P08.12 Tolerability and initial efficacy of convection-enhanced delivery of combinatorial IL-13RA2 and EphA2 targeted cytotoxins to dogs with spontaneous intracranial malignant gliomas

Abstract

INTRODUCTION: The interleukin 13 receptor alpha 2 (IL-13RA2) and EphA2 receptor are attractive molecular targets for malignant gliomas, being expressed in ~90% of canine and human malignant gliomas, and virtually absent in normal brain. We have generated potent IL-13 and ephrin-A1 (eA1)-based cytotoxins containing modified Pseudomonas exotoxin A or Diphtheria toxin targeted to IL-13RA2 and EphA2 receptors, respectively. We have begun a Phase I clinical trial in dogs with spontaneous gliomas, a faithful animal model of human disease, to determine tolerability of a IL-13RA2-and EphA2-targeted cocktail delivered using MRI-monitored convection enhanced delivery (CED). METHODS: Inclusion criteria for canines included clinical signs of forebrain dysfunction, histopathologically confirmed glioma demonstrating immunoreactivity to IL-13RA2 and/ or EphA2, and provision of written owner consent. A 3+3 dose-escalation design was used, with cohorts administered 0.05, 0.1, or 0.2 μg of each cytotoxin/ml of infusate. An inverse CED planning method, using a shape fitting algorithm generated from patient specific, segmented MRI/CT images, simulated optimum cannula placement and target coverage prior to treatment. CED was performed using reflux-preventing cannulae to co-administer cytotoxins with gadolinium-albumin at rates of 1-5 μl/minute; this allowed intraoperative MRI visualization of infusate distribution. Dose-limiting toxicities (DLT) are defined by the development of grades 3, 4, or 5 adverse events within 28 days of infusion. Clinical, laboratory, and brain MRI examinations were performed on days 14, 28, 42, 84, and 180 following treatment. Objective tumor responses were defined using RANO and volumetric criteria. RESULTS: Nine CED infusions have been performed in eight dogs with unresected gliomas (GBM, n=3; anaplastic oligodendroglioma, n=3; anaplastic mixed glioma, n=1; and anaplastic astrocytoma, n=1). The median target volume was 5.63 cm3 (0.69 to 11.4 cm3). The median duration of infusion was 2 hours (1.5-10 hours). The median actual target coverage was 74% (38-94%). MRI monitoring facilitated intraoperative cannulae revisions that allowed continued target infusion after observation of ventricular leakage or infusate reflux in 3/9 procedures. Durable clinical and partial tumor volumetric responses (52-95% decreases) have been observed in 3/6 dogs with ≥12 weeks of follow-up. Four dogs have died of progressive disease, with tumor necrosis in infused regions evident in post-mortem examinations in these cases. DLT have not been observed. CONCLUSIONS: Improvements in CED cannula design, therapeutic planning, and MRI monitoring allow for safe and effective intratumoral delivery of IL-13RA2 and EphA2 targeted cytotoxins. This study provides preliminary evidence of the efficacy of these cytotoxins in a spontaneous animal model.