Prognostic and Predictive Biomarkers in Patients With Coronavirus Disease 2019 Treated With Tocilizumab in a Randomized Controlled Trial

Abstract

Objectives: To explore candidate prognostic and predictive biomarkers identified in retrospective observational studies (interleukin-6, C-reactive protein, lactate dehydrogenase, ferritin, lymphocytes, monocytes, neutrophils, d-dimer, and platelets) in patients with coronavirus disease 2019 pneumonia after treatment with tocilizumab, an anti-interleukin-6 receptor antibody, using data from the COVACTA trial in patients hospitalized with severe coronavirus disease 2019 pneumonia.

Design: Exploratory analysis from a multicenter, randomized, double-blind, placebo-controlled, phase 3 trial.

Setting: Hospitals in North America and Europe.

Patients: Adults hospitalized with severe coronavirus disease 2019 pneumonia receiving standard care.

Intervention: Randomly assigned 2:1 to IV tocilizumab 8 mg/kg or placebo.

Measurements and main results: Candidate biomarkers were measured in 295 patients in the tocilizumab arm and 142 patients in the placebo arm. Efficacy outcomes assessed were clinical status on a seven-category ordinal scale (1, discharge; 7, death), mortality, time to hospital discharge, and mechanical ventilation (if not receiving it at randomization) through day 28. Prognostic and predictive biomarkers were evaluated continuously with proportional odds, binomial or Fine-Gray models, and additional sensitivity analyses. Modeling in the placebo arm showed all candidate biomarkers except lactate dehydrogenase and d-dimer were strongly prognostic for day 28 clinical outcomes of mortality, mechanical ventilation, clinical status, and time to hospital discharge. Modeling in the tocilizumab arm showed a predictive value of ferritin for day 28 clinical outcomes of mortality (predictive interaction, p = 0.03), mechanical ventilation (predictive interaction, p = 0.01), and clinical status (predictive interaction, p = 0.02) compared with placebo.

Conclusions: Multiple biomarkers prognostic for clinical outcomes were confirmed in COVACTA. Ferritin was identified as a predictive biomarker for the effects of tocilizumab in the COVACTA patient population; high ferritin levels were associated with better clinical outcomes for tocilizumab compared with placebo at day 28.

Trial registration: ClinicalTrials.gov NCT04320615 NCT04363736.

Conflict of interest statement

Drs. Tom, Bao, Tsai, Carrasco-Triguero, and Cai received funding from the Biomedical Advanced Research and Development Authority (BARDA) under OT number HHSO100201800036C. Drs. Tom, Bao, Tsai, Qamra, Summers, Carrasco-Triguero, McBride, Rosenberger, Lin, Stubbings, and Cai disclosed they are employees of Genetech/Roche. Drs. Tom and Cai disclosed that they have a patent pending to Genetech for biomarkers for predicting response to an interleukin (IL)–6 antagonist (P36367-US). Drs. Bao, Tsai, and van der Leest received funding from Roche/Genetech. Dr. Bao received support for article research from BARDA. Drs. Bao and Tsai disclosed a patent pending for a method for treating pneumonia, including coronavirus disease 2019 pneumonia with an IL-6 antagonist (EFS ID 38946141). Drs. Bao and Blyth disclosed government work. Drs. Bao, Tsai, Qamra, Summers, Carrasco-Triguero, McBride, Stubbings, Haslem, and Cai disclosed the off-label product use of tocilizumab (Actemra). Dr. Qamra received funding from Hoffman-La Roche Canada. Dr. Summers received funding from Roche Products Ltd. Drs. McBride, Rosenberger, and Lin disclosed they own stock/stock options of Genetech. Drs. McBride, Rosenberger, and Cai disclosed work for hire. Dr. Stubbings is an employee of F Hoffmann-La Roche AG. Dr. Blyth received funding from Rocket Medical UK Ltd. Dr. Carratalà’s institution received funding from Gilead and Roche. Dr. François received funding from Aridis, AM-Pharma, Asahi Kasei, Inotrem, GlaxoSmithKline, Enlivex, Polyphor, Takeda, Transgene, and Biomérieux. Dr. Benfield received funding from Novo Nordisk, Simonsen, GlaxoSmithKline, Pfizer, Gilead, Lundbeck, Kai Hansen Foundation and personal fees from GlaxoSmithKline, Pfizer, Boehringer Ingelheim, Gilead, and Merck Sharp & Dohme outside the submitted work. Dr. Bonfanti received funding from Viiv, Gilead, and Janssen. Dr. van der Leest received funding from Bristol Myers Squibb, Merck Sharpe & Dohme, AbbVie, Boehringer Ingelheim, and AstraZeneca. Dr. Luyt’s institution received funding from Roche; he received funding from Correvio, Bayer Healthcare, Aerogen, ThermoFisher Brahms, Merck Sharpe & Dohme, Carmat, and Biomérieux. Dr. Luyt reports a grant from Roche to the Institute of Cardiometabolism and Nutrition, Sorbonne Université, Hôpital de la Pitié Salpêtrière, Assistance Publique - Hôpitaux de Paris for the COVACTA trial and a grant from Correvio and personal fees from Bayer Healthcare, Aerogen, ThermoFisher Brahms, Merck Sharp & Dohme, and Biomérieux outside the submitted work. Dr. Kheradmand received support for article research from the National Institutes of Health. Dr. Rosas received funding from Boehringer Ingelheim, Bristol Myers Squibb, and Immunomet. He reports a grant from Roche for the COVACTA trial and a grant and personal fees from Genentech/Roche outside the submitted work. Dr. Cai’s institution received funding from F. Hoffman-La Roche Ltd and the U.S. Department of Health and Human Services, the Office of the Assistant Secretary for Preparedness and Response. The remaining authors have disclosed that they do not have any potential conflicts of interest.

Copyright © 2021 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the Society of Critical Care Medicine and Wolters Kluwer Health, Inc.

Figures

Figure 1.

Correlation between baseline biomarkers and clinical outcomes (A) and biomarkers and baseline covariates (B). p values are based on Pearson correlation unadjusted for covariates, placebo arm only for A and for placebo and tocilizumab arms for B. *p ≤ 0.05; **p ≤ 0.01. CRP = C-reactive protein, IL-6 = interleukin-6, LDH = lactate dehydrogenase, %Lym = % lymphocytes, %Neut = % neutrophils, STRAT = stratification factor.

Figure 2.

Efficacy outcomes. Placebo-adjusted, scaled prognostic model (A) and predictive model (B). Model adjusted for the stratification factors of region (Europe, North America) and mechanical ventilation (yes, no) as well as age, sex, baseline antiviral use (yes, no), and baseline steroid use (yes, no). Vertical dashed lines indicate p = 0.05. Outcomes are at day 28 and have been aligned, so that bars pointing in the same direction have the same direction of effect. Last observation carried forward was used for ordinal scale score at day 28. CRP = C-reactive protein, IL-6 = interleukin 6, LDH = lactate dehydrogenase, %Lym = % lymphocytes, %Neut = % neutrophils

Figure 3.

Ferritin as a predictor for efficacy outcomes. A, Day 28 all-comers, B, unadjusted tertiles model-predicted probability of death to day 28, C, unadjusted tertiles model-predicted probability of mechanical ventilation to day 28 (in patients who did not receive mechanical ventilation at randomization), and D, cumulative incidence function plot of hospital discharge and death to day 28. Seven-category ordinal scale: 1, discharged or ready for discharge; 2, non-ICU hospital ward, not requiring supplemental oxygen; 3, non-ICU hospital ward, requiring supplemental oxygen; 4, ICU or non-ICU hospital ward, requiring noninvasive ventilation or high-flow oxygen; 5, ICU, requiring intubation and mechanical ventilation; 6, ICU, requiring extracorporeal membrane oxygenation or mechanical ventilation and additional organ support; and 7, death. Ferritin values were log-transformed (sample size, n = 364). Ferritin tertile cutoff values were 3.63 pmol/L (minimum), 1,480.77 pmol/L, 3,150.29 pmol/L, and 75,299.67 pmol/L (maximum) in an unadjusted model showing predicted probability and 95% CI (B, C). Cumulative incidence was determined by Aalen-Johansen estimator, with an arbitrary median cut point of 7.7 on logarithmic scale (D).

Figure 4.

Predictive biomarker assessment for the subset of patients in ordinal scale categories 4 and 5 at baseline in COVACTA (n = 157). A, Death by day 28 according to baseline ferritin levels. Ferritin values were log-transformed. Shown are predicted probabilities for death within each tertile of ferritin (tertile cutoff values were 3.63 pmol/L [minimum], 1,480.77 pmol/L, 3,150.29 pmol/L, and 75,299.67 pmol/L [maximum]) showing predicted probability and 95% CI based on an unadjusted model (calculated based on all-comers) fit using samples restricted to those for patients in ordinal scale categories 4 and 5 at baseline. B, Time to hospital discharge by baseline interleukin (IL)–6 levels. Figure shows the cumulative incidence function for time to hospital discharge based on the median IL-6 cutoff value.