Abstract
Treatment of R/R AML and MDS remains an unmet clinical need. We have previously demonstrated the efficacy of haploidentical NK cells and IL-2/IL-15, administered after lymphodepleting chemotherapy, to achieve remissions in 30-50% of patients with advanced AML. We have learned that this individualized therapy is limited by cost, variability, specificity, and difficulties exporting beyond centers capable of on-site manufacturing. To harness the utility of endogenous NK cells, we have designed a novel Tri-specific Killer Engager (TriKE, GTB-3650) for the treatment of CD33-expressing AML/MDS. We hypothesize that inducing NK cell-mediated, AML-specific targeting along with IL-15 stimulation will lead to remission. This is a phase I, first-in-human study of GTB-3650, comprised of a linear sequence with a novel camelid single domain nanobody to engage CD16 on NK cells, a single-chain variable fragment (scFv) to engage CD33, and wild-type IL-15 to act as a co-stimulus to NK cells in the immune synapse. This second-generation camelid TriKE was found to be at least 10-40 times more potent in animal studies than the first-generation TriKE, which contained an anti-CD16 scFv version of the NK cell engager.
Patients with CD33+ R/R AML/MDS were eligible if they had adequate organ function and an absolute lymphocyte count ≥ 200 cells/µL. The primary endpoint is to define the maximum tolerated dose of GTB-3650. One treatment cycle consists of a 72 hr continuous infusion “block” delivered once weekly for two consecutive weeks, followed by a two-week recovery period. The objective is to administer a minimum of 2 treatment cycles in the absence of rapidly progressing disease or unacceptable toxicity. Two patients are enrolled per dose cohort, with FDA-required staggered enrollment for safety.
Four patients with R/R AML have enrolled, with two receiving 1.25 µg/kg/day and 2.5 µg/kg/day, respectively. All patients had advanced disease with high-risk features (two TP53, one monosomy 7, and one secondary AML) and failing their prior therapy with hypomethylating agents and venetoclax, allogeneic transplantation, and other therapies. At the end of two cycles, two patients had stable disease (1 in each dose cohort), and two patients progressed and only received 1 cycle. The treatment has been well tolerated with no treatment-emergent adverse effects or dose-limiting toxicities identified. Importantly, there was no evidence of cytokine release syndrome or fevers. Enrollment is continuing at the next dose level of 5 µg/kg/day.
Peripheral blood was collected before and at several time points after each treatment block. A robust immunologic signature was observed in patients with stable disease who received two cycles of therapy. Before therapy, the absolute number of NK cells was low, with a resting phenotype. After 48 hours of each continuous infusion, absolute lymphocytes and NK cells diminished in the blood as they become activated, as marked by CD69 expression. Four days after completing the 72-hour block, 1 week after the infusion was started, CD69 returned to baseline. This corresponded to a rebound in highly proliferating NK cells in a GTB-3650 dependent manner. Absolute NK cells remained elevated for the two-week rest period of each cycle. These cells lost their Ki-67 marker of proliferation but maintained natural cytotoxicity and CD16 expression. Ex vivo, they were highly functional for antibody-dependent cellular cytotoxicity in the presence of exogenously added GTB-3650 against CD33+ HL60 targets. A similar immune response was seen after cycle 2 of therapy in the patients with stable disease. Bone marrow was collected before and after each cycle and at day 10 (during the continuous infusion therapy) in cycle 1 in some subjects. In subject 3, Ki-67+ NK cells markedly increased to 89% in the marrow compared to 8% in the screening marrow. CD69+ NK cells increased to 66% from a baseline of 32% and there was a slight increase in marrow CD8+ T cell proliferation without a change in activation.
In conclusion, GTB-3650 administration is well-tolerated at the tested doses, and treatment leads to marked NK cell activation and proliferation in vivo, which can be sustained for weeks. Marrow infiltration of NK cells was found during infusion. We have already achieved a biomarker signature of safe NK cell activation, and dose escalation will continue at 5, 10, 25, 50, and 100 µg/kg/day, with updates provided at the annual meeting.
