Low rates of chronic graft-versus-host disease with ruxolitinib maintenance following allogeneic HCT Free

Graft-versus-host disease (GVHD) is the major immunologic complication limiting long-term success of allogeneic hematopoietic cell transplantation (HCT). Specifically, chronic GVHD (cGVHD), which typically occurs in the later months after HCT, can impair quality of life and increases the risk for both morbidity and mortality in HCT recipients. Posttransplant cyclophosphamide (PTCy) has revolutionized GVHD prophylaxis, expanding HCT accessibility in patients who lack an HLA-matched donor,^1,2 and is now considered the standard approach to GVHD prevention in HLA-matched HCT after reduced-intensity conditioning (RIC) because of its effectiveness in preventing acute GVHD (aGVHD) and cGVHD.^3,4 Nevertheless, there is room to further lower the incidence of cGVHD, and potential concerns, such as early toxicity, poor graft function, and delayed immune recovery, exist with PTCy. Outside the United States, the use of antithymocyte globulin products is standard in unrelated donor HCT, although infectious complications are a concern.^5,6 Thus, development of novel GVHD prevention regimens remains important, as having multiple safe and effective approaches allows the needs of the expanding and diverse HCT population to be met.

Ruxolitinib is an oral inhibitor of Janus kinase 1 and 2, approved by the US Food and Drug Administration and the European Medicines Agency for the treatment of corticosteroid-refractory aGVHD and cGVHD, respectively.^7-9 However, clinical data regarding the ability of ruxolitinib to prevent GVHD are limited.^10,11 In the current study, we investigated the clinical effect of prolonged ruxolitinib administration following HCT.

This prospective, open-label, multicenter, phase 2 trial was conducted at 7 clinical sites in the United States, with the Massachusetts General Hospital serving as the coordinating center. The study was approved by the institutional review board at each center. Informed consent was obtained from all enrolled participants. The trial was conducted according to the Declaration of Helsinki and registered at ClinicalTrials.gov (#NCT03286530).

Participants were older adults (aged 60-80 years) with acute myeloid leukemia in first remission or myelodysplastic syndrome and were enrolled before HCT. Donors were HLA-matched related or 7/8 or 8/8 HLA-matched unrelated. All participants received RIC regimens, peripheral blood stem cell grafts, and GVHD prophylaxis with tacrolimus (Tac) and methotrexate (MTX, 5-10 mg/m² on days +1, +3, and +6 ± 11). Ruxolitinib therapy began between day +30 and 100 following HCT after meeting pretreatment criteria: donor engraftment (donor chimerism ≥70%), disease remission confirmed by bone marrow biopsy, adequate hematopoietic parameters (hemoglobin ≥8 g/dL, platelets ≥50 000/μL), and the absence of progressive aGVHD (aGVHD requiring >1 mg/kg per day of corticosteroids or the addition of immunosuppression beyond corticosteroids). Ruxolitinib was initiated at 10 mg twice daily and administered continuously in 28-day cycles for up to 24 cycles. Ruxolitinib dosing was reduced by 50% for concurrent administration of strong cytochrome P450 3A4 (CYP3A4) inhibitors or creatinine clearance <40 mL/min. If grade ≥3 treatment-related adverse events (AEs) occurred, ruxolitinib was held until the AE improved to grade ≤2, and then restarted at a 50% dose reduction. Ruxolitinib dosing could be reescalated if the AE completely resolved. Participants requiring a delay ≥2 weeks were removed from protocol therapy.

The primary objective was to evaluate the GVHD-free, relapse-free survival (GRFS) at 1 year after HCT, defined as the absence of grade 3 to 4 aGVHD, cGVHD requiring systemic therapy, relapse, or death. The goal was to improve the GRFS rate to 45% with ruxolitinib from a historical rate of 29% with Tac/MTX.¹² With a goal accrual of 64 participants receiving the intervention, ruxolitinib would be considered worthy of further study if ≥24 participants were event free at 1 year, which gave 91% power at 1-sided type I error rate of 0.09. Key secondary objectives included characterization of treatment-related toxicities, the cumulative incidence of aGVHD and cGVHD, disease relapse, and nonrelapse mortality, as well as estimates of overall survival and progression-free survival.

Seventy-eight participants were enrolled before HCT. All participants underwent HCT; 15 participants did not initiate ruxolitinib (Figure 1). The baseline characteristics of the 63 participants who received ruxolitinib are shown in Table 1.

The median follow-up of survivors was 41 months (range, 10-61 months). The median start day of ruxolitinib after HCT was day 45 (range, 32-93 days). At the time of analysis, the median number of cycles of treatment was 24 (range, 2-24). Four participants remain on therapy; the reasons for discontinuation include completed treatment (n = 33), disease relapse (n = 15), adverse events (n = 8), and physician discretion (n = 3). Ruxolitinib-related AEs are listed in supplemental Table 1 (available on the Blood website). Severe (grade ≥3) cytopenias, regardless of attribution to ruxolitinib, occurred during treatment: neutropenia (n = 18), thrombocytopenia (n = 17), and anemia (n = 14), with 11 participants (17%) experiencing recurrent severe cytopenias. Twenty-two participants (35%) required a dose reduction of ruxolitinib during their treatment course; however, only 6 participants discontinued ruxolitinib because of cytopenias. The equivalent ruxolitinib dose (adjusted for concurrent administration of strong CYP3A4 inhibitors) received by participants is shown in supplemental Figure 1. Eight participants experienced grade ≥3 infectious events (supplemental Table 2), with most events occurring during the 30-day follow-up after ruxolitinib discontinuation while being treated for relapsed disease. No invasive fungal infections were observed.

Clinical outcomes are listed in supplemental Table 3. GRFS at 1 year, the primary end point, was 70% (95% confidence interval [CI], 57%-80%). Among participants who received ruxolitinib, the 6-month cumulative incidence of grade 2 to 4 aGVHD was 14% (95% CI, 7%-24%), and the 6-month cumulative incidence of grade 3 to 4 aGVHD was 4.8% (95% CI, 1.3%-12%), with 4 cases of grade 2 to 4 GVHD occurring before initiation of ruxolitinib. Although the incidence of all cases of cGVHD was 40% (95% CI, 28%-52%) at 2 years, most cases were mild. The incidences of moderate-severe cGVHD and cGVHD requiring systemic therapy at 2 years were 16% (95% CI, 8.2%-26%) and 13% (95% CI, 5.9%-22%), respectively (Figure 2A). The 2-year cumulative incidence of nonrelapse mortality was 4.8% (95% CI, 1.2%-12%), with only 1 death attributed to GVHD. The 2-year cumulative incidence of disease relapse was 27% (95% CI, 17%-38%). The 2-year overall survival and progression-free survival were 76% (95% CI, 63%-85%) and 68% (95% CI, 55%-78%), respectively (Figure 2B). Additionally, 2-year cGVHD-free, relapse-free survival was 62% (95% CI, 49%-73%). In univariable analysis (supplemental Table 4), fludarabine/melphalan conditioning was associated with the development of moderate-severe cGVHD (subdistributed hazard ratio, 8.49; 95% CI, 1.08-66.5; P = .04).

We observed that the prolonged administration of ruxolitinib following allogeneic HCT was associated with a low incidence of clinically significant cGVHD. Although the observed incidence of overall cGVHD (40% at 2 years) is expected for calcineurin inhibitor-based approaches, most of the cases were mild. The rates of clinically significant cGVHD in this study (moderate-severe cGVHD at 2 years, 16%; cGVHD requiring systemic therapy at 2 years, 13%) are lower than what has been achieved with Tac/MTX alone, and are in line with experiences with PTCy.^4,13,14 This improvement in cGVHD prevention was achieved without apparent compromise of the graft-versus-leukemia effect, as the incidence of disease relapse on this trial is comparable with standard experiences with RIC HCT. In our experience, ruxolitinib administration was feasible and associated with low rates of toxicities and severe infections, despite the prolonged course over the first 2 years after HCT. We believe these results strongly support further investigation of the use of ruxolitinib for GVHD prophylaxis, as it may represent an effective alternative to PTCy or antithymocyte globulin–based platforms, especially in older patients with comorbidities. Reports of ruxolitinib use in the setting of HCT for myelofibrosis or as aGVHD prophylaxis have demonstrated the feasibility of the approach, but with planned completion of ruxolitinib within 6 months of HCT.^10,11,15,16 A novel finding from the current trial is that by continuing ruxolitinib throughout the first 2 years of transplant, low rates of clinically significant cGVHD were observed, suggesting that duration of therapy is a key consideration to optimizing ruxolitinib’s efficacy in GVHD prophylaxis.

A clear advantage of PTCy over other GVHD prophylaxis approaches is the effective prevention of both aGVHD and cGVHD. Multiple approaches have demonstrated efficacy in preventing aGVHD without significantly impacting cGVHD,^17,18 whereas recent investigational approaches have sought to specifically enhance protection from cGVHD.¹⁹ In the current trial, our dosing strategy allows for evaluation of cGVHD-related outcomes. However, we are unable to properly evaluate aGVHD outcomes, as (1) the median start date was day +45 and (2) progressive aGVHD was a treatment exclusion criterion, occurring in 4 participants. A separate phase 2 multicenter trial of ruxolitinib initiated during conditioning and continuing through the peri-HCT period and for 12 months after in patients with myelofibrosis demonstrated low rates of severe aGVHD (grade 3-4, 2.4%),²⁰ suggesting that early ruxolitinib administration may effectively prevent aGVHD in addition to clinically significant cGVHD. Because of the above-mentioned limitations in evaluating aGVHD prevention, the interpretation of the primary end point of GRFS at 1 year is difficult. Thus, we analyzed the composite end point cGVHD-free, relapse-free survival, in which cGVHD requiring systemic therapy, relapse, and death are events.²¹ The 2-year CFRS was 62%, comparing favorably with other HCT approaches.^5,22 

We acknowledge the limitations of the current study. The trial only enrolled HLA-matched HCT, so data regarding the use of ruxolitinib in 7/8 HLA-mismatched HCT are absent. As discussed above, initiating ruxolitinib earlier in the transplant course, such as starting on day −1, would provide the opportunity to truly evaluate prevention of aGVHD in addition to cGVHD. Additionally, the optimal dosing and duration of ruxolitinib administration is unknown. The current study investigated at an initial dose of 10 mg twice daily, while mandating dose adjustments for concurrent medications and toxicities, and continued ruxolitinib for up to 24 months. It is unknown if shorter durations may be sufficient. However, data suggest that short durations (until day 30 or day 100) do not achieve the desired impact on GVHD prevention.¹⁶ Finally, although the incidence of severe infections was low, ruxolitinib was not administered peritransplant when infectious risk may be higher, and immune reconstitution was not evaluated in the current study. The above issues will be evaluated in Blood and Marrow Transplant Clinical Trials Network 2203, a randomized phase 3 trial that will compare the investigational arm (Tac/MTX/ruxolitinib) with standard of care (PTCy/Tac/mycophenolate mofetil) in the setting of HLA-matched, RIC allogeneic HCT. This trial is positioned to evaluate the promising results of the current trial and support the development of additional safe and effective approaches to GVHD prevention.

Figures and graphical abstract were created with use of BioRender.com.

Clinical trial funding and ruxolitinib were provided by Incyte.

Contribution: Z.D., Y.-B.C., and G.S.H. designed the study; Z.D., L.W.K., S.M.O., S.E.D., M.W., B.D., M.A.S., S.V., S.A., J.C., A.E.-J., M.J.F., S.M., R.A.N., P.V.O., T.R.S., and Y.-B.C. recruited patients to the study and collected clinical data; Z.D., H.T.K., and G.S.H. analyzed the data; H.T.K. and Z.D. interpreted the analysis results; Z.D. wrote the first draft of the manuscript; and all authors approved the final version of the manuscript and submission of the manuscript.

Conflict-of-interest disclosure: Z.D. reports research support from Incyte, RegImmune, Taiho Oncology, and Kura Oncology; and consulting fees or honoraria from Sanofi, Incyte, Inhibrx, PharmaBiome AG, Ono Pharmaceutical, RegImmune, MaaT Pharma, Forte Biosciences, and Medexus Pharmaceuticals. B.D. reports institutional research funding from Janssen, Angiocrine, Pfizer, Poseida, MEI, Orcabio, Wugen, Allovir, Adicet, Bristol Myers Squibb, Molecular Template, and Atara; and consulting fees or honoraria from MJH BioScience, Arivan Research, Janssen, ADC Therapeutics, Gilead, GlaxoSmithKline, Caribou, Roche, and Autolus. M.A.S. reports serving as a Data Safety Monitoring Board board member for Marker Therapeutics, Sorrento Therapeutics, and Kura Oncology; reports serving as a clinical trial adjudicator for GlaxoSmithKline and Novo Nordisk; and reports consulting fees or honoraria from Incyte, Sanofi, and StemLine. S.A. reports research support from Incyte and AltruBio; and honoraria from Incyte. R.A.N. reports research support from Incyte; consulting fees or honoraria from Sanofi and AbbVie; and a family member with equity in Vertex Pharmaceuticals. T.R.S. reports serving on a Data Monitoring Committee for Bluebird Bio; reports serving on Data Safety Monitoring and Adjudication Committees for Syneos Health; reports serving on a Scientific Review Committee for Ossium Heath; reports serving on a Scientific Advisory Board for Qihan Biotech; and reports consulting fees from Syneos Health. Y.-B.C. reports consulting fees from Incyte, Vor, CSL Behring, MaaT Biotherapeutics, and Ironwood; trial committee participation for Novo Nordisk, Editas, Alexion, and Daiichi; and equity in ImmunoFree and Phesi. The remaining authors declare no competing financial interests.

Correspondence: Zachariah DeFilipp, 55 Fruit St, Boston, MA 02114; email: zdefilipp@mgh.harvard.edu.