One of the highest profile clinical studies over the last few years has been the START trial of Stimuvax, also known as L-BLP-25 or tecemotide, an immunotherapy that looked promising in a randomized phase II trial that led to a subsequent phase III trial that administered Stimuvax or placebo after chemo and radiation for locally advanced (stage III) NSCLC, as described more in this post about STIMUVAX and the START trial from early 2007.
Over the next five years, the level of anticipation over the trial was remarkable. Though oncologists specializing in lung cancer were hopeful about it, financial analysts and patient/caregiver communities became exceptionally focused on the trial, many people becoming essentially convinced it was a “can’t miss” trial long before the trial was completed.
Unfortunately, we received news at the beginning of this year that the START trial failed to meet its primary endpoint of significantly improving overall survival with Stimuvax. Without any specific information of results beyond this, I did a video post to address the implications of the negative START trial, primarily noting that a negative trial for Stimuvax doesn’t mean that other immunotherapies aren’t promising and won’t work.
At ASCO 2013, we actually saw the data from the trial, which led me to question whether these results should be boiled down to just being considered “positive” or “negative”, effective or useless. Specifically, the trial showed a modest, 3.3 month longer median overall survival in the 829 recipients of Stimuvax compared to the 410 recipients of placebo (the randomization was 2:1 for the active treatment), but there was a whopping 10.2 month difference in favor of Stimuvax recipients when the analysis was limited to the 65% of patients on the trial who had previously received their chemo and radiation concurrently, as opposed to sequentially.
When these data were presented at the ASCO meeting, I noted that a 10.2 month difference in overall survival is rather remarkable, even if it was a subset analysis (806 patients) of a larger trial. That’s quite far from a complete strikeout. In fact, I found it very encouraging. However, when you look at the survival curves in the lower figure, you can appreciate that the curves just happen to separate hugely in the middle, then converge by 4 years out.
Unfortunately, this means that Stimuvax isn’t curing patients, which is what we really hoped might happen. Instead, a subset of patients may be doing better for a few months, but it’s not converting more potentially curable patients from “not cured” to “cured”.
Word on the street is that Merck KGA, the company developing Stimuvax, or tecemotide, is planning a START trial redux that only enrolls patients who received concurrent chemo/radiation. We don’t know why the results seem to be different and more favorable in these patients, but this is the preferred and more curative treatment approach, so there may be some significant difference by which the locally advanced NSCLC patients who are good candidates for concurrent chemo/radiation truly benefit from this immunotherapy. However, even if a subsequent trial shows a survival difference, it will be important to clarify whether this therapy is actually curing patients who wouldn’t otherwise be cured or “merely” prolonging survival for several months. The latter is still a helpful development, but it’s not as valuable as curing more patients with lung cancer.