Go, Ronald S.

PURPOSE Time to event is an important aspect of clinical decision making. This is particularly true when diseases have highly heterogeneous presentations and prognoses, as in chronic lymphocytic lymphoma (CLL). Although machine learning methods can readily learn complex nonlinear relationships, many methods are criticized as inadequate because of limited interpretability. We propose using unsupervised clustering of the continuous output of machine learning models to provide discrete risk stratification for predicting time to first treatment in a cohort of patients with CLL. PATIENTS AND METHODS A total of 737 treatment-naïve patients with CLL diagnosed at Mayo Clinic were included in this study. We compared predictive abilities for two survival models (Cox proportional hazards and random survival forest) and four classification methods (logistic regression, support vector machines, random forest, and gradient boosting machine). Probability of treatment was then stratified. RESULTS Machine learning methods did not yield significantly more accurate predictions of time to first treatment. However, automated risk stratification provided by clustering was able to better differentiate patients who were at risk for treatment within 1 year than models developed using standard survival analysis techniques. CONCLUSION Clustering the posterior probabilities of machine learning models provides a way to better interpret machine learning models.

AbstractHemolysis is an unintended sequel of temporary or permanent intracardiac devices. However, limited data exist on the characteristics and treatment of hemolysis in patients with cardiac prostheses. This entity, albeit uncommon, often poses significant diagnostic and management challenges to the clinical cardiologist. In this article, we aim to provide a contemporary overview of the incidence, mechanisms, diagnosis, and management of cardiac prosthesis‐related hemolysis.

Abstract We retrospectively reviewed the utility of N-terminal prohormone of brain natriuretic peptide (NT-proBNP) and transthoracic echocardiogram (TTE) in diagnosing cardiac involvement in patients with biopsy-proven systemic immunoglobulin light chain amyloidosis seen at the Mayo Clinic between 1 January 2006 and 30 December 2015. We analyzed 2 cohorts: patients undergoing endomyocardial biopsy for suspicion of cardiac involvement (cohort 1) and patients who had serum NT-proBNP and comprehensive echocardiographic evaluation at diagnosis (cohort 2). Of 179 patients undergoing endomyocardial biopsy (cohort 1), 173 (97%) had evidence of amyloid deposition, with 159 having NT-proBNP performed at the time of the procedure. The NT-proBNP was elevated (>300 pg/mL) in all 159 patients (sensitivity, 100%; median NT-proBNP, 4917 pg/mL; range, 355-69 541). The left ventricular ejection fraction, interventricular septal thickness, and strain rate were abnormal in 89/168 (53%), 102/64 (61%) and 92/95 (97%), respectively. Among cohort 2 (n = 342), 259 (76%) had an elevated NT-proBNP, of whom 237 (92%) had an abnormality detected on TTE. Of 83 patients with normal NT-proBNP <300 pg/mL, 27 (33%) had an abnormality on TTE (all with borderline strain rate −18% to −15%). Only 5/27 patients were considered to have possible early cardiac involvement and none had any other diagnostic or classical features of amyloidosis on TTE. The combination of NT-proBNP and comprehensive echocardiographic evaluation can diagnose cardiac amyloidosis negating the need for endomyocardial biopsy. A negative NT-proBNP rules out clinically meaningful cardiac involvement and may obviate the routine use of TTE in patients with a low clinical suspicion of cardiac amyloidosis.