In a post several months ago, I described the results of a trial from Japan, designated V-15-32, that directly compared Iressa to Taxotere as a second line therapy. Although overall comparable, the study showed that Japanese patients receiving Iressa had a higher response rate, but despite that had a lower median and one year survival. Although Iressa overall has not shown the same degree of clinical benefit as the very similar drug Tarceva, these results were perhaps a bit surprising because EGFR inhibitors have been most effective in Asia, where a higher proportion of lung cancer patients are never-smokers and/or have an EGFR mutation. But the Japanese trial had the problem that more patients randomized to Taxotere received a potentially effective therapy (different agent with potential activity in previously treated patients) after the trial drug than patients on the Iressa arm.

One of the trials presented at the Presidential Symposium at the recent World Conference on Lung Cancer in Seoul, Korea, was called the INTEREST trial (I don’t remember what the acronym stands for, except that it was a major stretch), led in North America by my friend Ed Kim from MD Anderson Cancer Center. This trial (abstract here) had almost the exact same design as the Japanese study, again randomizing previously treated patients to either Iressa 250 mg daily or Taxotere IV every three weeks.

(Click to enlarge)

The INTEREST trial was run around the world, accruing 1466 patients from 149 centers in 24 countries, making it the largest trial in previously treated patients with NSCLC that has ever been conducted. The Taxotere dose was also 75 mg/m2 every 3 weeks, the FDA-approved standard in the US based on a proven survival benefit in North American trials, while the Japanese V-15-32 trial used 60 mg/m2, which is the standard in Japan and appears to be a dose with the same general toxicity and effectiveness as a higher dose in North American/European populations. Otherwise, the INTEREST trial asked the same question as the Japanese trial but in a broader audience. Specifically, one of the reasons it was so big is that it was looking to see whether the two approaches could be proven to have the SAME survival/clinical benefit, which requires more patients than showing that one is better than another. The trial was also designed to look at whether patients with EGFR gene amplification as measured by the FISH test.

The vast majority, over 80%, received this therapy as second line treatment as opposed to later therapy. Never-smokers comprised 20% of both arms, and just over 20% on the trial were Asian patients. A little more than 50% of patients had adenocarcinomas. As expected for such a big trial, the two arms were balanced in their randomization.

The results overall were very similar for the two treatments. The response rate was slightly higher with Iressa, 9.1% vs. 7.6%, consistent with what you’d expect outside of Asia. The median survival was about 8 months in both groups. Importantly, about 50% of the patients on the two arms received potentially effective later therapies, so this was not an imbalance as occurred in the Japanese trial. This may have been a factor in why the two treatments produced the same survival here but not in the smaller Japanese trial. It's also possible that the larger, worldwide INTEREST trial produced a clearer picture of the overall story.

What was most interesting was that there weren't subgroups that really did much better with one treatment than another (the small proportion of patients who had received more than one prior therapy did better with Taxotere, but it's not clear what this really means or if it makes sense -- small groups can give quirky, less reliable results). Asian patients, never-smokers, and even the patients with EGFR gene amplification or mutations all did about the same whether they received chemo or Iressa. Nor did the patients with K-ras mutations, who might have been expected to do particularly poorly with Iressa (see prior post on ras mutations and EGFR resistance), also didn't do any better or worse with Iressa vs. Taxotere.

Of factors other than survival, quality of life was rated a little higher by recipients of Iressa. They also had less frequent side effects overall, and they were different between EGFR inhibitor therapy and chemo. Taxotere gave more decreased blood counts, fatigue, nausea, hair loss, neuropathy, mouth sores, fevers, constipation, and muscle aches; Iressa led to more rash, dry skin, and diarrhea.

What does this really mean? Reviewing the overall evidence about EGFR inhibitors and taxotere or alimta in previously treated NSCLC, my conclusion before this trial was already that they were pretty much equivalent. I did think Iressa was a little less active than Tarceva, at least at the 250 mg dose, so it's impressive to see the weaker sibling EGFR inhibitor fare as well as established chemo -- after all, this same dose of Iressa failed to show a survival benefit compared to a placebo in a progressing patient population after first line treatment of advanced NSCLC. I was also somewhat surprised to see that the biomarkers for EGFR and ras didn't distinguish between patients who would do better with Iressa or Taxotere, so maybe that won't pan out as an important way to tailor therapy; clinical factors like Asian race or never-smoking status also showed very similar results for the two groups.

Based on these results, it's possible that Iressa will claw its way back into the US market. However, I still haven't seen any reason to favor Iressa over Tarceva, except for milder side effects (likely because the 250 mg dose is similar to being on a lower dose of tarceva than the standard 150 mg daily, for better or for worse). The results from the INTEREST trial are indeed interesting, but I don't think they will change my approach to treating lung cancer patients.