Tyrosine kinase inhibitor for CML: all the same?

In this issue of Blood Advances, Matsumura et al¹ from the Japan Adult Leukemia Study Group (JALSG) address an important question in the management of patients with chronic myeloid leukemia (CML) in chronic phase receiving initial therapy with tyrosine kinase inhibitors (TKIs). The design is unique in that it is the first study, to our knowledge, to directly compare, in a randomized setting, the efficacy of 2 second-generation TKIs (2G-TKI).

The results are important as they demonstrate the near-equal efficacy of nilotinib and dasatinib at all end points. These included the primary end point (molecular response 4.5 [MR4.5] by 18 months) and other secondary response and survival end points (progression-free, event-free, and overall survival). That the efficacy is similar with these 2 agents is not unexpected. Previous studies comparing dasatinib² and nilotinib³ with imatinib demonstrated similar relative improvements in early response and major MR (MMR), and MR4 and MR4.5 rates at various times with the 2G-TKI (see table). The safety profile, reported only through grade ≥3 adverse events (AEs) in the JALSG study, is comparable with the expected differences in the frequency of some AEs, with more myelosuppression and pleural effusion with dasatinib, and more elevation of liver and pancreatic enzymes with nilotinib. A direct comparison in a randomized trial as provided by Matsumura et al is important for confirmation. It stands out that efficacy seems to be better than expected based on historical data (per the JALSG design, the expected MR4.5 rates by 18 months were 21% and 9%, with nilotinib and dasatinib, respectively; the actual rates were 32.6% and 30.8%, respectively). This may suggest that the outcomes have improved over time, perhaps reflecting the growing expertise in the management of patients with all TKIs. However, the different outcomes compared with expectations highlight the hindrances of comparisons with historical controls, making this study particularly important. The equivalence in efficacy with nilotinib and dasatinib thus reinforces the suggestion that the selection of 2G-TKI in frontline therapy may be guided by factors other than the expected efficacy. These include the possible risks of known AEs with each drug, schedule of administration, convenience, and costs (particularly as some drugs become generic) among others. Although not directly addressed in this study, the results also suggest that a change in TKI from 1 2G-TKI to another based on the lack of achievement of a deep molecular response, a scenario frequently seen in clinical practice, is unlikely to have any real benefit for most patients and may only expose them to AEs that they had not previously experienced.

There are some questions, not addressed in this study, that remain open related to frontline therapy in CML. One of these is the dose. The doses used in the JALSG study are standard (ie, nilotinib 300 mg twice daily and dasatinib 100 mg once daily); therefore, the results reflect what is the standard of care. There is, however, growing interest in the use of lower doses of dasatinib. Studies using lower doses of dasatinib have demonstrated similar efficacy as reported with 100 mg daily, perhaps with lower toxicity.^5-7 Interestingly, the incidence of pleural effusion grade 3 to 5 was only 4.9% in the JALSG. Because only grade 3 to 5 AEs were collected, the overall incidence is unknown. However, for the most concerning pleural effusion events, the incidence was as low as reported with lower doses (0%, 2%, and 13.2% in 3 studies, respectively).^5-7 The incidence of pulmonary hypertension and arterio-occlusive events is also low in the JALSG. Because of the way the AEs were collected, they might be underrepresented, but they are still likely to be lower than expected. The JALSG study also does not report dose adjustments or actual dose intensity; therefore, we cannot assess whether dose reductions contributed to this lower rate of the most concerning events. The rate of treatment discontinuation by 36 months (dasatinib 35%) was comparable with what is reported by 5 years with dasatinib 100 mg daily (39%)² and also for nilotinib (JALSG 33.4% and Evaluating Nilotinib Efficacy and Safety in Clinical Trials-newly diagnosed patients [ENESTnd] 40.1%).³ 

Another question is whether we could expect the same equivalence with bosutinib, the other 2G-TKI available as frontline therapy. The results of a randomized study of bosutinib vs imatinib⁸ suggest a similar relative benefit compared with imatinib, as reported in similar randomized studies of dasatinib or nilotinib (see table). Based on these studies, it is reasonable to assume that a similar efficacy would be observed with bosutinib as with nilotinib and dasatinib. We also expect a known difference in the toxicity profile characteristic of bosutinib, with a higher incidence of diarrhea and elevation of liver enzymes (both mostly grade 1-2). A direct comparison of bosutinib with dasatinib and/or nilotinib would evidently be ideal to confirm this predicted equivalence but is perhaps not necessary or realistic.

A more challenging question that has recently emerged is the possible role of asciminib, given the recent results of yet another randomized trial in which asciminib produced a higher rate of response compared with imatinib or all available TKIs.⁹ In that study (ASC4FIRST), the comparison of asciminib with 2G-TKI was a secondary end point and not powered for efficacy. There is a numerical improvement in the rate of MMR at 48 weeks with asciminib (66% vs 57.8%; see table). The most attractive efficacy benefit would be an improvement in the eligibility and/or success rate for treatment-free remission. Early responses are predictive of the probability of deep molecular responses,¹⁰ but a longer follow-up is needed for confirmation. The safety profile of asciminib is another potential worthy benefit. In ASC4FIRST, the rate of treatment discontinuation with asciminib, a partial surrogate for the safety profile, was approximately half that for 2G-TKI (and for imatinib). Other studies are more specifically addressing the question of asciminib vs 2G-TKI to further define the possible value.

We have many options for the initial therapy in CML. 2G-TKI have established value compared with imatinib, and the results from this randomized study propose that 2G-TKI are 1 option with 3 faces separated more by safety profile and other features. It remains to be confirmed with a longer follow-up if we have in asciminib a third tier with further clinical value to patients.

Conflict-of-interest disclosure: J.C. received consulting fees from Novartis, Pfizer, Sun Pharma, Takeda, Terns Pharmaceuticals, and Bio-Path Holdings, and research support to his institution from Novartis, Sun Pharma, and Ascentage. L.V. declares no competing financial interests.