Background
Quantitative Gleason grading appears to be a reliable prognostic parameter and provides broader risk stratification then the traditional Gleason grading in patients with prostate cancer (PCa) treated with radical prostatectomy (RP).
Objective
To determine if quantification of Gleason pattern (GP) 4 for targeted and systematic biopsy (TBx + SBx) cores together with further clinical variables can identify the lowest quantitative GP 4 fraction on RP.
Design, setting, and participants
A total of 548 patients underwent TBx + SBx of the prostate and then RP, with pathology revealing Gleason score 3 + 4, 4 + 3, or 4 + 4 disease.
Intervention
TBx + SBx of the prostate followed by RP.
Outcome measurements and statistical analysis
GP 4 fraction thresholds of ≤5%, ≤10%, ≤15%, ≤20%, and ≤25% were compared between the TBx + SBx and RP specimens. The sensitivity, specificity, negative predictive value (NPV), positive predictive value (PPV), and accuracy for predicting the GP 4 fraction in the RP specimen were determined. Logistic regression models were used to establish a probabilistic relationship between various combinations of clinical and biopsy variables and the GP 4 fraction in the RP specimen.
Results and limitations
GP 4 fractions of ≤5%, ≤10%, ≤15%, ≤20%, and ≤25% was observed in 33%, 49%, 58%, 65%, and 70% of patients on TBx, and 18%, 41%, 53%, 63%, and 70% of patients on RP, respectively. The sensitivity, specificity, NPV, PPV, and accuracy were 75%, 67%, 91%, 39%, and 74% for a TBx GP 4 fraction of ≤5%, and 65%, 85%, 65%, 85%, and 79% for a TBx GP 4 fraction of ≤25%, respectively. A model combining quantified TBx + SBx GP 4 with clinical parameters demonstrated the highest diagnostic accuracy. Limitations include the retrospective study design.
Conclusions
Our results demonstrate that the combination of MRI-TBx + SBx and GP 4 quantification allowed precise detection of a low fraction of GP 4 when using RP specimens as the reference standard. Moreover, we found that clinical variables including Prostate Imaging-Reporting and Data System score without biopsy are limited in detection of low GP 4 fractions.