"I've seen numbers that 4-6% of the worldwide CF population (70,000) has the G551D mutation. Let's call it 5%, but let's be conservative and say only 4% treat (and that no patients take VX-770 off-label for other mutations). So 4% (2800 patients) treat at $180,000/year (again, I think this is conservative for an ultra orphan drug), giving Vertex revenues of $504 million/year. Let's say that's $400 million/year gross profit, subtract $50 million for SG&A (VX-770 will sell itself, IMO) and $140 million for tax, leaving $210 million/year in income from VX-770."
I've changed my thinking on a couple of points. I was definitely off on gross margin. I was thinking of typical pharmaceutical margins, where 80% might be expected. With ultra orphan drugs selling at such high prices, the cost of goods becomes negligible. (Genentech, with a number of orphan drugs, used to have gross margins around 91%, despite the fact that biologics are expensive to manufacture.) So I'm going to say VX-770 will have gross margins of 95%, and gross profit of ~$475 million year.
Also, Vertex makes it sound like they might get European approval for other gating mutations in addition to G551D. I'm just going to take a wild guess at how many extra patients that brings in. Let's say 30,000 of the CF patients are in Europe, and 10% have non-G551D gating mutations. That extra 3000 patients could more than double the sales and gross profits numbers I list above.
You produced another masterpiece, congratulations.
I would only like to point out that 770 will also be used together with developing correctors. In the next 5-10 years correctors such as 809, 661, or competitor drugs will be developed to treat 508 CFers. Since correctors need 770 to potentiate CFTR, the market for 770 will expand to treat 70% of all CFers as well. This means that the market for 770 will expand 1750% as soon as a corrector becomes available. 770 merits a high multiple.
also, Vertex is currently negotiating with FDA additional testing required. Posters for upcoming CF Conference show that all gating mutations tested had extremely positive results from VX770. I cannot imagine vx770 not immediately being used off label by another 6-10%, and the label expanded in a pretty short time period.