CDER published a study titled “First-Cycle Complete Response Rates for CDER NDA and BLA Applications” (here). I read and reread it a few times and I am still confused. It appears that FDA published this report to counter industry’s stated belief that the number of Complete Response Letters (CR) have increased over time. This is not a new belief and has been repeated over the past few years and now the FDA’s report contradicts that position – well, at least they say it does.
Their definitions of what a CR is and how they are issued and when they are deemed to be after the first review cycle is a bit hard for me to follow, and it seems like smoke and mirrors to me. My view has always been more simplistic; for instance, when an NDA or BLA is filed, if the first action is a CR, that should mean that the CR was issued in the first review cycle, like it is for an Abbreviated New Drug Application when the initial review is completed. That seems simple and meaningful and very straightforward. But, as expressed in this report, CDER has other views regarding the CR actions that may occur over the course of the NDA or BLA application review cycle and states: “Replicating estimates in some of these reports can be challenging due to insufficient documentation of data sources, analytical units, inclusion/exclusion criteria, and methodological approaches. Investigating and analyzing first-cycle CR and approval rates require some familiarity with nuances in FDA’s regulatory data and reports as well as the business rules applied in various publicly available reports”. After their torturous explanation of these “nuances”, it is still not obvious to me how CDER’s explanation defines how the metric of a first cycle CR or, for that matter, how a first cycle application approval actually is reported.
Industry has had some major concerns regarding the seeming increase in the issuance of first cycle CR letters over the past few years, and, for me, this report still has me wondering exactly what the metric represents given the opaque explanation provided in the report. Maybe it is just me or maybe the algorithm described is too obtuse for me to comprehend. Let me know what you think at r.pollock@lachmanconsultants.com.
