Although consolidation taxotere after concurrent chemo and radiation therapy (CT/RT) emerged as the preferred treatment approach for about 2/3 of American oncologists over the last few years, this was predicated on an incomplete story. We received information from an additional two studies this year, and now it’s a big mess.

In the wake of the small but impressive SWOG 9504 trial (abstract here) that I described in my last post and that suggested a survival benefit from consolidation taxotere, the next trial that needed to be done was a direct test of taxotere. SWOG wanted to do such a trial, but the government agency, the Cancer Therapy Evaluation Panel (CTEP) that oversees cancer cooperative group trials didn’t let SWOG do that trial, because CTEP felt the question was too passé. Instead, they felt that we should presume consolidation taxotere is the new standard, and then test a sexy new drug like Iressa as a maintenance therapy after concurrent CT/RT and taxotere. This trial took shape as SWOG 0023, with the design shown here:

(Click to enlarge)

As I described in a prior post, the trial closed early, after a trial of Iressa in the advanced NSCLC setting showed no significant benefit compared to placebo. This closure was because the Data Safety Monitoring Board (DSMB) overseeing the trial reviewed early data and saw that there was no way that the Iressa arm would do better than the placebo arm. My friend Dr. Karen Kelly, now heading the heme/onc program at Kansas University Medical Center, presented very early data on the trial at ASCO 2005 (abstract here), when the overall survival for all enrolled patients on the trial was 19 months, shy of the 26 months seen in the SWOG 9504 trial:

There were several important points from her initial presentation, including the fact that less than half of the patients enrolled made it through CT/RT and consolidation taxotere, and also that the patients who received maintenance Iressa did worse (not quite statistically significantly so, but close):

So this established that there wasn’t any role for Iressa (or, arguably, Tarceva) as a maintenance therapy in an unselected stage III NSCLC population, but it didn’t say anything about the value of Taxotere, because everyone got it.

But meanwhile, the Hoosier Oncology Group based in Indiana, or HOG, as it’s affectionately known, conducted its own trial that directly tested taxotere in the consolidation setting. This trial accrued 203 patients with unresectable stage III NSCLC to receive cisplatin/etoposide/RT, just as SWOG does it, and then randomized patients to either no further therapy or consolidation taxotere for three cycles:

Like the SWOG trial, it closed early, when the DSMB reviewed the preliminary study results and found that it would not show a significant survival benefit for the taxotere arm. Like the SWOG trial, many patients who started the trial came off along the way because of disease progression or toxicity issues, leaving just 147 of the 203 to be randomized to observation or taxotere consolidation after CT/RT. Another friend in the lung cancer community, Dr. Nasser Hanna from Indiana University, presented the results at ASCO 2007 (abstract here). He showed that there was no significant benefit in terms of either progression-free or overall survival for the group that received taxotere. It wasn’t that the trial showed a modest benefit that wasn’t enough to be statistically significant in a relatively, small and underpowered trial (which it was), but rather that the trial showed no hint of a benefit from taxotere consolidation:

Unfortunately, taxotere was associated with significantly greater toxicity in many parameters, such as infections, severe pneumonitis, hospitalizations, and treatment-related deaths:

Dr. Karen Kelly then presented an update of the SWOG 0023 data (abstract here), which now show that the detrimental effect of Iressa is statistically significant with longer follow-up:

This worse survival in recipients of Iressa wasn't from increased toxicity, but rather appeared to be more from a higher rate of cancer recurrence with Iressa, which we really can't explain.

So where does this leave us? Still grappling for the truth, I'd say. You can see that some patients on the S0023 trial did remarkably well: the patients who received Iressa maintenance therapy had a median survival of 35 months (versus 23 months for the recipients of Iressa maintenance therapy), but that's an artificially high number because this included just the people who showed no progression through chemoradiation and then consolidation therapy, so it cherry picked the most favorable patients. What we can see is that the overall results for the entire SWOG 0023 trial were not very different from the resilts on the HOG trial, and that there is no evidence that adding extra chemo after two cycles of cisplatin-based chemo with definitive concurrent RT improves survival.

After several years of giving consolidation chemo, I and many oncologists have developed a sense that more treatment must be better, because we know that the majority of people who receive chemoradiation without additional chemo still develop recurrent cancer. But we don't have a better treatment approach, and that will leave some people still giving consolidation chemo (perhaps trying different therapies rather than taxotere, which can be challenging in terms of side effects), others stopping after just 6-7 weeks of concurrent CT/RT, and others reviewing options and perhaps treating unresectable stage III NSCLC patients in a more individualized way (suggesting more treatment only for the particularly healthy patients). New trials, like the Stimuvax vaccine study described in a prior post are also testing very novel alternatives as consolidation therapy. In the meantime, I think we're going to see a lot more variability in the treatment recommended for patients with locally advanced, unresectable NSCLC.