ITH of the primary tumor	The proportion of patients harbouring variable levels of ITH of the primary	8/15 patients with multi-regional samples from the primary tumor, 53%
ITH of the liver metastasis	The proportion of patients harbouring variable levels of ITH of the liver metastatic lesion	9/16 patients with multi-regional samples from liver metastases, 56%
Genetic heterogeneity between primary and matched metastatic	The proportion of patients harbouring mutations detected either in the primary or the matched metastatic tumor but not in both	9/17 patients with matched primary and metastatic tumor samples, 53%
De novo mutations in liver metastases	The proportion of patients harboring de novo mutations in liver metastases not found in the primary tumor, including potentially actionable variants indicating dynamic clonal evolution	6/17 patients with matched primary and metastatic tumor samples, 35%
Detection of cfDNA mutations	Proportion of patients in which tumor mutations were detected in cfDNA pre- operatively Proportion of patients in which tumor mutations were detected in cfDNA post- operatively	8/16 patients, 50% 2/16 patients, 12.5%
Potentially actionable mutations	9/18 (50%) and 17/18 (94%) patients could benefit from repurposing of already approved drugs and agents under clinical or pre-clinical development respectively Overall, an average of 3.2 actionable mutations per patient were identified.	12/17 (71%) patients with detectable mutations harbored potentially actionable alterations not ubiquitously shared by all tumor samples, indicating the need for spatiotemporal sampling to increase therapeutic accuracy