AUTHORS
Margery Gang, MD, MA, Monica Thakar, MD2, Parameswaran Hari, MD, MBBS3, Kevin Ng, MS4, Michael C Wu, PhD5, David G Maloney, MD, PhD6, Carolyn Keever-Taylor, DPhil7, Subramaniam Malarkannan, PhD8, Shelly Heimfeld, PhD9, Rainer F Storb, MD10, Brenda M Sandmaier, MD11
AFFILIATIONS
1 Fred Hutchinson Cancer Center, Seattle, Washington, United States, University of Washington, Seattle, Washington, United States
2 Fred Hutch Cancer Center, Seattle, Washington, United States, University of Washington, Seattle, Washington, United States
3 Medical College of Wisconsin, Milwaukee, Wisconsin, United States
4 Fred Hutchinson Cancer Center, Seattle, Washington, United States
5 Fred Hutchinson Cancer Center, Seattle, Washington, United States
6 Fred Hutchinson Cancer Center, Seattle, Washington, United States, University of Washington, Seattle, Washington, United States
7 Medical College of Wisconsin, Milwaukee, Wisconsin, United States
8 Versiti, Milwaukee, Wisconsin, United States
9 Fred Hutchinson Cancer Center, Seattle, Washington, United States
10 Fred Hutch Cancer Center, Seattle, Washington, United States, University of Washington, Seattle, Washington, United States
11 Fred Hutchinson Cancer Center, Seattle, Washington, United States, University of Washington, Seattle, Washington, United States
ABSTRACT
Introduction: The use of post-transplant cyclophosphamide (PTCy) in HLA-haploidentical HCT (haplo-HCT) provides potent in vivo T- and NK-cell depletion, significantly reducing severe graft-versus-host disease (GVHD). However, survival is limited by increased risk of early relapse, potentiated by PTCy-mediated depletion of early NK cells crucial for immune surveillance. This is particularly evident in the non-myeloablative setting (Flu/Cy/TBI 2Gy), when the balance of long-term disease control shifts more heavily to the graft-versus-leukemia (GVL) effect rather than conditioning intensity. To address this, we developed a phase I/II trial evaluating a prophylactic donor NK infusion after haplo-HCT (NCT00789776) to safely replace this NK loss. Initial results were previously presented (Thakar, et al. ASH 2016) and demonstrated safety and feasibility. Here, we report correlative NK cell studies and long-term follow-up (>10 years from final enrollment) to confirm the durability of this approach.
Methods: Participants (pts) with high-risk hematologic malignancies underwent non-myeloablative haplo-HCT using marrow. GVHD prophylaxis consisted of a single dose of PTCy (50 mg/kg, day [D]+3), followed by tacrolimus and MMF (D+4). On D+7, pts received a fresh donor NK cell infusion (either 2.5 or 5 x 106/kg, per dose escalation) using non-mobilized PBMC that underwent CD3 depletion/CD56 selection (Miltenyi CliniMACS). Phase II primary endpoints were cumulative incidence of relapse (CIR), non-relapse mortality (NRM), and grade III-IV acute GVHD (aGVHD). Exploratory endpoints included engraftment, chimerism, OS, PFS, and chronic GVHD (cGVHD). For exploratory KIR ligand alloreactivity correlatives, donors and recipients were grouped as C1, C2, and/or Bw4 based on available high-resolution HLA typing. NK alloreactivity was determined by the “missing ligand” model. HLA-KIR genotyping was performed at Versiti.
Results: Between 10/2008-6/2016, 40 pts with median age 46 (8-75) years were treated. Diagnoses included ALL (n=12), AML (n=8), MDS (n=4), BPCDN (n=2), CLL (n=1), HL (n=6), MM/PCL (n=4), and NHL (n=3). The median number of prior therapies was 2.5 (range, 1-9), many having prior HCTs (allo=7; auto=11), and the majority having intermediate (53%) or high (33%) Disease Risk Index scores. 55% were in MRDneg CR, 18% in MRDpos CR, and 27% had active disease.
Thirty-five pts received the phase II dose (5 x 106/kg NK cells) and included in primary efficacy analysis. The 1-year relapse incidence was 29 (95% CI=15-44)%, 200-day NRM was 9 (2-21)% and grade III aGVHD <100 days was 6 (0.7-19)%, with no grade IV events. Neutrophil engraftment was achieved in 89% pts, with median time to engraftment 14 (range, 11-38) days. Median D28 donor chimerism was 100% for both CD3 (n=31) and CD56 (n=21), respectively; 20% (n=7) experienced graft rejection. 1-year OS and PFS estimates were 71 (53-84)% and 63 (45-77)%, respectively. The probability of 1-year cGVHD was 20 (9-35)%. The probability of ≥grade 3 or higher infections by D+100 was 17 (7-34)%.
In updated analyses, long-term outcomes are reported. Among survivors, median follow-up was 6.1 (range, 4.0-12) years. At 5 years, OS was 43 (26-58)%, PFS was 31 (17-47)%, CIR was 43 (26-59)%, and NRM was 26(13-42)%. The 5-year incidence of cGVHD was 23 (11-38)%. Primary causes of death included disease relapse/progression (n=12), infection (n=5), secondary malignancy (n=2), cGVHD (n=1), and cardiogenic shock (n=1).
Finally, exploratory correlative analyses were done to assess the impact of NK alloreactivity on post-HCT outcomes in 32 available pts. Of these, we identified 8 donor/recipient pairs with “missing ligand” (NK alloreactive) and 24 “no missing ligand” status, respectively. Preliminary analyses show that the 1-year CIR was similar in recipients of “NK alloreactive” vs. “no missing ligand” HCTs (25% [3-58%] vs. 25% [10-44%], P=0.7). Data collection is ongoing and these analyses will be updated.
Conclusions: With this extended follow-up, our data confirm the safety, feasibility, and durable survival associated with donor NK cell infusion after nonmyeloablative haplo-HCT for pts with high-risk hematologic malignancies. These findings support this approach as a promising platform to enhance early GVL effects, promote long-term disease control, and warrant further investigation.
DISCLOSURES
Margery Gang, MD, MA: No relevant conflicts of interest to declare.