Title

Prostate Biopsy and Radical Prostatectomy Gleason Score Correlation in Heterogenous Tumors: Proposal for a Composite Gleason Score.

Document Type

Article

Publication Date

9-1-2015

Publication Title

The American journal of surgical pathology

Abstract

When prostate biopsy cores are separately identified in multiple containers, current recommendations are to grade each specimen individually. For treatment algorithms, the highest Gleason score (HGS) is typically used as the overall score, even if a lower score predominates. This practice has the potential to misrepresent the overall cancer in the entire gland for some patients and place them in a higher-grade group. We compare a novel composite Gleason score (CGS), integrating grade patterns from contiguous positive biopsy sites, with HGS to determine correlation with the radical prostatectomy (RP) Gleason score (GS). One hundred needle biopsy cases from 2008 to 2012 with >2 GSs in a biopsy set (eg, 3+3=6, 3+4=7, and 4+3=7) or more than a 1-step difference in GS (eg, 3+4=7 and 4+4=8 without 4+3=7) were analyzed. Grades were assigned using both methods (HGS and CGS) and compared with RPGS. Grade groups I to V were used to define downgrade and upgrade. Comparing HGS with RPGS, 31% remained the same and 69% had a change in GS (87% downgraded and 13% upgraded). Comparing CGS with RPGS, 59% remained the same and 41% had a change in GS (10% downgraded and 90% upgraded). Of the 2 methods, the CGS showed better overall correlation with RP (P2 grades are present in a biopsy set. CGS has a significantly lower rate of downgrade and predicts the RPGS more accurately than HGS.

Medical Subject Headings

Adult; Aged; Algorithms; Biopsy, Large-Core Needle; Humans; Lymphatic Metastasis; Male; Middle Aged; Neoplasm Grading; Neoplasm Staging; Neoplasm, Residual; Predictive Value of Tests; Prostatectomy; Prostatic Neoplasms; Reproducibility of Results; Treatment Outcome; Tumor Burden

PubMed ID

26274028

Volume

39

Issue

9

First Page

1213

Last Page

1218

Share

COinS