Ruxolitinib Improved Overall Response in the Treatment of Glucocorticoid-Refractory or -Dependent Chronic GVHD

Compared with a control therapy, ruxolitinib increased overall response and improved failure-free survival for patients with glucocorticoid-refractory or -dependent chronic graft-versus-host disease.

Using ruxolitinib (Jakafi) to treat patients with glucocorticoid-refractory or -dependent chronic graft-versus-host disease (GVHD) led to a significantly increased overall response, failure-free survival, and symptom response compared with the control therapy, according to the results of the phase 3 REACH3 study (NCT03112603).

At week 24, overall response was greater in patients who were randomized to the ruxolitinib group (49.7%) than patients in the control group (25.6%; odds ratio [OR], 2.99; 95% CI, 1.86-4.80; risk ratio, 1.93; 95% CI, 1.44-2.60; P < .001). Specifically, 11 patients in the ruxolitinib group and 5 patients in the control group achieved a complete response.

“Our trial showed that among patients with moderate or severe chronic GVHD in whom glucocorticoids produced an inadequate response, ruxolitinib was superior to control therapies, as evidenced by a greater overall response, longer failure-free survival, and greater reduction in symptoms,” wrote the investigators.

A total of 329 patients were enrolled in the open-label REACH3 trial between July 11, 2017, and November 18, 2019 and were randomized 1:1 to receive either ruxolitinib (n = 165) or the control therapy (n = 164). Ruxolitinib was given twice daily at a dose of 10 mg and the control therapy was selected by investigators from a list of 10 commonly used treatments considered as the best available options.

The median patient age was 49 years (range, 12-76) and most patients (61.1%) were male. Investigators reported that 42.9% of enrolled patients had moderate chronic GVHD, while 56.5% of patients had severe chronic GVHD. Also, 71.4% had glucocorticoid-refractory chronic GVHD, and 28.6% had glucocorticoid-dependent disease.

A majority of the control therapy used included extracorporeal photopheresis (34.8%), mycophenolate mofetil (22.2%), and ibrutinib (Calquence; 17.1%), and about half of the patients included received calcineurin inhibitors during the trial period.

Further data indicated that patients in the ruxolitinib group had a longer failure-free survival rate vs patients in the control group, with a median failure-free survival of more than 18.6 months versus 5.7 months (HR, 0.37; 95% CI, 0.27-0.51; P < .001). Moreover, the probability of failure-free survival at 6 months was estimated to be higher with ruxolitinib (74.9%; 95% CI, 67.5%- 80.9%) than with the control therapy (44.5%; 95% CI, 36.5%-52.1%).

Almost all patients were included in the safety analysis, including 165 patients in the ruxolitinib group and 158 patients in the control group who received at least 1 dose of treatment with ruxolitinib. The median duration of exposure to therapy up to day 179 was 25.6 weeks (range, 0.7-25.6) for the ruxolitinib group and 24.0 weeks (range, 0.6-25.6) for the control group.

Any grade adverse effects (AEs) at week 24 occurred in 97.6% of patients in the ruxolitinib group compared with 91.8% of patients in the control group. Grade 3 or higher AEs occurred in 57.0% and 57.6% of each cohort, respectively. Common grade 3 or higher AEs in both the ruxolitinib and control groups included thrombocytopenia (15.2% vs 10.1%), anemia (12.7% vs 7.6%), neutropenia (8.5% vs 3.8%), and pneumonia (8.5% vs 9.5%).

Serious AEs were observed in 55 patients and 58 patients in the ruxolitinib and control groups, respectively. At the point of data cutoff, 31 patients in the ruxolitinib group and 27 patients in the control group had died. The majority of deaths were attributed to complications caused by complications from chronic GVHD, disease treatment, or both (13.3% vs 7.9%) in both groups, respectively.


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