Abstract
Objective
To identify whether thymosin-α-1 (Tα1) is effective in patients with Coronavirus disease 2019 (COVID-19) and to determine a suitable population for Tα1 treatment.
Methods
We included studies with ≥10 cases and adults (aged ≥18 years) with laboratory-confirmed SARS-CoV-2 infection, data on mortality or length of hospitalization, disease severity, and study location, while excluded pregnant and breastfeeding women and minors. Publications were searched from November 1, 2019, to July 5, 2023, in six databases, including PubMed, Web of Science, Embase, Cochrane Library, China Knowledge Resource Integrated Database, and Wanfang Database. We separately utilized Newcastle-Ottawa Scale and Cochrane handbook methodology to evaluate risk of bias and used Review Manager (version 5.4, Cochrane Collaboration, Copenhagen, Denmark) to present and synthesize results. Relative risks (RR) and Standardized Mean Difference (SMD) with 95% confidence intervals (CI) were analyzed for dichotomous variables and continuous variables, respectively.
Results
Nine studies (participants = 5417) were included. No significant differences were found in mortality (nine studies; n = 5417; RR = 0.95; 95% CI: 0.56, −1.60; p = .84; I2 = 90%) or length of hospitalization (four studies; n = 3688; SMD = 0.16; 95% CI: −0.38, −0.69; p = .57; I2 = 96%) between patients with COVID-19 who did and did not receive Tα1. Participants were divided by the severity of the disease (serious and non-serious) and study location. Among the serious group, the incidence of death among patients who received Tα1 treatment was 0.67 times that of patients who did not receive Tα1 treatment (four studies; n = 1230; RR: 0.67; 95% CI: 0.58, −0.77; p < .00,001; I2 = 0%). There was no significant difference in length of hospitalization between the groups (two studies; n = 410; SMD = 0.66; 95% CI: −0.06, −1.38; p = .07; I2 = 87%). Among the non-serious group, compared to not having Tα1 treatment, receiving Tα1 treatment reduced hospitalization length (two studies; n = 3670; SMD = −0.28; 95% CI: −0.41, −0.14; p < .0001; I2 = 51%), while no significant difference in mortality (three studies; n = 3775; RR = 1.06; 95% CI: 0.22, −5.03; p = .94; I2 = 89%). Moreover, there was no significant difference between subgroups when divided by study locations (Studies within China: seven studies; n = 5263; RR = 1.14; 95% CI: 0.64, −2.04; p = .65; I2=92%; Studies outside of China: two studies; n = 154; RR = 0.41; 95% CI: 0.14, −1.24; p = .11; I2 = 51%).
Discussion
For patients with serious types of COVID-19, Tα1 significantly decreased mortality, which supports the utilization of Tα1 in patients with severe and critical types of COVID-19. Moreover, regarding hospitalization length, patients with non-serious COVID-19 who used Tα1 reduced their hospitalization length compared to those that did not use Tα1. However, these results have high heterogeneity and limited generalizability.
Introduction
Coronavirus disease 2019 (COVID-19) and its pathogen, severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), have caused worldwide serious disease with high pathogenicity, infectivity, and morbidity. 1 Three years into the COVID-19 pandemic, research has suggested that, in addition to viral contagiousness, immune dysregulation is also an independent symptom of disease exacerbation. 2
The immune system is divided into innate and adaptive systems. The former plays a vital role in the antiviral process as it limits viral replication within infected cells, creates an antiviral condition in the local tissue environment, and activates the adaptive immune response. The latter is central to antiviral treatment success and consists of B cells that produce antibodies, CD4+ T cells that have helper and effector functionalities, and CD8+ T cells that kill infected cells. 3
The synergistic effect between innate and adaptive immune responses is essential to control viral dissemination. 4 However, abnormal immune status is a common phenomenon in the deterioration of patients with COVID-19. Lymphopenia presents as declined amounts of CD4+ T cells, CD8+ T cells, and NK cells,5,6 and cytokine storm (CS), due to increased proinflammatory cytokines and chemokines levels from multiple pathological conditions.5,7–10 As a result, immunotherapies seem like a probable option for COVID-19 treatment.
Thymosin-α-1 (Tα1) is a type of endogenous polypeptide that is extracted from the thymic tissue and has broad applications for various diseases such as cancers, autoimmune disorders, and viral infectious diseases.11–13 Tα1 has immunomodulation effects through increasing T cell generation and anti-inflammatory cytokines including Interleukin-10 (IL-10), enhancing the host’s defense response, and reducing proinflammatory cytokines including tumor necrosis factor-α (TNF-α).14,15 Consequently, thymosin is a promising therapeutic drug option for those who are infected with SARS-CoV-2. Therefore, we aimed to investigate whether Tα1 is effective in patients with COVID-19 and to determine a suitable population for Tα1 treatment.
Methods
Eligibility criteria and selection process
The inclusion criteria covered the following provisions: (1) experimental, observational, or retrospective research studies of cases that explored the clinical effects of Tα1; (2) 10 or more cases included in each study; (3) research results included either mortality or length of hospitalization; (4) laboratory-confirmed SARS-COV-2 infection via throat swab, sputum, or lower respiratory tract samples by high-throughput sequencing or real-time reverse transcriptase polymerase chain reaction; (5) adults (aged ≥ 18 years) with no restrictions on gender and nationality, and (6) grouping criteria for subgroup analysis. Grouping criteria for subgroup analysis were as follows: (1) severity of the disease including serious versus non-serious groups and (2) locations in which studies were conducted. Serious versus non-serious groups were defined as patients with severe and critical versus mild and moderate COVID-19, respectively. When studies contained different levels of disease severity, grouping depended on the proportion of patients with severe and critical COVID-19; patients were classified as serious if the proportion included or exceeded 50%. Studies were divided into two groups based on where studies were conducted, including within and outside of the People’s Republic of China (PRC).
Studies were excluded if any of the following criteria were met: (1) had fragmented data or did not meet evaluation standards; (2) the subjects were pregnant or breastfeeding women or minors; (3) had confounding factors that could not be ruled out; (4) were animal experiments, duplicated studies, case reports, conference studies, reviews, expert opinions, dissertations, invalid studies, or papers without full text.
Two reviewers screened each study independently to determine eligible studies. When the two reviewers had differing opinions, a third researcher was involved in the selection process and solved the conflicts by consensus discussion.
Information sources and search strategy
By using the keywords, “COVID-19”, “SARS-CoV-2”, “2019 novel coronavirus”, “Corona-Virus-2019”, and “thymosin”, “thymosin Alpha 1 (Tα1)”,” thymalfasin”, “thymus peptide”, two reviewers searched articles in the following databases from November 1, 2019, to July 5, 2023, without language restrictions: PubMed, Web of Science, Embase, Cochrane Library, China Knowledge Resource Integrated Database (CNKI), and Wanfang Database. The title, abstract, and full text were checked successively to evaluate whether the research was appropriately themed for the scope of our study. Furthermore, reference lists of the included studies were examined to identify if there were additional studies omitted from our search. We adhered to an “umbrella” review methodology. Moreover, this review protocol was registered with PROSPERO (CRD42023440802).
Data collection process and data items
Two reviewers extracted the following details from the included studies into a data extraction sheet: the name of the first author, publication date, location of the research, research design type, sample size, Tα1-treatment group assignment, data about the patients (such as average age, gender, and severity of disease), and two outcomes of interest (mortality and length of hospitalization). Mortality was defined as the number of patients who died because of COVID-19. Length of hospitalization was defined as the time from hospital admission to discharge due to COVID-19. Finally, the last reviewer checked the tables.
Study risk of bias assessment
To evaluate the quality of the included articles, two researchers with systematic evaluation training separately assessed the quality of the papers using the Newcastle–Ottawa Scale (NOS) for cohort studies and Cochrane handbook methodology for randomized controlled trials (RCT).16,17
The NOS assessment had eight parts, including adequate definition of cases, representativeness of cases, selection of controls, definition of controls, comparability of cases and controls on the basis of the design or analysis, ascertainment of exposure, same method of ascertainment for cases and controls, and non-response rate. Based on the NOS score, the quality of the studies was divided into low (1–3 scores) moderate (4–6 scores), and high levels (7–9 scores).
Cochrane handbook methodology contained seven items, including allocation concealment, random sequence generation, blinding of participants and personnel, blinding of outcome assessment, selective reporting, incomplete outcome data, and other sources of bias. There were three outcomes of this approach: low risk of bias (the study has low risk of bias for all domains for a given outcome), some concerns (the study raises some concerns for risk of bias in at least one domain for a given outcome, but is not categorized as high risk of bias for any domain), and high risk of bias (the study has high risk of bias in at least one domain for a given outcome or the study has some concerns for bias for multiple domains in a way that substantially lowers confidence in the result).
If two researchers gave different marks, a third researcher participated in the evaluation process and solved the conflicts by consensus discussion. During this process, authors’ names and affiliations are anonymized for impartiality.
Effect measures and synthesis methods
We utilized the Review Manager (version 5.4, Cochrane Collaboration, Copenhagen, Denmark) for this meta-analysis. Relative risks (RR), standardized mean difference (SMD) and means with standard deviations (SD) or medians with interquartile ranges (IQR) were analyzed for dichotomous and continuous variables, respectively. Additionally, 95% confidence intervals (CI) were calculated, if available. The degree of heterogeneity in the studies was assessed by I2 statistics. When selecting between a fixed-effect or random-effect model, we chose the former when there was little heterogeneity among research participants. Commonly, p < .05 or I2 < 50% suggested low heterogeneity among the studies; in contrast, p≥ .05 or I2 ≥ 50% was regarded as a sign of notable heterogeneity.
Reporting bias assessment and certainty assessment
Sensitivity analyses or subgroup analyses were performed to explore provenances with high heterogeneity. Additionally, we drew funnel plots to assess publication bias.
Results
Study selection
In six databases, we found 829 articles in total (Web of Science n = 95; PubMed n = 85; Cochrane Library n = 29; Embase n = 243; CNKI n = 358; Wan Fang n = 19). Duplicate articles (n = 287) were ruled out first. Titles and abstracts of the remaining articles (n = 542) were inspected for appropriate inclusion and exclusion criteria. Of these, 23 articles were selected, and the full text was reviewed independently by two researchers. Afterwards, 14 papers were excluded from our study for the following reasons: no full text available, invalid assignment, incomplete data, no accurate outcomes, previously published meta-analyses. Ultimately, 9 studies were included in our meta-analysis18–26 (Figure 1). Flow chart of the literature search.
Study characteristics
Details of the included studies.
Risk of bias in studies
Quality if included cohort studies according to the Newcastle-Ottawa scale.
★ Indicates study scored 1 point.
☆ Indicates study scored 0 point.
Risk of bias graph of included RCTs according to Cochrane handbook methodology.
Risk of bias summary of included RCTs via Cochrane handbook methodology.
Mortality
All included studies provided entire mortality data, which allowed us to conduct a meta-analysis to compare the experimental (patients with Tα1 treatment) and control (patients without Tα1 treatment) groups (Figure 4). There was no significant difference in incidence between the experimental and control groups (nine studies; n = 5417; RR = 0.95; 95% CI: 0.56, −1.60; p = .84; I2 = 90%). To explore reasons for heterogeneity, we performed two subgroup analyses based on disease severity (Figure 5) and the nations in which the studies were conducted (Figure 6). Heterogeneity did not improve with these sensitivity analyses. Mortality among patients with versus without thymosin-α-1 (Tα1) treatment. Subgroup analysis of mortality among patients with versus without thymosin-α-1 (Tα1) treatment based on disease severity. Subgroup analysis of mortality among patients with versus without thymosin-α-1 (Tα1) treatment based on nations where the studies were conducted.
In addition to two studies without specific records of disease severity, participants in the remaining studies were divided into serious and non-serious groups based on disease severity. Among those with non-serious COVID-19, there was still no significant difference in the incidence of mortality between the experimental and control groups (Figure 5, three studies; n = 3775; RR = 1.06; 95% CI: 0.22, −5.03; p = .94; I2 = 89%). However, among those with serious COVID-19, those that received Tα1 had 0.67 times the risk of mortality compared to those who did not receive Tα1 treatment (Figure 5, four studies; n = 1230; RR = 0.67; 95% CI: 0.58, −0.77; p < .00,001; I2 = 0%).
We divided the nine studies into those conducted within and outside of the PRC. No significant difference in mortality was found between the two groups (Figure 6; seven studies within China: n = 5263; RR = 1.14; 95% CI: 0.64, −2.04; p = .65; I2 = 92%; two studies outside of China: n = 154; RR = 0.41; 95% CI: 0.14, −1.24; p = .11; I2 = 51%).
Length of hospitalization
Although five studies provided data on the length of hospitalization, one study was excluded due to insufficient information. Only one set of the studies
19
reported data as mean ± SD; therefore we estimated the mean ± SD for the remaining studies using the sample size, median, range, and IQR.27–30 The forest plot shows that there was no significant difference in length of hospitalization between experimental and control groups (Figure 7, four studies; n = 3688; SMD = 0.16; 95% CI: −0.38, −0.69; p = .57; I2 = 96%). Additionally, there was no statistically significant difference in length of hospitalization between treatment groups among those with serious COVID-19 (Figure 8, two studies; n = 410; SMD = 0.66; 95% CI: −0.06, −1.38; p = .07; I2 = 87%). Nevertheless, among those with non-serious COVID-19, those who were treated with Tα1 had a shorter length of hospitalization compared to those who did not receive Tα1 treatment, (Figure 8, two studies; n = 3670; SMD = −0.28; 95% CI: −0.41, −0.14; p < .0001; I2 = 51%). Length of hospitalization among patients with versus without thymosin-α-1 (Tα1) treatment. Subgroup analysis of length of hospitalization among patients with versus without thymosin-α-1 (Tα1) treatment based on disease severity.
Discussion
The clinical course of COVID-19 can be categorized as mild, moderate, severe, or critical. Serious COVID-19, including severe and critical types, is usually characterized by severe respiratory disease, multiple organ dysfunction, immune cell exhaustion, and local or systemic hyperinflammation.31–33 In this period, the pathophysiologic damage to the body derives more from the host’s own immune system rather than from the virus itself. Consequently, immunomodulatory treatment becomes pertinent.
Immunomodulatory therapies that have been approved for use in European countries include cytokine antagonists (tocilizumab and sarilumab), kinase inhibitors (baricitinib), plasma therapies (convalescent plasma), and others (dexamethasone and prednisolone). 31 However, the consequences of ignoring the hazards of these drugs could be detrimental. For example, restraining ideal antiviral interferons by baricitinib may support SARS-CoV-2’s immune escape strategies. 34 In addition, adverse effects of long-term use of glucocorticoids are widely known. In addition, variability in antibody titers and neutralizing capacity is an inherent limitation of plasma-derived therapies. 35
Produced in the thymus and brain, Tα1 mainly works in both the thymus and peripheral tissues. 36 With the ability to control the progression of immunity, inflammation, and illness, 37 Tα1 has sweeping applications for managing chronic hepatitis B and C, cytomegalovirus infection, sepsis, chronic obstructive pulmonary disorder, HIV/AIDS,38–40 by inducing Th lymphocytes, activating cytotoxic T lymphocytes, and maintaining immune homeostasis in viral infection. 41 Moreover, Tα1 has been shown to be well-tolerated and safe in patients diagnosed with the diseases previously mentioned. 42 Hence, Tα1 is a potential and promising treatment strategy against viral, fungal, and bacterial infectious diseases, 43 including SARS-CoV-2. 44
For this reason, we studied the efficacy of Tα1 to treat patients with COVID-19 and concluded that Tα1 could reduce mortality for serious types of COVID-19 but had no obvious effects for non-serious types. Furthermore, patients with non-serious COVID-19 who had Tα1 treatment had further reductions in length of hospitalization compared to those who did not have treatment. This may be attributed to the complex pathogenesis of SARS-CoV-2. Previously, it was suggested that adverse clinical outcomes of severe disease conditions might be strongly associated with CS 45 produced by the abnormal induction of the immune response to the infection.46,47 As research developed, more mediators in patients’ circulatory systems have been inspected, 48 such as elevated kynurenines, enhanced complement system, reduced lipids, decreased platelet-derived chemokines proplatelet basic protein (PPBP; also called macrophage-derived growth factor) and platelet factor 4 (PF4), and the change of the acute phase proteins (APPs).49–52 Consequently, as CS are only one part of the imbalanced immune response, viral sepsis may be more appropriate to describe the much more intricate host-pathogen interaction, particularly for severe and critical COVID-19 cases. 48
As immunological factors play a more important role in the latter course (severe and critical COVID-19) than in the earlier course (mild and moderate COVID-19), it is logical that Tα1 is more successful at reducing mortality in serious cases compared to non-serious cases. Patients who survive by using Tα1 need more time to recover, leading to longer average hospitalization for these serious types. These theories agree with our meta-analysis results.
Immune dysregulation is a critical factor of serious COVID-19 pathogenesis, which needs explicit and specific treatment as soon as possible. Although immunomodulatory approaches are promising, more research is needed on how to minimize adverse effects and determine the most suitable administration time. Additionally, the safety and validity of Tα1 to treat COVID-19 requires investigation in clinical trials. 53
This review has some limitations. First, there was a lack of high-quality RCTs among the included studies. Second, most of the studies eligible for inclusion were conducted in the PRC, which could result in bias. Third, when we study the relationship between the thymosin treatment and the length of hospitalization, we used imprecise data by estimating the sample size, median, and IQR range, which may cause errors in the results. Fourth, as the original data did not provide sufficient details, the grouping strategy for subgroup analysis lacked precision. Ideally, group classification should be based on individual characteristics, rather than the overall characteristics of each study. Fifth, generalizability is limited due to high heterogeneity in the results; therefore, the results should be interpreted with caution.
Conclusion
In this systematic review and meta-analysis, we demonstrate that, for patients with serious types of COVID-19, Tα1 could significantly decrease risk of mortality, which supports the utilization of Tα1 in these patients. Moreover, regarding the effect on the length of hospitalization, patients with non-serious COVID-19 who were treated with Tα1 had reduced length of hospitalization compared to those that were not treated with Tα1.
Footnotes
Author contributions
Da Chen conceived the study hypothesis. Pu Wang and Da Chen designed the study procedure. Pu Wang and Changhong Wang worked on the literature search, literature screening and data extraction. Da Chen and Pu Wang analyzed the data. Pu Wang wrote the first draft. Da Chen and Changhong Wang revised the paper. All authors approved the submitted and final versions.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Systematic review registration
PROSPERO (CRD42023440802).
