Can't See the Trees for the Forest

• Email
• Share
• Print

I'd like to introduce a new word into the lexicon: meta-stent -- a noun meaning information used to prop open the door of closing stent sales.

Last week that information came in the form of a meta-analysis published in the medical journal The Lancet. The article concluded that drug-coated and bare-metal stents present similar mortality risks. Further, the report said that Johnson & Johnson's (NYSE: JNJ) Cypher drug-coated stent was associated with less late-stent thrombosis -- blood clots -- and heart attacks than Boston Scientific's (NYSE: BSX) rival product Taxus. Medtronic and Abbott Laboratories emerged without a press release -- neither the kind that touts nor the kind that defends -- from the article since their stents weren't included in the analysis.

Meta-analysis looks at data from multiple clinical trials to find results not seen in the individual trials. With statistics being a numbers game, the larger number of subjects in the combined results can reveal statistically significant data not seen in individual trials. Meta-analysis has been in the news recently since Dr. Steven Nissen found an increased risk of heart attacks in patients taking GlaxoSmithKline's (NYSE: GSK) Avandia.

The meta-analysis published in The Lancet is even more complex in that it not only combines data from multiple trials, but it compares Cypher to Taxus by seeing how well each does against bare-metal stents. While it's fun to make those types of comparisons -- my Foolish colleague Brian Lawler did it just the other day -- he and I would both agree that it's nothing more than a general comparison.

The problem is that patients taking a placebo -- in this case bare-metal stents -- don't respond exactly the same in each trial. For these trials, the bare-metal stents had different complication rates -- those in the Cypher trials reclogged more often than those in the Taxus studies -- which makes it difficult to compare the two drug-coated stents.

For companies, the problem with meta-analyses is they have no control over the data release. The study gets published when the researchers are done, and companies have to scramble to do damage control or tout the data if it's positive for the company. That can often leave other drugs the sales force sells up a creek without a paddle.

Whether this meta-analysis will help Johnson & Johnson's falling stent sales is anyone's guess. If the sales force can spin the data well enough, it might be able to turn things around, but doctors may be able to see holes in the study that are much larger than the stents they're implanting.