Last week I ran a three part series about maintenance therapy:
The series prompted this thoughtful response from one of our readers:
I got curious about the McCarthy et al. article that was the basis for much of the recent maintenance discussion so I got a copy of the full text. There are a couple of points that I want to share with you.
It finally dawned on me (duh) why all the clinical studies focus on medians rather than means (averages). The median can be reported as soon as half the subjects pass the specified end point (e.g., disease progression) whereas calculation of the mean would require that all the patients pass that end point. In many cases this would take an unreasonably long time so results from the study couldn’t be reported in a timely way. You’ve no doubt understood this for a long time but it was a new discovery for me.
The McCarthy et al. study was designed to look at time to disease progression and was not specifically intended to look at overall survival. Again this was probably motivated by concerns about timeliness of the results. In the 2012 paper more than half of the subjects in both arms were still alive so median survival could not be calculated.
Even though median survival isn’t available, it’s still possible to look at survival in other ways. For example, the paper reports that 3 year survival in the lenalidomide arm was 88% while the figure for the placebo arm was 80%. But if we take this one step further we can use the graphs in the paper to see that the lenalidomide arm didn’t hit 80% survival until approximately 48 months (it’s hard to be exact taking numbers off the graph). That’s a full year after the placebo arm and suggests that overall survival may be significantly greater in the lenalidomide arm when a median can be calculated. Further out on the graph (that is, past 48 months) it looks like the two curves start to come closer together, but I suspect the data is unreliable because not enough patients are that far along in the study. The study had a staggered start, adding subjects through time, so each patient is on his or her own timeline determined by time since transplant.
Finally it’s important to consider that the study was unblinded after only 18 months of follow-up, and at that time 86 of the 128 patients in the placebo arm without disease progression chose to start taking lenalidomide. So there really is no longer a pure placebo arm of the study since 67% of the subjects who hadn’t progressed began to take lenalidomide. As the authors said “The increase in time to progression led to early study unblinding, and despite the crossover, benefits with respect to progression and overall survival were seen in patients receiving lenalidomide maintenance.”
Holt was kind enough to forward me a link to a free preview summary of the study, courtesy of The New England Journal of Medicine, which normally charges for access to their content. CLICK HERE to view.
Reader perspective is so valuable, don’t you think? The comments and emails I receive add so much to our site.
This isn’t the first time Holt has contributed. Here’s a link back to Holt’s story that I ran back in July:
Thanks again, Holt! You make a good point. There does seem to be an overall survival benefit from using maintenance following stem cell transplantation.
Does that translate to other forms of maintenance therapy? Hard to tell, but my guess is yes.
But I’m not ready to give-up on the perspective that extended drug-free holidays aren’t a good idea for some patients. The study Holt sites does show an overall survival benefit statistically. But if a patient stays healthy and doesn’t allow their myeloma to unexpectedly become active and sneak-up on them, doesn’t that alter individual outcomes?
A lot of this should get sorted-out in the next few years. Here’s hoping that each and every one of us are here to read it!
Feel good and keep smiling! Pat