USA got improperly made flu vaccine - By Julie Schmit, USA TODAY

[Guest guest]

Posted 2/9/2005 10:55 PM Updated 2/9/2005 11:17 PMhttp://www.usatoday.com/news/health/2005-02-09-vaccine-problems_x.htm USA got improperly made flu vaccineBy Julie Schmit, USA TODAYThe British plant responsible for the recent flu vaccine shortage sent some improperly made vaccine, by mistake, to U.S. consumers three years ago, according to a 2003 government inspection report recently given to Congress.Whether that vaccine stayed potent and sterile through its expiration date could be questioned, USA TODAY found. That was because the plant first deviated from approved manufacturing methods, then failed to document whether that would harm the vaccine's quality over time.The report doesn't say how many doses were involved, and there's no proof the vaccine harmed anyone. But the mistake — along with other deficiencies in vaccine production at the Liverpool plant — is likely to increase skepticism about the Food and Drug Administration's oversight of the plant and the vaccine industry. The FDA discovered the mistakenly released vaccine during a plant inspection in mid-2003, more than a year after millions of people got flu shots for the 2001-2002 season and a month before the plant's current owner, Chiron, bought it from PowderJect Pharmaceuticals. The FDA's report became public last fall during Congress' probe of the vaccine shortage. A close examination reveals many unreported details about the plant's vaccine, most of which was for the U.S. market:• Four out of five batches of tested vaccine made for the 2001-2002 flu season lost potency too soon. • PowderJect managers considered — but decided against — recalling flu vaccine in 2002 because of concerns about its potency.• Some required quality-control tests before 2003 were "simply overlooked," the company told FDA inspectors, and it didn't do others on time. In two unspecified years before 2004, the company sent the FDA test vaccine samples labeled as one kind of flu strain when they were actually another. • Plant operators repeatedly failed to fully and properly investigate problems. Despite these failings, the FDA largely accepted the company's word that it would correct deficiencies, even though the plant had received a bad FDA review in 1999, and was being relied upon for more of the USA's flu vaccine.Last year, the plant's 48 million doses, half the USA's expected flu-shot supply, were scrapped after British regulators suspended the plant's license when about 8% of its vaccine turned up contaminated. British officials cited deficient manufacturing conditions. The FDA then said it couldn't be sure any of the vaccine was safe.Chiron is upgrading the plant and hopes to provide flu shots to the United States this fall.The story of the Liverpool plant is one of lax FDA oversight of an industry in which millions of Americans put their trust every year, FDA critics and some pharmaceutical industry experts say. If the FDA had done more to ensure plant upgrades after the 2003 inspection, this season's shortage might have been prevented, says Rep. Henry Waxman, D-Calif., an FDA critic. The FDA disputes that.The 2003 report, which focuses on the 2001-2002 vaccine, also "raises serious questions" about the quality of vaccine the plant produced three years ago, Waxman says: "I see no clear evidence that individuals were harmed, but it is certainly a situation we do not want to repeat." The FDA says the plant's vaccine that flu season — and in others except for 2004 — met its requirements for safety and effectiveness. No vaccine was distributed that didn't pass the manufacturer's and FDA's tests, the agency says. But three drug-industry experts, including a former FDA investigator, who reviewed the report for USA TODAY say they cannot be sure that all the 2001-2002 vaccine met government standards. They don't think the FDA can say for certain, either. That's because the way the vaccine was made and tested was deficient in too many ways for anyone to be certain all of it met federal potency and safety standards through its expiration date, they say. None of them would have felt comfortable taking it themselves. "I don't feel confident that the plant's vaccine was safe and effective based on its non-compliance with good manufacturing practices," says Sarah Sellers, a drug safety consultant and FDA adviser who reviewed the report. Flu vaccine problems1999FDA concerns: 17Warning letter: YesConcerns included:High levels of microorganisms found in partially processed vaccine, requiring refiltering.No data proving refiltering didn't harm vaccine's quality.Inadequate probe into why some vaccines flunked sterility tests.2001FDA concerns: 31Warning letter: No2003FDA concerns: 20Warning letter: NoConcerns included:High levels of microorganisms found in partially processed vaccine, requiring refiltering.No data proved refiltering didn't harm vaccine's quality.No formal probe into root cause of high levels of microorganisms going back to 2000.Inadequate probe into why some vaccine flunked sterility tests.2004FDA concerns: 14Warning letter: YesConcerns included: Some vaccine is likely contaminated because of inadequate condtions as it moves from one part of the process to another - a weakness noted in 2003 inspection that wasn't adequately corrected.Company's probe into why some vaccine flunked sterility tests was inaccurate.High levels of microorganisms are found in partially processed vaccine, same as for vaccine produced in 2000 and 2001.Source: FDA Testing misses some problems Testing of vaccine is just one check to ensure that it's safe and effective. But it isn't a fail-safe method, because only samples of vaccine are tested, and problems can be missed, says Denise Dion, a former 18-year FDA investigator now with consulting firm EduQuest. She also reviewed the report.That's why regulators also check to make sure vaccines are made using good manufacturing practices, as defined by the government. Those include everything from having proven procedures to strong quality control, testing and record-keeping. Drugs made without good practices can be declared adulterated even if tests show samples are fine. Without good manufacturing practices, the World Health Organization says, it "is impossible to be sure that every unit of a medicine is of the same quality as the units ... tested in the laboratory." In mistakenly releasing the improperly made vaccine three years ago, the plant violated good manufacturing practices, the FDA says. "If you look at what's in FDA's own documents, it's stunning they didn't get a warning letter or something worse" after the 2003 inspection, says Gordon Richman, a former director of quality strategy for GlaxoSmithKline's manufacturing. He is now an FDA-regulatory consultant at EduQuest. He reviewed the report separately from colleague Dion.Warning letters demand fixes under threat of further action, such as license suspension. Instead, the FDA let the plant make fixes voluntarily, Waxman says, even though its own investigators recommended official action against the plant, he says. His office reviewed several thousand pages of documents, not all of which have been released to the public. The House Committee on Government Reform, which includes Waxman, plansa hearing todayon the flu-shot shortage and future challenges.The FDA says it allowed the plant to make voluntary fixes because it had made some corrections and planned others. In an e-mail response to USA TODAY, it also said the vaccine-making process includes safeguards, such as testing at various stages. Because of that, "violations in certain good manufacturing practices do not necessarily or routinely translate into unsafe products," it says. The FDA declined requests for interviews with FDA officials and didn't answer all questions via e-mail. Chiron declined comment on vaccines made at the plant before it owned the plant.Longtime supplier The plant has provided Fluvirin flu vaccine to the USA since 1988, the FDA says. In 1999, the FDA gave the owner then, U.K.-based Medeva Pharmaceuticals, a warning letter for violating good manufacturing practices. Among the FDA's concerns: Some of the plant's partially processed vaccine contained unusuallyhigh levels of bioburden — bacteria and other microorganisms. It had to be refiltered to remove the organisms. All flu vaccine starts out with bacteria in it because it's made using eggs, which contain bacteria. The manufacturing process is supposed to get rid of all detectable bacteria. Refiltering is discouraged because it can harm a vaccine's quality and might indicate something is wrong with the plant's system that needs to be fixed. Medeva took issue with some of the FDA's findings but promised many fixes, FDA documents show.In March 2001, when the FDA re-inspected the plant, PowderJect owned it. That inspection report was not made public. But the 2003 report says FDA inspectors in 2001 told plant operators that refiltering and reprocessing were not appropriate for U.S. vaccine.Yet, when FDA investigators returned in 2003, they found just that: Batches of vaccine during the summer of 2001 had been refiltered to successfully lower bacteria levels. They were mixed with other batches to make the final vaccine.The report says refiltered vaccine went into three lots sent to U.S. consumers. A vaccine lot typically has hundreds of thousands of doses, industry experts say.In addition to violating FDA rules, the plant lacked sufficient data to show that the reworking didn't harm the vaccine's stability, the 2003 report indicates, which would include its ability to stay potent and sterile through its expiration date and use by consumers. "Since there was no supporting data, it (the vaccine) may be OK but we don't know," says Lina Patel, a former Bayer AG scientist who is now a stability expert at Diosynth Biotechnology.When an FDA investigator asked about the reworked vaccine in 2003, PowderJect's head of quality first said refiltered vaccine wasn't sent to the USA. When the investigator found otherwise, the quality chief said the plant had FDA approval to do so, the 2003 report says. When the plant couldn't produce a letter to that effect, officials said they assumed they could, because they'd given the FDA a new operating procedure that included refiltering. Finally, plant executives said the shipment of that vaccine to the USA was a mistake, the report says. The company promised corrective actions. In 2001-2002, the plant supplied about one-quarter of the USA's flu vaccine.Potent problems In addition to mistakenly releasing vaccine, the plant struggled with potency.That might have been especially true for the 2001-2002 flu season. Then, four out of five lots of final vaccine predesignated for testing failed potency checks.A failure rate that high "is not typical" of drugmaking plants, says consultant Patel. "It's not good," she adds.Failure in a test lot could be important, because it might indicate problems with others. In 2002, for instance, vaccine maker Aventis Pasteur, now Sanofi Pasteur, voluntarily recalled some vaccine protecting against meningococcal disease after samples from four lots failed routine potency tests.The 2003 inspection report doesn't indicate consumers got subpotent vaccine. But PowderJect was sufficiently concerned to consider a recall. The company's report justifying its decision has not been released to the public. The FDA inspectors said the plant's investigation into the potency failures was incomplete, partly because even in 2003 it had failed to identify the cause of the problem. In addition, the FDA says the plant didn't do some required tests and didn't tell the FDA soon enough about others. In one instance, that meant in time to protect consumers, if needed.FDA investigator Robert Jennings noted that some "potency failure information that was likely to impact ... the current flu season, was not reported to the FDA."The report notes several testing and reporting mishaps. It's not clear which one he was referring to.In one, the plant was supposed to test a lot one, two, three and six months after it was manufactured. It didn't. That wasn't reported to the FDA. When FDA inspectors asked in June 2003 why the tests weren't done, the company's response "was that they were simply overlooked," wrote FDA biologist Jonathan McInnis. When the vaccine was tested at the seven-month point in May 2002, it failed.In another case, one lot of vaccine failed a potency test in February 2002. But the FDA didn't learn of it until four months later. It was supposed to be reported within 45 days. The test result was close to passing, and according to the FDA report, the plant "reportedly did not consider" it a failure. Three FDA officials who wrote the 2003 report declined comment or did not return telephone calls. The fourth could not be reached.New leadership Chiron has replaced several of the plant's top managers since last fall. Two of them were at the plant in 2003, and one dating back to at least 1999.The FDA is working to approve other flu-shot vaccine suppliers, but the Chiron plant is crucial. It was one of only two plants, along with one owned by Sanofi Pasteur, fully licensed last year to make flu-shot vaccine for U.S. consumers. That was also true at the time of the 2003 inspection — a factor that might have influenced the FDA not to come down too hard on the plant for fear of losing a key supplier, several industry consultants say."I'm sure the FDA had that in mind," says David Webster, pharmaceutical consultant with Webster Consulting Group. He says the FDA constantly struggles to balance the need for vaccine against ensuring manufacturers meet high standards. "It is not an easy decision," he says.Nor is it an easy task. The Liverpool plant became one of two U.S. flu-shot vaccine makers after two others, in 2000 and 2002, quit the U.S. market. One had its U.S. plant shut by the FDA after repeated problems. The other said it got out because the business did not justify costly plant upgrades needed to meet FDA requirements.