Medicine: Genentech (expensive) vs Genentech (cheap)?

Finally, anyone with an even mild interest in health management and policy should be following Genentech's new drug for macular degeneration. So Genentech dumped tons of money into making a new drug for that purpose, and a zillion dollars later, come up with Lucentis, a monoclonal antibody that will cost about 2000 bucks per dose and require monthly dosing. The only problem is that, while developing Lucentis, folks started wondering if one of Genentech's other drugs, Avastin, from which Lucentis is derived, might be helpful in its own right.
Avastin is a cancer drug which inhibits tumor angiogenesis. You can guess what happened next:

Before Lucentis was approved, as doctors started hearing the results of its clinical trials, many retina specialists began using Genentech’s cancer drug Avastin off-label as an eye treatment, at a cost of $20 to $100 a dose.

Both Avastin and Lucentis work in a similar way, and some eye doctors say they believe Avastin is equally as effective. However Avastin has not been tested in rigorous clinical trials as Lucentis has. So interest has been growing for a trial to compare the drugs.

This case demonstrates why having a prescription drug plan attached to Medicare might have some tremendous upside, in that the federal government now has a vested interest in controlling drug costs which will indubitably be in conflict with the interests of pharma and biotech firms. And that's something new, because the government doesn't really do ANYTHING to get in the way of pharma and biotech firms, besides making sure they don't directly poison the population. Incentives for innovation are good, but making profits with bad innovation (such as when you spend a zillion dollars developing a new drug for a new indication when one of your own damn proprietary products treats the disorder already) shouldn't be encouraged. Science depends on serendipity, and this time, serendipity bit Genentech in the ass. Don't feel bad, serendipity has made Genentech zillions of dollars in the past.

And just imagine, that Genentech wasn't interested in funding this research itself! It's wonderful to see NIH taking a role researching A) off-label usages of already approved drugs (generally left to the drug companies, who will only do so when having approval helps their profit margins, not when medicine needs further information), and B) trying to contain costs for a drug of prime interest to the Medicare population.