Optimizing Pancreatic Cancer Management with Next Generation Imaging and Liquid Biopsy

PROJECT SUMMARY: Pancreatic ductal adenocarcinoma (PDA) will become the 2nd leading cause of cancer deaths in the United States by 2030. Most patients with PDA present with nonresectable/metastatic disease, and systemic chemotherapy is the anchoring treatment in these patients. Surgery is an option in the minority of patients (~30%) who present with localized disease, though most develop early recurrence after a highly morbid surgery due to occult metastases. Therefore, neoadjuvant therapy (NAT) is an emerging standard approach, but is only beneficial if the selected systemic therapy is effective. Indeed, for all patients with PDA, metastatic or otherwise, the duration of effective systemic therapy is the most important factor in their survival. There is a critical unmet need for accurate and timely assessment of treatment response in order to 1) get patients on effective systemic therapy as soon as possible and keep them on it as long as possible, 2) expeditiously discontinue toxic, costly and ineffective therapies, and 3) facilitate evidence-based personalized clinical decisions regarding curative- intent surgery. Current management of PDA relies principally on computed tomography (CT) and tumor markers (CA19-9). However, these tools are not sensitive enough and are too slow for adjudicating benefit in patients with rapidly lethal metastatic disease, and for identifying suitable candidates most likely to benefit from surgery after NAT. We now have cutting-edge tools for more precise quantification of disease burden at the molecular and metabolic levels. We have previously shown that mutant KRAS circulating tumor (ct)DNA can be detected with high sensitivity in PDA, tends to drop rapidly with effective therapy, and may be a more dynamic predictor of therapy response than CA19-9. We have also shown that metabolic imaging with hybrid integrated 18-fluoro- deoxyglucose (FDG) positron emission tomography/magnetic resonance imaging (PET/MRI) improves the detection of subtle metastases that are occult on CT, early response assessment in patients with nonresectable/metastatic PDA, and prediction of pathological response to NAT in patients who undergo resection. Building on these promising results, we hypothesize that the appropriate combination of KRAS ctDNA and PET/MRI biomarkers will enable timely assessment of the clinical utility of therapy in PDA patients. In Aim 1, we will define thresholds for early chemotherapy switch in unresectable/metastatic PDA using dynamic and quantitative changes in KRAS ctDNA and FDG PET/MRI biomarkers for use in future prospective trials. In Aim 2, we will construct, test and validate a model of surgical benefit or futility for patients with potentially resectable PDA using dynamic KRAS ctDNA and FDG PET/MRI response data. Our overarching goal is to integrate reliable biomarkers that can accurately guide therapy and enable precision medicine to improve outcomes of patients with this deadly disease.