Armed and SAVVY: CAR T cells combat AML Free

In this issue of Blood, Geyer et al¹ report their planned interim analysis, including well-considered biological correlative study, of the pilot phase of their clinical trial (CLEAR-AML; clinicaltrials.gov identifier: NCT06017258), testing CD371-targeted/interleukin-18 (IL-18)–secreting chimeric antigen receptor (CAR) T-cell therapy in 5 patients with relapsed or refractory acute myeloid leukemia (AML).

The efficacy of CAR T-cell therapy for AML has been limited by several disease- and host-specific environmental factors, including the lack of a uniformly expressed AML-defining surface antigen, poor host T-cell fitness, an inhospitable disease microenvironment, rapid disease progression that is misaligned with CAR T-cell production time, and the potential for inflammatory cytokine–stimulated AML proliferation. Discovery of the optimal CAR T-cell features to eradicate AML without promoting disease progression or stimulating therapy-limiting toxicity is a significant challenge. Therefore, the results of this study are particularly exciting, with 3 of 5 patients showing an anti-AML response following a single infusion of CD371 (CLEC12A, CLL-1)–targeted CAR T cells armed with IL-18 secretion. CD371 is a type II transmembrane glycoprotein with high expression on AML and leukemia initiating cells² and thus, is an AML-associated target utilized for a myriad of CAR T cell and other antibody-based immunotherapies. The second-generation CAR structure used in this study includes a fully human CD371-binding domain to minimize immunogenicity and an intracellular costimulatory CD28 with amino acid substitutions positioned to modulate PI3K binding (SAVVY). CD28 sequence modification in the context of CAR T-cell design is an approach previously reported as improving T-cell function in preclinical study.³ IL-18 is a T- and natural killer (NK)-cell–activating inflammatory cytokine that has been engineered for production by CAR T cells as a strategy to boost local antitumor activity.^4-6 Notably, the safety of this strategy has recently been reported in a clinical trial testing IL-18 secreting CAR T-cell therapy for lymphoma.⁷ 

All patients who received the CD371/SAVVY/IL-18 CAR T-cell product had heavily pretreated and refractory AML. Most patients had circulating blasts at the time of leukapheresis, and all had detectable disease prior to CAR T-cell infusion, highlighting the challenges of manufacturing autologous products for patients with rapidly progressive and brittle disease. Despite this, 3 out of 5 treated patients achieved a morphological leukemia-free state (2/5 MRD-negative) when evaluated 4-weeks postinfusion. It may be significant that the 3 patients who responded to CD371/SAVVY/IL-18 CAR T-cell therapy, were previously treated with an allogeneic transplant. Although cell products were manufactured from patient’s apheresis components, donor cell origin of the manufactured CAR T cells was verified. Further detailed study of product characteristics will be illuminating. If poor patient-derived product quality is identified as relevant, serious consideration should be given to accelerated development of a similar, but allogeneic, product. Unfortunately, 2 of 3 responding patients ultimately relapsed. None of the tested relapse samples had any change in detected CD371 expression, implicating a lack of functional CAR T-cell persistence as determinant. It is intriguing to consider a multiple dosing strategy with closely monitored pharmacokinetics; yet, despite a relatively low infused CD371/SAVVY/IL-18 CAR T-cell number in the first 2 patients (3 × 10⁵ cells per kg), dose-limiting toxicity occurred.

All patients experienced cytokine release syndrome (CRS), with greater than grade 3 toxicity in 2 patients. One patient with a history of cranial radiation and prior central nervous system disease developed grade 3 immune-effector cell–associated neurotoxicity syndrome (ICANS). All observed CRS and ICANS resolved with directed therapy and supportive care. More concerning, significant hematotoxicity with prolonged marrow hypoplasia occurred and several severe infections were recorded. The cause of prolonged cytopenias post CD371/SAVVY/IL-18 CAR T-cell therapy is likely multifactorial and may be due to direct off-cancer on-target cytotoxicity, indirect inflammatory marrow suppression, and/or preexisting poor marrow reserve. Whether this effect is primarily driven by CAR T-cell activity or localized IL-18 secretion is unknown, but consideration of the likelihood of these events regardless of CAR T-cell therapy in the treated patient population is warranted.

A significant strength of this study is the extensive correlative and exploratory analyses conducted to measure CD371/SAVVY/IL-18 CAR T-cell expansion and IL-18 levels, while also studying changes in the CAR T-cell phenotype and the immune microenvironment. These paired analyses are critical for this study (and all studies) of novel, engineered cell therapies to understand the mechanisms that underpin CAR T-cell functionality, especially against AML.⁸ CAR T-cell peak expansion occurred at 7 to 14 days postinfusion with a maximal recorded CAR T-cell persistence of 56 days. Short persistence is likely intrinsic to this specific product because no evidence of humoral rejection was detectable in any patient. Cytokine levels were tracked postinfusion, with IL-18 and interferon gamma peaking early (<10 days postinfusion) in responding patients. Measurement of IL-18 binding protein was not reported. Using 5’-cellular indexing of transcriptomes and epitopes–seq for single-cell analyses of longitudinal patient biospecimens, a persistent cytotoxic effector memory CD8⁺ CAR T-cell phenotype was identified as associated with antitumor response. CAR-positive and CAR-negative T cells expressed a diversity of T-cell receptors, verifying polyclonality. Notably, increasing proportions of activated NK cells⁹ were observed following CD371/SAVVY/IL-18 CAR T-cell therapy, again predominantly in responding patients. It will be interesting to determine whether the emergence of this population is secondary to NK-cell binding of activating ligands on the AML blasts¹⁰ or because of direct stimulation of NK cells by CAR T-cell–derived IL-18. Regardless of the exact mechanism, the increase in cytotoxic NK cells in patients treated with CD371/SAVVY/IL-18 CAR T-cell therapy is likely a unique secondary effect that amplifies the anti-AML activity of this product.

Overall, treatment of patients with AML refractory to standard therapy using CD371/SAVVY/IL-18 CAR T cells is a novel approach. Modulation of the AML-associated immune microenvironment, including activation of cytotoxic NK cells, is an important advance in this challenging disease. Although the observed response rate is certainly thrilling, these responses were not sustained and incomplete disease clearance together with a lack of functional CAR T-cell persistence were likely contributory to disease relapse. Even so, the data from this small cohort of patients are supportive of further testing and development of CD371/SAVVY/IL-18 CAR T-cell therapy guided by paired biological study as treatment for patients who otherwise have extremely limited therapeutic options.

Conflict-of-interest disclosure: C.L.B. has been awarded and has pending patent applications describing the use of engineered T and NK cells as therapeutics. M.T.Z. declares no competing financial interests.