Topical ruxolitinib for chronic cutaneous graft-versus-host disease: a single-centre, randomised, double-blind, intra-patient-controlled phase 2 trial.

backgroundOral ruxolitinib, a Janus kinase (JAK) 1/2 inhibitor, is approved for both acute and chronic graft-versus-host disease (cGVHD), whereas the topical formulation is approved for atopic dermatitis and vitiligo. Topical corticosteroids remain the mainstay for cutaneous cGVHD but their use is limited by adverse effects. We aimed to evaluate the clinical activity and safety of ruxolitinib 1·5% cream for epidermal or superficially sclerotic cutaneous cGVHD.

methodsIn this single-centre, randomised, double-blind, intra-patient-controlled phase 2 trial at Memorial Sloan Kettering Cancer Center (NY, USA), patients aged 12 years or older with history of allogenic haematopoietic stem cell transplantation, with clinically confirmed, histologically confirmed, or both, cutaneous cGVHD involving at least 2% body surface area were randomly assigned to the body side receiving topical ruxolitinib 1·5% cream twice daily for 28 days; placebo vehicle was applied to contralateral lesional skin as a control. The primary endpoint was the absolute day-28 between-side difference in affected lesional body surface area (BSA), recorded as percentage of total BSA on each assigned side and analysed in evaluable patients; baseline between-side BSA differences were controlled for with adjusted mixed-effects regression. Only patients who completed the day 28 visit were included in the primary endpoint analysis; all patients who received any study treatment were included in the safety analysis. The trial is closed to enrolment and registered with ClinicalTrials.gov, NCT03954236. The trial is complete.

findingsBetween June 28, 2019, and Sept 8, 2022, 24 patients were randomly assigned, with a median follow-up of 128 days (IQR 90-280). Median age was 47·5 years (IQR 32·5-67·5); 13 (54%) were female and 11 (46%) were male. At day 28, 23 patients were evaluable (one patient did not start treatment after screening visit due to intensive care unit admission). Mean affected lesional body surface area changed from 14·3% (SD 7·9) at baseline to 6·2% (6·7) with ruxolitinib and from 14·5% (8·0) to 10·4% (9·1) with placebo vehicle (difference -4·2 percentage points, 95% CI -6·8 to -1·5; p=0·0030); after adjustment for baseline between-side body surface are difference, the difference was -4·0 percentage points (-5·5 to -2·4; p<0·0010). Grade 3-5 treatment emergent adverse events were CNS leukaemia, sinusitis, and lung infection (one [4%] each); no treatment-related deaths occurred.

interpretationTopical ruxolitinib showed significant clinical activity and favourable harms profile in active cutaneous cGVHD. Ruxolitinib cream could represent a steroid-sparing, skin-directed alternative or second-line therapy, expanding treatment options for patients with cutaneous cGVHD.

fundingIncyte.