BackgroundCarbohydrate antigen 125 (CA125) is an indicator of inflammation, immune response, and impaired cardiac function. The aim was to investigate whether CA125 behaves as a biomarker of severity and poor clinical outcomes in hospitalized patients with coronavirus disease 2019 (COVID-19).MethodsSerum CA125 [Elecsys CA125 II assay-(Roche Diagnostics GmbH)] was measured in stored biobank samples from COVID-19 hospitalized patients between 01 March 2020 and 17 October 2021. Multiple logistic regression models were built to explore the association between CA125 and clinical outcomes [in-hospital all-cause mortality, need for invasive mechanical ventilation (IMV), or non-invasive respiratory support (non-IRS)], estimating odds ratios (ORs; 95% CI). The gradient of risk of CA125 was evaluated by fractional polynomials.ResultsA total of 691 patients were included, median age of 63 years (50–76), men (57.2%), with high comorbidity. At admission, 85.8% had pneumonia. Median CA125 was 10.33 U/ml (7.48–15.50). The in-hospital mortality rate was 7.2%. After adjusting for confounding factors, CA125 ≥ 15.5 U/ml (75th percentile) showed an increased risk of death [OR 2.85(1.21–6.71)], as age ≥ 65 years, diabetes, and immunosuppression. Furthermore, CA125 as a continuous variable was positive and significantly associated with the risk of death after multivariate adjustment. The mean hospital stay of the patients with CA125 ≥ 15.5 U/ml was longer than the rest of the study population.ConclusionCA125 in the first 72 h of hospital admission seems a useful biomarker of mortality in hospitalized patients with moderate–severe COVID-19. If our findings are confirmed, the wide availability of this biomarker would make easy its widespread implementation in clinical practice.