Importance of predicting CML therapy failure

In this issue of Blood, Zhang et al¹ propose a scoring system to predict the risk of treatment failure with tyrosine kinase inhibitors (TKIs) used in the first-line treatment of chronic myeloid leukemia (CML). This scoring system is expected to become at least as important as those already in use, such as Sokal’s and European Treatment and Outcome Study long-term survival (ELTS), which are able to predict overall survival (OS).

The reason is simple. Nowadays, the OS of patients with BCR::ABL1-positive CML is almost the same as that of a control population without this disease.² This has been achieved with the use of several BCR::ABL1 TKIs that have been introduced into CML therapy over the last 25 years. Starting with imatinib, several new generations of TKIs have been developed and classified as second-generation (nilotinib, dasatinib, bosutinib, radotinib, flumatinib) or third-generation (ponatinib, olverembatinib) TKIs and the more recently developed STAMP (specifically targeting the ABL myristoyl pocket) category, such as asciminib.³ These drugs differ in their potency of inhibition of the BCR::ABL1 TK, but also in their toxicity profile, with the more potent ones and those with simultaneous inhibition of multiple other kinases being associated with a higher number of side effects and toxicities. They also differ in that imatinib and the second-generation TKIs approved for first-line use are now generic and, depending on location and generic availability, maybe less expensive, whereas the third-generation TKIs and STAMP are more expensive and are currently only approved for third-line use or, in the case of ponatinib, second-line use in patients with the T315I mutation. However, based on the results of the Asc4First trial, in which asciminib showed superior efficacy compared with other investigator chosen first-line TKIs and a good safety profile, the drug is now awaiting approval as a first-line treatment.⁴ 

We know that the few patients who still die from CML-related causes die after progression from chronic phase to more advanced phases of the disease, particularly blast crisis. This occurs most commonly in patients who fail their first assigned TKI therapy. The most widely used criteria for failure are those established by the European Leukaemia Net in 2020, although a new edition of these recommendations is expected to be published in a few months, and the response criteria may be slightly modified following the publication of the article by Lauseker et al.^5,6 

After their previous publication,⁷ which was limited to patients treated with imatinib, Zhang et al in this new study show that the rate of treatment failure in the high- and intermediate-risk groups identified by their scoring system is lower in patients treated with second-generation TKIs than in those treated with imatinib, and even more discriminating compared with the ELTS and Sokal’s scores. As a result, this new scoring system could be very helpful for clinicians to make the right choice in selecting the more appropriate first-line TKI for a given patient.

Although we know from randomized trials comparing imatinib and second-generation TKIs as first-line therapy that the long-term OS of patients is indeed the same with both categories of drugs, it is very important to identify the risk of treatment failure with imatinib or second-generation TKIs. Treatment failure requires switching to another TKI if possible, which may be stressful and alarming for patients, as any change in TKI therapy is associated with at least a 50% risk of resistance or intolerance to the new drug.

Of course, this does not necessarily mean that all patients in the high- and intermediate-risk groups should be treated with second-generation TKIs or that all patients in the low-risk group should be treated with imatinib. There are many other aspects to consider, such as the patient’s profile, the toxicity associated with the use of a particular TKI, the patient’s preference, and also the patient’s ability and desire to achieve the molecular response criteria recommended to try to discontinue TKI therapy and achieve treatment-free remission in a shorter period. Nevertheless, this new scoring system, when validated in other populations characterized by a higher median age compared with that of Chinese patients with CML, will become another tool in the hands of clinicians to properly inform the patient about the risk and possible outcome of therapy with a specific TKI and to decide on the most appropriate first-line treatment for the individual patient.

Finally, this scoring system may become even more important in the future when new patented TKIs such as asciminib or others are approved as first-line therapy, but at a cost that will certainly be much higher than that of generic imatinib and second-generation TKIs. The ability to predict the risk of failure and the potential benefits of specific therapies in CML and to calculate the true cost-benefit ratio of a specific treatment is now becoming extremely important, particularly in low- and middle-income countries, as the number of patients with CML increases, both from the newly diagnosed and those on treatment for life, and the burden of their assistance is rising. This scoring system may therefore become very useful for this task in this evolving context.

Conflict-of-interest disclosure: G.S. declares no competing financial interest.