Cost-effectiveness of systemic therapies for metastatic pancreatic cancer

Abstract

Purpose: Gemcitabine and capecitabine (gem-cap), gemcitabine and erlotinib (gem-e), and folfirinox (5-fluorouracil-leucovorin-irinotecan-oxaliplatin) are new treatment options for metastatic pancreatic cancer, but they are also more expensive and potentially more toxic than gemcitabine alone (gem). We conducted a cost-effectiveness analysis of these treatment options compared with gem.

Methods: A Markov model was constructed to examine costs and outcomes of gem-cap, gem-e, folfirinox, and gem in patients with metastatic pancreatic cancer from the perspective of a government health care plan. Ontario health economic and costing data (2010 Canadian dollars) were used. Efficacy data for the treatments were obtained from the published literature. Resource utilization data were derived from a chart review of consecutive metastatic patients treated for pancreatic cancer at Princess Margaret Hospital, Toronto, Ontario, 2008-2009, and supplemented with data from the literature. Utilities were obtained by surveying medical oncologists across Canada using the EQ-5D. Incremental cost-effectiveness ratios (icers) were calculated.

Results: The icers for gem-cap, gem-e, and folfirinox compared with gem were, respectively, CA$84,299, CA$153,631, and CA$133,184 per quality-adjusted life year (qaly). The model was driven mostly by drug acquisition costs. Given a willingness-to-pay (wtp) threshold greater than CA$130,000/qaly, folfirinox was most cost-effective treatment. When the wtp threshold was less than CA$80,000/qaly, gem alone was most cost-effective. The gem-e option was dominated by the other treatments.

Conclusions: The most cost-effective treatment for metastatic pancreatic cancer depends on the societal wtp threshold. If the societal wtp threshold were to be relatively high or if drug costs were to be substantially reduced, folfirinox might be cost-effective.

Keywords: Pancreatic cancer; chemotherapy; cost-effectiveness.

FIGURE 1

The Markov model.

FIGURE 2

Probabilistic sensitivity analysis. (A) Incremental cost-effectiveness scatterplot (n = 10,000). The slope of the broken line represents a willingness-to-pay (wtp) threshold of $100,000 per quality-adjusted life-year (qaly). Each point is a simulated result of the probability sensitivity analysis. Points below the line are considered cost-effective with a wtp threshold of $100,000 per qaly. Only 5% of the simulations fell below that line. The oval denotes a region where 95% of the simulations occur. (B) Cost-effectiveness acceptability curve. This plot shows the probability of various metastatic pancreatic cancer treatments being cost-effective at various wtp thresholds. Gem = gemcitabine; Cap = capecitabine; Erl = erlotinib.

FIGURE 3

Cost-effectiveness plane. The red triangle point that represents gemcitabine plus erlotinib (GemErl) lies to the right of the line joining the blue diamond point [gemcitabine plus capecitabine (GemCap)] and the light blue square point (folfirinox), meaning that GemErl was dominated by extended dominance. qaly = quality-adjusted life-year.