• Decrease font size
  • Return font size to normal
  • Increase font size
U.S. Department of Health and Human Services

Vaccines, Blood & Biologics

  • Print
  • Share
  • E-mail

Section Contents Menu


Information Request Memo - Flublok


Food and Drug Administration
Center for Biologics Evaluation and Research
Office of Compliance and Biologics Quality
Division of Manufacturing and Product Quality
To:                  File STN 125285/0/41
From:             Deborah Trout, Committee Member, OCBQ/DMPQ, HFM-675
Through:        Carolyn Renshaw, Branch Chief, OCBQ/DMPQ, HFM-675
Subject:          Review of responses to comments 1, 3, 7, and 8 of the November 12, 2010 information request submitted March 9, 2011 by Protein Sciences Corporation (hereafter referred to as PSC).
Action Due Date: There is no action due date at this time. The application review is currently under a complete response letter.
Recommendation: request additional information to resolve issues identified below.
Review and Comments
1. Regarding Item 1a, rHA (B) downstream process validation:
  • -----(b)(4)------;
  • -----(b)(4)-----;
  • ---(b)(4)---;
  • (b)(4);
  • ---(b)(4)---
  • ---------(b)(4)-------
3. Regarding Item 1c, fill validation:
3a. Please provide your final shipping report for bulk drug substance. This should include the CoA for each bulk shipped, potency results before and after shipment, and associated investigations/deviation reports.
The final bulk drug substance shipping report R-09-073 was included in Attachment 1of their response. Three shipments were performed at temperatures of -(b)(4)-. Parameters measured and acceptance criteria included: -------------------------------------------------(b)(4)------------------------------------, and potency by SRID pre and post shipment. PSC states all parameters in the study were met: integrity of the (b)(4) was maintained, temperatures were maintained between -(b)(4)-, and potency results pre and post shipment appear comparable. The batches included in this study were produced for use in validation studies and were not intended for human use therefore certain tests not relevant for the purpose of this study (such as testing for -------------------(b)(4)-----------------) were not included in the Certificate of Analysis.
Certificate of Analysis for the batches that were shipped were included in Attachments 2, 6, and 7 of the amendment. PSC claims were no investigations or deviations associated with the shipping validation.
Although PSC claims there were no deviations associated with the shipping validation there appeared to be a temperature excursion according to the Temptale data for trip number 49214689. PSC should clarify why this temperature excursion was not considered a deviation.
3c. No data are provided to support consistency in fill. Please provide analysis of vial-to-vial variability of lot release specifications for a statistically relevant number of vials, including the rationale for number and location of vials selected for testing.
Consistency in fill was evaluated in the three process validation runs PV1 -----(b)(4)-----, PV2    -------(b)(4)------ PV3 ---(b)(4)--- performed October 2008, February 2010, and July 2010, respectively. One vial was taken from the beginning, middle and end of the fill line during the filling of three runs of FluBlok. PSC indicates that given the acceptable results from the mixing study of the blended formulated bulk it is expected that the beginning, middle, and end of the fill are representative of the entire fill. The subset of attributes that were tested ((b)(4), identity, and potency) were chosen as attributes that are indicative of proper mixing and fill uniformity. Fill uniformity is supported by the consistency in results from these three tests. PSC did not perform all release tests as it was determined that these three tests were sufficient to demonstrate uniformity of the fill.
PSC states that there was no statistical relevance applied to the number of samples taken, as sampling beginning, middle, and end of a lot is a common industry practice, particularly when no differences across a homogeneous mixed lot are expected.
PSC’s response is unacceptable. Sampling plans should be based on statistical criteria as per 21 CFR 211.84(b). PSC should provide their corrective action plan to address the lack of statistically valid data to support uniformity of fill.
3g. Please submit complete information (including 3rd run) to validate the maximum hold time for your ----(b)(4)----.
Hold time for three batches of Influenza Vaccine final ---(b)(4)--- for FluBlok was evaluated in the three process validation runs PV1 (---(b)(4)---), PV2 (-----(b)(4)-----) and PV3 (-----(b)(4)-------) performed October 2008, February 2010, and July 2010, respectively. PSC claims that for runs 1, 2, and 3 the -------------------------(b)(4)-----------------------------, potency, and protein identity testing were reviewed and met all acceptance criteria to validate a hold period of -----(b)(4). PSC states that despite some invalid SRID preparations at (b)(4) the valid data indicate that (b)(4) is an acceptable hold time. PSC discussion with Hospira claim that the invalid results are assay (SRID) related and not process related. Because of the post sampling time limits for testing specified in the validation protocol, retesting by SRID was not possible.
PSC claims that H3 rHA invalid SRID results in fill --(b)(4)-- were due to high relative potency, and all were clustered on a set of gels. Rings from the reference prep were noticeably smaller than for the others. Excess vacuum was identified as a problem for H3 A/Victoria/210/2009 X-187 (lot H3-Ag-1003) reference vials, which could result in loss of the reference antigen when the vial is opened for reconstitution. Hence, the root cause for the smaller rings in this set is loss of reference antigen during reconstitution. Instructions for reconstitution without disturbing the lyophilized reference antigen will be clarified in the SOP.
H1 rHA invalid SRID results in fill --(b)(4)-- were due to failing the acceptance criterion for the correlation coefficient R of the regression of the reference standard (ring size vs. dilution). The cause was small rings for some reference antigen dilutions. PSC attribute this to the poorly matched reagents for H1 rHA A/California/07/2009. CBER is working with PSC to qualify SRID reagents for commercial release of H1 rHA derived from A/California/07/2009, as described in their response to Comment 8f in the amendment.
PSC’s response is incomplete. Please provide the investigations associated with all invalid SRID results and their plans for collecting additional data to support that potency is not deteriorating during the (b)(4) hold time.
3h. Please submit all cleaning results and a complete list of results from release testing for PV#3. Provided your alternative potency test is stability-indicating, alternate potency test reagents may be used to validate the fill but this product cannot be used for distribution until testing with CBER approved reagents demonstrate potency specifications are met.
Cleaning validation was completed for the three process validation runs documented in the results tables below for PV1 (----(b)(4)----), PV2 (----(b)(4)----) and PV3 (----(b)(4)----) performed October 2008, February 2010, and July 2010, respectively. Additional details of the runs were included in response to Comment 3d of this amendment. The cleaning validation results of the filling equipment at Hospira for FluBlok formula number ---(b)(4)--- have been reviewed and found to be acceptable. The results verify that the Standard Operating Procedure for cleaning of the appropriate line and the equipment preparation batch record are acceptable for cleaning equipment after the manufacturing of FluBlok.  Validation was performed according to Hospira protocols PQP0183.06 and PQP0183.06M6. The (b)(4) filling equipment at Hospira was cleaned according to specific line SOPs and the equipment preparation and (b)(4) batch record for FluBlok. Following the cleaning process, sites were sampled with a unique swab as specified in PQP0183.06 and PQP0183.06M6. An additional swab served as a negative control blank, and a WFI sample was pulled as a negative control. These samples were delivered to the Hospira chemistry laboratory for analysis. The swab samples were tested according to Hospira SOP PD-135-RPT-008.07 and evaluated against the pre-determined specification.
The cleaning data was reviewed and appears acceptable.
7b. Unless loss of potency of H3 produced in 2009 and 2010 is attributed to specific manufacturing conditions or methods, steps need to be taken to identify conditions to improve stability of drug substance and drug product. You state that you are indeed committed to a formulation program to improve stability. Please describe the work that you have done to date and provide your plans to implement any formulation changes for either drug substance or drug product.
Once formulation improvements are identified that result in increased shelf life, PSC will use a risk-based approach to determine a path forward to implement the change. Factors to be considered include:
  • Safety profile of excipient
  • Effect of excipient on immunogenicity of the product
  • Impact on the process and process validation
  • Potential effect of excipient on validated test methods
  • Potential effect of excipient on extractable and leachable compounds from rHA monovalent bulk drug substance and FluBlok drug product packaging and process contact materials
Risk assessment and discussions with FDA will determine whether additional pre-clinical or clinical studies are required.
PSC will also need to reassess cleaning validation of product contact equipment after the addition of ---(b)(4)---.
7c. You state that H3 batch --(b)(4)-- was originally included in the 2009 stability study but had to be replaced with another lot due to a leak in the sample storage (b)(4). Please provide the investigation report describing this incident.
PSC states that the deviation was judged not to have an impact on the 2009 stability study because the lost batch (---(b)(4)---) was replaced by batch --(b)(4)--, which had similar 3-month relative potency to that of ---(b)(4)---. The remainder of the study will be completed with batch   ---(b)(4)---. The report of the investigation of this event (Deviation #10-153) was attached as Attachment 1 in the amendment. Investigation of the event indicated that the inlet port of the storage (b)(4) had been loose, allowing leakage of sample. A corrective action was implemented and amended to the SOP QC0065 to assure that in the future, all storage (b)(4) are stored in a secure fashion and that the security of each (b)(4) is verified by a second observer.
A deviation was not opened at the time of the leaked (b)(4) and replacement. SOP QC0065 was not approved or activated at the time of the deviation, and no other deviation to the analytical method or other SOP occurred. The leaked (b)(4) and replacement was a deviation from the protocol describing the study (P-09-010), which states that deviations from the protocol will be described in the report. Under the systems now in place, a deviation would be opened for a leaked and replaced (b)(4) or any other divergence from SOP QC0065. A deviation was opened after the fact to assess the impact of the deviation and to specify and document preventative actions.
PSC explanation for why an investigation was not initiated for over one year is unacceptable. It appears as if an investigation was initiated only after FDA requested one. PSC should explain why their stability protocol/study would allow such a practice and why would their Quality Unit approve a protocol with this provision considering they were cited on the FDA Form 483 in October 2009 for just such a practice. In addition, PSC should provide evidence that they have implemented all the corrective actions noted in their response date November 13, 2009 to FDA Form 483 item 2.
Process Validation Addendum
PSC has completed three process validation batches for trivalent influenza vaccine. PSC claims that the third process validation run (PV3) was invalid due to a number of invalid potency results unrelated to the capability of the process to consistently produce quality product. Root causes for these invalid results have been identified and corrective actions identified. To enable production of batches of up to --(b)(4)-- vials per lot, PSC proposes to prepare one conformance batch at a larger (e.g., (b)(4)) scale post-licensure. The (b)(4) conformance batch will use the process understanding and knowledge gained through the three (b)(4) process validation runs. The critical process parameters and critical quality attributes for the (b)(4) scale batch as were established at the (b)(4) scale are not expected to be different.
The formulation mixing process will be performed in a ------------(b)(4)----------- instead of a       ---(b)(4)---. The mixing in the (b)(4) will use an ---------(b)(4)-----------. The (b)(4) will be           -----------------------------------(b)(4)---------------------------------. Additional cleaning validation studies will be performed on this (b)(4). The use of a -------------------(b)(4)-------------------- is not expected to impact critical process parameters or quality attributes.
No changes are planned for the filling process and no impact is expected on the filling process due to the scale up. The process will use the same filling equipment, filling line, and filling room. However, PSC will evaluate the impact of additional --------------------------------(b)(4)--------------------------------------. Additionally PSC will perform ------------------------------------------------------------------------------------------(b)(4)---------------------------------------------------------------------------. Release specifications for Drug Product will remain the same for the (b)(4) as for the ------(b)(4)------.
In addition to using the (b)(4) batch as a replacement conformance batch as part of the FluBlok process validation, PSC will also perform continuous process verification to assure that the process remains in a state of control. The continuous process verification will include monitoring of critical process parameters (-------(b)(4)----------), testing each batch for critical quality attributes, and trending of critical quality attributes.
PSC’s process validation addendum is unacceptable. PSC will have to perform 3 runs at the (b)(4) scale. In addition we do not accept the proposal to substitute one (b)(4) scale up run for the failed process validation post licensure. PSC will need to resolve the issues associated with their potency assay and perform an additional run to support licensure.