Hematology · Medical Literature
Low-dose ruxolitinib for graft-versus-host disease prevention in haploidentical haematopoietic stem-cell transplantation: a multicentre, open-label, randomised, controlled, phase 3 trial.
Abstract
Acute graft-versus-host disease (GVHD) remains a major cause of morbidity and mortality after haploidentical haematopoietic stem-cell transplantation (HSCT). Ruxolitinib, a Janus kinase (JAK) 1/2 inhibitor with established activity in steroid-refractory GVHD, has shown promise for prophylaxis in early studies, but there is little evidence from randomised trials in the haploidentical HSCT setting. We investigated whether, within a backbone of antithymocyte globulin, calcineurin inhibitor, and short-course methotrexate, replacing mycophenolate mofetil with low-dose ruxolitinib could reduce acute GVHD after haploidentical HSCT. In this multicentre, open-label, randomised, controlled, phase 3 trial conducted at five centres in China, eligible patients were aged 12-70 years, had haematological malignancies for which allogeneic HSCT was indicated, had a Karnofsky performance status of at least 70 or a Lansky score of at least 70 for patients younger than 16 years, and were undergoing their first myeloablative haploidentical HSCT. Patients were randomly assigned (1:1) to receive antithymocyte globulin, a calcineurin inhibitor, short-course methotrexate, and low-dose ruxolitinib, or standard prophylaxis with antithymocyte globulin, a calcineurin inhibitor, short-course methotrexate, and mycophenolate mofetil. Randomisation was stratified by centre and patient age (<40 years vs ≥40 years) using computer-generated permuted blocks with random block sizes. Oral ruxolitinib was started on day 1 at 5 mg twice daily for patients weighing at least 50 kg and 5 mg once daily for those weighing less than 50 kg, continued up to day 60, and then tapered to day 90 in the absence of grade II-IV acute GVHD. The primary endpoint was the cumulative incidence of grade II-IV acute GVHD by day 100. Efficacy and safety were analysed in the modified intention-to-treat population, which included all randomly assigned patients who received at least one dose of the assigned prophylaxis, excluding those with major protocol deviations. This trial is registered with ClinicalTrials.gov, NCT04838704, and is complete. Between April 1, 2021, and Dec 28, 2023, 215 patients were randomly assigned, of whom 206 were included in the modified intention-to-treat population (103 in the ruxolitinib prophylaxis group and 103 in the standard prophylaxis group). The median recipient age was 40 years (IQR 24-48), 98 (48%) patients were female and 108 (52%) were male, and all participants were Chinese. Among surviving patients, the median follow-up was 26·4 months (IQR 20·0-33·3). By day 100, grade II-IV acute GVHD occurred in seven (cumulative incidence of 6·8% [95% CI 1·9-11·7]) of 103 patients in the ruxolitinib prophylaxis group and 38 (36·9% [27·5-46·3]) of 103 patients in the standard prophylaxis group (subdistribution hazard ratio 0·15, 95% CI 0·07-0·34; p<0·0001). The most common grade 3-4 adverse events in the ruxolitinib prophylaxis group and the standard prophylaxis group were thrombocytopenia (17 [17%] vs 11 [11%]), neutropenia (14 [14%] vs 9 [9%]), anaemia (12 [12%] vs 9 [9%]), and cystitis (7 [7%] vs 10 [10%]). Serious adverse events occurred in 31 (30%) patients in the ruxolitinib prophylaxis group and 36 (35%) patients in the standard prophylaxis group. No fatal adverse events occurred in the ruxolitinib prophylaxis group. Three deaths occurred in the standard prophylaxis group: pulmonary infection (n=1), sepsis, bacteraemia, or fungaemia (n=1) and transplant-associated thrombotic microangiopathy (n=1). No deaths were considered related to the assigned study drug. Low-dose ruxolitinib, used in place of mycophenolate mofetil within a backbone of prophylactic antithymocyte globulin, calcineurin inhibitor, and short-course methotrexate, reduced grade II-IV acute GVHD by day 100 after haploidentical HSCT and had manageable toxicity. These findings support further evaluation of targeted JAK1/2 inhibition as part of GVHD prophylaxis. National Natural Science Foundation of China.
Clinical relevance
Acute GVHD remains a leading cause of morbidity and non-relapse mortality following haploidentical HSCT, a modality increasingly used when matched donors are unavailable. Despite standard prophylaxis with ATG, calcineurin inhibitors, and MMF, grade II–IV acute GVHD rates remain unacceptably high (~37%), underscoring the urgent need for more effective prophylactic strategies. This phase 3 RCT provides the first high-level randomised evidence supporting JAK1/2 inhibition with low-dose ruxolitinib as a prophylactic agent in the haploidentical HSCT setting.
Key findings
1. The cumulative incidence of grade II–IV acute GVHD by day 100 was 6.8% (95% CI 1.9–11.7%) in the ruxolitinib group versus 36.9% (95% CI 27.5–46.3%) in the standard MMF group — a statistically and clinically significant reduction (sHR 0.15, 95% CI 0.07–0.34; p<0.0001). 2. The absolute risk reduction for grade II–IV acute GVHD was approximately 30 percentage points, yielding a number needed to treat (NNT) of approximately 3–4 patients to prevent one case of acute GVHD. 3. The most common grade 3–4 adverse events in the ruxolitinib versus MMF groups were thrombocytopenia (17% vs. 11%), neutropenia (14% vs. 9%), and anaemia (12% vs. 9%), suggesting a modest but manageable increase in myelosuppression. 4. Serious adverse events occurred in 30% of ruxolitinib-treated patients versus 35% in the standard prophylaxis group, indicating no overall increase in serious toxicity burden with the experimental regimen. 5. No fatal adverse events were attributed to ruxolitinib; three deaths occurred in the MMF group (pulmonary infection, sepsis/bacteraemia/fungaemia, and TA-TMA), with none considered related to study drug in either arm. 6. The median follow-up among surviving patients was 26.4 months (IQR 20.0–33.3), providing meaningful intermediate-term safety and efficacy data across a modified ITT population of 206 patients (103 per arm).
Practice implications
Clinicians performing myeloablative haploidentical HSCT in patients aged 12–70 years with haematological malignancies and adequate performance status (Karnofsky/Lansky ≥70) should consider substituting MMF with low-dose ruxolitinib (5 mg BID for ≥50 kg; 5 mg QD for <50 kg) within a standard ATG/calcineurin inhibitor/short-course methotrexate backbone, initiating on day +1 and continuing through day 60 with taper to day 90 in the absence of grade II–IV acute GVHD. Monitoring for cytopenias — particularly thrombocytopenia and neutropenia — is warranted given modestly higher rates of grade 3–4 myelosuppression in the ruxolitinib arm. Prospective evaluation of long-term endpoints including relapse incidence, chronic GVHD, graft-versus-leukemia effect, and overall survival in non-Chinese populations will be essential before broad international adoption of this regimen.