Hoedt, Amanda M. De

Abstract Background: Missing risk factors, whether random or not measured at all, across different hospitals and patients provide challenges, both for the developers of online clinical risk tools and the patients trying to use such tools. This paper provides a development and end-user solution to the commonly encountered limitations of clinical risk tool based decision making due to missing information.Methods: Six state-of-the-art logistic regression approaches accommodating missing data were compared using prostate cancer data from ten North American and European cohorts from the Prostate Biopsy Collaborative Group (PBCG). An additional large European PBCG cohort was withheld for external validation, where calibration-in-the-large (CIL), calibration curves, and area-underneath-the-receiver-operating characteristic curve (AUC) were evaluated. Ten-fold leave-one-cohort-internal validation further validated the optimal missing data approach.Results: Among 12,703 biopsies from 10 training cohorts, 3,597 (28%) had clinically significant prostate cancer, compared to 1,757 of 5,540 (32%) in the external validation cohort. In external validation, the available cases method that pooled individual patient data containing all risk factors input by an end-user had best CIL, under-predicting risks as percentages by 2.9% on average, and obtained an AUC of 75.7%. Imputation had the worst CIL (-13.3%). The available cases method was further validated as optimal in internal cross-validation and thus used for development of an online risk tool. For end-users of the risk tool, two risk factors were mandatory: serum prostate-specific antigen (PSA) and age, and ten were optional: digital rectal exam, prostate volume, prior negative biopsy, 5-alpha-reductase-inhibitor use, prior PSA screen, African ancestry, Hispanic ethnicity, first-degree prostate-, breast-, and second-degree prostate-cancer family history.Conclusion: Developers of clinical risk prediction tools should optimize use of available data and sources even in the presence of high amounts of missing data and offer options for users with missing risk factors.

4 Background: Management of patients with a postoperative rising prostate-specific antigen (PSA) level is complex. Additional local treatment such as salvage radiation therapy (SRT) may be sufficient for many patients but some may require concurrent systemic therapy in order to delay or prevent metastatic disease. As PSA recurrence on its own is a poor surrogate for metastatic disease we hypothesized that the Decipher genomic classifier (GC), a validated predictor of metastasis may be able to better distinguish those patients where additional therapy is beneficial from those where SRT on its own is likely sufficient. Methods: Genomic classifier (GC) scores were calculated from 170 prostate cancer patients, who received SRT at the Veteran Affairs Medical Center Durham, Thomas Jefferson University and Mayo Clinic, between 1990 and 2010. SRT was defined as administration of RT with Pre-RT PSA levels > 0.2 ng/ml. GC and CAPRA-S scores were compared using survival c-index, competing-risks and Cox regression analysis for the prediction of metastasis. Results: Survival c-index for predicting metastasis 5 years post SRT was 0.85 (95% CI: 0.73-0.88) for GC and 0.63 (95% CI: 0.49-0.77) for CAPRA-S. The cumulative incidence of metastasis at 5 years post-SRT was 2.7%, 8.4%, and 33.1% for low, average, and high GC scores (p < 0.001) and 16.9%, 2.3% and 17.2% for low, average and high CAPRA-S scores (p = 0.113). In univariable analysis only GC, extraprostatic extension, path GS and Pre-RT PSA were significant predictors of metastasis. In multivariable analyses with clinical risk factors or the CAPRA-S risk model, GC was the only independent predictor of metastasis with a HR of 1.63 (1.22-2.18, p < 0.001) for a 10% unit increase in risk score. Conclusions: In patients treated with postoperative SRT for PSA recurrence, GC is a powerful predictor of metastasis. Patients with low Decipher have excellent prognosis with SRT and may avoid concurrent hormonal therapy. Patients with high Decipher risk are at highest risk for metastatic disease and SRT failure and may benefit from intensified systemic therapy.