In a talk at ASCO 2007, I was asked to present some commentary on a couple of phase II, single arm trials of patients with ED-SCLC that were reported by two different cancer cooperative groups in the US, each adding the anti-angiogenic agent Avastin (bevacizumab) to standard chemotherapy options in this setting. One trial, CALGB 30306 by Ready and colleagues (abstract here), added Avastin (15 mg/kg) every three weeks to a chemo regimen of weekly cisplatin and irinotecan (camptosar, CPT-11), each given two weeks out of a three week cycle, for up to 6 cycles, with no "maintenance" avastin alone after stopping the chemo. The second, ECOG 3501 by Sandler (the same Alan Sandler who led the advanced NSCLC trial ECOG 4599 that led to the FDA approval of Avastin in lung cancer) and colleagues (abstract here), combined Avastin at the same dose every three weeks with cisplatin and etoposide, stopping the chemo after four cycles, but continuing with maintenance avastin alone until patinets showed progression. Interestingly, these exact regimens, including the same schedules and doses of the chemo drugs, were compared to each other in a study by Nasser Hanna and colleagues that was published in 2006 (abstract here), so the performance of these chemo regimens in this phase III trial (that showed no significant differences in activity) can serve as a benchmark of what we should expect the chemo to do without avastin. Here's a summary of the two trials side by side, along with the general profile of the patients in each trial:

(Click to enlarge)

As is typical for other lung cancer trials, patients with a history of coughing up blood (hemoptysis) or with evidence of brain metastases were not eligible for these studies. Each enrolled a little more than 60 patients.

As with any trials that include avastin, side effects were a significant focus, particularly the question of whether patients experienced any life-threatening or fatal bleeding episodes. Fortunately, that wasn't the case, and although there were a few patients who died during treatment on each of these trials (issues like infection, multi-organ failure, and other unusual complications seen in perhaps up to 5-10% of patients on ED-SCLC trials), the rates of serious and even fatal complications on these chemo/avastin combination trials were no greater than what we typically see in trials of chemo alone in this setting. The fact that no episodes of serious pulmonary bleeding were seen was particularly notable because one question with avastin has been whether the bleeding sometimes seen in NSCLC is due to the squamous histology (appearance under the microscope, a particular subtype) or the fact that squamous tumors are usually central in the chest, not out toward the chest wall, where more adenocarcinomas are found. SCLC tumors are almost always central, so the fact that bleeding wasn't seen suggests that it has much more to do with the squamous subtype than the central location of the tumor. And we can conclude that adding avastin to chemo was pretty safe, without any obvious new issues compared to the chemo without avastin.

In terms of outcomes, the results were pretty encouraging, but nothing you could consider a blockbuster for either of the studies. It's hard to judge the response rate against the Hanna trial of the same chemo regimens alone, because nearly 1/3 of the patients on both arms of that trial didn't have responses confirmed (an emerging requirement on clinical trials), so they weren't counted as responders. But the numbers in terms of responses and stable disease are pretty encouraging, as are the survival numbers, which are a little better than we tend to see with large trials of chemo alone, but they aren't eye-popping:

The results for all of these situations actually are less impressive than the median survival of more than a year in an important Japanese trial of cisplatin/irinotecan in first-line treatment of ED-SCLC published in the New England Journal of Medicine by Noda and colleagues (abstract here), but we've never seen results like that in the US, and it's becoming clear that there are meaningful differences between populations in Asia and the US in terms of how cancers behave and how they respond to chemo (more on that in a future post). Because of that, it makes more sense to be cautious about comparing trials in different continents and genetic populations.

Putting all of this together, the results with avastin added to chemo seem to be a little better than the results with chemo alone, but these are also pretty small phase II trials, which very often show results far better than the larger phase III experience (for instance, carbo/taxol/avastin had a median survival of 18 months in non-squamous patients in the phase II trial experience in advanced NSCLC (abstract here), and ended up with a median survival of 12 months using the same exact drugs and schedule in the phase III trial that followed it (abstract here). Are these results good enough to run a larger phase III trial of chemo alone vs. chemo/avastin in SCLC? Arguably, it's reasonable, but these results aren't so amazing that CALGB or ECOG are stumbling over themselves trying to move a larger trial forward. They're considering it, and I'd consider it likely that a trial of chemo +/- avastin is likely to be run in the US over the next few years, unless an ongoing SWOG trial of cisplatin/etoposide vs. cisplatin/irinotecan shows that the irinotecan arm looks as good in the US as it did in Japan and does better than these chemo/avastin trials.

You may ask where avastin fits into LD-SCLC, and the answer is that this work is being held up because of concerns about the safety of combining avastin with radiation or chemoradiation for lung or esophageal cancer. I'll talk more about that very soon.