Rationale: cGVHD is a major complication of allogeneic hematopoietic stem cell transplantation (HSCT). Although a calcineurin inhibitor plus corticosteroids has remained the standard of care (SoC) for moderate/severe cGVHD, efficacy is limited. ECP, in which autologous buffy coat leukocytes are exposed to ultraviolet A light after treatment with 8-methyoxypsoralen (UVADEX) and then reinfused, yields excellent responses with reliable safety in T-cell-mediated diseases, including corticosteroid-refractory cGVHD. Response criteria for cGVHD assessment were modified in 2005 with the advent of the NIH Consensus Criteria. We conducted a randomized phase II study evaluating the addition of ECP to SoC in patients (pts) with new-onset NIH moderate to severe cGVHD.

Methods: Sixty pts ≥18 years with new-onset (≤3 years of HSCT) moderate (intent to treat [ITT] population, n=28) or severe (ITT population, n=25) cGVHD per NIH Consensus Criteria were randomized 1:1 to 26 wks of SoC with corticosteroids and cyclosporine A (CsA) or tacrolimus (Tac) vs SoC + ECP between 2011 and 2015. Corticosteroids were started at 1.0 mg/kg prednisone or equivalent daily, not to exceed 2.0 mg/kg. Doses were tapered to 0.5 mg/kg/day by wk 8 (±1 wk), 0.25 mg/kg/day by wk 16 (±1 wk), and 0.125 mg/kg/day by wk 24 (±1 wk); this dose was maintained until wk 28. CsA/Tac dosing was consistent with local institutional practice. Pts received ECP as follows: wk 1, 3 treatments; wks 2-10, 2 treatments/wk; wks 11-18, 2 treatments/wk q2w; and wks 19-26, 2 treatments/wk q4w. Pts were assessed every 2 or 4 wks for 28 wks; optional long-term follow-up (LTFU) was conducted in some pts. The primary endpoint was overall response rate (ORR), defined as clinically-assessed complete response (CR) or partial response (PR) by NIH Consensus Criteria at wk 28. Response was assessed by investigators and non-independently by blinded assessors. Secondary endpoints included total skin score (TSS) changes, investigator-assessed cGVHD, quality of life (QOL), cumulative corticosteroid dose, and safety. Failure-free survival (FFS; lack of GVHD recurrence, new immunosuppressive therapy, relapse, or death) was evaluated in LFTU. The study design was prospective and randomized; it had no prespecified hypothesis to test and was not powered to do so.

Results: Of the 60 pts, 53 (SoC + ECP, 29; SoC, 24) were eligible for efficacy. ITT analysis showed the ORR was 74.1% in the SoC + ECP arm and 60.9% in the SoC arm using the 2015 NIH Consensus Criteria for cGVHD at wk 28 by blinded assessors. Investigator-assessed ORR was 66.7% and 56.0% in the SoC + ECP and SoC arms, respectively, indicating moderate agreement with the NIH criteria. ORRs were not remarkably different across the moderate (SoC + ECP vs SoC: 75% vs 50%) and severe (SoC + ECP vs SoC: 72.7% vs. 69.2%) subgroups as measured by blinded assessor. In contrast, investigator-measured ORRs in the moderate and severe subgroups were SoC + ECP vs SoC: 66.7% vs 60%, and SoC + ECP vs SoC: 72.7% vs 40%, respectively. Mean change in TSS from baseline to wk 28 and mean cumulative corticosteroid dose up to wk 28 were similar between arms. Among the 28 LTFU pts (SoC + ECP, n=16; SoC, n=12), median FFS was 11.9 and 7.8 mos in the SoC + ECP and SoC arms, respectively. All 60 pts were eligible for safety. Overall, 96.6% and 90.3% of pts in the SoC + ECP and SoC arms, respectively, experienced treatment-emergent adverse events (TEAEs). The number of total events was higher in the SoC arm. The most frequently reported TEAEs in SoC + ECP pts were hypertension, cough, dyspnea, and fatigue; the most frequently reported TEAE in SoC pts was back pain. No TEAEs were considered to be related to UVADEX or the ECP instrument. Overall, 17 pts experienced 35 serious adverse events (SAEs); events in 13 pts were considered related to SoC treatment. No SAEs were considered related to the ECP instrument or UVADEX. Six pts from each arm were withdrawn from the study due to SAEs. Four deaths occurred in the SoC + ECP arm; none were considered to be related to the instrument, ECP therapy, or UVADEX. Of those pts, 2 had moderate and 2 had severe cGVHD.

Conclusion: These results indicate that ECP with UVADEX is safe and well-tolerated in pts with new-onset NIH-graded moderate to severe cGVHD. However, a remarkable short-term treatment benefit could not be confirmed using the current descriptive design based on a small sample size. Further evaluation of ECP in more pts with longer follow-up is warranted.


Jagasia: Therakos: Consultancy, Research Funding; Mallinckrodt: Consultancy; Janssen: Consultancy, Research Funding. Greinix: Bristol Myers Squibb: Consultancy, Speakers Bureau; Sanofi: Consultancy, Speakers Bureau; Mallinckrodt: Consultancy, Speakers Bureau; Novartis: Consultancy, Speakers Bureau; Amgen: Consultancy, Speakers Bureau; Celgene: Consultancy, Speakers Bureau. Scheid: Sanofi: Consultancy, Honoraria; Novartis: Consultancy, Honoraria; BMS: Consultancy, Honoraria; Amgen: Consultancy, Honoraria; Janssen: Consultancy, Honoraria; Celgene: Consultancy, Honoraria; Takeda: Consultancy, Honoraria. Socié: Alexion Pharmaceuticals, Inc.: Consultancy. Chin: Mallinckrodt: Employment. Boodee: Mallinckrodt: Employment. Mitri: Mallinckrodt: Employment.

Author notes


Asterisk with author names denotes non-ASH members.