Abstract
Background The 2019 novel coronavirus (COVID-19) emerged in Wuhan in December 2019 and has since rapidly spread across China and to the globe. This virus shows unusually high transmission rate and unique clinical characteristics, with key pathological mechanism remaining unclear. Here, we analysed the laboratory data based on clinical samples from COVID-19 patients, in parallel comparison with non-COVID-19 pneumonia cases, in an attempt to elucidate the key pathological features of COVID-19 during its infection of the human body.
Methods We analysed biochemical indices and lymphocyte subpopulation distribution in the samples of COVID-19 infected patients, and compare the data with those from the pneumonia cases not caused by COVID-19. The study enrolled 125 patients, comprising 103 COVID-19-infected patients and 22 non-COVID-19 pneumonia cases. Laboratory findings were extracted from electronic medical records. Blood was collected for lymphocyte subsets and cytokines determination by flow cytometry analysis. The differences in biochemical indices and lymphocyte subsets distribution between COVID-19-infected and non-COVID-19 pneumonia patients were analyzed. Correlation analysis was performed between leukocyte subgroups count and biochemical indexes relating to organ injury in COVID-19 infected patients (exclude asymptomatic infected cases).
Results We observed significant differences, including abnormal biochemical indices (CRP, LDH, AST, eGFR and sodium ion concentration) and reduced lymphocyte subsets count, between the COVID-19-infected and non-COVID-19-caused pneumonia cases. Correlation analysis indicates that the count for lymphocyte subsets-but not that for neutrophils and monocytes-exhibits a significant negative correlation with biochemical indices relating to organ injury, in the COVID-19 infected patients.
Conclusions The present study indicates significantly different clinical features between COVID-19-caused and non-COVID-19-caused pneumonia, especially in terms of lymphocytopenia and organ injury. Notably, correlation analysis demonstrates that tissue damage in COVID-19 patients is attributed to virus infection itself rather than uncontrolled inflammatory responses (“cytokine storm”). These findings provide new insights for developing efficient therapeutic strategies against COVID-19 infection.
Introduction
Since its initial outbreak in the Chinese city Wuhan in early December 2019, 2019 novel coronavirus (COVID-19 or 2019-nCoV) has spread throughout China and globally [1], with the total number of reported cases approaching 100,000. Analysis of clinical data available is in urgent demand for revealing the mechanisms of tissue damage and further devising more specific and efficient treatment strategies.
The COVID-19 virus is a novel beta coronavirus based on the data from gene sequencing [2]. Compared with other beta coronaviruses, such as SARS-CoV and MERS-CoV, COVID-19 infection leads to significant differences in clinical symptoms, according to the latest studies [3], including remarkably higher occurrence of non-fever patients, fewer upper respiratory and gastrointestinal symptoms, a different profile of increased blood cytokines, more frequent multi-organ injury, and lower death rate [3-5]. These differences, along with its unusual high transmission rate, suggest a different pathological mechanism underlying its infection of the human body.
In this study, we analyzed the clinical and laboratory data of most of the COVID-19-infected patients in the China city Nanjing, aimed at revealing the unique pathological features of COVID-19 cases in comparison with non-COVID-19-caused pneumonia cases. Importantly, through correlation analysis of the changes in blood cell subgroups and key blood biochemical indexes of tissue injury, we have discovered unexpected clues of the pathogenic process following COVID-19 infection, which may provide valuable information for designing efficient approaches of treatment.
Methods
Patients
A study based on 103 COVID-19 infected cases and 22 non-COVID-19 pneumonia cases from the Second Hospital of Nanjing (Tangshan Branch, 1 Kangfu Road) was performed. All COVID-19 infected cases were diagnosed based on the World Health Organization (WHO) interim guidance [6] and confirmed positive for COVID-19 nucleic acid in the respiratory samples via real-time reverse-transcriptase polymerase-chain-reaction (RT-PCR). All non-COVID-19 pneumonia cases samples were negative, based on the absence of 2019-nCoV nucleic acid in the respiratory tract specimen. The RT-PCR assay was performed according to the protocol established by the WHO [7]. All patients with COVID-19 infection were divided into four groups according to the Diagnosis and Treatment Plan for Novel Coronavirus Pneumonia Cases (Provisional) (6th Ed.): severe (including severe and critical), common (non-severe), mild, and asymptomatic infected group [8]. This study was approved by the ethics committee of the Second Hospital of Nanjing, with written informed consent obtained from each patient.
Blood samples and laboratory data collection
Blood samples from patients were collected for laboratory assessments and flow cytometry according to the doctor’s instruction. All patients didn’t receive any treatment before blood sampling. Laboratory data were obtained from electronic medical records. Laboratory assessments comprised the following indexes: complete blood count, blood biochemical test, liver and renal function, electrolytes, C-reactive protein (CRP), procalcitonin, lactate dehydrogenase and creatine kinase.
Flow cytometry analysis
Peripheral blood mononuclear cells (PBMC) were collected from EDTA anticoagulant whole blood via erythrocyte lysis. PBMCs were incubated with a panels of fluorescence-labeled antibodies (CD45/CD3/CD4/CD8, CD45/CD3/CD16CD56/CD19) or corresponding IgG controls (Biolegend, San Diego, CA, USA) and rinsed with 1*PBS. Finally, the samples were examined by a flow cytometry analyzer (BD FACSCalibur, BD Biosciences, San Jose, CA, USA).
Statistical analysis
The results are expressed as means ± SD. Data were statistically analyzed using Prism SPSS 19.0 (SPSS Inc., Chicago, IL, USA) and assessed for normality or homogeneity of variance. For normally distributed continuous variables, differences between multiple groups were compared using one-way ANOVA with Dunnett’s tests or, if appropriate, using one-way ANOVA with post-hoc Bonferroni correction. Continuous variables that were not normally distributed were compared using Kruskal-Wallis test. Spearman correlation analysis between the blood cell subgroups and the blood biochemical indexes were conducted using SPSS 19.0. A value of p < 0.05 was considered significant. NS = not significant.
Results
Laboratory findings of COVID-19 infected patients
A total of 125 patients were included in the study (Table 1), consisting of 103 COVID-19 infected patients (55 males and 48 females) with a mean age of 44.388±18.878 and 22 non-COVID-19 pneumonia cases (11 males and 11 females) with a mean age of 36.727±9.948. The cohort of COVID-19 infected cases were further divided into the following four groups: severe (aged 67.875±12.229, 5 males and 3 females), non-severe (aged 44.845±16.799, 36 males and 35 females), mild (aged 39.444±14.328, 6 males and 3 females), and asymptomatic infected group (aged 32.667±22.68, 11 males and 11 females).
Clinical characteristics and laboratory findings of patients with 2019-nCoV infection (n=103) and non-COVID-19 pneumonia patients (n=22).
The complete blood count results indicated that the number of lymphocytes and platelets were significantly decreased in both severe and non-severe COVID-19 patients, in comparison with non-COVID-19-infected pneumonia patients, suggesting that COVID-19 patients could have lymphopenia and thrombocytopenia. In contrast, no significant differences in the quantities and proportions of monocytes and neutrophils were observed among these groups. Severe COVID-19 patients had significantly higher levels of C-reactive protein (CRP), compared with non-COVID-19-infected pneumonia patients. Blood biochemistry detection found that several indices related to organ injuries, including lactose dehydrogenase (LDH), aspartate aminotransferase (AST), estimated glomerular filtration rate (eGFR) and sodium ion concentration, where significantly different between severe COVID-19 patients and non-COVID-19-infected pneumonia patients.
Circulating lymphocyte subset profile of COVID-19 patients
To further explore the mechanism, and since lymphopenia was observed in COVID-19 patients, we performed flow cytometry to analyze the composition of lymphocyte subpopulations in the peripheral blood from patients. As shown in Table 2, the count of total CD45+ lymphocytes, CD3+ lymphocytes, CD4+ T cells, CD8+ T cells and CD19+ B cells in both server and non-severe COVID-19 patients were significantly lower than those in non-COVID-19-infected pneumonia patients. However, the differences of the proportions of CD3+/CD45, CD4+/CD45, CD8+/CD45, CD4+/CD8+ and CD16CD56+/CD45+ and CD4+CD8+ T cells count were only observed between severe COVID-19 patients and non-COVID-19-infected pneumonia patients. Meanwhile, no difference of CD16CD56+ NK cells count was observed among different groups.
Results of lymphocyte subpopulations in peripheral blood of patients with 2019-nCoV infection (n=103) and non-COVID-19 pneumonia patients (n=22).
Correlation analysis between leukocyte subgroups count and biochemical indexes
To examine the role of inflammatory responses in organ injury of COVID-19 patients, we performed Spearman’s rank correlation coefficient (ρ) analysis between leukocyte subgroups count and biochemical indexes relating to organ injury in COVID-19 patients exclude asymptomatic infected cases (Table 3). LDH, AST and sodium ion concentration showed significant correlations (P<0.05) with total lymphocyte and all lymphocyte subsets count; meanwhile, creatinine kinase (CK), total bilirubin, eGFR and potassium ion concentration only exhibited significant correlations with part of lymphocyte subpopulation count. Among the lymphocyte subsets, CD4+ Th cells count showed significant correlations with most biochemical indices (Figure 1). In contrast, no significant correlation was found between biochemical indexes and the counts for monocytes and neutrophils. In general, correlation analysis implied a significant negative correlation between lymphocyte subsets count and organ injury degree.
Spearman correlation analysis between biochemical index levels and the count of lymphocyte subpopulations, neutrophil and monocyte in patients with 2019-nCoV infection (exclude asymptomatic infected patients) (n=88).
Spearman’s rank correlation coefficient (ρ) analysis between CD4+ T cells count and biochemical indexes. A value of p < 0.05 was considered statistically significant.
Discussion
Our most striking finding is that lymphocytopenia, which is common in acute infections[9], is particularly significant in COVID-19 infections[3]. Others have demonstrated the correlation between the lymphocytopenia and the clinical severity[10]. Most previous reports only compared the severe groups against non-severe patients or heath controls; to further study this clinical feature, in our study, we compared the differences in the subgroups of lymphocytes harvested from COVID-19 patients with those from pneumonia patients not infected by COVID-19. We noticed that lymphocytes also decreased in the non-severe COVID-19 patients, suggesting that it might not only be a consequence of the infection but is probably a critical factor driving the development and deterioration of the disease. Among the lymphocyte subgroups, B cell exhibited the most significant differences from the non-COVID-19-infected pneumonia patients. Since B cells are responsible for the humoral immunity against the invaders via producing antibodies[11], the impotent response of B cells may fail to restrict the virus expansion and release of free virions in vivo, which promotes the infection.
Another new and notable feature discovered is that the change in the proportion of the CD45+ lymphocytes was insignificant, compared with non-COVID-19 patients. This finding, first, demonstrates that the infection impacts on different lymphocyte subgroups evenly, and more interestingly, suggests that abnormality in cells that regulate the adaptive immunity may be responsible for the reduction of lymphocytes. For instance, H5N1 infection causes lymphocytopenia through dendritic cell (DC) dysfunction [12], suggesting the function of key antigen-presenting cells can be drastically impaired during the infection. According to a study in rats, the myeloid cells also express the COVID-19 receptor ACE2[13]; thus, we should consider the possibility of the virus infecting these cells, which is supported by the significant decrease in the platelet counts in both severe and non-severe COVID-19 patients.
Moreover, recent reports indicate that the infection can potentially damage other organs than the lung, including the kidney, heart, and the testis [14-15]. We also notice changes in multiple biochemical indexes relating to organ injuries, including LDH, CK, AST, and eGFR. In SARS-Cov cases, the multi-organ injury was considered as a result of the inflammatory reaction (inflammatory cytokine storm). However, the early report about COVID-19 cases found that the profile of cytokine in the blood of patients was different from that from SARS-Cov cases: not only pro-inflammatory cytokines but also anti-inflammatory cytokines (IL-4, IL-10, etc.) increased in the COVID-19 cases[4]. Additionally, all recent reports demonstrated that high fever cases (>38 °C) were significantly fewer (even in severe cases) in COVID-19 patients than in SARS-CoV or MERS-CoV cases[3]. These features suggest that the consequences in the COVID-19 patients might not be mainly due to the inflammatory reaction, especially in the non-severe cases. Instead, the damage was likely to be caused by the virus itself. Therefore, we performed a correlation analysis between the number of lymphocyte subgroups and the biochemical indexes. We found that most of the indexes relating to organ injuries were negatively correlated with the number of lymphocytes, further highlighting the virus infection – rather than the inflammatory reaction – as the possible cause of the multi-organ injury. Another supportive finding is that the difference in the number of monocytes, neutrophils, and NK cells is not significant in our analysis; while theoretically, these innate immune cells are the main player in a “cytokine storm”. Interestingly, the most significant correlation happened between the injury indexes and CD4+ Th cell counts. Because CD4+ Th cells are critical for the regulation of both cellular immunity and humoral immunity[16], it is reasonable that these cells are most sensitive to the total anti-virus immune responses. This finding further suggests that the damage of the tissue is from the virus itself. Consistently, studies on the tissue distribution of ACE2 suggest that the virus receptor is widely expressed in human tissue including the digestive tract, kidney, testis, and others [17].
We highlight that the above finding is important for the design of treatment regimens. Current treatments do not suggest using specific drugs to improve the activity of immune systems, such as the GM-CSF or other immune stimuli, in fearing of the exaggeration of inflammatory reactions and cytokine storm[18]. However, based on our clinical findings, we should seriously consider using immune-activating treatments for COVID-19 patients, which will be helpful to compensate the dysfunctions in the adaptive immune system and accelerate virus clearance process in vivo. Considering the high possibility of secondary infection induced by excessive inflammation in the severe cases, this strategy should be safer to apply in non-severe patients with the aim to lower the occurrence of severe cases.
Taken together, our analysis in the present study, based on clinical data, demonstrate significant differences between pneumonia caused and not caused by COVID-19 – especially including the lymphocytopenia. Our results further suggest that the tissue damage to COVID-19 patients might be mainly due to virus infection itself rather than uncontrolled inflammatory reactions. These findings are valuable for the development of more efficient treatments.
Data Availability
The data used to support the findings of this study are available from the corresponding author upon request.
Conflict of interest
None declared.
Funding
This study was funded by the Young Medical Talents of Jiangsu Province (QNRC2016056), the fifth period “333 high-level personnel training project” in Jiangsu Provience (2016-III-0084), the National Natural Science Foundation of China (31971309 and 31671031), the Jiangsu Province Funds for Distinguished Young Scientists (BK20170015).
Data Availability Statement
The data used to support the findings of this study are available from the corresponding author upon request.
References
