Perioperative Hepatic Arterial Infusion Pump Chemotherapy After Resection of Colorectal Liver Metastases


Key Points

  • Hepatic arterial infusion improved overall survival among all patients and most subgroups, including those receiving preoperative or adjuvant modern chemotherapy.
  • The benefit seemed to be particularly marked in patients with node-negative primary disease and in those with lower clinical risk scores.

A propensity score analysis in a prospectively maintained database at Memorial Sloan Kettering Cancer Center (MSKCC) has shown that use of perioperative hepatic arterial infusion pump chemotherapy after complete resection of colorectal liver metastases is associated with a marked improvement in overall survival compared with systemic chemotherapy without hepatic arterial infusion. The study was reported by Koerkamp et al in the Journal of Clinical Oncology.

Study Details

The study involved 2,368 consecutive patients who underwent complete resection of metastases between 1992 and 2012, including 785 (33%) who received hepatic arterial infusion and 1,583 (67%) who did not. All patients who received hepatic arterial infusion also received perioperative systemic chemotherapy. Systemic chemotherapy regimens varied over time, as determined by treating physicians on the basis of guidelines, chemotherapy history, and ongoing clinical trials.

Modern chemotherapy was defined as regimens containing oxaliplatin or irinotecan. All patients receiving hepatic arterial infusion were scheduled to receive fluorouracil with or without additional chemotherapy (eg, irinotecan or oxaliplatin). The hepatic arterial infusion pump was typically removed after 2-year follow-up, with earlier removal occurring in rare cases of patient discomfort or pump loss of function. Overall, 2,193 patients (93%) received preoperative or adjuvant systemic chemotherapy, with 1,196 patients (51%) receiving both.

The propensity score analysis for overall survival matched hepatic arterial infusion and non–hepatic arterial infusion patients for seven known independent prognostic factors: sex, age, year of resection, presence of extrahepatic disease, number of resected or ablated tumors, size of largest resected tumor, and margin status.

Patients receiving hepatic arterial infusion were more likely to have advanced disease (such as N2 disease; P = .02), higher number of tumors (eg, mean = 3.5 vs 2.4; P < .001), synchronous colorectal liver metastases (60% vs 45%, P < .001), higher rates of two-stage resection (5.6% vs 0.5%, P < .001), greater use of intraoperative ablation (17.3% vs 6.1%, P < .001), treatment during the “recent period” (2003–2012; 67% vs 49%, P <.001), and treatment with preoperative modern chemotherapy (48% vs 28%, P <.001). Patients who did not receive hepatic arterial infusion group were older (median = 62 vs 56 years, P <.001) and had a higher rate of extrahepatic disease (9.7% vs 5.4%, P < .001).

Survival Outcomes

Median follow-up was 55 months. Median overall survival was 67 months in the hepatic arterial infusion group vs 44 months in the no–hepatic arterial infusion group (P < .001), despite more advanced disease in the former. Overall survival was 52.9% vs 37.9% at 5 years (P < .001) and 38.0% vs 23.8% at 10 years (P < .001). The propensity score–adjusted analysis yielded an adjusted hazard ratio of 0.67 (P < .001).


Median overall survival was 67 months vs 47 months (P < .001) among the 1,442 patients who received modern systemic chemotherapy, 68 months vs 42 months among 879 not receiving modern chemotherapy (P < .001), 72 vs 51 months (P < .001) among 1,295 treated between 2003 and 2012, and 60 vs 40 months (P < .001) among 1,073 treated between 1992 and 2002. Median overall survival was 77 vs 45 months (P < .001) among 812 who received preoperative modern chemotherapy and 55 vs 43 months (P = .004) among 1,556 who did not and 68 vs 49 months (P < .001) among 1,124 who received adjuvant modern chemotherapy and 59 vs 41 months (P < .001) among 1,244 who did not.

Among subgroups according to disease characteristics, marked differences in overall survival were observed among 908 patients with node-negative colorectal cancer (median = 129 vs 51 months, P < .001) and among 1,252 with low clinical risk scores of 0 to 2 (median = 89 vs 53 months, P < .001). Other subgroups for which median overall survival with hepatic arterial infusion exceeded that with no hepatic arterial infusion by 24 months included patients aged 65 years and those with metachronous tumors, those with a disease-free interval of 1 year, and those with ≤ 4 tumors resected.

There was a significant difference among 2,173 patients without extrahepatic disease (median = 68 vs 46 months, P <.001), but not among 195 patients with extrahepatic disease (median = 37 vs 33 months, P = .92). The only other subgroups in which hepatic arterial infusion was not associated with a significant survival advantage were patients with a positive resection margin (n = 207; median = 37 vs 28 months, P = .21) and those with clinical risk scores of 4 (n = 196; median = 37 vs 42 months, P = .29) or 5 (n = 44; median = 28 vs 21 months, P = .35).

The investigators concluded: “Patients who received [hepatic arterial infusion] had a median [overall survival] of approximately 2 years longer than patients without [hepatic arterial infusion]. The strong association was independent of the use of modern systemic chemotherapy and remained in propensity score analysis. Patients with node-negative primary tumors or a low clinical risk score seemed to benefit most from [hepatic arterial infusion].”

The study was supported by the Dutch Cancer Society and the National Cancer Institute.

Bas Groot Koerkamp, MD, PhD, of Erasmus MC Cancer Institute and Memorial Sloan Kettering Cancer Center, is the corresponding author of the Journal of Clinical Oncology article. 

The content in this post has not been reviewed by the American Society of Clinical Oncology, Inc. (ASCO®) and does not necessarily reflect the ideas and opinions of ASCO®.