Varicella is a self-limiting and relatively mild disease of childhood, although it is frequently more severe and complicated among the immunocompromised rheumatology patients on immunomodulator therapies. Once logged in to your MyAccess profile, you will be able to access your institution’s subscription for 90 days from any location. In spite of rapid natural healing, those receiving adenine arabinoside over the first five days had accelerated clearance of virus from vesicles (P = 0.01), and cessation of new vesicle formation (P = 0.004), and a shorter time to total pustulation (P = 0.001). 10, Rm. Histologic sections of the lacrimal gland biopsy specimen documented the presence of a necrotizing dacryoadenitis, with a dense infiltrate composed of neutrophils; histiocytes; small, cytologically bland lymphocytes;and lesser numbers of plasma cells and eosinophils. A vaccine to prevent herpes zoster (HZ) in adults ⩾60 years of age with healthy immune systems was recently approved by the US Food and Drug Administration. This vaccine is contraindicated in persons with certain immunodeficiency states or who are receiving immunosuppressive therapy.
On the basis of studies of the varicella vaccine in healthy and immunosuppressed children and studies of HZ vaccine in healthy adults before its licensure, a series of strategies are proposed for evaluating the live HZ vaccine in immunosuppressed persons. In addition, the use of other vaccines, including heat-inactivated or replication-defective varicellazoster virus to prevent HZ in immunocompromised persons, is also discussed. Financial support: Division of Intramural Research, National Institute of Allergy and Infectious Diseases. These patients are more likely to develop disseminated HZ or multidermatomal disease. Dissemination to various organs, including the lung, liver, brain, and spinal cord, can occur. Patients with advanced HIV infection may develop recurrent or relapsing HZ, as well as verrucous lesions that persist for months . Prevention of varicella and HZ in immunocompromised patients would reduce the morbidity of these diseases.
Varicella-zoster virus (VZV) immunoglobulin and acyclovir are available for postexposure prophylaxis to prevent varicella in immunocompromised persons exposed to persons with varicella. A safe and effective vaccine for immunocompromised persons could prevent much of the morbidity associated with HZ. Vaccination with the live attenuated Oka virus is used to prevent disease in healthy persons who are exposed to varicella. Varicella vaccine has also been safely given to selected children with leukemia  or HIV infection [4, 5] and to liver or intestinal transplant recipients [6, 7]. Current recommendations  state that persons with impaired humoral immunity can be vaccinated, and vaccination should be considered for HIV-infected children with CD4 T cell percentages ⩾15%. Vaccination of children with leukemia whose disease is in remission and who have not received chemotherapy for at least 3 months should be considered only when antiviral therapy and expert guidance is available.Varicella vaccine is contraindicated in patients receiving highdose immunosuppressive therapy and those receiving ⩾2 mg/kg (or a total of ⩾20 mg/kg/day) prednisone when given for ⩾2 weeks. Combination measles-mumps-rubella-varicella vaccine is contraindicated in persons with primary or acquired immunodeficiency.
The Shingles Prevention Study (SPS)  showed that vaccination of older healthy adults with a high-potency live varicella vaccine reduced the burden of illness due to HZ, the incidence of postherpetic neuralgia, and the incidence of HZ. Although the vaccine was less effective in reducing the incidence of HZ in persons ⩾70 years of age than in those 60–69 years of age, the vaccine was more effective in reducing the severity of illness in the older subjects. The dose of vaccine used was ∼14 times that of the varicella vaccine used in the United States; a larger dose of vaccine has been associated with a longer duration of immunity in VZV-immune elderly persons . Persons who received the vaccine had more reactions at the injection site, including erythema, pain, swelling, or pruritus, than did those who received placebo, but there was no increase in the rate of serious adverse events with the vaccine. Although rashes did occur in persons who were vaccinated, all of the rashes that were analyzed were due to wild-type, not vaccine, virus. A number of differences between the varicella and HZ vaccines are important to note when considering vaccination of immunocompromised persons to prevent HZ (Appendix A). The higher dose of the HZ vaccine might be associated with a higher likelihood of adverse effects in mildly or moderately immunocompromised persons.
However, because virtually all persons receiving vaccine to prevent HZ would already have been infected with VZV and should have some memory T cells, these persons might have fewer rashes than do persons who were never infected with the virus. Studies of leukemic children receiving 2 doses of varicella vaccine showed that rashes, some of which contained vaccine virus, were almost always observed after the first dose of vaccine . Studies of healthy adolescents and adults receiving 2 doses of the varicella vaccine also showed a much lower rate of rash after the second dose of vaccine . The induction of immunity with VZV vaccination might be more likely in persons with preexisting memory T cells to VZV than in persons receiving the vaccine for the first time. Finally, although varicella vaccine has been used in immunocompromised persons (see above), there are no published studies of the HZ vaccine in immunocompromised persons. Live varicella vaccine. The Shingles Prevention Study  excluded patients who were immunosuppressed as a result of malignancy, HIV infection, immunosuppressive or cytotoxic chemotherapy (e.g., cancer chemotherapy or treatment for organ transplant recipients), or corticosteroid therapy (⩾800 µg/day of beclomethasone dipropionate or its equivalent).
Such patients with impaired T cell immunity are thought to be at greater risk for adverse effects from the vaccine and less likely to respond to the vaccine. Patients with skin cancer or other neoplasms that were stable in the absence of chemotherapy were not excluded. Optimally, the vaccine could be given to patients who are not yet immunocompromised but who will be given immunosuppressive therapy in the next several weeks to months. Such patients might be undergoing organ transplantation or have a recent diagnosis of a connective tissue disorder and would receive immunosuppressive therapy in the near future. The live Oka vaccine virus might be tested in selected patients who have impaired cellular immunity and were not included in the Shingles Prevention Study. The varicella vaccine is considered for HIV-infected children with CD4 T cell percentages of ⩾15% . In addition, a recent study showed that the varicella vaccine was well tolerated and often induced VZV-specific immune responses in HIV-infected children with CD4 T cell percentages ⩾15% and CD4 T cell counts ⩾200 cells/µL .
Therefore, asymptomatic or mildly symptomatic adults with HIV infection at risk for HZ who have CD4 T cell percentages ⩾15% and CD4 T cell counts ⩾200 cells/µL might be vaccinated with the HZ vaccine in controlled studies. Because the vaccine was not evaluated in persons receiving moderate doses of corticosteroids or other moderately immunosuppressive therapy, such patients might also be evaluated in future studies. However, because the dose of vaccine given is ∼14 times the dose of the varicella vaccine, there might be a higher rate of adverse effects than that seen with the varicella vaccine. The use of a live attenuated varicella vaccine to prevent HZ would be contraindicated in persons who have moderately to severely impaired cellular immunity who might develop symptomatic, progressive infection with vaccine virus. Vaccination of moderately or severely immunocompromised patients with live vaccine should be performed in carefully monitored clinical trials in which both the safety and immunogenicity of the vaccine are observed (Appendix B). These studies should include analysis of VZV-specific cellular immunity before, during, and after vaccination, as well as close attention to adverse effects, especially the development of rashes after vaccination. Suspicious rashes should be tested by polymerase chain reaction for detection of vaccine virus, and patients whose rashes contain VZV should be treated with antiviral therapy and followed closely.
Vaccination with the live virus vaccine might be less hazardous in immunocompromised persons with detectable cell-mediated immunity to VZV. A number of studies have examined VZV-specific immune responses to live VZV vaccination of healthy older subjects [18, 19]. Several of these studies showed a boost of virus-specific cellular immune responses with live virus vaccine. Responder-cell frequencies, which measure proliferation of serially diluted peripheral blood mononuclear cells (PBMCs) in response to VZV antigen, have been useful for assaying the cellular immune response to the virus. Persons with detectable levels of responder cells before vaccination were 4–6 times as likely to respond to the vaccine at 3 months, compared with those without detectable responder-cell frequencies before vaccination . Other tests of cellular immunity, such as lymphocyte proliferation assays, assessment of production of cytokines by PBMCs, or skin tests in response to VZV antigens have also be used . Inactivated varicella vaccine.
Safer vaccines would involve the use of a heat-killed virus vaccine or subunit vaccines. These nonreplicating vaccines might be less effective because they are less likely to present antigens in the context of major histocompatibility complex (MHC) class I and, therefore, might stimulate lower levels of virus-specific CD8 T cell responses than do live vaccines. An early study showed that healthy VZV-seropositive adults who received live or heat-inactivated VZV vaccine developed similar titers of virus-specific antibody responses at 6 weeks . Another study compared vaccination of 80 healthy persons >55 years of age with a single dose of 4000 pfu of live VZV vaccine versus a similar dose of heat-killed vaccine . Both viruses induced similar levels of VZV antibodies, virusspecific T cells, and production of interferon (IFN)-γ by PBMCs stimulated with VZV antigen at both 3 months and 1 year after vaccination. Persons who had greater responder-cell frequencies to VZV before vaccination had the highest responder-cell frequencies after vaccination. Compared with the killed virus vaccine, the live VZV vaccine induced higher levels of MHC class I cytotoxic T cells but similar levels of NK cell-dependent lysis at 3 months after vaccination .
A follow-up study using the same dose of live and heat-killed vaccine in 167 healthy older adults (mean age, 66 years) showed that both vaccines boosted VZV antibodies and VZV responder-cell frequencies at 3 months, but the level of VZV antibodies and IFN-γ production by PBMCs returned to baseline at 1 year, whereas the responder-cell frequency was still elevated in both groups at 1 year . The half-life of the boost in virus-specific responder cells was 17.5 months after vaccination with live virus and 21.3 months with inactivated virus, but the difference was not significant. Redman et al.  randomized autologous or allogeneic bone marrow transplant recipients to receive a heat-inactivated varicella vaccine or placebo. Fourteen patients received 1 dose of heat-inactivated vaccine 1 month after bone marrow transplantation, and 14 received placebo. PBMCs were stimulated with VZV antigen, and tritiated thymidine uptake was measured to determine the stimulation index. Although the stimulation index was higher in patients who received the inactivated vaccine than in those who received placebo (12.2 vs.
4.8) at 3 months after transplantation, there was no effect on the incidence of HZ in vaccine recipients compared with control subjects (38% vs. 36%). In a subsequent study, 24 patients received 3 doses of heat-inactivated vaccine at 1, 2, and 3 months after transplantation and were compared with 23 patients who received placebo. The stimulation index in the vaccinated group was higher than in those who received placebo (8.6 vs. 5.3) at 5 months, and the severity of HZ was reduced in the vaccinated subjects compared with the control group at 1 year. However, the incidence of HZ was not reduced in vaccine recipients compared with control subjects (23% vs. 22%).
A second randomized control trial of heat-inactivated varicella vaccine was undertaken in 119 patients scheduled to undergo autologous hematopoietic cell transplantation for Hodgkin or non-Hodgkin lymphoma . Unlike the prior study by Redman et al. , vaccine or placebo was given within 30 days before transplantation, as well as at 30, 60, and 90 days after transplantation. At 1 year after transplantation, there was a significantly lower rate of HZ among patients given the vaccine (13%) than among those given placebo (33%). The VZVspecific stimulation index in the vaccinated group was significantly higher (42.8) than that in the placebo group (21.3) at 1 year after transplantation. The mean percentage of CD4 cells that expressed intracellular IFN-γ or tumor necrosis factor-α in response to inactivated VZV at 6 months after transplantation was higher in those who received the vaccine than in those who did not. Adverse effects of the vaccine were generally mild and included pain, induration, and erythema at the injection site.
The authors postulated that vaccination before transplantation induced the production of VZV-specific memory T cells, some of which may have survived the preconditioning regimen and were restimulated by vaccination after transplantation. This study indicates that multiple doses of a heat-inactivated VZV vaccine can reduce the rate of HZ and enhance cellular immunity to the virus in an immunocompromised population. Other vaccines. Other approaches to vaccinating immunocompromised patients against HZ might also be tried. A sequential regimen of inactivated vaccine followed by live virus vaccine might be considered in an effort to prime the immune system before live virus is administered. A similar sequential approach was used for poliovirus vaccination during the transition from an all-live to an all-inactivated poliovirus vaccine program for healthy children in the United States. A subunit vaccine consisting of a viral protein (or proteins) might be used for vaccination instead of killed virus.
Such a vaccine might cause fewer injection-site reactions, because the amount of cellular proteins (which are present in live or heatinactivated VZV vaccines) would likely be reduced. A number of VZV gene products, including the immediate early 4, 62, and 63 proteins and glycoproteins C, E, and I are known to be targets for cytotoxic T cells . Unfortunately, it is unknown which of these or other viral proteins are necessary for protection against HZ. The addition of adjuvants or other delivery systems, such as presentation of VZV antigens by dendritic cells, would likely improve the cellular immune response to subunit or inactivated virus vaccine. A dendritic cell vaccine would be difficult to administer, because it would likely require isolating dendritic cells from the vaccine recipient, pulsing the cells with VZV antigens, and injecting them back into the vaccine recipient. An alternative approach would be to use replication-defective VZV. A number of VZV mutant viruses have been constructed from the Oka vaccine virus that knock out essential viral gene products that are required for virus replication [23–25].
These viruses should be able to infect cells and present nearly all of the viral proteins to the immune system but not replicate and cause disease. Unlike a killed virus vaccine, a vaccine containing such mutants might induce higher levels of MHC class I-restricted CD8 T cell responses, which should enhance cellular immunity to the virus. Replication-defective vaccines have the potential to recombine with wild-type virus and, therefore, are considered to be less safe than inactivated vaccines. Supplement sponsorship. This article was published as part of a supplement entitled “Varicella Vaccine in the United States: A Decade of Prevention and the Way Forward,” sponsored by the Research Foundation for Microbial Diseases of Osaka University, GlaxoSmithKline Biologicals, the Sabin Vaccine Institute, the Centers for Disease Control and Prevention, and the March of Dimes. Financial support: Division of Intramural Research, National Institute of Allergy and Infectious Diseases. Supplement sponsorship is detailed in the Acknowledgments.
The opinions expressed here are those of the author and do not necessarily reflect those of the National Institute of Allergy and Infectious Diseases. The virus titer in the HZ vaccine is ∼14-fold higher than the virus titer in the varicella vaccine ; therefore, vaccination with the HZ vaccine could result in more adverse effects than that with the varicella vaccine. Because rashes are less common after the second dose of vaccine than after the first dose in both healthy adults  and immunocompromised children , a history of varicella with preexisting memory T cells to the virus suggests that immunization of immunocompromised persons to prevent HZ might be safer than immunization to prevent varicella. Because prior exposure to varicella results in development of memory T cells to the virus , vaccination of immunocompromised vericella-zoster virus (VZV)-seropositive persons to prevent HZ might elicit better immunity than does vaccination of VZV-seronegative persons to prevent varicella.