CD4 and CD8 Lymphocyte Counts as Surrogate Early Markers for Progression in SARS-CoV-2 Pneumonia: A Prospective Study

Abstract

Background: COVID-19 pathophysiology and the predictive factors involved are not fully understood, but lymphocytes dysregulation appears to play a role. This paper aims to evaluate lymphocyte subsets in the pathophysiology of COVID-19 and as predictive factors for severe disease.

Patient and methods: A prospective cohort study of patients with SARS-CoV-2 bilateral pneumonia recruited at hospital admission. Demographics, medical history, and data regarding SARS-CoV-2 infection were recorded. Patients systematically underwent complete laboratory tests, including parameters related to COVID-19 as well as lymphocyte subsets study at the time of admission. Severe disease criteria were established at admission, and patients were classified on remote follow-up according to disease evolution. Linear regression models were used to assess associations with disease evolution, and Receiver Operating Characteristic (ROC) and the corresponding Area Under the Curve (AUC) were used to evaluate predictive values.

Results: Patients with critical COVID-19 showed a decrease in CD3+CD4+ T cells count compared to non-critical (278 (485 IQR) vs. 545 (322 IQR)), a decrease in median CD4+/CD8+ ratio (1.7, (1.7 IQR) vs. 3.1 (2.4 IQR)), and a decrease in median CD4+MFI (21,820 (4491 IQR) vs. 26,259 (3256 IQR)), which persisted after adjustment. CD3+CD8+ T cells count had a high correlation with time to hospital discharge (PC = -0.700 (-0.931, -0.066)). ROC curves for predictive value showed lymphocyte subsets achieving the best performances, specifically CD3+CD4+ T cells (AUC = 0.756), CD4+/CD8+ ratio (AUC = 0.767), and CD4+MFI (AUC = 0.848).

Conclusions: A predictive value and treatment considerations for lymphocyte subsets are suggested, especially for CD3CD4+ T cells. Lymphocyte subsets determination at hospital admission is recommended.

Keywords: CD4+ T cells; COVID-19; lymphocyte subsets; predictive value; severity.

Figure 1

Boxplot showing association of CD3+CD4+ count (top-left), CD4+MFI (top-right), and CD4+/CD8+ ratio (bottom-left) with COVID-19 evolution. p-values were derived from the F-test of a linear model.

Figure 2

Dotplots showing association of stay length in hospital with: CD4+MFI in COVID-19 patients that did not change non-critical clinical status (CD4 + MFI); and CD3+CD8+ T cell count in patients who reached critical condition during their hospitalization (CD3+CD8+ count). Association was assessed using a linear model controlling by age, gender, and time from symptoms onset. p-values were derived from a F-test of the linear model. Association was measured using Partial Correlation Coefficients (PCC) and their 95% confidence intervals. Dot lines represent the slope estimated by the linear model. Values are displayed after a-priori correction by the rest of covariates in the model.

Figure 3

Receiver Operating Characteristics (ROC) and their corresponding Area Under the Curve (AUC) assessing the ability of CD3+CD4+T cell count (CD3+CD4+A), CD4+MFI (CD4_MFI) and CD4+/CD8+ ratio (RATIO48) to predict COVID-19 critical clinical evolution. AUC intervals at 95% confidence were computed using bootstrap. Total accuracy, sensitivity, and specificity are displayed for the optimal threshold, defined as the ROC point closest to the top-left part of the plot (perfect sensitivity and specificity).