455 Correlation and prognostic value of pathological features, biological and clinical data before treatment, in high metastatic risk carcinoma of the prostate

Abstract

From 1987, 380 patients (pts) with high metastatic risk carcinoma of the prostate (T1/T2 with WHO grade 3, T3/T4 all grades) were treated in a randomized multicentric trial comparing curative radiotherapy alone with radiotherapy plus adjuvant LHRH analogue. The histology is currently reviewed by 2 reference pathologists. The initial PSA value, the UICC T-classification, the WHO histological grade (G) (local and review diagnostic), and the Gleason score (GS) attributed by the reference pathologist (sum of the 2 gleason grades) were investigated as potential prognostic factors for relapse free survival amongst the 175 pts already reviewed by the referees. Local control and survival were not considered in the analysis because too few events have been recorded so far. The median FU was 2 years. The potential prognostic factors were investigated first by univariate analysis, (Kaplan Meier estimate and logrank test), and subsequently in a multivariate model (Cox regression model). The median age was 70 (51–80), the median GS was 7 (2–10), 50% of the pts were G2, 30% were G3, and T1–T2 were only 8.5% of the pts. For histological grade, we found a good concordance between the local and review diagnostic (71% of concordant cases). In the univariate analysis for relapse free survival, G (local diagnostic) was predictive (<i>P</i>=0.001), G (review diagnostic) had a border line prognostic value (<i>P</i>=0.2), and GS had a significant influence (<i>P</i>=0.002). In the multivariate analysis, G only (local diagnostic) remained discriminant for relapse free survival (<i>P</i>=0.02). In conclusion this study confirmed the prognostic values of G and GS. However in the multivariate model GS did not provide any additional information to G. Surprisingly neither the PSA nor the T-class were found to be significant. This could be mainly explained by the selection of pts. The review process will be completed before the presentation and the present analysis will be accordingly updated.