OBJECTIVES: In cost-effectiveness analysis in oncology, relative effects of competing treatments are rarely available from head-to-head trials. For tyrosine kinase inhibitors ponatinib and bosutinib, indicated for third-line CP-CML, efficacy data are derived from single-arm trials, making unfeasible the application of traditional network meta-analysis approaches. Matching-adjusted indirect comparison (MAIC) is a novel method that can be used to overcome an incomplete evidence network. METHODS: An MAIC analysis was conducted to compare efficacy in the subgroup of third-line CP-CML patients in the ponatinib trial (PACE) and bosutinib trial (NCT00261846; published by Khoury 2012). Individual patients treated with ponatinib in the PACE study were assigned weights obtained from a logistic regression model for the propensity of being treated with bosutinib vs. ponatinib. The mean baseline characteristics in the reweighted PACE data were expected to closely match those reported for patients treated with bosutinib. The weights were used to produce matching-adjusted estimates of ponatinib best response rates, in terms of complete (CCyR) and partial (PCyR) cytogenetic response, complete hematologic response (CHR), and no response (NR). RESULTS: Baseline characteristics, including gender, race, median age, median time since diagnosis, ECOG performance status, and T315I mutation at study entry of the adjusted ponatinib cohort matched those of bosutinib. The effective sample size from PACE decreased from 97 to 69 patients, as a consequence of matching. Best response rates in the matched-adjusted ponatinib cohort (CCyR: 61.34%, PCyR: 8.46%, CHR: 18.19%, and NR: 12.01%) were consistent with unadjusted results (CCyR: 64.95%, PCyR: 6.19%, CHR: 17.53%, and NR: 11.34%), with considerably more ponatinib patients in deeper response categories versus bosutinib (CCyR: 24.07%, PCyR: 8.33%, CHR: 37.93%, and NR: 29.66%). CONCLUSIONS: Adjusted ponatinib response rates in third-line CP-CML were similar to unadjusted outcomes, suggesting that the higher efficacy observed in ponatinib versus bosutinib trial is not due to cohort imbalances.