I don't know if there is a good source of that information. But, we can guess. Gilead Sciences had a $86M IPO warchest in 1992. Four years later, Tamiflu was patented along with several other products. Lets just guess on the large side that Tamiflu development required half of thier money. So, $43M to get it patented. At that point Roche bought the rights. I don't know what Roche paid for FDA approval for that drug. But I have dealt with Roche for other products and Roche told me that they typically pay $7M to $10M for each approval. Probably since Tamiflu is a big drug, they paid more for more research. Until someone links real information, I'll just guess $20M. Note: the federal goverment picks up about half of the expenses, so lets add on another $20M. That gives us $43M + $20M + $20M = $83M as a very rough guess. If anyone has concrete numbers, I'll gladly update.
Now, there may have been other expenses. The goverment may have given Gilead grant money to help develop it. The government was also planning to spend $1B to stockpile it (does anyone here know how much was actually spent?). Not coincidently, a nameless goverment employee at the time received tens of millions of dollars in his stock share increase from this announcement. I think the "Tamiflu was expensive to develop" argument is a bit overused. They already got a ~15x return on the non-government expenses.
My bias in case you are wondering: I develop medical tests for a living and my wife works at a well-known pharmaceutical company.
I was about to post the same thing. You must be able to itemize, you must have massive expenses that aren't covered by insurance, and the insurance premiums aren't usually deductable in this category.