As I described in my last post, there is a strong consensus that overlapping chemotherapy (CT) and radiation therapy (RT) provides greater efficacy, meaning higher survival rates, than a sequential, non-overlapping approach for stage III, unresectable NSCLC. Beyond that, it’s a bit of a mess, with a wide range of choices and no clear “right” choice.

We know that most of the larger trials with CT/RT that have defined our standards have used cisplatin with older partner drugs like vinblastine or etoposide (also sometimes referred to as VP-16). One key advantage of these drugs is that they can actually be given at full dose with radiation at the same time. This means that the chemo can effectively combat potential micrometastases in the bloodstream, as well as provide radiosensitizing benefits, making the RT more effective in the area in which RT is directed (the "radiation field"). In contrast, other commonly used regimens during RT, most commonly carboplatin and paclitaxel, are given at low doses every week, which we don’t believe are systemically effective doses – in other words, although lower doses may make RT more effective, the attenuated doses aren’t enough to kill micrometastatic tumor cells that may lead to later distant (outside of the RT field, such as bone, liver, adrenals, brain, and other places metastases may settle).

Although the long-term survival rates with CT/RT have improved modestly over the past decade or two, we can all agree that there’s LOTS of room to try to improve. One of those ways might be to add more chemo before or after the 6-7 weeks of concurrent CT/RT, particularly if that gives us a better chance of giving good, systemic doses of chemo able to eradicate more distant cancer cells. I’ll cover the ups and downs of follow-up, or consolidation, chemo now.

The Southwest Oncology Group, or SWOG, one of the US-based cancer cooperative groups that includes several hundred participating sites across North America, has led this approach. Back in 1990, they started a 50 patient trial (abstract here) in which they gave a total of four cycles of cisplatin/etoposide, SWOG’s gold (or some people would say just old)standard chemo for stage III NSCLC, along with full dose RT to 61 Gray for the first two cycles. The results were OK, with a median survival of 15 months and three year survival of 17%, fine but not spectacular in the context of the preceding trials like the Dillman study (abstract here) I had already described in my prior post. What SWOG followed up with in the mid-1990s was the idea of keeping the cisplatin/etoposide/RT backbone for the first two cycles, but then switching from two more cycles of the same chemo to three cycles of single-agent taxotere (75 mg/m2 every three weeks), a different chemo with a distinct mechanism of action that might be more effective than using more of the same chemo the cancer has already seen. The trial enrolled 83 eligible patients with stage IIIB NSCLC (“dry IIIB”, with no malignant pleural effusions) and captured the attention of the lung cancer world despite the fact that this was only a relatively small phase II trial because the investigators saw a median overall survival of 26 months (remember, the preceding trials were showing median survivals of 15-17 months) and a 5 year survival of 29% (abstract here).

(Click to enlarge)

The key thing to note from the curves is that there is a tail on the right side of them, a plateau, that is well above the bottom line. We like plateaus on survival curves because they mean that people aren’t having recurrences and dying even over a longer period of follow-up – plateaus suggest cures. Now, truth be told, there were a few treatment-related deaths, about 5%, primarily from severe pneumonitis, which is inflammation of the lung tissue, generally in the region that got radiated. It’s still debated whether taxotere increases the rate and severity of pneumonitis, since people get plenty of pneumonitis weeks to months after CT/RT is completed (right during the time that they’d be getting taxotere) – the available evidence doesn’t really support that, but I am quite careful with taxotere in this setting and have often stopped patients before the full three cycles were given when I see them getting pneumonitis, since I believe it may worsen with each subsequent cycle, like adding fuel on a smoldering fire.

The initial reports from the SWOG 9504 trial emerged about 6 years ago, and people debated whether these results were so favorable that this approach should be the standard of care. On the favorable side, the more recent study had the same eligibility and enrolled from the same places as the preceding but much less exceptional SWOG 9019 trial, with the only difference being that the taxotere consolidation replaced the third and fourth cycles of cisplatin/etoposide:

Well, OK, the other variables were that there was a five year difference in when these trials were conducted and during that time, there may have been improvements in things like supportive care (use of growth factor support like G-CSF, better tolerance of chemo because of better anti-nausea drugs, etc.) and steadily increasing use of PET scans that may have upstaged patients who might have been classified as stage I, II, or IIIA without the benefits of PET scans.

Some sensible, smart people felt that the evidence wasn’t strong enough to change our management for stage III unresectable NSCLC and give consolidation taxotere to everyone. The trial on which this was all based had only 83 patients who may have been heavily selected and may not represent the real world experience. After all, the patients on the SWOG trial had to have very good lung function, a higher bar than most preceding trials required, so maybe this was just representative of how well the fittest of the fit patients with unresectable stage IIIB NSCLC could do. Moreover, the S9504 trial didn’t have a comparator arm, and the good results could potentially be explained in other ways. Nevertheless, in the absence of a better alternative, and feeling that this was the best chance to improve cure rates, I and the majority of US oncologists made consolidation taxotere our preferred way to treat healthier patients with unresectable stage III NSCLC. I had a few patients with quite severe pneumonitis, and many patients had pretty significant drops in their blood counts, but I didn’t have patient deaths, and I felt like something in the range of a third of my patients were doing really well.

But then something troubling happened. We got new information at ASCO 2007, and it really caused us to step back and reconsider what we thought we know. I’ll cover that in my next post.