Standardized AgNOR analysis as a prognostic parameter in endometrial carcinoma, endometrioid type

OBJECTIVE: To evaluate interobserver,er reproducibility of histologic grade in endometrial adenocarcinomas of endometrioid type (EC), to assess the re lationships between nuclear grade and the amount of argyrophilic nucleolar organizer region (AgNOR) proteins and to determine the prognostic value of AgNOR proteins and the main clinicopathologic parameters. STUDY DESIGN: Architectural and nuclear grading were independently assessed by two pathologists in 64 formalin-fixed, paraffin-embedded surgical sampl es of EC obtained from all equal number of patients (age range, 38-84 years ; mean, 63.5). Interobserver agreement was determined using the kappa stati stic; discrepant cases were reviewed, and a consensus was reached. Standard ized AgNOR analysis teas performed according to the guidelines of the Commi ttee on AgNOR Quantification, measuring the mean area of AgNORs per nucleus (NORA) by an image analysis system. RESULTS: The kappa values for interobserver agreement were substantial for architectural grading and moderate for nuclear grading. When NORA values we re compared to the nuclear grade assessed by different observers, the most significant linear correlation (r = .723, P < .001) was found for the nucle ar assessment obtained by consensus of the two pathologists. Moreover, stat istical analysis allowed discrimination of architectural grade 1 from grade 2 and 3 EC. By the Kaplan-Meier method, the prognosis was worse for patien ts with higher NORA values (> 4.212 mum(2)), while, by Cox multivariate ana lysis, AgNOR quantity emerged as an independent prognostic variable. CONCLUSION: Use of standardized AgNOR analysis may be an additional and obj ective tool in the assessment of histologic grade as well as a reliable met hod of determining prognosis in EC.