Initial excitement about the success of the ABL inhibitor STI-571 (Gleevec) has been tempered by reports of drug resistance. Wolf-K. Hofmann and colleagues have now developed a microarray-based approach to predict which patients will become resistant to the drug — before they've even started to take it.

Gleevec was initially developed to treat chronic myelogenous leukaemia (CML). CML cells have a chromosomal translocation — the Philadelphia chromosome (Ph) — that encodes the BCR–ABL fusion protein, a dysregulated form of the ABL tyrosine kinase. A subset of acute lymphoblastic leukaemias (ALLs) are also Ph+. Might Gleevec be an effective therapy for Ph+ ALL? Unfortunately, people with Ph+ ALL are frequently resistant to Gleevec from the outset (primary resistance) or develop resistance shortly after beginning therapy (secondary resistance). Hofmann and colleagues used oligonucleotide microarrays to find out whether differences in gene expression can distinguish Gleevec-sensitive from Gleevec-resistant ALL, using bone-marrow samples taken from patients before and during treatment with Gleevec.

Their analysis reveals 95 genes that are expressed differentially, before treatment, in Gleevec-sensitive patients compared with patients who have primary resistance. A further 56 genes changed their expression levels during treatment of those patients who developed secondary resistance. Secondary resistance was associated with overexpression of Bruton's tyrosine kinase (BTK) and two mitochondrial ATP synthetases (ATP5A1 and ATP5C1), as well as downregulation of the pro-apoptotic gene BAK1 and the cyclin-dependent kinase inhibitor INK4B. This provides some testable hypotheses as to how Gleevec-resistant cells overcome their need for BCR–ABL, and some potential targets for overcoming Gleevec resistance in those patients who develop it. In the future, it might be possible to test ALL patients for resistance before they begin therapy, and give them appropriate drugs if resistance does develop.