Search

Predicting COVID-19 severity is difficult, and the biological pathways involved are not fully understood. To approach this problem, we measured 4,701 circulating human protein abundances in two independent cohorts totaling 986 individuals. We then trained prediction models including protein abundances and clinical risk factors to predict adverse COVID-19 outcomes in 417 subjects and tested these models in a separate cohort of 569 individuals. For severe COVID-19, a baseline model including age and sex provided an area under the receiver operator curve (AUC) of 65% in the test cohort. Selecting 92 proteins from the 4,701 unique protein abundances improved the AUC to 88% in the training cohort, which remained relatively stable in the testing cohort at 86%, suggesting good generalizability. Proteins selected from different adverse COVID-19 outcomes were enriched for cytokine and cytokine receptors, but more than half of the enriched pathways were not immune-related. Taken together, these findings suggest that circulating proteins measured at early stages of disease progression are reasonably accurate predictors of adverse COVID-19 outcomes. Further research is needed to understand how to incorporate protein measurement into clinical care.

Squamous cell lung cancer (SqCLC) is a type of non-small cell lung cancer strongly associated with cigarette smoking and with known targetable mutations, however some patients do not have matching targeted drugs. The S1400I Phase III randomized LungMap sub-study of nivo+ipi versus nivo accrued 275 (252 eligible) previously-treated patients with stage IV SqCLC and absence of matched mutations (NCT02154490).Here, to investigate the longitudinal (baseline, C2 week 3, C4 week 7, C5 week 9) serum protein changes associated with treatment or response, we performed Olink proximity extension assay in 160 patients (561 samples) using an immuno-oncology panel of 92 analytes. We utilized mixed linear and joint models to identify differentially expressed proteins between treatment arms (nivo vs. nivo+ipi), response and quantify the effect of other potential prognostic factors. This approach quantifies the effect of covariates such as smoking status, demographics, prior radiation therapy, and metastases. We jointly modeled expression and survival to identify changes in proteins over time. The thresholds for significance in differentially expression tests were a log fold change of at least 0.5 and a false discovery rate of under 0.05 (FDR as a multiple testing adjustment method). The joint models used thresholds of log fold change of more than 1 and hazard ratio of more than 1. Results revealed increases in 40 serum proteins after treatment with either nivo alone or combined with ipi, including CXCL9, CXCL10, CXCL11, CXCL13, IL6, IL8, IL10, IFNg, and soluble PD-L1 and PDCD1. nivo+ipi treatment showed greater increases in IL8 and CXCL13 post-treatment compared to nivo alone. In addition, circulating IL6, CXCL13, and MIC-A/B were higher in patients with stable or progressive disease compared to those with objective response after treatment. The joint model revealed a worse hazard ratio with increases in 10 soluble proteins, including IL6, IL8, and CXCL13. Finally, for patients with similar values of IL8, those treated by combination had a better survival outcome than those treated by nivo alone. Using a data-modeling approach, we identified significant longitudinal changes in 40 serum proteins out of 92 tested in patients with SqCLC treated with nivo or nivo+ipi. Increases in serum inflammatory markers IL6 and CXCL13 were associated with worse disease. Although overall survival was not statistically different between arms, our modeling approaches suggested that IL-8 monitoring may help identify patients benefiting more from the combination vs. nivo alone. Citation Format: Edgar Gonzalez-Kozlova, Hsin-Hui Huang, Mary Redman, Roy Herbst, Scott Gettinger, Luda Bazhenova, Hui Xie, Manishkumar Patel, Kai Nie, Jocelyn Harris, Kimberly Argueta, Karen Kelly, Ethan Cerami, James Lindsay, Joyce Yu, Roshni Biswas, Stephen Van Nostrand, Radim Moravec, Diane Marie Del Valle, Seunghee Kim-schulze, Sacha Gnjatic. Dynamic changes in circulating protein levels reveal an association between ipilimumab and nivolumab combination treatment (SWOG Lung-MAP S1400I trial) with outcomes in squamous cell lung cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 5026.

9044 Background: While Immune checkpoint blockade (ICB) is standard treatment for lung cancer there are limited biomarkers that predict benefit, pharmacokinetic on-treatment activity or explain progression. S1400I was a randomized Phase III trial of nivolumab(N)+ipilimumab(I) versus N (NCT02154490, PMID 34264316) for ICB naïve, previously treated stage IV or recurrent squamous cell lung cancer. We performed circulating serum protein analysis of serial blood specimens from patients enrolled in S1400I to evaluate if serum proteins levels changed over time, changes differed by treatment arm, and if they were associated with overall survival. Methods: 561 serial blood specimens (baseline, weeks (wk) 3, 7, 9, and progression [PD]) from 160 of 252 eligible patients enrolled to S1400I were analyzed for 92 immuno-oncology analytes with the Olink proximity extension assay. Protein levels were normalized with use of internal controls and quantified as log2 protein expression (denoted as NPX). Linear mixed models evaluated change in expression from baseline at each time point (wks 3,7,9 and PD), and NPX differences at baseline, wk3, and PD depending on best objective response. A Cox model was used to evaluate the association between baseline NPX and survival. Overall survival and longitudinal cytokine expression were jointly modeled using a linear mixed model to estimate dynamic biomarker changes in NPX and a Cox model for survival. The joint models for time-varying NPX values included a random intercept and modeled time using a natural spline with three knots. Significance was defined as P < 0.05. Results: Serum proteins PCDC1, CXCL9, and CXCL10 were increased from baseline at wks 3,7,9 and at PD. CCL19 was increased at wks 3 and 7 but not at wk 9 and PD. IL10 and IFNγ were increased at wk 3 but subsequently returned to baseline. Change in CXCL13 from baseline to PD was larger for N+I versus N. Baseline CCL23, CSF-1, IL6, and MUC-16 were associated with shorter survival (HR > 1). Joint modeling of survival with cytokines showed an increased risk of death (HR > 1) with higher longitudinal serum levels of CXCL13, MMP12, CSF-1, and IL8. Patients achieving objective response had higher IL4 and LAMP3 and lower IL6 and IL8 at baseline and wk 3 compared to non-responders. Conclusions: Measurements of blood circulating soluble proteins represent easily accessible biomarkers that may be useful as indicators of outcome, and that will need to be prospectively confirmed. Clinical trial information: NCT02154490.