Statistical and empirical evaluation of histopathologic reviews for quality assurance in the Eastern Cooperative Oncology Group.

Central pathology review for quality assurance in cooperative clinical cancer trials has been an accepted practice for over a decade. However, the actual value of such pathology review has never been statistically evaluated or the need defined in a comprehensive manner. Pathology exclusions in 35 completed and ongoing Eastern Cooperative Oncology Group (ECOG) trials were analyzed. Ineligibility rates ranged from 0% to 16.9%. The lowest ineligibility rates occurred in breast cancer, small cell and non-small cell lung cancer, squamous cell carcinoma of head and neck, and gastrointestinal carcinoma protocols. The highest rates occurred in rare cancers such as malignant thymoma, endocrine carcinomas, and in the sarcoma-mesothelioma area. Simulated prototypical trials involving an aggressive and an indolent cancer were examined to evaluate the precision with which treatment differences would be measured when pathologically ineligible cases were included. Analysis of these models indicated that in trials in which the pathology exclusion rate is greater than 10%, slide review is prudent. When the exclusion rate is 5% to 10%, the necessity for pathology confirmation depends on the scientific objective of the trial. If less than 5% pathology exclusions characterize a clinical trial, routine pathology review is not justified. Routine histopathologic review for quality assurance in cooperative groups is not always necessary and may be eliminated in studies of the more common cancers. The practical value of reallocating limited pathology resources for investigative studies in cooperative groups is significant.