How Roche Hopes to Sow Confusion in Test of 2 Blindness Drugs

A long, cynical chapter in the history of Roche (RHHBY)'s Genentech unit will come to an end Sunday when the National Eye Institute reports the results of a major clinical comparison of two drugs for blindness, Avastin and Lucentis. Researchers are expected to announce that the two drugs, both made by Roche/Genentech, are "roughly as effective" as each other.

That's a problem for Roche because Avastin costs only $50 a dose whereas Lucentis costs $2,000. Avastin isn't indicated for age-related macular degeneration; Lucentis is.

For years, Genentech and Roche have pretended that the two drugs are significantly different and that the elderly -- and the taxpayers and insurance companies who pay for their care -- should stick with the pricey Lucentis and not use the cheap Avastin.

Roche earned about $1.5 billion selling Lucentis last year. In theory, that entire franchise could go up in smoke if reimbursers decide that AMD patients should use Avastin first. Here's how that Lucentis franchise has worked out for U.S. taxpayers:

In 2008, Medicare paid for 480,000 injections of Avastin to treat macular degeneration and 337,000 injections of Lucentis, according to a study led by Dr. Philip Rosenfeld of the University of Miami. Yet Medicare paid only $20 million for the Avastin compared to $537 million for the smaller number of Lucentis injections.

I use the word "die" only in a mathematical sense, the study actually shows that the risk of death in users of both drugs is 11 percent greater for Avastin patients. Of course, AMD is a disease of the elderly, and the elderly die fairly frequently. Luckily for Roche, a few more of them died while on Avastin than Lucentis.

The Lucentis death study ought to be taken with a pinch of salt given that all of its five authors have been on Roche's payroll, and the study itself was funded by Roche. (The NIE study is independent, and we don't yet know what secondary results lurk within it.)

In addition, the abstract of the Roche study only reports "relative" risks of death or cranial hemorrhage, which might look a lot less scary when expressed in absolute terms. For instance -- and these numbers are merely illustrative -- if Lucentis patients face a 0.1% chance of death, the risk for Avastin patients is just 0.11% (11% greater than the risk in the Lucentis group). Don't count on the mainstream media to clarify that anytime soon.

Roche doesn't care about any of that. It knew the NEI study was coming out soon and needed some clinical info to counteract it ahead of time. The authors' pro-Roche bias can't be proved unless someone repeats the study independently, and as no one will bother, Roche can maintain the fig leaf that the drugs are different and that it is in some way riskier to save everyone money by using the cheaper drug.

Related:

• Why Roche Is Excited by a Blindness Drug's Flaws But Not Its Effectiveness
• The Avastin Cancer Party Comes to an End as Hangover Kicks in
• Why a Charity for the Blind Opposes a Cheap Drug for Sight Loss (Hint: Big Pharma Cash Is Involved)