The Sokal Index is interesting in that, although it was developed prior to the tyrosine kinase inhibitor (TKI) era, it has been shown to be predictive of response in patients with chronic-phase CML who are on TKI therapy. In addition to the Sokal Index, we have the European Treatment and Outcome Study scoring system, but there have not been many updates since its development in 2011. I do hope that there will be molecularly defined approaches to risk stratification in the future that incorporate the presence and nature of additional genomic mutations aside from BCR-ABL1. And, ideally, incorporating this information will be even more predictive than some of our current clinical predictors. 

There is also a concerning issue regarding the presence of other chromosomal abnormalities in CML. I recently had a patient with a second Philadelphia chromosome in 1 cell out of 20, and nothing else to suggest anything other than chronic-phase CML. Within 1 year, the patient had disease progression and several additional chromosomal changes. This case illustrates the challenges associated with identifying those outliers who will experience substantial disease progression in the future. Part of the issue is that it is likely that the patient population responsible for disease progression or persistence is a very small proportion of the total population at diagnosis. Trying to define that population and characterize it are challenging; however, I do believe that genetic approaches hold promise.

Sometimes it can be interesting to calculate a patient’s European Treatment and Outcome Study score, although, generally, you can simply look at someone and know whether they are at the highest risk. Typically, the patients who tend to present to my acute leukemia service are the high-risk patients, and I have not found carefully stratifying them into the risk systems to be very helpful. Importantly, some of these patients may have blast crisis, which warrants further attention with risk stratification.

However, I think that we need a more genetic approach to stratification because that is where the answer lies. Some individuals with CML will inadvertently let their disease progress because they are able to tolerate the symptoms; these patients are young and fit, and they come in with high white blood cell counts. For such individuals, we might not immediately realize that they are high risk with our current systems, but I do not think that it is helpful to carefully stratify them along modern risk scores. Ultimately, I believe that more genetic characterization is an area in which the field could certainly improve.

There are several risk stratification systems that have proven to be useful in acute myeloid leukemia, myelofibrosis, and many other diseases. In contrast, the systems that we have for CML were either developed prior to second-generation TKIs or they do not take into account additional genetic markers. There is ample opportunity for the further development of more relevant systems for CML, by comparison. 

Some diseases could benefit from the additional incorporation of molecular insights, and CML is one of them. As part of our initial diagnosis of patients who come in with a high white blood cell count, I will check the polymerase chain reaction for BCR-ABL only because I have high suspicion that they have CML. But I always check the next-generation sequencing panel, and the presence of certain mutations seems to be associated with worse outcomes, but we do not currently have a way of incorporating these additional genetic mutations into our risk stratification systems. 

It is disconcerting that we do not have a robust method of identifying patients with CML who are the most at risk of not doing well. We also lack a good way of predicting the length of therapy that is required for obtaining deep remission and for the successful discontinuation of treatment. The currently available predictors are very rudimentary, so I believe that we would all like to see development in this area.