Resistance to imatinib mesylate in CML: all BCR-ABL mutations “are created equal but some are more equal than others”

On May 2001, imatinib mesylate (Glivec, Gleevec) made the cover of Time as the magic bullet to cure cancer. In spite of the rather sensationalist and oversimplified style of the announcement, the drug did in reality represent a landmark for targeted therapy in neoplasia. The idea that a chemical compound could be designed and refined to fit into and block a specific domain of an oncoprotein was in itself a vindication of the decades spent on research into the molecular biology of these proteins. The fact that the idea worked in practice, with the designer compound turning out to be an overwhelmingly successful drug, looked indeed like magic. For the treatment of chronic myeloid leukemia (CML), a new era had indeed begun with this selective inhibitor of the Bcr-Abl tyrosine kinase. Unfortunately, however, imatinib is somehow a victim of its own success: its target specificity and its “snug fit” into the Abl kinase pocket provide the ideal scenario for the leukemic population to evade its action, if some leukemia cells can produce Bcr-Abl molecules with mutant amino acids that directly or indirectly affect imatinib binding. As with conventional antibiotic resistance in bacteria, the mutant clone will be able to outgrow the wild-type sensitive cells, and so the leukemia will no longer respond to imatinib.

True to such a somber forecast, shortly after the encouraging results of the first clinical trials were announced, there came the first of reports of resistance to imatinib occurring as a consequence of point mutations in the Bcr-Abl kinase domain. To date, at least 19 different mutations have been identified in cells from patients who became refractory to imatinib treatment (Goldman and Melo, N Engl J Med. 2003, in press). Since the original, preimatinib leukemia clone in all these cases comprised wild-type BCR-ABL–positive cells, the point mutations in the newly emerged resistant clone seem to be the obvious “cause” for the imatinib-resistant phenotype. But are they always? In this issue, Corbin and colleagues (page 4611) sound a warning note. They show that, whereas several mutants detected in relapsed patients have in fact a significantly lower in vitro sensitivity to imatinib's antiproliferative and antiphosphorylation effects, others remain fully inhibited by the drug. In these cases, the resistant phenotype may be due to yet another abnormality, maybe another coexisting but unidentified point mutation that could even lie outside the kinase region (Azam et al, Cell. 2003;112: 831-843), or maybe an abnormality downstream of Bcr-Abl leading to a Bcr-Abl–independent mechanism of resistance.

The question of how the mutant cell clone is selected to become the predominant one at the time of relapse when the mutation itself does not confer a growth advantage in the presence of the drug remains to be explained. Equally intriguing is the opposite question: why are some of the mutations in amino acids predicted to be important and shown to be responsible for in vitro resistance (Azam et al) not actually seen in patients? And similarly, why are there some amino acid residues known to be in contact with imatinib that are never apparently mutated in either in vitro or in vivo resistance? Could it be that all these mutations would also result in the loss of kinase activity and transformation effect? These and other related issues tell us that Bcr-Abl still has a trick or two up its sleeve. Or in its pocket!