As I've described in a prior post, one of the most consistent findings in the work with the EGFR inhibitors Iressa (gefitinib) and Tarceva (erlotinib) is that never-smokers are far more likely to demonstrate a response and survival benefit than patients who do smoke or did smoke. Here, for instance, is the set of survival curves separated by smoking status for the large randomized trial of tarceva vs. placebo in previously treated patients with advanced NSCLC (abstract here):

(click to enlarge)

We really haven't explored why that as, and that's partly because there is a lot we don't know. We do know that never-smokers are far more likely to have mutations than current or ex-smokers with lung cancer, but it's really not all or nothing. The likelihood of having an EGFR mutation is higher in ex-smokers than in current smokers, and it's also a gradually higher likelihood with a longer period since patients quit smoking, and with less overall tobacco exposure in terms of pack-years, the product of the number of packs smoked per day (PPD) x the number of years smoked, so 2 PPD x 20 years would be a 40 pack-year history. You can see that between never-smokers and major smokers are a population of patients who fall in between in terms of their smoke exposure and likelihood of having an EGFR mutation (abstract here):

But it's probably not all about EGFR mutations, which are still only seen in a minority of patients with lung cancer, somewhere in the 5-15% range in the US (more in Asia). It's quite possible that one of the more important issues may be the metabolism and different pharmacokinetics (the way a drug is processed by the body, specifically the changing levels in the blood over time) between smokers and non-smokers, which may mean that what is an optimally effective dose in a smoker may be higher than the optimal dose in a non-smoker.

The idea behind this is that EGFR inhibitors like tarceva are metabolized by a group of enzymes in the liver, called the P450 system, that are induced by cigarette smoke. So some of the key questions we need to ask are whether drug levels of EGFR inhibitors are different between smokers and non-smokers, and if so whether drug levels actually matter for optimal activity of these agents. This boils down to a question of whether for smokers it would be better to use higher doses of EGFR tyrosine kinase inhibitors (and I'll limit to discussing tarceva, since this is the only one that has shown an overall survival benefit thus far and is commercially available in the US). The effects on these enzymes should really be affected only by current or recent smoking, such as in the last 6-12 months at most, so metabolism of people who quit one or more years before treatment should process tarceva like a non-smoker. We would need to focus on other explanations than metabolic differences for more remote smokers.

Hamilton and colleagues from OSI Pharmaceuticals, the manufacturers of tarceva, published a study on the blood levels of tarceva in smokers and non-smokers (abstract here) that demonstrated that never-smokers have trough (low point) blood levels of tarceva that are about twice as high as smokers who receive tarceva at 150 mg. They also showed that giving smokers 300 mg of tarceva produced a higher peak level than giving never-smokers 150 mg, but the area under the curve, which is the measure of the total amount of exposure of the drug in the bloodstream (my calculus teacher assured me there was someday going to be a use for that measurement!) is the same when giving 300 mg of tarceva to an active smoker vs. giving 150 mg to a never-smoker.

(150 mg for both groups)

(150 mg in never-smokers vs. 300 mg in smokers)

In addition, an analysis of the side effects of patients on the BR.21 trial I mentioned above actually demonstrated that current smokers seemed to experience less severe side effects than former or never-smokers (former smokers with slightly less notable side effects than never-smokers, but less pronounced than the current smokers):

I've described in another post how the degree of rash, which seems to correlate with the level of drug in the blood, also has been associated with better survival outcomes with EGFR inhibitors, particularly tarceva.

So, we can see that there are differences in blood levels of tarceva seen in active smokers vs. non-smokers, and that there are differences in side effects associated with different dose levels. But what we don't know yet is whether increasing the dose of tarceva in smokers leads to better results. OSI Pharmaceuticals has been running a small trial testing the idea of dose escalation, and we expect to hear about it at the ASCO meeting in early June, 2007. The Southwest Oncology Group (SWOG) is developing a larger trial of this idea as well. In the meantime, we don't have any good evidence yet that increasing the dose to the point of achieving a rash improves outcome, and I really haven't seen or heard of cases of patients who did not respond to tarceva at 150 mg but did better after going up to a higher dose. We still don't have significant experience using higher doses of tarceva than 150 mg, so I am not recommending using higher doses of tarceva than standard unless or until we see more information. But we are trying to get some explanations of how to use this class of drugs best, not only in narrower populations like never-smokers and those known to have a tumor with an EGFR mutation, but also in the much larger proportion of patients who might improve from no or a very minimal benefit to a greater one if we study issues like optimal dose more carefully. And we should be learning more soon.