The Food and Drug Administration seems to be stepping up its evaluation of sites used in clinical trials. In December, Johnson & Johnson (NYSE: JNJ) failed to get its antibiotic ceftobiprole by the agency because of site-monitoring issues, and now Pfizer (NYSE: PFE) is the latest victim.

Last week, the company received a warning letter from the FDA for failing to ensure proper monitoring of a 2007 clinical trial that tested its antipsychotic Geodon in children. At least 13 children reportedly got too high of a dose.

Not getting approved for use in children doesn't seem like much of an issue. Geodon's patent protection is set to expire in 2012. Between the time it will take the FDA to finish reviewing the application and the time it will take Pfizer to ramp up marketing to child psychiatrists, there wouldn't be that much time to boost sales. No big deal, right? Unfortunately, no.

As trade publication Pharmawire points out, Pfizer wasn't necessarily running the study to increase sales by marketing the drug for use in kids. The main purpose of the trial was to keep sales for a little longer -- the drug brought in more than $1 billion worldwide in 2009 -- because the FDA grants a six-month exclusivity extension to drugs that have been tested in children.

Almost every blockbuster drug uses this strategy to gain the extension, because sales for the extended monopoly easily pay for the clinical trial. Bristol-Myers Squibb's (NYSE: BMY) Ablify and Eli Lilly's (NYSE: LLY) Zyprexa, also antipsychotics, both have pediatric extensions, as do big drugs in other indications like Pfizer's Lipitor and Novartis' (NYSE: NVS) Diovan.

The FDA doesn't have to approve the drug for use in children for the company to get the extension; the trial doesn't even have to come back positive. It only has to be completed in a "manner consistent with good scientific practices."

Oops, looks like Pfizer might have just lost its extension. If the drugmaker can't convince the FDA that the trial meets the requirements and Geodon faces generic competition six months earlier than Pfizer had hoped, that could cost the company about $400 million in revenue, based on 2009 sales.

Pfizer might have enough time to run another clinical trial to gain the pediatric extension, but it would be cutting it close. Considering its new reputation with the FDA for not monitoring properly, Pfizer certainly doesn't want to rush anything.