Low-dose ruxolitinib as first-line therapy for acute graft-versus-host disease (aGVHD) increases the risk of Epstein-Barr virus (EBV) reactivation but not posttransplant lymphoproliferative disease (PTLD)

Abstract

Background

Ruxolitinib (RUX) impairs antiviral immunity in patients with hematologic malignancies. For steroid-refractory acute graft-versus-host disease (aGVHD), studies have yielded conflicting results regarding whether RUX treatment increases the risk of Epstein–Barr virus (EBV) infection. However, in newly diagnosed aGVHD, the use of RUX raises concerns about the potential risk of EBV infection. This study aimed to assess the incidence of EBV reactivation and EBV-related posttransplant lymphoproliferative disease (EBV-PTLD) in patients receiving a regimen of 5 mg/day RUX combined with 1 mg·kg⁻¹·d⁻¹ methylprednisolone as first-line therapy for aGVHD.

Method

In this retrospective cohort study, allogeneic hematopoietic stem cell transplantation (allo-HSCT) recipients were consecutively screened at the Chinese PLA General Hospital. Patients who developed grade II–IV aGVHD were retrospectively categorized on the basis of first-line treatment. The RUX plus steroid (RUX-steroid) group (5 mg/day RUX + 1 mg·kg⁻¹·d⁻¹ methylprednisolone) included patients who received this dose in a phase 2 dose-escalation study of RUX, the intervention arm of a phase 3 trial evaluating this regimen as first-line therapy for aGVHD, and patients who were deemed unsuitable for high-dose steroids owing to active infections. The steroid regimen group (2 mg·kg⁻¹·d⁻¹ methylprednisolone) included patients treated in routine clinical practice and those in the control arm of the same phase 3 trial. EBV reactivation, EBV-PTLD incidence, and other transplantation outcomes were assessed. T-cell reconstitution was also analyzed at day 100 post-transplant.

Results

A total of 487 patients treated between January 2018 and December 2023 were included in this study, with a median follow-up of 26.6 months (range, 2.8–69.9). Among these patients, 105 received RUX (5 mg/day) plus methylprednisolone (1 mg·kg⁻¹·d⁻¹) as a first-line treatment for grade II–IV aGVHD and were categorized into the RUX-steroid group. Moreover, 115 patients who received methylprednisolone alone (2 mg·kg⁻¹·d⁻¹) formed the steroid group. The median duration of RUX exposure was 177 days (range, 17–447 days). The cumulative incidence of EBV reactivation within 180 days was significantly greater in the RUX-steroid group (82.9%) than that in the steroid group (70.2%) (hazard ratio [HR]: 1.43, 95% CI: 1.03–1.97; P = 0.03). The cumulative incidence of PTLD was comparable between the RUX-steroid group (5.0%) and the steroid group (4.5%) (HR: 1.12, 95% CI: 0.33–3.83; P = 0.86). Multivariate analysis revealed that the use of a matched-sibling donor (MSD) was significantly associated with a decreased risk of both EBV reactivation (p < 0.001) and EBV-PTLD (P = 0.034). The 2-year nonrelapse mortality (NRM) was 14.9% in the RUX-steroid group, comparable to 13.7% in the steroid group (HR: 1.01, 95% CI: 0.50–2.03; P = 0.98).

Conclusion

RUX was associated with an increased risk of EBV reactivation in patients after allo-HSCT but not with PTLD. Regular monitoring of the EBV viral load to enable timely intervention remains essential in these patients.

Graphical abstract

Highlights

• •Low-dose ruxolitinib (RUX) added to first-line acute graft-versus-host disease (aGVHD) therapy tended to increase Epstein–Barr virus (EBV) reactivation.
• •Compared with standard-dose steroid therapy, low-dose RUX added to first-line aGVHD therapy resulted in a similar incidence of EBV-related posttransplant lymphoproliferative disease (EBV-PTLD).
• •Regular EBV monitoring with timely intervention is essential in patients receiving RUX after allogeneic hematopoietic stem cell transplantation (allo-HSCT).

Introduction

Allogeneic hematopoietic stem cell transplantation (allo-HSCT) is an established approach that provides the only potentially curative option for various malignant and nonmalignant hematologic diseases.^1,2 Infection with Epstein–Barr virus (EBV) is a common complication in allo-HSCT recipients and may cause life-threatening posttransplantation lymphoproliferative disorder (PTLD).³ At a dosage of 10 mg/day, RUX exhibits significant efficacy in treating steroid-refractory acute graft-versus-host disease (aGVHD) — a major complication of allo-HSCT — but may increase the risk of EBV infection.4, 5, 6

Ruxolitinib (RUX) impairs dendritic cell and T-cell functions, leading to reduced cytokine production that results in impaired control of silent infections and an increased risk of viral reactivation, such as herpes zoster and hepatitis B.7, 8, 9 A growing body of clinical evidence supports this mechanistic hypothesis for EBV reactivation. Rumi et al. reported a significant reduction in T lymphocytes and an inverse correlation between the EBV-specific T-cell response and the RUX dose.¹⁰ Several cases of EBV-related lymphoproliferative disorder have been documented in patients with myelofibrosis receiving RUX.¹¹ Most notably, in a cohort of 137 patients with GVHD, RUX use was identified as a significant risk factor for EBV reactivation.⁶ These studies suggest that RUX may interfere with immunosurveillance against EBV, emphasizing the importance of assessing the risk of EBV infection in posttransplant patients receiving ruxolitinib.

When RUX was incorporated into the first-line treatment for aGVHD, we considered that hematopoietic reconstitution is more fragile in the early postengraftment stage, making patients more susceptible to the hematologic toxicity of the drug. On the basis of these findings, a phase 2 dose-finding study was conducted, and the results suggested that RUX at 5 mg/day, rather than 5 mg or 10 mg twice daily, is both effective and safe when combined with corticosteroids.¹² When a lower dose was administered, concerns about the risk of EBV infection and EBV-PTLD remained. In a cohort of 200 patients who underwent allo-HSCT at our center between March 2014 and December 2017, Gao et al. reported a cumulative incidence of nearly 60% for EBV infection in patients with aGVHD.¹³ In contrast, our previous study combining 5 mg/day RUX with 1 mg·kg⁻¹·d⁻¹ methylprednisolone as initial treatment for aGVHD revealed that the cumulative incidence of EBV infection was 87.5%, which seemed much higher than that in patients receiving glucocorticoids alone.¹²

Here, we conducted a retrospective cohort study involving consecutive allo-HSCT recipients at our center from 2018 to 2023 and investigated the cumulative incidence and risk factors for EBV infection and EBV-PTLD, as well as other transplantation outcomes and T-cell reconstitution. This study focused on grade II–IV aGVHD patients who were initially treated with a combination of RUX (5 mg/day) and methylprednisolone (1 mg·kg⁻¹·d⁻¹).

Methods

Study design and participants

In this retrospective cohort study, patients who underwent allo-HSCT was consecutively screened at the Chinese PLA General Hospital between January 2018 and December 2023. The only inclusion criteria was having malignant or nonmalignant hematologic diseases and undergoing allo-HSCT. The exclusion criterion was a second transplantation or engraftment failure. Among patients who developed grade II–IV aGVHD, treatment was retrospectively categorized on the basis of the first-line regimen received. The RUX plus steroid (RUX-steroid) group (5 mg/day RUX + 1 mg·kg⁻¹·d⁻¹ methylprednisolone) included (1) individuals enrolled at our center in a phase 2 dose-escalation study of the above combination regimen, of whom only those receiving 5 mg/day RUX were included¹²; (2) participants in the intervention arm of a multicenter randomized controlled trial (RCT) evaluating this combination as first-line therapy for aGVHD, limited to those recruited at our center¹⁴; and (3) non-trial patients who received the regimen due to active infections that were deemed unfit for standard-dose corticosteroids (2 mg·kg⁻¹·d⁻¹). The steroid group (2 mg·kg⁻¹·d⁻¹ methylprednisolone) included patients from the control arm of the same RCT and those who received standard steroid therapy as routine care outside the trial period. The screening period of this retrospective study overlapped with the enrollment periods of the above two clinical trials, ensuring complete coverage of eligible trial participants.

Patients were retrospectively categorized into four groups: non-aGVHD, grade I aGVHD, and RUX-steroid or steroid groups among those with grade II–IV aGVHD. The cumulative incidence of EBV reactivation and EBV-PTLD within 180 days was analyzed across all the groups, along with other outcomes, namely, cytomegalovirus (CMV) reactivation, hemorrhagic cystitis, nonrelapse mortality (NRM), disease relapse, overall survival (OS), disease-free survival (DFS), and moderate to severe chronic graft-versus-host disease (cGVHD).

This study was approved by the Ethics Committee of Chinese PLA General Hospital (Approval No. S2019-177-01) and was conducted in accordance with the principles of the Declaration of Helsinki. Written informed consent was obtained from all patients before transplantation.

Procedures

All patients received myeloablative conditioning regimens, along with GVHD prophylaxis consisting of cyclosporine A, mycophenolate mofetil, and short-term methotrexate.¹⁵ Antithymocyte globulin (ATG; thymoglobulin, rabbit) was administered to all patients. In patients who underwent haploidentical donor stem cell transplantation (HID-SCT) and unrelated donor stem cell transplantation (URD-SCT), ATG was administered at a total dosage of 10 mg/kg (1.5 mg/kg on day −5, 2.5 mg/kg on day −4 to −3, and 3.5 mg/kg on day −2). In patients who underwent matched-sibling donor stem cell transplantation (MSD-SCT), a total dose of 5 mg/kg was administered (1.5 mg/kg on day −5 and 3.5 mg/kg on day −4). Patients who received granulocyte colony-stimulating factor (G-CSF) mobilized unmanipulated peripheral blood stem cells on the day of graft collection from an HLA-haploidentical related donor, an HLA-matched sibling donor, or an HLA-matched unrelated donor for allogeneic transplantation. The target mononuclear cell (MNC) count was 5 × 10⁸/kg, and the CD34+ cell count was 2 × 10⁶/kg per recipient body weight.

Lymphocytes and T cells were measured every other week for the first 3 months, monthly up to 6 months, and every 3 months until 1 year after HSCT. The absolute number of T cells was measured using TruCOUNT tubes (BD Biosciences) in EDTA-treated whole blood after neutrophil engraftment.

aGVHD was graded on the basis of the modified Glucksberg criteria.^16,17 Patients received 2 mg·kg⁻¹·d⁻¹ methylprednisolone (steroid group) or 5 mg/day RUX combined with 1 mg·kg⁻¹·d⁻¹ methylprednisolone (RUX-steroid group) as the first-line treatment for grade II–IV aGVHD. Once aGVHD resolved, RUX was subsequently tapered and discontinued over the course of 3 months. If the above first-line treatment failed, a second-line treatment of basiliximab, RUX, sirolimus, or mesenchymal stromal cells could be considered.

Monitoring, definitions, and treatment of EBV infection

EBV-DNA in plasma was monitored by real-time quantitative PCR, with testing conducted twice weekly from neutrophil engraftment to +1 month, once weekly from +2 to +3 months, and once every two weeks from +4 to +6 months posttransplant. The frequency of testing was increased to twice weekly upon the appearance of symptoms or signs of PTLD or when a previously negative result turned positive. The reporting threshold for EBV-DNA positivity in plasma at our center is 1000 copies/mL.

EBV reactivation was defined as a viral load of ≥10,000 copies/mL in a single plasma sample or ≥1000 copies/mL in two consecutive samples. This threshold prompted the initiation of preemptive therapy with rituximab. Rituximab was administered at a dosage of 375 mg/m² once weekly and continued until EBV-DNA became undetectable in two consecutive tests. The diagnosis of EBV-PTLD was derived from the Sixth European Conference on Infections in Leukemia (ECIL-6) guidelines.¹⁸ Probable PTLD was defined as lymphadenopathy, hepatosplenomegaly, or other end-organ manifestations accompanied by high EBV-DNA level and the absence of other documented causes. Proven PTLD was defined as meeting histological criteria for EBV-PTLD in tissue specimens accompanied by symptoms and/or signs from the affected organ. Treatment for EBV-PTLD includes rituximab, EBV-specific cytotoxic lymphocyte infusion, or chemotherapeutic regimens. Considering the risk of aGVHD, the dosage of immunosuppressive agents should generally not be reduced within 60 days after MSD-SCT or 90 days after HID/URD-SCT. CMV reactivation was defined as a CMV-DNA level ≥10000 copies/mL in a single test or ≥1000 copies/mL in two consecutive tests in plasma. None of the recipients received CMV prophylaxis with letermovir during the study period.

Statistical analyses

Statistical analyses were performed using the R statistical software ‘EZR’¹⁹ and SPSS 26.0 software (IBM Corporation, Armonk, NY, USA). A P value < 0.05 was chosen as the threshold for significance.

Comparisons between groups for quantitative data were performed using the independent samples t test or the Wilcoxon rank sum test. The chi-square test or two-sided Fisher's exact test was selected for categorical data between the two groups according to the total sample size and minimum theoretical frequency.

The cumulative incidences of EBV reactivation, EBV-PTLD, CMV reactivation, hemorrhagic cystitis, NRM, relapse, and moderate to severe cGVHD were estimated using Gray's method, considering competing events.²⁰ Death without the event of interest was considered a competing event for EBV reactivation, EBV-PTLD, CMV reactivation, and hemorrhagic cystitis. With respect to relapse, NRM was considered a competing event. For NRM, relapse was considered a competing event. Relapse and death without moderate to severe cGVHD were considered competing events for moderate to severe cGVHD. For group analyses according to aGVHD status (non-aGVHD, grade I aGVHD, grade II–IV aGVHD treated with steroids only, and grade II–IV aGVHD treated with steroids plus RUX), patients who experienced EBV reactivation, EBV-PTLD, CMV reactivation, or hemorrhagic cystitis before the onset of aGVHD were excluded from the corresponding analyses.

In the univariable and multivariable analyses, EBV reactivation, EBV-PTLD, CMV reactivation, hemorrhagic cystitis, NRM, relapse, and moderate to severe cGVHD were assessed using the Fine–Gray proportional hazards regression model with time-dependent covariates, whereas OS and DFS were evaluated using the Cox proportional hazards regression model with time-dependent covariates. The following variables were included in the univariable analysis: recipient age, recipient sex, diagnosis, status at HSCT, HCT-CI score, conditioning regimen, donor type, MNC count, CD34+ cell count, donor age, ABO match, and sex match. The occurrence of aGVHD, grade III–IV aGVHD, first-line treatment with RUX for aGVHD, CMV reactivation, EBV reactivation, and EBV-PTLD reactivation were also included in the univariate analysis as time-dependent covariates. Only variables with p values < 0.1 in the univariable analysis were included in the multivariable analysis.

Propensity score matching (PSM) for selected variables (recipient age, disease, type of donor and ATG dosage) was performed using logistic regression with the nearest neighbor method. The caliper settings were set to less than 0.05 to restrict the distance between matched units. A standardized mean difference (SMD) threshold of less than 0.1 was determined as an indicator of reduced bias between treatment cohorts.

Results

Patient characteristics

In this retrospective study, 505 patients were consecutively screened at the Chinese PLA General Hospital from January 2018 to December 2023. Eighteen patients were excluded—sixteen who underwent a second transplantation and two with primary engraftment failure. A total of 487 patients were included in the final analysis, with follow-up continuing until September 2024 (median, 26.6 months; range, 2.8–69.9 months). Among them, 220 patients did not develop aGVHD, 47 developed grade I aGVHD, 115 patients with grade II–IV aGVHD received 2 mg·kg⁻¹·d⁻¹ methylprednisolone (steroid group), and 105 with grade II–IV aGVHD received an initial treatment of 5 mg/day RUX combined with 1 mg·kg⁻¹·d⁻¹ methylprednisolone (RUX-steroid group). The flowchart of this study is shown in Figure 1. The RUX-steroid group included 5 patients from our previously reported single-arm dose-finding trial, 88 patients from a recent phase 3 randomized control trial (RCT), and 12 patients who could not tolerate 2 mg·kg⁻¹·d⁻¹ methylprednisolone due to active infections at aGVHD onset.^12,14 The steroid group consisted of 88 patients enrolled in the control arm of the same phase 3 randomized controlled trial and an additional 27 patients who received standard 2 mg·kg⁻¹·d⁻¹ methylprednisolone as part of routine clinical care outside the trial recruitment period.

Figure 1.

Patient selection and grouping flowchart based on first-line therapy for grade II–IV aGVHD.

Steroid group: patients who received methylprednisolone alone (2 mg·kg⁻¹·d⁻¹) as the first-line therapy for grade II–IV aGVHD; RUX-steroid group: patients who received ruxolitinib (5 mg/day) plus methylprednisolone (1 mg·kg⁻¹·d⁻¹) as the first-line therapy for grade II–IV aGVHD. aGVHD: Acute graft-versus-host disease; RUX-steroid: Ruxolitinib plus steroid therapy.

The baseline demographic and transplantation characteristics of the total cohort are shown in Table 1, and no significant differences in these baseline characteristics were noted between the RUX-steroid group and the steroid group. A total of 99.0% of the patients were serologically EBV positive, and their donors were also serologically EBV positive. No patients were serologically EBV-negative with EBV-positive donors.

Table 1.

Baseline characteristics of patients in the non-aGVHD group, grade I aGVHD group, steroid group, and RUX-steroid group.

Variable	Total	non-aGVHD	aGVHD I	aGVHD II–IV
				Steroid group	RUX-steroid group	P value	χ2/U value
No. of patients	487	220	47	115	105
Age of recipients (years)	37 (6–73)	42 (6–73)	33 (9–60)	34 (12–60)	35 (10–67)	0.177	5400.500
Gender
Male	329 (67.6)	138 (62.7)	31 (66.0)	85 (73.9)	75 (71.4)	0.679	0.171
Disease
AML	252 (51.7)	114 (51.8)	24 (51.1)	56 (48.7)	58 (55.2)	0.380	8.413
MDS	50 (10.3)	28 (12.7)	5 (10.6)	6 (5.2)	11 (10.5)
CML	5 (1.0)	2 (0.9)	1 (2.1)	0 (0)	2 (1.9)
ALL/LBL	133 (27.3)	54 (24.5)	12 (25.5)	41 (35.7)	26 (24.8)
NHL	14 (2.9)	7 (3.2)	3 (6.4)	2 (1.7)	2 (1.9)
MDS/MPN	8 (1.6)	3 (1.4)	1 (2.1)	2 (1.7)	2 (1.9)
MPAL	9 (1.8)	4 (1.8)	0 (0)	3 (2.6)	2 (1.9)
AA	14 (2.9)	8 (3.6)	1 (2.1)	4 (3.5)	1 (1.0)
MM	2 (0.4)	0 (0)	0 (0)	1 (0.9)	1 (1.0)
Status at HSCT
CR	379 (77.8)	167 (75.9)	38 (80.9)	95 (82.6)	79 (75.2)	0.199	4.632
Non-CR	52 (10.7)	22 (10.0)	4 (8.5)	10 (8.7)	16 (15.2)
Untreated	42 (8.6)	23 (10.5)	4 (8.5)	6 (5.2)	9 (8.6)
Nonmalignancy	14 (2.9)	8 (3.6)	1 (2.1)	4 (3.5)	1 (1)
Pretreatment with ATG
5 mg/kg ATG	103 (21.1)	72 (32.7)	7 (14.9)	11 (9.6)	13 (12.4)	0.503	0.448
10 mg/kg ATG	384 (78.9)	148 (67.3)	40 (85.1)	104 (90.4)	92 (87.6)
Conditioning regimen
Bu based	463 (95.1)	208 (94.5)	45 (95.7)	108 (93.9)	102 (97.1)	0.338	1.320
TBI based	24 (4.9)	12 (5.5)	2 (4.3)	7 (6.1)	3 (2.9)
Type of donor
Haplo	330 (67.8)	116 (52.7)	34 (72.3)	96 (83.5)	84 (80)	0.774	0.513
Sibling	103 (21.1)	72 (32.7)	7 (14.9)	11 (9.6)	13 (12.4)
Unrelated	54 (11.1)	32 (14.5)	6 (12.8)	8 (7.0)	8 (7.6)
Age of donor (years)	35 (8–65)	33 (8–65)	30 (11–54)	35 (10–58)	37 (9–60)	0.063	5160.5
ABO match
Match	251 (51.5)	105 (47.7)	26 (55.3)	58 (50.4)	62 (59.0)	0.237	4.233
Major mismatch	99 (20.3)	50 (22.7)	8 (17.0)	20 (17.4)	21 (20.0)
Minor mismatch	95 (19.5)	44 (20.0)	7 (14.9)	29 (25.2)	15 (14.3)
Bidirectional mismatch	42 (8.6)	21 (9.5)	6 (12.8)	8 (7.0)	7 (6.7)
Gender match
FF	56 (11.5)	25 (11.4)	7 (14.9)	15 (13.0)	9 (8.6)	0.051	7.756
FM	114 (23.4)	45 (20.5)	7 (14.9)	26 (22.6)	36 (34.3)
MF	102 (20.9)	57 (25.9)	9 (19.1)	15 (13.0)	21 (20.0)
MM	215 (44.1)	93 (42.3)	24 (51.1)	59 (51.3)	39 (37.1)
Graft
MNC ( × 108/kg)	10.6 (2.8–33.5)	10.6 (3.7–33.5)	10.1 (4.3–24.5)	10.1 (4.7–22.6)	11.2 (2.8–28.3)	0.064	5087
CD34+ ( × 106/kg)	4.8 (0.5–22.4)	4.9 (1.7–15.8)	5.2 (2.0–22.4)	4.8 (0.5–14.3)	4.6 (1.4–14.8)	0.968	6018.500
EBV serostatus
D-/R-	0 (0)	0 (0)	0 (0)	0 (0)	0 (0)	0.227	2.211
D-/R+	5 (1.0)	3 (1.4)	0 (0)	0 (0)	2 (1.9)
D+/R-	0 (0)	0 (0)	0 (0)	0 (0)	0 (0)
D+/R+	482 (99.0)	217 (98.6)	47 (100.0)	115 (100.0)	103 (98.1)
CMV serostatus
D-/R-	0 (0)	0 (0)	0 (0)	0 (0)	0 (0)	0.887	0.633
D-/R+	15 (3.1)	8 (3.6)	0 (0)	3 (2.6)	4 (3.8)
D+/R-	5 (1.0)	1 (0.5)	1 (2.1)	2 (1.7)	1 (1.0)
D+/R+	467 (95.9)	211 (95.9)	46 (97.9)	110 (95.7)	100 (95.2)
aGVHD grade
Non-aGVHD	220 (45.2)	220 (100.0)	0 (0)	0 (0)	0 (0)	0.900	0.210
Grade I	47 (9.7)	0 (0)	47 (100.0)	0 (0)	0 (0)
Grade II	176 (36.1)	0 (0)	0 (0)	93 (80.9)	83 (79.0)
Grade III	27 (5.5)	0 (0)	0 (0)	14 (12.2)	13 (12.4)
Grade IV	17 (3.5)	0 (0)	0 (0)	8 (7.0)	9 (8.6)

Categorical variables are presented as numbers (percentages), and continuous variables are presented as medians (ranges). Differences between categorical variables were assessed using the χ² test, and differences between continuous variables were assessed using the Mann-Whitney U test. Serostatus was categorized as follows: D^- (donor seronegative), D⁺ (donor seropositive), R^- (recipient seronegative), and R⁺ (recipient seropositive). Gender matching was categorized as follows: FF (female donor to female recipient), FM (female donor to male recipient), MF (male donor to female recipient), and MM (male donor to male recipient).AA: Aplastic anemia; aGVHD: Acute graft-versus-host disease; ALL/LBL: Acute lymphoblastic leukemia/lymphoblastic lymphoma; AML: Acute myeloid leukemia; ATG: Anti-thymocyte globulin; BU: Busulfan; CML: Chronic myeloid leukemia; CR: Complete remission; EBV: Epstein–Barr virus; Haplo: Haploidentical; MDS: Myelodysplastic syndrome; MDS/MPN: Myelodysplastic/myeloproliferative neoplasm; MM: Multiple myeloma; MNC: Mononuclear cell; MPAL: Mixed-phenotype acute leukemia; NHL: Non-Hodgkin lymphoma; RUX-steroid: Ruxolitinib plus steroid therapy; TBI: Total body irradiation.

EBV reactivation and PTLD

For the total cohort, the cumulative incidence of EBV reactivation within 180 days was 68% (95% CI: 63.7%–72.0%), with a median onset on day 49 posttransplantation (95% CI: 44–55). The cumulative incidence of EBV reactivation within 180 days in the RUX-steroid group was 82.9% (95% CI: 73.3%–89.3%), which was significantly higher than that in the steroid group (70.2%; 95% CI: 59.7%–78.5%) (HR: 1.43, 95% CI: 1.03–1.97; P = 0.03) [Figure 2A]. The cumulative incidence of EBV reactivation in patients without aGVHD was 60.6% (95% CI: 53.7%–66.7%), and that in patients with grade I aGVHD was 56.8% (95% CI: 39.0%–71.1%). Multivariate analysis revealed that MSD-SCT was the only factor significantly associated with a decreased risk of EBV reactivation by day 180 [Table 2]. Although the RUX-steroid regimen and longer duration of RUX exposure were associated with a higher cumulative incidence of EBV reactivation in the univariable analysis (RUX-steroid: HR: 1.72, 95% CI: 1.29–2.30, p < 0.001; duration: HR: 1.01, 95% CI: 1.00–1.02, p < 0.001), these associations were not statistically significant in the multivariable analysis.

Figure 2.

Cumulative incidence of EBV reactivation and EBV-PTLD in the non-aGVHD group, grade I aGVHD group, steroid group, and RUX-steroid group.(A) Cumulative incidence of EBV reactivation in the non-aGVHD group, the grade I aGVHD group, the steroid group, and the RUX-steroid group; (B) cumulative incidence of EBV-PTLD in the non-aGVHD group, the grade I aGVHD group, the steroid group, and the RUX-steroid group. Non-aGVHD group: Patients without aGVHD occurrence; grade I aGVHD group: Patients with grade I aGVHD; RUX-steroid group: Patients receiving ruxolitinib (5 mg/day) plus methylprednisolone (1 mg·kg⁻¹·d⁻¹) as the first-line therapy for grade II–IV aGVHD; and steroid group: patients receiving methylprednisolone alone (2 mg·kg⁻¹·d⁻¹) as the first-line therapy for grade II–IV aGVHD.aGVHD: Acute graft-versus-host disease; CI: Confidence interval; EBV: Epstein–Barr virus; PTLD: posttransplant lymphoproliferative disease; RUX-steroid:Ruxolitinib plus steroid therapy.

Table 2.

Risk factors for EBV reactivation and EBV-PTLD: univariate and multivariate analyses in the total cohort.

Variable	EBV reactivation				EBV-PTLD
	Univariate		Multivariate		Univariate		Multivariate
	HR (95% CI)	P value	HR (95% CI)	P value	HR (95% CI)	P value	HR (95% CI)	P value
Age (years)	0.99 (0.98–1.00)	0.033			0.99 (0.97–1.01)	0.450
Gender
Male	1				1
Female	0.81 (0.64–1.02)	0.071			0.68 (0.33–1.39)	0.290
Disease
Myeloid malignancies	1				1
Lymphoid malignancies	1.26 (1.01–1.57)	0.044			0.81 (0.40–1.64)	0.560
Aplastic anemia	1.33 (0.66–2.69)	0.420			1.80 (0.41–7.96)	0.440
Status at HSCT
Non-CR	1.02 (0.70–1.48)	0.930			1.19 (0.47–3.02)	0.710
Others	1				1
HCT-CI
<2 scores	1				1
≥2 scores	0.98 (0.76–1.25)	0.860			0.85 (0.39–1.85)	0.690
Conditioning
Bu based	1				1
TBI based	1.09 (0.72–1.66)	0.670			2.22 (0.80–6.12)	0.120
Type of donor
Sibling	1		1		1		1
Haplo	2.12 (1.58–2.84)	<0.001	2.11 (1.56–2.86)	<0.001	3.03 (0.93–9.83)	0.065	2.74 (1.10–9.01)	0.038
Unrelated	1.75 (1.14–2.68)	0.010	1.75 (1.12–2.73)	0.014	6.20 (1.70–22.57)	0.006	5.49 (1.46–20.65)	0.012
Age of donor (years)	1.01 (1.00–1.02)	0.005			1.02 (1.00–1.05)	0.044
ABO match
Match	1				1
Major mismatch	1.05 (0.80–1.38)	0.730			1.45 (0.67–3.14)	0.350
Minor mismatch	1.06 (0.80–1.40)	0.700			1.19 (0.52–2.74)	0.680
Bidirectional mismatch	0.96 (0.65–1.42)	0.850			1.32 (0.46–3.79)	0.610
Gender match
FF	1				1
FM	1.49 (0.96–2.30)	0.072			1.17 (0.30–4.50)	0.820
MF	1.69 (1.10–2.60)	0.018			1.29 (0.34–4.96)	0.710
MM	1.91 (1.28–2.84)	0.002			2.05 (0.62–6.78)	0.240
MNC ( × 108/kg)	0.99 (0.97–1.01)	0.500			0.91 (0.84–0.99)	0.023
CD34+ ( × 106/kg)	1.01 (0.97–1.05)	0.580			0.94 (0.79–1.11)	0.470
Duration of ruxolitinib exposure (days)	1.01 (1.00–1.02)	<0.001			1.02 (1.01–1.03)	<0.001
aGVHD
No	1				1
Yes	1.69 (1.35–2.13)	<0.001			0.68 (0.35–1.31)	0.244
aGVHD grade III–IV
No	1				1
Yes	0.76 (0.45–1.30)	0.319			0.73 (0.17–3.09)	0.670
Ruxolitinib in the first-line treatment
No	1				1
Yes	1.72 (1.29–2.30)	<0.001			0.64 (0.25–1.66)	0.358
CMV reactivation
No	1				1
Yes	2.03 (1.50–2.74)	<0.001			1.53 (0.75–3.10)	0.242
EBV reactivation
No					1
Yes	NA	NA			14.31 (5.66–36.18)	<0.001

Gender matching was categorized as follows: FF (female donor to female recipient), FM (female donor to male recipient), MF (male donor to female recipient), and MM (male donor to male recipient). aGVHD: Acute graft-versus-host disease; BU: Busulfan; CI: confidence interval; CMV: Cytomegalovirus; CR: Complete remission; EBV: Epstein–Barr virus; EBV-PTLD: EBV-related posttransplant lymphoproliferative disease; Haplo: Haploidentical; HCT-CI: Hematopoietic cell transplantation-specific comorbidity index; HR: Hazard ratio; MNC: Mononuclear cell; TBI: Total body irradiation.

Specifically, among the five patients with disparate EBV serostatus (donor EBV-negative and recipient EBV-positive), three did not develop aGVHD, of whom two experienced EBV reactivation. The remaining two patients were in the RUX-steroid group, and one of them developed EBV reactivation. Importantly, none of the above five patients developed EBV-PTLD. For patients with different underlying diseases, MSD-SCT remained significantly associated with a decreased risk according to the multivariate model. After further stratification by donor type, multivariate analysis revealed that aGVHD occurrence was an independent risk factor for EBV reactivation in recipients of HID-SCT or URD-SCT. In the MSD-SCT group, however, an HCT-CI score ≥2 was an independent risk factor [Table S1].

By day 180, 40 patients had developed EBV-PTLD, and the cumulative incidence for the total cohort was 8.2% (95% CI: 6.0%–10.9%). All patients who developed PTLD had prior EBV reactivation and received preemptive rituximab. The PTLD characteristics are presented in Table 3. In total, 15 patients were diagnosed with histologically proven PTLD, and 25 patients were diagnosed clinically. The cumulative incidence of PTLD within 180 days in the RUX-steroid group was 5.0% (95% CI: 1.8%–10.4%), comparable to that in the steroid group at 4.5% (95% CI: 1.7%–9.5%) (HR: 1.12, 95% CI: 0.33–3.83; P = 0.86) [Figure 2B]. The cumulative incidence of PTLD in patients without aGVHD was 7.7% (95% CI: 4.7%–11.8%), and that in patients with grade I aGVHD was 8.9% (95% CI: 2.8%–19.5%). Multivariate analysis revealed that MSD-SCT was significantly associated with a decreased risk of PTLD [Table 2]. A longer duration of RUX exposure was associated with a higher cumulative incidence of EBV-PTLD in the univariable analysis (HR: 1.02, 95% CI: 1.01–1.03; p < 0.001); however, this association was not statistically significant in the multivariable analysis. All patients with EBV reactivation received preemptive rituximab, and in this subgroup, multivariate analysis still revealed that MSD-SCT was significantly associated with a decreased risk of PTLD [Table S2].

Table 3.

Characteristics of patients with EBV-PTLD.

Characteristic	Number
Extra lymphatic involvement
Yes	5
No	35
Ann Arbor lymphoma stage
I	9
II	16
III	9
IV	0
Unknown	6
Histologically proven PTLD	15
CD20 expression
Positive	11
Negative	3
Unknown	1
EBER expression
Positive	13
Negative	2
Clonality
Monoclonal	9
Polyclonal	2
Unknown	4
Morphology
Monomorphic	4
Polymorphic	8
Unknown	3
Nonhistologically proven PTLD	25
EBV reactivation with
Lymphadenopathy	18
Hepatosplenomegaly	10
Fever	20

EBER: Epstein–Barr encoding region; EBV: Epstein–Barr virus; EBV-PTLD: EBV-related posttransplant lymphoproliferative disease; PTLD: Posttransplant lymphoproliferative disease.

Additionally, propensity score matching was performed as a sensitivity analysis between the steroid group and the RUX-steroid group. Of the 220 eligible patients, 34 (15%) were excluded because of unmet matching criteria, resulting in 186 patients (93 per group) being included in the final analysis. The median age was 34.5 years (range, 10–67 years), with 133 (72%) males and 53 (28%) females. Multivariate analysis revealed that HID-SCT and aGVHD occurrence were significantly associated with an increased risk of EBV reactivation. With respect to EBV-PTLD, MSD-SCT remained significantly associated with a decreased risk according to the multivariate model [Table S3].

T-cell reconstitution

Three patients (one in the steroid group and two in the RUX-steroid group) whose aGVHD onset and initiation of first-line treatment occurred after day 100 posttransplant were excluded from the T-cell reconstitution analysis. Compared with patients without aGVHD or with grade I aGVHD, those with grade II–IV aGVHD had significantly lower absolute lymphocyte (p < 0.001), total T-cell (p < 0.001), CD4⁺ T-cell (p < 0.001), and CD8⁺ T-cell (p < 0.001) counts on day 100 [Figure 3A–D]. The absolute count of CD4⁺ T cells in the RUX-steroid group (median 90 cells/μL, IQR: 40–159 cells/μL) was comparable to that in the steroid group (median 95 cells/μL, IQR: 41–178 cells/μL) on day 100 (p = 0.743) [Figure 3C]. The absolute counts of total T cells and CD8⁺ T cells were also comparable between the RUX-steroid and steroid groups on day 100 [Figure 3B and D].

Figure 3.

T-cell reconstitution on day 100 in the non-aGVHD group, grade I aGVHD group, steroid group, and RUX-steroid group.(A) Absolute lymphocyte counts on day 100 in the non-aGVHD group, the grade I aGVHD group, the steroid group, and the RUX-steroid group; (B) absolute T-cell counts on day 100 in the four groups; (C) absolute CD4⁺ T-cell counts on day 100 in the four groups; (D) absolute CD8⁺ T-cell counts on day 100 in the four groups. Non-aGVHD group: Patients without aGVHD occurrence; grade I aGVHD group: Patients with grade I aGVHD; RUX-steroid group: Patients receiving ruxolitinib (5 mg/day) plus methylprednisolone (1 mg·kg⁻¹·d⁻¹) as the first-line therapy for grade II–IV aGVHD; and steroid group: patients receiving methylprednisolone alone (2 mg·kg⁻¹·d⁻¹) as the first-line therapy for grade II–IV aGVHD. aGVHD: acute graft-versus-host disease; RUX-steroid: Ruxolitinib plus methylprednisolone.

Transplant outcomes

In the total cohort, no CMV-seronegative donor/recipient pairs (D^-/R^-) were observed. Twenty patients had CMV serostatus mismatches (D^-/R⁺ or D⁺/R^-), with a cumulative incidence of CMV reactivation within 180 days of 80.0% (95% CI: 4.7%–98.4%), whereas 467 patients were D+/R+, with a cumulative incidence of 43.2% (95% CI: 38.7%–47.5%) (HR: 2.12, 95% CI: 1.03–4.36; p = 0.041). The cumulative incidence of CMV reactivation within 180 days was comparable between the RUX-steroid group and the steroid group (HR: 1.18, 95% CI: 0.81–1.71; p = 0.39) [Figure 4A]. Univariate analysis revealed that the administration of RUX was associated with increased CMV reactivation (HR: 1.87, 95% CI: 1.40–2.49; p < 0.001). Multivariate analysis suggested that URD-SCT, HID-SCT and pretransplant diagnosis of lymphoid malignancies were significantly correlated with increased CMV reactivation [Table S4]. The cumulative incidence of hemorrhagic cystitis within 180 days was slightly greater in the RUX-steroid group (33.7%, 95% CI: 24.5%–43.1%) than in the steroid group (29.8%, 95% CI: 21.3%–38.8%), but the difference was not statistically significant (HR: 1.18, 95% CI: 0.73–1.92; p = 0.51) [Figure 4B]. Univariate analysis revealed that the administration of RUX was also associated with increased hemorrhagic cystitis (HR: 2.06, 95% CI: 1.39–3.07; p < 0.001). Multivariate analysis revealed that busulfan-based conditioning and HID-SCT were significantly associated with increased hemorrhagic cystitis [Table S4].

Figure 4.

Transplantation outcomes in the non-aGVHD group, grade I aGVHD group, steroid group, and RUX-steroid group. (A) Cumulative incidence of CMV reactivation in the non-aGVHD group, the grade I aGVHD group, the steroid group, and the RUX-steroid group; (B) cumulative incidence of hemorrhagic cystitis in the four groups; (C) cumulative incidence of nonrelapse mortality in the four groups; (D) cumulative incidence of relapse in the four groups; (E) overall survival in the four groups; (F) disease-free survival in the four groups; (G) cumulative incidence of moderate to severe cGVHD in the four groups. Non-aGVHD group: Patients without aGVHD occurrence; grade I aGVHD group: Patients with grade I aGVHD; RUX-steroid group: Patients receiving ruxolitinib (5 mg/day) plus methylprednisolone (1 mg·kg⁻¹·d⁻¹) as the first-line therapy for grade II–IV aGVHD; and steroid group: patients receiving methylprednisolone alone (2 mg·kg⁻¹·d⁻¹) as the first-line therapy for grade II–IV aGVHD. aGVHD: Acute graft-versus-host disease; cGVHD: Chronic graft-versus-host disease; CI Confidence interval; CMV: Cytomegalovirus; DFS: Disease-free survival; NRM: Nonrelapse mortality; OS: Overall survival; RUX-steroid: Ruxolitinib plus steroid therapy.

The 2-year NRM in the total cohort was 12.4% (95% CI: 9.6%–15.6%). The 2-year NRM was 14.9% (95% CI: 8.7%–22.6%) in the RUX-steroid group, which was similar to the 13.7% (95% CI: 8.0%–20.9%) observed in the steroid group (HR: 1.01, 95% CI: 0.50–2.03; p = 0.98) [Figure 4C]. The 2-year NRM was 11.0% (95% CI: 7.2%–15.7%) in patients without aGVHD and 9.0% (95% CI: 2.8%–19.7%) in patients with grade I aGVHD. Univariate analysis revealed that the administration of RUX was not associated with increased NRM (HR 1.28, 95% CI: 0.71–2.29; p = 0.408). Multivariate analysis revealed that MSD-SCT was a significant predictor of a decreased incidence of NRM. Additionally, EBV reactivation and PTLD were not identified as independent risk factors in the multivariate model [Table S4]. The 2-year cumulative incidence of relapse was 24.7% (95% CI: 20.7%–28.8%) in the total cohort. There was no significant difference in the cumulative incidence of relapse between the RUX-steroid group and the steroid group (HR 1.05, 95% CI: 0.60–1.82; p = 0.87) [Figure 4D]. The administration of RUX was not correlated with an increased cumulative incidence of relapse according to univariate analysis (HR: 0.86, 95% CI: 0.54–1.35; p = 0.5). Multivariate analysis revealed that HID-SCT was associated with a significantly lower cumulative incidence of relapse [Table S4].

The 2-year OS was 70.9% (95% CI: 66.3%–75.0%) for all patients, 70.2% (95% CI: 63.0%–76.4%) for patients without aGVHD, 78.3% (95% CI: 62.1%–88.2%) for patients with grade I aGVHD, 72.2% (95% CI: 62.7%–79.8%) for patients in the steroid group, and 67.4% (95% CI: 57.3%–75.6%) for patients in the RUX-steroid group [Figure 4E]. Multivariate analysis revealed that older patient age, noncomplete remission (non-CR) prior to transplantation, and CMV reactivation were significantly related to decreased OS [Table S4]. The 2-year overall survival for patients with PTLD was 63.8% (95% CI: 46.4%–76.8%). Five patients died within three months after the PTLD diagnosis. Two of them died from hematologic malignancy relapse, whereas three died from sepsis. One patient received the RUX-steroid regimen prior to PTLD diagnosis, one was treated with the steroid regimen for grade II–IV aGVHD, one had no history of aGVHD, and two experienced grade I aGVHD. Multivariate analysis confirmed older age and CMV infection as independent risk factors for OS among PTLD patients [Table S5].

There was no significant difference in DFS between the RUX-steroid group and the steroid group (HR: 1.04, 95% CI: 0.67–1.61; p = 0.859) [Figure 4F]. Multivariate analysis revealed that older patient age and non-CR prior to transplantation were significantly associated with decreased DFS [Table S4]. The cumulative incidence of moderate to severe cGVHD was comparable between the RUX-steroid group and the steroid group (HR: 1.09, 95% CI: 0.45–2.61; p = 0.85) [Figure 4G]. Multivariate analysis revealed that the occurrence of grade III–IV aGVHD was a factor that significantly influenced increased moderate to severe cGVHD [Table S4].

Discussion

In this retrospective cohort study, the risks of EBV reactivation and EBV-PTLD were evaluated in patients receiving low-dose RUX (5 mg/day) combined with steroids (1 mg·kg⁻¹·d⁻¹) as the initial treatment for grade II–IV aGVHD. Compared with patients in the steroid group, patients in the RUX-steroid group had a significantly higher cumulative incidence of EBV reactivation (82.9% vs. 70.2%, p = 0.03). The cumulative incidence of EBV-PTLD within 180 days was similar between the RUX-steroid group and the steroid group (5.0% vs. 4.5%, p = 0.86). Multivariate analysis revealed that the administration of RUX as a first-line treatment was not an independent risk factor for either EBV reactivation or EBV-PTLD. Nevertheless, the observed trend toward increased EBV reactivation suggests that close monitoring of the EBV viral load remains essential for patients treated with RUX, even at relatively low doses.

Activated Janus kinase (JAK) is required for T effector cell responses in different inflammatory processes, and its blockade can potently reduce aGVHD.²¹ JAK1 is essential for responses to interferon (IFN), interleukin-10 (IL-10), IL-2/IL-4, and IL-6, whereas JAK2 is involved in responses to single-chain receptors, the IL-3 cytokine family, and IFN-γ receptors. JAK1/2 inhibition may increase viral risk by impairing these immune pathways.^7,22 RUX broadly suppresses immune function by impairing dendritic cell differentiation and function, reducing the number and maturation of NK cells, and decreasing the number of circulating regulatory T cells.^7,23 These effects dampen both innate and adaptive responses, impair virus clearance, and alter cytokine production, contributing to the increased risk of opportunistic viral infections. RUX has been linked to increased risks of infection with adenovirus, herpes zoster and hepatitis B.^7,8,24 Given these immunosuppressive effects, RUX may also promote EBV infection. Moreover, high levels of signal transducer and activator of transcription-3 (STAT3) maintain EBV latency type III through positive regulation of BamHI-Q promoter activity, suggesting an additional mechanistic link between JAK inhibition and EBV reactivation.25, 26, 27 However, whether RUX increases the risk of EBV infection remains unclear, as comparative studies on this topic are lacking. González Vicent et al. reported a mild increase in EBV infection (13%) among 22 patients with steroid-refractory GVHD following the administration of RUX (median dosage, 10 mg/day; range, 2.5–20).²⁸ Lebon et al. reported that RUX (10–20 mg/day) was a significant risk factor for the first instance of EBV reactivation after GVHD in a study that included 137 GVHD patients.⁶ In contrast, the REACH2 study reported similar rates of EBV infection between the RUX and control groups.²⁹ This study represents the first comparative study focusing on EBV infection after the use of RUX. In addition to RCT participants, patients from a prior single-arm study and off-label use were also included to better reflect real-world practice. Our results indicated that a lower dosage of RUX (5 mg/day) was still related to an increasing trend of EBV reactivation. This may be explained by the pharmacokinetics of RUX, which inhibited cytokine-induced pSTAT3 in a dose-dependent manner. At a dose of 5 mg, pSTAT3 inhibition was approximately 40%, whereas at 10 mg, it increased to approximately 54%. These findings suggested that the immunosuppressive effect of RUX at 5 mg remained substantial.³⁰ In addition, the incidence of hemorrhagic cystitis was slightly greater in the RUX group than in the steroid group, likely because of weakened antiviral immunity, which facilitates BK polyomavirus reactivation and subsequent hemorrhagic cystitis.³¹

Although 5 mg/kg RUX administered in the first-line setting for aGVHD was associated with an increasing trend of EBV reactivation, it did not appear to influence the incidence of EBV-PTLD. This may be partly attributed to the use of preemptive rituximab therapy based on EBV-DNA level monitoring, with the goal of preventing progression to EBV-PTLD.^18,32 Beyond this clinical factor, the JAK-STAT pathway plays a dual role: while its inhibition may promote viral reactivation by disrupting latency, it is also required for B-cell transformation and survival during lymphomagenesis.³³ Furthermore, analysis of immune reconstitution revealed that the absolute counts of T-cell subsets in the RUX-steroid group were not significantly lower than those in the steroid group on day 100, suggesting that the impact of low-dose RUX may be more focused on functional aspects rather than quantitative suppression. In summary, the effect of low-dose RUX on the risk of EBV-PTLD appears to be relatively mild.

Numerous studies have investigated the risk factors for EBV reactivation and EBV-PTLD following transplantation. Among these, the administration of ATG has been identified as the most important factor.³⁴ In this study, all patients received ATG, providing a uniform background to explore other potential risk factors. Under these conditions, multivariate analysis revealed that MSD-SCT was the only factor significantly associated with a reduced risk of EBV reactivation and EBV-PTLD, which is likely because patients receiving HID-SCT or URD-SCT were given a higher cumulative ATG dose (10 mg/kg) than the MSD-SCT group (5 mg/kg). These findings are consistent with prior evidence that ATG has a clear dose‒response relationship with its immunosuppressive effect.³⁵ In addition, multivariate analysis revealed the type of donor as the sole independent predictor of EBV reactivation, with aGVHD and RUX use not remaining significant. These findings suggest that the profound alloreactivity inherent in HID-SCT or URD-SCT drives both the incidence of aGVHD and the consequent need for prolonged RUX therapy. Thus, aGVHD and RUX act as intermediate steps—rather than independent drivers—on the causal pathway from donor-derived immune disparity to impaired EBV surveillance and reactivation.

Our previous study revealed that the one-year cumulative incidence of EBV-DNAemia was 44.1%, which seemed much lower than that in this study.¹³ This difference may be attributed to the higher rate of ATG use (100% vs. 62.5%) and greater proportion of HID-SCT (67.8% vs. 55.0%) in the current cohort than in the previous study.^36,37

One limitation of this study was its retrospective and nonrandomized design. Selection bias, baseline imbalance and residual confounding cannot be excluded completely but were minimized through consecutive patient selection and multivariate analyses. Another limitation of this study was that it was a single-center study, which may limit the generalizability of our findings. Certain potential confounding factors—such as prophylactic antiviral use, immunosuppressive protocols, and limited variability in baseline EBV serostatus—were not fully accounted for. The potential long-term effects of low-dose RUX on immune reconstitution remain uncertain because of the limited follow-up for immune function assessment. Larger multicenter studies are needed to validate our findings; however, such efforts are currently limited by the lack of standardized EBV monitoring and treatment thresholds across centers.

In conclusion, the retrospective analysis revealed a trend toward increased EBV reactivation with low-dose RUX; however, it was not associated with an increased incidence of EBV-PTLD. These findings support the continued use of EBV DNA level monitoring in patients receiving RUX, coupled with timely intervention when indicated.

Authors contribution

Kun Qian: Data curation, Methodology, and Writing–original draft; Shuangyi Xing: Data curation, Visualization, and Writing–original draft; Yujun Wei: Data curation and Formal analysis; Lili Wang: Project administration and Resources; Jingjing Yang: Visualization and Formal analysis; Fei Li: Project administration and Resources; Zhenyang Gu: Project administration and Resources; Nan Wang: Formal analysis; Liping Dou: Conceptualization, Supervision and Funding acquisition; Daihong Liu: Conceptualization, Supervision and Writing–review and editing. All the authors critically revised and approved the final version of the manuscript.

Ethics statement

This study was conducted in accordance with the Declaration of Helsinki and approved by the Ethics Committee of Chinese PLA General Hospital (No. S2019-177-01). Written informed consent was obtained from all patients before transplantation.

Data availability statement

The datasets generated and analyzed during the current study are available from the corresponding author upon reasonable request.

Declaration of Generative AI and AI-assisted technologies in the writing process

The authors declare that generative artificial intelligence (AI) and AI-assisted technologies were not used in the writing process or any other process during the preparation of this manuscript.

Funding

This work was partially supported by grants from the National Key R&D Program of China (Nos. 2023YFC2507800 and 2021YFA1100904), the National Natural Science Foundation of China (Nos. 82270162, 82270224, and 82200169), the Beijing Natural Science Foundation of China (No. 7222175), the Military Medical Support Innovation and Generation Program (No. 21WQ034), the Special Research Fund for Health Protection (No. 24BJZ30 and 21BJZ30), the Beijing Nova Program Cross-cutting Project (No. 20230484407), the Logistics Independent Research Program (No. 2023hqzz09), the Innovation Science Foundation of Youth Program of Chinese PLA General Hospital (No. 22QNFC021) and Capital Funds for Health Improvement and Research (No. 2024-2-5063).

Conflict of interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Managing Editor: Peng Lyu

Appendix A. Supplementary data

The following is the Supplementary data to this article: