Cost-effectiveness analysis of docetaxel versus weekly paclitaxel in adjuvant treatment of regional breast cancer in New Zealand.

BACKGROUND: There have been recent important changes to adjuvant regimens and costs of taxanes for the treatment of early breast cancer, requiring a re-evaluation of comparative cost effectiveness. In particular, weekly paclitaxel is now commonly used but has not been subjected to cost-effectiveness analysis. AIM: Our aim was to estimate the cost effectiveness of adjuvant docetaxel and weekly paclitaxel versus each other, and compared with standard 3-weekly paclitaxel, in women aged ≥25 years diagnosed with regional breast cancer in New Zealand.
METHODS: A macrosimulation Markov model was used, with a lifetime horizon and health system perspective. The model compared 3-weekly docetaxel and weekly paclitaxel versus standard 3-weekly paclitaxel (E1199 regimen) in the hospital setting. Data on overall survival and toxicities (febrile neutropenia and peripheral neuropathy) were derived from relevant published clinical trials. Epidemiological and cost data were derived from New Zealand datasets. Health outcomes were measured with health-adjusted life-years (HALYs), similar to quality-adjusted life-years (QALYs). Costs included intervention and health system costs in year 2011 values, with 3% per annum discounting on costs and HALYs.
RESULTS: The mean HALY gain per patient compared with standard 3-weekly paclitaxel was 0.51 with weekly paclitaxel and 0.21 with docetaxel, while incremental costs were $NZ 12,284 and $NZ 4,021, respectively. The incremental cost-effectiveness ratio (ICER) of docetaxel versus 3-weekly paclitaxel was $NZ 19,400 (purchasing power parity [PPP]-adjusted $US 13,100) per HALY gained, and the ICER of weekly paclitaxel versus docetaxel was $NZ 27,100 ($US 18,300) per HALY gained. In terms of net monetary benefit, weekly paclitaxel was the optimal strategy for willingness-to-pay (WTP) thresholds >$NZ 27,000 per HALY gained. However, the model was highly sensitive to uncertainty around survival differences, while toxicity-related morbidity had little impact. Thus, if it was assumed that weekly paclitaxel and docetaxel had the same efficacy, docetaxel would be favoured over weekly paclitaxel.
CONCLUSION: Both weekly paclitaxel and docetaxel are likely to be cost effective compared with standard 3-weekly paclitaxel. Weekly paclitaxel was the optimal choice for WTP thresholds greater than $NZ27,000 per HALY gained (PPP-adjusted $US 18,000). However, uncertainty remains around relative survival benefits, and weekly paclitaxel becomes cost ineffective versus docetaxel if it is assumed that the two regimens have equal effectiveness. Reduced uncertainty about the relative survival benefits may improve decision making for funding.