CSL fails to take proper corrective actions.

A May 5-15, 2008, inspection revealed 14 violations at influenza vaccine manufacturer CSL Biotherapies.

Investigators Omotunde Osunsanmi and Joan Adamo faulted the Victoria, Australia, facility for CAPA-related deficiencies. They noted that no corrective action had been taken for an increase in the viable air monitoring results in the Egg Room, and out-of-specification investigations conducted into WFI endotoxin were deficient.

Further, inadequate corrective action was taken after a "breach in percentages of inoculum eggs rejects." A deviation report noted that from January to April 2008, the percentage of rejected eggs fell outside the alert limit. However, the firm failed to determine the possible cause of the increase in the rejection rate before removing the alert limit.

The investigations found no deviation report, documentation of investigation, immediate corrective action or CAPA into water system test point excursions from November 2007 to January 2008. Documentation on the OOS investigation stated "inappropriate set limits. for the water tested as the likely cause. However, the FDAers stated, "There is no documentation that other areas of the water system that could be causing the excursions were investigated," such as frequency of maintenance.

The FDA team found that documentation of deviation reports was inaccurate in some instances. Eight out of 10 out-of-limit viable air excursions occurred on different dates than those documented in the deviation report. Deviations that occurred on different days involving different batches were documented on the same report, they added. Incomplete investigations of the water system were documented under OOS investigations, and there was a lack of documentation of deviation reports for all water system excursions as required by the firm's procedures.

Osunsnanmi and Adamo also wrote that the company had "no procedure in place for reporting of Afluria Vaccine Biological Product Deviations to FDA/CBER."

Checklist: CSL

* Corrective actions not taken

* Inaccurate deviation reports

* Deficient cleaning processes

* Inadequate storage of materials

The inspection revealed that there was "no documented sterile filtration study that specifically defines the maximum sterile filtration time of the Influenza Vaccine monovalent batches," the 483 reported. The firm's research and development protocol and a validation report from 2002 "concluded that more data was required for the determination of sterile filtration time. However, no additional study has been performed." Further, the investigators found no recommendation in the validation report or in any other documentation for cleaning revalidation and/or monitoring of validated cleaning parameters; and there was no maximum sterile filtration time in an inactivated influenza vaccine batch processing record from May 2, 2007, nor in a procedure of the same date.

The FDAers also cited CSL for deficiencies in the preparation and testing of the prepared viral influenza vaccine used at the egg inoculation stage of the influenza vaccine manufacturing process. There was no procedure for sampling and testing the eggs, they wrote; CSL had no documented study for the expiration date assigned to the flu virus. There was no requirement in the manufacturing batch record that unused vials of the virus strain be discarded.

"The validation of the stability of the sucrose solution used in zonal centrifugation of the concentrated harvested allantoic fluids is inadequate," the investigators observed. They noted that two out of three validation runs failed to meet specifications. The investigation of the sucrose to meet acceptance criteria failed to include a review of the formulation of the solution to assure consistency among operators during formulation, the team stated.

In regard to qualification of the inoculation machines, the FDAers observed that there was no procedure in place for revalidation of the inoculation volumes since the machines were tested Dec. 21, 2007.

The inspectors also found the company's protocol for facility cleaning and environmental monitoring and the validation report for the influenza viral vaccine seed laboratory cleaning to be deficient. "The qualification failed to meet acceptance criteria that the cleaning and environmental monitoring of the IVV facility for a three-week period is consistently in compliance with environmental requirements under operating conditions," they wrote. After several validation runs "the firm was unable to achieve three consecutive cleaning validation runs that meet acceptance criteria."

The cleaning validation study dated Dec. 17, 2007, was also inadequate, the 483 reported. Two out of three runs for one type of filter and all three runs for a different filter failed the filter test acceptance criteria post cleaning and rinsing. "As such, there is no assurance that the cleaning agent residue can be successfully removed," the investigators stated. Although filter integrity failures had been noted, the report continued, "numbers of uses have not been assigned to multiple-use process product filters and manufacturing filters." The team noted a number of deviations, including failure of filters to pass integrity testing during processing of monovalent vaccine and lack of documentation that filters were replaced after integrity failures.

Storage of in-process materials was inadequate, the FDAers added. "Separation of Northern Hemisphere and Southern Hemisphere Monovalent Pooled Harvests (MPH) was inadequate to prevent potential product mix-up," they reported. The inspectors observed that the two MPHs were stored in "very close proximity with lack of clear differentiation between the two sets." Further, MPHs that were quarantined pending results from QA analysis were stored next to MPHs that had passed through the manufacturing process.

No documentation, such as cancellation of batch records, could be provided to the inspectors to prove that CSL did not process or manufacture 13 monovalent batches of one lot number and nine monovalent batches with different lot numbers. The issue was attributed to a lack of egg deliveries, the 483 reported. However, "no documented evidence was provided as to the reasons why the eggs were not delivered."

Growth promotion studies for purchased and in-house growth media were deficient because there was "no inclusion of environmental isolates in the growth promotions that are conducted, including growth promotion studies for aseptic media simulations," the 483 concluded.