Dendreon: "Riskless"?

When Dendreon received the FDA's Complete Response Letter in May 2007, there were two things that were said to be the reason for the non-approval of Provenge.

One was that they needed more clinical trial data; and the at-the-time ongoing IMPACT trial was going to fulfill that need.
The second was a Form 483 from the CMC section. Basically, something in the manufacturing part wasn't up to snuff.

In April we learned that the IMPACT trial succeeded, and wildly so. Unfortunately, the company was never very forthright on exactly what caused the Form 483. They said it was something that could be corrected; but as investors in Discovery Labs (DSCO) know, that is cold comfort: DSCO has now received THREE 483s and is still not approved.

Yesterday Deutche Bank upgraded Dendreon, saying their research indicated the 483 was for two main reasons: barcoding and documentation issues. If this is true, then these truly are easily fixed things and there would be NO remaining regulatory risks for Dendreon.

Regards,
Trond

Deutche:
We believe CMC issues will likely not delay FDA approval. There is some concern remaining CMC issues related to Provenge could delay approval. One investor concern is that there is risk to the Provenge manufacturing process, which could translate into a delay in FDA approval. This concern stems from the FDA’s 2007 complete response letter, in which additional information with respect to the chemistry manufacturing and controls (CMC) section of the BLA was requested.
Management believes it has addressed the issues brought up in the 483. While Dendreon expects it will have another inspection, the company does believe it has addressed the issues raised in the original 483 issued following a PAI (pre-approval inspection) in 2007. It has stated that it has kept the manufacturing facility in Morris Plains, NJ “PAI validated”, meaning that the company has been running it at lower levels so that it remains a validated class 100 space and meets a number of specifications. The company has also stated that it has had preliminary discussions with the agency on most of the 483 issues. Key issues described in the 483 related to bar-coding samples & documenting rules In our view, the issues raised in the 483 from the inspection that took place from February 12 to February 17 appear very mild. We believe the most important comment on the 483 was related to the request for more data on the handling of multiple samples in the same room as well as quality control steps used for maintaining the identity of the samples (i.e. bar codes).
The rest of the issues relate primarily to Standard Operating Procedure (SOP) documents that
basically describe various rules for certain procedures (the inspectors believed certain additions needed to be made in the documents). We believe that the issues raised in the 483 were likely easily addressed by the company, as in various conversations with management it has discussed using a bar-code system. A detailed list of issues raised in the 483 is available in the figure on the next page. In our opinion, there is likely little risk to manufacturing process for BLA approval
The manufacturing process and control systems for Provenge have been in place for many years and have been utilized during the clinical trial (over 1000 patient samples have been made in clinical trials). We believe the company has likely addressed any outstanding issues the FDA had in the original 483 because it has had time to do so and because the issues seem relatively easy to fix. Dendreon has had the advantage of having the FDA review its CMC section before and has had two years to implement any issues raised. What else gives us confidence? Many steps of the process are not unique to Provenge. The FDA has had experience managing patient-specific products such as diagnostic tests that have quality control measures in place that are similar to those for Provenge. The leukapheresis process is also well known by the FDA and many centers conduct the process on a daily basis for other procedures. Diagnostic testing frequently uses bar coding and quality assurance procedures to keep samples from getting mixed up. Finally, the antigen activation steps that take place at the NJ facility are not that complex, in our view. We believe the greatest challenge will be scaling up the process in time for launch and getting more physicians set up with infusion ability (to be discussed in a later section) Given patient awareness and lack of other therapies, we believe many patients will want Provenge. In addition to the ~30,000 new patients with metastatic disease every year, there are approximately ~70,000 existing metastatic patients who may want therapy. We believe the current plant will only provide for ~10,000 patients at peak and may only be able to supply ~2500 patients in the first 6-9 months of launch.