KRAS Mutations in Unresectable Colorectal Liver Metastases Linked to Reduced Response Rates

Based on a single-institution study, investigators were able to correlate outcomes for unresectable colorectal liver metastases with the addition of hepatic arterial infusion pump therapy to systemic therapy to the presence of a mutation in KRAS.

A retrospective review examining the effects of KRAS mutations on tumor response in unresectable colorectal liver metastases revealed worse overall outcomes and lower rates of conversion to resection with combination hepatic arterial infusion (HAI) pump therapy and systemic chemotherapy when compared against wild-type KRAS, according to a presentation at the Society of Surgical Oncology 2021 International Conference on Surgical Cancer Care.1

Despite these findings, the investigators determined that patients with KRAS mutations still derived benefit from receiving HAI.

“The effects of KRAS mutations on response to HAI therapy in patients with unresectable colorectal liver metastases is not well known,” Hordur M. Kolbeinsson, MD, of Spectrum Health/Michigan State University General Surgery Residency, Grand Rapids, MI, said during a presentation of the data. “We hypothesized that KRAS-wild­ type patients would have a more pronounced response to HAI therapy compared with patients with KRAS mutations.”

Kolbeinsson explained the rationale for this investigation, including results of a case-control study comparing combination systemic chemotherapy plus HAI versus systemic chemotherapy alone that demonstrated better survival outcomes in patients receiving HAI.2 Additionally, the presence of a KRAS gene mutation in colorectal cancer is associated with worse prognosis for liver metastases as well as shorter survival following resection.3,4

The single-institution, retrospective cohort study included all patients with unresectable colorectal liver metastases treated with HAI between August 2017 and September 2020. Patients were stratified by their mutation status and the primary end point was objective response rate (ORR) with secondary outcomes measures of response overall magnitude and the rate of conversion to respectability. The definition of respectability in this study was 8 or fewer lesions that could be resected while maintaining 2 continuous uninvolved liver segments. Standardized mutational pattern testing was done for KRAS, NRAS, and BRAF mutations during the study timeframe.

HAI therapy used was floxuridine 0.12 mg/kg daily adjusted for ideal body weight and dexamethasone administered in 28-day cycles given in conjunction with systemic chemotherapy.

Of the 25 patients meeting the criteria for inclusion, 13 had a wild-type KRAS gene and 11 had a KRAS mutation; 1 had a BRAF mutation and was excluded from the analysis. Prior chemotherapy was noted in all but 1 patient, with 14 having experience with 1 prior regimen and 10 experiencing 2 prior regimens.

Patients in the KRAS-mutant group had a median age of 57.3 years (+/- 12.8), 28% were men, the majority were White (91%), and the median BMI was 27.8 (range, 20.3-31.1). Patients in the KRAS wild–type group had a median age of 55.0 years (+/- 11.8), 69% were men, the majority were White (85%), and the median BMI was 27.3 (range, 21.3-40.1).

Notably, median operative duration was somewhat longer in the KRAS mutant group at 250 minutes (range, 197-474) versus the KRAS wild­–type group at 228 minutes (range 122-201) as was length of stay at 7 days (range, 3-13) and 3 days (range, 2-11), respectively. The corresponding median number of liver lesions was 11 (range, 3-59) and 12 (range, 1-54).

The median ORR in the KRAS mutant and wild-type groups were 64% and 100%, respectively (P = .03), and met the prespecified threshold for significance of P-value less than 0.05. The median overall magnitude of response was 58% for patients with mutations and 70% for patients with wild-type KRAS (P = .04). Almost two-thirds of patients (62%) in the wild-type group converted to respectability with HAI, but only 18% of those in the KRAS mutation–positive group had the same benefit (P = .05).

“This is hardly unexpected given what we know of the effects of KRAS mutations on overall outcomes in colorectal cancer,” Kolbeinsson said. “Nevertheless, the objective response rate in the KRAS-mutant cohort was 64%, which has to be considered a favorable response given this was third-line treatment for 4 out of the 11 patients in that group.”

Kolbeinsson pointed out that while these data have important implications for patient education and prognostics, they should not be used to guide patient selection for HAI chemotherapy.

“Future studies should investigate the most appropriate timing of hepatic artery infusion chemotherapy in the setting of unresectable colorectal liver metastases,” Kolbeinsson said.


1. Kolbeinsson HM. KRAS Mutation Predicts Magnitude of Response and Outcomes in Hepatic Arterial Infusion Pump Therapy of Unresectable Colorectal Liver Metastases. Presented at: Society of Surgical Oncology 2021 International Conference on Surgical Cancer Care; March 18-19, 2021; Virtual. Abstract 39

2. Dhir M, Jones HL, Shuai Y, et al. Hepatic Arterial Infusion in Combination with Modern Systemic Chemotherapy is Associated with Improved Survival Compared with Modern Systemic Chemotherapy Alone in Patients with Isolated Unresectable Colorectal Liver Metastases: A Case-Control Study. Ann Surg Oncol. 2017;24(1):150-158. doi: 10.1245/s10434-016-5418-6

3. Tosi F, Magni E, Amatu A, et al. Effect of KRAS and BRAF Mutations on Survival of Metastatic Colorectal Cancer After Liver Resection: A Systematic Review and Meta-Analysis. Clin Colorectal Cancer. 2017;16(3):e153-e163. doi: 10.1016/j.clcc.2017.01.004

4. Gholami S, Kemeny NE, Boucher TM, et al. Adjuvant Hepatic Artery Infusion Chemotherapy is Associated With Improved Survival Regardless of KRAS Mutation Status in Patients With Resected Colorectal Liver Metastases: A Retrospective Analysis of 674 Patients. Ann Surg. 2020;272(2):352-356. doi: 10.1097/SLA.0000000000003248