Retrospective, Multi-Center Analysis of Ruxolitinib Tapering after Successful Treatment of Steroid Refractory Chronic Graft Versus Host Disease

Introduction:

Ruxolitinib (RUX) has been approved for 2 ^nd line therapy or beyond for steroid-refractory chronic graft-versus-host disease (cGVHD) based on the REACH3 study demonstrating superior efficacy of RUX with respect to greater overall response and failure-free survival (FFS) vs. best available therapies. Despite this, there is no real-world data to evaluate whether the tapering speed used in this study is clinically relevant, or results in successful RUX taper without flare. Accordingly, the present retrospective, multi-center study provides data on RUX tapering in patients who had been successfully treated with RUX for steroid-refractory cGVHD.

Patients and Methods:

109 patients treated with RUX to treat chronic GVHD at 4 Canadian transplant centers (Toronto, Calgary, Saskatoon, Quebec City) were enrolled, with patient characteristics in Table 1. The overall response was defined as complete (CR) or partial response (PR) as defined by the 2014 NIH consensus criteria. RUX tapering was started at the treating physician's discretion, while maintaining CR, PR or no change (NC) after complete tapering of steroid. When cGVHD flare is noted during gradual RUX tapering, RUX dose was escalated back to the previous dose level based on physician's discretion. FFS after RUX tapering (R-FFS) was calculated from RUX taper day 1 until either GVHD flare or death from any cause. For risk factor analysis of GVHD flare-up, Cox's proportional hazard model was applied for R-FFS at 24 months.

Results:

Of 109 patients enrolled, 70 attempted RUX tapering at median 12.9 months (95% CI [8.9-16.4 months]) following start of RUX therapy after achievement of CR (n=29), PR (n=28) or NC (n=22) with complete discontinuation of steroid in 16 (23%). Daily RUX dose at the time of RUX taper was 30mg (n=4, 5%), 20mg (n=59, 79%), 15mg (n=4, 5%), and 10mg (n=8, 11%).

Time from start of RUX taper to daily RUX dose of 20mg, 15mg, 10mg and complete discontinuation was 8 ± 6 days, 81 ± 23 days, 185 ± 38 days, and 204 ± 30 days, respectively. In terms of RUX tapering speed, RUX dose was reduced, on average, by 5mg/day every 41 days. With a median follow up duration of 12 months in those who attempted RUX taper, 22 (31%) experienced a cGVHD flare at a median 19.4 months after RUX tapering attempt, while 41 pts (59%) were able to wean RUX completely and 7 pts (13%) remained on tapering schedule. Median time to successful taper was 6.8 months (range 4.6-73.7 months).

The median R-FFS for RUX taper (95% confidence interval [CI]) to flare was 4.2 years (2.1 - 6.3), while estimated R-FFS at 1 year was 89.1% (0.785 - 0.947%) and at 3 years was 56.0% (39.4-69.7%). Univariate analysis showed that duration of RUX ≥ 434 days prior to taper was associated with lower risk of R-FFS (Hazard ratio [HR] 0.323, 95% C.I. [0.095 - 1.093]), while non-matched related donor was associated with higher risk of R-FFS with HR of 2.826 (1.011 - 7.902). However, multivariate analysis could not confirm these findings.

There were 9 deaths during RUX taper, including sepsis (n=3), cardiac arrest (n=1), COVID pneumonia (n=1), leukoencephalopathy (n=1) and undetermined causes (n=3), which were not directly related to RUX taper.

Conclusions:

Our study provides real-world data that (1) RUX can be successfully tapered with no associated GVHD flare at 12 months median follow up in a majority of this patient population, (2) faster tapering (by 5mg every 6 weeks) demonstrates comparable efficacy to 5mg every 8 weeks (as per REACH3) and (3) factors such as GVHD severity, number of GVHD organ involvement or CR achievement prior to taper were not associated with R-FFS.