from studies using a second 2GTKI after failure of the first 2GTKI are sparse, usually limited to
small numbers of patients and often confounded by failing to distinguish patients with
intolerance or resistance. The data that are available suggest that the achievement of CCyR,
irrespective of durability or long-term tolerability, are around 20%. Increasingly patients with
resistance to a 2GTKI are switching directly to ponatinib where the probability of CCyR is of
the order of 50%, although the same caveats of longer-tern efficacy and tolerability referred
to above, remain. Data were presented at the 2020 ASH meeting from the ASCEMBL study, in
which patients who had failed two or more prior TKI for any combination of resistance and
intolerance were randomised to asciminib or bosutinib. The endpoint of the study of major
molecular response at 24 weeks was somewhat demanding in a group pf patients who had
already failed at least two TKI, but was achieved in 25% of patients given asciminib compared
to 12% of those on bosutinib. Similarly CCyR rates at 24 weeks were almost twice as high in
the asciminib cohort at 41% compared to 24% on bosutinib. Although there is clearly no direct
comparison with ponatinib, it would seem that the efficacy of asciminib in this situation will
be similar to that of ponatinib and that both ponatinib and asciminib are more effective after
failure of a 2GTKI than an alternative 2GTKI. It is a logical conclusion therefore, that if the
2GTKI is used as first-line therapy, then a third-generation TKI (3GTKI) will become the
preferred second-line option for those who are resistant and allo-SCT becomes the third-line
option for patients who fail, for resistance or intolerance, the 3GTKI.
The use of allo-SCT is more problematic than the simple administration of an oral drug for
obvious reasons. The patients have to be transplant eligible, in terms of donor availability and
fitness for allografting, often but not always associated with age. However with the improved
outcome of haploidentical transplant, and the ability to effectively use reduced intensity
conditioning in patients with CML in chronic phase, where an increased increase in the risk of
relapse can be managed with donor lymphocyte infusions, allo-SCT becomes a realistic option
for the majority of patients who fail TKI. Consideration for allo-SCT should also be given to
younger patients who may be responding adequately but not optimally to a 3GTKI, say those
who have reached CCyR or MMR but who lack a continuing decline in BCR-AL1 transcript
levels, and who face decades of treatment with a drug with a troublesome side=effect profile.
In conclusion there are some patients and some situations where allo-SCT is the
recommended third-line treatment option, and failure to recognise this group exposes these
patients to the risk of disease progression and death. We recommend that a search for a
suitable donor should be initiated at the time of identification of either of resistance to a 2GTKI
or the emergence of a T315I mutation.