Herpes simplex viruses type 1 and 2 (HSV-1/2) and varicella zoster virus (VZV) are ubiquitous α-herpesviruses able to establish life-long infection and to reactivate under certain circumstances, such as immunosuppression. Solid organ transplant (SOT) recipients are more prone to experiencing reactivation of α-herpesviruses as compared to the general population. The clinical spectrum may range from minor mucocutaneous forms—orolabial or genital vesicular lesions or localized herpes zoster (HZ)— to disseminated disease with central nervous system (CNS) and visceral involvement [1–3]. In addition to older age, use of valganciclovir as prophylaxis against cytomegalovirus (CMV) and increase of immunosuppressive therapy due to previous rejection episodes [1, 4–6], the factors governing the development of post-transplant HSV/VZV reactivation remain poorly characterized.

Single nucleotide polymorphisms (SNP) in genes coding for immune molecules confer a differential susceptibility to viral pathogens. The contribution of host genetics is highlighted after SOT due to the additive effect of iatrogenic immunosuppression. Therefore, SNP genotyping has emerged as a complementary tool for risk stratification in this population [7].

Herpesviruses are able to modulate adaptive T-cell-mediated responses to maintain latency, with CMV as the most notorious example [8]. The expression of co-inhibitory T-cell receptors plays a relevant role in the virus-host interaction [9–11]. Cytotoxic T-lymphocyte antigen 4 (CTLA-4) and its homologous CD28 are two immunoglobulin superfamily members with a shared ability to bind CD80/B7.1 and CD86/B7.2 but opposed biological functions. CTLA-4 suppresses T-cell receptor signaling, contracts the expanded T-cell populations by inhibiting T-cell proliferation and interleukin-2 secretion, and promotes the suppressive functions of Tregs [12–14]. Different SNPs in the CTLA4 gene have been accordingly investigated in the context of cancer or autoimmune diseases [15] or infection, such as hepatitis C [16] or dengue [17].

The effect of genetic polymorphisms in CTLA4 on the risk of infectious complications in the specific setting of SOT has been assessed in some previous studies [18–21]. Jiang et al. reported a protective role for the GG genotype of rs231775 on the recurrence of hepatitis B virus infection after liver transplantation [19]. Other study revealed that the presence of the mutant genotypes of rs231775 and rs3087243 were associated with a lower CMV disease-free survival in kidney transplant recipients (KTRs) as compared with heterozygous and wild genotypes [20]. In addition, a meta-analysis established a correlation between two CTLA-4 SNPs and the risk of post-transplant infection [21]. Of note, this previous research yielded some discrepant results in the sense of the association found (protective or deleterious).

Due to the alleged impact of CTLA4 SNPs on the host susceptibility and the lack of specific data, we aimed to explore the effect of two CTLA-4 SNPs (rs5742909 and rs231775) on the incidence of α-herpesvirus infection (HSV/VZV) in a cohort of KTRs.

We have shown an association between the presence of the minor allele of rs231775 in the CTLA4 gene and the susceptibility to α-herpesviruses in KTRs. Homozygous carriers of the G allele faced a more than three-fold increase in the incidence of HSV-1/2 and VZV infection, typically in form of mucocutaneous disease and unidermatomal HZ secondary to viral reactivation. This impact was still significant after controlling for well-established risk factors, such as valganciclovir prophylaxis or over-immunosuppression due to recent treatment for acute rejection [1, 4–6].

The CTLA4 (rs231775) SNP consists of a nonsynonymous A/G substitution that implies the change from threonine to alanine, which lead to lower expression levels of membrane-bound CTLA-4 [23, 24]. In keeping with this effect, the GG genotype has been associated with a lower mortality in sepsis patients, a finding presumably attributable to a less pronounced sepsis-associated immunoparalysis [25]. Limited evidence is available regarding the risk of post-transplant infection. In pediatric heart transplant recipients, Ohman et al. reported a significant (albeit modest) association between AA/AG genotypes and the late occurrence of viral infection at the univariable level, but not in the adjusted Cox model. Although the authors did not provided data on specific viral pathogens, it is likely that most episodes were due to primary infection rather than reactivation, in view of the age of the cohort [26]. Iravani Saadi et al. performed a meta-analysis on the basis of 9 studies that found a protective effect for the A allele of the rs231775 SNP (odds ratio: 0.77; 95% CI: 0.59–0.95). Unfortunately, no details on the type of SOT or infection were provided, or whether the genotyping was performed in the donor or the recipient, which limited the possibility of drawing clear conclusions [21].

We are not aware of previous studies that have investigated the effect of genetic polymorphisms in CTLA4 on the risk of post-transplant HSV-1/2 or VZV infection. Thus, the present results should be considered hypothesis-generating only. The mechanistic explanation is not straightforward, since the rs231775 G allele has been shown to reduce the inhibitory function of CTLA-4 through decreased cell surface expression and ligand affinity [23, 24]. This should result in the improved immune control of latent α-herpesvirus infection. Nevertheless, the frequency of the GG genotype of CTLA4 (rs4553808) SNP—which is mapped within the promoter region and also alters its transcription rate—was significantly higher in Chinese KTRs that developed viral infection as compared to those without [18]. It may hypothesized that a lower baseline CTLA-4 expression on polyclonal activated T-cells would render more effective the induction of phenotypically exhausted virus-specific CD8⁺ T-cells, which is one of the immune evasion tactics displayed by HSV and VZV [27, 28]. This susceptibility would be specific for α-herpesviruses, since we have found no association between CTLA4 SNPs and the incidence of CMV infection (data not shown).

Our study is limited by the relatively low number of KTRs that developed infection and the lack of severe cases. Since no data on the baseline HSV-1/2 serostatus was available, we cannot rule out that some episodes were secondary to primary infection rather than reactivation, which would imply a differential role for virus-induced immune evasion. In addition, we lack granular data on the receipt of immunosuppressive therapy before transplantation. Nevertheless, neither the presence of glomerulonephritis as ESRD nor previous SOT (as two surrogate markers for pre-transplant immunosuppression) had an apparent impact on the event of interest. Most episodes of shingles and orolabial HSV infection were diagnosed solely based on clinical findings, and some of them by GPs (although with prompt referral to the transplant outpatient clinic). However, previous studies have reported that GPs have good clinical judgment for the diagnosis of herpes zoster [29]. Finally, the assessment of the confounding effect associated to the use of valganciclovir prophylaxis may have limited by the relatively low number of KTRs in this subgroup.

Future investigations should provide a functional insight into the immune and cellular mechanisms eventually involved in the association observed between CTLA4 polymorphisms and susceptibility to α-herpesvirus infection among KTRs. In the current setting of increasing availability of the HZ/su vaccine for the immunocompromised population, it might be worth exploring whether carriers of the risk-genotype would additionally benefit from extended antiviral prophylaxis during the early post-transplant period.