There is a substantial need to develop accurate methods for predicting individual response to chemotherapy. Imaging of radiolabeled chemotherapeutic drugs with PET provides one approach to this problem. However, the usefulness of PET with drugs that are rapidly broken down in vivo, such as 5-fluorouracil (FU), is limited by the inability of the methodology to directly identify the molecular association of the radiolabel. Our long range objective is to develop PET as a tool for predicting response to chemotherapy and individualizing regimens for treating neoplastic disease. Our intermediate range objective is to develop clinically-practical PET techniques for predicting tumor response to FU. In order to be effective, FU must be taken up by tumor cells and anabolized to fluoronucleotides, which in turn interfere with DNA and RNA synthesis. Investigators at Heidelberg, Germany, recently reported positive correlation between tumor response to FU and tumor retention of radiolabel from [18F]FU as measured in PET in 17 patients undergoing treatment for metastatic colon cancer. While very promising, this study did not address the influence of recirculating, labeled catabolites of [18F]FU on the PET images, and the imaging technique used provides little information about the kinetics of FU in tumors. We are developing a new approach in which biomodulation is used to create an in vivo imaging technique capable of measuring tumor transport and metabolism of FU per se. Specifically, we pretreat with ethynyluracil (EU), a potent inhibitor of FU catabolism, to prevent degradation of [18F]FU and thereby improve the sensitivity of PET to those aspects of FU pharmacokinetics most closely related to tumor response. We have shown in preclinical studies with a rat colon tumor model that cellular uptake and anabolism of FU can be accurately measured by mathematical modeling of data obtained with our """"""""PET/[18F]FU+EU""""""""technique.
The specific aims of the proposed study are (1) to evaluate our """"""""PET[18F]FU+EU"""""""" technique for predicting tumor response to a form of therapy in which the technique is very closely related, viz., FU modulated by EU (""""""""FU+EU""""""""); and (2) to repeat the Heidelberg study """"""""PET/[18F]FU"""""""") in a larger group of patients with colon cancer. The timeliness of Aim 1 is enhanced by current Phase II clinical trials of FU+EU and new indications of the importance of FU catabolism as a mechanism of tumor resistance to the drug. While PET and [18F]FU may provide information related to thymidylate synthase (TS), the primary target of FU, the imaging techniques does not measure TS inhibition directly. In the proposed study, TS concentration in tumor, a known correlate of tumor response, will be measured via tumor biopsy prior to therapy. Multiple regression analysis will be used to evaluate combination of parameters derived from PET and ex vivo TS assays as predictors of tumor response.
