As a follow-up to my last post and an end to this extended discusison of locally advanced NSCLC before moving to other topics, I'll just cover some more recent work on the topic of chemo followed by chemo and radiation for stage III NSCLC. In that post, I showed that over the past several years, we hadn't seen the promise of this induction strategy in small trials translate into especially favorable results in larger multicenter trials.

More recently, we've seen the CALGB continue to study this approach, in trial 39801, presented by Dr. Everett Vokes from the University of Chicago at ASCO several years ago and then published very recently (abstract here). This trial really directly tested the value of induction chemo, because the trial randomized just over 360 patients with unresectable stage III NSCLC to either weekly carbo/taxol and concurrent RT for 7 weeks, or two cycles of carbo/taxol (one day every three weeks at a higher dose) followed by the same weekly chemo and concurrent RT. The trial showed no significant differences between the two arms, no improvement in survival, progression-free survival, or distant failures (which we might presume would be lower with the induction chemo strategy, because I had previously mentioned that many of us are concerned that low-dose weekly carbo/taxol may not treat micrometastases effectively). These results indicated that adding the higher doses of chemo didn't add anything but more side effects, which included significantly more low white blood cell counts and more pneumonitis (inflammation of the lung tissue). Here's the summary of the trial:

At the time of his initial presentation of the trial at ASCO a couple of years ago, Dr. Vokes noted that the 12-14 month median overall survival was pretty disappointing, and he raised the question of whether that might be because a carboplatin-based regimen was used instead of a cisplatin-based regimen, since trials in stage III NSCLC that use cisplatin-based chemo and RT together have recently shown median survivals in the 19-21 month range. I think there's a real possibility that this is the case, but some others have suggested that the lower survival is because somewhat less fit patients were channeled into this trial than the cisplatin-based trials. Some of the investigators were concerned that just seven weeks of low-dose carbo/taxol with RT wouldn't be enough chemo treatment, so more marginal performance status patients went on this trial, while the healthier ones might have been directed to more aggressive approaches. So here's an example where selection bias may have led to lower survivals than you would otherwise expect. However, I still prefer to give cisplatin-based chemo in the curative setting, if the patient can tolerate it.

Another trial presented at ASCO 2007 by Kim and colleagues from Korea also tested induction therapy (abstract here). This study enrolled 134 stage III NSCLC patients who either received cisplatin/taxol and concurrent RT for 6 weeks with no other treatment, or two cycles of cisplatin/gemcitabine as induction chemo, followed by the same cisplatin/taxol and RT:

(Click to enlarge)

It's not a large trial, but it showed no hint of benefit for the induction approach. In fact, the induction chemotherapy arm had a significantly worse progression-free survival and a strong trend toward a worse overall survival:

Why would that happen? Potentially because they saw the same trend that I described with the LAMP trial in my last post: the patients who went from chemo to chemoradiation had a much harder time getting through all of their planned chemoradiation, with only 69% of patients getting all six cycles of chemo with RT in the group starting with induction chemo, vs. 86% in the group that just did concurrent chemoradiation.

So that's where we are. No hard evidence that additional chemo before or after a mere 6-7 weeks of chemo with RT improves survival. I suspect that part of the problem is that this is challenging treatment, and it's a point where people can be limited by the side effects of treatment pretty readily when their performance status isn't great and the treatment is rough. I believe it's quite possible that the fittest patients may do better to get more chemo, but those struggling to just get through 6-7 weeks of chemo/RT together may be better served by concentrating on the highest yield treatment and avoiding treatments that push them into the danger zone where treatment can be as threatening as the disease. But in the meantime, I've taken a step back from my presumptions and now can say just that my preference is for two cycles, or about 6 weeks worth, of cisplatin-based chemo, and concurrent RT. The rest may not buy you much, if anything.

Next we shift to something different, a new approach to personalizing cancer treatments.