CMC Method Validation Memo - Addendum 2 -Ruconest

MEMORANDUM

From:       Wayne Hicks, Ph.D.; HFM-343; LBVB, DH, OBRR, CBER; (301) 451-6547, FAX (301) 402-2780.

Through:  Abdu Alayash, Ph.D.; HFM-343, LBVB, DH, OBRR, CBER; (301) 827-3813, FAX (301) 435-4034.
                   Basil Golding, MD.; OMPT/CBER/OBRR: (301) 496-4396

Subject:   Final review of methods/assay validation section for clinical studies5.3.1 of STN 125495 Pharming Group N.V. Biological License Application
                   
To:            Nannette Cagungun, RPM; HFM-380, RPMB, DBA, OBRR, CBER; (301) 451-3985
                   Elena Karnaukhova, Ph.D.: HFM-343, LBVB, DH, OBRR, CBER

Action recommended: The validation of methods used for clinical studies are adequate for the purpose of conducting the clinical studies.

Summary:  

Submission date: 4-15-2013
 CBER receipt date: 4-16-2013
 Type of submission: Original BLA
 Sponsor: Pharming Group N.V.
 Product: Recombinant human C1 esterase inhibitor, rhC1INH,
 Proprietary name: Ruconest
Indications: Treatment of acute attacks of hereditary angioedema (HAE) in adult and adolescent patients

Testing Center:
 -------------(b)(4)--------------------
 ------(b)(4)------
 ------(b)(4)------
 -------------(b)(4)---------------------

Module 5.3.1 of the original BLA STN 125495 contains reports of the method validation procedures for Ruconest for clinical trials of, Recombinant human C1 esterase inhibitor, rhC1INH. The validation process was carried out by ---------(b)(4)----------. Ruconest is a recombinant form of C1INH purified from the milk of rabbits raised in a closed colony.  The proposed indication of Ruconest is treatment for acute attacks of angioedema in patient with HAE.  Ruconest has been authorized for marketing in the European Union. The -------------(b)(4)---------------- of C1INH has been shown by the sponsor to be identical to human C1INH.

Validation of Analytical Procedures: (Section 5.3.1.4)

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

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

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

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

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

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

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

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

Redact 22 pages (b)(4)

1) Reviewer Question: In validation report VAL-R-03-061, there are two entries for the rhC1INH concentration of ---(b)(4)--- designated as VS4. The entry line for VS4 has a high overall CV% of (b)(4) that falls outside of the (b)(4) C.V. for the acceptance criteria. The provided explanation is that this sample is suspected of actually being a VS6 sample with a concentration of ---(b)(4)---.  Please provide data that support this explanation.

Pharming Response:  The results obtained at all concentration levels were comparable -----------------(b)(4)------------------- run for the VS4 level, which was significantly (b)(4) than the other runs in the cohort. As the results for the -----(b)(4)----- for the other VS levels were not much (b)(4) than the results for other runs within a cohort, an error was suspected. The results of the VS4 sample in ----------------------(b)(4)------------------------ were comparable to the VS6 results in the --------------------(b)(4)------------. Therefore, it was suspected that a VS6 vial was taken inadvertently for the VS4  --(b)(4)-- analyses. However, a formal investigation of these discordant results was not conducted and no further data are available to support this hypothesis.

Because the validation samples were analyzed in ------(b)(4)------, the analysis of results was completed without inclusion of the aberrant results for the VS4 level      -----(b)(4)-----. Please refer to the response to Question 2 for additional discussion of the VS4 level results.

2) Reviewer Question: In validation report VAL-R-03-061 there were (b)(4) measurements made for a concentration of ---(b)(4)--- of rHC1INH VS4, and VS4a. A bias was not calculated for sample VS4a. The precision method validation measurements for sample VS4 gave values that fell outside of the acceptance specifications. In the accuracy studies, however, VS4 appears to give acceptable values and VS4a is outside of these specifications. Please provide an explanation for the inconsistencies between these two methods.

Pharming Response:  The accuracy studies in VAL-R-03-061 Section 8.5.8 [Sequence 0000] contain an error in the designation for the VS4 sample, which should be VS4a. In Section 8.5.8. the value -----------------------------------------------(b)(4)-----------------------------------------------------------------------------, which corresponds to the complete ---(b)(4)--- VS4 cohort (see Section 8.5.4 and Section 8.5.5 for VS4 and VS4a, respectively). Pharming apologizes for the error and resulting misinterpretation of results.

As shown in the table in Section 8.5.8, from the ----(b)(4)---- C1INH result of VS4a and the ----------------------------------------
-------------------------------------------------------------------------------------------------(b)(4)--------------------------------------
----------------------------------------------------------------------------------------------------------. It was concluded that both the ----------(b)(4)---------- were acceptable for the ---------(b)(4)----------- of VS4a (i.e. without the results of ----(b)(4)----) and sufficient to conclude that the method when performed as written is capable of providing reliable results.

3) Reviewer Question: In validation report VAL-R-03-064 the results of the precision studies show an overall trend of -----------------------------(b)(4)-------------------------. The data indicates that the assay does not provide reliable results for the IgM anti-rhC1INH antibodies at -----------------------------(b)(4)----------
---------------------. The results in section 6.4.2 state that the ----------------------(b)(4)----------------- precision was acceptable at all levels which does not agree with the manufacturers stated acceptance standard of a C.V. (b)(4). Please provide an explanation as to why the acceptance criteria were amended to include these results.

Pharming Response: During the initial validation, acceptance criteria were only set for the overall precision. Because the standard deviation (SD) was found to be comparable at all levels, it was concluded that ---------------(b)(4)------------- would not improve the assay (------------------(b)(4)--------------------) and the variation --(b)(4)-- runs was comparable to the variation -(b)(4)- runs (i.e. it is not required to measure all samples from a subject in the -(b)(4)-) (VAL-R-03-064, Section 6.4.2 [Sequence 0000]). It was recognized that the precision of the IgM assay was outside the acceptance criteria, but the performance was considered adequate for the purpose of the assay at the early stage of clinical development.

Subsequently, an additional validation study was performed (VAL-R-03-137 [Sequence 0000]), to ensure that the assay would be accurate at the cut-off point of (b)(4) and at higher concentrations (for samples containing high levels of rhC1INH-specific IgM antibodies. The precisions for ---------(b)(4)--------- met the criteria (-----(b)(4)------, respectively).

4) Reviewer Question: In validation report VAL-R-03-065 the sponsor should provide a justification for the arbitrary (b)(4) cutoff level. The sponsor should clarify how the validation samples were prepared for the ---(b)(4)--- study, specifically what dilution(s) were used to generate the data for the positive controls for IgG, IgA, and IgM. 

Pharming Response: The cut points determined during the study were ---------------------------(b)(4)------------
--------------. These values were obtained with (b)(4) plasma from 30 healthy volunteers, which can be expected to be devoid of any antibodies to either plasma-derived C1INH or recombinant human C1INH. In contrast, many patients to be included in the clinical studies with Ruconest were likely to be exposed to plasma- derived C1INH products (Cinryze, Cetor, or Berinert) in the recent or more distant past and antibody responses to these products, often of low magnitude, have been reported (Craig, 2011). Additionally, the presence of IgM antibodies to endogenous C1INH were reported in HAE patients, even patients who were nave to C1INH products (Varga, 2011). Therefore it was expected that the cut point determined for plasma samples from healthy volunteers would result in an excess of positive samples with no clinical relevance. Hence, the cut points were set to slightly higher values, namely ----(b)(4)----
 ---------------(b)(4)---------------.

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

5)   Reviewer Question: In validation report VAL-R-03-065 the acceptance criteria stated by the sponsor for overall C.V for IgG is (b)(4). The experimentally based overall C.V. for the Q.C. low sample is (b)(4) which falls outside of the stated acceptance criterion. Please explain this discrepancy.

Pharming Response: In the Validation Report, the CV was reported out to (b)(4) decimal place. However the final result is determined against the validation acceptance criteria and reported with the same number of significant figures. Therefore, the CV for ----------------------(b)(4)---------------------- and was considered within the validation acceptance criteria.

6)   Reviewer Question: In validation report VAL-R-03-065 specificity study for the IgM antibody appears to be established at ---(b)(4)---. This was later adjusted      ------(b)(4)------ based on conversations between the testing center, -(b)(4)-, and the sponsor. Please provide an explanation for this change in the cutoff level.

Pharming Response: Please refer to the response to Question 4.

7) Reviewer Question: In validation report VAL-R-03-065 the studies for precision results show an overall trend of ------------------------(b)(4)------------------. The data indicates that the assay does not provide reliable results for the IgM anti-C1INH antibodies at ----------------------------(b)(4)------------------------. The results in section 6.4.2 state that the ----------------------(b)(4)----------------------- was acceptable at all levels which is outside the manufacturers stated acceptance standard of a C.V--(b)(4)--. Please provide an explanation for this exception to the stated acceptance criteria.

Pharming Response:  During the initial validation, acceptance criteria were only set for the overall precision. Because the standard deviation (SD) was found to be comparable at all levels, it was concluded that ---(b)(4)--- measurements would not improve the assay (i.e. --(b)(4)-- measurement is sufficient) and the variation -(b)(4)- runs was comparable to the variation -(b)(4)- runs (i.e. it is not required to measure all samples from a subject in the --(b)(4)--) (VAL-R-03-065, Section 6.4.2 [Sequence 0000]). It was recognized that the precision of the IgM assay was outside the acceptance criteria, but the performance was considered adequate for the purpose of the assay at the early stage of clinical development.

Subsequently, an additional validation study was performed (VAL-R-03-138 [Sequence 0000]), to ensure that the assay would be accurate at the cut-off point of (b)(4) and at higher concentrations (for samples containing high levels of C1INH-specific IgM antibodies.  The precisions for ----------(b)(4)----------- met the criteria      (-----(b)(4)----, respectively).

8)  Reviewer Question: In all of the (b)(4) based assays the following statement appears, During the validation no criteria did apply to the results of the validation samples. Please clarify this statement.

Pharming Response:  This statement is located in the protocol section and refers to the acceptance of validation batch results. This statement means that during validation studies, individual validation sample results that appear unexpected or discordant in relation to other validation sample results are not retested or otherwise invalidated. Consequently, all results obtained during the validation are included in the evaluation of the assay, providing an accurate representation of the overall performance of the assay.

9)   Reviewer Question: In validation report VAL-R-03-135, page 11 the sponsor states that all calibration, quality control and validation samples were analyzed in ---------(b)(4)---------- samples are insufficient.  Meaningful statistics such as C.V. and means cannot be calculated using ---(b)(4)---.

Pharming response:  For calibration, quality control and validation samples of each test(e.g. IgG), -----------------------------(b)(4)--------------------------------) were analyzed, each in --------------(b)(4)--------------------. The results from all aliquots were used for the overall statistics, including calculation of overall CV.  For the         ---(b)(4)--- determinations, the results for the ------------------------(b)(4)----------------------- is used for calculation. The (b)(4) results should be comparable to yield a valid result: the CV (i.e. the correct term for (b)(4) would be relative error) was calculated for this purpose only. A preset criterion of (b)(4) applied for the relative error of the  ---(b)(4)--- measurements.

10) Reviewer Question: In validation report VAL-R-03-137 please explain why anti-C1INH antibody containing plasma samples were used for this validation of an additional test for anti-rhC1INH IgM antibodies.

Pharming response: No human antibodies specific to rhC1INH were available. During method development it was demonstrated that IgM anti-C1INH antibodies extracted from an ----------------------------------(b)(4)--------------------------------------- (5.3.1.4, Bioanalysis - Section 1.2.1 [Sequence 0000]). Consequently, it was concluded that the available anti-C1INH antibodies extracted from these (b)(4) plasma samples can serve as positive control for the anti-rhC1INH antibody assay.

11) Reviewer Question: In validation report VAL-R-03-137 section 6.2 the reported-values for precision, C.V. were ------(b)(4)------ for the QC med and QC low samples respectively is confusing.  The chart below reports values for QCmed2 and QChigh2; none of the precision C.V.s in this table agree with the reported values in section 6.2.  Please provide an explanation for these contradictory compositions in the QCmed and QC low samples from section 6.2.

Pharming response: The QCmed and QClow samples are not the same as the QCmed2 and QClow2 samples, as was intended to be conveyed by the designation 2. A        ------------------(b)(4)------------------ were also prepared in order to compare the assay performance of this validation with the initial validation (VAL-R-03-064 [Sequence 0000]). The individual results of these QCs are presented in VAL-R-03-137 Attachment 9.1 (Sequence 0000); the summarized data are provided in VAL-R-03-137 Section 6.2 and Attachment 9.2 (Sequence 0000). This validation was conducted to assess the performance of the assay at two additional levels: --------------------(b)(4)--------------------. The results of these validation samples are presented in VAL-R-03-137 Section 6.3 (Sequence 0000) (individual results in Attachment 9.1, summarized results in VAL-R-03-137 Attachment 9.3 [Sequence 0000]). As the results in Section 6.2 and 6.3 concern different samples, the C.V.s mentioned in these two sections are not related.

12) Reviewer Question: It appears that the protocol for Val-R-03-137 and Val-R-03-138 are the same, although the results and data are different. Please clarify the differences among these two validation methods.

Pharming Response: These two validations are for the -------------(b)(4)--------------- versions of C1INH IgM assays. VAL-R-03-137 [Sequence 0000] concerns the validation of an anti-rhC1INH IgM assay and VAL-R-03-138 [Sequence 0000] concerns the validation of an anti-C1INH IgM assay. The Validation Protocols were essentially identical.  The difference between the two assays concerns the ------------------------------------------------(b)(4)--------
------------------.

13) Reviewer Question: In validation report VAL-R-03-174 please clarify whether this discussion refers to two different assays or one?  It appears that the ---------------------(b)(4)------------------------------- is used in a separate assay to validate the results of the assay in which ----------------------(b)(4)------------------------- were used.  However, if this is the case why were -------------------(b)(4)--------------------- used as the -------(b)(4)------- when --------(b)(4)-------- were used as the --------------(b)(4)-----------------?

Pharming Response:  In validation report VAL-R-03-174 [Sequence 0000] the sensitivity of the antibody assay to recombinant human C1 inhibitor (rhC1INH) is assessed, using the validated method for detection of IgA, IgG and IgM antibodies to rhC1INH (VAL-R-03-064 [Sequence 0000]). Hence, the discussion refers to one assay, in which rhC1INH is --------------------------------------------------------------------------------------------------------------------------------
----------------------------------------------------------------------------------------------(b)(4)-----------------------------
----------------------------------------------------------------------------------------------------------------------------------
-------------------------------------------------------------------------------------------------------------
-------------------------------------------------------------------------------------------
---------------------------------------------------------------------------------------------------------------------(b)(4)----------
-------------------------------------------------------------------------------------------------------------------------------------
---------------------------------------------------------------------------------------------------------------------------------------
-----------------------------------------------------------------------------------------------------------------------------------------
---------------------------------------------------------------------------------------------------------------------------------------
----------------------------------------------------------------------------------------------------------.

14) In validation report VAL-R-03-175 the data in -----(b)(4)---- assay results table in section 9.2.4 for the QChigh sample shows that for (b)(4) samples tested some data values are missing.  Please explain how these samples were treated in the final analysis of the testing results.  Please also explain why the results for the QC low samples are inconsistent with this same assay when applied to IgA, and IgM subtypes. 

Pharming Response: In validation report VAL-R-03-175 [Sequence 0000] it is indicated that ----------------------------------------------------------------------------------------------------------------------------------
---------------------------------------------------------------------------------------------------
---------------------------------------------------------------------------------------------------------(b)(4)----------------
-------------------------------------------------------------------------------------------------------------------------------
-------------------------------------------------------------------------------------------------------------------------------
--------------------------------------------------------------------------------------.

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

 (b)(4)

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

15) In validation report VAL-R-03-175 the sponsor states in section 6.4 that for all validation samples a % difference (b)(4) is obtained. The sponsor then states that a cutoff value (b)(4) is sufficient to establish that an individual sample contains rhC1INH specific antibodies.  Please clarify this statement.

Pharming Response: ---------------------------------------------------------------
-----------------------------------------------------(b)(4)---------------------------
--------------------------------------------------------------------------------------------------.

16) In validation report VAL-R-03-207 why were calibration, control and validation samples analyzed in ----(b)(4)---- and used to calculate the mean, standard deviation, C.V., and bias?

Pharming Response: --------------------------------------------------------------------------------------
------------------------------------------------------------------------------------------------------------------------------------------
------------------------------------------------------------------------------------------------------------------------
------------------------------------------------------------------------------------------------------------------------------------------
------------------------------------------------------------------------------------------------------------------(b)(4)----------------
-----------------------------------------------------------------------------------------------------------------------------------------
---------------------------------------------------------------------------------------------------------------------------------------------
---------------------------------------------------------------------------------------------------------------------------------------------
------------------------------------------------------------------------------------------------------
------------------------------------------------------------------------------------------------------
-----------------------------------------------------------------------------------------------------------------------------.

17)  In validation report VAL-R-03-207 why were calibration, control and validation samples analyzed in ---(b)(4)--- and used to calculate the mean, standard deviation, C.V., and bias?

Pharming Response:  ---------------------------------------------------------------------------------
------------------------------------------------------------------------------------------------------------------------------------
-------------(b)(4)------------------------------------------------------------------------------------------------------
----------------------------------------------------------------------------------------------------------------------------------.

18) Please explain why robustness was not validated for any of the (b)(4) based assays.

Pharming Response:  Robustness was included in method development, as part of the optimization of critical steps such as --------------------------------------------------------------------------------------------------------------------(b)(4)-------------
----------------------------------------------------------------------------------------------------------. Additionally, for the validations at least (b)(4) technicians were involved. Consequently, --------(b)(4)--------------- variation is covered in the validation.

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

Critical parameters that could affect robustness were varied during assay development and assay results are related to an internal control to ensure that minor variations in the execution of the assay do not affect the study results.

Addendum:

Section 5.3.1 that addresses the methods/assay validation studies section is lacking a neutralizing assay for the rhC1INH, and is also lacking a method for the assay of rhC1INH antibodies that has been validated using a standard rhC1INH antibody with proven specificity to the recombinant form of the C1 esterase inhibitor.  In both cases Pharming uses antibodies that have been validated for C1INH and have been demonstrated to cross react with rhC1INH. The antibodies used for validation are isolated from -------------------------------------------------------------(b)(4)----------------------------------------------------------. This was previously the subject of an information request that was submitted on Oct 23, 2013 regarding validation report VAL-R-03-137. This is our understanding of the most recent response (response received on December 16, 2013) to our information request on this matter that Pharming committed to the development and validation of an analytical method that can establish the presence ,or absence of neutralizing antibodies to rhC1INH.  Pharmings design of the suggested assay includes a ---------------------------------------------------------------------------------------------------------(b)(4)---------------------------------------------------------.  We have no objection to the inclusion of this assay as a means of assaying anti-C1INH antibodies that cross react with rhC1INH.  However, we would also like to see an assay developed with standard rhC1INH antibodies that were either raised using rhC1INH as an antigen, or that were obtained from patients that were treated with rhC1INH.
