Another Viral Infection Linked to Posttransplant Diabetes Mellitus?

In this issue of Transplantation, Vinson et al1 published a large retrospective study incorporating data from the largest data repository for patients with COVID-19 in the United States, the N3C COVID Enclave, which includes 83 medical centers. They examined nondiabetic adult solid organ transplant recipients (SOTRs) between April 1, 2020, and March 31, 2023, with a minimum follow-up time of 6 mo. To evaluate the risk of developing posttransplant diabetes mellitus (PTDM), the authors compared those who developed COVID-19 within the first 180 d after transplantation with those who did not. The authors used 1:1 propensity score matching, following a COVID-positive SOTR with a COVID-negative SOTR who were both diabetes-free at the same point posttransplant. At 730 d of follow-up, 15% of all SOTRs developed PTDM, with 3.3%, 6.7%, and 11% developing PTDM at 90, 180, and 365 d, respectively. They use a multivariable Cox proportional hazards model to demonstrate that COVID-19 increases the hazard of PTDM by 37% at 2 y after transplantation. It is the first study to identify COVID-19 as a risk factor for PTDM. The strengths of this study include its nationwide design, use of a database with large comparison cohorts, minimal loss of patients with 1:1 matching, inclusion of a diverse and wide range of potential confounding factors, and a sensitivity analysis stratified by severity of infection and pre- and postvaccine periods. By evaluating the impact of severe infection, the authors were able to posit a possible exposure–outcome relationship, which strengthens the overall argument for causality. Limitations include the retrospective design, registry-based nature, the use of diagnostic coding and suspected infection for patient inclusion, significant differences in the type of immunosuppression between groups, lack of important medication data (ie, dose, periods of discontinuation), and an absence of data on COVID-19 treatment or graft function. Another significant limitation is that the multivariate model was unable to validate well-established risk factors for PTDM (ie, obesity, diabetogenic immunosuppressive agents). The authors also note that the overall incidence of PTDM in their cohort was on the lower side of most other reports (15% versus 10%–40%), which may indicate differences in baseline characteristics or outcome definitions.2 Vinson et al attempted to provide evidence for COVID-19 positioning itself in the causal pathway for the development of PTDM. Bias and confounding factors, both measured and unmeasured, are important considerations when interpreting observational nonrandomized data. Could there be any differences in unmeasured or possible risk factors for PTDM between the COVID-positive and COVID-negative cohorts in this study? Immunosuppressive medication use is documented qualitatively but not quantitatively, and it is unclear what the cumulative individual drug exposure was for each patient during the follow-up period. This would be important for diabetogenic medications (ie, corticosteroids, calcineurin inhibitors, mammalian target of rapamycin inhibitors).3 Notably, dexamethasone became the standard of care for the treatment of COVID-19 in July 2020 in patients hospitalized and receiving oxygen.4 Episodes of graft dysfunction were not documented and may have altered patient exposure to corticosteroids. COVID-19 itself has been associated with graft dysfunction and failure.5 Rates of cytomegalovirus infection are also not documented in this study. Cytomegalovirus has been associated with PTDM, graft dysfunction, and adjustments in immunosuppressive medication.6 Indeed, there have been multiple viruses linked with DM in the general population, including enteroviruses, hepatitis C virus, and HIV.7 Differences in hospitalization, healthcare utilization, and health-seeking behaviors among the COVID-19-positive cohort may have confounded findings by either increasing testing for COVID-19 and/or PTDM or being associated with unmeasured risk factors. However, it remains plausible that COVID-19 could play a causal role in the development of PTDM. New-onset DM post-COVID-19 infection has been observed in the general population at a similar rate.8 Proposed mechanisms for causing disease have included viral cytopathic effect causing beta-cell damage, autoimmune dysfunction of pancreatic beta-cells, and host response to infection causing increases in proinflammatory cytokines and counter-regulatory hormones leading to insulin resistance.9 There is also a bidirectional relationship between COVID-19 and DM whereby DM increases the risk of severe COVID-19, and new-onset DM is observed with COVID-19. Pancreatic beta-islet cells express angiotensin-converting enzyme 2 receptors, which bind SARS-CoV-2 and can undergo virus-mediated destruction.9 COVID-19 may also promote islet autoimmunity through a variety of mechanisms.9 Both findings can explain the observed presentations of ketosis-prone diabetes seen with COVID-19. Pathogenesis is complicated by the release of proinflammatory cytokines resulting in insulin resistance and pancreatic islet cell dysfunction. An additive effect of risk was not seen when comparing the incidence of new-onset DM in the general population with PTDM seen in SOTRs. This interesting finding may represent a blunting of proinflammatory cytokine-mediated effects and augmentation of virus-mediated effects in SOTRs. This could theoretically provide a similar increased hazard of new-onset DM in both general and SOTR populations. The impact of finding a causal link between COVID-19 and PTDM is not insignificant. PTDM can lead to increased micro- and macrovascular events, graft dysfunction and failure, infection complications, and reduced overall survival.2 This highlights the importance of effective prevention and early treatment strategies in COVID-19 in SOTRs. Immunization and education on respiratory hygiene (eg, mask wearing, hand washing) and avoiding sick contacts remain critical for this vulnerable population. Although this study failed to demonstrate a protective effect for vaccination status and development of PTDM, this may represent poor vaccine immunogenicity (both strength and duration) among SOTRs and the need for increased frequency of COVID vaccine administration in immunocompromised people, which has been a recent recommendation by the United States Centers for Disease Control and Prevention.10 Early antiviral therapy is important in preventing complications related to COVID-19 in vulnerable patients, including SOTRs. It remains to be seen whether this would have an impact on the development of PTDM. This work by Vinson et al identified COVID-19 as a potential independent risk factor for PTDM. Although more work is required to confirm their findings, this study emphasizes the importance of having a heightened clinical awareness and better screening for PTDM after COVID-19.