Purpose: Recently presented results from a randomized controlled trial show survival benefit of pemetrexed (Pem) maintenance therapy over placebo in patients with non-small cell lung cancer (NSCLC) (Belani Proc ASCO 2009). Here, we estimate the cost-effectiveness of Pem versus placebo in an overall NSCLC population and in histology subgroups.
Method: A 3-year semi-Markov model was developed in Excel to compare the impact of Pem plus best supportive care (BSC) to placebo plus BSC from the US payer perspective. Because hazard rates showed substantial temporal variation, 52 three-week cycles were modeled. Data from the trial provided clinical inputs used in the model including survival, progression, and serious adverse event rates. Medicare reimbursement rates were used to determine the cost of Pem. A retrospective claims database analysis was used to estimate costs of chemotherapy administration, lab monitoring, subsequent therapies, direct care for disease- and therapy-related morbidity, and end-of-life care. Non-medical and indirect costs were not included in the model. Because of the increasing importance of histology in identifying the most appropriate candidates for Pem therapy, this analysis defined three patient populations: 1) all patients with advanced NSCLC, 2) patients with nonsquamous NSCLC (adenocarcinoma, large cell, or other), and 3) patients with adenocarcinoma or large cell carcinoma. One-way and probabilistic sensitivity analyses evaluate the importance of uncertainty in various inputs and the overall variability. Simple discounting was performed at 3% per year.
Result: In all patients regardless of histology, using Pem led to an incremental cost per life-year gained (ICER) of $205,609 vs. placebo. In the subset of patients with nonsquamous NSCLC, the ICER is $122,364, while for the third population the ICER is $113,306. For the nonsquamous population, 9.4% of the Pem group vs. 5.6% of the placebo group are estimated to survive >3 years. Assuming per-patient costs and survival are constant after year two, the lifetime ICER for the nonsquamous group is $110,168. Only altering assumptions related to Pem cost, such as limiting the duration of maintenance, substantially change the estimated ICERs.
Conclusion: In an unselected advanced NSCLC population, Pem maintenance therapy is unlikely to be considered cost-effective; however, in histology subgroups the ICERs are comparable to other reimbursed cancer treatments. This analysis emphasizes the importance of histology in identifying appropriate patients for Pem maintenance chemotherapy.
Candidate for the Lee B. Lusted Student Prize Competition