Impact of increased use of adjuvant chemotherapy in non-small cell lung cancer: A population and economic assessment.

Abstract

e16630 Background: Lung cancer confers a substantial burden on patients and healthcare systems. Among patients with resectable non-small cell lung cancer (NSCLC), adjuvant chemotherapy (AC) has improved survival, but uptake has been incomplete. To determine the impact of increased use of AC, microsimulations were undertaken using the Cancer Risk Management (CRM) model developed by Statistics Canada and the Canadian Partnership Against Cancer. Methods: The CRM model uses inputs from Canadian demographics, risk factors, direct healthcare costs, and clinical practice data to assess incremental life expectancy (LE), health-adjusted life expectancy (HALE), and cost from the perspective of the Canadian healthcare system. Sensitivity analysis examined rates of use of AC over a range of 0-100% in 10% increments. The baseline regimen tested was 4 cycles of cisplatin and vinorelbine (PV) assuming a 50% uptake with a hazard ratio (HR) of 0.68 for overall mortality. Sensitivity analysis with HR=0.74 was performed. Alternate regimens included cisplatin and etoposide (given for 3 or 4 cycles (PE3, PE4) and assessing a HR range of 0.95 to 0.68), gemcitabine with cisplatin (PG) or carboplatin (CG) (HR=0.68) and baseline PV plus bevacizumab q3 weeks to 1 year (EPB). Results: In the PV baseline scenario, LE was 76.2 years, HALE was 66.2 years and the lifetime cost per patient was $32,290. Adjuvant chemotherapy remained cost-effective under the sensitivity analysis (HR=0.74). Each 10% increase in the use of AC resulted in an average increase of 3.7 months in LE and 2.7 months in HALE, and a lifetime cost savings of up to $360 per patient. Compared with baseline PV, the PE regimen was not cost-effective, with the exception of PE3 assuming HR = 0.68. PG and CG were less cost-effective than PV, at an additional lifetime cost of $589 and $1023 per patient, respectively. The EPB scenario would be cost-effective at $100,000 per QALY compared to baseline PV if the survival HR=0.55. Conclusions: The CRM model suggests greater use of PV AC in NSCLC would improve LE and net healthcare costs, with other drug pairs being less cost-effective. The HR required for bevacizumab to be cost-effective will be difficult to achieve in current trials.