Bioport struggles with anthrax vaccine issues.

Bioport, Lansing, MI, Team Biologics The agency's inspection of Bioport of Lansing, MI, in October 2000 did not go well for a firm embroiled in seemingly unending controversy, primarily over its Defense Department contract to manufacture anthrax vaccines, which led to a congressional investigation.

Last August, Bioport implemented a voluntary recall of the anthrax vaccine due to an expiration date mixup and, in the following month, a voluntary recall of human albumin because the firm could not assure the sterility of the substance.

FDA investigators Paula Trost, Marsha Major and William Tingley arrived in October and filled 17 pages on a 483. The EIR has not yet been released. Despite the numerous adverse findings, Bioport appears to have escaped receiving a warning letter.

Renovation efforts go awry

In 1998, the Pentagon helped fund a renovation of the facility while authorities with the state of Michigan sought a buyer. However, the renovation took five months longer than expected. The delay ran parallel to a serious cash flow problem for the operation and in August 1999, the Pentagon agreed to roughly double the price it would pay Bioport for the vaccine.

Spokeswoman Kim Root stated that the renegotiation of the price reflected in part the fact that the state of Michigan had subsidized the production of the vaccine when the facility was under the control of the Michigan Biologics Products Institute and that the renegotiated price more accurately reflected the cost of production (See October 1998 issue, Page 12).

The latest 483 cited Bioport for problems with the design and construction of three filling rooms that, along with environmental monitoring, cleaning and employee practices, did not assure the sterility of products filled in the suites.

The filling suites were not part of the firm's renovation. Root stated Bioport is currently contracting out the filling work and that the firm had not decided whether to continue contracting the work or reestablish filling capability in Lansing.

The FDAers wrote that seams around a hot air oven in one of the rooms were not adequately sealed as demonstrated by leakage of smoke to an adjacent room. The 483 also noted that the floors in one of the filling rooms were made of a material that was not easily cleaned and that the "walls have seams."

The investigators cited the fact that an in-line HEPA air filter for a hot air oven was located in a Class 10,000 area in one fill room as a deviation, and stated that when the hot air oven was in operation, the same room went out of specification for temperature and humidity."

Cleaning validation hit

FDA zapped the firm's cleaning practices for this filling room, noting that Bioport personnel used only one solution for cleaning curtains, walls, the floor and the doors on ovens and autoclaves. However, the 483 noted that Bioport could come up with no documentation that the solution "was effective in removing microorganisms from surfaces." The inspectors also hit Bioport for the elapsed time between cleaning and filling operations, but the number of elapsed days was purged.

The 483 itemized a list of three lots of anthrax vaccine, one of rabies vaccine and one lot of albumin that flunked "sterility testing for release or for stability testing."

According to Trost, Major and Tingley, Bioport's evaluation of the problem did not conclude the failures were related to testing error. The 483 also noted that the firm's personnel had not evaluated the environmental monitoring it had conducted as part of the investigation.

According to the 483, "several failure investigations state in the conclusion that the firm has proven anti-microbial effectiveness testing," and FDA insisted that Bioport did not address corrective actions relating to personnel...though the firm concludes in several incidences that the cause of the sterility failure is technician error.

Bioport ran into problems with its adverse event reporting system, too. According to the 483, the firm had not filed a 15-day report in connection with the death of a member of the military three months after inoculation with the anthrax vaccine.

According to FDA, the date of the fatality was June 14. The lot from which the vaccine involved was drawn was not one of the three cited for stability failures.

Root stated Bioport had received the report of this fatality the day the inspection began and that the firm had subscribed to VAERS shortly before the inspection commenced.

The 483 noted the autopsy showed the fatality was due to aplastic anemia and invasive aspergillosis and that while Bioport received a vaccine adverse event reporting system (VAERS) file on the incident, the firm could not document the date of receipt. The investigators also alleged that Bioport never investigated the death.

The investigators showed little enthusiasm for the company's OOS procedure, which allowed "two independent re-tests to be performed on a sample with initially OOS test results" without requiring the lab to determine a cause for the initial OOS outcome.

The retests were to be performed by the analyst who performed the initial test and a second analyst. The SOP stated that should these retests fail to confirm the initial result, the second analyst may run a series of further retests.

FDA still chary regarding retests

The 483 stated that a bulk lot of albumin came up short for protein specification and the firm's analysts went through the retests on the next two days. The agency purged the outcomes in each of the three tests.

According to the FDAers, Bioport personnel made "a second calculation to determine whether these three results were in the 95% confidence interval" and concluded that the calculation did not confirm the OOS results. The 483 noted the second analyst ran samples through the assay an unspecified number of times and that the average was documented.

On the same date of the first assay of the bulk form of this lot of albumin, an analyst conducted the same test on a finished version of the albumin from this lot. The course of events for the finished albumin paralleled the investigation and retest of the bulk.

The 483 hit Bioport because the firm had no report "addressing the fact that both the bulk and finished product of human albumin lot #1215 initially failed to meet protein specifications" and commented that the firm had no SOP for calculation of a 95% confidence interval.

Bioport, Lansing, MI, 10/26/00, Doc. 109017M, $8.50 plus retrieval.