Testing of two binary grading systems for FIGO stage III serous carcinoma of the ovary and peritoneum

A variety of histologic grading systems for ovarian carcinoma have been used, but there is no widely accepted system. Binary grading systems are inherently superior to the more common three-grade systems because they are more reproducible and they correspond to the number of options in the binary treatment decision for which grade is considered important: the use of or withholding of chemotherapy. One hundred thirteen unselected FIGO stage III serous carcinomas of the ovary and peritoneum were tested with two grading systems: a binary system recently proposed by investigators at MD Anderson Cancer Center (MDACC) and a new binary system we formulated at the Washington Hospital Center (WHC). Both of these systems are based on nuclear grade. The WHC system has a higher threshold of nuclear size for diagnosing high-grade tumors. The WHC system separated the cases into 89 high-grade and 24 low-grade tumors. The median survival rates were 30 and 49 months for high and low grade respectively, and the actuarial survival curves were not significantly different ( P > 0.10). The MDACC system separated the cases into 103 high-grade and 10 low-grade tumors. With this system, low-grade tumors were significantly more likely than high grade to be stage IIIA ( P < 0.05) and occurred at a mean age of 57 years compared to 65 years for high-grade tumors ( P < 0.05). Low-grade tumors were suboptimally debulked in 10% of cases compared to 27% for high-grade tumors ( P > 0.05). The median survival for high-grade tumors was 34 months, and the median for low grade has not been reached. The actuarial survival curves were not significantly different ( P = 0.065). The MDACC grading system appears more promising than the WHC system. The MDACC system separates a small (9% of advanced stage serous carcinomas) but distinctive well-differentiated tumor which usually has the appearance of invasive low-grade (micropapillary) serous carcinoma. The rarity of this tumor, however, will require a larger series to demonstrate prognostic value. The WHC system, which was designed to enlarge the low-grade group to a size that would be more meaningful in clinical practice, did not demonstrate a survival difference. The failure of the WHC system suggests that attempts to enlarge the low-grade group using histologic features alone are unlikely to be successful. The potential for confounding of grade with substage, volume of residual disease and patient age are issues that may impede determination of th