The 10-5 vote by a Food and Drug Administration advisory committee recommending approval of Johnson & Johnson's (JNJ -1.39%) diabetes drug canagliflozin sounds pretty decisive. But I wouldn't add the future revenue from the drug to your valuation just yet.

Canagliflozin, which will go by Invokana if it makes it to market, is a selective sodium glucose co-transporter 2 (SGLT-2) inhibitor. It's in the same class of drug as Bristol-Myers Squibb (BMY -7.86%) and AstraZeneca's (AZN 5.45%) dapagliflozin, which the FDA rejected last year over cancer concerns.

Invokana's problem has to do with a potential for increased cardiovascular risk. The advisory panel was asked about that issue specifically and only voted 8-7 that the drug isn't associated with an unacceptable cardiovascular risk.

After heart issues with GlaxoSmithKline's (GSK 0.05%) Avandia were discovered, the FDA required all drugmakers to run a cardiovascular study on diabetes drugs. If initial reports come out clean, the large study can be finished after the drug is on the market. But if there's a slight signal for cardiovascular risk, the company has to wait for the full data before approval.

With a pretty close vote on the cardiovascular risk, don't be shocked if the FDA errs on the side of caution, waiting until Johnson & Johnson produces the full data from its cardiovascular study due out in a few months.

Personally, I think the issue is probably nothing. The observed increase occurs in the first 30 days of taking the drug, which sounds more like a fluke observation than a serious problem. But there's no harm to the FDA to wait. It's not like this is a neglected area; there are plenty of drugs for diabetics to take until Johnson & Johnson can prove its drug is safe.

For Johnson & Johnson, the delay could cost the company billions since the patent clock keeps running through FDA reviews. Bristol-Myers and AstraZeneca, as well as Eli Lilly (LLY -1.63%), which has a SGLT2 inhibitor that just posted positive phase 3 data, certainly wouldn't mind a delay, though.