Utilization of Proteomic Measures for Early Detection of Drug Benefits and Adverse Effects.

Abstract

Recognition of benefits and adverse effects of therapies in earlier clinical trial phases could improve the safety, efficiency, and cost of clinical trials. Using four clinical trials representing a diverse set of diseases and drug classes (EXSCEL: exenatide/GLP-1 RA, SUGAR-DM-HF: empagliflozin/SGLT2i, PRADA: epirubicin/anthracycline, and AMPLE: abatacept/immunomodulator and adalimumab/TNF inhibitor), we hypothesized that previously validated proteomic measures for cardiometabolic outcomes could enable the detection of beneficial and adverse drug effects in fewer participants over a shorter follow-up period. Changes in SomaSignal^TM proteomic tests over time in response to treatment were assessed in the EXSCEL (baseline vs 1 year; once-weekly exenatide (EQW) (n) = 1812 vs control (n) = 1787), SUGAR-DM-HF (baseline vs 12 weeks and 36 weeks; empagliflozin (n) = 45 vs control (n) = 52), AMPLE (baseline vs 85 days and 1 year; abatacept (n) = 210, adalimumab (n) = 222), and PRADA (baseline vs 7-10 days and 3 months, n = 120) trial. Improvement of cardiovascular risk and cardiometabolic traits with EQW was detectable within 1 year (P = .002) in sample sizes significantly smaller than the original study. Cardio- and kidney-protective (P = .06, P = .037) effects of empagliflozin were detectable within 36 weeks in a small sample size (n ∼ 50). Abatacept and adalimumab treatment demonstrated significant improvements in cardiovascular risk (P ≤ .001, P ≤ .001) and cardiorespiratory fitness (P ≤ .001, P ≤ .001) within 85 days. In contrast, anthracycline treatment led to significant increases in heart failure mortality risk (P ≤ 0.001) and cardiovascular risk (P = .004) after the first cycle of chemotherapy treatment. These findings provide preliminary evidence that proteomics may provide a powerful tool for optimizing drug pipelines by predicting the effects of novel therapeutics in smaller, shorter trials.