Lymphocyte subsets early predict mortality in a large series of hospitalized COVID-19 patients in Spain.

The role of lymphocytes and their main subsets as prognostic factors of death in SARS-CoV-2-infected patients remains unclear, with no information obtained from patients outside China. We aimed to assess whether measuring lymphocyte subpopulations added clinical value to the total lymphocyte counting regarding mortality when they were simultaneously tested at hospital admission. Peripheral blood was analysed in 701 polymerase chain reaction (PCR)-confirmed consecutive patients by lysed-no washed flow cytometry. Demographic and clinical features were registered in electronic medical records. Statistical analysis was performed after a 3-month follow-up. The 112 patients who died were older and had significantly higher frequencies of known co-morbidities than survivor COVID-19 patients. A significant reduction in total lymphocytes, CD3+ , CD4+ , CD8+ and CD19+ counts and CD3+ percentage was found in the group of deceased patients (P < 0·001), while the percentage of CD56+ /CD16+ natural killer (NK) cells was significantly higher (P < 0·001). Multivariate logistic regression analysis showed a significantly increased risk of in-hospital death associated to age [odds ratio (OR) = 2·36, 95% confidence interval (CI) = 1·9-3·0 P < 0·001]; CD4+  T counts ≤ 500 cells/µl, (OR = 2·79, 95% CI = 1·1-6·7, P = 0·021); CD8+  T counts ≤ 100 cells/µl, (OR = 1·98, 95% CI = 1·2-3·3) P = 0·009) and CD56+ /CD16+ NK ≥ 30%, (OR = 1·97, 95% CI = 1·1-3·1, P = 0·002) at admission, independent of total lymphocyte numbers and co-morbidities, with area under the curve 0·85 (95% CI = 0·81-0·88). Reduced counts of CD4+ and CD8+ T cells with proportional expansion of NK lymphocytes at admission were prognostic factors of death in this Spanish series. In COVID-19 patients with normal levels of lymphocytes or mild lymphopenia, imbalanced lymphocyte subpopulations were early markers of in-hospital mortality.