How can we chop CML with an ASXL1? Free

In this issue of Blood, Shanmuganathan et al¹ report on the adverse prognosis conferred by the presence of ASXL1 mutations at the time of diagnosis in patients with chronic myeloid leukemia (CML), even when treated with more potent tyrosine kinase inhibitors (TKIs) including asciminib. Previous reports suggested poor outcomes in patients with ASXL1 mutations when treated with imatinib,² and later also when treated with second-generation TKI.³ 

CML had long been considered a “molecularly clean” disease in which all patients had the BCR::ABL1 rearrangement with no other known molecular abnormalities. This uniformity and uniqueness were assumed to largely explain the positive outcome with TKIs. In contrast, other hematologic malignancies (or blast-phase CML), which are molecularly more complex and in which targeted therapies, even when effective, produce only transient responses at best. It is now clear that CML is, even at diagnosis, molecularly more complex than previously recognized, with mutations in other cancer-related genes or Philadelphia chromosome–associated rearrangements in nearly a third of all patients at the time of diagnosis. These abnormalities, particularly ASXL1 mutation, had been recognized as predictors of poor outcome for patients treated with imatinib.² The expectation was that second-generation TKIs, and even more asciminib, could overcome this poor prognosis considering the improvement they provide in many measures of efficacy compared to imatinib. However, in their study, Shanmuganathan et al show that, unfortunately, the poor outcome of patients with ASXL1 mutations is essentially unchanged regardless of the TKI used. These patients with ASXL1 had a lower probability of achieving major molecular response, an inferior failure-free survival, and a higher probability of developing BCR::ABL1 kinase domain mutations.

These results raise important questions for which we may not yet have answers. First, whether all patients with CML should undergo next-generation sequencing at the time of diagnosis to identify patients with mutations in other genes. ASXL1 mutations are identified in approximately 10% of patients at diagnosis.^1-3 Thus, the yield of general screening would be expected to be low, but screening would identify a higher-risk population. This finding may have prognostic value, but there are currently no therapeutic implications, because no specific therapy can be recommended for these patients considering that, as shown by Shanmuganathan et al, none of the approved TKIs appear to alter the outcome. When adding the cost and uneven access to this technique, routine testing provides uncertain value. This will probably change in the future, but at the moment the European LeukemiaNet recommendations do not advise routine testing at the time of diagnosis.⁴ For patients with resistance to multiple TKIs, testing may add more value, as mutations in other genes are identified more frequently in this setting.⁵ This is likely because of the enrichment of patients with these mutations among those with poor outcomes that have required treatment changes. There is also the interesting observation of the emergence of these gene mutations during therapy in patients without evidence of them at diagnosis, although it appears that later-occurring mutations are more frequently in genes other than ASXL1, particularly in DNMT3A.³ Then comes the question of how to manage these patients. Despite their worse outcome compared with those without ASXL1 mutations, 55% to 59% achieve major molecular response by 12 months,^1-3 and it may increase to 65% at 24 months.³ Evidently, these patients need close monitoring, but this may still represent a favorable outcome for most patients, even when not as good as for those without these mutations. An intriguing observation comes from the FASCINATION study, in which patients with newly diagnosed CML were treated with asciminib in combination with adenosine triphosphate–competitive inhibitors, imatinib, dasatinib, or nilotinib. In this study, it was reported that the response rate for patients with ASXL1 mutations may be equivalent, if not better than for those without any mutations or with mutations in other genes.⁶ This is a preliminary observation in only 1 study and with small numbers of patients that requires confirmation, but it opens an area of research that may provide a path for an effective treatment for this group of patients. An alternative would be to consider performing a stem cell transplant for these patients. Although instinctively reasonable, there are currently no data to show the outcome of these patients with CML with ASXL1 mutations when receiving a stem cell transplant. In other entities, such as acute myeloid leukemia, patients with ASXL1 mutations have also been associated with a worse outcome after stem cell transplant compared to those with no such mutations, irrespective of the cytogenetic risk group.⁷ In the absence of data, it may not be justifiable to consider stem cell transplant as the standard for these patients at the time of diagnosis, but it should be considered in patients who have experienced resistance to 2 or 3 TKIs and who are found to have ASXL1 mutations. Finally, other mutations have not been shown to confer the same inferior prognosis when present at the time of diagnosis compared to ASXL1. This, however, may be in part because all other mutations are clustered together, possibly diluting the impact of uncommon but high-risk mutations. Much larger series are required to accumulate enough patients with various gene mutations to better answer these questions. Projects such as the one by the International CML Foundation Genomics Alliance⁸ aim to address this need by collecting data from a large number of patients from centers in countries all over the world.

The evolution of our understanding of CML and its translation into improved therapies and patient outcomes has been a fascinating story, and one with remarkable achievements. But we are unquestionably not done. We continue learning and identifying new complexities and areas of need that require additional investigation. This emphasizes the necessity of not only continued research in CML, but also the need for multicenter, international collaborations to address “niche” questions, and the need for continued funding for research in CML. In the words of the Uruguayan writer Mario Benedetti, “When we thought we had all the answers, suddenly they changed all the questions.”

Conflict-of-interest disclosure: J.C. has received grant or research support (to the institution) from Novartis, Sun Pharma, Ascentage, Tern Pharma, CytoAgents, Kuro Pharma, and Bristol Myers Squibb and is a paid consultant for Novartis, Pfizer, Sun Pharma, Takeda, Tern Pharma, Incyte, Ascentage, Enliven, Syndax, and Janssen.