Скачать 0.59 Mb.
DEPARTMENT OF HEALTH AND HUMAN SERVICES
CENTERS FOR DISEASE CONTROL AND PREVENTION
(cdc logo image)Centers For Disease Control and Prevention
Advisory Committee on Immunization Practices
February 21-22, 2007
Record of the Proceedings
February 21, 2007
WELCOME AND INTRODUCTION
HEPATITIS A POSTEXPOSURE PROPHYLAXIS
Introduction: Dr. Bell.
Study Results: Dr. Victor
General Update: Dr. Parashar
Disease Burden: Dr. Payne
VAERS: Ms. Haber
Data Interpretation: Dr. Patel
THIMEROSAL: REVIEWING THE EVIDENCE
Update: Dr. Allos
Information: Dr. Fiore
VFC Resolution: Dr. Calugar
Pandemic Vaccine Prioritization: Dr. Schwartz
FluMist Safety Data: Dr. Walker
February 22, 2007
DIPHTHERIA, TETANUS, PERTUSSIS, Hib, PENTACEL
IMMUNIZATION SAFETY OFFICE SURVEILLANCE UPDATES
HUMAN PAPILLOMAVIRUS VACCINE
Introduction: Dr. Gilsdorf
Quadrivalent HPV Vaccine: Dr. Barr
Bivalent HPV Vaccine: Dr. Dubin
Working Group Update: Dr. Markowitz
DEPARTMENT OF HEALTH AND HUMAN SERVICES
CENTERS FOR DISEASE CONTROL AND PREVENTION
Advisory Committee on Immunization Practices
February 21-22, 2007
Minutes of the Meeting
The Department of Health and Human Services (HHS) and the Centers for Disease Control and Prevention (CDC) National Center for Immunization and Respiratory Diseases (NCIRD) [proposed] convened a meeting of the Advisory Committee on Immunization Practices (ACIP). The meeting was held on February 21-22, 2007 at CDC’s Global Communications Center, Building 19, in Atlanta, Georgia. The list of participants is appended to the minutes as Attachment 1. [Note: the list of participants only includes persons who introduced themselves for the record, presented, made public comments, or registered prior to the meeting.]
At 8:11, Dr. Abramson welcomed everyone to the February 2007 ACIP meeting. Dr. Pickering introduced several international visitors, including Dr. Eibhlin Connolly and Dr. Jim Kiely from the Department of Health and Children in Dublin, Ireland, as well as Dr. Keiko Taya, Infectious Diseases Surveillance Center in Wakayama City, Japan, and Dr. Takehiro Togashi from Sapporo City University, Japan. He also introduced new representatives of ACIP liaison organizations. Dr. Tamara Lewis, American Health Insurance Plans replaces Dr. Andrea Gelzer. Dr. Stanley Grogg represents a new liaison organization: the American Osteopathic Association. Dr. Vesta Richardson from the National Immunization Council of Child Health Program in Mexico replaces Dr. Romeo Rodriquez. Dr. Harry Keyserling represents the Society for Healthcare Epidemiology of America. Dr. Geoff Evans (HRSA) was unable to attend the meeting, so Dr. Jevaji attended on Dr. Evans’ behalf. Dr. David Kimberlin from the American Academy of Pediatrics was also unable to attend. Dr. Patricia Whitley-Williams from the National Medical Association also sent her regrets.
Dr. Pickering noted that slide presentations would be posted on the ACIP Web site one to two weeks after the meeting. The ACIP home page is located at www.cdc.gov/nip/acip and is updated at frequent intervals to include meeting agendas, meeting minutes and presentations, ACIP recommendations, and other information related to immunization and ACIP activity.
Dr. Pickering confirmed that there was a quorum of ACIP members present and explained that the ACIP charter gives the Executive Secretary or his or her designee the authority to temporarily designate the ex officio members as voting members if necessary. This would occur in the event that there are fewer than eight ACIP members available or who cannot vote because of conflicts of interest. The ex officio members will be formally requested to vote when necessary. If this occurs, they will be asked to express and disclose any potential conflicts of interest. The goal in appointing members to the ACIP is to achieve the greatest level of expertise while minimizing the potential for actual or perceived conflicts of interest. The conflict of interest provisions state that members agree to forego participation in certain activities related to vaccines during their tenure on the committee. CDC has issued limited conflict of interest waivers for certain other interests that potentially enhance a member's expertise while serving on the committee. Members who conduct vaccine clinical trials or serve on data safety monitoring boards may serve as consultants to present to the committee on matters related to those vaccines. However, they are prohibited from participating in the deliberations or votes of the committee on issues related to those vaccines. Regarding other vaccines of an affected company, a member may participate in all discussions with the proviso that he or she abstains on all votes related to the vaccines of that company. All ACIP members must state their conflicts when they vote for a VFC resolution. ACIP members who may have a potential financial conflict of interest should make it known by disclosing all of their vaccine-related financial interests and work.
Dr. Abramson asked ACIP members to state any conflicts of interest. Dr. Carol Baker stated she had a conflict with Novartis Vaccines. Dr. Janet Gilsdorf stated she was an independent safety monitor on an NIH-sponsored influenza trial, for which she receives no compensation. Dr. John Treanor said his group was involved in clinical trails of influenza or pneumococcal vaccines for sanofi, Protein Sciences, Wyeth, Novartis and Merck. Dr. Lieberman stated that he was involved in clinical trials with Merck and MedImmune. Dr. Abramson stated that he served on a Data Safety Monitoring Board for an NIH-sponsored study of the use of oseltamivir in infants, and he received no compensation for this study.
HEPATITIS A POSTEXPOSURE PROPHYLAXIS
Dr. Beth Bell, National Center for HIV, Hepatitis, STD and TB Prevention (NCHHSTP), CDC
Dr. John Victor, PhD., Department of Epidemiology, School of Public Health, University of Michigan
Dr. Bell explained the current ACIP recommendation with respect to hepatitis A post-exposure prophylaxis calls for a single dose of IG as soon as possible but within two weeks of exposure; if hepatitis A vaccine also is recommended, it can be administered simultaneously with IG. The statement does address the question of the use of hepatitis A vaccine alone, saying results of an appropriately designed clinical trial comparing the post-exposure efficacy of vaccine with that of IG are needed to determine if hepatitis A vaccine without IG can be recommended. With this question in mind, about five years ago a trial was funded to compare the efficacy of hepatitis A vaccine to IG after exposure to hepatitis A virus, through a cooperative agreement with the University of Michigan’s School of Public Health.
Dr. John Victor presented the results of that study. As background, he showed post-exposure efficacy estimates for IG from the published literature. Most of the studies were done before the advent of serologic tests. For the most part, these efficacy estimates were quite high and support the notion that IG is at least 85 percent effective for post-exposure prophylaxis.
Unlike most pre-exposure trials, in a post-exposure study potential participants must be ascertained within a relatively short time period after exposure, which is logistically difficult. Identification of cases and their contacts for potential post-exposure prophylaxis requires a site with relatively high endemicity and high rates of hepatitis A.
The trial was conducted from October 2002 to April 2005 in Almaty, Kazakhstan, a very densely populated city of about 1.4 million people. Major hepatitis A outbreaks occur during the fall and winter periods and typically involve large numbers of young children. Over 95 percent of recognized cases of hepatitis A are hospitalized.
The study objective was to compare the efficacies of hepatitis A vaccine and IG in the prevention of laboratory-confirmed, symptomatic hepatitis A, when given within 14 days of exposure to a symptomatic index case. Eligible participants were household and day-care contacts of index cases identified from surveillance, ranging from 2 to 40 years of age. They had to be exposed to an index case within two weeks after index-case symptom onset, with no reported history of hepatitis A in the past and no receipt of hepatitis A vaccine or IG within the previous six months. Also, they had no reported medical diagnosis of chronic liver disease and no contraindications to either of the study interventions. The study interventions were hepatitis A vaccine, VAQTA, at the age-appropriate, licensed dose for pre-exposure protection and U.S.-manufactured IG at the standard post-exposure dust. Both interventions were administered intramuscularly in the deltoid in a scheme that maximized participant blinding.
Within households and day-care groups, contacts were randomized at a 1-to-1 ratio to receive either intervention. Contacts were blinded to the intervention, but physicians administering the interventions obviously could not be blinded because of the different sorts of dosing. Therefore, physicians who had not administered the intervention, conducted the follow up and they were blinded.
Susceptible contacts were followed weekly for eight weeks. At four and eight weeks post exposure, visits were made to interview contacts about hepatitis A-related symptoms and to collect further blood samples for testing. These visits coincided with the average and the end of the incubation period for hepatitis A. If there was serologic or biochemical evidence of infection or if a contact reported illness, a separate illness visit was triggered. Primary endpoints included illnesses occurring among contacts that met the following three criteria: positive for IgM anti-HAV; ALT level at least twice the upper limit of normal during an episode of illness with no other obvious cause; and one or more of the listed signs or symptoms of hepatitis A.
Because the purpose of the study was to determine if hepatitis A vaccine is at least equivalent to IG, a noninferiority hypothesis was tested, stated as follows: Among those initially seronegative contacts who receive each intervention within 14 days of exposure to an index case of hepatitis A, the proportions of laboratory-confirmed symptomatic hepatitis A with onset between 15 and 56 days post exposure will be similar in the two intervention groups.
Similarity was defined as follows: assuming equivalence, one would expect to find a relative risk of 1.0, but one cannot test a statistical hypothesis and prove the null in the traditional sense. Instead, the statistical hypotheses were restated in terms of confidence-interval bounds around the point estimate. A null hypothesis that defined vaccine as substantially worse than IG was specified and rejection of the null would mean the two interventions were equivalent. “Substantially worse” was defined as a one-sided 95 percent confidence interval upper bound of the relative risk of 3.0. In order to reject the null and conclude similarity, the confidence interval bound on the observed relative risk was required to be less than 3.0. This 3.0 margin was selected in a pretrial meeting of experts in the clinical, epidemiological, and laboratory aspects of hepatitis A and in statistics and vaccine field-trial design, based on what was felt to be both clinically relevant and statistically valid. Assuming that IG efficacy would be 90 percent, the confidence interval upper bound of 3.0 translates into a vaccine efficacy of at least 70 percent, so the vaccine efficacy estimate lower bound would be 70 percent. Although this confidence-interval bound defines the minimum vaccine efficacy, the point estimate of vaccine efficacy in this circumstance is greater than 84 percent.
In exploratory secondary analyses, probable hepatitis A was examined, defined as any symptom plus serologic evidence of infection and biochemical or virologic evidence of infection, through PCR testing. The subset of cases that were icteric was examined, as well as subclinical illnesses that were confirmed by serology and by biochemical or virologic criteria but were asymptomatic.
The results of the trial were as follows. During 29 months of enrollment, there were 4,524 enrolled household and day-care contacts of 920 index cases of hepatitis A; 2,272 were randomized to vaccine and 2,252 received IG. Because about two-thirds were immune to hepatitis A from previous exposure in their lifetime, this left 740 susceptible persons who received vaccine and 674 who received IG. In this modified intent-to-treat data set, a little over half of contacts were female and about 85 percent were household contacts. Since over 95 percent of primary endpoints and 98 percent of secondary endpoints observed in the study were among household contacts, there was actually very little day-care transmission. Both the index cases of contacts and the contacts themselves were on average relatively young, in early adolescence. Most contacts were immunized late in the two-week window of post-exposure prophylaxis. Over 86 percent received one of the study interventions in the second week post exposure.
During follow up, 29 vaccine and 22 IG recipients were found with an illness confirmed by serology and ALT elevation. The independent data monitoring committee determined that 26 vaccine and 18 IG cases actually met the criteria for primary endpoints. Most outcomes occurred during the second week of the post-exposure period. Illnesses occurred on average a few days earlier than the reported average incubation period for hepatitis A of 28 days, on days 24 and 25. There seemed to be a truncation of the incubation period since the latest case seen was at day 33 post exposure. The average age of cases among those who received vaccine was a bit lower than among IG recipients, though not statistically significant in the crude analysis. The peak ALT elevations measured at the time of illness among vaccine recipients were a bit higher than among IG recipients. Rates of PCR positivity, icteric illness, and frequency of gastrointestinal symptoms were similar in both groups.
In order to assure maximum robustness, the primary endpoint analysis was done on a per-protocol basis; for a non-inferiority study, it is important to remove all possible misclassification, since that biases toward the null. So 172 vaccine and 150 IG recipients were eliminated, leaving 568 vaccine recipients and 522 IG recipients. In the per-protocol data set, contacts had characteristics similar to those in the intent-to-treat analysis.
For the primary endpoint, 25 cases of laboratory-confirmed hepatitis A occurred among vaccine recipients and 17 occurred among IG recipients, yielding a relative risk among vaccine versus IG of 1.35. That is the point estimate for the relative risk; the one-sided 95 percent confidence interval upper bound of the relative risk was 2.4, which was well within the pre-specified margin of 3.0. The difference of the observed risks for the two groups was only 1.1 percent. So although vaccine might appear to perform slightly less well than IG, even though it met the criteria, the results show only a miniscule difference in the risk of hepatitis A in the two groups, especially compared to not treating at all.
For the secondary endpoint of icteric illness, a similar point estimate was found for the relative risk; even though the confidence interval upper bound is slightly higher, there were fewer cases that were icteric. Looking at all clinical and subclinical infections together, both the confidence interval and the point estimate are less than a relative risk of 2.0. Most cases were among children in the study, so that is where most of the inference was made. However, no major differences appear in estimates for adults.
Dr. Victor then put the relative-risk results in context and described implications for vaccine efficacy. The assumption of 90 percent IG efficacy implied an underlying secondary attack rate of 33 percent in the study population. If this assumption was correct, the study results would translate into a point estimate of vaccine efficacy of 86 percent and the lower bound of the 95 percent confidence interval would have been 76 percent. Therefore one can be quite confident that vaccine efficacy is high in this situation. Using a slightly lower secondary attack rate, the estimate of vaccine efficacy is still very good.
In summary, the efficacy of vaccine post exposure appears quite high and similar to that of IG. The risk of hepatitis A for vaccine recipients was never more than 1.5 percent greater than the risk for IG recipients for any of the endpoints looked at. There is some evidence that IG may be attenuating clinical illness based on the characteristics of the cases seen. There was no evidence that vaccine given in the second week after exposure resulted in lower clinical protection. Finally, contacts in households experienced the highest transmission rates in the study.
Dr. Bell provided some additional data and talked about potential implications. The question before the committee has to do with using hepatitis A vaccine alone post exposure. There are a number of potential benefits of being able to use vaccine, including long-term protection, ease of administration, acceptability, and availability. In addition, there is currently only one U.S. supplier of immune globulin and the cost has risen considerably over the last five or ten years, making it similar to that of hepatitis A vaccine. A single adult dose of IG now is about $20, a pediatric vaccine dose under government contract is about $12, and an adult dose is about $19. Another benefit of being able use hepatitis A vaccine is that it brings U.S. practice in line with many other countries that recommend vaccine as post-exposure prophylaxis.
Only one other clinical trial used hepatitis A vaccine post exposure and that was conducted in Italy in 1997. This study enrolled 212 household contacts, aged 1 to 40 years, of hospitalized hepatitis A cases. Contacts were randomized to receive either the GSK vaccine within eight days of symptom onset of the index case or no intervention. The outcome was IgM positivity as measured either at Day 14 or Day 45 post vaccination or post no intervention. The trial was stopped when the study results reached statistical significance. The estimated vaccine efficacy was 79 percent, with a very wide confidence interval. The outcomes identified included two in the vaccinated group, ages 10 and 11 years, both of whom were asymptomatic with normal ALTs. Since vaccination can induce IgM, it is possible that these two so-called vaccine failures, in fact, did not have hepatitis A. In the no-intervention group, there were 12 outcomes, aged 3 to 25 years, the majority with symptomatic hepatitis A and elevated ALTs. This study was of limited usefulness in the U.S., where there was already an efficacious intervention in IG, but it did provide good evidence that the vaccine did a reasonable job of preventing hepatitis A.
Unfortunately, results of the current study and previous information cannot answer all questions about how hepatitis A vaccine performs post exposure. Eligibility was restricted to age 40 and under in the Kazakhstan study. Most cases and study participants were fairly young, which is not surprising, given that essentially all adults older than 40 are immune to hepatitis A, as are many younger adults. Patients who reported a diagnosis of chronic liver disease were excluded, based on concern about the increased severity of hepatitis A in those people. People with other medical conditions were not explicitly excluded, but in general, the study population was young and healthy.
Only one of the two U.S.-licensed vaccines was used, so there is the question of the generalizability of the findings to both licensed vaccines. There is also the question of time since exposure. Going into the study, many might have thought that vaccine would not perform as well in the second week after exposure. Any recommendation the committee might make about the ability to use vaccine would represent off-label use because the manufacturers do not intend to apply to FDA for this as an indication. Finally, it is likely that currently available data are all that will be available. Additional studies addressing these areas with limited data are logistically unfeasible.
Dr. Bell shared some technical information about the antibody to hepatitis A virus. The minimum protective antibody concentration is unknown, and in fact, it may be below the assay-detection limit. For example, after IG administration, it is known that protection continues after antibody is no longer detectable. In vaccine immunogenicity studies, protection has been defined as the lower limit of detection of whatever assay was being used. This can vary from 10 to 33 mIU per ml, which complicates the precision with which one can look at the available pre-exposure immunogenicity data.
With that caveat, Dr. Bell presented a summary of data available from the pre-exposure studies about seroconversion after one dose of hepatitis A vaccine, which seem to be most relevant to how hepatitis A vaccine might perform post exposure. Ninety-five percent or more of children seroconvert by four weeks after the first dose. However, many of these studies were conducted using old formulations of the vaccine, which was a three-dose series, with the first dose being half of the currently licensed dose. That makes it very difficult to study response after one dose of what is used now. Thus there are no published data among children that report seroconversion at two weeks. The GSK package insert reports that 92 to 96 percent of children have seroconverted at two weeks, citing unpublished data.
Essentially all adults (over 18 years old) seroconvert by four weeks. Using the currently licensed formulations, one or two published studies and information in the package inserts quote anywhere from 70 percent to 100 percent seroconversion at two weeks after one dose. Only one published study was found that directly compares the response after a single dose of vaccine in younger versus older adults, and that study used the currently licensed formulation of the GSK vaccine. At Day 15, the study reported 90 percent seroconversion among younger adults versus 77 percent among older adults, and by one month, 97 percent were seropositive in both groups.
Data from the National Health and Nutrition Examination Survey conducted in 1988 to 1994 address the likelihood of adults and older adults in the United States being susceptible to hepatitis A. About 50 percent of 40- to 49-year-olds are immune, rising to 60 to 65 percent in older adults and reaching over 70 percent in adults age 70 or older. It is reasonable to conclude that the majority of adults remain susceptible to hepatitis A in the U.S.
The next issue addressed was the response to hepatitis A vaccine among people with medical conditions and immunocompromised people. The data are fairly limited with respect to response after one dose of vaccine and there are no data regarding response at two weeks; all the data are at four weeks. Among HIV-infected patients, the three relevant studies report a wide range of seroconversion after one dose, ranging from 10 to 78 percent. The percent positive is higher with a higher CD4 cell count, but in several studies a much lower than the usual 95 percent of adults seroconvert after one dose. Hepatitis A vaccination is recommended for patients with chronic liver disease, so there has been a fair amount of interest in studying the immunogenicity of the vaccine in these patients. In the three published studies, the percent positive at four weeks ranges from 63 to 93 percent. In the one study that used a control group, the percent positive was lower than controls. It appears that the response rate is lowest in patients with decompensated cirrhosis. Finally, a few studies in liver and kidney transplantation patients show a much lower percent positive at four weeks.
Dr. Bell mentioned two other considerations. The first was time since exposure. The Italian study cut off at seven days post exposure. The current study appears to remove theoretical concern about diminished vaccine efficacy with longer time since exposure; there was no difference between the groups in time since exposure, and most interventions were given fairly late in this two-week post-exposure limit. A second question is the comparability of the two U.S.-licensed vaccines. They are considered equivalent for pre-exposure use but there have been very few head-to-head comparisons. Because of problems with the different assays, it is difficult to make precise comparisons but the percent seroconverting appears to be similar.
Many other industrialized countries that recommend post-exposure prophylaxis have recommendations for using vaccine. In Canada, since 2000 the recommendation has been that vaccine without IG is preferred during the first 7 days after exposure, with a recommendation of IG for infants and immunocompromised persons. Apparently IG is more difficult to get in Canada than in the U.S. In the U.K., since 2001 the recommendation has been to use vaccine if the exposure has been within the previous seven days, based on the Italian study, and to continue to use IG if the exposure has occurred more than seven days previously and for people older than 50 years, cirrhotics, and people with chronic viral hepatitis. During 2000-2003, a survey was conducted of European countries with respect to viral hepatitis. France, Italy, and Belgium all reported using vaccine only for post-exposure prophylaxis. Only Sweden and perhaps Norway reported using only IG for post-exposure prophylaxis. The majority of the other countries used vaccine or vaccine and IG.
Dr. Bell then described the current epidemiology of hepatitis A in the United States. The provisional total number of cases in 2006 was about 3300, compared to 20,000 - 35,000 hepatitis A per year during the past several decades. Although communitywide outbreaks do occur not anymore, vaccine has been recommended in that context for the past decade. The common setting for IG use as post-exposure prophylaxis currently is among household and other close personal contacts. Outbreaks in childcare centers used to often require IG, but now these outbreaks are rare.
It is difficult to quantify exposure from an infected food handler, but surveillance data indicate that 3 to 7 percent of reported hepatitis A cases are food handlers. A study funded a few years ago in a couple of states found that about 5 percent of food handlers worked while they were infectious and were felt to pose a transmission risk. In the context of contact notifications, an average of 350 IG doses per episode were administered. Health departments do not systematically keep track of how much IG has been used such settings, but given the low number of cases these days, it is potentially on the order of tens of thousands of doses and not likely more than 100,000.
In summary, vaccine offers a number of advantages over IG, and the flexibility to use vaccine in some circumstances would be beneficial. The available data suggest that vaccine is efficacious post exposure, but not all populations were studied. The relevance of pre-exposure immunogenicity data is unclear and they do not suggest large differences in responses of children and healthy adults to vaccine. Some adults might not respond as briskly, but the clinical significance of this finding is quite unclear. It does appear that there are suboptimal responses at least in some immunocompromised persons. Since additional data are not likely to be forthcoming, the committee needs to balance the practical public health implementation considerations against the limitations of the available information.
|1 department of health and human services||Department of health and human services|
|Department of health and human services||Department of Health and Human Services|
|Department of health and human services||U. S. Department of health & human services|
|U. S. Department of health and human services||U. S. Department of health and human services|
|Department of health and human services||Department of health and human services|