Purpose: An increasing serum prostate-specific antigen (PSA) level is the initial sign of recurrent prostate cancer among patients treated with radical prostatectomy. Salvage radiation therapy (SRT) may eradicate locally recurrent cancer, but studies to distinguish local from systemic recurrence lack adequate sensitivity and specificity. We developed a nomogram to predict the probability of cancer control at 6 years after SRT for PSA-defined recurrence. Patients and Methods: Using multivariable Cox regression analysis, we constructed a model to predict the probability of disease progression after SRT in a multi-institutional cohort of 1,540 patients. Results: The 6-year progression-free probability was 32% (95% CI, 28% to 35%) overall. Forty-eight percent (95% CI, 40% to 56%) of patients treated with SRT alone at PSA levels of 0.50 ng/mL or lower were disease free at 6 years, including 41% (95% CI, 31% to 51%) who also had a PSA doubling time of 10 months or less or poorly differentiated (Gleason grade 8 to 10) cancer. Significant variables in the model were PSA level before SRT (P < .001), prostatectomy Gleason grade (P < .001), PSA doubling time (P < .001), surgical margins (P < .001), androgen-deprivation therapy before or during SRT (P < .001), and lymph node metastasis (P = .019). The resultant nomogram was internally validated and had a concordance index of 0.69. Conclusion: Nearly half of patients with recurrent prostate cancer after radical prostatectomy have a long-term PSA response to SRT when treatment is administered at the earliest sign of recurrence. The nomogram we developed predicts the outcome of SRT and should prove valuable for medical decision making for patients with a rising PSA level.

Recent studies of surgery for cancer have demonstrated variations in outcomes among hospitals and among surgeons. We sought to examine variations in morbidity after radical prostatectomy for prostate cancer. We used the Surveillance, Epidemiology, and End Results–Medicare linked data base to evaluate health-related outcomes after radical prostatectomy. The rates of postoperative complications, late urinary complications (strictures or fistulas 31 to 365 days after the procedure), and long-term incontinence (more than 1 year after the procedure) were inferred from the Medicare claims records of 11,522 patients who underwent prostatectomy between 1992 and 1996. These rates were analyzed in relation to hospital volume and surgeon volume (the number of procedures performed at individual hospitals and by individual surgeons, respectively). Neither hospital volume nor surgeon volume was significantly associated with surgery-related death. Significant trends in the relation between volume and outcome were observed with respect to postoperative complications and late urinary complications. Postoperative morbidity was lower in very-high-volume hospitals than in low-volume hospitals (27 percent vs. 32 percent, P=0.03) and was also lower when the prostatectomy was performed by very-high-volume surgeons than when it was performed by low-volume surgeons (26 percent vs. 32 percent, P<0.001). The rates of late urinary complications followed a similar pattern. Results for long-term preservation of continence were less clear-cut. In a detailed analysis of the 159 surgeons who had a high or very high volume of procedures, wide surgeon-to-surgeon variations in these clinical outcomes were observed, and they were much greater than would be predicted on the basis of chance or observed variations in the case mix. In men undergoing prostatectomy, the rates of postoperative and late urinary complications are significantly reduced if the procedure is performed in a high-volume hospital and by a surgeon who performs a high number of such procedures.

Prostate cancer (PCA) is the most common solid malignant disease and second leading cause of cancer death in men worldwide, with over a million new cases and approximately 300 000 deaths in 2014.^1,2 The United States Preventive Services Task Force (USPSTF) guidance against prostate-specific antigen (PSA) screening and the conflicting outcomes of 2 randomized clinical trials of PSA screening have created uncertainty in the use of PSA in clinical care.^3-5 With a dramatic increase in PCA detection after the inception of PSA testing in the United States, treatment rates increased and a fall in PCA mortality ensued. Nonetheless, evidence strongly suggests that most cancers detected, especially low-grade (LG) tumors, will remain indolent for the patient’s lifetime.^6-8

Objective: Machine learning methods are flexible prediction algorithms that may be more accurate than conventional regression. We compared the accuracy of different techniques for detecting clinical deterioration on the wards in a large, multicenter database. Design: Observational cohort study. Setting: Five hospitals, from November 2008 until January 2013. Patients: Hospitalized ward patients Interventions: None Measurements And Main Results: Demographic variables, laboratory values, and vital signs were utilized in a discrete-time survival analysis framework to predict the combined outcome of cardiac arrest, intensive care unit transfer, or death. Two logistic regression models (one using linear predictor terms and a second utilizing restricted cubic splines) were compared to several different machine learning methods. The models were derived in the first 60% of the data by date and then validated in the next 40%. For model derivation, each event time window was matched to a non-event window. All models were compared to each other and to the Modified Early Warning score, a commonly cited early warning score, using the area under the receiver operating characteristic curve (AUC). A total of 269,999 patients were admitted, and 424 cardiac arrests, 13,188 intensive care unit transfers, and 2,840 deaths occurred in the study. In the validation dataset, the random forest model was the most accurate model (AUC, 0.80 [95% CI, 0.80-0.80]). The logistic regression model with spline predictors was more accurate than the model utilizing linear predictors (AUC, 0.77 vs 0.74; p < 0.01), and all models were more accurate than the MEWS (AUC, 0.70 [95% CI, 0.70-0.70]). Conclusions: In this multicenter study, we found that several machine learning methods more accurately predicted clinical deterioration than logistic regression. Use of detection algorithms derived from these techniques may result in improved identification of critically ill patients on the wards.

PURPOSE: Although models exist that place patients into discrete groups at various risks for disease recurrence after surgery for prostate cancer, we know of no published work that combines pathologic factors to predict an individual's probability of disease recurrence. Because clinical stage and biopsy Gleason grade only approximate pathologic stage and Gleason grade in the prostatectomy specimen, prediction of prognosis should be more accurate when postoperative information is added to preoperative variables. Therefore, we developed a postoperative nomogram that allows more accurate prediction of probability for disease recurrence for patients who have received radical prostatectomy as treatment for prostate cancer, compared with the preoperative nomogram we previously published. PATIENTS AND METHODS: By Cox proportional hazards regression analysis, we modeled the clinical and pathologic data and disease follow-up for 996 men with clinical stage T1a-T3c NXM0 prostate cancer who were treated with radical prostatectomy by a single surgeon at our institution. Prognostic variables included pretreatment serum prostate-specific antigen level, specimen Gleason sum, prostatic capsular invasion, surgical margin status, seminal vesicle invasion, and lymph node status. Treatment failure was recorded when there was either clinical evidence of disease recurrence, a rising serum prostate-specific antigen level (two measurements of 0.4 ng/mL or greater and rising), or initiation of adjuvant therapy. Validation was performed on this set of men and a separate sample of 322 men from five other surgeons' practices from our institution. RESULTS: Cancer recurrence was noted in 189 of the 996 men, and the recurrence-free group had a median follow-up period of 37 months (range, 1 to 168 months). The 7-year recurrence-free probability for the cohort was 73% (95% confidence interval, 68% to 76%). The predictions from the nomogram appeared to be accurate and discriminating, with a validation sample area under the receiver operating characteristic curve (ie, a comparison of the predicted probability with the actual outcome) of 0.89. CONCLUSION: A postoperative nomogram has been developed that can be used to predict the 7-year probability of disease recurrence among men treated with radical prostatectomy.