Drs. Bob Doebele, Ross Camidge, and their colleagues at the University of Colorado just published an interesting and clinically relevant paper in Clinical Cancer Research that looked in detail at the mechanisms of resistance in ALK rearrangement positive patients to the ALK inhibitor XALKORI (crizotinib).   Evaluating 14 patients with a known ALK rearrangement who were resistant to XALKORI, either from first treatment (in the case of two patients), or after a period of response (for the other 12), they performed repeat tissue biopsies after resistance was documented, in order to determine whether the ALK rearrangement was still present, whether there were additional copies of the ALK gene, or whether there were new mutations documented that might now be an alternative mechanism of driving the cancer.

Of the 11 patients who had sufficient tumor tissue for analysis, they found a wide range of apparent mechanisms of progression.  Four of the 11 (36%) developed new mutations in the ALK gene that conferred resistance, and in two cases this was associated with a particular mutation(known as G1269A) that has been associated with resistance to crizotinib in lab-based work.  Two patients (18%), including one who had an ALK resistance mutation, were now found to have developed additional copies of the ALK gene in their cancer cells, which could overcome the inhibitory effect of crizotinib. Three patients (27%) no longer had the ALK rearrangement present in their tumor tissue, of whom one had developed a new EGFR mutation, one had developed a KRAS mutation, and one had developed an unknown mutation (they tested for a limited panel of known high yield ones). Another patient (9%, in case you haven't figured out the pattern) developed a KRAS mutation but continued to have an ALK rearrangement detectable.  The final two continued to have an ALK rearrangement present and no detectable mutation or other changes that could explain the development of resistance.  The summary of mechanisms of recurrence is as shown on the left in the figure below:

(click on image to enlarge)

Part B, on the right side of the figure, describes two potential mechanisms for acquired resistance.  The first is the emergence of a second cancer driver that is present and coexistent in the cancer cell with the mutation in question (left side of figure B), like the development of c-MET over-expression in EGFR mutation positive cancer cells. This occurs under evolutionary pressure of a treatment with an EGFR tyrosine kinase inhibitor (TKI) to produce a population of cancer cells that still have the driver mutation (EGFR, in this example) but have now developed a competing and overriding feature that confers resistance, whether it's a co-existent T790M mutation or increased c-MET expression, both leading to resistance to EGFR tyrosine kinase inhibitors).

In this paper, even though the numbers are small, the authors demonstrate that it's also possible for the evolutionary selection pressure of effective treatment to lead to the prevalence of a new and separate oncogenic (cancer-facilitating) driver mutation (right side of figure B) above).  Here, the idea is that there are different subsets of cancer cells.  For example, prior to treatment with a targeted therapy the cancer may be comprised of predominantly ALK rearrangement positive cancer cells with a small minority having a KRAS mutation, but that after ALK inhibitor therapy, those ALK positive cells die and leave a cancer growing that is comprised of a predominantly KRAS mutation positive cancer.

There are many potential implications of this work, although none is proven yet.  One is that there is a real value in repeat biopsies, at least in patients who have had great responses to targeted agents and/or identified mutations, to help shape subsequent treatment decisions.  Though rare, it's possible for a person to have a new EGFR mutation that could lead to effective treatment with an EGFR TKI.  The finding that some of these cancer cells no longer have an ALK rearrangement suggests that these patients aren't likely to benefit from a new ALK inhibitor or treatment with a heat shock protein inhibitor (which has appeared very promising for ALK positive patients).

We've seen examples of situations in which the cancer, or at least some areas of the cancer, have morphed into a new version, sometimes a different subtype of lung cancer entirely; in some cases, that may occur from induced fundamental changes in the cancer cells (model 1, above), and in other cases, that might occur because different cells become predominant (model 2, above).  Even looking at the small number of cases here, we can see a complex array of possibilities.  Our knowledge of what's possible is growing on a case by case basis, but it is helping direct us to literally individualized recommendations for patterns that are becoming cancer and patient-specific.

NEXT: New insights on T790M and EGFR TKI resistance.