Cost-effectiveness of systemic and regional chemotherapy for the treatment of patients with unresectable colorectal liver metastases.

BACKGROUND Management of unresectable colorectal liver metastases (CLM) can be by regional (hepatic arterial infusion [HAI]) or systemic chemotherapy, or by symptom control alone. In this study the costs of each type of management were related to clinical outcome in 134 patients with CLM. METHODS The costs (both in terms of health care and to society) and benefits (treatment-added survival and normal quality of life survival) of chemotherapy treatment of 85 patients (HAI with implanted pump: 51 patients; and systemic chemotherapy: 34 patients) were compared with those in 49 patients managed by symptom control only. RESULTS HAI chemotherapy cost the most (Pound Sterling 18,263 per patient) and symptom control the least (Pound sterling 2136 per patient). When survival was included, HAI was the most cost-effective treatment (health care cost per life year gained with HAI vs. systemic chemotherapy: Pound Sterling 24,604; systemic chemotherapy vs. symptom control: Pound Sterling 32,788), but there was no difference with regard to health care cost per normal quality of life gained. Societal costs incurred by lost work time and welfare payments during illness were higher for HAI (Pound Sterling 12,897) than systemic chemotherapy (Pound Sterling 9143) or symptom control (Pound Sterling 8090) because HAI-treated patients lived longer and, although working longer and contributing more productivity to society, lost more work days than other patients. CONCLUSIONS The least expensive management for CLM was symptom control, whereas systemic and HAI chemotherapies were equally cost-effective in producing added normal quality survival for health care resources expended. Although overall societal costs were higher for HAI than for either systemic chemotherapy or symptom control, the cost benefit was difficult to interpret because of uncertain attitudes regarding continued work during terminal illness.