A retrospective analysis was conducted to evaluate the treatment outcomes of 133 patients who underwent allo-HSCT and were diagnosed with aGVHD at our center between August 2014 and June 2023. The underlying diseases included aplastic anemia (AA), acute myeloid leukemia (AML), acute lymphocytic leukemia (ALL), myelodysplastic syndrome (MDS), and congenital hemophagocytic syndrome. The marrow transplant databases of the Second Affiliated Hospital of Dalian Medical University in China were screened for aGVHD after allo-HSCT diagnosis. The patient outcomes were monitored until the end of June 2024. This study was approved by the Ethics Committee of the Second Affiliated Hospital of Dalian Medical University.

All 133 patients underwent myeloablative (MA) conditioning regimen transplantation, except for those with AA. This included the treatment of acute leukemia and MDS based on the Beijing protocol, which utilized a modified BU/CY conditioning regimen as follows: cytarabine (2 g/m² q12h for 2 days; qd for fully matched donors), busulfan (3.2 mg/kg/day for 4 days), cyclophosphamide (1.8 g/m² for 3 days), and anti-thymocyte globulin (ATG) (2.5 mg/kg/day for 4 days; 2 days for fully matched donors) (24). Enhanced MA regimens, including decitabine, idarubicin, and clarithromycin, were administered according to the risk of primary disease and disease status before transplantation (25). For AA, the FLU/CY conditioning regimen, combined with ATG to suppress T cells, involved the administration of 30 mg/m² fludarabine for 4 days, 30–50 mg/kg cyclophosphamide for 2 days, and 2.5 mg/kg/d ATG for 4 days (26). Most transplantation recipients received ATG, cyclosporine A (CsA)/Tacrolimus (FK506), mycophenolate mofetil, and short-term methotrexate for GVHD prophylaxis.

The grading and staging systems of aGVHD were based on the Glucksberg or Mount Sinai aGVHD International Consortium (MAGIC) criteria (27, 28). Individuals were divided into two treatment groups. In the ruxolitinib/steroids combination group, patients were given methylprednisolone at a dose of 0.5–2 mg/kg/day, and ruxolitinib was administered orally at a daily dose of 10–15 mg. If patients with aGVHD responded to treatment by achieving partial response (PR) or complete response (CR) within 7 days, the methylprednisolone dosage was gradually tapered. If GVHD did not recur after steroid discontinuation, the ruxolitinib dose was gradually reduced. In the steroid-only group, patients were given methylprednisolone at a dose of 0.5–2 mg/kg/day, which was gradually tapered and discontinued after CR. The initial steroid dosage for both groups was determined based on the severity of the aGVHD. For mild patients (grade I and II aGVHD, only involving the upper digestive tract), the initial dosage was 0.5 mg/kg. For grade III/IV aGVHD involving the gastrointestinal tract or hyperacuteGVHD, the initial dosage of methylprednisolone was 2 mg/kg. For most grade II-III aGVHD, the dosage was 1 mg/kg. The steroid dosage in both groups were basically balanced. According to guidelines and relevant literature (12), during the treatment of aGVHD, both groups of CNI were adjusted to effective therapeutic concentrations (cyclosporine trough concentration to 150-250ng/ml, FK506 5-15ng/ml). After aGVHD achieves CR, stop steroid first, and then tappered gradually and eventually discontinued CNI. The CNI reduction plan for both groups is the same. Second-line therapy was initiated in both groups for patients with refractory aGVHD, defined as GVHD progression after 3 days of treatment, lack of improvement within 7 days, or failure to achieve CR after 14 days.

The primary endpoint was the overall response rate (ORR) to aGVHD treatment at 3, 7, 14, and 28 days post-intervention. ORR was defined as the percentage of patients in each group who attained either PR or CR without requiring additional immunosuppressive agents. CR was characterized by the complete resolution of aGVHD symptoms, whereas PR was defined as an improvement in at least one stage in a single organ without worsening in others. No response (NR) was classified as no improvement, worsening symptoms in any organ, or the emergence of new GVHD-related symptoms. Additionally, GVHD progression after 3 days of therapy or lack of improvement within 7 days was considered NR. Secondary endpoints included overall survival (OS), progression-free survival (PFS), failure-free survival (FFS), cumulative incidence of relapse (CIR), NRM, and safety. OS was defined as the time from aGVHD onset to death from any cause. PFS was defined as the time from aGVHD onset to relapse of the primary disease or death from any cause, whichever occurred first. Failure-free survival (FFS) refers to the time from aGVHD onset to disease relapse or progression, NRM, or the initiation of additional therapy for aGVHD. Safety endpoints were assessed based on the frequency of adverse events (AEs), defined according to the National Cancer Institute Common Terminology Criteria for Adverse Events (NCI-CTCAE 4.0).

Data were analyzed using STATA/SE 15.1 software (STATA Corp, College Station, Texas, USA). ORR, CR, and PR were analyzed using Pearson’s chi-square test. The correlation between ORR and different independent variables was calculated using regression analysis. The Kaplan-Meier method was used to estimate OS, PFS and FFS, and the Log-rank test was used to determine statistical significance. NRM and cumulative incidence of relapse were estimated by considering each other as competing risks. The reverse Kaplan-Meier method was used to calculate the median follow-up duration. Prognostic variables for OS were evaluated by univariate and multivariate analyses using Cox proportional hazard regression. Variables with statistical significance (p <0.1) in the univariate analysis were included in the multivariate analysis to adjust for potential confounding effects. A p-value <0.05 was considered statistically significant. The following variables were evaluated: sex, age at allo-HSCT (<60 years vs. ≥60 years), donor sex (female donor to male recipient vs. others), disease type (malignant vs. benign), donor type (cord blood, CBT vs. haplo-identical donor, HID vs. matched sibling donor, MSD vs. unrelated donor, URD), HLA type (fully match vs. haplo), blood type (match vs. mismatch), graft origin (CBT vs. PB+BM+CB vs. PB+BM vs. PB), hematopoietic cell transplant-comorbidity index (HCT-CI: 2 vs. 1 vs. 0), aGVHD grade (IV vs. III vs. II vs. I), aGVHD skin involvement (yes vs. no), aGVHD GI involvement (yes vs. no), aGVHD liver involvement (yes vs. no), day 3 ORR (yes vs. no), day 7 ORR (yes vs. no), day 14 ORR (yes vs. no), and day 28 ORR (yes vs. no). Pearson’s chi-square test was used to analyze AEs between different groups, including neutropenia, thrombocytopenia, cytomegalovirus (CMV) infection, Epstein-Barr virus (EBV) infection, renal toxicity, cardiac toxicity, digestive reactions, nerve toxicity, electrolyte disturbance, blood glucose disturbance, blood lipid disturbance, and hypertension.

The patient demographics and baseline disease characteristics are presented in Table 1 . This study included a cohort of 133 patients with aGVHD after allo-HSCT. Of these, 83 patients were classified into the ruxolitinib/steroid group, while 50 patients were in steroid-only group. The overall median age was 30 (0.5–67) years in the ruxolitinib/steroid group and 27.5 (1–61) years in the steroid-only group. The median body weight was 55 (9–100) kg vs. 52 (8.1–97) kg, respectively. The median number of infused mononuclear cells (MNC) was 9.41 × 10^8/kg (range: 4.68–20.4) in the ruxolitinib/steroid group and 10.31×10^8/kg (range: 4.27–28.3) in the steroid-only group. The median number of infused CD34+ cells was 6.63×10^6/kg (range: 2.29–22.7) and 6.01×10^6/kg (range: 0.99–21.7), respectively. The baseline demographic, transplantation-related, and disease-related patient characteristics-including recipient sex, disease type, donor type, donor sex, stem cell source, conditioning regimen intensity, recipient/donor blood type pair, GVHD prophylaxis, ATG use, HCI-CI, aGVHD skin involvement, aGVHD GI involvement, and aGVHD liver involvement-were comparable between the two groups. All patients underwent neutrophil engraftment, with no significant difference in the time to neutrophil or platelet engraftment between the two groups. Notably, grade I aGVHD was more common in the steroid-only group, while grade II aGVHD was more frequent in the ruxolitinib/steroids group (p=0.008).

The primary endpoint was ORR at 3, 7, 14 and 28 days post-intervention. As shown in Figure 1A and Table 2 , ORR on day 3 was comparable between the two groups (59% vs 58%; odds ratio [OR]=1.04, 95% confidence interval [CI]: 0.51–2.13, p=0.906). On day 7, ORR was significantly higher in the ruxolitinib/steroids combination group (86%) compared to the steroid-only group (68%; OR=2.78, 95% CI: 1.18–6.53 p=0.019; Figure 1B , Table 2 ). Similarly, ORR on day 14 was significantly greater in the ruxolitinib/steroids combination group (92%) compared to the steroid-only group (79%; OR=2.77, 95% CI: 0.98–7.87, p=0.05; Figure 1C , Table 2 ). The ruxolitinib/steroids combination cohort showed similar ORR on day 28 compared with the steroid-only cohort (94% vs 91%; OR=1.45, 95% CI: 0.37–5.69, p=0.593; Figure 1D , Table 2 ).

The median follow-up duration from aGVHD onset was 47 months (range: 1–80 months) in the ruxolitinib/steroids group and 90 months (range: 0.4–118 months) in steroid-only group. The 3-year OS rates were similar between two groups: 70.5% in the ruxolitinib/steroids group vs. 67.6% in the steroid-only group, both in Cox regression analysis (hazard ratio [HR]=0.86, 95% CI: 0.45–1.61, p=0.62) ( Table 3 ) and Kaplan-Meier survival analysis (Log-rank test, p=0.65. Figure 2A ). Table 3 shows the univariate and multivariate analysis results for risk factors influencing OS. Univariate analysis revealed the following: (i) ORR on day 7 and day 14 had a significantly positive effect on OS (HR=0.45, 95% CI: 0.23–0.88, p=0.02) (HR=0.37, 95% CI: 0.17–0.78, p=0.009), confirmed by Kaplan-Meier survival analysis ( Figures 2B, C ). (ii) Malignant hematologic disease had a significantly negative effect on OS (HR=3.26, 95% CI: 1.16–9.17, p=0.02), also confirmed by Kaplan-Meier survival analysis ( Figure 2D ). (iii) Other factors-including donor type (CBT vs. HID vs. MSD vs. URD), HLA type (fully matched vs. haploidentical), stem cell source (CBT vs. PB+BM+CB vs. PB+BM vs. PB), and HCI-CI-influenced OS ( Table 3 , Figure 3 ). In the multivariate analysis, only day 14 ORR (HR=0.37, 95% CI: 0.16–0.88, p=0.02), aGVHD grade (HR=3.21, 95% CI, 0.96–10.6, p=0.05), and HCI-CI (HR=2.43, 95% CI, 0.99–5.99, p=0.05) influenced OS ( Table 3 ). Additionally, we analyzed other major risk variables for OS ( Table 3 ). Recipient sex, age at transplantation, donor sex, recipient-donor blood pair, aGVHD skin involvement, aGVHD GI involvement, and aGVHD liver involvement had no statistically significant effect on OS. Though the 1-year PFS rates were similar between the two groups-74% in the ruxolitinib/steroids group and 72% in the steroid-only group, as shown in Kaplan-Meier survival analysis ( Figure 4A , log-rank test, p=0.927)—ORR on day 7 ( Figure 4B , log-rank test, p=0.03) and day 14 ( Figure 4C , log-rank test, p=0.001) demonstrated a significantly more favorable impact on PFS. The 1-year FFS rates were similar between the two groups, with 56.2% in the ruxolitinib/steroids group and 55.9% in the steroid-only group ( Figure 4D , log-rank test, p=0.47). Reaching ORR on day 7 ( Figure 4E , log-rank test, p=0.046) and day 14 ( Figure 4F , log-rank test, p=0.0004) had a significantly more favorable effect on FFS. Both groups had comparable 2-year cumulative incidence rates of relapse (ruxolitinib/steroids: 18% vs. steroid-only: 11%; Subdistribution HR, SHR=1.69, 95% CI: 0.68–4.24, p=0.257; Figure 5A ) and non-relapse mortality (ruxolitinib/steroids: 15% vs. steroid-only: 23%; SHR=0.60, 95% CI: 0.24–1.45, p=0.26; Figure 5B ).

In the subgroup of patients who received peripheral blood stem cell transplantation, the 3-year OS tended to be higher in the ruxolitinib/steroid combination group than in the steroid-only group (64.3% vs. 27.3%; log-rank test, p=0.046; Figure 6A ). Similarly, the 3-year PFS was superior in the ruxolitinib/steroid group compared to the steroid-only group (57.1% vs. 27.3%; log-rank test, p=0.05; Figure 6B ). Finally, the 2-year FFS was better in the ruxolitinib/steroid combination group than in the steroid-only group (21% vs. 9%; log-rank test, p=0.05; Figure 6C ).

The median duration of ruxolitinib exposure was 70 days (range: 9–730 days) among the 83 patients, and none of the 83 patients discontinued ruxolitinib due to side effects, especially hepatotoxicity. No ruxolitinib toxicity related death was observed. As shown in Table 4 , the most frequently observed AEs in the ruxolitinib/steroid group were neutropenia (32.5% vs. 12%, p=0.008), CMV infection (34.9% vs. 18%, p=0.036), and renal toxicity (20.4% vs. 6%, p=0.024). Other AEs, including thrombocytopenia, EBV infection, cardiac toxicity, nerve toxicity, electrolyte disturbance, blood glucose disturbance, blood lipid disturbance, and hypertension, were comparable between the two groups.