Lee Evaluates Chronic GVHD Therapies for Depth and Durability of Response

Catherine J. Lee, MD, MS

Associate Professor

Clinical Research Division

Fred Hutchinson Cancer Center

Seattle, WA

CASE SUMMARY

• A 48-year-old man underwent myeloablative conditioning and a matched unrelated donor hematopoietic cell transplant for acute myeloid leukemia (AML) with tacrolimus plus methotrexate as graft-vs-host disease (GVHD) prophylaxis.
• The donor was a 50-year-old woman with 3 children who is seropositive for cytomegalovirus.
• After day 22, acute GVHD of the skin emerged and was successfully treated with slow steroid taper.
• Results from bone marrow biopsy performed at 3 months post transplant showed AML in complete remission.
• He returned at 6 months post transplant with normal blood counts. He now presented with skin changes with hyperpigmentation, with approximately half of each arm showing lichen planus; superficial sclerotic features (able to pinch the skin) on lower trunk and lower extremities; 18% body surface area involved; and no decrease in forced expiratory volume in first second of expiration/diffusing capacity of the lung for carbon monoxide on pulmonary function test results.

Peers & Perspectives in Oncology: What data support the use of ibrutinib (Imbruvica) plus prednisone for chronic GVHD (cGVHD)?

LEE: The FDA approval was based on [findings from] a phase 1b/2 trial [NCT02195869]. This was a small trial [with] only 42 patients with steroid-dependent/ refractory cGVHD. This was a highly select population, meaning that all these patients had to have inflammatory features of cGVHD. They had either more than 25% of an inflammatory skin rash or they had to have oral involvement, [either of] the inflammatory features. There were only 3 patients with liver involvement at baseline and 2 patients with lung [involvement], and there were no patients with joint involvement. There were 18 patients who had sclerosis at baseline in the study. Patients had to have [GVHD that failed to respond] to 3 or fewer prior lines of therapy. The median number of lines failed in this cohort was 1.

In terms of efficacy, the primary end point was overall response rate [ORR] at 24 weeks.1,2 The ORR was 67%, defined as those having either a complete response [CR] or a partial response [PR]. The best ORR at median 26 months’ follow-up, which [meant the] patient could have had progression after that, was 69%. Sixty-nine percent retained their response for 20 weeks or longer.¹

Of the 18 patients who had sclerosis at baseline, 11 patients responded and 64% of patients were able to taper their steroids to less than 0.15 mg/kg per day.² The median dose on enrollment was about 0.31 mg/kg per day, so about 64% of the patients were able to decrease their steroids by at least 50%. There was a high response rate in patients who had GVHD of the skin and the mouth. There was an 88% overall response in the gastrointestinal [GI] tract. Among the 18 patients who had sclerosis, 11 of them responded and 39% of them had a complete response whereas 22% had a partial response. Common adverse events [AEs] in this trial included fatigue, diarrhea, muscle spasms, and bruising. There were no major hemorrhagic events, and atrial fibrillation was infrequent.

Ibrutinib was tested in the up-front setting…. It was ibrutinib plus prednisone vs placebo plus prednisone for newly diagnosed cGVHD. This was a phase 3 study. It enrolled patients who were 12 years or older. The primary end point was ORR at 48 weeks. It didn’t meet the primary end point nor did it meet any of the secondary end points, so it was a failed [treatment] trial; ibrutinib does not work in the first-line setting. The ORR in the first-line setting [41%] was even less than reported in the steroid-refractory setting.³

What led to the recent approval of axatilimab (Niktimvo) in cGVHD?

AGAVE-201 [NCT04710576] was the randomized phase 2 trial that was presented at the [65th] American Society of Hematology Annual Meeting & Exposition in December [2023 in San Diego, California], which confirmed the FDA approval of axatilimab for relapsed/refractory cGVHD after 2 prior lines of treatment. Inclusion criteria included [patients] 2 years or [older whose treatment] failed with 2 or more prior lines of therapy. They had to meet the definition of active GVHD per the 2014 NIH [National Institutes of Health] consensus criteria. They could have concomitant use of corticosteroids, calcineurin inhibitors, or mTOR inhibitors, and no additional systemic cGVHD therapy was allowed after enrollment; otherwise, it was considered a treatment failure. Patients were randomly assigned to 3 different doses, either 0.3 mg/kg every 2 weeks, 1 mg/kg every 2 weeks, or 3 mg/kg every 4 weeks.

The primary end point was ORR at 24 weeks or 6 months. This is the most common primary end point in cGVHD trials. The efficacy was met if the lower bound of the 95% CI exceeded 30%. Secondary and exploratory end points included traditional end points, including meaningful improvement using the modified Lee Symptom Scale looking at organ-specific response rates, duration of response, failure-free survival [FFS], overall survival [OS], and safety events. This was a global trial; a total of 241 patients were enrolled.

The primary end point was met in all groups, and the highest ORR of 74% was in the 0.3-mg dose group.⁴ The overall response was a little bit lower in the 1-mg dose group [67%] and the 3-mg dose group [50%].

There were complete responses in every single organ, which is encouraging because these are patients who have very refractory disease and [whose GVHD failed to respond to] either ibrutinib, ruxolitinib [Jakafi], or belumosudil [Rezurock] but were still getting CRs and PRs across all organs. The greatest percentage of CRs was in the GI tract, the esophagus, as well as the mouth. PRs were seen in some of these highly morbid forms of cGVHD, including the lungs, the joint and fascia, and the eyes, so this is exciting.

The most common AEs were dose-dependent transient laboratory abnormalities expected from the CSF1R receptor blockage. I enrolled a few of these patients in the trial, and I would see elevations in AST [aspartate aminotransferase], ALT [alanine aminotransferase], and alkaline phosphatase, and these are expected AEs [because of] the off-target actions of axatilimab on the liver. AEs leading to discontinuation of axatilimab occurred in 6% of patients in the 0.3-mg dose group, 22% in the 1-mg dose group, and 18% in the 3-mg dose group. The 0.3-mg dose was the dose that was shown to be most effective and safe from this study. One of the other common AEs that was seen with axatilimab was periorbital edema. This usually occurred by the third cycle, second and third cycle, and can resolve with time.

Could you describe the trial findings supporting the use of ruxolitinib?

The pivotal trial for ruxolitinib is the phase 3 REACH3 trial [NCT03112603]. This was a randomized trial of ruxolitinib vs best available therapy [BAT] in patients with steroid-refractory cGVHD per the NIH consensus [whose treatment] failed [with] 1 line of therapy, essentially steroids, for moderate to severe cGVHD and who were 12 years or older. They were eligible for enrollment and were randomly assigned to receive ruxolitinib 10 mg twice daily, which is the starting dose, or BAT, which included commonly used agents. Extracorporeal photopheresis and mycophenolate mofetil were the 2 most common BATs that were used in this cohort. The primary end point was ORR at 6 months, with key secondary end points including failure-free survival and the modified Lee Symptom [Scale] score. Other secondary end points included best overall response, duration of response [DOR], OS, nonrelapse mortality, steroid reduction, and relapse.

For the baseline characteristics of the patients, the patient groups were well-balanced between the ruxolitinib and the control group.⁵ The median age of the patients was around 49 years. Overall, 43% of the patients had moderate cGVHD, 56% of the patients had acute GVHD, and 17% of patients received ibrutinib in this study.

With this study, there was a primary analysis for the primary end point and there was an extended follow-up. The primary analysis was the ORR at 24 weeks. After the primary analysis, patients were allowed to cross over from the BAT to the ruxolitinib arm. Those patients were assessed 6 months after crossover to evaluate their response as well. These data are part of the extended follow-up from the 3-year final analysis. The failure-free probability at 12 months with ruxolitinib is 64% vs 29% in the best available therapy arm.6 The median FFS was 38 months in the ruxolitinib arm vs 5.7 months in the BAT arm. FFS was significantly superior in the ruxolitinib arm [HR, 0.361; 95% CI, 0.268-0.485].

In terms of DOR, the median DOR was 6.4 months in the BAT arm and was not estimable or not yet reached in the ruxolitinib arm. The probability of duration of response at 12 months was 70% in the ruxolitinib arm vs 40% in the BAT arm, again showing that ruxolitinib was superior to BAT. Median OS was not reached between BAT or ruxolitinib, so there was no difference in the risk of death between the arms [HR, 0.85; 95% CI, 0.54-1.33].

[Regarding] the overall response for the primary analysis and the crossover…49.7% was the ORR at 24 weeks for the primary analysis in the patients who were originally randomly assigned to the ruxolitinib arm. The ORR in the primary analysis to those randomly assigned to the best available therapy was 25.6%. When you look at the extended follow-up in the final analysis, which is the crossover analysis, 50% of patients had a response to ruxolitinib even after the crossover.

What’s significant here is that the ORR, even after the crossover, was very similar to [ORR in] those patients who were randomly assigned initially to ruxolitinib. It was the same thing with the best ORR [at any point]. The best ORR of those who went over to crossover analysis was similar or even a little bit higher [81.4%] than in those who were randomly assigned to ruxolitinib in the primary analysis [76.4%]. Compared with BAT, the overall response rate was superior.

In terms of overall response, 5.7% of patients had a CR [at week 24] and 7.1% of patients had a [best response of] CR. Most of the responses here are PRs to ruxolitinib [Figure⁶].

In terms of safety, this is from the 3-year final analysis and takes into account the patients who crossed over. The median duration of exposure to ruxolitinib is longer than in those who were exposed to best available therapy. The median duration of exposure to ruxolitinib was 52.9 weeks vs 24.1 weeks in the BAT arm. Those who crossed over from BAT to ruxolitinib had a total exposure of 109.9 weeks. What’s important about this is that patients in the ruxolitinib-treated arm had more AEs, and that’s likely because of the longer duration of exposure to ruxolitinib. The main cause of on-treatment death in the ruxolitinib group was death due to cGVHD. In those who crossed over from BAT to ruxolitinib, cGVHD was a cause of death but there were other causes of death including Aspergillus infection, cardiac arrest, hemorrhage, intracranial hemorrhage, pulmonary sepsis, and systemic mycosis.

Common AEs of ruxolitinib are anemia and possibly neutropenia and thrombocytopenia. Anemia is the most common AE that we see in the cGVHD population and more so in the acute GVHD setting rather than the cGVHD setting. Often, we have to dose reduce or hold therapy to allow for recovery.

What role does belumosudil serve in cGVHD?

Belumosudil is the inhibitor of the ROCK pathway. The FDA approval of belumosudil came from the phase 2 ROCKstar trial [NCT03640481] of patients who were 18 [years or] older who had active cGVHD that failed to respond to at least 2 lines of prior therapy. They were randomly assigned to either belumosudil 200 mg once a day or 200 mg twice a day.

The efficacy end point was met if the lower bound of the 95% CI was greater than 30%. The primary end point was ORR at 24 weeks. The primary end point was met for both arms. The lower end of the 95% CI was greater than 30% for both arms, with an ORR of 74% in the once-a-day cohort and 77% in the twice-a-day cohort.⁷ The once-daily dose was selected for the FDA approval, but we use belumosudil 200 mg twice a day when patients are on a proton pump inhibitor.

These were patients with very refractory cGVHD [whose disease] had failed a median of 3 lines of prior therapy. The median time from cGVHD to diagnosis was 29 months. Sixty-seven percent of patients had NIH-severe cGVHD, [52%] of patients had 4 or more organs involved at the time of enrollment, 34% of patients had prior ibrutinib, and 29% of patients had prior ruxolitinib.

Among all the different subcategories by the characteristics that they looked at, there was still a high ORR in the patients who received belumosudil. So 60% of the patients felt better based [on at least 7-point decrease] on the Lee Symptom Scale, 65% of patients were able to taper steroids, and among the 21% of patients who were able to completely get off steroids, 22% also were able to stop their calcineurin inhibitors, 20% were able to stop their sirolimus, and 21% stopped their mycophenolate mofetil. Those who received prior ibrutinib or ruxolitinib, particularly ibrutinib, had a very good response to belumosudil. Even those who had cGVHD for more than 29 months, greater than the 50th percentile, still had a pretty good response to belumosudil [68% ORR vs 83% in those with shorter duration of cGVHD].

If you look at response by organ, CRs were still seen in all organs. You still have a number of PRs, but CRs were achieved. A high percentage of CRs were achieved in the lower GI tract [61%], the upper GI tract [48%], the mouth [44%], and the esophagus [45%], as well as in the liver [31%]. You can see CRs in the lung [13%] as well as in the joints and fascia [20%]. These were good responses in a highly refractory population.

When you look at the durability of response to belu-mosudil or the FFS and OS, the 6-month FFS rate is around 75% and the 1-year FFS rate is 56%. The 2-year OS rate is very high at 89%.

In terms of safety, belumosudil was well tolerated. It was associated with low discontinuation rates secondary to AEs. The most common AEs of any grade were fatigue, diarrhea, nausea, cough, and upper respiratory tract infection. [Drug-related] reactivation of cytomegalovirus was not observed. In practice, I will often have my patients say that they feel a little nauseous. We just tell them to take belumosudil on a full stomach or to take an antiemetic prior to belumosudil, and this makes the drug much more tolerable.

_{DISCLOSURES: Lee previously reported advisory board participation for Bristol Myers Squibb, Incyte, Kadmon, Kite Pharma, and Sanofi; research funding from Incyte; steering board committee participation for Incyte; speakers’ panel participation for Kite Pharma; and consulting for Fresenius Kabi, Mallinckrodt Pharmaceuticals, and Sanofi.}

_REFERENCES

_{1. Miklos D, Cutler CS, Arora M, et al. Ibrutinib for chronic graft-versus-host disease after failure of prior therapy. Blood. 2017;130(21):2243-2250. doi:10.1182/blood-2017-07-793786}

_{2. Waller EK, Miklos D, Cutler C, et al. Ibrutinib for chronic graft-versus-host disease after failure of prior therapy: one-year update of a phase 1b/2 study. Biol Blood Marrow Transplant. 2019;25(10):2002-2007. doi:10.1016/j.bbmt.2019.06.023}

_{3. Miklos DB, Abu Zaid M, Cooney JP, et al. Ibrutinib for first-line treatment of chronic graft-versus-host disease: results from the randomized phase III iNTEGRATE study. J Clin Oncol. 2023;41(10):1876-1887. doi:10.1200/JCO.22.00509}

_{4. Wolff D, Cutler C, Lee SJ, et al; AGAVE-201 Investigators. Axatilimab in recurrent or refractory chronic graft-versus-host disease. N Engl J Med. 2024;391(11):1002-1014. doi:10.1056/NEJMoa2401537}

_{5. Zeiser R, Polverelli N, Ram R, et al; REACH3 Investigators. Ruxolitinib for glucocorticoid-refractory chronic graft-versus-host disease. N Engl J Med. 2021;385(3):228-238. doi:10.1056/NEJMoa2033122}

_{6. Zeiser R, Russo D, Ram R, et al. Ruxolitinib in patients with chronic graft-versus-host disease: three-year final analysis of efficacy and safety from the phase III REACH3 study. Blood. 2023;142(suppl 1):654. doi:10.1182/blood-2023-172590}

_{7. Cutler C, Lee SJ, Arai S, et al. Belumosudil for chronic graft-versus-host disease after 2 or more prior lines of therapy: the ROCKstar study. Blood. 2021;138(22):2278-2289. doi:10.1182/blood.2021012021}