The coronavirus disease 2019 (COVID-19) pandemic has accounted for more than 108 million documented cases and 2 million deaths worldwide as of February 2021. Patients with diabetes mellitus have a significantly higher risk for severe disease and COVID-19-related death than the general population, representing a very sensitive population group [1,2]. A vivid and ongoing discussion is continuing regarding the place of antidiabetic drug classes in the therapeutic management of COVID-19-infected patients with type 2 diabetes mellitus (T2DM) and whether antidiabetic medications predispose patients to infections [3].

Dipeptidyl peptidase-4 (DPP-4) inhibitors have attracted scientific interest due to the potential involvement of the DPP-4 enzyme (also known as the T-cell antigen CD26) in the pathophysiology of the infection, along with the anti-inflammatory properties of this drug class [4]. A previously published meta-analysis of six observational studies by Kow and Hasan [5] addressing the composite outcome of severe or fatal illness demonstrated that pre-admission use of DPP-4 inhibitors did not have a significant impact on prespecified outcomes. More recently, Bonora et al. [6] pooled data from seven observational studies published in the Journal of Endocrinological Investigation, demonstrating that the prior use of DPP-4 inhibitors was not associated with an increased risk of COVID-19-related death, having a rather neutral effect. The aim of the present study was to update the previously published meta-analyses.

We sought to determine whether DPP-4 inhibitors influence the risk for COVID-19-related death, which corresponds to the worst outcome of the disease course. We searched two major databases (PubMed and Cochrane Library) for published randomized controlled trials (RCTs) or observational studies enrolling adult patients with T2DM with prior use of DPP-4 inhibitors versus other antidiabetic drug classes, with confirmed COVID-19 infection, by applying the following search strategy: (sitagliptin OR vildagliptin OR alogliptin OR saxagliptin OR linagliptin OR anagliptin OR tenegliptin OR trelagliptin OR gemigliptin OR DPP-4 inhibitor) AND (COVID-19 OR SARS-CoV-2 OR covid19). We excluded studies of the pediatric population and studies enrolling patients with type 1 diabetes mellitus and patients with suspected COVID-19 infection without laboratory confirmation. We did not apply any limitation according to publication language or study sample size. We hand-searched the retrieved bibliographies to identify potentially eligible studies that were missed with the applied search strategy.

Two independent reviewers (D.P. and M.D.) extracted the data from the eligible reports using a pilot-tested data extraction form. We set a “hard” safety outcome of COVID-19-related death (regardless of occurrence in the outpatient or inpatient setting).

As we only assessed a dichotomous variable, the difference was calculated using risk ratios (RRs) with 95% confidence intervals (CIs), after implementation of the Mantel-Haenszel random effects formula. Statistical heterogeneity among studies was assessed using the I² statistic [7]. All analyses were performed at the 0.05 significance level using RevMan version 5.3 software (Nordic Cochrane Centre, Copenhagen, Denmark).

Our search yielded a total of 77 reports. After deduplication and screening, we retrieved 17 reports for potential inclusion in our meta-analysis. We finally included in our quantitative synthesis 10 observational studies, published until May 5, 2021, while no relevant RCTs were identified [8–17]. The detailed selection of studies is provided in detail in the corresponding flow diagram (Supplemental Fig. S1). Only two studies evaluated the effect of DPP-4 inhibitors in the inpatient setting [8,11], while the others assessed COVID-19 surrogate outcomes in diabetes patients who had previously used DPP-4 inhibitors in the outpatient setting [9,10,12–17].

We therefore pooled data from 10 observational studies. The main baseline characteristics across the included studies are summarized in Supplemental Table S1. The use of DPP-4 inhibitors compared to other antidiabetic drug classes produced a non-significant decrease of 3% in the risk for COVID-19 related death (95% CI, 0.67 to 1.41; I²=83%; P=0.89), as shown in Fig. 1. Of note, subgroup analyses according to the setting revealed that, when DPP-4 inhibitors were initiated in the inpatient setting, the result became significant (RR, 0.50; 95% CI, 0.34 to 0.71; I²=0%; P=0.0001), while outpatient use of this antidiabetic drug class was not associated with a significant effect on the risk for COVID-19 death (RR, 1.14; 95% CI, 0.78 to 1.66; I²=81%; P=0.50), as clearly depicted in Fig. 1. Testing for subgroup differences revealed a significant difference between inpatient and outpatient use of DPP-4 inhibitors (P=0.002).

Human DPP-4 has been identified as a transmembrane glycoprotein acting as a functional receptor for the spike protein of coronaviruses, including Middle East respiratory syndrome coronavirus (MERS-CoV) and severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), and playing a crucial role in immune regulation, as it is implicated in the activation of T-cells and upregulation of the CD86 and nuclear factor-κB pathways [18]. Therefore, it may be deduced that DPP-4 might be involved in the cytokine storm and the hyperinflammation observed among patients infected with SARS-CoV-2. It has been previously shown that DPP-4 inhibitors exert significant systemic anti-inflammatory effects, finding with direct implications for COVID-19, by decreasing the concentration of major inflammatory cytokines, such as interleukin-6 (IL-6) [19]. In addition, DPP-4 inhibition has been shown to significantly suppress pulmonary hyperinflammation in acute lung injury, as observed in a relevant experimental model, in which sitagliptin suppressed the lipopolysaccharide-induced lung injury in mice by decreasing the release of cytokines, such as tumor necrosis factor-alpha (TNF-α) and IL-6 by human lung microvascular endothelial cells [20]. To summarize, it seems that there is a strong pathophysiological basis supporting the hypothesis that DPP-4 inhibition can decrease major SARS-CoV-2 complications, such as acute respiratory distress syndrome, which represents the main cause of death in COVID-19 patients. Human mechanistic studies are required to confirm this hypothesis.

Regarding the observed significant effect of the inpatient usage of DPP-4 inhibitors on the risk for COVID-19-related death, the results are based upon only two studies, primarily driven by the results of the observational study conducted by Solerte et al. [8]. More specifically, Solerte et al. [8] demonstrated that sitagliptin treatment compared to standard of care resulted in a significant reduction in C-reactive protein (P=0.001) and procalcitonin (P=0.01) levels at follow-up (30 days post-admission), while no significant difference in levels of other inflammatory markers, such as ferritin and IL-6, was documented. The researchers suggested an acute anti-inflammatory and immunomodulatory effect of sitagliptin, potentially preventing multi-organ failure and associated complications. However, in another study conducted in the inpatient setting, Zhou et al. [11] showed that inpatient usage of DPP-4 inhibitors compared to control did not have a significant effect on neutrophil count, IL-6 and TNF-α, raising doubts regarding whether DPP-4 inhibitors have a true anti-inflammatory effect in COVID-19 patients in the acute setting. Despite these observations, it is important to keep in mind that our results are based only on two studies. Future relevant studies assessing the efficacy of inpatient administration of DPP-4 inhibitors will provide mechanistic insights that could potentially explain these results.

Collectively, despite the fact that current evidence is based upon observational studies, prior use of DPP-4 inhibitors has a neutral effect on COVID-19-related deaths among patients with T2DM, while inpatient administration might contribute to a decrease in the corresponding risk. Ongoing phase 3 RCTs will shed further light on the impact of this antidiabetic drug class on surrogate safety endpoints among patients hospitalized with COVID-19 (ClinicalTrials.gov Identifier: NCT04365517, NCT04371978). According to current knowledge, DPP-4 inhibitors should not be discontinued in infected patients.