Purpose: This study was designed as "proof of concept" for a drug development model utilising multi-tracer serial small animal PET imaging to characterise tumour responses to molecularly targeted therapy.
Methods: Mice bearing subcutaneous A431 human squamous carcinoma xenografts (n=6-8) were treated with the pan-Erb-B inhibitor CI-1033 or vehicle and imaged serially (days 0, 3 and 6 or 7) with [(18)F]fluorodeoxyglucose, [(18)F]fluoro-L: -thymidine, [(18)F]fluoro-azoazomycinarabinoside or [(18)F]fluoromisonidazole. Separate cohorts (n=3) were treated identically and tumours were assessed ex vivo for markers of glucose metabolism, proliferation and hypoxia.
Results: During the study period, mean uptake of all PET tracers generally increased for control tumours compared to baseline. In contrast, tracer uptake into CI-1033-treated tumours decreased by 20-60% during treatment. Expression of the glucose transporter Glut-1 and cell cycle markers was unchanged or increased in control tumours and generally decreased with CI-1033 treatment, compared to baseline. Thymidine kinase activity was reduced in all tumours compared to baseline at day 3 but was sevenfold higher in control versus CI-1033-treated tumours by day 6 of treatment. Uptake of the hypoxia marker pimonidazole was stable in control tumours but was severely reduced following 7 days of CI-1033 treatment.
Conclusion: CI-1033 treatment significantly affects tumour metabolism, proliferation and hypoxia as determined by PET. The PET findings correlated well with ex vivo biomarkers for each of the cellular processes studied. These results confirm the utility of small animal PET for evaluation of the effectiveness of molecularly targeted therapies and simultaneously definition of specific cellular processes involved in the therapeutic response.