Prospective evaluation of predictive factors of response to treatment and prediction in patients with advanced neoplasia of the biliary tract

Abstract

Biliary duct and gallbladder tumors lead to 3.320 annual deaths in US, according to American Cancer Society and is highly prevalent in Latin America. Curative treatment is based on surgery. Most patients present with unresectable disease with expected survival inferior to 12 months, even when treated with chemotherapy. Identification of those who will benefit from treatment is essential, focusing on tumor metabolism evaluation and biomarkers. Among potential biomarkers, the positron emission tomography (PET) using 18-Fluorodeoxiglucose is a well-established diagnostic method for detection of malignant tumors that seems to be useful for characterization of tumor response and prognostic prediction. Molecular markers could also play a role in response and prognosis prediction. , Objectives: To evaluate whether (1) reduction in the tumor's glucolitic metabolic activity detected by PET scan after first 2 cycles of palliative chemotherapy could precisely predict patients that will have a better response; whether (2) baseline molecular alterations and serum biomarkers can identify responders to treatment. Methods: We propose a quasi-experimental and exploratory prospective cohort of patients with unresectable and/or metastatic biliary duct cancer, candidates for palliative chemotherapy treatment with cisplatin and gemcitabine. Primary, endpoint is response rate defined by RECIST version 1.1 after 2 cycles of chemotherapy. Exploratory factors to be evaluated are the percentage of reduction in FDG Standardized Uptake Value [%SUV] uptake from baseline to 2 cycles after chemotherapy that predicts the biological behavior of tumor and better outcomes. The best cut off value for metabolic response of PET-CT scan that correlates with tumor response by RECIST will be calculated by a ROC curve. Immunehistochemistry staining will be used to study ERCC1, PTEN and EGRF expression, and specific mutations in BRAF, KRAS, EGFR, and PI3K genes, besides Ca19.9 serum levels. We will use summary statistics to describe results. Categorical variables will be compared pre- and post-treatment by Chi2 test. P values of less than 0.05 will be considered significant. Our planned sample, based on feasibility and convenience, is 50 patients to be accrued in one year. Our plan is to have these results published in up to 3 years. (AU)