Response and Survival Following Second-Line Therapy in 113 Patients with Steroid-Refractory Acute Graft-Versus-Host Disease

Background: Steroid-refractory acute graft-versus-host disease (SR-aGVHD) remains the major, non-relapse obstacle to successful hematopoietic cell transplantation (HCT). Evolving donor and cell sources, conditioning regimens and GVHD prophylaxis are altering the field of HCT. We evaluated response rates following second-line therapy for SR-aGVHD in this modern era.

Methods: We retrospectively analyzed day 28 response rates and clinical outcomes of 113 HCT patients (median age 39 years, range 1-75) who received second-line therapy for SR-aGVHD from 1998-2012 at the University of Minnesota.

Results: Patients received grafts from 43 related (38 matched, 5 mismatched), 31 unrelated (23 HLA 6-8/8, 8 HLA ²5/8) and 39 cord blood donors (23 HLA 6-8/8, 16 HLA ²5/8); 76 patients (67.3%) received myeloablative regimens. Second-line therapy consisted of ATG (44.2%), MMF (24.8%), steroid boost (22.1%) or other (8.8%). Patients receiving ATG had higher-risk GVHD Risk Scores (MacMillan et al, Br J Haem, 2012) at second-line initiation compared with patients receiving MMF or steroids.  At day 28, overall response (complete response [CR] / partial response [PR]) was observed in 32 patients (28%).  All therapies had similar responses.  In multivariate analysis, factors associated with day 28 CR/PR included GVHD Risk Score and days from onset of aGVHD to second-line therapy (Table); Six-month non-relapse mortality (NRM) was lower among patients receiving steroid boosts (24% [7-41%, 95% CI]) vs. ATG (64% [48-90%, 95% CI] with no differences vs. MMF or other therapies (P=0.01, Figure 1). In multivariate analysis, use of steroids vs. ATG was associated with reduced 6-month NRM (Table).  Overall survival (OS) at 6 months was increased among day 28 responders (68% [48-81, 95% CI]) vs. non-responders (NR) (45% [33-57, 95% CI]) (P=0.03, Figure 2).  Donor and cell source, conditioning regimen and age did not impact response, NRM or OS.

Conclusions: Effective therapy for SR-aGHVD remains inadequate.  Day 28 CR/PR was similar among second-line therapies. ATG recipients experienced higher rates of NRM but also exhibited higher-risk GVHD by Risk Score.  Identification of high-risk GVHD at initial onset using the GVHD Risk Score may identify patients likely to fail upfront steroids in need of additional or alternative agents.

Table: Factors associated with day 28 CR/PR and 6-month NRM

Factor	N	OR of CR/PR (95% CI)	P	RR of NRM (95% CI)	P
Second Line Rx
ATG*	50	1.0		1.0
MMF	28	2.1 (0.7-6.8)	0.20	0.5 (0.2-1.1)	0.07
Boost	25	1.9 (0.5-7.6)	0.36	0.2 (0.1-0.6)	<0.01
Other	10	0.2 (0.02-2.2)	0.19	0.4 (0.1-1.1)	0.08
GVHD Risk Score (at 2nd Rx)
Standard*	55	1.0		1.0
High	58	0.3 (0.1-0.9)	0.04	1.0 (0.5-1.7)	0.89
Time from Onset to 2nd Rx
²14 days*	63	1.0		1.0
>14	50	0.3 (0.1-1.0)	0.05	1.2 (0.6-2.6)	0.57

Figure 1: Six-month NRM is lower among steroid boost recipients compared to ATG

Figure 2: Six-month OS is increased among day 28 second-line responders.