As I mentioned in another post, one of the first branch points in the decision tree about what I recommend as treatment for fit patients with previously untreated advanced NSCLC is the question of eligibility for avastin. Although I do routinely recommend avastin for eligible patients, they aren't the majority; instead, I would estimate that population ineligible for avastin due to squamous histology, brain metastases (especially now that we routinely look for them with a head MRI), coughing blood, or other factors account for about 60% of my patients.

As I’ve described previously (prior post here), the available evidence is pretty definitive that the vast majority of platinum-based doublets are remarkably similar in their overall activity. There is a little evidence that cisplatin in a little bit more active (see prior post for details), but most oncologists feel that it’s clearly more challenging in terms of side effects, so carboplatin doublets are far more commonly used, including by me. I’ve historically favored the doublet of carbo/gemcitabine for many patients, because the main side effects are decreases in blood counts with usually very little hair loss, nausea/vomiting, and many other unpleasant side effects. I generally favor regimens that have mostly "paper toxicities" that you tell the patient about, compared with ones that the patient experiences and tells you about. But I’ve used several different carbo-based doublet regimens.

While up until recently there’s been little reason to select any one regimen over another, a very large trial comparing cisplatin/gemcitabine to cisplatin/alimta showed that while they were remarkably similar in activity overall, patients with non-squamous cancers did better with the alimta doublet, and those with squamous cancers did better with the gemcitabine combination (prior post here). There isn’t any evidence that other commonly used drugs like taxol or taxotere or navelbine work better in one subgroup than another. In truth, they haven’t been tested thorough in this regard, so we really don’t know, but the only thing that I’d really consider convincing is that alimta doesn’t work well at all in patients with squamous cancers and really has its benefit concentrated in the 2/3 of NSCLC patients with non-squamous cancers (prior post here).

Because of this, I am happy to use gemcitabine with a platinum in patients with squamous cancers, but many other combinations may be just as good. For patients with non-squamous cancers, it’s pretty much a dealer’s choice. Alimta may well soon be approved as a first line treatment with platinum for patients with non-squamous cancers, and it’s also a very good choice. I’d certainly consider it, but I have also been very happy with it as a second or third line agent, and I doubt there’s any real advantage to using it first line vs. using it second line for non-squamous patients. I think the carbo/alimta doublet is remarkably appealing from the standpoint of mild side effects and a convenient schedule of treatment for just an hour or so once every three weeks, but we can't presume that the impressive survival with the cisplatin/alimta regimen is the same with carbo/alimta. In fact, the evidence we have is that carbo/alimta has a very typical, unexceptional activity overall. We haven't gotten a breakdown of that combination by histology yet.

A new consideration is erbitux (cetuximab), the IV antibody against EGFR that showed a very modest but statistically significant survival benefit when added to cisplatin/navelbine in patients with advanced NSCLC who tested positive for EGFR protein by an immunohistochemistry (IHC) test (reviewed in prior post here). With a mere 1.2 month survival benefit vs. a weekly treatment with a problematic rash and sometimes other side effects, I’m inclined to consider it primarily for patients not eligible for or inclined to take avastin, which at least showed a 2 month improvement in median survival in the ECOG trial that led to its approval. I’d only be inclined to add erbitux for patients who tested positive for EGFR by IHC, and overall I’m just pretty ambivalent about the real value of erbitux – I think there’s a good chance that benefits will wash out for patients who get subsequent treatment (importantly, the FLEX trial that was positive for survival with erbitux was done mostly in Europe, where second line and later therapies are often not administered). Although I’ve been reluctant to do molecular testing for EGFR to help with decisions about tarceva, I’m starting to send the test for EGFR gene amplification by fluorescence in situ hybridization (FISH), since this MAY be helpful in separating out who will benefit more or less with erbitux (prior post here) . That’s not standard yet, but frankly I’m so on the fence about whether erbitux is worth pursuing that I’d like to use any information available to help with this decision.

Finally, there’s the question of what chemo to use with erbitux if you use it. There’s some modest evidence that appears convincing to me that it may add benefit to a platinum/gemcitabine regimen (prior post here), so I might do that for a patient with a squamous cancer, or any NSCLC. But right now, before erbitux has been formally FDA approved for lung cancer, I’m inclined to use the cisplatin/navelbine combination, both because I think it’s still a fine choice, and because I think it’s got the best chance of being covered by insurance companies still adapting to changing standards.

There are a few other “special case” scenarios that I might carve out and treat differently, such as lifelong never-smokers and patients with bronchioloalveolar carcinoma (BAC). I’ll cover what I really do for those patients in another post.