The relapse of acute myeloid leukemia (AML) after allogeneic hematopoietic stem cell transplantation (allo-SCT) confers a poor prognosis. The prevention of leukemia relapses becomes a reasonable strategy. Prognostic factors identification is critical in selecting high-risk patients for early interventions. In previous studies, both AML-related and transplant-related risk factors have been demonstrated.1

Cytomegalovirus (CMV) reactivation after human leukocyte antigen (HLA)-matched allo-SCT has been shown to reduce AML relapse rates.2-6 The underlying immune-mediated anti-leukemic mechanisms which are enhanced by CMV reactivation have been proposed. The CMV-induced adaptive natural killer (NK) cells characterized by the NKG2C+CD57+ expression on the CD56dim+ NK cell have been reported to be associated with lower AML relapse in allo-SCT settings.7-9 In addition, a subgroup of gamma-delta T cells (Vδ2neg γδ-T cell) responding to CMV reactivation was demonstrated after allo-SCT.10 Furthermore, the T-lymphocyte, CD56+ T cell subset and NK cell reconstitutions were accelerated early after SCT by CMV viremia.8,11,12 These escalated immune cells may contribute to the protection against AML relapses.

CMV-seropositive healthy blood donors have higher numbers of adaptive NK cells (NKG2C+) and CD56+ T-lymphocytes, compared to seronegative donors.13,14 The loss of relapse prevention by pre-emptive anti-CMV therapy in a region with a high CMV prevalence was reported.15 Therefore, the CMV-influent immune activation response in a high CMV-seropositive population may be different from impacts in low CMV rate countries. This study explored clinical factors, including CMV-reactivation, which were associated with transplant outcomes and characterized CMV-mediated immune response after HLA-matched myeloablative allo-SCT for AML patients in a high CMV prevalence country.

The retrospective cohort study enrolled AML patients who underwent HLA-matched myeloablative allo-SCT after complete remission at the King Chulalongkorn Memorial Hospital in Bangkok, Thailand from 2011 to February 2021. Thailand has a high CMV prevalence. The use of ATG-containing regimens was an exclusion criterion. The protocol was approved by the Institutional Review Board of the Faculty of Medicine, Chulalongkorn University, Bangkok, Thailand.

The transplantation protocol included a myeloablative conditioning regimen with cyclophosphamide plus total body irradiation (Cy-TBI), or busulfex plus cyclophosphamide (Bu-Cy), or fludarabine plus 4-day busulfex (Flu-Bu4). The graft-versus-host disease (GVHD) prophylaxis regimen was the calcineurin inhibitor and methotrexate, or post-transplantation cyclophosphamide plus cyclosporine and mycophenolate mofetil. The infectious prophylaxis protocol was comprised of acyclovir prophylaxis for herpes virus and fluconazole for primary fungal prevention or voriconazole for secondary aspergillosis prevention. No anti-bacterial and anti-CMV (letermovir) prophylaxis protocol was prescribed. The acute GVHD (aGVHD) grading system used the modified Glucksberg criteria and the chronic GVHD (cGVHD) assessment used the NIH 2005 criteria.

The post-transplantation CMV-viral load (VL) monitoring was scheduled weekly, following white blood cell engraftment for 2 to 3 months post-SCT. The reactivation was defined as plasma DNA of CMV-VL > 500 copies/ml, with detection at least two times. For patients taking high-dose corticosteroids, the criteria of plasma CMV-VL > 100 copies/ml was defined as positive reactivation. The preemptive therapy with ganciclovir or valganciclovir was initiated when CMV reactivation occurred. The clinically significant CMV infection was defined as CMV viremia leading to preemptive treatment, CMV syndrome or CMV disease.

The T and NK cell immunophenotypes were performed in thawed peripheral blood mononuclear cells from the available twenty-seven patients on day 90 post-SCT, in addition to five healthy volunteers. The single cell suspension was incubated with fluorochrome-conjugated monoclonal antibodies in the dark at 4°C for 30 minutes. Cells were washed 2 times and then analyzed on the BD FACSCanto II flow cytometer. The dead cells were excluded, using 7-amino-actinonycin D (7-AAD) staining. The data were analyzed on the FlowJo software. The fluorochrome-conjugated monoclonal antibodies used included: CD56-PE (clone REA196, MACS), NKG2C-VioBright FITC (clone REA205, MACS), CD57-APC (clone REA769, MACS), PD1-APC (clone EH12.2H7, Biolegend), CD3-APC-Vio770 (clone BW264/56, MACS) and TCRγδ-FITC (clone 11F2, MACS).

The clinical data were analyzed to calculate the cumulative incidence of leukemia relapse (CIR), relapse-free survival (RFS) and overall survival (OS). The RFS was defined from the date of stem cell infusion to the date of a relapsed disease or death and measured in days. The OS was defined from the date of stem cell infusion until death from any cause and measured in days. We used the Kaplan–Meier method to estimate the probabilities of RFS and OS up to 3 years. Thereafter, we explored factors associated with RFS and OS using the Cox Regression. The linearity of continuous covariates against the hazard function was assessed and, in the case of non-linearity, the covariate was modelled in quartiles; adjacent quartiles, collapsed in hazard ratio (HR) and 95%CI, were similar. Covariates with p < 0.1 were adjusted for, in a multivariate model and p-values < 0.05 were considered statistically significant. The data were analyzed using the Stata15 (Statacorp, College Station, TX, USA).

In this study, the strongest independent factor associated with an inferior RFS and OS after allo-SCT was transplantation beyond the first CR. The primary refractory to the 3+7 regimen also tended to yield a poor RFS and OS, although when controlled for >CR1 and cGVHD, the significance was lost. These findings are consistent with previous studies and support the linkage of these factors with chemotherapy resistance.1 The unfavorable cytogenetic risk was not detected as a significant factor in this cohort, possibly due to the limited number of patients. In addition to leukemic risk factors, the cGVHD was a significant prognostic factor associated with a better RFS and OS in our study, similar to many previous reports.1 This referred to the graft versus leukemia effect. The TRM was low in our cohort. The major cause of death was relapsed AML, which accounted for 88.9% of deaths.

In our center in Thailand, which has a high CMV prevalence, no difference in the CIR between the CMV-reactivation and non-reactivation was demonstrated. Due to the small number of subjects, a weak association cannot be excluded. Existing literature remains conflicting. Although a meta-analysis showed a lower relapse risk with the CMV replication,3 a large multicenter registry (n = 5,310) reported no preventive effect from the CMV reactivation in AML patients after allo-SCT.17 Heterogeneous results may depend on some variables that abrogate this advantage, for example, antithymoglobulin,18,19 reduced-intensity conditioning regimen 20 and antiviral treatment.15 The intensity of CMV-activated immune responses are possibly more important for relapse protection than the CMV itself. The early pre-emptive CMV therapy may be partly responsible for the lack of protection in our study. The low threshold for initiating the anti-CMV treatment in our cohort also led to 49.4% of patients having a clinically significant CMV infection. No CMV syndrome/disease and CMV-related mortality occurred. Although letermovir has been approved and shown efficacy and safety for CMV prevention in seropositive allo-SCT patients,21,22 the effects of this prophylactic strategy and CMV-related relapse protection requires more investigation. In Thailand, letermovir is currently not yet available.

The characterization of CMV-related immune response was performed in several blood samples. An increase in the CD56+ T lymphocyte and CMV-induced adaptive NK cells (NKG2C+CD57+) was detected without statistical significance in the CMV reactivation patients. The CD56+ T lymphocyte was previously described as the adaptive T cell subset that increased in response to the CMV replication after both an allo-SCT setting 8,23 and CMV-seropositive healthy donors.14 Consistent with our findings of NK cells, a recent study also reported no significant difference in the NK-cell reconstitution on day 100 between the CMV-reactivation and non-reactivation of CMV-seropositive patients, using the PBSC as the graft source.24 The major limitation of our study were small numbers for sample analysis and the lack of a functional analysis of the CMV cell-mediated immunity.25

The limitations of this study are the lack of minimal residual disease information and some gene mutational data. This study also conducted before the FLT3 inhibitor was combined with the treatment regimen at our center. In addition, the high probability of the survival outcome in this cohort may be influenced from some biases of baseline characteristics, such as young age, low proportion of poor-risk patients and use of myeloablative regimen.

In summary, our study found that transplantation beyond the CR1 and absence of cGVHD are powerful prognostic factors associated with an inferior RFS and OS. With a high CMV prevalence in the Thailand setting, no impact of the CMV reactivation on the relapse in AML patients undergoing an allo-SCT was found in our study.