A Bayesian, Phase II Randomized Trial of Extracorporeal Photopheresis (ECP) Plus Steroids Versus Steroids-Alone in Patients with Newly Diagnosed Acute Graft Vs. Host Disease (GVHD): The Addition of ECP Improves Gvhd Response and the Ability to Taper Steroids

Acute Graft-vs.-Host Disease (aGVHD) is the major cause of mortality and morbidity following allogeneic hematopoietic cell transplantation (AHCT). Steroids are the standard initial therapy yet only ~50-60% will achieve a response. Strategies to improve response and minimize steroid-exposure are needed. We report the results of a randomized trial of extracorporeal photopheresis (ECP) plus steroids vs. steroids- alone as initial therapy for aGVHD.

Methods: This was a Phase II, randomized, adaptive Bayesian design. Patients with histologically confirmed aGVHD and who received <72 hours of steroids were randomized to receive 2mg/kg methylprednisolone (MP) with or without ECP. ECP schedule was: days 1-14 (8 sessions), days 15-28 (6 sessions) and days 29-56 (8 sessions). The primary endpoint was treatment success at day 56 post-randomization defined as being alive, in remission, achieving a GVHD response without additional therapy, and on < 1mg/kg of MP at day 28 and < 0.5mg/kg on day 56. Randomization probabilities were based upon the probabilities of response in each arm, stratified by skin-only vs. visceral involvement. The primary endpoint was evaluated using a Bayesian predictive probability approach; i.e., at the end of the study, if the posterior probability that the success rate for either arm was greater than that for the other arm was > 0.80, that arm would be declared the preferred treatment.

Pts and Results: Between 2008 and 2014 eighty-one (81) pts were randomized to ECP + MP (51 pts) or MP-alone (30 pts). Median age was 54 yrs (range, 17-75); 42% had AML/MDS and 62% were male. HCT was performed with myeloablative conditioning in 69% and 63% had an unrelated donor. At enrollment, 90% of patients had aGVHD grade II, 10% had grade III/IV. Organ Involvement was: skin (86%), upper GI (22%), lower GI (22%) and liver (10%). The treatment arms were balanced for all baseline characteristics. The probability that ECP+MP had a higher success rate than MP-alone was 0.815 which exceeded the protocol-specified threshold of 80%, declaring ECP-arm the preferred treatment. The ECP-arm was more beneficial in pts with skin-only aGVHD (72% vs. 57% response rate) whereas visceral-organ involvement response rates were similar (47% vs. 43%). By day 56, 43% of the evaluable patients randomized to the ECP were on physiologic doses of steroids (≤0.1mg/kg) versus 30% for MP-alone arm (p=0.34). To study the immunologic effects of ECP, we examined the recovery of T-cell subsets by multi-parameter flow cytometry in a subset of pts randomized to the ECP+MP and MP-alone arm at enrollment and study conclusion. ECP+MP was associated with more robust recovery of CD4+ and CD8+ cells and significantly higher absolute number of regulatory T-cells (p=0.043), figure 1.

Conclusions: The results of this randomized, phase II trial indicate that the addition of ECP to steroids results in higher GVHD response and facilitates steroid-tapering in pts with newly diagnosed aGVHD. This combination appears most efficacious for pts with skin-involvement. Correlative studies implicate expansion of regulatory T-cells as a possible mechanism of action through which ECP exerts its immunologic effect.

Figure 1.

View largeDownload slide

Patient Characteristics and Outcomes

Figure 1.

View largeDownload slide

Patient Characteristics and Outcomes

Close modal

Figure 2.

View largeDownload slide

Mean Absolute Regulatory T-cell Count in subset of patients randomized to ECP+MP (red) vs. MP-alone (green) p-value= 0.043.

Figure 2.

View largeDownload slide

Mean Absolute Regulatory T-cell Count in subset of patients randomized to ECP+MP (red) vs. MP-alone (green) p-value= 0.043.

Close modal