Introduction

Chronic graft-versus-host-disease (cGVHD) is a multi-organ syndrome characterized by tissue inflammation and fibrosis. First-line treatment remains corticosteroids; however, the therapeutic landscape is shifting towards more targeted therapies [1, 2]. Recently there are three oral agents FDA-approved for the treatment of steroid-refractory (SR)-cGVHD: ibrutinib, ruxolitinib (RUX), and belumosudil (BEL) [2,3,4]. While these novel drugs show promising clinical efficacy, complete responses are rare, particularly in highly morbid forms of cGVHD [2, 3, 5]. The optimal approach to cGVHD treatment in the era of these newly approved agents is an area of active research, with studies proposing potential combination therapy for maximum effect [6, 7]. Belumosudil is a selective inhibitor of the Rho-associated coiled-coil containing protein kinase 2 that blocks production of proinflammatory cytokines, promotes T-regulatory cells, and inhibits fibrosis [1]. Following BEL’s FDA-approval, we noticed an increase in BEL combinations prescribed concurrently with the novel agent, RUX, at our institution. We sought to investigate BEL’s use in combination with novel and standard agents.

Methods

This post-FDA approval descriptive observational study of BEL was performed to assess real-world use of BEL combinations in SR-cGVHD patients at the Stanford University Medical Center. Adult patients with clinical SR-cGVHD who were prescribed BEL as an additional therapy after one or more prior lines of cGVHD therapy were included. Data were collected retrospectively from electronic medical records.

The primary outcome variable was ORR defined as best response (CR, PR, or clinical benefit) in each organ during the 12-month period after starting BEL. Responses were based on organ symptom scoring from the 2014 NIH Consensus Criteria for GVHD (7 patients) or global rating by physician assessment of clinical benefit, defined as an improvement in patient functioning and symptoms (all others) [8,9,10]. We assessed the ORR of two groups: BEL + standard agent, such as prednisone, tacrolimus, or sirolimus (BEL+STD), and BEL + ruxolitinib (BEL+RUX). We explored mean ORRs for each group within an organ system using stratification by numbers of evaluable patients to look for suggestive trends. Secondary outcomes were regimen overall survival (rOS), regimen non-relapse mortality (rNRM), and rate of infections [9]. The rOS, rNRM, and regimen follow-up time (rFT) begin on the date when the BEL combination is initiated and end on the dates of death from any cause, death without relapse/progression, and last follow-up, respectively.

We utilized the Statistical Package for Social Sciences for descriptive statistics plus the R package survival to compute point and interval estimates using Kaplan-Meier and Greenwood methods for rOS, cumulative incidence and infinitesimal jackknife methods for rNRM, and 50th percentile of the reverse Kaplan-Meier method for median rFT [11].

Results

Twenty-six patients were identified from July 2021 to July 2022. The median age was 54 years and the median number of prior GVHD therapies received was 3 [1–8]. Skin was the most affected organ (n = 22), of which 17 patients had sclerotic features, followed by mouth (n = 20), eye (n = 16), and joints/fascia (n = 16) (Fig. 1a). Fifteen unique BEL combinations were prescribed. For 17 patients, BEL was added for cGVHD worsening; for 9, no improvement of the prior regimen. Thirteen patients received a combination regimen with BEL+RUX. In these patients, BEL was added after RUX had been initiated (Fig. 1b). The median interval from RUX to addition of BEL was 10 months (range 1–20). In the BEL+STD group, five of the 13 patients had prior RUX.

Fig. 1: Demographic data and outcomes of patients who received BEL combination therapy.
figure 1

a Baseline patient characteristics. b List of all belumosudil (BEL) combination regimens in the cohort. c Best overall response rate (ORR) stratified by BEL combination regimen at one year.

Full size image

In our small sample, the suggestive trend was that mean ORR appears higher across all organ systems except skin and GI in BEL+RUX relative to BEL+STD (Fig. 1c). The median time to best response for the entire cohort was 123 days [13–435 days]. In this heavily pretreated population, there were no stringent CRs, but PRs and clinical benefit from baseline were evident in the first year after BEL was added. Skin was advanced with 82% of patients having sclerotic features at study start. Of note, five patients were taken off BEL within one year due to lack of clinical benefit.

There were 27 infectious events in the cohort. Of these, SARS-CoV-2 infections accounted for 40% (11/27) of all events (Fig. 1a). As of February 2024, three deaths occurred, two of which were attributed to GVHD complications, and one due to SARS-CoV-2 infection.

The median rFT for all patients was 24 months with 95% confidence interval (CI95) 22–25 months. The 24-month rOS probability was 80% with CI95 59–100% in BEL+STD, and 92% with CI95 79–100% in BEL+RUX. The 24-month rNRM probability was 20% with CI95 6%–70% in BEL+STD, and 8% with CI95 1–51% in BEL+RUX. There was no statistically significant difference between groups for rOS and rNRM. There were no primary disease relapses.

Discussion

This study is one of the first to demonstrate real-world use of BEL cGVHD combination therapy using both standard and novel agent RUX. The rationale to combine BEL’s immunomodulatory and anti-fibrotic effects with cGVHD agents that suppress other pathways of inflammation and immune activity in an attempt to achieve superior responses aligns with that of Pusic et al. who reported a retrospective study of 14 patients on a BEL+RUX combination for cGVHD [12]. Similar to Pusic et al., we found the combination of BEL+RUX to be well tolerated without excess toxicity. However, our study extends the addition of BEL to encompass many combinations with BEL that occurred early post-FDA approval, allowing us to look at safety and response patterns within the combinations of BEL+RUX as well as BEL+STD.

In the ROCKstar trial which led to BEL’s approval for cGVHD, BEL could be used concomitantly with standard of care (SOC) agents, such as calcineurin inhibitors, steroids, cellcept, or ECP [3]. Concurrent treatment with investigational cGVHD agents such as RUX was prohibited. The ROCKstar trial showed a best ORR of 74–77% with BEL+STD regimens, but did not delineate the details of the STD combinations or the duration of use of these agents. In contrast, our study occurred after the three FDA drug approvals (ibrutinib, RUX, BEL), which allowed us to examine differences in the cGVHD regimens using STD agent combinations with BEL versus novel agents with BEL (which at our center was RUX). We further delineated the various combinations which were given with BEL and these therapies were given for the duration of the observation period of at least one year. Our BEL combinations were well-tolerated and appeared to provide salvageable responses/clinical benefit in heavily pre-treated cGVHD patients, particularly notable in the BEL+RUX group. Even patients with highly morbid and long-term manifestations of cGVHD that involved joints/fascia, eye, lung, liver, and/or mouth organ systems appeared to benefit from BEL+RUX.

The limitations of this study are small sample size and retrospective observational study design, which did not allow us to directly compare efficacy of BEL+STD and BEL+RUX regimens. A larger prospective study is needed to increase precision of the rOS and rNRM estimates to provide evidence supporting one of the BEL combinations. Additionally, because this was real-world, assessment of cGVHD improvement tended to be dependent on individual physician global rating of clinical benefit more than NIH consensus criteria, thus limiting formal comparisons of efficacy [10].

Conclusion

In conclusion, our results suggest safety and possible efficacy of BEL+RUX combination therapy in treating refractory cGVHD. Future prospective studies should explore novel agent combinations, particularly those that target differing pathways with the goal of optimizing use of emerging FDA approved agents in steroid-free combinations for refractory cGVHD.