I was on the right side of the coin, but missed by a wide margin the magnitude that the Food and Drug Administration advisory committee would approve MannKind's (MNKD 0.72%) inhaled insulin Afrezza.

After the documents from the FDA came out on Friday, I suggested Afrezza might "squeak by with a positive recommendation from the panel." The vote was 13-1 in favor of approval for type 1 diabetes and 14-0 for type 2 diabetes.

That's far from a squeak.

Investors shouldn't forget that the FDA has the final say. It's hard to see the agency going against such a strong vote of confidence, but there is some precedence. The FDA rejected Merck's sugammadex -- back when it was owned by Schering-Plough -- after an advisory committee unanimously approved the drug. Over five years later, Merck still hasn't been able to get the drug approved.

At this point, there are three things that could keep MannKind from getting approved on April 15, the FDA's goal for making a decision.

1. It's only two weeks away
That may not be long enough time for the FDA to incorporate the committee members' comments into the label and work out post-marketing requirements with MannKind. The FDA has a couple of options. It could just ignore the goal since it's not an official deadline and take a couple of weeks to make its decision or it could issue a three-month extension.

Either way, it's a minor issue for MannKind. It was running low on cash, but has shares that it can sell in the open market if it hasn't done so already. Worse comes to worst, its billionaire namesake, CEO Al Mann, can cut the company a loan.

Source: MannKind.

2. There's still the conversion issue
MannKind made a change to the injected-inhaled insulin conversion between the clinical trials and the recommendation for the label. The conversion is important for type 1 diabetics switching from Eli Lilly's (LLY -1.81%) Humalog or Novo Nordisk's (NVO -1.93%) Novolog to Afrezza. The change obviously wasn't a big enough deal to affect committee members' votes -- or the difference between pharmacokinetic and pharmacodynamics went over their heads -- but the FDA was clear in its briefing documents and during its presentation to the committee that the new conversion was unacceptable.

It seems to me that MannKind is going to have to change back to the conversion used in the clinical trials or run a pharmacodynamics study to get the new conversion approved. In theory this shouldn't affect type 2 diabetics that are new to insulin, so the FDA could approve Afrezza for type 2 diabetics while MannKind ran a pharmacodynamics study on type 1 diabetics.

3. Risk-benefit analysis
There were no major safety issues revealed by the FDA, which was the biggest risk going into the advisory committee meeting since investors had only seen top-line data. But the FDA highlighted multiple times that the effects of Afrezza were "modest" compared to other diabetes treatments. Drugs with modest effects can still get approved, but they need squeaky-clean safety profiles.

For Afrezza, the concern is with putting a drug in the lungs that doesn't treat lung disease. The FDA is worried about lung cancer and potentially decreasing lung function over time. Both are hard to definitively rule out in a clinical trial designed to get a drug approved.

While all except one committee member ultimately believed the benefits outweighed the risk, many still expressed concerns about safety in their comments during the committee meeting. The FDA considers the comments as much, if not more, than it does the final vote. If MannKind had squeaked by with a positive vote, I'd argue the FDA might go against the committee like we saw with Orexigen's obesity drug Contrave. The committee voted 11-8 recommending approval, but the FDA rejected Contrave anyway.

With a near-unanimous recommendation for approval, it seems less likely -- though not impossible -- the agency will go against the committee's risk-benefit evaluation. More likely, the FDA will ask MannKind to run post-marketing studies to look at lung safety.

So will it get approved?
I think an approval on or before April 15 is unlikely because it takes time to cross the t's and dot the i's. The conversion issue has to be worked out. It's a fixable problem, but might result in a delay, at least for type 1 diabetes.

The big unknown is how the FDA will treat the lung safety issues. I certainly hope the FDA approves Afrezza. What's the point in holding advisory committee meetings if the agency is just going to ignore its outside advisors? But the FDA marches to the beat of its own drum and investors should factor in the potential for a surprise rejection especially given the FDA's tempered view of the efficacy data.

What do you think? Vote in the poll below and elaborate in the comments section.