A comparative evaluation of ultrasound molecular imaging, perfusion imaging, and volume measurements in evaluating response to therapy in patient-derived xenografts.

Most pre-clinical therapy studies use the change in tumor volume as a measure for disease response. However, tumor size measurements alone may not reflect early changes in tumor physiology that occur as a response to treatment. Ultrasonic molecular imaging (USMI) and Dynamic Contrast Enhanced-Perfusion Imaging (DCE-PI) with ultrasound are two attractive alternatives to tumor volume measurements. Since these techniques can provide information prior to the appearance of gross phenotypic changes, it has been proposed that USMI and DCE-PI could be used to characterize response to treatment earlier than traditional methods. This study evaluated the ability of tumor volume measurements, DCE-PI, and USMI to characterize response to therapy in two different types of patient-derived xenografts (known responders and known non-responders). For both responders and non-responders, 7 animals received a dose of 30 mg/kg of MLN8237, an investigational aurora-A kinase inhibitor, for 14 days or a vehicle control. Volumetric USMI (target integrin:α av β3) and DCE-PI were performed on day 0, day 2, day 7, and day 14 in the same animals. For USMI, day 2 was the earliest point at which there was a statistical difference between the untreated and treated populations in the responder cohort (Untreated: 1.20 ± 0.53 vs. Treated: 0.49 ± 0.40; p < 0.05). In contrast, statistically significant differences between the untreated and treated populations as detected using DCE-PI were not observed until day 14 (Untreated: 0.94 ± 0.23 vs. Treated: 1.31 ± 0.22; p < 0.05). Volume measurements alone suggested no statistical differences between treated and untreated populations at any readpoint. Monitoring volumetric changes is the "gold standard" for evaluating treatment in pre-clinical studies, however, our data suggests that volumetric USMI and DCE-PI may be used to earlier classify and robustly characterize tumor response.