This is a quick follow-up to this post from a couple of days ago. I’ve now read the two articles referred to, and agree that there’s some funny stuff going on here. I don’t really doubt the conclusions, but I think the authors really do have to explain some things.
Now that I’ve read them, I’ll identify the two articles referred to. They are (all links require subscriptions):
Klareskog L, van der Heijde D, de Jager JP, Gough A, Kalden J, Malaise M, et al. Therapeutic effect of the combination of etanercept and methotrexate compared with each treatment alone in patients with rheumatoid arthritis. Lancet 2004;363:675–81.
van der Heijde D, Klareskog L, Rodriguez-Valverde V, Codreanu C, Bolosiu H, Melo-Gomes J, Tornero-Molina J, et al. Comparison of Etanercept and Methotrexate, Alone and Combined, in the Treatment of Rheumatoid Arthritis. Arthritis & Rheumatism 2006;54:1063–1074.
The trial referred to is the “TEMPO” trial (I hate these acronyms!), which evaluated methotrexate (MTX), etanercept, and both for rheumatoid arthritis (RA); i.e., subjects got one, the other, or both.
The authors of the letter I referred to (Yazici and Yazici) basically got it right; there are indeed some funny things going on. It is true that the Lancet article did not note anywhere that this trial was designed as a three year study, though I do note that they say they’re presenting ’52 week results”, perhaps implying that there will be more data. They also do not note how investigators and patients were kept blind, and, what I find most concerning, the primary endpoint in the 2nd trial is not the same as that in the first trial. Though I’m not sure this throws the results into question, it is queer.
The other thing I note is that this study is sponsored by the manufacturer – Wyeth – that they are “responsible for the collection and analysis of data, and that this study was required “as a postapproval commitment to the European Agency for the Evaluation of Medicinal Products”; i.e. as a longer-term follow-up, presumably to monitor safety.
So, in general, I agree with the letter writers’ questions. I also have a few of my own. The authors say that to measure the progress of joint damage, the primary outcome was defined as “linear extrapolation of actual change from the baseline.” I’m not sure why they did this, rather than using the actual change. It seems weird to me. In addition, in the first paper they say that “estimated yearly total Sharp score [for joint damage] progression was defined as score at baseline divided by duration of disease for every patient.” Once again, I don’t understand why they’re dividing by duration of disease. There may be a good reason for this but the authors have not told us why.
So anyway, I agree that this is a weird trial, and I’d like to see the trial authors respond to this letter. I don’t really doubt the results. They found that subjects on both drugs did better generally than those on either one alone, and this makes sense. But I would like to know more about how the trial was designed and conducted.