T Cell Costimulation Blockade with Abatacept Nearly Eliminates Early Severe Acute Graft Versus Host Disease after HLA-Mismatched (7/8 HLA Matched) Unrelated Donor Transplant, with a Favorable Impact on Disease-Free and Overall Survival

Ben Watkins, Muna Qayed, Brandi Bratrude, Kayla Betz, Melanie Brown, James Rhodes, Shauna Sinclair, Yvonne Suessmuth, Alison Yu, Kyle Hebert, Marcelo C. Pasquini, Sung W. Choi, Jeffrey Davis, Christine Duncan, Roger Giller, Michael Grimley, Andrew C Harris, David A. Jacobsohn, Nahal Lalefar, Maxim Norkin, Michael A Pulsipher, Shalini Shenoy, Bruce Blazar, Amelia Langston, John T Horan and Leslie Kean


HLA mismatched unrelated donor (MMURD) HCT plays a critical role in the treatment of high-risk hematologic malignancies in ethnic minorities and others lacking a fully matched donor. However, the high incidence of severe (Grade III-IV) acute GVHD (AGVHD) and resultant mortality remain critical barriers to success. T cell costimulation blockade with CTLA4-Ig has been demonstrated to potently inhibit T cell activation, and is FDA-approved for rheumatoid arthritis and renal transplant. We previously completed a pilot study adding 4 doses of the 'abatacept' formulation of CTLA4-Ig to calcineurin inhibitor (CNI) + methotrexate (MTX) GVHD prophylaxis in URD HCT (BBMT 19:1638-49, 2013). Based on robust immune reconstitution and a low rate of AGVHD, we initiated a multicenter phase II trial of abatacept (10 mg/kg/dose, Days -1, +5, +14, +28) plus CNI/MTX ('ABA2') in patients undergoing URD HCT ( # NCT01743131).

ABA2 has 2 cohorts: (1) 8/8 URD (a randomized double-blind stratum vs CNI/MTX, current accrual =133 of 140 patients) and (2) 7/8 MMURD (a single-arm stratum with 43 patients, enrollment completed in Nov, 2016). The 7/8 stratum was non-randomized due to ethical concerns related to the high risk of death from severe AGVHD in 7/8 URD HCT in the absence of adjunctive immunosuppression. The 7/8 stratum thus underwent a pre-specified comparison to CIBMTR matched controls. Here we report the results from the 7/8 stratum with a follow-up of 227-1544 days (median 559).

The primary outcome was the incidence of severe (Gr III-IV) AGVHD prior to day +100. Key secondary outcomes were Gr II-IV AGVHD, chronic GVHD (CGVHD), graft failure, engraftment, transplant-related mortality (TRM), relapse, disease-free survival (DFS) and overall survival (OS). Key inclusion criteria were age >6, myeloablative conditioning and a high-risk heme malignancy. Key exclusion criteria were the use of ATG, alemtuzumab, or other T cell depletion, or a cord blood graft. Stopping rules were in place for TRM and PTLD, neither of which were met. We compared ABA2 to two CIBMTR matched control cohorts: (1) 7/8 URD HCT using CNI + MTX (n =127) or (2) CNI/MTX plus anti-thymocyte globulin ('+ATG', n = 128). Matching was based on age, performance score and disease stage, with similar baseline characteristics (Table 1).

42 of 43 ABA2 patients received all 4 abatacept doses, and all 43 successfully engrafted donor cells, with no delay in median neutrophil (17 days) or platelet (18 days) engraftment. ABA2 was associated with a marked reduction in the cumulative incidence of grade III-IV AGVHD at day 100, with only 2.3% Gr III-IV AGVHD in ABA2 vs 30% in CNI+MTX and 22% in +ATG (p =0.0070 and 0.025, Fig. 1A, 2A). Grade II-IV AGVHD was similar in ABA2 vs controls (44.9% in ABA2 vs 53% in CNI+MTX and 45% in +ATG, p= 0.08 and 0.36). CGVHD was also similar: 40% at 12 months in ABA2 vs 41% in CNI+MTX and 37% in +ATG (p =ns). Importantly, the reduction in severe AGVHD in ABA2 was achieved without significant impact on patient safety, resulting in striking improvement in 12 month TRM (12.5% vs 34% in CNI+MTX and 27% in +ATG (p = 0.002 and 0.04, Fig. 1B, 2B). Also noteworthy, there was no increase in relapse in ABA2, with 7.5% relapse at 12 months compared to 15% in CNI+MTX and 11% in +ATG (p=0.04 and 0.26). Detailed analyses of viral reactivation and other infectious outcomes are ongoing, but uncontrolled infections were not a major problem. This led to very promising rates of DFS and OS in ABA2: 81% 12 month DFS, which was better than both CNI+MTX (50%) and +ATG (63%), p=0.0003 and 0.016, Fig. 1C, 2C). The same was true for OS: 85.4% OS at 12 months in ABA2, vs CNI+MTX (57%) and +ATG (68%, p=0.0025 and 0.048 respectively, Fig. 1D, 2D). These survival advantages are stable thus-far through 2 years.

Our results strongly suggest that the administration of 4 doses of abatacept for GVHD prophylaxis is safe and highly effective at preventing severe AGVHD, lessens the risk for TRM and improves survival in MMURD HCT for heme malignancies. Our results also suggest that the administration of 4 doses of abatacept prior to day +30 may not be sufficient to impact CGVHD, thus trials to assess the effect of longer courses of abatacept are warranted. These results are the first to demonstrate significant efficacy of in vivo T cell costimulation blockade in enhancing outcomes after MMURD HCT, which could significantly expand the donor pool and improve outcomes for patients requiring these high-risk transplants.

Disclosures Norkin: Celgene: Honoraria, Research Funding. Pulsipher: Jazz Pharmaceuticals: Other: advisory board, education; CSL Behring: Other: advisory board Feb 2017; Chimerix: Other: Advisory board Dec 2015; Novartis Pharmaceuticals Corporation: Consultancy, Other: Phase II steering committee; Adaptive: Other: Advisory board 6/17. Kean: Juno: Research Funding; Regeneron: Research Funding; Kymab Ltd: Research Funding; Bristol Myers Squibb: Consultancy.

  • * Asterisk with author names denotes non-ASH members.