Testing for KRAS Mutations in NSCLC
Paul Bunn, MD: In general, in the beginning, if the driving alteration was KRAS, it’s unlikely that the tumor cells would have EGFRor ALK. On the other hand, sometimes in the presence of an EGFR-activating mutation or an EGFR drug, some of the cells will become resistant by developing KRAS mutations. Similarly, for ALK fusions, some cells, minimally, may become resistant by developing KRAS mutation. If you have a patient with an EGFR mutation who becomes resistant to the KRAS mutation, it will make sense in the future when we consider treating the patient with the KRAS G12C–specific drug.
For all these molecular alterations, 1 mechanism of resistance is they develop mutations that will prevent binding of the drug. In another way, it becomes resistant to development activation of another path, which may be KRAS. In some instances, there are other activating pathways such as the EGFR, ALK, ROS1, and so on.
Alexander Drilon, MD: In general we test for KRAS mutations in patients with non–small cell lung cancer, and we’ve done this broadly for all cases, recognizing that we identified this gene a long time ago and well before the discovery of EGFR, MET, and the other oncogenes such as ALK, ROS1, and RETfusions. We’ve done this testing in the earlier phases also, in part because we know that these hot spot mutations tend to be mutually exclusive with these other drivers of oncogenesis—meaning, when you find a KRAS mutation, there is a low likelihood that you might find another oncogenic driver. The 1 exception that we’ve learned about more recently is that KRAS mutations can sometimes co-occur with other drivers like MET exon 14 alterations, and the RAS mutation can be found either in a de novo or an acquired fashion. In general, for the other drivers, these tend not to coexist with those other driver states.
Paul Bunn, MD: In the old days, 1 strategy was to test KRAS first because it’s the most common, and that worked for others only if KRAS was negative. In my opinion, that’s not a particularly good strategy. Nowadays, basically every NGS [next-generation sequencing] panel will have KRAS mutations as part of the panel because it’s the most common alteration since every panel has KRAS mutations as part of the family nowadays.
If we do these tests sequentially, it will take a very long to find out how you’re going to treat a patient. Again, the old strategy was for some people to do KRAS first. If it was positive, that’s the end. If it was negative, look for other alterations.
Right now, the common alterations are in the panels. Some panels are larger than others. The foundation, for example, has 400 genes that they’re looking for mutations. In our particular University of Colorado panel, we have only 40 genes, because those are the genes that are most common for which there is a potential treatment. In the foundation panel, there are many genes that may be mutated, but there is nothing you can do clinically. Even now, for many of the KRAS mutations you find, there’s no sense to look for another if you have it, because there’s no specific treatment. Right now, the only specific treatments are basically based on G12C, for which there are some covalent drugs, binding drugs, that will prevent KRAS activation.
Jonathan Riess, MD, MS: My practice, particularly for stage IV metastatic lung adenocarcinoma, is to do broad genomic profiling up front, consistent with National Comprehensive Cancer Network Guidelines. The reason is that things have evolved from what we used to call the big 3: EGFR, ALK, and ROS1, where you can match to EGFR, ALK, and ROS1 TKIs [tyrosine kinase inhibitors], respectively. We have so many other agents now: MET exon 14 capmatinib is approved, BRAFV600E, dabrafenib-trametinib, RET fusion, selpercatinib. With all these additional mutations, I approach it from the basis of getting broad genomic profiling to try to find targets that give us the highest percentage chances possible to find targets to match to effective drugs. Beyond that, at University of California, Davis Comprehensive Cancer Center, we have clinical trials for HER2 [human epidermal growth factor receptor 2] and other molecular subtypes. It also broadens the scope to bring clinical trials to our patients.
I approach broad genomic profiling, particularly for lung adenocarcinoma. For squamous non–small cell lung cancer, particularly in patients who have a scant tissue specimen or have a never- or light-smoking history, I also pursue up-front genomic testing. At some point in squamous histology, I do get genomic testing if a patient has a heavy-smoking history. Often at University of California, Davis Comprehensive Cancer Center we get it as part of our Lung-MAP master protocol, which is a National Cancer Institute–sponsored clinical trial that looks at next-generation sequencing and includes squamous cell carcinoma to try to match the targeted therapy. The study helps support those getting that sequencing result by foundation medicine.
Transcript Edited for Clarity
