The article by Geisler et al.1 in a recent issue of Cancer explored an important area of inquiry, namely the correlation between mismatch repair (MMR) genes and microsatellite instability (MSI) in ovarian carcinomas. We thought it important to comment herein on some aspects that are germane to clinical relevance.
One important issue is the accurate estimation of the frequency of MMR gene mutations. The use of protein truncation test (PTT) as a mutation detection strategy, rather than full gene sequencing, leads to a much lower estimation of mutation frequency2 because sequence variants not leading to protein truncation are missed. Another limitation is the obtaining of normal DNA from ovarian snap-frozen tissue, which could lead to contamination with tumor cells as well as limitations in DNA quality.3 To our knowledge, the proportion of hereditary ovarian carcinomas resulting from hereditary nonpolyposis colorectal carcinoma (HNPCC) remains poorly defined,4 and large-scale, population-based ovarian carcinoma studies are needed to definitively answer this question.
A second important consideration is the stage distribution of the patient sample. For example, National Cancer Insitute-designated comprehensive cancer centers often overrepresent patients with more advanced disease. HNPCC carriers may have an improved prognosis, possibly related to an early stage of disease at the time of presentation.5 Among 80 HNPCC-associated ovarian carcinomas studied by Watson et al.,5 84% were Stage I or Stage II, which is significantly different from the rate of 30% observed in the general population. Therefore, if the study population was not representative of the true frequency of early-stage ovarian carcinoma, there is likely to be underrepresentation of patients with a hereditary predisposition.
A third design issue that may have resulted in the underestimation of MMR in ovarian carcinoma was the inclusion of both peritoneal and fallopian tube carcinomas. Of the 125 tumors examined in the study, only 107 were ovarian tumors, and the remainder were either primary tumors of the peritoneum or fallopian tube. Although ovarian carcinoma is within the spectrum of HNPCC-associated tumors, in which the high-frequency MSI phenotype is a hallmark feature, primary peritoneal carcinoma or fallopian tube carcinoma are not.
Future studies should include epidemiologic risk factors, information regarding family history, and detailed data concerning clinical characteristics of the study population. The ultimate clinical significance of MSI in ovarian carcinoma will require correlation of MSI and MMR expression with survival. Although the report by Geisler et al.1 provides valuable data regarding the association between MSI and MMR genes, further studies clearly are warranted.