Introduction

Preoperative, clinical risk profiles can influence biochemical recurrence-free (BSF) and metastatic disease-free (MSF) survival-rates in prostate cancer (PCa) patients that were primarily treated with radical prostatectomy (RP) and harboured aggressive disease (i.e. upgrading/upstaging). However, it remains unclear if the same phenomenon can be observed, when contemporary risk stratification tools, namely preoperative CAPRA relative to postoperative CAPRA-S risk-scores at RP, are used.

Methods

We evaluated 17,251 PCa patients treated with radical prostatectomy and pelvic lymph node dissection at two European tertiary care referral centres between 1991 and 2016. All patients were stratified into low(LR)-, intermediate(IR)- and high-risk(HR) according to pre- and postoperative CAPRA risk-scores (≤2, 3-5 and ≥6 points, respectively). Multivariable cox regression models were fitted to assess, whether preoperative CAPRA risk-score is an independent predictor of BSF and MSF. The model was quantified using the receiver operating characteristic-derived area under the curve

Results

Overall, 38.3 (n=6584), 45.3 (n=7806) and 16.4% (n=2825) were stratified as preoperative LR, IR and HR CAPRA, respectively. Within LR patients, 14.8% (n=976) upgraded to IR or HR CAPRA-S at RP, respectively. Similarly, within IR patients, 11.7% were upgraded to HR CAPRA-S at RP (n=912). The 5-yr BSF of CAPRA-S HR patients at RP that had preoperative LR profiles was 95.8% (95% CI 95.1-96.5%), relative to 90.6% (95% CI 89.2-92.0%) of those with preoperative IR/HR profiles (p <0.001). Similarly, 5-yr MSF of HR patients at RP with preoperative LR vs IR/HR profiles was 99.7% (95% CI 99.6-99.9) vs. 98.8 (95% CI 98.3-99.4%) (p=0.001). In MVAs assessing BSF and MSF, preoperative CAPRA risk-score was an independent predictor and increased the accuracy compared to CAPRA-S risk-score at RP alone. (AUC 76.8 vs. 78.8% and 82.6 vs. 84.4%).

Conclusions

In postoperative CAPRA-S risk-scores, the additional consideration of clinically derived, preoperative CAPRA risk scores show a significant influence on BSF and MSF. Specifically, BSF and MSF is lower in postoperative CAPRA-S HR patients, who had low preoperative CAPRA risk-scores. These results warrant attention when using existing nomograms and genetic markers developed for the prediction of adverse pathologic outcomes. Finally, our results strongly suggest that postoperative CAPRA-S should be combined with preoperative CAPRA risk-score whenever applicable.

Funding

none