GlaxoSmithKline surmounted the final safety hurdles for its human papillomavirus vaccine Cervarix at a landmark September 9 FDA advisory committee review of the adjuvanted product.

The sponsor, however, faces a remaining hurdle to convince FDA that its bivalent vaccine should carry labeling as efficacious in the prevention of cancers from HPV types beyond the two virus types which are the specific targets of the vaccine: HPV-16 and 18.

Securing a strong positive vote on safety issues (11-1) from the Vaccines & Related Biological Products Advisory Committee is a significant milestone for GSK and for the development of the next generation of vaccines across the industry.

Cervarix is the first product with a novel adjuvant (AS04) to get as far in development in the U.S. as a positive FDA advisory committee review.

The advisory committee endorsement of safety was probably delayed by at least two years from GSK's initial development schedule; but it is a strong endorsement. The recommendation that the vaccine with adjuvant is safe creates a solid base for FDA to conduct its final review of Cervarix and opens the door to an era for other vaccines with new adjuvants.

The extra steps that GSK took to answer safety issues are instructive to other vaccine makers pursuing novel adjuvants that may extend the effectiveness and duration of protection of vaccines. From that perspective, the GSK Cervarix BLA breaks the ice on adjuvant reviews and provides a roadmap for future sponsors.

On the efficacy side, GSK wants labeling for Cervarix to include reference to protection beyond the two valents specifically included in the vaccine. That is clearly important to help GSK in its effort to play catch-up to Merck's three-year lead in the U.S. with Gardasil and the added genital warts protection from that four-valent product.

GSK VP-North American Vaccine Development Barbara Howe, MD, summarized the importance of the broader indication in her closing remarks to the advisory committee.

"A second vaccine, one which has demonstrated protection on oncogenic HPV types beyond 16 and 18," would be important to prevent cervical cancer. Cervarix would provide an additional protection against 11-16% of cervical cancer cases because of its carry-over effect to other HPV viruses, she maintained. That number comes from the prevalence of cancers from HPV virus types 31, 33 and 45 globally.

"Cervarix with cross-protection," Howe declared, "will prevent more cancers and save more lives than a vaccine without cross-protection."

FDA has indicated skepticism on that point thus far in the BLA review. In the briefing documents prepared for the September 9 meeting, FDA pointedly noted that the sponsor is seeking an indication for some non-vaccine HPV types. The proposed FDA indication is limited to types 16 and 18.

The breadth of the indication appears to be the big issue in the final FDA review of the application. The product has a user-fee review deadline at the end of this month.

Norman Baylor, PhD, director of FDA's Office of Vaccines Research & Review, told the advisory committee that the agency needed the comments and advice of the advisory committee on whether there was evidence of cross protection on any, or specific, HPV viruses beyond 16 and 18. The agency "will have to evaluate this," he pointed out.

Committee Leans To Efficacy Against HPV-31

The chair for the Cervarix advisory committee session, John Modlin, MD, Dartmouth-Hitchcock, summarized the committee's view of those extended indications.

The committee feels, Modlin said, that there is good basis for protection from Cervarix for several strains beyond 16 and 18, but the committee would be very uncomfortable assuming that activity extends to a broad range of HPV types.

Modlin further observed support specifically for including activity against HPV-31 - which GSK says causes 4.5% of the cervical cancers in North America.

Temporary voting member Robert Seder, MD, Vaccine Research Center, National Institute of Allergy and Infectious Diseases, said the data on protection was "clinically most consistent with type 31." He wondered whether the effect derived from the MPL (monophosphoryl lipid A) adjuvant or the difference in the nature of the vaccine-like particles for HPV-16 and 18 used in the GSK product versus the VLPs in the Merck version. Seder noted that HPV-31 is phylogenetically related to HPV-16.

Advisory committee member Peter Gilbert, PhD, University of Washington, said "My reading is that there is clear evidence for protection against 31 but there is not clear evidence for protection against any other individuals." The data on protection against HPV-31 reached statistical significance in clinical trials.

FDA Acknowledges Activity On HPV-31

Prior to his summation of the advisory committee discussion, Modlin offered his own view that "the analyses for each individual type [of HPV infection 31, 33 and 45] all seemed to go in the same direction - that is towards some evidence of the degree of protection by the vaccine."

The committee chair said that suggests "that there probably is a biological effect. I don't think that we can precisely measure that ... I personally believe that that is the case."

Temporary advisory committee member Elizabeth Unger, PhD, MD, Centers for Disease Control, felt that the data on cross-protection was promising but not sufficient for specific recommendations.

There are "so many uncertainties," Unger said. "While it looks very promising, I think it is too early to specify a specific type" of virus where Cervarix is protective beyond 16 and 18.

FDA has acknowledged in briefing materials on the BLA review that the vaccine "may have an impact on reducing genital dysplasias ... related to HPV-31."

GSK VP and Director of the Global Clinical Development Center Gary Dubin, MD, told the committee that the cross-protection analyses done by the company "indicate the true vaccine efficacy against CIN2+ associated with non-vaccine types lies between 37-54%." The range of non-vaccine viruses beyond 16 and 18 includes 12 strains including the three prominent ones: 31, 33, and 45.

"So what does this additional protection mean in terms of public health impact?" Dubin posed during the companies presentation. "Since about 30% of cervical cancers are caused by non-vaccine types, we expect globally that Cervarix would provide an additional 11-16% protection against cervical cancer beyond its effect on HPV-16/18.

The agency and sponsor have a lot of work to do before the next deadline (since the user fee date for the resubmission is at the end of September).

In addition to a decision on the cross-protection labeling, the agency and sponsor have to confirm large Phase IV commitments.

The final decisions are significant but they are not impossible in the short time to the user fee deadline - especially in light of how far GSK has come to this point.

An 11-Year Development Program In U.S.

Cervarix has been in trials for U.S. licensure since 1998. The final discussion of safety on the adjuvant has been ongoing with FDA in earnest since October of last year, when the company met specifically with FDA on the adjuvant.

The messages to other adjuvant sponsors from the Cervarix review: be ready with worldwide meta-analysis data on an adjuvant in all vaccine formulations (not just in the disease in question); be prepared for a number of outside review organizations (including other U.S. government entities) to weigh in on data interpretation on side effects outside the condition addressed by the vaccine; and prepare for a major postmarketing safety effort.

GSK was asked to undertake a meta-analysis of all data on three vaccines that use the AS04 adjuvant (Cervarix, Fendrix and an HSV vaccine in late Phase III trials) in mid-2007.

The meta-analysis request preceded a "complete response" letter from FDA at the end of 2007. GSK eventually provided FDA and the advisory committee with meta-analysis data on 68,512 patients who had been exposed to vaccines with the adjuvant.

The request for the meta-analysis on the adjuvant in different vaccine formulations emphasizes the degree of caution by FDA regarding safety. The agency traditionally looks at vaccines in their final formulation for approval decisions. Faced with the new adjuvant, the agency asked for all the information it could get - even in other vaccines.

The final safety hurdles for the vaccine were questions (1) about a higher rate of spontaneous abortions seen in women taking the vaccine and (2) "numerical imbalances in musculoskeletal and neuroinflammatory events" in patients receiving the vaccine in clinical trials.

The spontaneous abortion issue was the one that brought in another government agency. In the middle of this summer, GSK sent data on the spontaneous abortions to the National Cancer Institute for a separate third-party analysis to confirm previous outside reviews of the spontaneous abortion data in trials. The NCI analysis, according to FDA, "could neither deny nor confirm an increased discrepancy in spontaneous abortion rates among vaccine recipients."

FDA similarly used outside advice from a panel convened by GSK and from a neurologist contacted by FDA directly to address the neuroinflamatory events in trials. FDA said the non-agency neurologist found that the "data was not sufficient to establish a link" but wanted monitoring in the postmarket setting.

At the advisory committee, temporary voting member Kenneth Noller, MD, Ob-Gyn Department at Tufts University School of Medicine, quieted the concerns about the spontaneous abortions.

Noller pointed out that the comparator product in the Cervarix pivotal trial may have had an unusual protective effect against spontaneous abortions. GSK used its hepatitis A vaccine Havrix (HAV) in the control group at an old dosage regimen to match the Cervarix three-shot dosing.

"I hope nobody takes this as a flippant remark because I don't mean it that way at all," Noller said, "but it seems to me there is an alternate interpretation in the difference in those rates of spontaneous abortion."

The HAV vaccine, Noller contended, "was protective against spontaneous abortion but the Cervarix was less so." Both products had lower rates of spontaneous abortions in the pivotal trial HAV (5%) and HPV (11.6%).

The background rates for spontaneous abortions are about 15-16%, Noller estimated. Neither of these vaccines met that threshold. "If you would just compare it to what you would expect, they are both considerably lower."

Noller was one of a minority of non-government employees on the committee for the Cervarix vote. Because of a number of absences and a large pool of temporary voting members, the committee was made up of eight government employees and six advisors from outside government.

The committee and FDA are looking to the sponsor for a substantial postmarketing effort. On the pregnancy issue, the company is preparing to do a registry in the U.S. GSK is already one year into collaborating on a pregnancy registry with the Health Protection Agency in the UK where Cervarix has 70% of the vaccinations in young adolescent girls for HPV.

GSK is also developing a Phase IV observational study on 100,000 women to screen for autoimmune diseases within 12 months after first vaccination and women vaccinated during pregnancy or who get pregnant within about two months of ending the vaccination schedule.

"We have interacted with several potential investigators who have access to [administrative claims] databases and they have run some checks in their databases to make sure they catch the adverse outcomes as well as the actual pregnancies," GSK Biologicals Clinical Safety & Pharmacovigilance VP Thomas Verstraeten, PhD, told the committee.

The FDA briefing documents indicate that GSK was initially thinking about a Phase IV study of 44,000.

The control group has been proposed to be patients receiving Sanofi-Pasteur's Menactra , which suggests that the trial may derive from the claims database study to the Vaccine Safety Datalink project undertaken to follow Menactra after safety issues developed in the postmarket period.

By asking for more information over the past two years, FDA has in effect set up its own postmarketing trial outside the U.S.

The company notes that more than 7 million doses have been distributed. That may be the final factor that helps FDA with the safety issue around the new adjuvant.

- Cole Werble ( ¹ c.werble@elsevier.com )

Cervarix Passes Safety Hurdles; FDA Ponders Breadth of Protection Claim

More from Archive

More from Pink Sheet