To evaluate the effectiveness of imatinib as first-line treatment for chronic myeloid leukaemia (CML) compared with interferon-alpha (IFN-alpha), hydroxyurea and bone marrow transplantation (BMT), and the cost-effectiveness of imatinib compared with IFN-alpha and hydroxyurea.
Selected studies and full-text articles were screened and rigorously selected. Survival was the key outcome measure. Surrogate outcome measures included haematological (blood) response and cytogenetic (bone marrow) response (CR). As no published cost-effectiveness studies were found that compared imatinib and IFN-alpha, an independent Markov model was constructed and this was compared with models submitted to the National Institute for Clinical Excellence by the manufacturer of imatinib.
Intention-to-treat analysis showed that imatinib was associated with complete CR at 12 months follow-up of 68% compared with 20% for the IFN-alpha plus Ara-C group. The estimated proportion of people taking imatinib who had not progressed to accelerated or blast phases at 12 months was 98.5%, and 93.1% for IFN-alpha plus Ara-C. Overall survival was not statistically significantly different. Withdrawal due to side-effects was 2% for imatinib and 5.6% for IFN-alpha plus Ara-C. Cross-over due to intolerance was 0.7% and 22.8% for imatinib and for IFN-alpha plus Ara-C, respectively. Quality of life was better in the imatinib group than the IFN-alpha group when assessed at 1, 3 and 6 months. Median survival across the four IFN-alpha versus hydroxyurea studies was 66 and 56.2 months, respectively. Median complete CR was 6% for IFN-alpha and 0 for hydroxyurea. Median withdrawal due to side-effects was 24% and 4% for IFN-alpha and hydroxyurea, respectively. Four out of the five studies comparing BMT and IFN-alpha showed a long-term survival advantage for BMT over IFN-alpha, but a short-term disadvantage. In four of the five studies comparing BMT and IFN-alpha, median survival had not yet been reached in the BMT groups in 6--10 years. Median survival in the IFN-alpha arms ranged from 5.2 to 7 years. The BMT group gained a survival advantage over IFN-alpha at 3--5.5 years. In the BMT group death due to transplant-related complications ranged from 36 to 45%. The incremental cost-effectiveness ratio (ICER) of imatinib compared with IFN-alpha from the independent model was GBP26,180 per quality-adjusted-life-years (QALY) gained and was relatively robust. Imatinib was less cost-effective than hydroxyurea with an ICER of GBP86,934.
Imatinib appears to be more effective than current standard drug treatments in terms of cytogenetic response and progression-free survival, with fewer side-effects. However, there is uncertainty concerning longer term outcomes, the development of resistance to imatinib, the duration of response and the place of imatinib relative to BMT. New issues are continually arising, such as optimal management pathways and combination therapies. Recommendations for research include: long-term follow-up data from the first- and second-line imatinib trials; investigation into specific subgroups, e.g. high-risk patients, the elderly, children or those eligible for BMT; long-term comparisons of imatinib with BMT performed in early stages of CML; the use of imatinib in combination with other therapies, and further detailed economic studies. Investigation of the impact of CML and imatinib on quality of life is also important.