Results of a Phase 1 Clinical Trial of Allogeneic Tumor-Primed Natural Killer Cells in AML

Abstract 946

We have previously reported that resting human NK cells can be primed to lyse NK-resistant tumor cells, including AML, by co-incubation with membrane preparations from the leukemia cell line CTV-1. Tumor-primed NK cells can be cryopreserved and remain in a primed state when thawed. Following translation of this production method to GMP-compliance we designed a phase 1 trial for AML patients. PROTOCOL: AML patients meeting the inclusion criteria were enrolled and they and their children HLA-typed and screened for medical suitability for apheresis. KIR mismatched donors in the GvL direction were preferred when available. PATIENT INCLUSION CRITERIA: >60 years in PR (blasts >5<20% in BM) after 2nd course of induction chemotherapy; >60 years in CR2 after re-induction chemotherapy; >60 years in PR or CR after 2 courses of chemotherapy with poor risk disease by conventional cytogenetic criteria; <60 years beyond CR2 but unsuitable for transplantation. CONDITIONING: Fludarabine 25mg/m2/day for 3 days plus a single fraction (2Gy) TBI on day 4. MANUFACTURE OF Tumor-primed NK: T-pNK cells were generated from single 2hr apheresis from haploidentical related donors in 14 cases; 1 donor declined and the patient was not treated. NK cells were isolated with anti-CD56 microbeads (CliniMACS, Miltenyi Biotec) with (n=2) or without (n=12) prior depletion of CD3+ cells. Purified NK cells were co-incubated with a lysate of CTV-1 cells (DSMZ repository) overnight. Lysate was removed by density gradient separation on day 2 and cells were cryopreserved in dosed aliquots and released for infusion within 15 days of manufacture. TREATMENT: Three groups of 5 patients were to receive 1×10^6/kg, 5×10^6/kg or 1×10^7/kg haploidentical related donor T-pNK as a single i.v. infusion 24 hours after completion of TBI. RESULTS: 15 patients were enrolled between July 2008 and Dec 2009 – Table 1. Mean donor T-pNK yield was 6.73×10^9 and CD56 purity 99.3%. T cell contamination was <10^4/kg. All products achieved QC release criteria. Median time from enrolment to product release was 46 days. 8/15 patients received the T-pNK therapy; 7 untreated due to relapse during product manufacture or failure to remit (6) and donor refusal (1). No infusional toxicity was observed. 7/8 patients experienced profound bone marrow suppression with median time of neutrophil recovery 55 days (19–101). Two patients who previously had an allogeneic transplant were salvaged with a CD34+ top up stem cell transfusion from their original donors. All experienced neutropenic fevers requiring antibiotic support with 2 of them requiring ICU admission. Due to the toxicities observed the dose escalation was abandoned and all treated patients received 1×10^6 cell/kg. The median duration of CR post T-pNK was 222 days (55–845) compared to previous CR median of 47 days (0–90) – Table 1. 7/8 patients experienced a longer CR post T-pNK than their previous CR; median increase in duration of CR was 242 days (range 54–815). CONCLUSIONS: T-pNK can be easily generated and administered with acceptable toxicity which was easily managed. In this non-controlled trial there was evidence of clinical efficacy but delays in donor screening and processing led to 40% attrition of enrolled patients prior to drug administration.