The Pill That Could · Resistance

Resistance, part 4

The CML toolkit, in 2026, has five oral kinase inhibitors approved for the disease: imatinib, nilotinib, dasatinib, bosutinib (a third second-generation drug, slightly different in its profile, that we have not paused on), ponatinib, and asciminib. (Five drugs that are oral and approved as monotherapy; bosutinib makes six if you count it separately, which the field generally does.) For most patients with most patterns of resistance, at least one of these drugs will produce a durable response. The question, in CML practice, has shifted from ‘is there a drug for this patient’ to ‘which drug, in which sequence, at what dose, weighed against what side-effect profile.’

This is a different question from the one Druker faced in his clinic in 1993. It is a better question.

* * *

What did the resistance story actually mean, looking back?

The first thing it meant is that the early concerns the skeptics inside Novartis had voiced — the worry that resistance might rapidly outpace any single inhibitor — had not been crazy. They had been right to be worried. CML cells did evolve resistance to imatinib. The evolution happened reproducibly, through specific mechanisms, in a substantial fraction of patients with advanced disease. It was the predictable thing to happen. It happened on the predictable timeline. It produced, in many patients, the predictable end — a relapse on a drug that no longer worked.

The second thing it meant, however, is that the resistance story was, in CML, manageable in a way it has not always been in other cancers. The mutations that arose in BCR-ABL were not arbitrary; they clustered in particular residues, they could be enumerated, and the field could design drugs that worked around them. The cancer's evolutionary repertoire was finite. Each successive generation of inhibitor blocked some additional set of resistance mutations, and patients could be cycled through the drugs as their disease evolved, and most patients could be kept in remission through the cycle. Resistance, in CML, became a chronic management problem rather than a terminal one. The disease did keep evolving, and the field did keep responding, and the pace of the back-and-forth turned out to be one the field could mostly maintain.

The third thing, harder to summarize, is that the resistance story shaped how the field thought about all subsequent targeted therapies. After CML, every targeted-therapy program assumed resistance would emerge — it was no longer a surprise, it was a planning parameter. Companies developing first-in-class inhibitors began thinking, from the outset, about second-generation versions. Trials were designed to monitor for emerging resistance mutations. The model of treatment shifted from ‘give the drug, hope it works forever’ to ‘give the drug, monitor for resistance, switch when needed.’ This was a more realistic model of what targeted therapy could do. It was also, in some other cancers, a more difficult one to execute, because the resistance landscape in those cancers turned out to be wider and messier than CML's.

We will come back to that in chapter eleven.

* * *

There is one more thing about the resistance story worth attending to before this chapter ends, because it is the thing that connects what happens in the cells of any one patient to the larger story of the disease.

When a CML patient is in chronic phase, on imatinib, doing well, with the Philadelphia chromosome essentially undetectable in the blood and bone marrow — a patient in deep cytogenetic and molecular response — it is easy to think of the cancer as gone. The disease is, by every measure available, suppressed. The patient feels well. The blood looks normal. The bone marrow, biopsied, shows a more or less normal-looking population of cells, with no obvious leukemic infiltrate.

But the disease is not gone. It is, in most patients, still there, just below the level of detection, in some small reservoir of cells that the drug has not killed. The drug suppresses the cells; it does not eradicate them. As long as the patient keeps taking the drug, the cells stay suppressed. If the drug is stopped, in most patients, the cells re-emerge and the disease comes back. Resistance mutations, when they appear, often appear in this small reservoir — mutations that were already there at low frequency before treatment began, or that emerged during treatment in cells that the drug was selecting against but not quite eliminating. The reservoir is the source from which both relapse and resistance can grow.

What this means, practically, is that imatinib in CML — even imatinib at its best, in patients in deep remission — is a drug that holds the disease in check without finishing it off. The disease is controlled but, in most patients, not cured. The patient has to keep taking the drug. If the patient stops, the disease comes back. If the disease comes back, it is sometimes still controllable with the same drug, and sometimes it is not, because the cells that emerge in relapse have, sometimes, acquired the resistance mutations that the drug couldn't block.

This is a different concept of what a successful cancer therapy looks like than what most of us, growing up, learned to expect. It is not the cure that the surgical-oncology model trains us to think about — take it out, and if the borders are clear, it's gone. It is something more like the management of a chronic infectious disease, or the management of a chronic metabolic disease: a daily intervention that keeps a process from happening, with the understanding that the process is still trying to happen and would resume immediately if the intervention stopped.

What this means, for the long-term outlook of a CML patient, is the subject of the next chapter.

The chapter is about what cure means in a disease that is well-managed but not eradicated, about what the long-term data show for patients who have been on imatinib for fifteen, twenty, or twenty-five years, and about whether there is anyone in CML — with the disease really, truly gone — or only patients in whom the suppression has been so deep, for so long, that the distinction has become unimportant.

The honest answer, as we will see, is somewhere in between. And the resistance story this chapter has told is part of why the answer is honest rather than triumphant.

All 13 chapters of The Pill That Could · Library · All courses