Rethinking Gleason pattern quantification in predicting metastasis: results of 20 years of follow‐up in the Rotterdam section of the European Randomized Study of Screening for Prostate Cancer

TL;DR

This study shows that the amount of Gleason pattern 3 in prostate cancer biopsies does not predict metastasis when the amounts of more aggressive patterns are known.

Contribution

The study reveals that Gleason pattern 3 length does not affect metastasis prediction when absolute lengths of patterns 4 and 5 are considered.

Findings

Absolute lengths of Gleason patterns 4 and 5 are significantly associated with metastasis-free survival.

The presence of invasive cribriform/intraductal carcinoma improves the model's discriminative ability.

Gleason pattern 3 length is not a significant predictor of metastasis-free survival.

Abstract

The Gleason grading system for prostate cancer (PCa) is based on the proportions of Gleason patterns (GP) 3–5. While pure GP3 has minimal metastatic potential, it is unclear whether GP3 quantity in the presence of GP4 and GP5 affects oncological outcomes.

To assess the predictive value of PCa biopsy GP lengths on long‐term metastasis‐free survival (MFS).

Prostate biopsies of 1,881 men with screen‐detected PCa who participated in the Dutch part of the European Randomized Study of Screening for Prostate Cancer (ERSPC) between 1993 and 2007 were revised for GP 3–5 length. Multivariable Cox regression analyses were used to evaluate the relationship between GP lengths and MFS truncated at 20 years, adjusting for clinical‐tumour stage (cT), prostate‐specific antigen (PSA), percentage positive biopsies and the presence of invasive cribriform/intraductal carcinoma (CR/IDC).

On multivariable analysis, ≥cT2, PSA, percentage positive cores and absolute length of GP4 and GP5 were all significantly associated with MFS. The discriminative ability was improved by adding CR/IDC to the model. Total GP3 length was neither associated with MFS in the model with (hazard ratio [HR] 0.99, 95% confidence interval [CI] 0.97–1.00, P = 0.3) nor without CR/IDC (HR 0.98, 95% CI 0.96–1.01, P = 0.2). A limitation is the lack of targeted biopsies.

GP3 length does not have an impact on the prediction of MFS in biopsies, once GP4/GP5 lengths are known. Although GP3 percentage is essential in Gleason grading, MFS is related to absolute GP4 and GP5 quantity rather than their proportion to GP3.

Example of prostate biopsy absolute lengths and proportions, and their impact on ISUP Grade Group assessment. While the absolute length of Gleason pattern 4 in all three pictured biopsies is the same (3 mm), the absolute length of Gleason pattern 3 differs, which leads to counter‐intuitive proportion‐based Grade Groups. Metastasis‐free survival is related to absolute Gleason patterns 4 and 5 quantity rather than their proportion to Gleason pattern 3.

Figures