key: cord-0834807-xqusmudh authors: Arulkumaran, Nishkantha; Snow, Timothy A.C.; Kulkarni, Adarsh; Brealey, David; Rickman, Hannah M.; Rees-Spear, Chloe; Spyer, Moira J.; Heaney, Judith; Garr, Edmund; Williams, Bryan; Cherepanov, Peter; Kassiotis, George; Lunn, Michael P.; Ambler, Gareth; Houlihan, Catherine; McCoy, Laura E.; Nastouli, Eleni; Singer, Mervyn title: Influence of IL-6 levels on patient survival in COVID-19 date: 2021-09-16 journal: J Crit Care DOI: 10.1016/j.jcrc.2021.08.013 sha: 1e940d86ea344118d0be8030e74e4d4f22eb1e77 doc_id: 834807 cord_uid: xqusmudh nan IL-6 levels are significantly higher among patients who develop critical illness or who subsequently die compared to patients with mild illness (p < 0.001). Levels of IL-6 on admission correlated well with CRP (r 2 = 0.398; P < 0.001). On hospital admission, IL-6 discriminated between eventual survivors and non-survivors (AUROC 0·67, p = 0.020) (Fig. 1) . Admission levels of IL-6 had better discriminatory value for development of critical illness or death compared to patients with mild illness (AUROC 0.78, p < 0.001). We found a good correlation between serum IL-6 and CRP; IL-6 being a key regulator of C-reactive protein (CRP) production. However, co-interventions, particularly the use of corticosteroids, affect CRP levels. This may explain the lack of association between the treatment effect of Tocilizumab with baseline CRP in clinical trials [5] . Furthermore, a significant proportion of our patients had IL-6 levels that were only marginally elevated, consistent with other studies. [6] No clinical Table 1 Baseline demographics, seroconversion rates and treatments used for patients subdivided by World Health Organisation (WHO) COVID-19 ordinal severity scale. WHO 4-5 (hospitalized with or without supplemental oxygen via nasal prongs or face mask), WHO 6-9 (hospitalized requiring non-invasive or invasive respiratory support), and WHO 10 (died following hospital admission). Ct value (cycle threshold) represents viral load, with higher values representing a lower viral load. (hospitalized with or without supplemental oxygen via nasal prongs or face mask (mild disease)), WHO 6-9 (hospitalized requiring non-invasive or invasive respiratory support (critical illness)), and WHO 10 (died following hospital admission). (a) IL-6 levels are significantly higher among patients who develop critical illness or who subsequently die compared to patients with mild illness (p < 0.001). (b). Levels of IL-6 on admission correlate well with CRP (r 2 = 0.398; P < 0.001). (c-d) Admission levels of IL-6 had discriminatory value for development of critical illness or death compared to patients with mild illness (AUROC 0.78, p < 0.001). In comparison, the ability of IL-6 to discriminate between survivors and non-survivors was less effective (AUROC 0.67, p = 0.020). Comparison of continuous data between groups was performed using the Kruskal Wallis. Pearsons correlation is used to assess correlation between IL-6 and CRP levels. Area under the receiver operator curve (AUROC) was constructed to ascertain the predictive value of cytokines for mortality. trial investigating the use of IL-6 receptor antagonists have stratified patients based on circulating levels of IL-6 levels [5] . Personalized medicine has struggled to find a foothold in critical care trials. Patient stratification by theragnostic biomarker levels will allow clinicians to identify those who may specifically benefit from treatment, and avoid possible harm in patients where treatment is likely to be futile. The funding source had no role in data collection, analysis, or interpretation; trial design; patient recruitment; or any aspect pertinent to the study. There has been no payment from a pharmaceutical or other company in the design of the study or writing of the manuscript. All authors have had access to the data in the study. This database has been used for previous publications, although addressing different hypotheses with different analyses, figures and tables: • Sex differences in immunological responses to COVID-19: A crosssectional analysis of a single centre cohort. Brit J Anaes. In press. • Therapeutic targets in COVID-19: A cross-sectional analysis of a single centre cohort. Crit Care Explorations. Accepted for publication All authors had access to data. NA, TACS, EN, and MS designed the study. TACS, AK, DB, HR, CRS, MoS, JH, EG, CH, BW, ML, PC, GK, LEM, and EN acquired study data. NA and TACS analysed the data. NA and TACS wrote the manuscript. MeS critically reviewed the manuscript. NA, TACS, and MeS have accessed, and verified, the data in this article. All authors approved the final version. MS reports grants and advisory board fees from NewB, grants from the Defence Science and Technology Laboratory, Critical Pressure, Apollo Therapeutics, advisory board and speaker fees (paid to his institution) from Amormed, Biotest, GE, Baxter, Roche, and Bayer, and honorarium for chairing a data monitoring and safety committee from Shionogi. Characterization of the inflammatory response to severe COVID-19 illness Immunomodulators in COVID-19: two sides to every coin Tocilizumab in patients admitted to hospital with COVID-19 (RECOVERY): a randomised, controlled, open-label, platform trial Effect of tocilizumab on clinical outcomes at 15 days in patients with severe or critical coronavirus disease 2019: randomised controlled trial Tocilizumab in COVID-19: a meta-analysis, trial sequential analysis, and meta-regression of randomized-controlled trials Cytokine elevation in severe and critical COVID-19: a rapid systematic review, meta-analysis, and comparison with other inflammatory syndromes We thank Chris Wilson and Atul Goyale at UCLH Biochemistry for their assistance with patients' sera identification.