Review and diagnosis of normal ovaries, their variances and neoplastic lesions. Evaluation of the technical quality of the slides and interobserver diagnostic concurrences

Abstract

A total of 407 cases were recorded under "ovarian lesions" in the department of histopathology KNH over a 5 year period, extending from 1st January 1986 to December 1990. These 407 case excluded all the lesions recorded as inflammatory . Out of ~07 recorded cases, 368 cases had either the previous histology slides retrieved or had the paraffin embedded tissue blocks retrieved and processed. Following a microscopic review of the 368 cases, 70 of these were omitted because they did not show any ovarian tissue or the required diagnosis as outlined. Two hundred and ninety eight cases we re the re fore recruited into the study Sixty percent of these were diagnosed as normal ovaries or their variances, that is corpora lutea and graafian follicles that were apparently considered abnormal ovaries and hence removed. Eight of this had no epithelial lining and were diagnosed as simple ovarian cysts. Forty percent of the cases were microscopically identified as neoplasm that are listed in the World Health Organization (WHO) classification namely epithelial tumours , germ cell tumours sex cord-stromal tumours and lymphoma Three of these neoplastic cases had two types of tumours identified, this included two cases with diagnoses of Brenner

tumour with mucinous cystadenoma and another case with a diagnosis of benign cystic teratoma and serous cystadenoma. In two of the above cases, the tissues demonstrating the two types of tumour, were submitted in separate bottles. The diagnoses of all the cases were independently made without prior knowledge of the original diagnoses. The second diagnoses were first made by the investigator and then reviewed by the supervisor before the final diagnoses were made. The second diagnoses were then compared with the original ones made by various pathologists within the department of human pathology on staff at the time when the tissues were submitted to the laboratory. Out of the slides examined 66% were found to be of reasonable technical quality, adequate for a definitive diagnosis. The remaining 34% were of a less than the desired quality and required an increased amount of time to achieve a diagnosis. There was an overall diagnostic interobserver variation of 21%