Today, as the FDA’s Oncologic Drug Advisory Committee (ODAC) adcomm debates and discusses whether to recommend approval of a Sandoz filgrastim biosimilar, (supported by the agency in their meeting materials), a few interesting items of note.
On pages 21-22 of the FDA briefing documents, an agency analysis found that, statistically, the commercial variety of EP2006 (Sandoz’ biosimilar) was lower in protein content than the comparator product (Neupogen) and was nonequivalent. FDA dismissed this as something that could be worked out with better manufacturing controls and asked Sandoz to correct it. Okay, but isn’t this exactly the kind of thing that can cause “poor responders” to filgrastim to have suboptimal responses? An important fact to consider when debating the value of differential nomenclature for biosimilars.
Here’s what the FDA had to say …
More concerning was that Sandoz was apparently confused in providing multiple lots of the biosimilar to the FDA for the protein testing. As a result, they thought they had provided 6 lots for testing that turned out to be only 4. Page 4 of the Addendum states:
“On November 25, 2014, Sandoz responded to the Agency request. In the response, Sandoz clarified that the actual number of EP2006 commercial drug product lots used in the statistical analysis referred to above is four lots instead of six lots. The six EP2006 commercial drug product lots initially considered were determined to be not independent because four of those six EP2006 commercial drug product lots were split-fill lots from two EP2006 bulk drug product batches, resulting in only four independent EP2006 commercial drug product lots.”
If Sandoz, a world-class company with a stellar record for cGMPs, in their highly reviewed and internally scrutinized licensing application confuses batches of their biosimilar, what can we expect in the real world after marketing? If Sandoz can’t track different lots of their experimental drug, what will the reality be in the real world? Another cry for sanity in the debate over nomenclature.
Biosimilars are here to stay -- and we need a nomenclature safety net.