The total percentage of biopsy cores with cancer improves the prediction of pathological stage after radical prostatectomy

OBJECTIVE: To examine whether the simple variable 'percentage of cancer-positive biopsy cores' is a significant predictor of true pathological stage after radical prostatectomy and can be used to improve pathological stage prediction by simple means. PATIENTS AND METHODS: In all, 375 patients had a radical prostatectomy for localized prostate cancer in two UK centres; 260 had complete preoperative staging information. Logistic regression was used and predicted probability graphs constructed to assess predictors of pathological stage. RESULTS: In this study, only PSA (P = 0.004) and percentage cancer-positive biopsy cores (P < 0.001) were significant predictors of pathological stage. The final model was an acceptable classifier for pathological stage (area under the receiver operating characteristic curve 0.76, specificity 85%, sensitivity 47%). A patient with a PSA of 10 ng/mL and one of six cores positive for cancer would have a predicted probability of extraprostatic disease of 20%, whereas the same patient with all six biopsy cores positive would have a predicted probability of extraprostatic disease of 80%. CONCLUSIONS: The percentage of cancer-positive biopsy cores significantly predicts the disease stage after radical prostatectomy. This variable is easy to obtain by the clinician and avoids the need to estimate the percentage of biopsy tissue infiltrated by cancer. This readily available information can easily be computed and may help to counsel patients about realistic expectations of organ-confined disease in relation to surgery as a treatment option.