Widespread Use of Varicella Vaccine Does Not Reduce Immunity to Zoster of Others

(See the Major Article by Carryn et al, on pages 413–21.)

For as long as there have been vaccines, there have been “anti-vaxxers.” The two seem to go together. It all began in 1796, when William Jenner inoculated a child with cowpox virus, which subsequently protected the boy from getting smallpox. The idea of inoculating people with cowpox, or as it is more properly called, the vaccinia virus, caught fire because the procedure was miraculously successful in preventing illness and death from smallpox. Jenner coined the name “vaccination” because the virus originated from a cow, for which the Latin word is vacca. Jenner’s vaccine became popular despite some ridiculous adverse publicity, such as colorful drawings of cow parts growing out of the bodies of individuals who had been vaccinated. Eventually the practice became standard in the world, and the anti-vaxxers had to give up on it. Smallpox was the first (and only) human viral infection to be eliminated from the world by the judicious use of a vaccine.

Almost immediately after a live attenuated varicella vaccine (Oka strain) was developed in the 1970s late mid-20th century by Takahashi, it was shown to be able to prevent severe or fatal varicella in a few children who had survived leukemia [1, 2]. At that time, nothing was more disturbing than seeing a child who had amazingly survived leukemia, which had previously been uniformly fatal, die of a disease as mundane as chickenpox. For some, it was considered close to miraculous that a virus could be attenuated and still maintain its ability to stimulate an immune response in an immunocompromised child. For others, it was met with criticism that the vaccine was potentially dangerous because varicella zoster virus (VZV) is a herpesvirus that establishes a lifelong latent infection after primary infection. Other objections to the varicella vaccine were that the immunity might not be long lasting, and that the elimination of chickenpox epidemics in children might deprive adults of needed immune boosting through exposure and lead to an increase in the incidence of zoster [3]. The efficacy of the varicella vaccine to prevent severe varicella in children with leukemia in remission had been shown by the Collaborative Varicella Vaccine Study Group [4]. The “anti-vaxxers” argued that because of the demonstrated efficacy of the varicella vaccine, the universal vaccination of children would cause chickenpox to disappear. This apparent success would mean that exogenous exposure to VZV would cease because humans are the only reservoir of this virus. The argument supposed that children needed to keep getting chickenpox to protect adults from the presumed scourge of waning immunity manifesting in outbreaks of zoster. Computational models based on this fear predicted epidemics of zoster in countries that adopted universal vaccination against varicella. This fear was taken seriously in many countries, such as the United Kingdom, where, even today, varicella vaccine is not licensed for use in healthy children. Every year, as a result, a small number of otherwise healthy children in the United Kingdom die of varicella despite the availability of antiviral drugs.

The incidence of zoster actually is increasing in many countries, including the United States (US), and the reasons for this increase are not clear. In the US, an increase in the incidence of zoster started in the 1950s, long before the development of the varicella vaccine and the adoption of its universal use to prevent chickenpox. The increase is undoubtedly multifactorial and probably related to improved diagnosis, an increase in the elderly population, and increasing numbers of immunocompromised individuals in our population. One additional factor may be stress, which is known to affect immune function negatively. Fortunately, successful vaccines have been developed for zoster that are very effective and safe for use, both in old people and immunocompromised individuals [5, 6].

Zoster is due to reactivation of latent VZV. Because VZV becomes latent in ganglionic neurons after either an episode of varicella or VZV vaccination, it can reactivate in immunized children. VZV can also reactivate silently, that is, without provoking a rash or other clinical manifestations [5]. Zoster without rash can also present as gastrointestinal problems, some of which are amenable to antiviral therapy [7, 8].

Because today most children in the US are immunized against varicella with the live attenuated vaccine, chickenpox has become rare in the US population. Interestingly, the incidence of zoster is actually lower in vaccinated children than in those who have experienced clinical varicella [9]. If exposure to varicella were important in protection from zoster, one would not predict zoster to be unusual in vaccinees. Interestingly, when zoster does occur in vaccinees, it is caused by wild-type VZV in about half of the cases [9]. These vaccinees might have experienced subclinical varicella prior to their vaccination, enabling wild-type VZV to establish latency before the introduction of the live attenuated vaccine virus. Alternatively, vaccination might protect an individual exposed to wild-type VZV against clinical disease, but not against infection; therefore, wild-type VZV might be able to replace latent vaccine virus and later reactivate to cause zoster.

Hope-Simpson hypothesized in the 1960s that after an episode of varicella, VZV can reactivate either with or without clinical manifestations [10]. If reactivation occurs, it will boost immunity to VZV that theoretically can protect against both varicella and zoster. This presumed phenomenon is termed “endogenous boosting.” Periodic reactivation of VZV in vaccinees may thus, in part, account for the long-term protection against VZV that is currently being observed in vaccinated populations.

Exposure to wild-type VZV does not seem to be required to maintain immunity either to varicella or to zoster. A study in France indicated that the incidence of zoster was similar in cloistered communities, who were not exposed to children, and communities living in towns with frequent exposure to children. Zoster was thus not more common in nuns and monks than it was in the outside world [11]. Several studies that have examined exogenous boosting have failed to find evidence supporting this concept or, if they have found it, the boosting is present only briefly [12–14].

Mathematical models have made valuable predictions about the future; however, models are only as good as the presumptions on which they are based. New evidence that was not included in a model may cause them to be proven wrong, and their predictions must be tested as time passes and more data are collected. Brisson and colleagues predicted, on the basis of computational modeling, that universal vaccination against varicella would cause severe increases in the incidence of zoster in adults [15]. This model does not take the phenomenon of “endogenous boosting” into account, and assumes that exogenous exposure to VZV shed by children with varicella is essential to maintain anti-VZV immunity and suppression of VZV reactivation and zoster. To date, no evidence has emerged to back the assumptions underlying the predicted epidemic of zoster, nor has the predicted outbreak of zoster occurred.

In contrast to epidemiologic evidence, it is very difficult to obtain accurate immunologic data on patients to examine the phenomenon of immunity to VZV. In the current issue of The Journal of Infectious Diseases, Carryn et al provide an astonishingly clever study of the VZV immunological data obtained from participants in a double-blind, placebo-controlled trial of a new recombinant zoster vaccine that was carried out to determine whether this vaccine was safe and effective in preventing zoster in older individuals. The zoster vaccine that had been studied was composed of VZV glycoprotein E and an adjuvant to increase immune responses. It was observed to have a startling high efficacy in preventing zoster in the elderly. In the vaccine study, blood had been obtained routinely, before and after vaccination, to test for antibodies and cell-mediated immunity to VZV in vaccine recipients and placebo controls.

The study had been carried out in 18 different countries, some of which used varicella vaccine whereas others did not. In those countries where varicella vaccine was widely used or universal, the incidence of clinical varicella was rare. No differences were found in antibody titers or cell-mediated immune responses to VZV in the aged, however, in the countries where varicella vaccine was nearly universal and varicella was rare, and those in countries where varicella remained a common childhood illness. These numerical data argue strongly against Brisson et al’s assumptions.

At this time in the US, it is the responsibility of the medical profession to fight “anti-vaxxers” in every way possible. Vaccines may not be perfect, but their employment is always preferable to enabling a serious or fatal infection to rage in a population. The current abundance of misleading, erroneous, or distorted information widely available on the internet makes it essential for healthcare professionals to present patients with the truth about vaccines. It is obvious that it will be much less expensive for patients to get a vaccine than to pay for hospitalization, intubation, and possible death from a vaccine-preventable disease. In doing so, the vaccine policies of every country should be updated to take account of new data, such as those of the current Carryn et al report.

Notes

Financial support. The authors are supported by the National Institutes of Health (grant numbers DK093094 and NS15547).

Potential conflicts of interest. All authors: No reported conflicts of interest.

All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.