This investigation intended to explore the relationship between VZV reactivation and patients’ long-term prognoses following allo-HSCT, including CIR, NRM, and OS. In the combined cohort, VZV reactivation was significantly linked with decreased relapse, decreased NRM, and increased OS. The multivariate regression analysis confirmed that VZV reactivation was an independent protective predictor of CIR, NRM, and OS. For disease-type-based subgroup analysis, whether in AML or ALL patients, the multivariate regression analysis confirmed that VZV reactivation was still an independent protective predictor of CIR and OS. Our findings provide a way to lessen recurrence following allo-HSCT.

The prognosis could be influenced by the patient's pre-transplant status, the transplant procedure, post-transplant complications, and donor status. In these two groups, the variations in patient demographic features, disease type, DRI, conditioning regimen, the incidence of aGVHD, the incidence of cGVHD, the incidence of CMV DNAemia and donor demographic characteristics were not statistically significant. Also, after multivariate analysis, VZV reactivation proved to remain significant and was the strongest risk factor. This boosted the reliability of the findings of this study.

Whether VZV reactivation substantially contributes to the graft-versus-leukemic effect is unknown. However, it is worth investigating whether cellular immunity induced by VZV reactivation has an anti-leukemic effect.

Firstly, the reactivation of CMV, another herpes virus, may induce cellular anti-leukemic immunity [25]. Much has been learned about the mechanisms of anti-leukemic effects induced by CMV reactivation after allo-HSCT, which propose that circulating Vδ2negγδ T-cells expand strongly and permanently, and these cells recognize leukemic blasts via their TCR, and CD8αα probably serves as a co-receptor in antigen recognition [10, 15, 26]. It has been well documented that post-transplant or immunocompromised patients with VZV reactivation had markedly decreased CD4+ T cells and surface CD28 expression but increased CD8 + T cells [27, 28]. This immunological response to VZV infection was not weakened by acyclovir treatment [29]. Moreover, CD8+ T cells, specifically, cytotoxic T cell lymphocytes, are essential components in the immune response to viruses, and these cells are also crucial effector cells in the immunological reactions against cancer [7,8,9].

Secondly, CMV infection may increase NRM which counteracts the benefit of the graft-versus-leukemic effect, and the graft-versus-leukemic effect may be influenced by several factors. A Japanese study involving 3,539 patients who underwent allo-HSCT demonstrated that CMV reactivation was significantly linked with reduced disease recurrence only in AML patients, but not in ALL, CML, or MDS patients. Additionally, CMV reactivation was strongly associated with increased NRM and overall mortality [15, 30]. But in the combined cohort in our study, VZV reactivation was not only significantly related to decreased CIR, but also decrease NRM. Due to the limited number of patients in this study, a disease-based subgroup analysis solely examined the AML patients and ALL patients, which revealed that the beneficial impact of VZV reactivation on disease recurrence and OS was not restricted to AML patients. Meanwhile, the Varicella-zoster virus serves as the only human herpes virus with a highly effective vaccine [18]. Unlike most vaccinations, which predominantly produce antibodies for protection, the varicella vaccine primarily produces cellular immunity [31]. Additionally, the Oka VZV vaccine strain could potentially elicit decades-long antiviral antibodies, lymphoproliferative responses, and cytotoxic T-cell responses. Furthermore, existing studies indicated that the inactivated VZV vaccine was safe for patients with haematological malignancies who had undergone chemotherapy or hematopoietic stem cell transplantation [32, 33]. Moreover, a credible assessment approach of VZV-specific cell-mediated immunity has been proposed to quantify the expression of the ifn-γ gene in 24-h-stimulated whole blood [34], which could aid in understanding the cellular immunity of VZV reactivation. However, VZV reactivation will damage the quality of life for the patients. Tatebe et al. pointed out that VZV reactivation following HSCT in children could be prevented by using low-dose acyclovir [35]. Therefore, as advised in the guidelines, routine prophylactic antiviral therapy should be provided to all patients following allo-HSCT. Our ultimate objective is to identify the varicella-zoster virus antigen, which could activate particular immune cells with antileukemic properties. These provide opportunities for the development of VZV-related oncology vaccines.

Additionally, even in AML patients, Bao et al. discovered that in the ATG-using cohort, T-cell depletion diminished the anti-leukaemia effects induced by CMV reactivation [36]. But in patients with VZV reactivation in this study, more patients used ATG than patients without, although the lack of statistical significance, which implied that VZV reactivation may result in robust anti-leukaemia effects, and the anti-leukaemia effects may be less affected by other factors.

Therefore, it is plausible and worthy to hypothesize that VZV reactivation may directly contribute to anti-leukemic effects. To confirm this, prospective studies are urgently needed to assess the enduring immunological response of patients with VZV reactivation following allo-HSCT. It is reasonable to postulate that VZV reactivation directly contributes to the substantial long-term antileukemia effect reported in this study.