Another side note is that he was also impeding the FDA approval process.
In order to get your generic approved you have to prove that it's biologically equivalent to what's on the market. Shkreli's distributor would only sell directly to end-users and would refuse to sell it to anyone doing equivalence studies on it. Thus it was effectively impossible for anyone else to prove bioequivalence and bring a competing product to market.
Also, any other company spending money to spin up production and get FDA approval would face its incumbent competitor suddenly dropping its price to cost. This lasts until the new entrant goes bankrupt or stops selling, and then the price jumps right back up to gouging territory.
The only way to eat the incumbent's lunch in a worthwhile way would be to start out larger, so you can absorb the startup costs, and find a way to produce at lower marginal cost.
The FDA approval process for generics is part of the problem there. It should be enough to do a chemical analysis showing that the molecules are the same. The current FDA requirements are akin to having to prove that your new vodka can get someone drunk just as well as vodka from other vodka makers. The active ingredient is ethanol. Ethanol made by you and ethanol made by AB InBev is chemically indistinguishable. As long as you can accurately report concentration, for the purposes of dosing, you shouldn't have to prove that ethanol still gets people drunk when you are the one making it. You should only need to prove that the amount of ethanol in your vodka matches the amount printed on the label.
So if your doctor prescribes 9.3 g ethanol, taken orally, once daily, your "compounding pharmacist" can give you a 1 oz. shot of any 80-proof beverage. It doesn't matter if it's name-brand top shelf or generic rail, so long as it's 40% alcohol by volume.
I'm going to have to disagree that you shouldn't only need to show chemical equivalence. You need to show biological equivalence.
A few years back the FDA approved a generic version of Wellbutrin. They didn't ask for bioequivalence testing and as a result, the formulation dumped all the drug into the body, drug levels spiked and patients got sick. The FDA eventually made the company reformulate the drug so that it was truly interchangeable with the branded drug.
There's more to the buproprion HCl fiasco. The active ingredient produced by the generics manufacturers was the same. The delivery mechanism was different. The approval for the 300 mg XL version of the pills was extrapolated from 150 mg XL pill data, rather than compared with the Wellbutrin XL 300 mg. There wasn't even a check for chemical equivalence there.
You could have dropped the pills into an artificial stomach and measured concentrations at 20 minute intervals. That tests chemical equivalence. You don't need to test bioequivalence until after that. If your synthetic stomach setup does not show chemical equivalence, that's when you need to get the animal test subjects.
Part of the problem here is that the FDA bundles several different tests into one.
- Does the molecule do harm to a human body? What are the side effects?
- Is the drug effective for a specific medical condition?
- How does a higher or lower dosage affect the observed results?
- How effective is a given delivery mechanism?
Those questions should really be answered separately, and in that order, for initial approval. Generics would then only need to re-answer question 4 if the delivery mechanism was different.
For limited application drugs (so annual sales smaller than Prozac... which is most of them), this effectively means there will only be one supplier.
I don't know a way to fix this that will satisfy most interested parties.