OBJECTIVES: To validate the Shared Equal Access Regional Cancer Hospital (SEARCH) nomogram to better risk stratify men with intermediate-risk pathology after prostatectomy (positive surgical margins, PSM, and/or extracapsular disease, ECE, without seminal vesicle or lymph node involvement) in a tertiary referral centre (the Duke Prostate Center, DPC). PATIENTS AND METHODS: We retrospectively analysed 485 men in the DPC cohort with PSM and/or ECE but without seminal vesicle or lymph node involvement. The predicted risk of biochemical progression-free probability at 1, 3 and 5 years was estimated by the SEARCH and updated Kattan postoperative nomograms. Calibration plots were generated and accuracy assessed with the concordance index. RESULTS: The SEARCH nomogram appeared to be well calibrated, with the highest-risk quartile having a predicted <60% progression-free probability at 5 years, vs >80% for the lowest risk. In comparison, overall external calibration appeared to be similar for the updated Kattan nomogram, although there was less separation between the highest- and lowest-risk quartiles. The SEARCH model had an overall predictive accuracy of 0.65, which compared favourably with the updated Kattan nomogram (0.57). CONCLUSION: In an external dataset, the SEARCH nomogram to predict progression-free probability for men at intermediate risk after prostatectomy was well calibrated and performed better than the updated postoperative Kattan nomogram.
OBJECTIVES: To validate the Shared Equal Access Regional Cancer Hospital (SEARCH) nomogram to better risk stratify men with intermediate-risk pathology after prostatectomy (positive surgical margins, PSM, and/or extracapsular disease, ECE, without seminal vesicle or lymph node involvement) in a tertiary referral centre (the Duke Prostate Center, DPC). PATIENTS AND METHODS: We retrospectively analysed 485 men in the DPC cohort with PSM and/or ECE but without seminal vesicle or lymph node involvement. The predicted risk of biochemical progression-free probability at 1, 3 and 5 years was estimated by the SEARCH and updated Kattan postoperative nomograms. Calibration plots were generated and accuracy assessed with the concordance index. RESULTS: The SEARCH nomogram appeared to be well calibrated, with the highest-risk quartile having a predicted <60% progression-free probability at 5 years, vs >80% for the lowest risk. In comparison, overall external calibration appeared to be similar for the updated Kattan nomogram, although there was less separation between the highest- and lowest-risk quartiles. The SEARCH model had an overall predictive accuracy of 0.65, which compared favourably with the updated Kattan nomogram (0.57). CONCLUSION: In an external dataset, the SEARCH nomogram to predict progression-free probability for men at intermediate risk after prostatectomy was well calibrated and performed better than the updated postoperative Kattan nomogram.