Background: Nonalcoholic steatohepatitis (NASH) is associated with increased cardiovascular outcomes. Assessment of the impact of NASH therapy on cardiovascular risk is an important element of NASH drug development but is challenging particularly in early phase trials. Aptamer-based proteomic profiles (Somalogic®) in serum have been used to develop and validate a risk score as a surrogate for cardiovascular (CV) risk. Hypothesis: Improvement in NASH histology will result in improved proteomic cardiovascular risk scores. Methods: A post-hoc analysis of proteomic profiles of serum samples, using the Somalogic® platform, from the Pioglitazone vs. Vitamin E vs. Placebo for Treatment of Nonalcoholic Fatty Liver Disease (PIVENS) trial was conducted. PIVENS was a 96-week trial of nondiabetic participants with (NASH). We applied the proteomic CV risk scores to samples from baseline, on therapy and end of treatment visits (n=7) visits and received liver histology results at baseline and 96 weeks on N = 209 (84.6%) study participants. Generalized linear and mixed models were used to assess the association between CV risk score, treatment arm and change in liver biopsy results. Results: Baseline scores were similar across groups (mean 0.19, SD 0.14). There was no association between treatment arms and changes in scores during therapy and end of treatment. However, improvement in histological markers of activity (lobular inflammation and NAFLD activity score) and fibrosis were associated with improved cv risk scores (Figure) (p< 0.05 for all). Conclusions: Improvement in hepatic inflammation, NAFLD activity score and fibrosis were associated with improved proteomic CV risk scores regardless of treatment provided. Additional prospective validation of these findings is warranted. Proteomic profiling can potentially be used to track changes in cardiovascular risk profile changes in response to therapy in the short term NASH treatment trials.

Proteins are effector molecules that mediate the functions of genes^1,2 and modulate comorbidities^3-10, behaviors and drug treatments¹¹. They represent an enormous potential resource for personalized, systemic and data-driven diagnosis, prevention, monitoring and treatment. However, the concept of using plasma proteins for individualized health assessment across many health conditions simultaneously has not been tested. Here, we show that plasma protein expression patterns strongly encode for multiple different health states, future disease risks and lifestyle behaviors. We developed and validated protein-phenotype models for 11 different health indicators: liver fat, kidney filtration, percentage body fat, visceral fat mass, lean body mass, cardiopulmonary fitness, physical activity, alcohol consumption, cigarette smoking, diabetes risk and primary cardiovascular event risk. The analyses were prospectively planned, documented and executed at scale on archived samples and clinical data, with a total of ~85 million protein measurements in 16,894 participants. Our proof-of-concept study demonstrates that protein expression patterns reliably encode for many different health issues, and that large-scale protein scanning^12-16 coupled with machine learning is viable for the development and future simultaneous delivery of multiple measures of health. We anticipate that, with further validation and the addition of more protein-phenotype models, this approach could enable a single-source, individualized so-called liquid health check.