
Statistical Review, December 18, 2009 - Flubok

 

Date                December 18, 2009

 

To                    Katherine Matrakas

                        Rakesh Pandey

                        Timothy Fritz                                                

                                               

From               Lev A. Sirota                         

                        Vaccines Evaluation Branch

 

Through          A. Dale Horne

                        Branch Chief, VEB, Division of Biostatistics

 

Subject           BLA 125285/0 (Protein Sciences Corporation)

                        BLA 125285/0.8

FluBlok Influenza Vaccine

 

 

c.c.                  Chronological File

                        A. Dale Horne

Henry Hsu

                        Christopher Egelebo

 

Executive Summary Statement and Conclusion

 

I did not find bioassay-related statistical issues in STN 125285 that exceed the level when submission may not be approved by the agency based on bioassay performance. 

 

Background 

 

This review covers statistical issues related to the bioassay in Module 3 Volume 1.9 of the original submission. It is also concentrated on answers presented by the applicant to CBERs questions presented in BLA 125285/0.8.

My comments are in italics.

 

 

Results and Comments to CBER

 

In assay validation part of Module 3 Volume 1.9 the applicant used sets of validation parameters in correspondence with ICH Q2A and Q2B guidelines. Those guidelines were recently aggregated into ICH Q2(R1), which inherited definitions of validation parameters from the earlier version.

 

The definitions of the assay validation parameters used by the applicant are in general in agreement with the above mentioned guidelines.

 

In Vitro assay for Adventitious Viral Contaminants and Detection of Adventitious Virus

 

----b(4)--------------------------------------------------------- -----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

 

Single Radial Immunodiffusion Assay 

 

Single Radial Immunodiffusion Assay (SRID) measures concentration of the hemagglutinin (HA) antigen. The SRID assay is used to determine potency of monovalent bulk drug substance and to confirm the potency of the formulated vaccine drug product. The applicant conducted validation experiments for specificity, linearity, precision and range. The applicant considers not feasible to obtain a theoretical SRID value from other product characterization assays, therefore they did not include parameter of accuracy in the validation exercise.

 

--b(4)-- ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

 

--b(4)--- ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

 

--b(4)---- ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

 

 

--b(4)-- --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

 

--b(4)-- ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

 

--b(4)--- ----------------------------------------------------------------------------------

 

The range by the ICH definition implies linearity accuracy and precision within specified bounds. The applicant defined range for linearity only, which may or may not be acceptable, given intended use.

 

I have checked selected calculations of standard statistics provided by the sponsor and did not find deviations that could not be explained by rounding.

 

Influenza Virus Hemagglutination Inhibition Assay

 

The applicant used standard set of validation parameters: specificity, precision (repeatability and intermediate precision), robustness, range, and linearity. The applicant concluded that all parameters tested meet the predefined acceptance criteria.

 

--b(4)-- ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

 

--b(4)-- -----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

 

This type of criteria does not offer substantial control over precision. The %RSD used in most bioassays is in general preferable and offers more power to distinguish good from poor precision. It should be noted though, that for HAI assays the criteria based on individual deviations being within 1 titer steps are pretty common. It can be explained by semi-qualitative nature of the HAI assays that may lead to quite high thresholds in terms of %RSD.

 

--b(4)-- ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

 

--b(4)-- --------------------------------------------------------------------------------------------------------------------------------------------------------------------

 

--b(4)-- --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

 

I complemented sponsors analysis with residual analysis, calculations of coefficients of determination and confidence intervals for slopes. My analysis does not reveal curvature. Coefficients of determination were ranged from b(4)-----------------. Point estimates for slopes were b(4)-- --------------------------------------------------------------------------------------------------------------------------------------------------------------------

 

The following CBERs questions were addressed in the BLA 125285/0.8 Amendment:

 

In the Validation Final Report Protocol 95026.00 for the In Vitro Assay for Adventitious Viral Contamination ----------(b)(4)----------, the first table on page 6 of 30 contains a column "%Confidence." What is the definition of the %Confidence in this table and how it was calculated?

 

-----------------------------------------------------------------------------------------------------------------------------------------------------------------(b)(4)----------------------------------------------------------------------------------------------------------------------------------------

 

----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------(b)(4)-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

 

----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------(b)(4)----------------------------------------------------------------------------------

1 Page Determined to be Not Releasable: (b)(4)