linia! immunology 217 (2020) 108486 Contents lists available at ScienceDisect Clinical Immunology Journal homepage: www.clsevier.comilocatelyelim Selective CD8 cell reduction by SARS-CoV-2 is associated with a worse ® prognosis and systemic inflammation in COVID-19 patients ein J.M. Urra"", CM. Cabrera”, L. Porras’, I. Rédenas* “tema, Haale Unrstero de Cal Rel, Sin “muna Facade Medina de Cade Ret Spt “Ie esc Hol General Unies de Cd Rel Sain ‘The ymphopenia exhibited inpatients with COVIDA9 hasbeen aseciated witha wort prognais in the development ofthe disease. "To understand the actors asecated witha worse evasion of COVID-19, we analyzed comedies, indicators of inflammation sich as CRP and the ratio of neutrophis/ymphoeytes, a8 well 3s the count of blo! eels with TIymphoeyte subtypes in 172 hasptlized patients with COVID-19 paeumoni. Patients were ‘Touped according to thelr needs fr mechanical venation ICU cae) of no Withia the comorbidities studied, obesity was the only associated with greater severity and ICU admission. Hoh dhe percentage aad the absolute aumber of neutrophils were higher in patients needing ICU ear than omiCU patents, whereas absolute Iynphocyte count and expecially the percentage of Iyphocyes, presented a deep deine in eicl patents. There wasn difference beeen the two group of patients for CDM T lymphocytes, neither in percentage of Imphocy noe in absolute number, however for CDS T-cells the diferences were significant for both parameters which were in decline in ICU patients. There was a firm ‘comelation between the highest values of inflammation indleatos wih the derease In percentage of CDS T-lymphocytes. This elect was not seen with CD4 cel. ‘Obey together with yaiphopena, especially whether preferentially flees to CD8 T- ymphoeytes, are factors that can predict por prognosis a patent with ovo, 1. Introduction Severe acute respiratory syndrome coronavirus (SARS-CoV’-2) which ‘causes the COVID-19 has rapidly evolved from an epidemic outbreak to ‘2 pandemic that affects virtually everyone. SARS-CoV-2 has a great similarity with to SARS-CoV and invades host humaa cells by binding to the angiotensin-converting enzyme 2 (ACE2) receptor (1). In addition the cellular serine protease TMPRSS2 is also required to properly pro ‘cess the SARS-COV-2 spike protein and facilitate host cell entry (2) ‘Although itis established that COVID-19 manifest itself mainly as an infection of the respiratory traet, COVID-19 behaves as a system disease affecting multiple organs including the gastrointestinal, cari ‘vascular, neurological, hematopoietic and immune system. SARS-CoV- 2 viremia affects the organs where ACE2 is expressed, From several days aftr the onset of symptoms, the infection becomes more systemic, affecting various organs and with a clear evidence of inflammation ‘development. Associated with the systemic involvement of the disease, the presence of lymphopenia is evident in many patients [2]. Lym. phopenia could be explained due to the direct lethal effect of SARS- ‘CoV-2 on lymphocytes, since expression of ACE2 in leukocytes has been described, although at low level [4]. Another possibility to explain lymphopenia is that the inflammation caused by the infection and the release of pro-inflammatory cytokines, such as TNF alpha and IL, could also induce apoptosis in lymphocytes (SI. This phenomenon has tbeen clearly demonstrated inthe sepsis. “The lymphopenia exhibited in patients with COVID-19 along with the rise in neutrophil leukocytes have been associated with a worse prognosis in the development of the disease. Indeed, in patients who rnceded intensive care units (ICU) and who presented acute respiratory Alistress syndrome, the Iymphocyte count levels were lower than those without these requirements [6]. Likewise lymphopenia has been asso: ciated with increased mortality and mechanical ventilation require- ments (71 ‘The objective ofthe present study was to investigate whether the subpopulations of Tymphocytes (CD4+ and CD8-+) are affected in a sreater way in Iymphopenia induced by SARS-CoV-2, as well a8 to determinate the associations with clinical features. To this end, we studied the lymphocyte populations, inflammation markers, as well as comorbidities in patients with COVID-19 pneumonia admitted in ICU tnd patients with a less severe condition (without request invasive rmechanial ventilation and without severe malti-organ involvement © Corresponding author at: immunology, Hospital General Universitario de Ciudad Real, 13005 Cludad Real, Spain. ‘mal are: jseratsescar ecmes UM. Uta} hnups://di.org/10.10167 im. 2020.108486 Received 4 May 2020; Received in revised frm 26 May 2020; Accepted 26 May 2020, ‘ailable online 29 May 2020 521-6616/ © 2020 Flsevier In. Alright reserved2, Patient selection and methods A retrospective case-control study was conducted in patents su fering from COVID-19 pneumonia admited to University Hospital of ‘Ciudad Real (Spain) from March 1 to Apel 15, 2020. total number of ‘one hundred and seventy two patients (N = 172) were investigated Hospitalized patents without acute respiratory failure (N = 145) were Included as the control group, and patients supporting mechanic ven tation at intensive care unit (CU-patiens, N= 27) represented the ‘group of cases. The infection by SARS-Cov-2 was confirmed by the real time (rt) reverse transcriptase (RT) polymerase chain reaction (PCR) from Abbott Laboratories (Abbott RealTime SARS-COV-2 assay, Abbott Park, llinois, U.S.A) from upper respiratory tract samples after ad- mission. Demographic and clinical features (comorbidities) obtained from patients medical records as well as laboratory determinations ‘were analyzed between cases and controls. The absolute number and percentage of lymphocytes and neutrophils were measured in an a: tomated analyzer of blood cell count (UniCel DxH800 Beckman Coulter Miami FL) from EDTA anticoagulated samples. For the flow eytometry assay of the lymphocyte subpopulations (Cb3+, C3 + CD4+ and (Cb3 + CDS+ lymphocytes) in peripheral blood samples the following ‘monoclonal antibodies were used: PerCP conjugated anti-human CD3, FITC ant-Cb4 and PE anti-C8 purchased from Becton Dickinson (BD San José, USA). Lymphocytes were acquired in a FACScan cytometer (BD) and analyzed with the CellQuest Pro software (BD). The pre- valence of categorical variables (gender and comorbidities) was ealcu lated by Chi? with the Fisher’ exact test, with 2 x 2 contingency tables. For the continuous variables (age at admission, laboratory findings, and lymphocyte subpopulations) the non-parametric Mann-Whitney U test was used, The influence of the parameters under study on the defined ‘groups of patients were determined by the odds ratios (ORs) with the ‘confidence intervals (Cs) at 95% in @ univariate regression analysis ‘The relationships between the percentage of Iymphocyte subpopula tions (CD3-+, CD3. + CD4+ and CD3 + CDS+) and laboratory de- terminations were analyzed by the Spearman's correlation coefficient linear regresion. All the statistical determinations were analyzed using ‘SPSS version 23.0 (SPSS Ine, Chicago Il, USA). The diferences were ‘considered statistically significant at p-values of less than 0.05. 3. Results 3.1, Glnica anaytcal and demographic characterises of patents ‘The demographic and clinical characteristics of the patients i troduced in the study are summarized in Table 1. We found differences between the groups of patients studied in terms of biochemical and blood cell parameters, in agreement with the results previously de- scribed in other series [7,5]. Asis reflected in Table 1, only the presence ‘of obesity showed significant differences between both groups of pa tients in terms of comorbidities, with an increased risk respect to the ‘control group in the univariate regression analysis performed (OR = 472, 95% CI 1.614-13.830, p = .005). The remaining co: morbidities did not show substantial differences between both groups ‘of patents, 3.2. Parameters associated with inflammation Inflammation is the clue determinant and primary basie mechanism resulting in disability and increased mortality in COVID-19 [2]. Sys- temic inflammation is associated with changes in cireulatin blood cells ‘quantity and composition. In the blood cell analysis, both the percen tage and the absolute number of neutrophils were higher in ICU pa tients, whereas on the other hand, absolute Iymphocyte count, and ‘especially the pereentage of lymphocytes, presented a deep decline in ‘critical patents (OR = 0.769, 95% Ci 0.687-0.861, p < .0001). The Yeutrophil -Lymphocyte ratio (NLR) and platele-lymphocyte ratio Canal nme 27 2020) 108486 ‘Table 1 Demographic, clnieal and analytical data of 172 patients with COVID-19 grouped by svery of illness according to ICU cae requirements. In quant tative variables the p vale was calcite by’ the nonparametric U Mann Witney test and in qualitative variables by the test Chi withthe Fisher’ exact text, with 2 2 2 contingency tables, cu pane New UW = 145) praloe waz Age ea) Goons ist S789 = 11018 Male 207419) S579) = Fens 7.59%) 1s) = Hypertension 1K) 7A(5109%) = Dates B296%) sim) = Dpstipidenia 5uss%) 5419726) 0 Obes 789%) 10(63%9) 002 com 20.4%) 15.0038) = Cartiovacub dca (148%) 240656) = Thrombote dsewe 137%) nese ts Cancer sane) 19 (110%) fe Aitimmine disse S185) ism) ™ Morty 622%) nase = Hp ome) 152 = 90 toss 40 oon DDiner(agmk) 7582 = 109222317 = sw1—< DOGOT Neuophts ssi 2 57 735 2 154 = nowt Newtopbls 103 125 = 46 Sisaa Spoon Lympbocteh ——«-7583-2 4739 71S = 290 ‘Slgmphopenla (281 39 oo Toon) Neutrophil” misma 72267 =o0001 Lnbeeyes alo Pinteew/inpbogtes 20762 26 aR STS ‘T eympbacts subpopulations Cos Meimmptecte! 603 = 12720 = 97 = ost Sms +309 70L0= 4085 0018 4 mph) 81-2 SS EDA ont Sins 22519 359 2000s (806 lymphocyte) 251 Bans ti 0009 [Abbreviations COPD, Chrone Obstructive Pulmonary Diese; CRP, Greactive Protein: ns not significa. (PLR) have emerged as excellent indicators of systemic inflammation 10}. AS is indicated in Table 1, NLR was higher in patients equicing ICU care, as well asthe biochemical indicator of inflammation, Cre active protein (CRP), demonstrating a higher rate of inflammation in the mast erties patients. 3.3. Tlymphocyte subpopulations When lymphopenia is analyzed in terms of studies of T lymphocyte subpopulations, as shown in Table 1, there was no difference berteen the two groups of patients studied for CD4 Tymphocytes, nether in percentage of lymphocyte nor in absolute number. However for CD8 T- cell the differences were significant for both parameters which are in decline in the most severe patients (Fig. 1). Alteration in D8 T Iym- phocyte population is partially transfered to the results observed in T lymphocytes and the Cb4/C8 ratio However, the decrease observed in CD8 Tells (OR = 0.956, 95% CI 0.918-0.996, p< .031) is less evident than the global data of lymphopenia. Similarly, as is shown in the univariate regression analysis (OR) in Fig. 2, the Tymphocyte populations showed that only the decrease in the CDS Iymphocyte subpopulation is associated with more severity in disease. Whether we seta random cutoff point for the absolute number of CDS cells at a value $100 cells, the univariate regression study will give us @