OBJECTIVES: To assess relative efficacy and safety of second-line treatments in chronic myeloid leukaemia (CML), a systematic review (SR) and network meta-analysis (NMA) feasibility study were conducted. METHODS: A SR was conducted in January 2015 (Embase, MEDLINE, Cochrane Library, Clintrials.gov and conferences) to identify comparative trials evaluating treatment outcomes in patients with CML previously treated with tyrosine kinase inhibitors. Eligible studies were examined to assess NMA feasibility. RESULTS: Twenty-three publications relating to six randomised controlled trials (RCTs) on second-line treatment met the eligibility criteria. Included studies compared either nilotinib (n=3) or dasatinib (n=1) with imatinib, or studied dasatinib at alternative doses (n=2). No comparative bosutinib or ponatinib studies were identified. Efficacy outcomes were reported using various definitions and different time points. Compared with nilotinib, significantly fewer imatinib treated patients with complete cytogenetic response (CCyR) at baseline, achieved complete molecular response (CMR) (23% vs 11%, p=0.02) by 12 and confirmed CMR (22.1% vs 8.7%, p=0.0087) by 24 months and in patients without major molecular response (MMR) at baseline, MMR by 12 (75% vs 36%, p=0.006) and 24 (83.3% vs 53.6%, p=0.0342) months. Compared with imatinib, significantly more dasatinib patients achieved CCyR (16% vs 40%, p=0.004; 18% vs 44%, p=0.0025), MMR (4% vs 16%, p=0.038; 12% vs 29%, p=0.028) and complete haematologic response (82% vs 93%, p=0.034; 82% vs 93%, p=0.0341) at 15 and 24 months, respectively. Interpretation of safety data was inconclusive due to its limited availability and treatment exposure differences. Even considering prospective non-RCTs, NMA was not feasible due to missing network links, significant differences between trial populations, and varying follow-up times. CONCLUSIONS: Review of all published comparative studies on second-line treatment of CML confirms that, based on direct efficacy results, dasatinib and nilotinib are the second line agents of choice. NMA comparing nilotinib and dasatinib was not feasible.