Magnetic resonance imaging of tumor response to stroma-modifying pegvorhyaluronidase alpha (PEGPH20) therapy in early-phase clinical trials
Andres Arias, James R Costello, Sunil R. Hingorani, Daniel D Von Hoff, Ronald L Korn, Natarajan Raghunand
Abstract
Abstract Pre-clinical and clinical studies have shown that PEGPH20 depletes intratumoral hyaluronic acid (HA), which is linked to high interstitial fluid pressures and poor distribution of chemotherapies. 29 patients with metastatic advanced solid tumors received quantitative magnetic resonance imaging (qMRI) in 3 prospective clinical trials of PEGPH20: HALO-109-101 (NCT00834704), HALO-109-102 (NCT01170897), and HALO-109-201 (NCT01453153). Apparent Diffusion Coefficient of water (ADC), T1, k trans , v p , v e , and iAUC maps were computed from qMRI acquired at baseline and ≥ 1 time point post-PEGPH20. Tumor ADC and T1 decreased, while iAUC, k trans , v p , and v e increased, on day 1 post-PEGPH20 relative to baseline values. This is consistent with HA depletion leading to a decrease in tumor extracellular water content and an increase in perfusion, permeability, extracellular matrix space, and vascularity. Baseline parameter values predictive of pharmacodynamic responses were: ADC > 1.46 × 10 −3 mm 2 /s (Balanced Accuracy (BA) = 72%, p < 0.01), T1 > 0.54 s (BA = 82%, p < 0.01), iAUC < 9.2 mM-s (BA = 76%, p < 0.05), k trans < 0.07 min −1 (BA = 72%, p = 0.2), v e < 0.17 (BA = 68%, p < 0.01), and v p < 0.02 (BA = 60%, p < 0.01). A low v e at baseline was moderately predictive of response in any parameter (BA = 65.6%, p < 0.01 averaged across patients). These qMRI biomarkers are potentially useful for guiding patient pre-selection and post-treatment follow-up in future clinical studies of PEGPH20 and other tumor stroma-modifying anti-cancer therapies.