Cytomegalovirus infection is associated with rapid NK differentiation and reduced incidence of relapse in HLA matched sibling donor transplant patients

Abstract

The effect of cytomegalovirus (CMV) infection on leukemia relapse and the potential mechanism remains controversial. In this retrospective study, we evaluated the association among CMV infection, NK reconstitution and clinical outcomes in consecutive patients with hematologic malignancy who underwent HLA matched sibling donor transplantation (MST). In total, 228 patients were enrolled in the study between January 2010 and December 2011. The cumulative incidence of CMV infection on day 100 post-HSCT was 13.6 ± 4.9%. The probabilities of OS and DFS were 45.4% vs. 71.7% (P = 0.004) and 43.9% vs. 64.2% (P = 0.050) in the patients with CMV infection and without CMV infection, respectively. The cumulative incidence of treatment-related mortality (TRM) and relapse at 5 years was 48.6 ± 9.6% vs. 11.5 ± 2.9% (P < 0.001) and 6.2 ± 4.3% vs. 29.2 ± 3.9% (P = 0.024) in the patients with CMV infection and without CMV infection, respectively. In the multivariate analysis, CMV infection was associated with higher TRM, lower OS, and lower DFS. In addition, we found that CMV infection may promote the recovery of the absolute number of NK cells and promote the differentiation of NK cells post-MST. In conclusion, CMV infection may promote the recovery and differentiation of NK cells and was correlated with a lower relapse rate post-MST.

CMV infection associated with decreased relapse rate post MST. CMV infection associated with the recovery and differentiation of NK cells post MST. Rapid NK cells recovery and differentiation associated with decreased relapse post MST.

Graphical Abstract

Graphical Abstract.

Introduction

Allogeneic hematopoietic stem cell transplantation (allo-HSCT) is an effective curative option for patients with hematological disease [1–3]. Treatment-related mortality (TRM) and relapse lead to the failure of allo-HSCT. Cytomegalovirus (CMV) infection remains the most common complication post-HSCT [4]. CMV infection is a double-edged sword for allo-HSCT patients. On the one hand, CMV infection can increase the risk of TRM after allo-HSCT. On the other hand, CMV infection can reduce the risk of leukemia recurrence in some transplantation settings [5–17]. Some recent reports suggest that CMV infection is associated with a prominently reduced risk of leukemia relapse in adult patients with myeloid malignancy [10, 14, 16, 18–20] and lymphoma [9]. In contrast, the protective effect of CMV infection on reducing leukemia relapse has not been demonstrated in patients with acute lymphoid leukemia (ALL) [17], lymphoma [17, 21], myelodysplastic syndrome (MDS) [22], and pediatric acute leukemia [23]. Therefore, the effect of CMV infection on leukemia relapse is still controversial in different transplantation settings and different hematological malignancies [12, 14, 24]. Moreover, the probable mechanism by which CMV infection reduces relapse remains unknown.

Natural killer (NK) cells are innate immune lymphocytes that respond to viral infection and tumor cells and are the first lymphocytes reconstituted after allo-HSCT [25–30]. Recent studies demonstrated that CMV infection after transplantation may promote NK education and trigger a distinctive NK-cell population expansion that expresses the NKG2C receptor, exhibits enhanced effector functions [31–36] and is associated with less leukemia relapse after HLA matched sibling or cord blood transplantation [33]. A previous study demonstrated that HCMV infection can drive rapid NK maturation characterized by the expansion of CD56^dimNKG2A^-KIR⁺ cells, even in the absence of NKG2C expression [36]. Meanwhile, the observations that NKG2A, KIR, and CD57 are independently correlated with proliferative responses along with the distinct expression patterns of these molecules on naïve and adult human NK cells suggest that the loss of NKG2A, sequential acquisition of KIRs, and CD57 expression reflect a continued differentiation process occurring in CD56^dim NK cells [37]. The effect of CMV infection on NK-cell differentiation after transplantation is not very clear. We hypothesized that CMV infection could promote NK-cell differentiation with the expansion of CD56^dimKIR⁺ CD57⁺ cells after allo-HSCT, which could reflect the response of NK cells against CMV infection.

To identify the influence of CMV infection on clinical outcomes and reveal the potential mechanism in patients with MSD transplantation, we retrospectively analyzed patients with MSD in our single center.

Material and methods

Study design

This single-center retrospective study involved consecutive patients with hematologic malignancy who underwent allo-HSCT. The present study was performed at Peking University People’s Hospital and was approved by the Ethics Committee of Peking University People’s Hospital. To evaluate the effect of CMV infection on relapse, patients diagnosed with hematologic malignancy who received MST from Jan 2010 to Dec 2011 were enrolled.

To explore the effect of CMV infection on NK maturation post-transplantation, this prospective study recruited patients with acute leukemia who underwent MST between July 2012 and May 2014. Patients aged 15–65 years with acute leukemia in complete remission (CR) diagnosed with non-Ph+ALL or T-ALL who received HLA matched myeloablative allo-HSCT were eligible for inclusion in this trial. Finally, 29 patients were enrolled.

The primary endpoint was the relationship between CMV infection and the cumulative incidence of relapse and NRM following HSCT. The second endpoint was the relationship between NK reconstitution and CMV infection.

Conditioning regimen and stem cell collection

All patients received matched sibling transplantation (MST) under a myeloablative conditioning regimen (MAC) without in vitro T-cell depletion. The conditioning regimen used in our center includes BU-based and total body irradiation (TBI)-based regimens. The patients received hydroxycarbamide (80 mg/kg) orally on day −10, a lower dose of cytarabine (2 g/m² per day) on day −9, busulfan (3.2 mg/kg per day) intravenously on days −8 to −6, cyclophosphamide (CY, 1.8 g/m² per day) intravenously on days −5 to −4, and Me-CCNU (250 mg/m²) orally once on day −3. The TBI-based conditioning regimen was composed of TBI (770 cGy for a single dose) on day −6 and cyclophosphamide (CY, 1.8 g/m² per day) intravenously on days −5 to −4; Me-CCNU (250 mg/m²) was administered orally once on day −3.

All subjects received fresh granulocyte colony-stimulating factor (G-CSF)-mobilized bone marrow (BM) and peripheral blood cells. Ganciclovir was administered during conditioning (through day −2), and acyclovir (400 mg twice a day) was given until the discontinuation of all immunosuppressive agents. The patients also received prophylactic drugs to prevent infection by fungi.

GVHD prophylaxis

The prophylaxis for acute GVHD (a-GVHD) comprised cyclosporine, mycophenolate mofetil (MMF), and short-term methotrexate (MTX). MMF was withdrawn when WBCs were engrafted post-HSCT.

CMV monitoring and preemptive treatment

Within 100 days post-HSCT, CMV and EBV infection was monitored weekly by plasma DNA testing using real-time PCR (RT-PCR) (kits from PG Biotech Co., Ltd., Shenzhen, China with a detection limit of 500 copies/ml). Once patients with CMV copies over 1000 copies/ml were diagnosed with CMV infection post-HSCT, intravenous ganciclovir or forscanet was given until two consecutive negative plasma PCRs were confirmed.

Definitions and endpoints

Neutrophil engraftment was defined as an absolute neutrophil count of 0.5 × 10⁹/L or more for 3 consecutive days. Platelet engraftment was defined as a platelet count of 20 × 10⁹/L or more for 7 consecutive days without transfusion. Acute and chronic GVHD were diagnosed and graded according to established criteria. Risk stratification was performed to differentiate the patients into the following two groups: AML and ALL in the first or second remission (CR1 or CR2), chronic myeloid leukemia (CML) in the first or second chronic phase (CP1 or CP2), and MDS with refractory anemia with or without ringed sideroblasts were defined as the standard-risk group, and other malignant diseases were defined as the high-risk group. The overall survival (OS) times were measured as the date of HSCT until death from any cause. Disease-free survival (DFS) was defined as the time from transplantation to relapse or death from any cause. The surviving patients were censored on the date of their last follow-up. Treatment-related mortality (TRM) was defined as death after allo-HSCT without disease progression or relapse. Relapse was defined by the appearance of morphological evidence of the disease obtained from testing samples from the peripheral blood, bone marrow, or extramedullary sites or the recurrence and sustained presence of pretransplantation chromosomal abnormalities.

Blood sample preparation and immune reconstitution monitoring

The 29 subjects who underwent MST were monitored for the reconstitution of NK cells. Blood samples were obtained on days 30, 60, 100, and 180 and 1 year after HSCT. All samples were treated with stains for cell surface markers to detect the recovery of CD56⁺ and T cells.

The monoclonal antibodies CD3-V500, CD3-Percp, CD4 PerCP, CD8-APC, CD56-PE, CD62 L-FITC, CD69-APC, NKP30-Alexa 647, CD158a-FITC, CD158b-FITC, CD158e-FITC (BD Bioscience, Mountain View, CA, USA), CD3-APC/Cy7, CD16-APC/CY7, CD56-APC/CY7, CD127-Brilliant violet, CD57-Pacific Blue (Biolegend, San Diego, CA, USA), CD25-PE-Cy7 (eBioscience, San Diego, CA, USA) (BD Bioscience, Mountain View, CA, USA) (Biolegend, San Diego, CA, USA), and NKG2A-PercP (R & D systems, USA) and their relative allotypes were used in individual 8-color flow cytometry assays to analyze the immunophenotypic recovery of CD56⁺ NK-cell subsets, i.e., CD56^dim and CD56^bright, as their relative allotypes were used in individual 7-color or 8-color flow cytometry assays.

Statistical methods

The data were up-to-date as of December 31, 2019. The data were censored at the time of relapse, TRM, or last follow-up. The demographic and clinical characteristics of the patients and the reconstitution of NK cells were compared between patients with CMV-positive and CMV-negative results using χ2 and Fisher’s exact tests for dichotomous variables. Student’s t test was used to analyze data with a normal distribution, and the Mann–Whitney U test was used to analyze data without a normal distribution.

The probabilities of OS and DFS were estimated using the Kaplan–Meier method. Univariate and multivariate time-dependent Cox proportional hazard models were used to examine the risk factors associated with relapse reported in previous studies. The cumulative incidences of TRM and relapse were estimated using a competing risks model. Relapse and TRM were treated as competing events. A P-value <0.05 in a two-sided test was considered indicative of significance. The multivariate Cox proportional model and survival analysis were performed with SPSS software (SPSS 22.0, Chicago, IL, USA). The cumulative incidence was calculated with R statistical software, version 3.6.0 (R Foundation for Statistical Computing, Vienna, Austria).

Results

CMV infection and clinical outcomes post MST

The patient characteristics are depicted overall and according to CMV infection in Table 1. From January 2010 to December 2011, 228 patients who received matched sibling transplantation (MST), including 153 male patients and 75 female patients, were enrolled in the study. The patients were diagnosed with AML (n = 106), ALL (n = 69), MDS (n = 18), CML (n = 31), CMML (n = 2) and lymphoma (n = 2). The median age of the patients was 39 years (range, 2–63). CMV infection within 100 days post-HSCT was observed in 37 patients (16.2%), with a median time of 44 days (range, 15–99) from transplantation. The cumulative incidence of CMV infection on Day 100 post-HSCT was 13.6 ± 4.9% in the patients who received MST. Only one patient was diagnosed with CMV-associated pneumonia and died. EBV infection was observed in 3 patients (1.3%), including 2 patients diagnosed with EBV viremia and 1 patient diagnosed with EBV-PTLD.

Table 1:

patients’ characteristics

Characteristics, n%	CMV+ (n = 37)	CMV− (n = 191)	P
Patient age, year
Median (range)	45 (20–63)	36 (2–62)	<0.001
Patient gender
Male	29 (78.4%)	123 (64.4%)	0.099
Female	8 (21.6%)	68 (35.6%)
Diagnosis
ALL	11 (29.8%)	57 (29.8%)	0.991
AML	18 (48.6%)	88 (46.1%)
MDS	3 (8.1%)	15 (7.8%)
CML	5 (13.5%)	27 (14.1%)
CMML	0	2 (1.1%)
Lymphoma	0	2 (1.1%)
Disease status
Standard risk	28 (75.7%)	159 (83.2%)	0.272
High risk	9 (24.3%)	32 (16.8%)
CMV serostatus
D+/R+	29 (78.4%)	139 (72.8%)	0.758
D+/R−	1 (2.7%)	9 (4.7%)
D−/R+	1 (2.7%)	10 (5.2%)
D−/R−	1 (2.7%)	8 (4.2%)
Not available	5 (13.5%)	25 (13.1%)
EBV reactivation
Yes	3 (8.1%)	9 (4.7%)	0.406
No	34 (91.9%)	182 (95.3%)
Relapse
Yes	6 (16.2%)	48 (25.1%)	0.243
No	31 (83.8%)	143 (74.9%)
Acute GVHD
Grades 0–1	21 (56.8%)	181 (94.8%)	<0.001
Grades 2–4	16 (43.2%)	10 (5.2%)
Chronic GVHD
Limited/extensive	23 (62.2%)	106 (55.5%)	0.056
None	14 (37.8%)	85 (44.5%)
TRM (cause)	15	18	0.003
CMV infection	1	0
aGVHD	2	0
cGVHD	6	1
Hemorrhage	0	4
Infection	3	11
Others	3	2

ALL, acute lymphocyte leukemia; AML, acute myeloid leukemia; MDS, myelodysplastic syndrome; CML, chronic myeloid leukemia; CMML, chronic myelomonocytic leukemia; acute GVHD, acute graft-versus-host disease; CMV, cytomegalovirus; EBV, Epstein-Barr virus; TRM, treatment-related mortality.

With a median follow-up of 1696 days (range, 312–3644) after HSCT, in the entire cohort, the probability of OS and DFS at 5 years was 65.0% (95% CI 57.6–72.4%) and 60.8% (95% CI 53.8–67.8%), respectively. The cumulative incidence of relapse at 5 years was 24.6 ± 13.5%. The cumulative incidence of TRM at 5 years was 19.6 ± 3.4%.

In the patients with CMV infection, only one patient died from CMV pneumonia, 2 patients died from aGVHD, 6 patients died from cGVHD, and 3 patients died from severe infection. However, among the patients without CMV infection, most patients (11/18, 61.1%) died from severe infection. The probability of OS was 45.4% (95% CI 27.0–63.8%) in the patients with CMV infection and 71.7% (95% CI 64.5–78.9%) in the patients without CMV infection (P = 0.004; Figure 1A). The probability of DFS was 43.9% (95% CI 26.1–61.7%) in the patients with CMV infection and 64.2% in the patients without CMV infection (95% CI 56.8–71.6%) (P = 0.050) (Figure 1B). The cumulative incidence of TRM in the patients with CMV infection was significantly higher than that in the patients without CMV infection (48.6 ± 9.6% vs. 11.5 ± 2.9%, P < 0.001; Figure 1C). The cumulative incidence of relapse at 5 years in the patients with CMV infection was significantly lower than that in the patients without CMV infection (6.2 ± 4.3% vs. 29.2 ± 3.9%, P = 0.024; Figure 1D). In the multivariate analysis, CMV infection was associated with a higher TRM and lower OS and DFS (Table 2).

Figure 1:

effect of CMV infection on the OS (A), DFS (B), NRM (C), and relapse (D) in patients post MST

Table 2:

multivariate analysis of risk factors for relapse, NRM, DFS, and OS for patients with MST

	Variables	HR (95%CI)	P
CIR	c-GVHD	0.315 (0.176–0.565)	<0.001
	CMV infection	0.282 (0.068–1.165)	0.080
TRM	c-GVHD	0.322 (0.153–0.680)	0.003
	a-GVHD	1.159 (0.580–2.642)	0.726
	CMV infection	7.049 (3.378–14.711)	<0.001
OS	c-GVHD	0.304 (0.180–0.512	<0.001
	Disease	0.072 (0.016–0.330)	0.001
	CMV infection	3.071 (1.746–5.402)	<0.001
DFS	c-GVHD	0.336 (0.212–0.533)	<0.001
	Disease	0.137 (0.031–0.604)	0.009
	CMV infection	2.122 (1.253–3.593)	0.005

CIR, cumulative incidence of relapse; TRM, treatment-related mortality; OS, overall survival; DFS, disease-free survival; CMV, cytomegalovirus; a-GVHD, acute graft-versus-host disease; c-GVHD, chronic graft-versus-host disease.

CMV infection associated with deceased relapse in AML patients after MST

In the patients diagnosed with AML (n = 106), the probability of OS was 48.1% (95% CI 21.1–75.1%) in the patients with CMV infection and 73.6% (95% CI 61.6–85.6%) in the patients without CMV infection (P = 0.003; Figure 2A). The probability of DFS was 40.2% (95% CI 13.8–66.6%) in the patients with CMV infection and 66.6% in the patients without CMV infection (95% CI 56.2–77.0%) (P = 0.092; Figure 2B). The cumulative incidence of TRM in the patients with CMV infection was significantly higher than that in the patients without CMV infection (59.1 ± 13.9% vs. 4.7 ± 2.7%, P < 0.001; Figure 2C). The cumulative incidence of relapse at 5 years in the patients with CMV infection was significantly lower than that in the patients without CMV infection (5.9 ± 5.7% vs. 33.4 ± 5.8%, P = 0.040; Figure 2D).

Figure 2:

effect of CMV infection on OS (A), DFS (B), NRM (C), and relapse (D) in AML patients post MST

Among the patients diagnosed with ALL, CML, MDS and other conditions, CMV infection did not influence the cumulative incidence of relapse.

CMV infection promoted NK-cell expansion after MST

In the MST group, 13 patients (13/29) experienced CMV infection. There were no differences in the absolute numbers of CD3+ T cells on days 30, 60, and 100 post-MST between the patients with or without CMV infection (Figure 3A). In the patients with CMV infection, the absolute numbers of NK cells on day 100 were significantly increased compared to those on days 30 and 60 (P = 0.001 and 0.03, respectively, Figure 3B). Therefore, the absolute numbers of NK cells on day 100 in the patients with CMV infection were higher than those in the patients without CMV infection post-MST (P = 0.002, Figure 3B), and the NK/T ratio on day 100 was significantly higher than that in those without CMV infection on day 100 (P = 0.001, Figure 3C), indicating that CMV infection stimulated greater expansion of NK cells post-MST.

Figure 3:

kinetics of T-cell and NK-cell reconstitution post MST. Comparisons of the absolute numbers of T cells (A) and NK cells (B) and the ratio of NK cells and T cells (C) between patients with or without CMV infection post-MST

CMV infection promoted NK-cell differentiation after MST

In the MST group, compared to the patients with CMV infection, the patients without CMV infection (n = 16) showed fewer CD56^dim subsets (P = 0.001 on Day 60, P = 0.010 on Day 90, Figure 4A), greater expression of KIR (P = 0.048 on Day 60 and P = 0.046 at 1 year, Figure 4B) and CD57 (P = 0.083 on day 60, P = 0.007 on day 90, Figure 4C) and, therefore, more mature CD56^dimKIR⁺CD57⁺ NK subpopulations (P = 0.012 on day 60, P = 0.033 on day 90, P = 0.025 at 1 year, Figure 4D) posttransplantation.

Figure 4:

kinetics of NK-cell differentiation post-MST. Comparisons of the percentage of CD56^dim NK cells (A), expression of KIR (B) or CD57 (C) on CD56^dim NK cells, and percentage of CD56^dimNKG2A−KIR⁺CD57⁺ NK cells (D) between patients with or without CMV infection post-MST. Comparisons of the percentage of CD56dimNKG2A−KIR⁺CD57⁺ NK cells (E) between patients with or without relapse post-MST

However, the antileukemia expression of CD107a by CD56dimKIR+CD57+ NK cells was comparable on days 60, 100, and 180 between the patients with or without CMV infection post-MST (data not shown).

In the MST group, 8 patients (8/29) experienced relapse. The CD56dimKIR+CD57+ NK subpopulations in the patients without relapse were significantly higher than those in the patients with relapse (P = 0.017 on day 100, Figure 4E) post-transplantation.

Percentage of CD56^dimKIR⁺CD57⁺ NK cells correlated with CD56^dimNKG2C⁺CD57⁺ NK cells

Previous reports have demonstrated that CMV could stimulate the expansion of adaptive NK cells after cord blood transplantation, and these NKG2C+ NK cells preferentially acquire the maturation marker CD57 and are enriched in the expression of educating killer immunoglobulin-like receptors (KIRs). In healthy donors and patients before transplantation, significant correlations were found between the percentages of CD56^dimNKG2C⁺CD57⁺ NK and CD56^dimKIR⁺CD57⁺ NK cells as shown in Figure 5. However, the percentage of CD56^dimNKG2C⁺CD57⁺ NK cells was consistently higher than that of CD56^dimKIR⁺CD57⁺ NK cells in the healthy donors and patients before transplantation, indicating that these two populations overlapped.

Figure 5:

correlations and comparisons of the percentage of CD56^dimNKG2A^-KIR⁺CD57⁺ and CD56^dimNKG2C⁺CD57⁺ cells in healthy donors (A, C) and patients before transplantation (B, D)

Discussion

Allogeneic hematopoietic stem cell transplantation (allo-HSCT) remains a powerful treatment strategy for malignant hematological disease [1, 38]. Although obvious improvements have been achieved in the prevention of post-HSCT relapse in the past few years, TRM and relapse are still the two main causes of treatment failure after HSCT [39].

CMV infection, as one of the most common and important complications after allo-HSCT, is the main reason for the high TRM previously reported [40]. However, the effect of CMV infection on relapse is controversial, and the mechanism unclear. In our present study, we found that CMV infection promotes the recovery of NK cells and the differentiation of NK cells post-MST. Meanwhile, we found that CMV infection significantly decreased the relapse rate after MST, especially among AML patients. Although a lower relapse rate was associated with CMV infection, the patients with CMV infection also had a high NRM; therefore, patients do not benefit from CMV infection in the setting of MST.

The discrepancies between CMV infection and leukemia relapse could be related to differences in treatment protocols, including the use of antithymocyte globulin (ATG), graft sources, and treatment of CMV reactivation. In a cohort of adult AML patients, the results showed that CMV replication was associated with a significantly reduced risk of relapse after HSCT [41]. Green et al. reported studies involving a large cohort including different hematological malignancies and confirmed that the antileukemia effect of CMV replication was only detected in patients diagnosed with AML [17]. The association between early CMV infection after HSCT and reduced AML relapse risk was also confirmed in recent studies [14, 20, 24, 42]. In addition, a protective effect was demonstrated in patients with a myeloablative conditioning regimen [14, 24] and patients treated with in vivo T-cell depletion therapy, such as ATG or alemtuzumab [24, 41]. Our results in the setting of MST were in accordance with the abovementioned previous results.

The antileukemia effect of CMV infection after HSCT in AML may depend on CMV-stimulated innate and adaptive immune responses to cancer cells. A recent retrospective study showed that in patients with CMV infection, CD56^brightCD16^dimDNAM1⁺ NK cells steadily increased until day 90. In patients without CMV infection, however, the increase in CD56^brightCD16^dimDNAM1⁺ NK-cell counts were not maintained after day 30, and these cells decreased in number. The recovery of CD56brightCD16dimDNAM1+ NK cells in CMV-infected patients provides valuable information for understanding the plausible immunologic mechanism of the graft-versus-leukemia effect [43]. Meanwhile, studies have focused on the expansion of CD56^dimCD57⁺NKG2C⁺ NK cells following CMV infection as a potential immunomodulatory mechanism [33, 35, 44]. However, direct evidence of the enhanced antileukemia effect of CMV-expanded CD56^dimCD57⁺NKG2C⁺ NK cells is lacking. Young et al. demonstrated that A2AR adenosine signaling suppresses NK-cell maturation in the tumor microenvironment, and ablating adenosine signaling was found to promote natural killer cell maturation and antitumor immunity and reduce tumor growth [45]. Cichocki et al. also demonstrated that GSK3 inhibition drives the maturation of NK cells and enhances their antitumor activity. The maturation of human natural killer (NK) cells, as defined by the accumulation of cell-surface expression of CD57, is associated with increased cytotoxic characteristics and TNF and IFN-gamma production upon target-cell recognition [46]. In this study, we first found that CMV infection was associated with the differentiation of NK cells, with increased KIR⁺ and CD57⁺ expression on CD56^dim NK cells and CD56^dimKIR⁺CD57⁺ NK cells. Meanwhile, the percentage of CD56^dimKIR⁺CD57⁺ NK cells correlated with, but was higher than, the percentage of CD56^dimCD57⁺NKG2C⁺ NK cells, suggesting that the differentiation of NK cells played important roles in the decreased relapse rate post-MST. Moreover, we found that in the patients without relapse, the absolute NK cells and the CD56dimKIR⁺CD57⁺ NK subpopulations were significantly higher than those in the patients with relapse. This finding may support our hypothesis that CMV infection could promote NK-cell reconstitution and differentiation and increase CD56dimKIR⁺CD57⁺ NK subpopulations to prevent relapse in MST patients. However, in the patients diagnosed with ALL and other conditions, CMV infection did not influence the cumulative incidence of relapse. A possible reason is that the graft-versus-leukemia effects of NK cells are more prominent in AML than ALL or other conditions that are intrinsically resistant to NK recognition [47, 48].In future experiments, more research should be performed to demonstrate the enhanced antileukemia capacity of CD56^dimKIR⁺CD57⁺ NK cells.

TRM is another main cause of treatment failure after HSCT. In agreement with other previous studies, the survival benefit of the decreased relapse risk was outweighed by the increased TRM and decreased OS in the patients with CMV infection after MST [14, 17, 19, 20]. The primary causes of TRM include GVHD, bacterial/fungal infection, respiratory insufficiency and renal failure. Similarly, in our present study, only 1 patient died from CMV pneumonia, and most patients died from GVHD and other infections among the patients with CMV infection.

GVHD is a major cause of mortality and morbidity post allo-HSCT. More recent studies demonstrated that aGVHD is a risk factor for subsequent CMV infection, and the use of corticosteroids was correlated with the reactivation of CMV [49, 50]. Our previous study also showed that aGVHD combined with CMV infection was highly associated with poor CD8+ T-cell IR and inferior post-MST outcomes, and there was a strong correlation between CMV reactivation and aGVHD [51, 52]. In the present study, the occurrence of aGVHD in the patients with CMV infection was also significantly higher than that in the patients without CMV infection, and 2 patients died from aGVHD.

Therefore, the higher TRM related to CMV infection may be due to the following possible reasons. First, CMV treatments lead to immunosuppression, which increases the risk of opportunistic infections. Second, CMV infection increases the occurrence of aGVHD and delays immune reconstitution, which may further contribute to the increased rate of TRM.

We should acknowledge that this study has some limitations. (1) This was a retrospective study of CMV infection and clinical outcomes in our center. (2) A CMV infection and NK-cell relationship was demonstrated in a limited number of patients.

In conclusion, CMV infection was associated with a decreased relapse rate in patients with MST. The potential mechanism is that CMV infection promotes NK-cell reconstitution and differentiation to decrease the relapse rate in patients with MST. However, CMV infection was also associated with higher TRM and lower OS in this situation. Further prospective studies need to be conducted to clarify this relationship, and how to decrease the NRM in patents with CMV infection post-MST remains an important research priority.

Glossary

Abbreviations

aGVHD

acute graft-verse-host disease

Allo

allogeneic

CMV

cytomegalovirus

HSCT

hematopoietic stem cell transplantation

MST

HLA-matched sibling transplantation

NRM

non-relapse mortality

OS

overall survival

IR

immune reconstitution

NK cells

Natural killer cells.

Funding

This work was supported by the National Natural Science Foundation of China (nos. 81621001, 81870140, 82070184), Beijing Municipal Science and Technology Commission (no. Z211100002921071), Peking University People’s Hospital Research and Development Funds (no. RDX2019-36, RDL2021-01).

Conflict of interests

The authors declare no financial or commercial conflict of interest.

Author contributions

Z.X.Y. was responsible for conceiving and designing the study; X.J., Z.X.S., L.M., C.Y.J.., W.Y., X.L.P., Z.X.H., L.K.Y., and H.X.J. performed the clinical examination; Z.X.Y. and H.T.T. contributed to analyzing and interpreting the data and writing the manuscript; and all authors read and approved the final manuscript.

Ethical approval

This study was approved by the Human Ethical Systems Review Committee of Peking University People’s Hospital.

Patient consent

All patients provided written informed consent before their entry into the study in accordance with the Declaration of Helsinki.

Data availability

The data presented in this manuscript are included in the paper.

Permission to reproduce (for relevant content)

Not applied.