A decision analysis model predicts the optimal treatment pathway for patients with colorectal cancer and resectable synchronous liver metastases

Background: The estimated 2400 Americans who annually present with colorectal cancer and simultaneous resectable liver metastases encounter a wide array of surgical and medical treatment options. Because of the large number of possible treatment sequences and the absence of clinical trials comparing these various pathways, there is no consensus on the optimal therapeutic strategy. Materials and Methods: To address this issue, a decision-making model was developed incorporating all possible combinations of the following treatments: colorectal resection, hepatic resection, simultaneous colohepatic resection, and systemic chemotherapy. Transition probabilities associated with each treatment were determined by systematic review of the literature. Variations in complication rates based on the extent of hepatectomy (minor: 1-2 segments vs. major: > 2 segments) were factored into the model. Sensitivity analyses were performed to identify threshold values for study variables that altered the optimal treatment pathway. Results: After 10,000 simulated patient trials with no bias toward any one initial treatment (ie, current practice conditions), the global calculated 5-year survival rate was 21%. For simulated patients with moderate hepatic tumor burden, only treatment sequences that placed systemic therapy before major hepatectomy resulted in improved 5-year survival projections (38% vs. 29%; P=.001; odds ratio, 1.82). Initial treatment with simultaneous colohepatic resection was only favored when the operative mortality rate was adjusted to < 0.5%. Conclusion: This detailed decision-making analysis predicts that the optimal treatment pathway for most patients with colorectal cancer and simultaneous resectable liver metastases is pre-operative systemic therapy followed by colohepatectomy or 2-stage resection. In the era of improved systemic therapies, major hepatic resection should be deferred until local and systemic disease can been addressed.