HIV and Associated Indicators of COVID-19 Cytokine Release Syndrome: A Meta-Analysis and Meta-Regression

The aim of this review was to evaluate the risk of COVID-19 cytokine release syndrome (CRS) with HIV infection and meta-regress for indicator covariates. Electronic databases, including Google Scholar, Cochrane Library, Web of Sciences (WOS), EMBASE, Medline/PubMed, COVID-19 Research Database, and Scopus, were systematically searched till February 30, 2022. All human studies were included, irrespective of publication date or region. Eleven studies, with a total of 2,005,274 detailing cytokine release syndrome defined by specific parameters, were included. To pool the estimate, a random-effects model with risk ratio (RR) as the effect measure was used. Moreover, publication bias and sensitivity analysis were evaluated followed by meta-regression analysis to account for any possible covariates. This systematic review, meta-analysis, and meta-regression trial was registered (CRD42021264761) on the PROSPERO register. HIV infection showed an increased risk for COVID-19 cytokine release syndrome (RR= 1.48, 95% CI (1.16, 1.88) (P=0.002)) with substantial heterogeneity (I2 > 80%) and a 4.6% cumulative incidence. The true effects size in 95% of all the comparable populations (prediction interval) fell between 0.67 to 3.29. HIV infection further showed an increased risk for intensive care unit (ICU) admission ((P<0.0001) (I² = 0%)] and mechanical ventilation (MV) ((P=0.04) (I² = 0%)) as the key indicators of cytokine release syndrome. Meta-regression analysis demonstrated that COVID-19 cytokine release syndrome was influenced by the year a study was published (R² = 0.55) and the region from where the study was conducted (R² = 0.11). On meta-regression analysis, the combined impact of all covariates in the model explained at least some of the variance in effect size (Q = 16.21, df = 6, P= 0.0127), and the proportion of variance explained by covariates on comparing the model with and without the covariates was 73 % and highly significant (Tau² = 0.1100, Tau = 0.3317, I² = 86.5%, Q = .99, df = 10, P<0.0001) (R² = 0.73). Our updated meta-analysis indicated that HIV infection was significantly associated with an increased risk for COVID-19 cytokine release syndrome, which, in addition, might be moderated by the year a study was published and the region in which the study was conducted. Further, the risk for intensive care unit (ICU) admission and mechanical ventilation (MV) were identified as the key indicators of cytokine release syndrome. We believe the updated data anchoring cytokine release syndrome will contribute to more substantiation of the findings reported by similar earlier studies.


Introduction And Background
About 38 million people living with HIV (PLWH) globally, including 1.7 million children, with a global HIV prevalence of 0.7% among adults [1], may have an increased risk of adverse outcomes from coronavirus disease 2019 (COVID- 19) infection as a result of HIV-associated immune dysfunction due to the associated cells' alterations and depletion [2]. There may also be a higher prevalence of comorbidities among PLWH that predispose them to adverse COVID-19 outcomes [3]. Conversely, PLWH may have more favorable outcomes due to increased health awareness or close medical follow-up and constant reviews with some specific antiretroviral agents under consideration as potential treatments for COVID-19 [4].
Severe COVID-19 disease manifested by fever and pneumonia, leading to acute respiratory distress syndrome (ARDS) has been described in up to 20% of COVID-19 cases. This is reminiscent of cytokine release syndrome (CRS)-induced ARDS and secondary hemophagocytic lymph histiocytosis (sHLH) observed in patients with SARS-CoV-2 [5], characteristics of CRS, including pulmonary inflammation, fever, and dysfunction of non-pulmonary organs. An increase of interleukin-6 in peripheral blood is a key risk factor and an early indicator of CRS in COVID- 19. Both antibody and T-cell responses are critical for the effective control and clearance of SARS-CoV-2. More severe COVID-19 disease correlates with lymphopenia and low T-cell concentrations [6]. COVID-19-associated CRS by HIV serostatus is not explicitly researched and most meta-analyses have focused on studies lacking comparator groups or they used a general population as controls unlike in the current study, which restricts the comparator as HIV negative in the same included study. The study aimed at evaluating the evidence on the risk of COVID-19 CRS in PLHIV using both earlier and recently published data, and a meta-regression to ascertain the extent to which this risk is modified by other possible covariates.
A portion of the content of this current article was previously posted to the Multidisciplinary Preprint Platform server under Preprints on COVID-19 and SARS-CoV-2 on May 13, 2022.

Study Design and Search Strategy
We utilized a systematic review to identify studies between April 1, 2020, and February 30, 2022, which described cytokine release syndrome in people living with HIV (PLWH) and compared them with HIVnegative people, and a meta-analysis approach, followed by a meta-regression, to ascertain the covariates associated with COVID-19 cytokine release syndrome. For the remaining articles, another author performed title/topic and abstract screening, with subsequent full-text review by two authors using a standardized data extraction form. Where disagreement was feasible, inclusion decisions were made by a third author. We also included preprints to capture the most recent and emerging evidence. Studies with 15 or fewer participants were excluded, as they were less likely to have the power to detect meaningful relationships. The quality of the studies was evaluated using the Newcastle-Ottawa Scale for observational studies [7].

Studies included, effect measures, and analysis
Observational studies reporting any possible indicator of COVID-19-related cytokine release syndrome in people with and without HIV were included in a meta-analysis. Specific relative risks (RRs) and hazard ratios were combined with a random effects model to account for the variability of the true effect between studies. To explore possible effect modifications, subgroup and meta-regression analyses were conducted for COVID-19-related CRS.

Study Selection Procedure
We identified 2285 records and included a total of 11 studies detailing cytokine storm syndrome as an outcome in our final analysis. The included studies were peer-reviewed, with some as preprints since the research quest sought to capture even the latest data and information. The 11 studies [8][9][10][11][12][13][14][15][16][17][18] reported and compared cytokine storm syndrome, defined by a specific parameter (such as intensive care unit admission) between HIV seropositive and seronegative persons. The procedure is shown in Figure 1.

Quality of evidence and risk of bias assessment
We assessed the quality of the included studies based on a modified version of the Newcastle-Ottawa Scale (NoS), which consists of eight items with three sub-scales, and the total maximum score of these three subsets is 9. We considered a study that scored ≥7 a high-quality study since a standard criterion for what constitutes a high-quality study has not yet been universally established. The studies assessed generated a mean value of 6.59 and as a result, the overall quality was found to be moderate (NOS score min: 5, max: 8).
There were common limitations among the included studies. Most were retrospective analyses of routinely collected clinical data, meaning the identification of COVID-19 cases was not systematic and depended on the local approach to screening and diagnosis with only five prospective cohort studies. This may have varied over time and between settings and may also differ between PLWH and the general population but in the case of this study though, the studies included both HIV-seropositive and seronegative populations' data. Across all studies, the numbers of HIV-seropositive and COVID-19 infections were relatively low. The Newcastle Ottawa Scale (NoS) information is shown ( Table 1) [17].

TABLE 1: Quality of evidence and risk of bias assessment
Newcastle Ottawa Scale (NoS) consisting of eight items with three sub-scales and a total maximum score of 9 The asterisk(*) numbers here depict the level of agreement in terms of each author's view. Using the tool, each study is judged on eight items, categorized into three groups: the selection of the study groups; the comparability of the groups; and the ascertainment of either the exposure or outcome of interest for case-control or cohort studies, respectively. Stars awarded for each quality item serve as a quick visual assessment. Stars are awarded such that the highest quality studies are awarded up to nine (9) stars.

Results
In this meta-analysis pool, 2,005,274 from 11 studies [8][9][10][11][12][13][14][15][16][17][18] with cytokine release syndrome diagnosed with COVID-19 were included utilizing the predefined given Centers for Disease Control and Prevention (CDC) reporting guidelines on COVID-19 diagnosis [19]. The cumulative COVID-19 cytokine release syndrome defining parameter was 48863 (2.4%). The total COVID-19-related CRS was 837(4.6%) and 48026 (2.4%) among the HIV seropositive and HIV-seronegative persons, respectively. The cumulative incidence of COVID-19-related cytokine release syndrome ranged from 1.5% to 40 % (average: 19 %). A summary of the studies included in this meta-analysis is available (    The prediction interval demonstrated the true effects size in 95% of all the comparable populations falling between 0.67 to 3.29, which depicted that, in some populations, the risk of COVID-19 cytokine release syndrome due to HIV infection is at one extreme of effect as low as 0.67 and as high as 3.29, thus necessitated accounting for any possible covariates ( Figure 4).

FIGURE 4: Distribution of true effects
The prediction interval of the true effects size in 95% of all the comparable populations using the random effects model.
The mean effect size is 1.48 with a 95% confidence interval of 1.17 to 1.88 while the true effect size of all comparable populations falls in the interval 0.67 to 3.29.

FIGURE 7: Pooled CRS subgroup analysis by specific indicators (parameters)
Note: Weights and the between-subgroup heterogeneity tests are from the Mantel-Haenszel model.

FIGURE 8: Sensitivity analysis on CRS indicator by HIV-serostatus
Note: Weights and the between-subgroup heterogeneity test are from the Mantel-Haenszel model.

Meta-Regression for Possible Moderators of COVID-19 CRS With HIV-Serostatus
The values of heterogeneity (I 2 ) in studies accounting for COVID-19 CRS was 87, which means the observed variance came from real differences between studies and, as such, can potentially be explained by study-

Discussion
The purpose of this study was to systematically review and conduct a meta-analysis using the most current data from studies on the incidence of COVID-19-related cytokine release syndrome relative to HIV serostatus, alongside the associated covariates via meta-regression. Further, it aimed at ascertaining the parameters defining cytokine release syndrome predicted by HIV infection and estimating the combined proportion effect of all covariates in studies detailing CRS.
Principally, the present meta-analysis found that HIV seropositive status was significant in predicting CRS by over 50%. Following sensitivity analysis of good-quality studies only, the risk for both COVID-19-related CRS was more significant. Overall, there was a high degree of heterogeneity among studies detailing COVID-19-related CRS, which greatly reduced following sensitivity analysis. The outcome remained significant on the inclusion of only good-quality studies suggesting these analyses represent true effects as per the generated prediction intervals. A high level of heterogeneity was only observed with the inclusion of a few studies in assessing the effect of HIV on COVID-19-related CRS, likely to substantial inter-study variation.
Egger's regression test indicated a low impact of publication bias on our results.
The finding that HIV clearly predicts more significantly the experience of cytokine release syndrome confirms previous findings [20]. The association of CRS with HIV seropositive status in the context of this current findings is biologically plausible as in normal circumstances, CRS is linked with acute respiratory distress syndrome (ARDS), which leads to COVID-19 severity prior to case fatality (death) [21].
CRS indicators were critical care services (ICU) admission, mechanical ventilation, increased intubation rates, elevated interleukin-6, clinical severity of COVID-19, and inpatient services. These trends are similar to other studies that demonstrated that HIV infection is associated with ICU admission, mechanical ventilation [14], intubation [22], interleukin-6 [23], and clinical severity of COVID-19 inpatient services [24]. In this current study, mechanical ventilation and ICU admission clearly showed an association with HIV seropositive status with a similar trend of increased risk [14], but this is contrary to another study that found no difference in HIV infection and non-infection [17].
Meta-regression analysis showed that the year (2020, 2021, and not 2022) and the region in which a study was conducted were associated with COVID-19-related cytokine release syndrome (P < 0.05), unlike the study setting sites. Generally, the combined impact of all covariates in the model explained at least some of the variance in COVID-19-related CRS, similar to existing findings in countries and region-related factors [25].
Some limitations were noted in our review and meta-analysis. The included studies did not put into categories clear HIV infection staging as per the WHO criteria [26], thus it made it impossible to conduct a subgroup analysis on PLWH based on that. Cytokine release syndrome is multifaceted and an acute systemic inflammatory syndrome characterized by fever and multiple organ dysfunction that is associated with chimeric antigen receptor (CAR)-T cell therapy [27], however, the study focused on the clinical outcomes that were defined by specific parameters such as ICU admission, though with the clear presumption that this would ensue due to the pathophysiology of the cytokine release syndrome [28].

Conclusions
Our study indicated a consistent and statistically significant effect of HIV on COVID-19-related cytokine release syndrome even after heterogeneity investigation all in the random effects model with Egger's intercept regression test indicating no major publication bias. ICU admission, mechanical ventilation, and intubation were the key CRS parameters predicted by HIV infection in COVID-19 patients. The proportion of variance explained by covariates was significant with the year a study was conducted, the region of the study population, and the study setting, either single or multiple center, being the major covariates associated with COVID-19-related CRS.
Public health interventions should be carefully tailored and implemented on PLWH and infected COVID-19 to reduce the risk of severity associated with cytokine release syndrome, a key predictor of COVID-19 case fatality. An intensive and regular focus is required to detect early occurrences of clinical conditions in similar viral pandemics or COVID-19 resurgence.

Conflicts of interest:
In compliance with the ICMJE uniform disclosure form, all authors declare the following: Payment/services info: All authors have declared that no financial support was received from any organization for the submitted work. Financial relationships: All authors have declared that they have no financial relationships at present or within the previous three years with any organizations that might have an interest in the submitted work. Other relationships: All authors have declared that there are no other relationships or activities that could appear to have influenced the submitted work.