Vaccine Safety : Part 2

by US Department of Health and Human Services

(Page 2 of 3)

The tests manufacturers must perform on each lot are spelled out in the Code of Federal Regulations or in the product license application. When the manufacturer sends the lot samples, along with the results of testing, to FDA, "we either test the lot sample ourselves or go with the manufacturer's documentation," says Jerome A. Donlon, M.D., Ph.D., director of CBER's office of establishment licensing and product surveillance. With vaccines for diseases that attack the nervous system, such as the live polio vaccine, "we test every lot because of the tremendous potential for harm," he explains.

Over the last 10 years, there have been only three vaccine recalls. One lot was recalled after FDA detected particulates; another was mislabeled. The third lot was recalled because of potential problems after an FDA inspection found violations of good manufacturing practices at the production plant.

Continuing Studies

Although clinical trials are carefully designed to uncover potential adverse reactions before FDA licenses a vaccine, "we obviously can't get all the information premarketing," says Susan Ellenberg, Ph.D., director of CBER's division of biostatistics and epidemiology. "You're never going to be able to do studies big enough to detect risks that might happen at a level of one in 100,000 or one in 1 million. We'd never get vaccines on the market. Still, such risks are important to detect because of the large population exposed. So we have to develop postmarketing surveillance programs."

For some vaccines, there are formal Phase 4 studies under way. At FDA's request, the manufacturer of the new chickenpox vaccine, licensed by FDA March 17, 1995, will monitor several thousand vaccinated children for 15 years to determine the long-term effects of the vaccine and possible need for a booster immunization. (See "First Vaccine Available for Chickenpox" in the September 1995 FDA Consumer.)

For most vaccines, the government relies on the Vaccine Adverse Event Reporting System (VAERS) to identify problems after marketing begins.

FDA and the national Centers for Disease Control and Prevention manage VAERS, a system the two agencies developed in response to the National Childhood Vaccine Injury Act of 1986. Anyone — physicians, vaccine manufacturers, patients, or the parents of a patient — can report to VAERS an adverse event that may be associated with any vaccine.

"What we're most interested in with VAERS is identifying any new problem, particularly serious problems, that might be so rare that it wasn't noticed or detected during the clinical trials," says Ellenberg.

However, many events that might be associated with vaccines go unreported.

"We don't have to have 100 percent reporting," says Donlon. Ellenberg agrees. Still, she adds, "We need enough reports to permit detection of rare events and to allow us to make reliable comparisons of reporting rates among vaccine lots. Our ability to do this improves if doctors make more reports and make them more timely."

Donlon points out that the report of an adverse event to VAERS is not documentation that a vaccine caused the event. He says doctors shouldn't make that judgment.

"Just report it," he says, "even if you've never seen it before. Maybe many others around the country are seeing the same thing."

For example, a mother recently called FDA because her child's hair had fallen out each time the child received a dose of the hepatitis B vaccine. The mother said she asked the pediatrician whether the vaccine could have caused the hair loss, but the pediatrician was sure that couldn't be the case. In fact, after the second dose of the vaccine and subsequent hair loss, the doctor was preparing to do a scalp biopsy to determine the cause.

A search of the VAERS database found 45 cases of hair loss after hepatitis B vaccination. Of those, 15 cases were like this one, in which hair loss happened after each of two doses of the vaccine.

"That's called 'positive rechallenge,'" says Ellenberg, "and it gives you a much stronger belief that the event was actually due to the product. Now, hair falling out is not a life-threatening event. But if people are aware it could happen, then they won't be imagining the worst, and invasive, unnecessary tests may not have to be performed."

Besides identifying previously unknown adverse events, VAERS is an important tool for monitoring individual lots of vaccines. "We don't expect there to be problems with vaccine lots," says Ellenberg, "because the regulations are very stringent. But even though we don't expect to find anything, we look [at the reports to VAERS] every week, and if there really was a problem with a lot, we could move very rapidly to get that lot off the market.

"One of the first things we do if we see a lot that has an elevated number of adverse events is look at its "sister" lots, the other lots that came from the same larger bulk lot. If those lots also have high rates of adverse events, it would raise our level of suspicion that there might be a problem. If, however, the other lots had average or even low rates, we would feel that this is just more likely chance variation."

Another key factor to assess the significance of the number of adverse events is the size of the lot.

"A lot with hundreds of thousands of doses is going to be associated with more events than a lot with tens of thousands of doses," says Ellenberg.

VAERS is designed to detect signals or warnings that there might be a problem rather than to answer questions about what caused the adverse event, according to "Research Strategies for Assessing Adverse Events Associated with Vaccines," a 1994 report by the Institute of Medicine. These signals can lead to hypotheses about causality, which can then be tested by other methods, such as epidemiologic or laboratory studies.

About the Author

The Department of Health & Human Services is the United States government's principal agency for protecting the health of all Americans and providing essential human services, especially for those who are least able to help themselves. The department includes more than 300 programs, covering a wide spectrum of activities.