A very small, but incredibly expensive mistake…

In early August, the FDA sent Adventrx Pharmaceuticals a Complete Response Letter (CRL) for their drug Exelbine which basically said “Hey, thanks for the NDA, but based on what you sent us, we’re not going to approve your drug.”  Now CRLs can take many forms, all the way from “we have some concerns about your manufacturing controls”, which can often be resolved in a matter of months, to “your trial design is so mess up that you’ll have to repeat the whole thing”, which is often a death knell for biotech start-ups who don’t have a lot of cash.

Whenever I read about a company receiving a CRL, I’m always interested learn why their application was rejected.  CRLs are not publically released (this is a bit of a hot topic lately, since many think they should be), but companies always issue press releases when they receive them since the information they contain is obviously very material to the value of the company.  When I read Adventrx Pharmaceuticals’ press release, I cringed:

The FDA determined that it could not approve the Exelbine NDA in its present form.  In particular, the complete response letter noted that, based on inspections at clinical sites, the authenticity of the drug products used in the pivotal bioequivalence trial (Study 530-01) could not be verified, which placed the results of the trial into question.  The letter stated that the bioequivalence trial will need to be repeated to address this deficiency.

Ouch!  Basically what the letter is stating is that the data from the clinical trial in question (Study 530-01) is being thrown out because the FDA doesn’t believe that the trial sponsors could tell the drugs they were using apart.  After doing some searching on clinicaltrials.gov, I found the trial here.  This was a bioequivalence study where the company’s product, Exelbine, was being compared to a reference product, Navelbine, to make sure that they both behaved the same way in patients.  The reason this trial was run is that if bioequivalence could be proven, Adventrx Pharmaceuticals could use Navelbine data in support of its application.

The reason why this is so painful to see is that it can be so easily avoided.  The FDA has very strict guidelines concerning what type of testing is required to prove what you say you are giving patients is actually what you are giving patients.  It goes further than just making sure you ordered the right drug; trial sponsors are usually required to test a cetain percentage of the packages they receive (be it vials or bottles) to ensure that what is written on the label is actually what is in the bottle.  And following that, once the drugs are taken from their packaging and prepared for administration to patients, they need to be carefully tracked so that no mix-ups occur.  It appears that the clinics running this trial failed to do that.

What makes it even more painful, is that it’s likely that nothing is wrong with the clinical trial.  In the press release, the company states:

Of note, Exelbine and the reference product come in different package presentations, require different preparation procedures and have different physical characteristics.  Based on the different characteristics between the study drugs, the Company believes it is unlikely that study sites would confuse the two study drugs or fail to recognize which drug was being administered to a patient.

The company’s arguement is pretty weak (“hey, we didn’t follow procedure, but come on!  only an idiot would mix those two drugs up”), but they are most likely right.  However, it doesn’t matter (and it shouldn’t matter).  If you don’t follow FDA guidelines for trial design you can be guarenteed you won’t be getting approval.

The only solace I can provide is that it was relatively a small phase I trial.  With an estimated 28 patients enrolled, it would likely cost the company only a few hundred thousand dollars and a year or so of their time to repeat it.  However, it appears they may not bother:

“In the meantime, our resources and focus are on ANX-188 and ANX-514, which we believe are the long-term value drivers for our company.  Our cash and equivalents of $40.7 million at July 31, plus cost savings from delaying or potentially discontinuing the Exelbine program, will provide us the capital to continue to advance both of these programs,” Mr. Culley added.

Lesson learned I guess…