INTRODUCTION AND OBJECTIVE:
Under current guidelines, radical prostatectomies (RP) for unfavorable intermediate risk prostate cancer (UIR PCa) patients are deferred . However, high risk criteria (pT3-4/pN1 stage and/or GGG IV-V) will be present in some unfavorable intermediate risk prostate cancer (UIR PCa) patients. These individuals should be prioritized for definitive therapy. We attempted to identify them.
A total of 2,024 UIR PCa patients treated with radical prostatectomy were abstracted from an institutional database (2000-2018). Multivariable logistic regression models predicting high risk criteria (pT3-4/pN1 stage and/or GGG IV-V) in IR PCa were developed, validated and directly compared with NCCN IR PCa stratification.
All 2,024 patients were randomly divided between development (n=1,012, 50.0%) and validation cohorts (n=1,012, 50.0%). The development cohort was used to fit a new model (PSA level, clinical T stage, biopsy Gleason Grade Group (GGG), cumulative tumor length in available biopsy cores [all p<0.001]) to predict high-risk criteria according to RP pathology, using. At RP, the rates of pT3-4/pN1, GGG IV-V and of both were 52.2, 3.7 and 55.9%, respectively. Within the validation cohort, the new model outperformed any combination of NCCN UIR PCa criteria in accuracy (70.1 vs 63.0%), calibration and decision curve analyses. The new model derived probability cutoff of 50.0% for presence of pT3-4/pN1/GGG IV-V identified 317 (31.1%) of 1,012 patients, in whom the rate of pT3-4/pN1/GGG IV-V was 65.9%, instead of average 44.6%.
We developed a new model that can accurately identify a subset of NCCN UIR PCa patients with significantly higher proportion of high-risk criteria (pT3-4/pN1 stage and/or GGG IV-V). These individuals should be prioritized for definitive therapy during COVID-19-pandemic.
Source of Funding:
There was no external financial support for this study. The authors declare that they have no conflict of interest