a premalignant lesion. In virtually every other common gynecologic cancer (cervical, vaginal, vulvar, and endometrial) there appears to be such an entity. The group at Mt. Sinai Hospital in New York has described "abnormalities" in the ovarian surface epithelium, such as dysplastic changes, metaplastic changes, and inclusion cysts, which, they believe, represent a premalignant condition. But this concept has never been widely accepted. One strategy has been to compare the changes in the ovarian epithelium of women at high risk for ovarian cancer with those in the ovaries of women at normal risk. A recent study from Fox Chase Cancer Center observed a higher frequency of such lesions in the ovaries prophylactically removed because of family history of ovarian cancer than in ovaries removed for other reasons.1 In that study, the pathologists were apparently unblinded to the patients’ history. In the present study from the United Kingdom, the pathologists were blinded to such information, and the study and control groups were more closely matched. It is important to note that Stratton et al found no differences in the ovarian epithelial changes between the two groups. There is at least one other unpublished large series on women who underwent prophylactic oophorectomy because of high risk of ovarian cancer in which the findings agree with those of the U.K. investigators. The findings of this study strongly suggest that we have not been able to define a precursor lesion for ovarian cancer.


1. Salazar H, et al. J Natl Cancer Inst 1996;88:1810-1820.

A premalignant condition has been defined for all the following gynecologic malignancies except:

    a. cervical cancer.
    b. ovarian cancer.
    c. endometrial cancer.
    d. vulvar cancer.
    e. vaginal cancer.