Early lymphopenia as a predictor of COVID-19 outcomes: A multicenter cohort study.

Introduction: Lymphopenia is recognized as a biomarker for predicting outcomes in coronavirus disease (COVID-19). However, the optimal timing for its observation remains uncertain. We investigated the association between early lymphopenia and COVID-19 prognosis, as well as the relationship between lymphocyte count trends and disease outcomes.

Methods: We analyzed data from the J-RECOVER study, a multicenter retrospective cohort study in Japan, encompassing patients with COVID-19 between January and September 2020. The patients were categorized into lymphopenia (LP) (<800 cells/μL) and non-lymphopenia (NL) (≥800 cells/μL) groups based on the lymphocyte counts between days 1 and 4 post-onset. They were further divided into "persistent," "recovered," "exacerbated," and "stable" groups based on lymphocyte counts between days 7 and 10. The primary outcome was the in-hospital mortality. The Cox proportional hazard regression was used for the analysis.

Results: Of 995 enrolled patients, 212 patients (21.3%) were classified into the LP group. LP was significantly associated with in-hospital mortality (hazard ratio [HR] 2.32, [95% CI 1.39 to 3.87], -value 0.001). In both the LP and NL groups, lower lymphocyte counts between 7 and 10 days-categorized as the "persistent" and "exacerbated" groups-was associated with in-hospital mortality (HR 4.65, [95% CI 2.07 to 10.47], -value <0.001, and HR 5.59, [95% CI 2.24 to 13.97], -value <0.001, respectively).

Conclusions: Early lymphopenia is predictive of poor prognosis in patients with COVID-19. A declining lymphocyte count trend post-onset further indicates disease deterioration.