Edgardo S. Santos Castillero, MD, FACP: How has the development of tyrosine kinase inhibitors [TKIs] been from the third-generation drugs, and how have we ended up today with a third-generation osimertinib? You can take us to that pathway in the last 10, 12 years.
John V. Heymach, MD, PhD: It’s hard to believe. It’s even longer than that that we’ve been testing these drugs. I treated my first patient with gefitinib in 1999, believe it or not, when we had those early trials going on, and that preceded the discovery of EGFR mutations. Of course, the first drugs that were in the clinic for EGFR were gefitinib and then erlotinib, so we think of those as the first-generation drugs. They’re certainly active.
The hallmarks of the first-generation drugs are the following. First, they have toxicities related to inhibiting wild-type EGFR. Those toxicities are familiar to any of us who treat patients with EGFR inhibitors. This is the acneiform rash and commonly some diarrhea. They often get fingernail issues like paronychia, and then some other rare things, such as adverse effects in the hair and in the cornea mucous membranes and other rarer adverse effects as well. This all has to do with the fact that wild-type EGFR is present in hair follicles; it’s present in a lot of other tissues. Here’s an interesting tribute point: EGFR was initially discovered by Stanley Cohen. In fact, he won a Nobel Prize for this because it was found to be a factor that was critical for teeth erupting in mice. It was key in the differentiation of certain epithelial surfaces. In mice, when you knock this gene out, they get a funny phenotype with their hair; it’s called a wavy phenotype. That wavy phenotype has been known for a long time. Mice don’t get the acne; they get other skin and hair effects.
Those were the first-generation drugs, and resistance emerges. Typically, these drugs will give a PFS [progression-free survival] on the order of 9 to 12 months, and the most common alteration-associated resistance for these is the T790M. They call this the gatekeeper mutation, and this prevents the binding of the first-generation drugs. The first-generation drugs are irreversible inhibitors, so if they get a gatekeeper mutation, they can’t bind avidly enough into that pocket, and that gatekeeper mutation prevents them from inhibiting the receptor.
The second-generation drugs are dacomitinib and afatinib. They’re a bit more potent, and they can bind irreversibly, but T790M still often emerges, and they still have a lot of wild-type activity. Both afatinib and dacomitinib have a lot of rash, and they have a lot of diarrhea associated with them. It’s manageable in the setting of the clinical studies. If you look at outcomes for the second-generation agents vs. the first-generation agents, they’re a bit better. Afatinib has been tested against gefitinib head-to-head; it was a bit better. Dacomitinib was tested against the first-generation drugs, which was the ARCHER1050 study, and it prolonged overall survival in that study. The median PFS was about 15 months, so that was a real difference, but it came at the cost of a fair amount of skin toxicity. It was manageable but not trivial.
This brings us to the third-generation drugs. The big differences with the third-generation drugs are 2-fold. The first is that they bind irreversibly, so they have a covalent linkage. The details here are that they form a covalent linkage with the cysteine near the binding pocket: It’s called the C797. The reason that’s important is because, if that cysteine gets mutated, then third-generation drugs don’t work anymore, so 1 mechanism of resistance to third-generation drugs is the C797S mutation.
The advantage of third-generation drugs, of which osimertinib is the only 1 that we have approved right now, is that they have a much better specificity. They inhibit the mutants much better than they inhibit the wild-type receptor, so they have much less skin toxicity and less diarrhea, and they’re better tolerated overall. Osimertinib is a well-tolerated drug. It’s more active, and it will inhibit the T790M mutation, so you don’t get the same mechanism of resistance.
Interestingly, we [at the University of Texas MD Anderson Cancer Center] and other groups have recently reported about how you get resistance to third-generation drugs. You often get bypass mechanisms. You get a different kinase that’s active like MET amplification or RET fusions, and there are a number of others that can happen. HER2 [human epidermal growth factor receptor 2] can get activated, and so forth, in which case, you may have to switch therapies. There are other mutations that can emerge. The important study that was done was called the FLAURA study, and that compared osimertinib with the first- and second-generation drugs: erlotinib and gefitinib. It was substantially better for PFS, with about 18.9 months vs. about 10 months, so there was a big difference. As we said, most of that benefit was seen in the exon 19 deletions, not in the L858R, and it prolonged overall survival as well. Third-generation drugs are clearly an advance with better tolerability.
Another factor I didn’t mention is that there was better CNS [central nervous system] penetration for osimertinib. There was much more activity in the CNS. You don’t get T790M as a mechanism of resistance, but you do get other mechanisms of resistance. One of the challenges that emerges is if you get resistance to osimertinib. What other targeted agents do you go to? We don’t have any other approved agents if you get osimertinib resistance. In some people’s minds, they consider, “Maybe I should start with a first-generation drug or a first-generation drug in combination,” like the ramucirumab studies we talked about earlier. If you get resistance, you can use osimertinib and keep that in your pocket for later. There are pros and cons for those 2 approaches.
That’s how the field has evolved from first-generation to second-generation and now to third-generation drugs. It’s been a clear advance with clear progress as we’ve gone forward.
Transcript edited for clarity.