Three years after the world's first cervical-cancer vaccine was hailed as a public-health breakthrough, Gardasil is facing renewed questions about its safety and value.

In today's Journal of the American Medical Association, federal researchers analyze 12,424 voluntary reports of post-vaccination "adverse events" ranging from headaches to deaths. They conclude that only two complaints - fainting and dangerous blood clots -- are more common than expected and may be related to the immunization.

But an accompanying editorial points out that many questions about Gardasil remain - key among them, whether it really will reduce the toll of cervical cancer.

Another opinion piece in JAMA looks at Merck & Co.'s marketing strategy, contending the company coopted professional medical societies to promote and recommend the vaccine.

Merck - already on the defensive over Gardasil's second-quarter sales, which slumped sharply in the United States and worldwide - said in a statement that "we welcome continued study and discussion" of the product's safety.

"The bottom line is that Gardasil has a very positive benefit-risk profile," Richard M. Haupt, head of Merck's clinical program for the vaccine, said in an interview.

Gardasil, a series of three shots, protects against two strains of the sexually transmitted human papillomavirus (HPV) that cause 70 percent of cervical-cancer cases. Gardasil also wards off two other HPV strains that cause 90 percent of genital warts in men and women.

Next month, advisers to the Food and Drug Administration will consider whether to recommend expanding Gardasil's current approval for females ages 9 to 26 to include males of the same age. Vaccinating men would reduce HPV transmission to women and protect men from very rare HPV-related anal, head, and neck cancers.

The FDA panel will also review data on would-be rival Cervarix, the cervical-cancer vaccine for which GlaxoSmithKline is seeking FDA approval. Cervarix does not protect against genital warts.

Although the theory behind these vaccines is sound - if you prevent HPV infection, you prevent cervical cancer - the reality is more complex, infectious-disease expert Charlotte Haug wrote in the JAMA editorial.

No one knows whether Gardasil will require booster shots, or whether suppressing the two main cancer-causing HPV strains will drive rarer, equally dangerous strains to fill the void. That is why vaccinated women still need to get Pap smears, the routine screening test that can catch precancerous changes years before they turn malignant.

Most of the world's 250,000 cervical-cancer deaths each year occur in poor countries. Screening has made the disease relatively rare in the United States, where there are 10,000 new diagnoses and 3,700 deaths each year.

In Haug's view, the overall benefit of vaccination is so uncertain that a regularly screened woman "should be willing to accept only a small risk of harmful effects."

'A blind eye'

The size of Gardasil's risk is a matter of controversy.

Judicial Watch, a conservative public-interest group, obtained some of the FDA's adverse-event reports, and in a recent news release accused the agency of turning "a blind eye to what seems to be an extremely serious public health problem."

And the Internet is loaded with tales of women who say Gardasil made them or their daughters sick.

Jodi Speakman of Northeast Philadelphia says her daughter Victoria Ventimiglia, 18, developed a host of disabling symptoms - including non-epileptic seizures, migraines, diarrhea, tremors, and numbness - in March 2008, a month after her second shot.

"There are thousands of girls with the same symptoms," Speakman said yesterday. "We're not looking for Gardasil to be pulled from the market. But maybe they should do more research to understand why some of these girls have greater sensitivity."

Merck's view

Merck has several studies underway around the world to further assess Gardasil's safety and effectiveness, said Haupt, of Merck, which is based in Whitehouse Station, N.J.

Meanwhile, Haupt said, the JAMA analysis of adverse-event reports "reinforces the safety profile."

That analysis, conducted by the U.S. Centers for Disease Control and Prevention, is reassuring but limited, said Barbara Slade, the researcher who led the study.

In the three years since Gardasil was approved, Merck has distributed more than 23 million vaccine doses. The FDA, meanwhile, has received 12,424 voluntary adverse-event reports from patients, doctors, and Merck. Of these, only 772, or 6 percent, were serious - resulting in death, disability, or hospitalization.

But many of the reports had too little information to be verified or medically evaluated, so they were eliminated from the analysis.

The CDC researchers investigated the most common verifiable reports - both serious and not serious - including rashes, pancreatitis, autoimmune reactions, and Guillain-Barre Syndrome, a paralyzing illness that has been linked to some other vaccines.

The adverse events were no more common with Gardasil than with other vaccines - suggesting nothing more than a coincidental connection - except for fainting (1,896 reports) and dangerous blood clots (56 reports).

For some people, the sight of a needle is so distressing that it can cause blood pressure to plummet, triggering fainting, also called syncope. But the reaction is so common with Gardasil that the FDA in June revised the product label to stiffen warnings and note that fainting may be accompanied by jerking movements and "seizure-like activity." Patients should be kept off their feet for 15 minutes after the shot to prevent injuries, the new label warns.

The FDA and CDC are also discussing whether studies are needed to assess the blood-clot risk.

Even so, Slade said, Gardasil's "benefits seem to outweigh the risks."

Haug, the editorial writer, had a different view. She was not reassured by the CDC's finding that most of the women with serious blood-clot cases - including two who died - were at increased risk of clots because they used birth-control pills, smoked, were overweight, had a history of surgery, high blood sugar, or other factors.

"These are extremely prevalent risk factors," Haug, editor of the Journal of the Norwegian Medical Association, said in an interview. "Are we saying these people should not get Gardasil? I'm not seeing that in the [study's] conclusion."

The third article in JAMA, an opinion piece by Sheila and David Rothman of Columbia's Mailman School of Public Health, faults Merck - and professional medical groups such as the American College of Obstetricians and Gynecologists and the American College Health Association - for collaborating on educational programs that "did not provide balanced recommendations on risks and benefits."

Medical groups and CDC advisers, for example, recommended vaccinating all 11- and 12-year-old girls, reasoning that this age group has not yet been exposed to the sexually transmitted HPV. Some states set off a storm of controversy when they considered mandating it.

The recommendation opened a huge market to Merck, yet the company advertised Gardasil as an anticancer vaccine, downplaying the "uneasy" notion that sex was involved, the Rothmans contend.

Merck's Haupt called their article "a mischaracterization of our efforts."

"We're obviously committed to preventing HPV-related diseases," Haupt said. "Education is a critical component of that. In no way did we influence [medical groups'] decision-making" or edit what they produced.