25 Years of Varicella Vaccination in the United States

Varicella is an illness with a characteristic maculopapular, vesicular rash that, during the prevaccine era, infected almost everyone in the population during childhood. The causative agent is varicella-zoster virus (VZV), an alphaherpesvirus. After the primary varicella infection, VZV establishes latency in sensory ganglia and may reactivate years or decades later to cause herpes zoster (HZ). Most children with varicella recover without sequelae and historically the disease was generally regarded as benign. However, varicella can lead to serious complications and deaths in healthy as well as immunocompromised persons [1–3]. Additionally, varicella during the first 20 weeks of pregnancy can lead to a collection of severe birth defects and complications for the fetus manifested as congenital varicella syndrome [4].

The recognition of the severity potential of varicella emerged in the late 1960s when effective new treatments for children with leukemia evolved to include systemic steroids, chemotherapeutics, and radiation, which greatly improved survival, only to have many of these immunocompromised children develop severe and even fatal varicella for which there was no effective therapy. What was long considered a relatively mild rite of passage in childhood became newly recognized as a serious, potentially lethal illness in susceptible immunocompromised hosts. Around this time in the United States, varicella-zoster immune globulin for passive immunization was developed to prevent severe varicella in immunocompromised children, particularly for those with leukemia [5]. This approach was successful but only when a close exposure to varicella had been recognized and within a limited interval after exposure (96 hours).

Dr Michiaki Takahashi shocked the scientific world when, in 1974, reported that he had developed a live attenuated vaccine to prevent varicella [6]. Most immediate reactions to this information were unfavorable in the western world, including in the United States. The idea of developing a live varicella vaccine had seemed both remote and potentially dangerous because of the risk for latency of the vaccine virus and later reactivation to cause HZ. Additionally, there was concern that vaccine-induced immunity might be only transient, so that, theoretically, vaccinated children might lose immunity to VZV and develop varicella as adults when varicella was more severe or could complicate pregnancy. If the fear of varicella in leukemic and other immunocompromised children had not been so dire, it is possible that the varicella vaccine would not have been developed.

In Japan, Takahashi and his group demonstrated that the vaccine was safe, well tolerated, and protected both healthy children and adults from developing varicella after close exposures to VZV in households or in hospital settings [6–9]. The development group also safely vaccinated small numbers of children with leukemia in remission and children with nephrotic syndrome and demonstrated that these children were protected from developing varicella after exposure to VZV [6, 7, 10, 11]. In the 1970s, when these studies were completed, the only serologic test available for measuring antibodies was complement fixation, which was not sensitive enough to measure immunity after vaccination, so there was no ability to demonstrate a serological correlate of protection.

Following the experience with the vaccine in Japan, in February 1979, a meeting was organized at the National Institutes of Health (NIH) in the United States to discuss “experimental herpesvirus vaccines.” Takahashi’s live attenuated varicella vaccine was featured; the data were reported by Takahashi himself. In fact, no other successful vaccines were discussed at that meeting. Even today, we have yet to successfully develop vaccines for other herpesviruses. Experienced virologists from across the United States attended the meeting, which resulted in a recommendation to develop and study the varicella vaccine in the United States. A competitive NIH process resulted in the establishment of the Collaborative Varicella Vaccine Study Group; over the next 5 years, this group showed that live attenuated varicella vaccine could be safely given to children with leukemia in remission [12, 13]. Vaccination resulted in development of antibodies that were correlated with protection against household exposures to varicella (antibodies were measured by fluorescent antibody to membrane antigen [FAMA] test, validated to indicate protective immunity to varicella); 85% of exposed children did not develop varicella although they were exposed to siblings with the illness. None of these vaccine recipients (>500) had problems with HZ during the study [13, 14]. The publication of these data by mid-1980s helped to pave the way to study varicella vaccine in healthy children in the United States [15]. Clinical trials using varying doses of varicella vaccine were conducted in the United States and Europe during the 1980s [16, 17].

As other causes of infectious disease burden in children in the United States were controlled through vaccination (eg, measles, diphtheria, poliomyelitis), varicella assumed more importance as a preventable cause of mortality and morbidity. Anticipating licensure of varicella vaccine in the United States, a heightened effort was made in the early 1990s to more fully document varicella epidemiology in the US general population. Varicella is highly infectious with secondary attack rates between 61% and 100% among susceptible household contacts [18]. The disease, therefore, affected nearly all persons in the United States by young adulthood with an estimated 4 million varicella cases (average incidence 15.0/1000 population/year) occurring annually [19]. The highest incidence was in preschool and early elementary school aged children [19, 20]. Seroprevalence studies confirmed high VZV-specific immunoglobulin G (IgG) antibody in the United States population among children aged 6–11 years (86%) rising progressively to ≥99% among persons aged 30 years or older [21]. Annually, varicella cases resulted in 10 500–13 500 hospitalizations (4–5 hospitalizations/100 000 population) during 1988–1995 [2, 22, 23]. Similarly, during 1990–1994, an average of 145 death certificates per year listed varicella as either underlying or contributing cause of death (0.6 deaths/1 million population) [3]. Prevaccine, approximately 44 cases of congenital varicella syndrome were estimated to have occurred per year. The disease burden from varicella was highest in children, who accounted for more than 90% of the cases, two-thirds of the hospitalizations, and about half of the deaths. In addition, an important burden for society was the work loss from parents to care for their children sick with varicella [24].

Based on a documented continuing disease and societal burden, the United States was the first country to implement a routine varicella vaccination program following licensure of the live, attenuated varicella vaccine (VARIVAX, Merck & Co, Inc) in 1995. Policy recommendations from the American Academy of Pediatrics (1995) and the Advisory Committee on Immunization Practices (1996) focused on routine vaccination of children with 1 dose at age 12–18 months with catch-up vaccination of susceptible children before age 13 years and vaccination with 2 doses of susceptible adults in designated high-risk groups (health care personnel and family contacts of immunocompromised persons) [25, 26]. In 1999, the policy was updated to include vaccination requirements for childcare and school entry, postexposure prophylaxis, and all susceptible adults at increased risk for exposure or transmission [27]. There were concerns raised before vaccine implementation that a universal childhood varicella vaccination program could shift the varicella burden to older ages and result in more severe disease and infections during pregnancy [28, 29]. Another concern was that declines in varicella cases and circulation of VZV would reduce exogenous boosting and produce a rapid increase in HZ in adolescents and adults [30].

Impact of the vaccine on varicella was evident in all age groups within 5 years of implementation [31] and as the 1-dose program attained high vaccine coverage among children aged 19–35 months—reaching 85% in 2003—there were substantial reductions in incidence, hospitalizations, impatient and outpatient medical expenditures, and deaths [32–34]. Postlicensure monitoring demonstrated that the vaccine was safe [35, 36] and highly effective at preventing moderate and severe disease (median vaccine effectiveness 97%) [37]. Despite progress in controlling varicella in the United States, the moderate vaccine effectiveness (approximately 82%) for prevention of varicella of any severity afforded by 1 vaccine dose [38, 39] did not provide sufficient herd immunity to interrupt community transmission of VZV, especially in settings with high contact rates such as schools [40, 41]. Although severe cases were reduced by approximately 90%, the decline in the number of varicella cases reached a plateau between 2003 and 2006 and outbreaks continued to be reported, albeit fewer and smaller than in the prevaccine era [32]. These outbreaks, which proved difficult to control because most students were already vaccinated according to existing recommendations, placed a financial and resource burden on state and local health departments and were disruptive for schools and parents. Varicella in vaccinated children, termed breakthrough varicella, usually has a modified presentation with fewer skin lesions (<50) that are more commonly maculopapular than vesicular [42]. These characteristics posed clinical and diagnostic challenges and stimulated the development of new techniques for specimen collection from nonvesicular lesions and for testing to improve laboratory confirmation [43]. Although usually mild, breakthrough cases are infectious [44] and because of challenges in diagnosis, patients are not always isolated, creating opportunities for transmission of the virus.

To further improve varicella control in the United States, vaccine policy was revised in 2007 to a routine 2-dose schedule in childhood (12–15 months and 4–6 years) with catch-up vaccination for children who had received 1 dose [45]. Licensure of the measles, mumps, rubella, and varicella vaccine (MMRV) in 2005 simplified implementation of the second dose policy. A second dose of varicella vaccine among children produced an improved humoral and cellular immune response that correlated with improved protection against disease [38, 46]. Additionally, antepartum assessment of evidence of immunity in women of childbearing age and postpartum vaccination, as appropriate, were recommended to prevent severe consequences of varicella infection during pregnancy, including infection of the fetus.

Expansion of the vaccination program to include a second dose led to further declines in incidence, outbreaks, hospitalizations, and deaths. The decline in reported cases was steepest in the age groups recommended for the second dose [47, 48]. The improved varicella control with decreased community transmission and risk of exposure provides added protection to groups not protected directly by vaccination. Twenty-five years after the introduction of the varicella vaccine in the United States, disease has declined overall by >97%; declines have occurred in all age groups and have been greatest (99%) among persons <20 years of age born during the vaccination program in whom severe disease outcomes have been nearly eliminated. An additional benefit of the varicella vaccination program, not fully appreciated initially, is the impact on HZ. Data show a lower risk for HZ among both healthy and immunocompromised children (approximately 80% lower risk among healthy vaccinated children vs unvaccinated children) [14, 49]. At the population level, this is reflected in stepwise declines in HZ incidence as the age groups became dominated by persons born during the varicella vaccination program and, therefore, vaccinated [50, 51]. Importantly, an increase in HZ among adults attributable to the varicella vaccination program, as predicted by some modelers, was not observed, and moreover, the pattern of decline in HZ incidence in children provides tentative reassurance that, over time, rates of HZ will decline as vaccinated children age [50, 51].

In this supplement of the Journal of Infectious Diseases, we present the United States experience with the varicella vaccination program, summarizing the successes and challenges in implementing and monitoring the program over 25 years, including establishing varicella surveillance, monitoring varicella and HZ epidemiology, monitoring vaccine coverage, safety and effectiveness, clinical and laboratory issues, and economic impact. Ongoing tracking of varicella and HZ, especially severe outcomes, and duration of immunity will be essential as the program continues, as well as developing more sensitive laboratory assays to measure vaccine-induced immunity and improve diagnosis of the mild, modified breakthrough disease. Since 1995, the United States varicella vaccination program has led to prevention of more than 91 million varicella cases, 238 000 hospitalizations, and almost 2000 deaths, with an impressive return on investment with net societal savings of more than $23 billion [52]. It is a public health achievement made possible by the efforts of thousands of nurses, physicians, pharmacists, and public health staff who implemented this immunization program. We dedicate this supplement to them and to the memory of Dr Takahashi who bravely pioneered the development of varicella-zoster virus vaccines.

Notes

Disclaimer. The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention.

Financial support. No financial support was received for this work.

Supplement sponsorship . This supplement is sponsored by the Centers for Disease Control and Prevention, Atlanta, GA, USA.