606P - An Estimation of the Population Survival Benefit of First-Line Chemotherapy for Gastrointestinal Cancer

Abstract

ABSTRACT Aim: Randomized clinical trials and meta-analyses describe the benefit of chemotherapy and combined chemotherapy and radiotherapy for specific subgroups of major gastrointestinal (colorectal, pancreas, oesophagus, stomach, liver and gallbladder) cancer patients with selected patient and disease characteristics. This study estimates the overall survival benefit for the whole population of gastrointestinal cancer patients in Australia if evidence-based guidelines for chemotherapy were followed Methods: Decision trees with evidence-based indications for chemotherapy have been previously defined. Additional branches were defined in cases where new sub-groups benefit from chemotherapy. For all defined indications, the highest level of clinical evidence available was identified. Multiple electronic citation databases were systematically queried, including Medline and the Cochrane Library. In cases where there were multiple sources of the same level of evidence, hierarchical meta-analysis was performed. The benefits of chemotherapy were estimated for 1, 5 and 10-year survival. To assess the robustness of our estimates, sensitivity analyses were performed. Results: The estimated 1-year, 5-year, and 10-year absolute population-based overall survival benefits of optimally utilized chemotherapy for gastrointestinal cancer in Australia are 6.5%, 4.1% and 4.0%, respectively. They are summarized in the below table. GIT cancer 1-yr benefit 5-yr benefit 10-yr benefit Colon 6.0% 4.4% 5.0% Rectum 4.2% 8.2% 9.0% Oesophagus 4.4% 0.6% 0.0% Stomach 11.2% 3.1% 4.0% Pancreas 10.5% 0.2% 0.0% Liver 6.1% 0.0% 0.0% Gallbladder 10.6% 1.0% 0.0% Total GIT Cancer 6.5% 4.1% 4.0% Conclusions: Chemotherapy provides a modest survival benefit to the gastrointestinal cancer patient population in Australia when it is used in accordance with guideline recommendations. Disclosure: All authors have declared no conflicts of interest.