A Phase I Study of IGN523, a Novel Anti-CD98 Monoclonal Antibody in Patients with Relapsed or Refractory Acute Myeloid Leukemia (AML)

CD98 is a heterodimeric cell surface protein that is implicated in the pathogenesis of many hematological and solid malignancies. Elevated tumor CD98 expression has been correlated with poor prognosis. Previous studies demonstrated that CD98 functions in integrin signaling and amino acid transport processes that support tumor cell proliferation, anchorage independence, invasion, and metastasis. IGN523 is a humanized monoclonal antibody targeting CD98 with multiple mechanisms of action including antibody-dependent cellular cytotoxicity (ADCC), direct cell death induction, and interference with the normal amino acid transport function of CD98. IGN523 elicits potent anti-tumor activity against a variety of leukemic xenograft models (Hayes et al., Int. J. Cancer 2015). Here we present the results of a Phase I dose escalation study of IGN523 in adult patients with relapsed or refractory AML.

IGN523 was administered intravenously (IV) as a single agent to 19 patients at weekly doses ranging from 0.1 mg/kg to 20 mg/kg, including 3 single-patient cohorts starting at 0.1, 0.3, and 1 mg/kg, 6 patients treated at 10 mg/kg, 3 at 20 mg/kg, and 7 at 15 mg/kg. Median age was 71 years (range 23-78). Patients received a median of one 4-week cycle of treatment (range <1-10 cycles).

The most common treatment-related adverse events observed were part of infusion-related reactions (IRRs) occurring primarily with the first infusion, and were generally Grade 1 or 2 (CTCAE V4.03), including chills (52.6%), fever (pyrexia or febrile neutropenia, 36.9%), nausea (26.3%), and vomiting (26.3%). Premedication with acetaminophen, diphenhydramine, and hydrocortisone was instituted after a dose-limiting toxicity (DLT) of significant IRR with the first dose was observed in 1 of 3 patients treated at 10 mg/kg, and the cohort was expanded to 6 patients. Treatment-related Grade 3 and 4 AEs occurring in >1 patient included febrile neutropenia (26.1%), pyrexia (10.5%), and altered mental status (confusional state or mental status changes in 15.8% and 10.5% respectively). DLTs all occurred with the first dose of IGN523 and were observed in 1 of 6 patients at 10 mg/kg, 2 of 3 at 20 mg/kg, and 2 of 7 at 15 mg/kg. The maximum tolerated dose (MTD) of IGN523 administered weekly was 10 mg/kg.

PK studies of IGN523 in all patients demonstrated evidence of target-mediated drug disposition at doses below 10 mg/kg/week, as pre-dose (trough) CD98 receptor occupancy (%RO) levels tended to be < 90%. By the 10 mg/kg dose cohort, pre-dose %RO levels were 50 to > 90% during the first few weeks of dosing, and consistently > 90% after 4-8 weeks of dosing. At 15 and 20 mg/kg/week doses, %RO was consistently > 90% from the first dose. Therefore, dose levels of 10 mg/kg/week and above should be suitable for most patients to maintain >90% receptor occupancy over time. Anti-drug antibodies (ADA) were not detected in any patient.

Evidence of transient anti-leukemic activity was observed in 3 patients, including improvement in platelet count and reduction in marrow blast percentage, early reduction and control of blasts, and an episode of tumor lysis and peripheral blast count reduction. No complete or partial responses were observed.

In conclusion, IGN523 was safe when administered at 10 mg/kg IV weekly with DLTs of significant first-dose infusional toxicity that may be manageable with prophylactic steroids or a modified dosing regimen. IGN523 demonstrated modest evidence of anti-leukemic activity as a single agent. Preclinical data demonstrating greater than additive activity of IGN523 when combined with cytarabine and significant monotherapy and combination activity with standard chemotherapy in patient-derived xenograft models of NSCLC and colorectal cancer support further development of IGN523 in combination with standard therapies in AML and solid tumors.