Second-generation TKIs, such as dasatinib and nilotinib, have been approved for first-line therapy for CML as well as second-line therapy for those with imatinib failure [5]. Dasatinib, a dual Src/ABL1-kinase inhibitor, is active against many imatinib-resistant mutants except the T315I [8, 90]. Nilotinib, with N-methylpiperazine moiety incorporated into imatinib, is a more potent compound against most of the clinically relevant BCR–ABL1 mutants, except the T315I mutant [8, 90, 91].

Ponatinib is a pan-TKI effective against the T315I mutation and has recently been approved for the treatment of patients who failed previous TKI therapy [5, 77, 78]. However, its use has to be carefully monitored against its toxicity since serious cardiovascular, cerebrovascular and peripheral vascular adverse events have been reported in the study, especially in the presence of risk factors including hypertension, hyperlipidaemia and diabetes [99].

Allosteric inhibitors, such as ABL-001, which target sites on BCR-ABL1 other than the ATP-binding site, have been shown to prevent TKI resistance in preclinical models, and the results of ongoing phase I studies are eagerly awaited [www.​clinicaltrials.​gov identifier: NCT02081378].

Combining TKIs with compounds targeted against different downstream signalling pathways has been widely studied in preclinical and clinical studies [57]. JAK2 inhibitors induced apoptosis and decreased survival of CML stem cells in cultures and mouse models, when given together with imatinib [58, 93]. Early-phase clinical trials of ruxolitinib in combination with TKIs in CML with residual disease have been carried out.

Another approach is to combine mammalian target of rapamycin (mTOR) inhibitors with TKIs in treating TKI-resistant mutants [79]. Rapamycin has been shown to be effective in overcoming imatinib resistance in cell lines with BCR-ABL1 amplification or ABL1-kinase mutation [40, 80, 81, 100]. Allosteric mTOR inhibitors, everolimus and temsirolimus, have also been studied in phase 1 trials for CML.

Histone deacetylase inhibitors (HDACi), such as suberoylanilide hydroxamic acid (SAHA), suppress BCR–ABL1 levels in TKI-resistant CML cells in vitro and in vivo [94, 95]. Panobinostat (LBH589) has recently been shown to eliminate quiescent CML LSCs that are otherwise resistant to elimination by imatinib in mouse models. Clinical studies of panobinostat in combination with imatinib are being carried out in CML patients with residual disease [82].

DNA methylation inhibitors (DNMTi) or demethylating agents have also been tested in CML patients with TKI resistance. Early clinical studies of decitabine treatment showed modest response in CML patients who developed resistance to imatinib but are now being tested in combination with TKIs [101, 102].

Therapy based on microRNA (miRNA) may also be another promising epigenetic approach as BCR–ABL1 regulates miRNA profiles and miRNA has targets on downstream signalling pathways in CML, such as PI3K/AKT/mTOR pathway [103].

Wnt/β-catenin pathway can be an attractive novel target for the elimination of CML stem cells [87, 96]. A study in mice showed that indomethacin, a cyclooxygenase inhibitor, reduced β-catenin and inhibited proliferation of LSCs. The hedgehog pathway is another potential target since smoothened (SMO) antagonists are found to effectively decrease replication in CML cell lines in vitro [88, 89].

An alternative approach is the use of arsenic trioxide, which impairs BCR-ABL1 stability and causes depletion of CML LSCs through autophagosomal degradation [104, 105]. Early-phase clinical trials with those small molecules in combination with TKIs for resistant CML patients are being carried out [www.​clinicaltrials.​gov identifier: NCT00250042 and NCT01397734].

Modification of inflammatory cytokines, leukotrienes and prostaglandins by targeting against lipoxygenase and cyclooxygenase pathways in bone marrow environment has been shown to be a feasible approach to suppress proliferation and induces apoptosis of human CML cells and mouse models [73, 74].

Since hypoxia and up-regulated HIF1a in bone marrow environment have been implicated in TKI resistance, targeting HIF1a and associated genes may be a potential strategy to overcome resistance and eliminate CML LSCs [76].

HSP90, a chaperone molecule for oncoproteins such as BCR-ABL1 can be inhibited by 17-allylamino-17-demethoxygeldanamycin (17-AAG) or tanespimycin [85, 86]. While studies using tanespimycin have been initiated, results of these studies are not publicly available.

Another potential therapeutic approach is inhibition of autophagy. Autophagy is a mechanism used by cancer cells to resist apoptosis and is characterized by intracellular formation of autophagosomes and breakdown/recycling of cell components [106]. CML cells can undergo autophagy to avoid apoptosis induced by TKIs and develop resistance to treatment. Compounds such as chloroquine and clarithromycin inhibit essential autophagy and increase the sensitivity of CML stem cells to treatment with TKIs [83, 84, 107].

With the advent of TKIs, patients with CML now enjoy increased survival with a reasonable quality of life. However, most patients will likely have to remain on long-term therapy with imatinib and second-generation TKIs. Even among those who achieve durable molecular response with imatinib, cessation of therapy results in recurrent disease in a significant majority. Accordingly, the outcomes of similar studies employing second-generation TKIs are underway.

The next lap of CML research will be directed at eliminating the CML LSCs and providing effective cure for CML off any drug therapy. With the current knowledge and understanding of CML pathogenesis, several targetable pathways have been identified. In practical terms, such studies will have to balance the possibility of long-term cure against toxicity.