QOL, Efficacy, and Cost Considerations Help Drive TKI Treatment Selection in CML

From first-generation TKIs such as imatinib (Gleevec), second-generation treatments such as bosutinib (Bosulif), nilotinib (Tasigna), and dasatinib (Sprycel), and even third- and fourth-generation iterations, the arsenal of TKIs at the disposal of clinicians, and deciding which TKI is best for patients with chronic myeloid leukemia (CML) comes down to a few key factors, according to Hagop Kantarjian, MD.

In an interview with OncLive^®, Kantarjian discussed his thought process behind TKI treatment selection for patients CML in clinical practice. He outlined the specific factors that help him choose between TKIs across different generations like imatinib, dasatinib and asciminib (Scemblix). Additionally, Kantarjian touched on the November 2025 FDA approval of generic dasatinib tablets for CML and acute lymphoblastic leukemia, and how he feels it will impact clinical practice, and how novel formulations could affect TKI selection.^1,2

Kantarjian is a professor in the Department of Leukemia and Samsung Distinguished Leukemia Chair in Cancer Medicine both at The University of Texas MD Anderson Cancer Center in Houston; as well as a non-resident fellow in health policy at the Rice University Baker Institute.

OncLive: How do you approach frontline TKI selection for CML? What factors do you consider?

Kantarjian: We have several BCR-ABL TKIs which are available, we call them first generation, like imatinib because it was first discovered. Then we have the second-generation TKIs, like bosutinib and nilotinib. They are called second-generation because we developed them in the second wave, and they are more potent. Then, there's the third generation TKIs, which were also developed later. [Third generation TKIs] are much more potent and they have the property of suppressing the BCR:ABL1 T315i clone. The third generation TKIs are divided into 2 categories, one that still targets the ATP kinase domain like all the others. An example of this is ponatinib [Iclusig], which is probably the most potent BCR-ABL TKI, but it [has associated toxicities] at the FDA-approved dose of 45 mg daily.

Then, there are the drugs that selectively target the ABL myristoyl pocket, or STAMP inhibitors. An example of a STAMP inhibitor is asciminib, which is quite good, and the toxicity profile is being developed still. There are also 4 investigational drugs, 2 of them are like ponatinib: ELVN-001 and olverembatinib. Then there are 2 STAMP inhibitors, which are TERN-705 and TGRX-678.

We have plenty of drugs, and you have to think [about] the aims of frontline therapy. There are four aims. The first one is making patients with CML live a normal life and to normalize survival. People refer this as this as operational cure, I like to call it functional cure, meaning that if they take the pill every day through their life, and they will live a normal life.

The second aim is what is referred to as treatment-free remission. If [patients with CML] achieve the deep molecular response of 4-log reduction of the disease or MR4, and it lasts for 2 years and we stop the treatment, then the success rate of not needing treatment or a treatment-free remission will be 50%. If you take the TKI for a durable, deep molecular response of 5 years or more and you stop, the treatment-free remission rate is about 80%. This is what I call a molecular cure.

There are two additional aims. One is to reduce the early and long-term adverse effects [AEs] of TKIs, this is where we have to be cautious with the new TKIs. We had the example of nilotinib, where at 5 years we thought it was a very safe drug then at 10 years we found out that it caused serious cardiovascular problems in about 25% of the patients. You have to be cautious about the early and long-term toxicities [of TKIs].

The final aim of therapy is to make the treatment be cost-effective. If you're using a new treatment, how much value does it add in terms of either survival, treatment-free remission, or AEs? For example, if imatinib is S350 per year, you're paying 1$ per day for the treatment. The generic dissemination through Cost Plus Drugs at 50 mg per day is $1500 per year, thus you're paying $5 per day. If you look at the patented drugs, like asciminib, you’re paying close to $360,000 per year. If you have a choice of paying $1, $5, or $1000 per day, that new drug has to justify itself in terms of the treatment value, which I don't see yet.

To summarize, how do I choose frontline therapy? If there is a patient where survival or living a normal life is the end point, then [I would] probably [choose] imatinib. [Imatinib] will provide a near-normal life at a price of $1 a day. If I have a young patient or a patient with high-risk disease where I want to suppress the disease better, or I want to achieve a deep molecular response faster, then the second-generation TKIs could be used. This is where dasatinib at 50 mg per day would probably be the safest and best treatment value. People say, well, the patients are not paying; actually, the patients do pay out of pocket. Patients who do not have insurance pay and the government pays for this. These are important considerations now.

Finally, if you want to reduce toxicities, we do not know if the new drugs, particularly asciminib, have a lesser incidence of long-term toxicities compared with imatinib or the second-generation TKIs.

Which safety considerations are most important to remember with currently available TKIs for CML?

We look at the pretreatment characteristics of patients. For example, in a patient with chronic lung disease, we avoid using dasatinib because there's a higher risk of pleural effusions. In a patient with diabetes or with cardiovascular problems, we avoid nilotinib because it can cause a worsening of the diabetes and cardiovascular problems. In a patient with kidney problems or liver problems or GI problems, we avoid bosutinib since it can worsen the kidney problem and can cause enterocolitis or liver dysfunction. These are the choices for the frontline therapy and the patients, and in general, if we manage the patients well, the incidence of resistance is only 1% per year.

The incidence of toxicities and prohibitive toxicity if we adjust the dosages [of TKIs] is approximately 1.5% per year. [Additionally,] the treatment discontinuation should not be more than 2.5% per year. However, when you look at the randomized studies, they report treatment discontinuations of 5% to even 15% or 20% per year, which I don't understand. [This may be due to] the patients and the physicians now having a lower threshold for changing TKIs, because there are so many of them available.

What is important to know about the FDA approval of generic, small molecule dasatinib for CML?

First, people assume based on historical data that when you have a generic drug, the price drops very quickly, and that's not the case anymore with the current market forces. In the past, we said that whenever we have 4 to 5 generics, the price of the generic will fall from over $100,000 to about less than $5000. [This drop in price] has not been the case. What we have seen surprisingly, for example with imatinib, is that when it became generic in 2017, the price of the generics did not drop down below $50,000 per year, which it should have now. [The price of generic imatinib] dropped down only when the companies that bypass the conglomerate structure, such as Cost Plus Drugs, came up with the generic. This is what forced all the generics to go down, Cost Plus Drugs and companies, which are not part of the vertical conglomerates in health care, are the ones, which are really reducing the price of generics. The same thing happened with the dasatinib, there were several generics, and [the price] went down to $1500 per year only when Cost Plus Drugs produced a generic that bypassed the health care conglomerates. If you look at Nilotinib, it's still at a high price, it's still at about $7000 to $8000 per year, and that's because it was not adopted by a generic company like Cost Plus Drugs. When we talk about the new generics, you have to know the source.

What is your take-home message about TKIs and treatment sequencing in CML?

My take-home message is that patients with CML today will live their normal life, provided they are treated and managed optimally, can access the TKIs in terms of cost and geography, and that those TKIs are safe. Almost all the patients will live their normal life, but they have to be monitored closely, and they must have their treatment modified at the sign of resistance, which I view as a BCR-ABL transcripts going over 1% after 1 year or later of therapy.

The other take home message is when you encounter toxicities, do not change the TKIs. If [the TKI] is a manageable toxicity, reduce the dose of the same TKI, if it is a prohibitive toxicity that can potentially cause organ damage, then you can change the TKI and use a lower dose of the newer TKI rather than the full dose.

My take-home messages are very hopeful ones. CML is functionally curable in most patients, and it is molecularly curable in a fraction of the patients. I would like to see the molecular cure increase with the new research, but there should be no patients with CML who would die from the disease, other than those with truly resistant disease or cross-intolerance to multiple TKIs. [For patients with cross-intolerance to multiple TKIs] there is always the option of a one-time curative allogeneic stem cell transplant, which is a bit risky, but it is potentially of good value. Among the patients who develop true resistance to all the TKIs available, we have to consider the transplant as an option in these patients.

References

1. Alembic Pharmaceuticals Limited announces USFDA Final Approval for Dasatinib Tablets, 20 mg, 50 mg, 70 mg, 80 mg, 100 mg, and 140 mg. News release. Alembig Pharmaceuticals. November 7, 2025. Accessed June 23, 2026. https://alembicpharmaceuticals.com/webfiles/media/2025-2026/Press-Release-Dasatinib-Tablets.pdf
2. Sprycel. Prescribing information. Bristol Myers Squibb. Updated July 2024. Accessed June 23, 2026. https://packageinserts.bms.com/pi/pi_sprycel.pdf