Current status, development and challenges in the treatment of chronic graft-versus-host disease

Haploidentical transplantation represented by the “Beijing Protocol” is leading to rapid growth in the number of allogeneic hematopoietic stem cell transplantations (allo-HSCT) [1], 2]. With the development of transplant technique, many early transplant complications such as acute graft-vs.-host disease (GVHD) have been controlled effectively [3], [4], [5], [6] and a large number of patients can achieve long-term post-transplant survival. However, the occurrence of chronic GVHD (cGVHD) increases accompany with the extension of survival, which can cause disability and mortality and is critical for long-term survivors after allo-HSCT.

According to the international guidelines, first-line treatment of cGVHD is prednisone with the initial dose of 1 mg/kg/day [7]. Particularly, if prednisone is less than 1 mg/kg/day at the onset of cGVHD, the dose can be increased. However, the overall response rate (ORR) for prednisone was merely 40 %–50 %. So how to improve the ORR of first-line treatment is important.

Several randomized controlled trials (RCTs) had identified that whether prednisone combine with other immunosuppressants (i.e., “prednisone+X”) could improve the ORR of first-line treatment for cGVHD. For example, in the study of Koc et al. [8], extensive cGVHD patients were randomized to receive prednisone plus cyclosporine and prednisone alone, however, cyclosporine was associated with a poorer disease-free survival (DFS, 53 % vs. 59 %, p=0.03) and it only reduced the risk of avascular necrosis. Thus, cyclosporine was only recommended for the cGVHD patients who are at high risk for glucocorticoid-usage-related complications or those who is already on corticosteroid treatment when cGVHD occurrence. Besides of cyclosporine, Sullivan et al. [9]. reported that prednisone plus azathioprine did not improve ORR, and it was associated with a poorer overall survival (47 % vs. 61 %, p=0.03) and a higher non-relapse mortality (NRM) rate (52 % vs. 37 %, p<0.0001) compared with prednisone alone in extensive cGVHD patients. Martin et al. [10] reported that prednisone plus mycophenolate mofetil showed a higher rate of treatment failure (70 % vs. 64 %, p=0.03) and had a trend of poorer OS and a poorer relapse-free survival compared with prednisone alone. In addition, prednisone plus novel anti-cGVHD agents, such as ibrutinib or belumosudil (closely early at 2025), could also not further improve the ORR and survival compared with prednisone alone. Thus far, no solid evidence supported to use “prednisone+X” as first-line treatment for cGVHD.

Several reasons may cause the failure of combination therapy. Firstly, cGVHD patients often show multi-organ involvement but combined therapy may only effective for particular organ which could not improve the overall ORR. Secondly, fibrosis was the common characteristics of moderate to severe cGVHD but prednisone plus other immunosuppressants might not effectively reverse the organ fibrosis. Thirdly, several drugs (such as azathioprine and ibrutinib) could induce myelosuppression and increase the risk of infection and NRM. In future, strategy of “prednisone+X” could be further performed in cGVHD with specific organ involvement, such as obliterative bronchiolitis or scleroderma.

Because of the unsatisfactory efficacy of first-line treatment, the second-line treatment was important. Ruxolitinib, an oral selective JAK1/JAK2 inhibitor, which inhibits pro-inflammatory signaling and shifts T-cell subsets to an anti-inflammatory phenotype, has been recommended as the first therapeutic choice for steroid-resistant/dependent cGVHD by the European Society for Blood and Marrow Transplantation [7]. In its successful RCT (REACH3 study), ruxolitinib showed a better ORR at week 24 (49.7 % vs. 25.6 %), a better best overall response (BOR, 76.4 % vs. 60.4 %), a better failure-free survival (>18.6 months vs. 5.7 months) and a better duration of response (not-reached vs. 6.24 months) compared with those of best available treatment groups [11]. However, the satisfactory response depends on the sufficient doses (i.e., 10 mg twice daily) and duration of therapy (≥6 months), but the side effect of ruxolitinib, particularly severe myelosuppression and viral activation (e.g., cytomegalovirus), may lead to dose reduction or stop the treatment too early which will negatively impact the efficacy of ruxolitinib. In addition, the response rate for particularly organ such as lung cGVHD was still lower than 10 % after ruxolitinib treatment [11].

Rho-associated coiled-coil containing kinase 2 (ROCK2) inhibitor can restore immune homeostasis by downregulation of both T helper 17 (Th17) and T follicular helper (Tfh) cells and upregulation of Regulatory T (Treg) cells. In addition, it can disrupt fibrotic progression by blocking myocardin-related transcription factor-mediated transcription of profibrotic markers, attenuating myofibroblast differentiation and collagen deposition. Thus, ROCK2 inhibitor has become one of the most important second-line therapies for cGVHD particularly for those with organ fibrosis. In the ROCKstar study which enrolled patients with cGVHD who required additional treatments after failure of previous 2–5 lines of treatments, the overall ORR was 76 % after belumosudil treatment with a median response time of 5 weeks. Particularly, the ORR could be as high as 68 % for those who had previously received ruxolitinib [12]. Because belumosudil has weaker inhibition on cytotoxic T lymphocyte, it leads to a lower risk of infection and relapse. It also shows a weak myelosuppression effect. These characteristics help to use belumosudil persistently and prevent loss of response, and it is also suitable to the patients with a higher risk of relapse. On the contrary, for the cGVHD patients in the stage of active inflammation, belumosudil may not strongly enough to control the disease and combined it with other T cell inhibitors (e.g., ruxolitinib) is reasonable.

Rovadicitinib, which is both the JAK and ROCK inhibitor, has showed excited efficacy in its phase 1b/2a study. The BOR in total cohort was 90.9 %, which is 96.6 % and 80 %, respectively for the 10 mg and 15 mg twice daily group. The responses were rapid, with 72.4 % and 66.7 % of responders achieving a response at 4 weeks, respectively. Particularly, all six patients with lung cGVHD achieved partial response (PR) with lung responses with a >10 % increase in a forced expiratory volume in the first second (FEV1) [13]. The phase 3 randomized study of rovadicitinib for cGVHD patients who required other treatments after failure of previous 2–5 lines of treatments is ongoing in China (NCT06682169).

Colony stimulating factor-1 receptor (CSF-1R)-driven macrophage signaling plays a critical role in inflammation and fibrosis processes. By blocking the CSF-1R pathway, axatilimab achieved a high response rate (74 %) in patients with recurrent or refractory cGVHD who failed prior second-line therapies in the AGAVE-201 trial, representing an important therapeutic method for cGVHD [14]. Pimicotinib, an oral, highly selective, and potent small molecule inhibitor of CSF-1R, also shows potential in the treatment of cGVHD. In the preliminary disclosed results of its phase 1b/2a study, the ORR has already reached 64 % although there were still about 40 % of patients in the recommended 20 mg cohort have not reached the week 24 visit. In addition, pimicotinib showed a favorable safety in this study, and the slightly elevations of the hepatic enzyme mostly because it inhibits Kupffer cell function and reduces the clearance of these enzymes [15]. The anticipated final results of pimicotinib will be announced in December 2026.

In summary, the treatments for cGVHD are still challenge, including the low ORR for first-line treatment and the low complete response rate (<10 %) for second-line treatment. This suggests that even using different second-line treatment, patients would have inevitable impairments of organ functions when they show resistant to steroid treatment. In future, combined therapy based on the pathogenesis mechanism, particularly with the novel anti-fibrosis drugs (e.g., CSF-1R inhibitor), may help to further improve the outcomes of cGVHD patients (Figure 1).

Figure 1:

cGVHD treatment landscape: from first-line to second-line. ORR=overall response rate, DFS=disease-free survival, OS=overall survival, NRM=non-relapse mortality, RFS=relapse-free survival, JAK=janus activated kinase, ROCK=rho-associated kinase, CSF-1R=colony-stimulating factor 1 receptor, BOR=best overall response, FEV1=forced expiratory volume in 1 s, cGVHD=chronic graft-vs.-host disease.