A <scp>Phase I</scp> dose‐escalation study of <scp>DCLL9718S</scp>, an antibody‐drug conjugate targeting <scp>C</scp>‐type lectin‐like molecule‐1 (<scp>CLL</scp>‐1) in patients with acute myeloid leukemia

Abstract

Acute myeloid leukemia (AML) is an aggressive malignancy with especially dismal outcomes in relapsed or refractory patients. C-type lectin-like molecule-1 (CLL1) is a transmembrane glycoprotein expressed on the surface of AML blast cells.1 CLL1 is expressed on committed myeloid cells in bone marrow, but is absent on normal HSCs. Its expression is detectable on neutrophils and monocytes/macrophages but there is no known expression on non-hematopoietic stem cells (HSC) or tissues.2 The absence of CLL1 expression on HSC may translate into less myelosuppression than other AML targets. Reduced myelosuppression may increase the therapeutic window while also enabling better combinability with standard AML therapies. DCLL9718S is a THIOMAB™ antibody-drug conjugate (TDC) consisting of a humanized monoclonal IgG1 anti-CLL1 antibody (MCLL0517A) linked to two pyrrolobenzodiazepine (PBD) dimer drugs via a cleavable disulfide linker. The development of DCLL9718S was under a Research Collaboration and License Agreement with Spirogen Ltd., a member of the AstraZeneca group. The PBDs bind covalently in the minor groove of DNA and form inter-strand and intra-strand cross-linked adducts as well as mono-adducts.3, 4 Following binding of the TDC to target cells expressing CLL1, the TDC is internalized, degraded in the lysosome, and the active PBD dimer drug is released, resulting in DNA damage and cytotoxicity.4 Patients ≥18 years with relapsed or refractory AML, Eastern Cooperative Oncology Group (ECOG) Performance Status score 0–2, adequate end-organ function, and who were were willing and able to undergo pre-treatment and on-treatment bone marrow aspirates were eligible. The primary objectives were to assess the safety and tolerability of DCLL9718S and to determine the maximum tolerated dose (MTD). Secondary objectives included analysis of pharmacokinetics, pharmacodynamic biomarkers, target expression profile, and preliminary assessment of anti-tumor efficacy (see detailed inclusion criteria, DLT definitions, and endpoints in supplemental methods). Efficacy was measured by the rate of complete remission (CR), complete remission with incomplete blood count recovery (CRi), and complete remission with incomplete platelet count recovery (CRp) per the International Working Group Response Criteria for AML,5 assessed at the end of Cycles 1 and 3, then every three cycles. Eighteen safety evaluable patients were assigned to five escalating dose level cohorts (dose levels 10–160 μg/kg). Baseline characteristics and molecular markers at screening are presented in Table 1. Patients received DCLL9718S (10–160 μg/kg) given in 21-day cycles with DCLL9718S dosed on Day 1 of each cycle. The median number of DCLL9718S treatment cycles administered was two (range 1–4). Most patients experienced at least one AE, irrespective of attribution (Table S1). Half of the safety evaluable patients experienced at least one AE that was considered related to the study drug. Twelve patients (67%) experienced at least one AE of Grade ≥3 intensity; the most common were febrile neutropenia (six patients [33%]) and pneumonia (five patients [28%]). No DLTs were reported during the 21-day DLT assessment window. Two patients had Grade 4 AEs (thrombocytopenia) assessed as related to the study drug. The SAEs experienced by ≥25% of patients were febrile neutropenia (six patients [33%]) and pneumonia (four patients [22%]). No deaths were reported as related to DCLL9718S. Two patients in the 160 μg/kg dose developed elevated liver function tests. One patient experienced a Grade 3 increased ALT, Grade 2 increased AST, and Grade 1 increased alkaline phosphatase, without concomitant bilirubin elevation (all related to study drug) on Cycle 1 Day eight, but had a Grade 1 bilirubin elevation on Day 17 (unrelated) that resolved on its own on Cycle 1 Day 19. The patient then developed progression of disease and was taken off protocol on Cycle 1 Day 19. The ALT/AST elevations had both improved to Grade 1 at that time. The second patient experienced Grade 3 AST/ALT, bilirubin, and alkaline phosphatase increases on Cycle 1 Day eight, initially thought to be due to the concomitant administration of posaconazole (Figure S1). Posaconazole was discontinued. The AST/ALT elevations resolved on study Day 23, though bilirubin and alkaline phosphatase remained elevated for several weeks. The delayed recovery raised the concern that a causal relationship to DCLL9718S could not be ruled out. A maximum tolerated dose was not identified; based on the hepatic events observed at the highest dose and lack of anti-leukemic activity dose escalation was stopped. The mean acPBD maximum concentrations (Cmax) occurred immediately after the infusion and increased with dose; acPBD PK showed a multi-exponential decline. The acPBD and the total antibody analytes demonstrated dose-dependent rapid clearance of ADC from circulation and large IIV (up to 182% CV) across the tested doses. Increases in Cycle 1 doses generally resulted in an increase in systemic exposure for all the three analytes; the Cycle 1 AUC0-21d increased disproportionately with dose for all the three analytes suggesting non-linear PK. The unconjugated PBD was consistently low. Minimal accumulation was observed for the acPBD, total antibody and unconjugated PBD analytes upon repeated dosing on the q3w schedule and steady-state appeared to be reached within the first dose in Cycle 1. The summary of PK parameters after the first dose in Cycle 1 are shown in Figure S2 and Table S2. No post-baseline ADA evaluable patients tested positive for ADA. Among the 18 patients who received DCLL9718S, no patients achieved an objective IWG2003 CR or PR response. Response data was missing in three of these 18 patients (two patients in 160 μg/kg and one patient in 40 μg/kg) due to death or discontinuation of treatment before assessments (Figure 1). Expression of CLL1 was detected in all patients except one (95%, 17/18), with varying expression intensity. While the majority of patients showed an expected unimodal CLL1 expression, a bimodal expression of CLL1 with positive and negative population on CD34+ blasts was observed in some AML samples (3/18 patients). We were not able to assess whether the CLL1 bimodal expression and intensity had an impact on the clinical activity since no responses were observed. DCLL9718S is the first anti-CLL1 ADC to enter the clinic. Preclinical evidence suggested that as compared to CD33 which is expressed on normal HSC, the degree and duration of marrow suppression with an anti-CLL1 ADC would be reduced. Also, DCLL9718S was designed with a highly potent PBD payload with the potential to have anti-leukemic activity even in chemotherapy resistant AML cells.6, 7 The study was stopped during the dose escalation phase based on an assessment of safety, efficacy and pharmacokinetic data. With the limited data, the cause of the hepatic injury cannot be definitely determined, but may plausibly be attributed to the PBD payload. Emerging data from other ADCs in development in multiple indications suggest that hepatic injury may be a class effect of the PBD or modified PBD-like payloads.8 Although anecdotal, both patients on this study with hepatic AEs had relatively higher DCLL9718S exposures (Cmax/AUC) for all the three measured analytes. Hepatic injury by direct targeting of the liver by DCLL9718S is considered unlikely as CLL1 is not known to have any hepatic expression. Other mechanisms of hepatic injury could include non-specific uptake of antibody in the liver, metabolism of the unbound PBD in the liver, covalent binding of the PBD dimer to the hepatic tissues, or possible bystander killing if AML blasts are present in the liver. Ultimately though the mechanism of hepatic toxicity remains speculative. Evaluation of CLL1 as a potential target in AML was incomplete as dose escalation was stopped due to toxicity unlikely to be attributable to the fact that DCLL9718S targets CLL1. We cannot determine whether dose levels were sufficiently high to drive antileukemic activity. Rapid clearance was observed in the majority of patients suggestive of targeted mediated clearance indicating engagement of CLL1 and internalization of DCLL9718S. Expression of CLL1 was observed in 95% of patients on study. These data support that CLL1 remains a viable target to pursue in AML with future ADCS, bispecific antibodies and CART approaches. As recently reported, a subset of AML patients have bimodal CLL1 expression.9 In these patients with bimodal CLL1 AML blast expression, a subpopulation of blasts were CLL1 negative. No evidence of selective clearance of CLL1 positive blasts was observed in this study. Given the limited tolerability and anti-tumor activity of DCLL9718S in this study, the program will not move forward in Phase II trials. In general, drugs (e.g., ADCs) with a highly potent payload, a narrow therapeutic index, and high PK variability pose significant challenges to development. CLL1 as a target in AML continues to remain viable and the data presented here helps to elucidate its role, guide future development and potential pitfalls that should be considered with future CLL1 directed therapies in myeloid malignancies. We thank the patients and their families who took part in the study, as well as the staff, research coordinators, and investigators at each participating institution. Writing and editing assistance was provided by Genentech, Inc. This work was supported by Genentech, Inc. Genentech was involved in the study design, data interpretation, and the decision to submit for publication in conjunction with the authors. Naval Daver: Has received research funding from Daiichi-Sankyo, Bristol-Myers Squibb, Pfizer, Gilead, Sevier, Genentech, Astellas, Daiichi-Sankyo, Abbvie, Hanmi, Trovagene, FATE, Amgen, Novimmune, Glycomimetics, and ImmunoGen and has served in a consulting or advisory role for Daiichi-Sankyo, Bristol-Myers Squibb, Pfizer, Novartis, Celgene, AbbVie, Astellas, Genentech, Immunogen, Servier, Syndax, Trillium, Gilead, Amgen and Agios. Amandeep Salhotra: Research funding from: BMS/Celgene. Ad Board: Syros. Consultancy: Kadmon. Joseph Brandwein: Consulted for and received honoraria from Pfizer, Celgene/Bristol-Myers Squib, Jazz, Abbvie, Astellas, Amgen, Taiho, Roche, Novartis, Teva. Nikolai Podoltsev: Consulted for and received honoraria from Alexion, Pfizer, Agios Pharmaceuticals, Blueprint Medicines, Incyte, Novartis, Celgene, Bristol-Myers Squib, CTI BioPharma, PharmaEssentia. Received research funding (all to the institution) from Boehringer Ingelheim, Astellas Pharma, Daiichi Sankyo, Sunesis Pharmaceuticals, Jazz Pharmaceuticals, Pfizer, Astex Pharmaceuticals, CTI biopharma, Celgene, Genentech, AI Therapeutics, Samus Therapeutics, Arog Pharmaceuticals, Kartos Therapeutics. Daniel Pollyea: receives research funding from Abbvie and serves as a consultant for Abbvie, Celgene, Genentech, Novartis, Karyopharm, Syndax, Takeda, Bristol Myers Squibb, Syros, Kiadis. Joseph Jurcic: Research funding (institutional) from: AbbVie, Arog Pharmaceuticals, Astellas Pharma, Celgene, Daiichi-Sankyo, Forma Therapeutics, Genentech, Kuro Oncology, PTC Therapeutics, Syros Pharmaceuticals. Consultant for: AbbVie, AstraZeneca, Celgene/BMS, Daiichi-Sankyo, Incyte, and Novartis. Sarit Assouline: serves as a consultant for Abbvie, Astra Zeneca, Roche, Janssen. Karen Yee: Consulted for and received honorarium from Novartis, F. Hoffmann La Roche, Takeda, Pfizer, TaiHo, Bristol-Myers Squib/Celgene, Paladin, Astex, and Otsuka. Received research funding from Astex, Novartis, Forma Therapeutics, Jazz, Onconova, F. Hoffmann La Roche, Genentech, and Tolero. Mengsong Li: Employee of Genentech, Inc., shareholder of F. Hoffmann La Roche, Ltd. Tony Pourmohamad: Employee of Genentech, Inc., shareholder of F. Hoffmann La Roche, Ltd. Divya Samineni: Employee of Genentech, Inc., shareholder of F. Hoffmann La Roche, Ltd. Teiko Sumiyoshi: Employee of Genentech, Inc., shareholder of F. Hoffmann La Roche, Ltd. Anjali Vaze: Employee of Genentech, Inc., shareholder of F. Hoffmann La Roche, Ltd. Randall Dere: Employee of Genentech, Inc., shareholder of F. Hoffmann La Roche, Ltd. Connie Ma: Employee of Genentech, Inc., shareholder of F. Hoffmann La Roche, Ltd. James Cooper: Employee of Genentech, Inc., shareholder of F. Hoffmann La Roche, Ltd. Designed of the study: Randall C. Dere. Provided patient data: Naval Daver, Amandeep Salhotra, Joseph M. Brandwein, Nikolai A. Podoltsev, Daniel A. Pollyea, Joseph G. Jurcic, Sarit Assouline, Karen Yee, Teiko Sumiyoshi, Connie Ma. Analyzed data: Naval Daver, Joseph G. Jurcic, Mengsong Li, Tony Pourmohamad, Divya Samineni, Teiko Sumiyoshi, Anjali Vaze, Randall C. Dere, Connie Ma, James Cooper. Writing of the paper: all authors. All authors reviewed the manuscript critically and approved the final version of the manuscript. Qualified researchers may request access to individual patient level data through the clinical study data request platform (https://vivli.org/). Further details on Roche's criteria for eligible studies are available here (https://vivli.org/members/ourmembers/). For further details on Roche's Global Policy on the Sharing of Clinical Information and how to request access to related clinical study documents, see here (https://www.roche.com/research_and_development/who_we_are_how_we_work/clinical_trials/our_commitment_to_data_sharing.htm). Appendix S1. Supporting Information. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.

Topics & Concepts