This is not an idle question, for Lowy and Schiller’s conflict disclosure forms the basis for an alternative to Zerhouni’s narrative, one that spotlights the unusually self-contained set of Department of Health and Human Services (DHHS) activities that surrounded HPV vaccine development. This alternative narrative is more of a business story than a scientific one, a narrative in which commercial interests were inextricably linked to matters of life and death. In this narrative, Gardasil is perhaps the leading example of a new form of unconstrained government self-dealing, in arrangements whereby DHHS can transfer technology to pharmaceutical partners, simultaneously both approve and protect their partners’ technology licenses while also taking a cut of the profits. Literally and figuratively, DHHS has the authority in such situations to allow its business partners to get away with murder for the greater good, effectively granting its private business partners a license to kill.
 
DHHS officials have their own language for such arrangements.  They call them public-private partnerships, and DHHS agencies have gotten progressively more aggressive about pursuing them. NIH, for example, launched its own “Program on Public-Private Partnerships” in 2005, shortly before Gardasil’s launch. On the web-site describing this program, the NIH program managers concede that the kind of technology transfer involved with Gardasil carries unavoidable ethical risks, acknowledging that “The potential for conflict of interest exists any time the NIH and NIH staff engage with non-Federal entities to achieve mutual goals.” They provide little more than a pro forma solution for such conflicts, however:  any concerned NIH staffers are encouraged to “contact their Deputy Ethics Counselor.”
 
It’s important to shed light on this alternative narrative as a counterpoint to the heroic story promoted by Gardasil’s many sponsors. An uninformed observer might like to assume that the responsible agencies of DHHS care not at all about commercial opportunities and exclusively attend in a disinterested fashion to the issues of health and safety that would naturally concern any consumer of vaccine products.
 
But that assumption would be incorrect. By taking a commercial perspective on Gardasil’s development and regulation, one is forced to confront a new and disturbing question. How is disinterested vaccine safety governance even remotely possible when DHHS employees stand as heroes at the head of the parade when a new vaccine is invented within its walls, while agency leaders are leading the cheering section, approving the new product’s launch, making the market for the product with its recommendations and then turning around to cash multi-million dollar checks? In order to better understand the real lessons of Gardasil under the harsh light of the business interests at work, let’s take a closer look at how the Merck-NIH partnership on Gardasil was forged.
 
Conflicts of interest in vaccine development and regulation
 
As the world’s largest single sponsor of biological research, NIH frequently funds research with commercially valuable outcomes. When that R&D generates potentially valuable inventions, NIH submits patent applications to the U.S. Patent and Trademark Office (USPTO) and actively pursues the approval of those patents, which when granted become valuable commercial property for DHHS, the patents’ owner. Since NIH has neither the authority nor the capability to pursue product commercialization efforts, in order to encourage private companies to invest in conducting the necessary clinical trials, NIH’s Office of Technology Transfer (OTT) was created to grant commercial licenses for such DHHS patents to commercial partners, including vaccine manufacturers. When new products invented at NIH clear the requisite regulatory hurdles at the Food and Drug Administration (FDA) and reach the market, OTT then shares in the profits. They also distribute the rewards back to the scientific teams whose products have succeeded in reaching the commercial stage: when license fees flow into OTT’s coffers, the Federal employees who invented the technology are entitled by NIH policy to a share of the royalties.
 
From a technology development standpoint, such commercial arrangements are the result of an intentional public policy; in fact they resulted from an Act of Congress. The Bayh-Dole Act of 1980 was written with the express purpose of making it easier for federally-funded academic research to receive patent protection that would allow the ready licensing of the fruits of commercially valuable R&D to private businesses. At the time, the concern of Congress was that federally funded inventions too often languished within the academy because businesses had insufficient incentive to invest in clinical trials, since these inventions were often unsupported by the powerful competitive protection afforded by an exclusive patent license.
 
The policy worked. Within the research universities that receive the vast majority of federal funding, Bayh-Dole has had the desired effect and has enabled university technology transfer offices all over the world to generate billions of dollars of licensing revenue in the last few decades--especially in the life sciences--by licensing patents from federally-funded university research to corporate partners. Bayh-Dole has effectively turned research into big business for many universities and transformed technology transfer offices into important profit centers at academic institutions all over the world.
 
But when technology licensing takes place within federal agencies, Bayh-Dole creates an entirely different problem: an unprecedented web of conflict, one in which the same departments that are tasked with regulating the health and safety of medical products are also profiting from them. As Lowy and Schiller conceded in their review article disclosure, this conflict of interest came into play directly on Gardasil: both men are named inventors on the technology that makes Gardasil possible; NIH filed for and received patents on their invention of the VLP technology; DHHS is the owner of the patent family that protects the commercial rights to the invention; in order to bring the product to market, OTT licensed the vaccine technology to Merck; and as Merck has generated billions in Gardasil revenue, OTT has received millions in Gardasil profits.
 
But DHHS is also responsible for regulating Gardasil in numerous ways. The FDA reviewed the clinical trials in which Gardasil was tested in human populations and passed judgment on Gardasil’s safety. An Advisory Committee on Immunization Practices (ACIP) of the Centers for Disease Control and Prevention (CDC) decided whether or not to recommend Gardasil for young women and children. The FDA and CDC together now conduct the surveillance to decide whether or not Gardasil is proving safe in larger populations. And as some families are now beginning to seek compensation based on claims that Gardasil caused injury in some of its recipients, the division of the Health Resources and Services Administration (HRSA) that oversees the Vaccine Injury Compensation Program (VICP) will soon sit in judgment as to whether, to whom, and how much compensation will be provided to Gardasil’s victims.
 
As you can see in the chart below, all of this activity is supervised in a single department by one Cabinet official, the Secretary of Health and Human Services. The sole non-governmental agency involved in this commercial enterprise is Merck’s Vaccine Division.  In effect, the Merck-DHHS partnership leaves the business side to Merck while DHHS is solely responsible for

1. Creating the market for Gardasil by funding commercial research, supervising the conduct of clinical trials, judging the outcome of those trials and promoting a policy of universal vaccination;
2. Collecting the license fees that result from Gardasil revenues from Merck and other vaccine manufacturers and then distributing these financial benefits to Federal employees; and
3. Deciding whether or not to protect the policy decisions and profit streams of their sister DHHS agencies through postlicensure safety monitoring and vaccine injury compensation rulings.

Is this good government at work or an example of the medical-industrial complex run amok? In this investigative series, Age of Autism will take a look at how DHHS agencies have managed Gardasil in all three of these sequences. We’ll start by taking a closer look at the NIH patent portfolio and the associated license fees that have been flowing into NIH coffers since 2006. (Click chart to see original slide.)



 
Celebrating the invention of a new market
 
Lowy and Schiller are both employed by the National Cancer Institute (NCI). One of the largest of the NIH institutes, NCI was established in 1937 by Franklin Delano Roosevelt. For many decades, NCI has been the agency at the forefront of the so-called “War on Cancer.” Perhaps the earliest inspiration for the both the Cancer War and the Gardasil program began during the 1960s, when NCI researchers first began looking in earnest at viruses as a potential cause for cancer. In 1961, NCI leaders created the Laboratory of Viral Oncology to begin the search for cancer-causing viruses; in 1962 the Human Cancer Virus Task Force was first convened; and by the end of the decade, enthusiasm over this research was part of the scientific momentum that persuaded President Richard Nixon to launch the War on Cancer in 1971. Unfortunately for Nixon’s legacy, and for most subsequent cancer victims, the War on Cancer has famously failed to find a cure for cancer or to validate theories of viral causation in the vast majority of human cancers.
 
But starting in the 1980s, the two exceptions to this litany of failure—hepatitis B virus and the human papillomavirus--led to the launch of two blockbuster new vaccine products. The infant hepatitis B vaccine was developed in the 1980s and launched in 1991 with an ACIP recommendation that all American infants be vaccinated on the first day of life. And after 1984, when Harald zur Hausen first pinpointed the role of certain strains of human papillomavirus in cervical cancer, the work on another anti-cancer vaccine could begin. By the early 1990s, laboratories all over the world were racing to develop the first HPV vaccine.
 
Lowy and Schiller’s NCI team were among the four most active research teams in this race, all of whom were aggressively filing patents on their HPV inventions. Along with a third NCI colleague, Reinhard Kirnbauer, Lowy and Schiller filed their first application for a patent entitled “Self-assembling recombinant papillomavirus capsid proteins” on September 3, 1992. Since then--and after splitting the original application into 29 “children” in the form of numerous “divisionals”, “continuations” and “continuations-in-part”--nine patents from that family have been granted, as well as four from a branch of the family tree entitled “chimeric papillomavirus-like particles.” The ability of the novel “L1 proteins” described in their patent to “self-assemble” into virus-like structures, which when deployed in a vaccine solution could stimulate a protective immune response against HPV, formed the essence of their invention. Although OTT doesn’t specify the royalty-bearing patents, the commercially valuable technology that Merck has licensed likely comes from this group of nine “self-assembling recombinant papillomavirus capsid proteins” patents: US5437951, US5709996, US5716620, US5744142, US5756284, US5871998, US5985610, US7220419, and US7361356.
 
The NCI team was among the leaders in HPV technology, but the race to make a commercially viable HPV vaccine involved several other research teams from all over the world. Most notable among these were the University of Queensland in Australia, Georgetown University and the University of Rochester. In addition to NCI’s filings, each of these university-based research teams filed their own patents; eventually, Merck and GSK got into the act as well. Like many promising areas of technology, the HPV patent landscape became large and crowded in a short period of time.
 
Amid this blizzard of activity, the USPTO’s Bureau of Patent Appeals and Interferences (BPAI) had to step in to sort out whether these competing patent applications interfered with each other and to distribute the credit, making a series of hotly contested decisions that were ultimately appealed to the Court of Appeals for the Federal Circuit (CAFC), the nation’s most powerful patent court. By 2007, all the BPAI and CAFC rulings had come in and the respective contributions of all four groups were conclusively allocated for commercial purposes. The team led by Ian Frazer at the University of Queensland received credit for the being the first to propose the idea of using VLP technology for a vaccine, since their application was filed on July 20, 1992, just six weeks earlier than the NCI team’s. But thanks to their unique technology of “self-assembly,” most of the invention claims of the NCI patent family remained intact as well; Lowy and Schiller’s invention has since been generally accepted as a critical advance in the wave of new technology that made Gardasil possible. In terms of the distribution of financial reward, both Rochester and Queensland have reported receiving royalty income for their HPV inventions (in undisclosed amounts) in addition to the revenues reported by OTT.
 
As the technology transfer officials at OTT were paving the way for the financial benefits from Gardasil to flow back to NIH, Lowy and Schiller were benefiting in other ways as well, especially when it came to scientific credit. Throughout much of 2006 and 2007, they received awards from many quarters for their role in developing Gardasil’s “virus-like particles.” Their joint awards included the Dorothy P. Landon-AACR Prize for Translational Cancer Research in April 2007 and the 2007 Novartis Prize for Clinical Immunology. In addition, Lowy by himself received the Daniel Nathans Memorial Award in September 2007 and the American Cancer Society’s Medal of Honor for Basic Research in October 2007. 
 
In addition to these awards, on September 19, 2007, Lowy and Schiller received what was perhaps their crowning honor. That’s when the Partnership for Public Service awarded the pair the “Federal Employees of the Year Service to America Medal.” According to its sponsors, “The Service to America Medals have earned a reputation as one of the most prestigious awards dedicated to celebrating America’s civil servants. Often referred to as the 'Oscars' of government service," they are more commonly known in government circles as the “Sammies.” Upon receiving his crowning honor, Lowy was interviewed for the NIH Record and professed the requisite modesty in its October 2007 edition, saying “We are simply symbols of the many people who have made critical contributions to understanding the relationship between papillomavirus infection and cervical cancer.”
 
If Lowy was modest, the top brass at NIH could barely conceal their pride over their employees’ accomplishments.  According to the Partnership for Public Service, “Lowy and Schiller’s 20-year partnership has been a boon to the nation’s health and for the advancement of scientific discovery.”
 
Collecting the licensing fees
 
Alongside the science and policy celebrations, the business side of the Merck-NIH partnership proceeded with a bit less fanfare and with a different kind of currency. Once their patent was approved, OTT could then turn to extracting their share of the benefits from their commercial partners’ new products, which in the case of HPV vaccine included sales first from Merck’s Gardasil product and later from GlaxoSmithKline’s Cervarix. Merck reached the market first in 2006, but GSK followed shortly thereafter in 2007. As each company began collecting revenue from their new vaccines, OTT began collecting royalties. The table below shows Age of Autism’s analysis of how Merck and GSK’s revenues may have flowed into OTT’s coffers. 

	Gardasil Revenue ($M)	Cervarix revenue ($M)	NIH Top 20 Revenues ($M)	HPV Rank in NIH Top 20	HPV Revenue: estimated at 1% license fee ($M)
2006	235	--	NA	NR
2007	1,481	20	71 (est)	#4	15
2008	1,403	229	77.4	#2	16
2009	1,108	292	75.7	#1	14

Both Merck and GSK itemize revenue for Gardasil and Cervarix in their quarterly and annual earnings statements. Their annual results are summarized in the first two columns of the table. For Merck, Gardasil has been a blockbuster success, yielding a cumulative total of over $4 billion in revenue through year end 2009. By contrast, GSK’s revenues have been growing more slowly and have not yet reached a cumulative total of half a billion dollars.
 
For their part, OTT does not itemize their HPV license revenues. However, they do report their total royalty revenue as well as the cumulative revenue from their “top 20” technology licenses since 2007. These top 20 licenses have been worth over $70 million annually in profits for NIH in the last three years, and HPV licenses have soared to the top of those rankings quickly. Last year, OTT reported that HPV licensing was its top revenue generator. OTT doesn’t disclose exactly how much the Gardasil and Cervarix royalties contribute to NIH, but if we make the assumption that their patent licenses entitle them to 1% of the HPV vaccine revenues of their partners (an assumption that appears reasonable based on the available data), then we can safely estimate that OTT has been collecting somewhere in the range of $15 million per year from Lowy and Schiller’s invention.
 
In addition to their numerous scientific awards for their discoveries, Lowy and Schiller have received cash distributions from NIH based on their patents. As Federal employees, they are each eligible to receive a share of patent royalties up to $150,000 per year and Gardasil’s success has guaranteed that they would receive the maximum reward. That means that since FDA’s approval in 2006, each man has earned roughly a half million dollars in royalty revenue.
 
* * *
This is the DHHS vision of public private partnership at work. Contrary to the rhetoric, these partnerships aren’t simply a high-minded collaboration of scientific visionaries, but rather a large commercial enterprise with extraordinary profits at stake: an enterprise from which NIH receives credit and money and based on which its corporate partners build multi-billion dollar businesses.
 
How does such a partnership affect the incentives of regulators whose job it is to make sure the products are safe? It’s not obvious that they do. Just because DHHS has a financial stake in Gardasil doesn’t necessarily mean that every subsequent decision its employees make is corrupt, part of some nefarious conspiracy to kill young women for money. Indeed, HPV royalty revenues of $15 million represent just a small fraction of a DHHS budget that rose to well over $700 billion in 2009. In the larger scheme of things, DHHS revenues on Gardasil are just a small drop in a very large bucket.
 
Far more likely to play a role, however, in public-private paternerships like the Gardasil vaccine are the insidious cultural pressures that emerge in a supremely political organization like DHHS.  Can we really expect the Secretary of HHS to take his or her FDA Director to task for implementing lax standards on vaccine approval when the Director of NIH is simultaneously praising the “heroic” researchers who invented the product in the first place? Is it more likely that CDC will apply extra caution in their vaccine policy recommendations when its sister agency is involved or will they be more likely to activate the fast track in their process of making recommendations for Gardasil? What we have observed so far merely suggests the potential for bias in the regulation of products in which DHHS holds a direct stake. In the next part of our series, Age of Autism will investigate the question of whether or not there have been actual patterns of bias in the ways in which regulators at FDA and CDC have conducted their duties with respect to Gardasil.
--

If the FDA trial standards were truly to enforce a high standard of safety, they would require the comparison of Gardasil’s safety profile to a true saline placebo. But Merck performed no such analysis and CBER permitted them to apply a lesser safety standard of safety analysis. As a result, CBER issued its BLA approval without any idea whatsoever of the true risks of Gardasil. Not surprisingly, most of the comparisons between adverse outcomes for those receiving doses of Gardasil and those exposed to an aluminum adjuvant “placebo” showed little evidence of injury risk from Gardasil.

Unfortunately, the conclusion that Gardasil was therefore safe was horribly wrong.

A different view of the Gardasil trial data

Based on the data provided in CBER’s review of the Gardasil trials, it is possible to piece together an alternative view of Gardasil’s adverse event profile by examining three separate populations: 1) the subjects who received actual doses of Gardasil (over 96% got all three doses);  2) the subjects who received a “placebo” containing an aluminum adjuvant (over 98% got 225 micrograms of amorphous aluminum hydroxide sulfate) formulated in a carrier solution that made it visually indistinguishable from the full vaccine; and 3) the subjects who received only doses of a carrier solution. For the Gardasil and aluminum adjuvant groups, safety results were collected in two ways: a smaller set of reported outcomes was measured for the entire trial group (the “general safety population”) and a smaller group (the “detailed safety population”), including the entire carrier solution group, followed a more detailed protocol. The respective sizes of these safety assessment groups are shown below. Unfortunately, the small relative size and somewhat unmatched profile of the carrier solution group reduces the statistical power of a comparative analysis across the three groups: the age profile of this carrier solution trial was younger (9-15 years of age) than the other four trials (10-26 years of age, with the bulk falling between 16-23 years old); and less female (54%) than the Gardasil recipients (over 90%) and the aluminum adjuvant recipients (100% female). Nevertheless, the results of this three way comparison are the closest thing we have to a valid, non-passive safety analysis; and they show striking differences in safety profiles, none of which can be attributed to sample bias.


There are several ways in which the CBER trial review permits a comparative safety analysis across all three groups. The first is by comparing immediate adverse events at the injection site: events such as pain, swelling, “erythema” (redness of the skin), hemorrhage and pruritis (itching). These events are highly specific and show up in the first few days; they can, however, vary quite a bit in terms of severity. The Gardasil trials reported their results for these injection site adverse events in the “detailed study population” within five days after any vaccination visit. The comparison of these outcomes is shown below (using a scale that keeps the ratios between the rates of the adverse events constant).

 
As the chart shows, the vast majority of the Gardasil (81%) and aluminum adjuvant (75%) groups reported some kind of adverse event, most of which involved some kind of pain. By contrast, less than half of the carrier solution group (45%) reported an adverse event. This pattern continues in almost all of the individual categories, with the Gardasil group showing the largest rate of local reactions, followed closely by the aluminum adjuvant group and then with a clear drop off in the frequency of adverse events in the carrier solution group. On a retrospective basis, all but one of the reduced risks for the carrier solution group were statistically significant.

The most striking difference between the three groups is in the area of “serious adverse events.” Although less frequent than minor instances of pain or swelling at the injection site, these serious events were disturbingly common in the groups exposed to active substances. Nearly 5% of the Gardasil recipients had a serious adverse event, well over six times the rate of the carrier solution group. And more than 2% of the aluminum “placebo” recipients had severe reactions, more than three times the rate of adverse events in the carrier solution group. Based on this finding alone, it’s hard to defend the choice to classify Merck’s adjuvant as an “inert” placebo.

A second approach to comparative safety analysis involves examining the adverse events that caused the participants to withdraw from the trial in a two week period after any vaccine visit. These withdrawals included a range of adverse reactions, only a small fraction of which the investigators designated as “severe.” But sudden deaths (which need not be specific to the vaccine) were also included. The comparison of the discontinuation rates in the three groups is shown below.

 
Overall, the rate of discontinuation was low, at less than half a percent. But in the carrier solution group not a single recipient chose to drop out of the trial. In addition, there were three discontinuations after two weeks due to deaths in the Gardasil group and one such death in the aluminum adjuvant groups, whereas there were zero deaths at any point in the carrier solution group. Seven discontinuations (four in the Gardasil group and three in the aluminum adjuvant group) were due to other severe adverse events. These are obviously small numbers, and the deaths were dismissed by the reviewers as unrelated to vaccination. And in fact, the rate of discontinuation in the Gardasil and aluminum adjuvant groups was nearly identical. As a result of this similarity on outcomes, the CBER reviewers dismissed any effect of vaccination on withdrawal decisions, in all likelihood because the vast majority of the officially designated “placebo” group was exposed to the aluminum adjuvant.

A third approach to a comparative safety analysis takes a longer view of adverse events, using data for serious adverse events over a twelve month period after the beginning of the trial. The FDA review includes voluminous data on these events, but one of the easiest to measure is simply the overall rate of serious adverse events. The trial data show rates for such serious events that were similar between the Gardasil and placebo group. Indeed the rate of serious adverse events in the Gardasil group (1%) was actually lower than the placebo group as a whole (1.1%). Not surprisingly, however, this result was driven entirely by a high rate of serious adverse events in the aluminum adjuvant group. When one examines the rate of serious adverse events in two distinct placebo groups, the rate of serious adverse events in the aluminum adjuvant group rises even higher, to 1.27%, while the rate in the carrier solution group comes out at zero. This comparison is shown below.

 
A final approach to safety assessment takes the extensive twelve-month data on the medical conditions in all trial subjects and examines the longer term adverse events in specific categories of interest. Several such categories show disturbing patterns. Autoimmune conditions like arthritis, lupus and thyroiditis were sharply higher in the Gardasil group when compared to the overall “placebo” group and were even noted by the FDA reviewer as a source of concern. These occurred at a rate of over 1 in 1000 in the Gardasil group; there were, however, zero reported cases of autoimmune disorders in the carrier solution group. As in the two week analysis, death rates over twelve months were higher in the Gardasil and aluminum groups. By contrast, the carrier solution group had no deaths in the longer period. The chart below shows the results for the twelve month analysis.

 
How much of the low rate of adverse events in the carrier solution group (officially designated the “018 Protocol”) was due not to real differences in outcome but rather to sample bias, the fact that the population for the 018 Protocol was younger and less purely female than the other four trial populations? The short answer is, not very much. There are several ways to test the effect of sample bias. These include: comparing the adverse event rate in the 018 Gardasil group to the Gardasil groups in the other four protocols (higher adverse events show the 018 population was more vulnerable); comparing differences in adverse event rates between boys and girls in the 018 Gardasil group (higher adverse events in boys also show the 018 population was more vulnerable); and comparing differences in rates between the 9-12 year olds and the 13-15 year olds in the 018 Gardasil groups (higher adverse events in younger subjects show the 018 group was more vulnerable). If anything, most of these comparisons suggest the use of the carrier solution group understates the adverse event rate for Gardasil. For example, the younger subjects in both 018 groups had a higher rate of injection site adverse events and the 018 Gardasil group also had a higher rate of severe adverse events than the other groups. Only the findings on deaths and discontinuations (which were most frequent in the Protocol 018 Gardasil boys and 13-15 year olds) might have been influenced by sample bias.

The FDA downplayed deaths during the clinical trial

When it came to the most serious adverse event of all, death, the FDA review effectively gave Gardasil a free pass. They failed to mention, of course, that the deaths in their “placebo” group actually received the entirety of the vaccine’s contents excepting the VLPs. Nevertheless, they did report briefly on each individual case of death. In cases of death due to traumatic events like motor vehicle accidents, however, no details were reported (could a seizure or heart attack while driving have caused some traumatic events?). In most of the biologically-related deaths, they found reasons not to make any connection to Gardasil or to blame the victims’ behavior (“they were on birth control pills”) or family history (“the family had a history of arrhythmia”). Here is the FDA reviewer’s summary of the deaths in the trial.

There were 10 deaths in the Gardasil recipients (0.8%), and 7 deaths in the placebo group (0.7%). The majority of the deaths were due to trauma in both groups. These deaths did not appear related to vaccine administration.

In each treatment group, there was a death related to a deep vein thrombosis and/or pulmonary embolism, and both subjects were on hormonal contraceptives. The Gardasil recipient with this event had symptoms of leg pain prior to the first vaccination. The other Gardasil recipients who died included one subject with pancreatic cancer 578 days after dose 3, and one young male who died of arrhythmia 27 days after dose 1. This latter subject had a strong family history for arrhythmia. These events did not appear related to administration of the vaccine.

Even if all of these deaths could be explained away one way or another, this certainly sounds like a lot of deaths for such a young, and overwhelmingly female, group (16 of the 17 deaths were in females; one 15 year-old male Gardasil recipient died of a heart attack). What kind of death rate is normal for young women? The trials provide no such reference rate, but such statistics are readily available. Carnegie Mellon has a web-site called “Death Risk Rankings” (see HERE) that provides an interactive tool for calculating death rates within a wide range of demographic categories. For American females in the age range of the Gardasil trials (9-26 years of age), the rates are as follows: 2.75 per 10,000 in 10-19 year olds and 5.03 per 10,000 in 20-29 year olds. [Note: the majority of trial subjects were from the U.S. and Europe. European deaths rates from young women are 30% lower than American death rates, making this a conservative comparison].

Out of 11,778 Gardasil recipients, over 90% of them young women between the ages of 9 and 23, one would expect an annual death rate to be a mix of the rates for the two reference groups, or less than 4 per 10,000 in an entire year.  But in the trials, there were three “sudden deaths”, i.e. deaths that occurred within just the two weeks of the Gardasil injections, in a review period of less than forty-five days. That’s a death rate close to ten times higher than would be expected such a short period. And the overall Gardasil death rate of 8.5 per 10,000 (10 deaths out of 11,778) for the 12 month period of the trial is more than twice what one would expect. The FDA review evinced little concern over this high death rate, preferring instead to compare the deaths in the Gardasil group to that of the “placebo” group. But as one can see from the chart below, the death rate in the aluminum adjuvant group was higher than the reference groups as well.

 
In short, the CBER review of Gardasil condoned the use of an immunologically active placebo and not an inert solution. So instead of a adopting an increased measure of diligence in light of the potential for bias due to the DHHS conflict of interest in Gardasil, it appears that FDA permitted Merck to use a lower standard of safety. Only by using the single set of trial data in which the placebo solution was relatively (although not entirely) inert, can one assess the impact of this relaxed standard. Based on an analysis of this data, one is drawn to an inescapable conclusion.

Gardasil was not safe.

ACIP’s recommendation for universal vaccination of young females

If the FDA was less than diligent in its review of safety profile of the anti-cancer vaccine invented at its sister agency at NIH, how did its counterparts at CDC compare in terms of their own decision processes? Faced with a choice either to take a deliberate course, allowing a period of observation to follow Gardasil’s BLA approval, or to rush Gardasil into widespread use, CDC’s approach provides another standard of comparison for DHHS’s conduct. Would their key decision-making group, the Advisory Committee on Immunization Practices (ACIP) choose the deliberate or the hasty path?

There is very little ambiguity in this answer. ACIP wasted absolutely no time in recommending Gardasil for universal use among young women. Indeed, it would have been hard for them to move any faster. In June 2006, almost immediately after FDA approval, ACIP recommended the HPV vaccination. An account in the March 23, 2007 edition of the CDC publication, Morbidity and Mortality Weekly Report (MMWR) showed that ACIP was preparing for near instantaneous approval even before the FDA’s final reviews were completed.

The Advisory Committee on Immunization Practices (ACIP) HPV vaccine workgroup first met in February 2004 to begin reviewing data related to the quadrivalent HPV vaccine. The workgroup held monthly teleconferences and meetings three times a year to review published and unpublished data from the HPV vaccine clinical trials, including data on safety, immunogenicity, and efficacy. Data on epidemiology and natural history of HPV, vaccine acceptability, and sexual behavior in the United States also were reviewed. Several economic and cost effectiveness analyses were considered. Presentations on these topics were made to ACIP during meetings in June 2005, October 2005, and February 2006. Recommendation options were developed and discussed by the ACIP HPV vaccine workgroup. When evidence was lacking, the recommendations incorporated expert opinion of the workgroup members. Options being considered by the workgroup were presented to ACIP in February 2006. The final recommendations were presented to ACIP at the June 2006 ACIP meeting. After discussions, minor modifications were made and the recommendations were approved at the June 2006 meeting.

The date of the BLA approval for Gardasil was June 6, 2006. In the June 29 ACIP meeting, just 23 days after the FDA’s decision, ACIP gave Gardasil its formal support. The vote was unanimous, with two of the fifteen members abstaining due to their financial involvements with Merck.

But not only was the vote unanimous, the mood in the meeting was celebratory. Numerous vaccine safety advocates attended the June meeting due to their concerns over the influenza vaccines. One attendee recounted the reaction to the Gardasil decision. “After the vote the place erupted in applause. There was hand-shaking and back-slapping. It seemed kind of odd and inappropriate to us.” Asked why it seemed inappropriate, the observer explained that the concern arose, “because they were so clearly cheering the recommendation. It was clear and absolutely a celebratory reaction.”

The safety discussion was almost exclusively informed, of course, by the FDA’s flawed trial data. Not surprisingly, few concerns were raised. Here’s what the ACIP minutes had to say about Gardasil’s safety profile.

The clinical trial program places strong emphasis on evaluating the safety profile of GARDASIL®. Of ~21,464 subjects, ~11,000 received detailed safety follow-up and the remainder received serious adverse experiences in medical history and pregnancy follow-up. The incidence of overall adverse events (AEs), injection-site AEs, and low-grade fevers >100°F was slightly higher in the GARDASIL® group compared to the placebo group. Systemic AEs were comparable between the two groups. Serious AEs and discontinuation due to adverse experiences were extremely rare. [emphasis added]

As a result of this accelerated process, ACIP made a series of recommendations. The first placed Gardasil on the recommended list of childhood vaccines. “ACIP recommends routine vaccination of females 11-12 years of age, with three doses of the quadrivalent HPV vaccine. The vaccine series can be started as young as nine years of age at the discretion of the provider.” In the context of the clinical trials, this was an extremely aggressive recommendation. A scant 85 pre-pubescent girls, nine years of age or younger had received Gardasil in any trial (matched with only 48 controls) and only three percent of the Gardasil trail recipients were in the range of ACIP’s target population of 11-12 year olds. Not only was ACIP basing its recommendation on flawed safety analysis, it was extending its recommendations to include groups who lay outside of even this biased assessment.

But ACIP didn’t stop there. They also recommended a catch-up vaccine for all young women, even those who would have been sexually active for many years, who had contracted at least one strain of the virus, cleared it, and therefore received diminished benefit from vaccination. “ACIP recommends vaccination for females 13-26 years of age who have not been previously vaccinated. Ideally, vaccine should be administered before onset of sexual activity, but females who are sexually active should still be vaccinated.”  From a commercial perspective, this recommendation multiplied Gardasil’s profit potential for Merck and its NIH partners, creating a near term target market that was seven times larger than just the routine market of 11-12 year- olds.

There was little effective restriction placed on Gardasil’s market potential. And only the yeast protein in the carrier solution was cited as a safety concern. “Vaccination should be deferred until after moderate or severe acute illnesses improve,” read the ACIP recommendation. “A history of hypersensitivity or severe allergic reaction to yeast or any other vaccine component should be classified as a contraindication. Initiation of the vaccine series should be delayed until after completion of the pregnancy.”

The momentum for an aggressive roll-out was strong. Representatives from nine different organizations gave formal statements in support of Gardasil during the public comment period. In the meantime, and with a strong push from Merck, some state officials stood in line to move even faster than ACIP. In February of 2007, Republican Texas Gov. Rick Perry bypassed the legislature and mandated Gardasil for all 11- and 12-year-old girls in the state.

The commercial results were powerful and immediate. Merck reported its first revenues from Gardasil in the second quarter of 2006 (presumably from sales after the June BLA approval), and its revenues began to climb rapidly: $70 million in the third quarter, $155 million in the fourth quarter, all leading to a blockbuster year of 2007 in which Gardasil recorded revenues of $1.5 billion. The financial bonanza had begun in earnest.

A few years later, Merck would attract criticism for its aggressive marketing of Gardasil during this period. Sheila and David Rothman wrote a sharp critique in the Journal of the American Medical Association (hardly a radical hotbed of vaccine consumerism) in which they neglected the conflicts described here but noted the extreme measures that Merck adopted.

The marketing of this vaccine broke with traditional practices. Heretofore, vaccines had been identified by the disease they were preventing (measles, mumps) or by their creators (Salk or Sabin). This HPV vaccine followed a different model. It was identified by a trade name, Gardasil, and promoted primarily to “guard” not against HPV viruses or sexually transmitted diseases but against cervical cancer. The marketing campaign that followed, according to Merck’s chief executive officer, proceeded “flawlessly.” In 2006, Gardasil was named the pharmaceutical “brand of the year” for building “a market out of thin air.”

But the Rothmans’ critique would do little to delay or disrupt the launch. At the time of the DHHS rush to market there were few dissenting voices and none of them were heard at ACIP.

One lone voice stood out.  In March 2007, just as the vaccine was reaching its peak revenue numbers, one of the doctors who had guided the clinical trials voiced an objection. On March 14, 2007, an article in a small newspaper in Fort Wayne, Indiana reported on an interview with one of the scientists involved in the clinical trials. The scientist was named Diane Harper and she expressed dismay at the ACIP recommendation.

"Giving it to 11-year-olds is a great big public health experiment," said Diane M. Harper, who is a scientist, physician, professor and the director of the Gynecologic Cancer Prevention Research Group at the Norris Cotton Cancer Center at Dartmouth Medical School in New Hampshire.

"It is silly to mandate vaccination of 11- to 12-year-old girls There also is not enough evidence gathered on side effects to know that safety is not an issue."

Harper didn’t have much to gain from the commercial success of Gardasil, but she also was taking considerable risks by breaking ranks with her colleagues. One can only imagine how things would have been different if she had been in charge of the review process at FDA rather than running one branch of the clinical trial. In the midst of such a widespread degradation in regulatory ethics and standards, it’s interesting to consider why she made that choice.

The Fort Wayne reporter, Cindy Bevington, was frank as to why Harper was telling her story to them rather than a larger media outlet.  “For months, Harper said, she's been trying to convince major television and print media to listen to her and tell the facts about the usefulness and effectiveness of this vaccine.” Why was an inside critique of the Gardasil promotion campaign not already big news? "No one will print it," Harper said.

Over the coming months, the assessment of adverse effects of Gardasil would be transferred to a different group within DHHS, from CBER and ACIP to the “postlicensure safety surveillance” groups within FDA and CDC. At this point, the deaths and serious adverse events would leave the realm of closely held statistics within a vaccine manufacturer’s actively monitored trial sample and into the realm of passive surveillance in the general population; soon watchdog and vaccine safety groups like National Vaccine Information Center and Judicial Watch issued critical analyses, see HERE and HERE. And Harper’s concerns over the inadequate safety data would prove prophetic. Why had Harper broken ranks so early? When Bevington asked Harper why she was speaking out despite the momentum to the contrary, her answer was refreshingly simple.

“I want to be able to sleep with myself when I go to bed at night.”

In contrast to Jessica’s sudden death, and just a few weeks before the New York Post headline and article, another family went public with their daughter’s plight, in a blog named “Jenny’s Journey.” In their introductory blog post, Jennifer Tetlock’s parents relayed an urgent request. Their daughter wasn’t dead, but she was dying from what Jenny’s doctors had theorized was a rare degenerative neurological condition, an unusual form of early onset amyotrophic lateral sclerosis (or ALS, popularly known as Lou Gehrig’s disease). The bloggers, Barbara Mellers, Philip Tetlock, and Barbara Shapiro, were in no sense activists, and weren’t eager to join what they later termed the “anti-Gardasil movement.” What they wanted most of all was to find a way to save their daughter’s life. “One of the major things that would help her doctors figure out what to do”, they wrote on June 6, 2008, “is to find other people like Jenny (called "comparables")--people that share her medical condition and perhaps have had luck with certain treatments.”

The list of Gardasil victims who have gone public--parents of young women like Jessica Ericzon and Jennifer Tetlock--provides only a fragmentary view of the death toll associated with Gardasil. Many more deaths have been reported to the Vaccine Adverse Events Reporting System (VAERS) in cases where the family has chosen not to go public with their tragic loss.  Among the short list of publicized cases, most simply dropped dead like Jessica Ericzon within days of receiving a dose of the vaccine; these cases most closely resembled the three cases of sudden death from Gardasil reported during the clinical trials. Cases of clear “comparables” to Jenny Tetlock, young women who could satisfy Jenny’s parents’ quest, were less common. Nevertheless, there were a number of these publicly reported cases in which a Gardasil shot seemed to trigger a downward spiral of ill health-- encompassing a diverse range of symptoms--that would culminate in death (many of which came on suddenly as well).

The table below summarizes the connection between Gardasil and ten deaths that have been publicly associated with the vaccine. All of these stories have been reported elsewhere, most of them assembled in a memorial web-site called “The Truth About Gardasil.” You can find go to this web-site to read more about the stories of many of these young women (see HERE).

	Name	Date of First Vaccine	Date of Death	Reported Cause, Complications and Timing of Death
1	Santana Valdez	Dec, 2006?	Aug  31, 2007	Sudden death, with airway papillomatosis, less than 4 months after 3rd dose
2	Jenny Tetlock	Mar, 2007	Mar 15, 2009	Juvenile amyotrophic lateral sclerosis with onset after 1st dose
3	Brooke Petkevicius	Mar, 2007	Mar 26, 2007	Sudden death, with seizure and pulmonary embolism, 2 weeks after 1st dose
4	Jessica Ericzon	Jul, 2007	Feb 12, 2008	Sudden death, 1 day after 3rd dose after neurological symptoms
5	Jasmin Soriat	Sep, 2007	Oct 12, 2007	Sudden death, with “respiratory paralysis,” less than 1 month after 1st dose
6	Amber Kaufman	Mar, 2008	Apr 7, 2008	Sudden death, with seizure, “cardiac disturbance of undetermined etiology,” 1 week after 2nd dose
7	Christina Tarsell	Jun, 2008	Jun 23, 2008	Sudden death, 2 weeks after 3rd dose after symptoms of dizziness and fatigue
8	Moshella Roberts	Apr, 2008	Apr  5, 2008	Sudden death, 4 days after 1st dose
9	Megan Hild	May/Jun, 2008?	Nov 15, 2008	Sudden death, 2 months after 2nd dose after “severe headaches” and “severe stomach pain”
10	Jasmine Renata	Sept, 2008	Sept 21, 2009	Sudden death 6 months after 3rd dose after increasing cardiac and neurological symptoms

These public reports provide varying degrees of detail regarding cause of death. In one case (one where substantial detail has been reported), a young woman named Brooke Petkevicius who died suddenly after her first dose of Gardasil showed symptoms remarkably similar to a case report detailed in CBER’s clinical trial review. The February 6, 2008 edition of East Bay Express (see HERE), a San Francisco Bay Area newspaper, provided the following account.

In early 2007, as the pharmaceutical giant Merck began promoting its new vaccine Gardasil as protection against cervical cancer, Brooke Petkevicius was a nineteen-year-old freshman at UC Berkeley. She had seen the ads for the vaccine, and discussed getting it with her mother, whose gynecologist also had recommended it. On March 12, Brooke received the first of three doses. Two weeks later, she dressed to go running with a friend. As they reached the elevator, Brooke suddenly collapsed against the wall and had a seizure.

"She started shaking a lot," recalled the friend, Kristin Bietsch. "And her eyes went glazy a little bit." An ambulance rushed Petkevicius to the hospital, but doctors couldn't save her. Her autopsy indicated that she was killed by a pulmonary embolism, or blood clot, which had blocked the artery between her heart and lungs. "She had a whole bunch of little floating clots in her system," said her mother, Debra Sonner, recalling what doctors told her at the time.

Was Brooke’s death just a random coincidence or were there clues from the early safety reviews that anticipated her tragic death? One needn’t look far for comparable stories: the following account comes from one of the three sudden deaths after Gardasil reported in the clinical trials.

This 22 year old non-smoking white female subject had symptoms of leg pain prior to the vaccination (11/15/02), and was seeing a masseur for this complaint. She was also on hormonal contraceptives. The subject was vaccinated with her first dose of Gardasil on -------. On --------, Day 19 Postdose 1, the subject experienced suspected deep vein thrombophlebitis (DVT) of the left leg and consulted her own general practitioner. On --------, Day 20 Postdose 1, the subject experienced severe chest pain and was taken to the emergency room (ER). The subject subsequently experienced a suspected acute massive pulmonary embolism of severe intensity and was admitted to the intensive care unit (ICU). Echocardiography was performed and showed normal aorta and no thrombosis in the vena cava. Abdominal ultrasound was performed with no abnormal findings. On the same day, the subject died of acute massive pulmonary embolism and deep vein thrombosis of the left leg. The autopsy report confirmed the diagnosis of acute massive pulmonary embolism and deep thrombophlebitis of the left leg and also revealed an incidental finding of acute ischemic renal failure.

As this account suggests, the CBER review eventually explained away this death as a coincidence, an unfortunate side effect of taking birth control pills in a situation where the victim suffered from a pre-existing condition. So as soon as reports of similar deaths began entering the VAERS system, the CDC found ways to dismiss comparable cases such as Brooke Petkevicius by pointing a finger at birth-control pills as well. In a June 2007 report on the early deaths from Gardasil that were reported to the Vaccine Adverse Event Reporting System (VAERS), CDC dismissed Brooke’s death as yet another coincidence. “Preliminary data indicate that the two women [including Brooke], who died of blood clots were taking birth-control pills, and blood clots are a known risk associated with birth-control pills. All four deaths are being fully investigated but none appear to be caused by vaccination," claimed CDC. With regard to Gardasil, CDC wrote in its defense, "Since more than 5 million doses have been distributed, some deaths will occur coincidentally following vaccination (but not due to vaccination)."

Blaming the victim and citing coincidental death following vaccination are two well-known tactics that long ago became part of the DHHS playbook and CDC is not alone in deploying this tactic. In addition to the pulmonary embolism described above, the CBER review dismissed the two other cases of sudden death following Gardasil. One such case was a 15-year old boy who died of a heart attack less than a month after his first dose of Gardasil; yet CBER reported “the autopsy was inconclusive, but there was a strong family history of arrhythmia.” The other case was a 21 year-old woman who died with a convulsion four days after her third Gardasil dose; the CBER review again blamed this victim for her death, reporting that “this subject had a history of seizure disorder and anxiety. She suffered a seizure 4 days after dose 3, and was noted to have cocaine in her urine.”

Natural skepticism aside, making sense of individual cases like the ten public death reports and the three sudden deaths in the Gardasil trials is tricky business. Without intensive medical investigations one may never find definitive proof of harm from Gardasil. And, of course, the vast majority of Gardasil recipients have survived their vaccination series with no discernible lasting effects. There can be little dispute, however, that Merck has an enormous incentive to downplay obstacles to a profitable new product like Gardasil. And unfortunately, as we’ve seen in the clinical trial cases, FDA appears to have shared Merck’s bias, acting more like an equity participant in a DHHS “public-private partnership” than a conscientious guardian of the public trust; Its CBER reviewers effectively turned a blind eye to troubling signals as they granted Merck its BLA for Gardasil. Inevitably, however, a persuasive critique of vaccine safety monitoring for a blockbuster vaccine like Gardasil needs to move away from the realm of anecdote and into the realm of statistics. As we move beyond the review of individual cases, we’ll take a closer look at whether or not DHHS officials displayed notable biases in their analyses of Gardasil’s adverse effects in larger populations and how those analyses have been criticized by others.

Lack of diligence in postlicensure safety surveillance

Responsibility for what public health officials call “postlicensure safety surveillance” falls to a small set of DHHS departments. Two of these are the FDA’s Vaccine Safety Branch (VSB) and the CDC’s Immunization Safety Office (ISO). After CBER approves a vaccine and ACIP recommends it, the baton within DHHS passes next to VSB and ISO. In the passing of this baton, as stipulated previously, the presence of a conflict of interest does not mean that regulatory activity will necessarily reflect bias, negligence or lack of diligence on the part of the next group of regulators: Each department’s work deserves to be judged on its own merits. But in light of what appears to be a clear pattern of bias in the prelicensure activities of DHHS, it’s reasonable to approach an assessment of postlicensure activities with some skepticism. What, then, does the public record of postlicensure surveillance activity say about the presence of absence of bias and how VSB and ISO have done their jobs in assessing Gardasil’s safety?

The main public output of the FDA and CDC groups’ work has so far come in a single report published in the August 19, 2009 issue of the Journal of the American Medical Association (JAMA). In that paper, ISO’s Barbara Slade and four of her colleagues from CDC joined together with seven FDA colleagues to publish the first-ever analysis of the VAERS data on Gardasil. Not surprisingly, the JAMA paper gave Gardasil a free pass; in the process the authors joined the chorus of DHHS celebration for the breakthrough of its home-grown anti-cancer vaccine. “Vaccination with [Gardasil] has the potential to decrease the global morbidity and mortality of HPV-associated diseases, including cervical cancer. After hepatitis B vaccine, which can prevent liver cancer, [Gardasil] is only the second vaccine licensed with an indication to prevent cancer.” And although they acknowledged the possibility of injury due to blood clots like those that killed Brooke Petkevicius, Slade et al argued that the data from VAERS led to the same conclusion as the positive review from their colleagues at CBER. “The postlicensure safety profile presented here is broadly consistent with safety data from prelicensure trials.”

As for the specific question of Gardasil deaths, Slade et al acknowledged that there had been deaths associated with Gardasil. But they dismissed the VAERS death reports as not frequent enough to worry about.

Causes of death included 4 unexplained deaths, 2 cases of diabetic ketoacidosis (1 complicated by pulmonary embolism), 1 case related to prescription drug abuse, 1 case of juvenile amyotropic lateral sclerosis, 1 case of meningoencephalitis (Neisseria meningitidis serogroup B), 1 case of influenza B viral sepsis, 3 cases of pulmonary embolism (1 associated with hyperviscosity due to diabetic ketoacidosis), 6 cardiac-related deaths (4 arrhythmias and 2 cases of myocarditis), and 2 cases due to idiopathic seizure disorder. The PRR [statistics speak for “proportional reporting ratio”, or the VAERS death rate relative to the “background” expected death rate] for deaths in 6- to 17-yearolds was 1.4 (Χ2=0.42, P=.52). The PRR for deaths in 8- to 29-year-olds was 1.2 (Χ2=0.01, P=.92). Neither of these met the screening criteria for signal detection. [emphasis added]

Criticism of the JAMA analysis came quickly (some preceded the paper’s publication) and from several different quarters. One of the most trenchant attacked the relevance of the VAERS case data itself and came from the parents of a Gardasil victim. Like so many parents of vaccine injured children, Jennifer Tetlock’s parents had become deeply dissatisfied by the diligence of federal officials in evaluating Gardasil’s safety. Jennifer’s adverse reaction, as originally diagnosed, was among the least common adverse events; Slade et al reported it as an isolated case. But in April 2009, Jenny’s parents publicly voiced their suspicion that their doctor’s original diagnosis of juvenile amyotrophic lateral sclerosis may have been misleadingly narrow. They argued that “world-class immunologists suspect that Jenny had a potentially treatable autoimmune disorder mimicking ALS, possibly triggered by the Gardasil vaccination.” If true, this interpretation of her reaction placed Jenny’s death in a broader category of a severe autoimmune reactions from Gardasil. As reported previously in part 2, this kind of autoimmune reaction was a risk that the clinical trials showed to be quite common.

More broadly, Jenny’s parents struck out at CDC’s failure of postlicensure diligence.
 
The CDC does not inspire confidence, so we conducted our own shoestring search to determine whether Jenny was alone. We created a website (jenjensfamilyblogspot.com). Although this website has only drawn 40,000 visitors, it has out-performed the federal government in finding girls ominously similar to Jenny (current score is: Jenny site 2; CDC’s VAERS: 0).

One does not need to be a statistician to see how unlikely it is that these two other girls are the only cases out there—or how frightening it is that we already know of three documented cases of girls (those two plus Jenny) who developed ALS within several months after their vaccinations. After all, if the odds of ALS in teenaged girls are 1 in 3 million and we found 3 in only 40,000, it is very possible that many other of the 6 million girls vaccinated have already developed severe neurological collapse, like Jenny.

Jenny’s parents may have been uncertain about the proper diagnosis of Jenny’s reaction, but they were not alone in their criticism of the statistics underlying the FDA/CDC analysis. In a December 2009 letter to JAMA, Drs. Vicky Debold (Full disclosure: Debold is a director of the Age of Autism sponsor SafeMinds) and Eric Hurwitz criticized the Slade analysis. Debold and Hurwitz identified numerous flaws in the CDC/FDA report, arguing that: a) cases of autoimmune diseases such as Guillain-Barré syndrome were systematically underreported; b) the method for obtaining background rates of disease could include vaccine injury and were thus inappropriate to use in comparisons; and c) the denominator for the case population used to generate the PRRs was grossly overstated. On the last point, Debold and Hurwitz noted that Slade et al had mistakenly used “total vaccine doses distributed” as the denominator for disease rates instead of doses  administered (or for that matter the number of women receiving doses), a choice that “systematically inflates the ratio’s denominator.”

Debold and Hurwitz noted the larger policy problem created by these failures in postlicensure surveillance and by Slade et al’s low standard of diligence. “Federal officials have cited this study as evidence that [Gardasil] ‘is a safe and effective vaccine,’” they noted. “However, we consider that conclusion to be unwarranted because the study draws inferences from data likely to be systematically biased.”

Diane Harper, one of the researchers involved in the Gardasil trials, agreed with Debold and Hurwitz. In an August 2009 interview with CBS News reporter Sharyl Attkisson, she also criticized Slade et al, arguing to Attkisson that “the risks of vaccination are underreported in Slade's article, as they are based on a denominator of doses distributed from Merck's warehouse. Up to a third of those doses may be in refrigerators waiting to be dispensed as the autumn onslaught of vaccine messages is sent home to parents the first day of school. Should the denominator in Dr. Slade's work be adjusted to account for this, and then divided by three for the number of women who would receive all three doses, the incidence rate of serious adverse events increases up to five fold" [emphasis added].

Harper also agreed with Jenny Tetlock’s parents’ suggestion that Slade et al were understating the number of deaths in their risk assessment, the numerator.  Harper told Attkisson, "Parents and women must know that deaths occurred. Not all deaths that have been reported were represented in Dr. Slade's work, one-third of the death reports were unavailable to the CDC, leaving the parents of the deceased teenagers in despair that the CDC is ignoring the very rare but real occurrences that need not have happened if parents were given information stating that there are real, but small risks of death surrounding the administration of Gardasil."

So how should one weigh the net impact of all these claims and criticisms? Certainly, although there are serious health risks, including death, associated with Gardasil, there are also potential benefits, including reduced rates of cervical cancer. Slade’s analysis doesn’t provide the number of individuals receiving vaccine in the first two years after the Merck vaccine’s introduction, but based on the statistics she provides, it’s reasonable to assume that something less than 10 million young women received doses of Gardasil (23 million doses distributed with a target of 3 doses per subject). A balanced review of the overall benefits of Gardasil would compare the frequency and consequences of vaccine injury in this population, based on the most objective rate of adverse events, with the vaccine’s purported benefits.

If we compare the death rate (these rates were calculated in part 2 of this series) in the Gardasil trial group of 8.5 per 10,000 to an expected death rate in young women of 3.9 per 10,000, we get an extra death risk from Gardasil of roughly 4-5 per 10,000 annually. Applied to a population of 4 million young women per year, that would come to a total of close to 2,000 extra deaths per year that were caused by Gardasil during the period of its launch. This is an extraordinarily high rate and may possibly be explained away by bad luck and coincidences, as the CBER review contended. But it is the only active surveillance analysis ever done on a population exposed to Gardasil.  By contrast, the calculated death rate from VAERS reports gives a much lower number, over 100 times lower, a ratio that many consumer advocates (and certainly Jennifer Tetlock’s parents) believe reflects more on the poor quality of the VAERS database than the actual risk of death from Gardasil.

But if Gardasil is successful in preventing deaths from cervical cancer, is it possible that society is still coming out ahead overall? The America Cancer Society estimates that there were 4,000 deaths in 2009 from cervical cancer. If Gardasil could prevent a significant number of these deaths in the future, then it’s possible that the benefits of the vaccine might exceed the risks. Unfortunately for the vaccine program, the deaths caused by Gardasil are immediate and the preventable deaths from cervical cancer are many years away.

Unfortunately for the public, in assessing the reliability of any estimates of Gardasil’s future benefits there are many unknowns: we don’t know how long the immunity against HPV will last; we don’t know whether costly booster shots and a new round of adverse events will be necessary; and, most important of all, we don’t yet have any idea whether or not Gardasil will succeed in preventing a meaningful number of cervical cancer deaths. We know it provides effective immunity against two of the most common cancer-causing strains of HPV. But it’s entirely possible that, just as soon Gardasil suppresses these strains of HPV, new cancerous strains will arise to take their place leaving overall cervical cancer rates unchanged. The dirty little secret of the war on cervical cancer is that public health officials have no current way to judge how and whether widespread vaccination with Gardasil will affect the actual rate of cervical cancer and they won’t be able to make that judgment for many years.

A similar situation holds in the case of a similarly multi-strain infection and its associated vaccine formulation: in this case the many different strains of streptococcus pneumoniae, the bacterial species responsible for invasive pneumoccoal disease (or IPD) and the so-called pneumococcal vaccine. In order to combat IPD, a leading cause of bacterial meningitis, a multi-strain vaccine called Prevnar was introduced by Wyeth in 2000. Prevnar quickly became one of the most commercially successful vaccines of all time, bringing in nearly $3 billion in revenue in 2008. But in asking whether Gardasil will actually work against cervical cancer, it’s useful to ask whether a similar vaccine product like Prevnar has been successful in actually preventing IPD?

A recent study from Massachusetts on its effectiveness suggests that Prevnar, contrary to all expectations, did not reduce the incidence of IPD in the state. Instead, although the forms of IPD caused by strains in the vaccine went down after Prevnar’s introduction, IPD cases caused by other strains (some of which were even more dangerous than the original strains) rose almost immediately and in opposite proportions, to keep the rate of IPD in Massachusetts constant. In plain language, Prevnar, the most commercially valuable vaccine in history, created no health benefits whatsoever. It just didn’t work.

Will this be the case with Gardasil? We will only know with any certainty what Gardasil’s benefits will be after many years, even decades of use. Right now, it’s too early to tell how many of the 4,000 annual deaths from cervical cancer might be prevented. But several things are clear in the near term: Gardasil has already  injured an unknown number of young women, with these injuries likely including deaths that may number in the hundreds, possibly (based on the only data available) as many as 2,000 per year; Gardasil also has left many more young women with crippling chronic conditions like Guillain-Barré syndrome, arthritis and autoimmune thyroid conditions.

Diane Harper has also argued that assessments of the net benefits of Gardasil have overstated its value. She told Sharyl Attkisson in her interview that, "the risks of serious adverse events including death reported after Gardasil use in the JAMA article by CDC’s Dr. Barbara Slade were 3.4/100,000 doses distributed.” This rate is substantially lower than our estimates here. Nevertheless, Harper remained concerned that Gardasil’s risks outweighed its benefits. “The rate of serious adverse events is on par with the death rate of cervical cancer. Gardasil has been associated with at least as many serious adverse events as there are deaths from cervical cancer developing each year.”

In light of the many years required to prevent cases of cervical cancer, it’s clear that the current cost of Gardasil outweighs its current benefit. Moreover, it may take many years to realize any net benefit to society from Gardasil; the crossover point where realized benefits exceed the costs incurred is far away and uncertain. At the moment, therefore, there is only one net benefit that is certain: the benefit to Merck’s bottom line. And, of course, the bottom line of Merck’s business partner at NIH.

Revolving door culture

DHHS has a clear conflict of interest with respect to Gardasil at the institutional level since it shares directly in Gardasil’s profits. We’ve also seen now that this conflict of interest is echoed by (and possibly sustains) a pervasive pattern of regulatory bias in favor of Gardasil during multiple stages of the decision process. But as the Gardasil body count rises, one natural question one might ask is why, at a personal level, more DHHS officials haven’t taken the principled stand of Diane Harper, who spoke up against the ACIP Gardasil recommendation because “I want to be able to sleep with myself when I go to bed at night.”

In the case of several senior officials involved in overseeing key DHHS decisions during the Gardasil era, some portion of that answer is provided by their subsequent career moves. Indeed, these moves reveal a cultural problem that is in many ways more troubling than the direct Gardasil financial conflicts: a pervasive pattern of senior officials cashing in on their careers in public service in order to obtain lucrative corporate and consulting jobs. The career moves of these senior officials show that a virtual revolving door between regulators and the pharmaceutical, vaccine and biologics companies they are supposed to regulate erodes any meaningful sense in which these officials truly serve consumer interests, especially when it comes to product safety. This revolving door provides the cultural foundation that undergirds some of the more egregious institutional conflicts.

A short account of the recent careers of just a few of the officials involved in regulating Gardasil shows this revolving door in action.

• Mike Leavitt was named on December 13, 2004 as Secretary of DHHS, where he subsequently was responsible most of the critical regulatory decisions involving Gardasil. In January 2009, he left HHS and formed Leavitt Partners, a Washington DC consulting firm that helps its client “enter new markets, enhance the value of their products and navigate dynamic regulatory and reimbursement systems.” In his consulting work, Leavitt could certainly teach his clients about Gardasil and how they could follow Merck’s example in forging a model “public-private partnership.”

• Julie Gerberding was named Director of CDC on July 3, 2002 and served in that role until she resigned on January 29, 2009. Gerberding watched over Gardasil policy at CDC during the period of FDA review in which ACIP put Gardasil on a fast track for approval in June 2006. She also was in charge of the oversight for CDC’s postlicensure safety activities during much of the period leading up to Slade et al’s JAMA submission when portions of the VAERS analysis were reviewed with ACIP. Less than a year after leaving public service, on December 21, 2009, Merck announced Gerberding’s appointment as President of the Merck Vaccine Division, effective January 25, 2010, the minimum interval allowed for a Federal official to assume a position at a company they used to regulate. Gerberding, who once regulated Gardasil, is now directly responsible for its growth and profitability.

• Karen Goldenthal was the Director of the Division of Vaccines and Related Products Applications within CBER, the FDA division responsible for approving Gardasil’s BLA in June 2006. In 2007, shortly after Gardasil’s approval, Goldenthal left CBER to become Executive Director of PharmaNet Consulting. PharmaNet is “a global, drug development services company, provides a comprehensive range of services to the pharmaceutical, biotechnology, generic drug, and medical device industries.” Other FDA executives have taken leadership positions at PharmaNet, including William Egan, former head of Vaccine Research and Review at FDA, now a Vice President in PharmaNet’s consulting practice.

These departures provide just a few small examples of a pervasive exodus of FDA officials, many of whom leave FDA in order to provide advice to pharmaceutical companies on how to make their way successfully through the pre- and post-licensure processes. And when it comes to vaccines, there is a specific market for former CBER officials to coach vaccine manufacturers on how to get their BLAs approved and their launches more profitable. Like PharmaNet, a consulting company called the Biologics Consulting Group (BCG) shows how active the revolving door between FDA and industry has become. Here’s how BCG describes itself.

Biologics Consulting Group, Inc. (BCG) is a team of consultants who provide national and international regulatory and product development advice on the development and commercial production of biological, drug and device products. Our staff consists of experts in regulatory affairs, product manufacturing and testing, pharmacology/toxicology, facility inspections, statistics, program management, and clinical trial design and evaluation. Many of our consultants are former CBER, CDER, and CDRH reviewers. [emphasis added]

In an environment so stepped in both direct and indirect conflicts of interest, is it any wonder that Gardasil regulators have leaned so steeply in favor of industry while overlooking serious safety concerns?

The early victims of Gardasil look for justice

The final phase of regulatory activities surrounding Gardasil have yet to play themselves out. These involve the process of adjudicating claims of injury and death due to the vaccine. The DHHS agency responsible for this work is the Health Resources Services Agency (HRSA), which houses the Division of Vaccine Injury Compensation, the group responsible for managing the Vaccine Injury Compensation Program (VICP), more commonly known as “vaccine court.” In light of the pattern of bias we’ve observed in the approach other DHHS agencies have taken to Gardasil, one might reasonably question the prospects for fair treatment of Gardasil victims in vaccine court. Can we really expect the director of HRSA to encourage a fair and generous compensation policy on Gardasil when her colleagues over at NIH are profiting from the patent license, CDC is actively promoting its use and her former colleagues at FDA are providing consulting services to companies helping them avoid regulatory pitfalls and keep their profits intact?

As in the case of its sister agencies, the presence of conflict of interest does not necessarily mean that HRSA will demonstrate bias, negligence or failures of diligence in their approach to Gardasil. In advance of any record of decisions, however, it’s simply too early to tell how HRSA will respond. In other controversial areas such as the Autism Omnibus Proceeding, petitioners have been deeply disappointed in their treatment at the hands of the vaccine court. And it seems likely that Gardasil families are destined for their own day in vaccine court: HRSA provides “table injuries” that provide compensation for a short list of outcomes on a few vaccines, but there are no table injuries yet specified for Gardasil. (HRSA has commissioned the Institute of Medicine to develop such a list for the entire category of HPV vaccines. A candidate list of injuries can be found in the latest working list from the “Committee to Review Adverse Events of Vaccines”, see HERE ) So as Gardasil petitioners find their way into the VICP process, HRSA officials will be setting Gardasil injury compensation policy for the first time.

And the Gardasil girls are coming to seek justice. In an early action, on March 13, 2010, the parents of Jennifer Tetlock filed the following petition with the VICP.

The above captioned Petitioners request compensation under the National Vaccine Injury Compensation Program, 42 U.S.C. 300aa-10 et seq. (Supp. 1996), for the death of minor, Jennifer Tetlock who received the third series of the Gardasil vaccination on March 1, 2007 from James Cuthbertson, M.D. in Berkeley, California and thereafter suffered from atypical amyotrophic lateral sclerosis (ALS)-like lower motor neuron disease which was caused in fact by the above stated vaccination.

Will Jennifer Tetlock receive justice? It’s hard to imagine how any agency so inextricably linked to the Gardasil program--from invention to approval and recommendation to protection and profit—can possibly be trusted to be a fair arbiter of guilt and innocence.

But one can always hope. After all, the guardians of vaccine safety in DHHS have children themselves. And like Diane Harper, they also need to sleep with themselves when they go to bed at night.