Abstract
Background Multiple myeloma (MM) is a neoplastic disease of plasma and the second most common hematological cancer. The malignancy is incurable, but the introduction of both lenalidomide and bortezomib has improved survival outcomes for patients with MM. Now there is a need for a treatment for patients who become refractory to lenalidomide and bortezomib. Pomalidomide has shown efficacy and acceptable safety, but since health care resource are scarce, it is important to know if pomalidomide is good value for money using common benchmarks for cost-effectiveness of the National Health Service (NHS) in the United Kingdom (UK), which is £30,000 per QALY gained. Because information about the effectiveness of pomalidomide is scarce, information about the patient population that most resembles the population that will receive pomalidomide is determined. Methods A Markov model was built to assess the cost-effectiveness of pomalidomide plus low-dose dexamethasone compared to high-dose dexamethasone. The usefulness of information on lenalidomide and bortezomib for the model on pomalidomide was tested by patient similarity through a survival analysis. Kaplan-Meier survival curves of lenalidomide and bortezomib were reconstructed and parameters for the Weibull equation were estimated through maximum likelihood estimation in order to reconstruct the survival. Results from the survival analysis show that only the information of trials on lenalidomide seem reliable to use in the model for pomalidomide. Data on survival from pomalidomide plus low-dose dexamethasone and from high-dose dexamethasone was obtained from the MM-003 trial. The model has a 10 years time horizon. The model is based on the different health states the patient can experience: progression-free state, progressed state, and death. In each cycle, a patient can transfer to another stay or remains in the same state. All states include costs and a determined quality of life. The Markov model calculates the total costs, life years, and QALYs gained over the full time horizon. Results The incremental gain in life-years is 0.38 years. The costs of treating patients with pomalidomide plus low-dose dexamethasone is almost 6 times as high than treating patients with high-dose dexamethasone (£99,134 versus £17,420). Therefore, the undiscounted ICER is £105,787 per life-year gained. By including quality of life in the model, the ICER of pomalidomide becomes more unattractive (£216,373 per QALY gained). With a standard threshold value for the ICER of £30,000 per QALY gained, pomalidomide would not be considered cost-effective. Uncertainty Most data inputs of the model are uncertain. One-way sensitivity analysis was performed to show the impact of single parameters on the ICER. The cost and utility parameters of pomalidomide showed the greatest impact. A probabilistic sensitivity analysis (PSA) was performed to assess the robustness of the deterministic model. All parameters were altered according to their distribution. Cost parameters were assumed to have a gamma distribution, utility parameters a beta distribution. Uncertainty surrounding the highly correlated survival parameters was obtained by a Cholesky decomposition assuming a bivariate Normal distribution. Results of the PSA showed almost no variation from the deterministic model. A threshold analysis was performed to seek the appropriate costs of pomalidomide for an ICER under the NHS threshold. It was found that for no price of pomalidomide, the ICER would be acceptable. Discussion There was no individual patient level data available to build a micro simulation model, nor information about individual patient characteristics that may influence the outcomes in terms of the ICER. The results of the sensitivity analyses are largely based on the choice of the distributions of parameters. Transferability of the model to other settings may be difficult, because other countries may have no threshold ICER or take another perspective in their analyses. Because pomalidomide can be considered as an end-of-life drug and there is only a small amount of patient eligible for receiving pomalidomide, other considerations than the ICER could play a role in the decision-making process. Conclusion The cost-effectiveness of pomalidomide plus low dose dexamethasone compared to high dose dexamethasone for patients with relapsed multiple myeloma refractory to both lenalidomide and bortezomib in the NHS setting is £216,373 per QALY gained. Pomalidomide is not considered cost-effective with a standard threshold value for the ICER of £30,000 per QALY gained.