Both standard chemotherapy and EGFR tyrosine kinase inhibitors (TKIs) have been approved in NSCLC, and other anti-EGFR agents like erbitux/cetuximab and vectabix/panitumumab are also commercially available for treating other cancers and are being studied in lung cancer. Iressa was previously approved as a single agent in previously treated patients with advanced NSCLC, and Tarceva is now available but approved also as a single agent therapy. However, some oncologists give EGFR inhibitor therapies in combination with standard chemo. I don't favor that approach, but I think it will be helpful to review the issue and my rationale for avoiding concurrent chemo and EGFR inhibitor therapy, at least until there is some evidence suggesting a benefit of combined therapy.

It's fair to say that treatment combinations are a key cornerstone of medical oncology. The idea of combining anti-cancer drugs with different mechanisms of action and non-overlapping side effects has been central in our field for decades, and combinations are now standard therapy that often are far superior to single-agent approaches. Targeted therapies like EGFR inhibitors, specifically Iressa and Tarceva initially, generally have very different side effects than chemo and were shown to be active as single agents in the second and third line settings and beyond for chemo-pretretaed patients. So there was a great deal of hope that we could improve clinical outcomes for advanced NSCLC patients by adding Iressa and/or Tarceva to standard chemo doublets in previously untreated patients. There were several clinical trials done to test this idea, and the entire lung cancer community was disappointed to see no benefit for combined chemo and EGFR TKI therapy.

The first trials that combined standard chemo and EGFR TKIs was the INTACT-1 and INTACT-2 paired trials. They had the same design, each giving previously untreated advanced NSCLC patients standard chemo doublet (cisplatin/gemcitabine in INTACT-1, and carboplatin/paclitaxel in INTACT-2) either with a placebo or with one of two doses of Iressa (250 mg or 500 mg by mouth daily) for UP TO 6 CYCLES, then stopped chemo and kept patients on their oral treatment (placebo or active drug) alone. The cisplatin/gemcitabine-based trial was conducted outside of the US, as this regimen is most commonly used elsewhere around the world, and the carboplatin/paclitaxel-based trial was conducted primarily in the US; each trial enrolled over 1000 patients. The design of the two trials is shown below:

(click to enlarge)

Unfortunately, both showed the same result, of no benefit at all in terms of better response rate or survival from addition of Iressa to standard chemo (the abstract for INTACT-1 is here, and the abstract for INTACT-2 is here). Survival curves showing no difference between Iressa and placebo arms in both trials is below:

In trying to determine why there wasn't really any good hint of benefit, one idea was that the chemo and Iressa may compete with each other. Supporting this was an analysis where the investigators looked specifically at how the different groups did when looking at results only after 90 days or longer after the start of treatment. This analysis also included just the patients with adenocarcinomas. What you can see is that the curves begin to separate in favor of Iressa after about 8 months, so after chemo is done (18 weeks to get through 6 cycles), patients on Iressa did better than patients on placebo, as shown below:

Well, the combination of chemo with Iressa could have been negative because of interference between chemo and EGFR TKI therapy, or maybe Iressa just wasn't effective enough. After all, since then we've found that Tarceva beat placebo in chemo-pretreated patients with advanced NSCLC, but Iressa didn't, even though they share the same mechanism of action. Well, we can answer that question by looking at the trials with Tarceva that were designed nearly the same as the INTACT trials with Iressa. TALENT and TRIBUTE each enrolled over 1000 patients with advanced NSCLC to receive first line chemo for up to 6 cycles with placebo or Tarceva at 150 mg by mouth daily. As with the INTACT trials, they differed only in that the TALENT trial was done outside of the US, using cisplatin/gemcitabine, while the TRIBUTE trial was done in the US with carboplatin/paclitaxel as the chemo backbone. The trial design for both is shown below:

Well, both trials again showed no hint of a benefit overall for patients receiving chemo and tarceva together vs. chemo and placebo (TALENT abstract here, TRIBUTE here). Here's the survival curve for the TALENT trial (carbo/taxol with or without tarceva), which I find interesting because the tarceva arm is BELOW (as in worse survival) the placebo arm in the first few months, which is when the chemo and tarceva overlap:

And when the investigators analyzed the two arms AFTER 6 months (i.e., after chemo was done), the tarceva arm did better:

So with both Iressa and Tarceva, there's the nagging hint that the benefit from these drugs is present when NOT overlapping with chemo.

I know there are a lot of people who do combine chemo and EGFR inhibitors, including experts at some major cancer centers. I'm not saying they're wrong (at least until we get more evidence to prove it), but look at the results from patients on the TRIBUTE trial with or without EGFR mutations:

The solid red line represents the majority of people, who don't have EGFR mutations, and received carbo/taxol with tarceva. Look at them compared to the solid blue line of patients with no EGFR mutation who got chemo alone. Hmmm....that red line with tarceva is looking worse than placebo, don't you think?(What's also interesting is the fact that patients with EGFR mutations did just as well with chemo alone as they did with chemo and tarceva, but that's another story).

The only place where there is some evidence of chemo and tarceva looking good is in the never-smokers on the TRIBUTE trial, who had a more than doubling of survival with chemo + tarceva compared with chemo + placebo:

That's impressive, but the real question is whether they would have done just as well or better with tarceva alone. Look at where the curves separate and where the tarceva recipients do better, which I've marked with an arrow, because it looks to me like it's about the point when the chemo is done:

I'd contend that the never-smokers did so well despite getting chemo at the same time, not because of it. Regardless, a trial is being done in never or light smokers that is comparing chemo/tarceva to tarceva alone, so I hope we'll clarify which is the better approach. If I'm wrong, I'll be happy just to have an answer, but my money's on no combo benefit or even a detrimental effect.

There's actually test tube evidence that some forms of chemo, particularly taxanes like taxol and taxotere, have an antagonistic (interfering) relationship with EGFR tyrosine kinase inhibitors in terms of the damaging effects on cancer cells. That's still investigational, but it provides some explanation for why EGFR inhibitors and chemo can both the active against lung cancer cells, but not work well when you give them together. Some trials are now giving chemo and tarceva but holding the tarceva for a couple of days before and after the chemo is given. This is called "pharmacodynamic separation" and is being studied at the University of California at Davis, one of the leading lung cancer centers in the country, and also in New York by Dr. Roman Perez-Soler at Montefiore Medical Center, who has been a leading researcher with tarceva from the very beginning.

I would also say that it's not clear that these problems apply to antibodies against the EGFR inhibitors like erbitux and vectabix. They've been studied more in colon cancer and can be given with chemo, seemingly with added benefit. This class of drugs has been studied much less in lung cancer, but we'll get more info about them in the next year or two.

As always, I welcome your thoughts, comments, questions, and objections about this very controversial area of lung cancer treatment.