[Concordance of the Gleason score in prostate multibiopsy and definitive histology]

Due to an insufficient mean agreement between the Gleason score (GS) revealed from multibiopsy and definitive histology after radical prostatectomy (RP) of merely about 45 %, a modification of the GS including an elimination of GS 2-4 was -accomplished in 2005. The aim of the present study was to evaluate the concordance of GS and WHO grading in biopsy and definitive histology and to -determine parameters influencing the diagnostic accuracy of the biopsy and the prognosis.

Within a 10-year-period before modification of the GS, radical prostatectomy was performed in 856 patients (study group, SG; mean age 64.2 years). The grade of agreement between GS and WHO grading in biopsy and definitive histology was calculated by kappa statistics (kappa) (for the complete and single time -periods). Furthermore, we assessed the univariable and multivariable influence of different preoperatively available parameters on disease-free survival (DFS). The mean follow-up period was 39 months (range: 10-139 months).

Undergrading of GS and WHO grading decreased continuously within the three time -periods in favour of a higher agreement regarding the histological results revealed from biopsy and definitive histology. However, we found only a poor to moderate agreement in the complete time period (kappa values of 0.354 for GS and 0.404 for WHO grading) that - with regard to both parameters - was improved by an increased number of biopsy cores taken. PSA value, clinical -tumour stage, number of positive cores (dichotomised at 34 %), annual RP case load (dichotomised at 75), and GS revealed an independent significant influence on DFS. Patients with GS 2-4 in the biopsy exhibited an upgrade to GS > or = 7 in only 5.7 %, and -showed, -independent of the definitive histology, a significantly better prognosis in comparison with patients presenting with a higher GS.

The results of the present study again suggest the independent prognostic impact of the GS revealed from biopsy. However, the concordance with the GS in the definitive histology remains deficient and is improvable by taking a higher number of biopsy cores. Although the elimination of GS 2-4 might be comprehensible for the pathologist's purpose, it results in a considerable loss of pretherapeutic prognostic information.