The last topic in our discussion of the evolving field of optimal treatment for locally advanced NSCLC is the potential role for induction chemotherapy before radiation, or, more commonly concurrent chemo/radiation (CT/RT). In my recent post, we covered the new findings from ASCO 2007 that our common practice of giving concurrent CT/RT followed by consolidation chemo, most commonly with taxotere, is not necessarily the best treatment for most patients with unresectable stage III NSCLC. Perhaps giving chemo before the concurrent CT/RT is a better approach. After all, the trial that established a role for chemo in stage III NSCLC back in 1990 first gave chemo, followed by radiation (summarized in another recent post).

Since then, while some investigators have used the SWOG approach of initial CT/RT followed by more chemo, others have used the reverse, starting with chemo and then continuing on to concurrent CT/RT. Some of these trials have been pretty promising. Dr. Everett Vokes, who chairs the Lung Cancer Committee for the Cancer and Leukemia Group B (CALGB) published a trial that gave two cycles of cisplatin with any of three different partner drugs -- taxol, navelbine, or gemcitabine -- followed by the same combination at lower doses with concurrent chest RT (abstract here). The trial was designed as shown, and as you can see from the results slide, all three of these combinations performed similarly, although there was a lot of toxicity when gemcitabine was combined with radiation (at the time it wasn't known how extremely radiosensitizing gemcitabine is). Here's the trial design and results:

(Click to enlarge)

Those results are all pretty good. Another trial, done by Dr. Mark Socinski and his colleagues at the University of North Carolina (abstract here), gave two cycles of carbo/taxol followed by weekly carbo/taxol along with concurrent RT to an escalating dose, beyond the typical 60 Gray (Gy) range and all the way up to 74 Gy (they're very good at radiation therapy there and do a lot of pioneering work at UNC):

The results looked quite impressive, with 40% of patients on the trial alive three years later, but this was a single institution trial with just 62 patients. We'd need to see how something similar to this approach works in the real world, with more participating centers.

The Locally Advanced Multimodality Protocol, or LAMP trial, gave us some insight about this (abstract here). This trial randomized 268 patients with stage III unresectable NSCLC to chemo followed by radiation, induction followed by concurrent CT/RT, or initial CT/RT followed by consolidation chemo, with the chemo being carboplatin/paclitaxel for all three arms:

The trial closed early, because it was becoming increasingly problematic to enroll people on a trial with sequential chemo and radiation as we became convinced that concurrent CT/RT was superior. But in an early analysis of the data, the arm that got induction chemo before concurrent CT/RT was the arm that was found to do the worst and closed before the other two. Here are the final results for the trial:

But why did the induction arm do so poorly? Perhaps it had to do with the fact that patients who went from chemo to CT/RT were less likely to get as much overlapping chemo with their radiation, compared with those who started with the overlapping CT/RT and then went on to chemo alone, as shown here:

My interpretation was that it's easier to start with the harder concurrent approach and then go to the easier chemo alone. Perhaps it's especially hard to get through chemo, then move on to something even harder. A lot of people can't do it.

So that's where we were until very recently. Some new trial results have emerged, and we'll end our discussion of locally advanced NSCLC with that next post.