Key points

Herpes zoster (HZ), also called zoster and shingles, is caused by reactivation of latent varicella zoster virus (VZV) that usually occurs decades following initial exposure. Before 1995, most people were infected as children and developed the clinical syndrome of chicken pox, although some individuals have anti-VZV antibodies without recall of clinically evident disease. In 1995, a live-attenuated vaccine against VZV (Varivax, Merck) was introduced in the United States and became part of the routine childhood vaccine schedule, with the first dose administered at 13 to 15 months of age. Approximately 98% of the adult population in the United States has serologic evidence of primary infection with varicella and is therefore at risk for the later development of zoster. Although the vaccine is approximately 85% effective, many children develop clinical or subclinical infection with wild-type virus, and are still at risk for later HZ. Although rare, localized or disseminated OKA (vaccine-strain) varicella, and even later HZ can develop in Varivax-vaccinated children. Clinically, zoster typically presents as a unilateral vesicular rash limited to 1 or 2 dermatomes, usually in the thoracic, cervical, and ophthalmologic regions. The rash usually lasts 7 to 10 days, during which the virus may be transmissible through airborne particles. Complications of zoster include postherpetic neuralgia, bacterial superinfection, and disseminated zoster with meningoencephalitis. The incidence of zoster increases with age, presumably because cell-mediated immunity (CMI) wanes with advancing age. Studies of VZV CMI in unvaccinated individuals estimate an approximate 2.7% to 3.9% decrease with each year of age after 60 years, whereas VZV-specific antibody levels remain essentially unchanged. The age-adjusted incidence of zoster in the general population in the United States is estimated to be 3.6 to 4.4 per 1000 person-years, with rates highest among elderly individuals, reaching 11.0 per 1000 person-years in women more than 65 years of age. Women seem to be more susceptible to HZ reactivation than men, with an approximate 25% to 30% higher incidence among women at all ages. Studies have suggested that the sex difference in HZ incidence may be explained by a lower frequency of VZV-specific memory T cells in women.

The Centers for Disease Control and Prevention (CDC) published guidelines for the use of the US Food and Drug Administration (FDA)–approved live-attenuated HZ vaccine (Zostavax, Merck), which is the identical OKA/Merck strain of the Varivax vaccine, but has 14 times the potency to account for the immunosenescence of advancing age and prior exposure to VZV. This document outlines recommendations as well as relative contraindications for the use of the zoster vaccine in adults in the United States, particularly for immunocompromised individuals. Note that these recommendations are based on theoretic concerns, and not based on human data of adverse events of zoster vaccine administration in immunocompromised persons. Guidelines suggest that the vaccine may not be contraindicated for people taking low-dose corticosteroids (<20 mg prednisone daily equivalent), methotrexate (≤0.4 mg/kg/wk), azathioprine (≤3.0 mg/kg/d), or 6-mercaptopurine (≤1.5 mg/kg/d). The decision to vaccinate individuals who are concurrently receiving biologic therapies, particularly tumor necrosis factor inhibitors (TNFi), should be made on an individual basis, but, if possible, the vaccine should be given 4 weeks before institution of biologic therapy. There is no mention of vaccinating individuals who are currently receiving mycophenolate mofetil, which is one of the most commonly used medications for moderate-to-severe systemic lupus erythematosus (SLE).

Unlike influenza, the antibody response to varicella virus does not protect against reactivation of latent virus. It is presumed that the immunoglobulin (Ig) G antibody response serves to protect the host from reexposure of exogenous virus at sites of inoculation after contact with infected individuals. Therefore, the presence of VZV-specific IgG antibodies correlates with past exposure to varicella or effective vaccination, and protects against primary infection. Seroconversion following vaccination is associated with protection from primary vaccination, and loss of detectable antibodies places the individual at risk for breakthrough infection.

In contrast, it seems that intact VZV-specific T cell–mediated immunity is critical to protection from reactivation of latent varicella infection of sensory ganglia following initial infection. In healthy elderly adults, the incidence of HZ reactivation increases with age despite fairly consistent anti-VZV antibody titers. Individuals with T-cell immunodeficiencies, either congenital, acquired immunodeficiency syndrome related, or via immunosuppressive medications, are at increased risk for inability to resolve acute infection, development of life-threatening infection, or frequent episodes of viral reactivation. However, individuals with B-cell immunodeficiency syndromes do not seem to have increased susceptibility to infection or reactivation. At present, the most common assay for VZV-specific cell-mediated immunity in both clinical trials and observational studies is the interferon gamma (IFNγ)–secreting enzyme linked immunosorbent SPOT (ELISPOT) assay, a modification of the enzyme-linked immunosorbent assay that enumerates cytokine production by single cells following stimulation. In an immunologic substudy of the randomized, placebo-controlled Shingles Prevention Study, all 1395 participants had detectable VZV IgG antibodies at baseline, whereas 18.5% had undetectable IFNγ ELISPOT assays. The decline in VZV-specific IFNγ ELISPOT response was approximately 3.9% per increased year of age.

Systemic lupus erythematosus

Several studies have shown that zoster is more common, and can present with more severe manifestations, among immunocompromised patients, including those with SLE and rheumatoid arthritis (RA) with various degrees of immunosuppressive medications. Evidence suggests that the incidence of zoster may be increased in these autoimmune diseases even in the absence of immunosuppressive medications. This increase is possibly caused by an inherent deficiency in cell-mediated or other immunity associated with the disease process. However, distinctions between susceptibility from immunosuppressive medications and susceptibility caused by intrinsic immunologic defects inherent in autoimmune diseases remain elusive.

The autoimmune disease with the most compelling epidemiologic evidence of increased risk of HZ is SLE. Most estimates of the incidence of HZ in SLE have been derived from retrospective single-center studies. Studies have estimated the annual incidence of zoster to range from 6.4/1000 person-years to 32.5/1000 in patients with SLE. This is nearly 2-fold to 10-fold more than the estimated rate of zoster in the general population of 3.4/1000 person-years to 4.4/1000 person-years. In a large study of governmentally and commercially insured adults, the incidence rate of HZ among people with SLE ranged from 15.2 to 24.6 cases per 1000 person-years, with increased rates distributed across all age groups. This rate is significantly higher than is found for healthy individuals aged 61 to 70 years (8.5/1000 person-years) and aged 71 years and older (10.6/1000 person-years); individuals for whom the shingles vaccine is indicated. Furthermore, age-adjusted and gender-adjusted standardized incidence rates for HZ in SLE approach 20/1000 person-years; nearly 4-fold higher than healthy individuals (5.5/1000 person-years) or those with diabetes mellitus (7.5/1000 person-years). In juvenile-onset SLE, estimates approach 58.7/1000 person-years.

Studies are mixed regarding relative roles of SLE disease activity and immunosuppressive medications in increasing risk of HZ. Some studies have found that certain immunosuppressive medications, such as glucocorticoids, azathioprine, and cyclophosphamide, increase the risk of HZ, as would be expected based on the degree of immunosuppression caused by these agents. Among patients with SLE, use of prednisone (hazard ratio [HR], 2.29; 95% confidence interval [CI], 1.24–4.23) and mycophenolate mofetil (HR, 5.0; 95% CI, 1.4–17.6) within the past 6 months placed patients at higher risk for developing HZ. Another case-control study of HZ in SLE performed in a nationwide sample from Taiwan found that use of any immunomodulatory or immunosuppressant medication, including hydroxychloroquine, increased the risk of HZ in a dose-dependent manner. Studies are generally mixed, with some showing HZ incidence increasing with immunosuppression, whereas others found a high rate of HZ among patients with SLE with quiescent disease on little to no immunosuppression. In another study of large administrative claims data, the crude incidence of HZ among all patients with SLE was estimated to be 15.2/1000 person-years compared with 4.8/1000 person-years in the general population. Although the use of immunosuppressant therapy among patients with SLE increased the incidence of HZ further, even those patients with SLE not receiving immunosuppressants had increased HZ compared with the general population (12.2/1000 person-years), suggesting that immunosuppressants play a smaller role in the increased risk of HZ imposed by SLE.

Very few studies have evaluated the role of VZV-specific immunity in SLE in relation to risk of HZ reactivation. One study of 24 Korean patients with SLE, all anti-VZV IgG positive, found statistically significantly lower VZV-specific IFNγ-releasing CD4 cells (0.043% ± 0.009%) than in 12 healthy subjects (0.126% ± 0.025%; P = .007). Cell-mediated responses to cytomegalovirus among patients with SLE were higher, suggesting that the lower response to VZV may be more specific than simply a general reduction in surveillance against viral infections. However, rates of HZ in this cohort were not reported.

One group published the results of a pilot study of CMI response to the live-attenuated HZ vaccine (Zostavax) in 10 female patients with SLE with quiescent disease and on minimal immunosuppressant medications (following CDC guidelines) compared with responses in healthy women. Although not designed for statistical power, the study found that patients with SLE mounted a demonstrable, but lower-magnitude, VZV-specific response than healthy women at 2, 6, and 12 weeks after vaccination. The proportion of subjects in each group with a greater than 50% increase in VZV-specific CMI compared with baseline was similar at all time points following vaccination, with no VZV-related adverse events or flares of underlying SLE. Patients with SLE seemed to have a blunted CMI response to VZV exposure, and this may, in part, contribute to the increased rates of HZ seen in this population. However, reasons for the blunted response have yet to be explored.

Systemic lupus erythematosus

Several studies have shown that zoster is more common, and can present with more severe manifestations, among immunocompromised patients, including those with SLE and rheumatoid arthritis (RA) with various degrees of immunosuppressive medications. Evidence suggests that the incidence of zoster may be increased in these autoimmune diseases even in the absence of immunosuppressive medications. This increase is possibly caused by an inherent deficiency in cell-mediated or other immunity associated with the disease process. However, distinctions between susceptibility from immunosuppressive medications and susceptibility caused by intrinsic immunologic defects inherent in autoimmune diseases remain elusive.

The autoimmune disease with the most compelling epidemiologic evidence of increased risk of HZ is SLE. Most estimates of the incidence of HZ in SLE have been derived from retrospective single-center studies. Studies have estimated the annual incidence of zoster to range from 6.4/1000 person-years to 32.5/1000 in patients with SLE. This is nearly 2-fold to 10-fold more than the estimated rate of zoster in the general population of 3.4/1000 person-years to 4.4/1000 person-years. In a large study of governmentally and commercially insured adults, the incidence rate of HZ among people with SLE ranged from 15.2 to 24.6 cases per 1000 person-years, with increased rates distributed across all age groups. This rate is significantly higher than is found for healthy individuals aged 61 to 70 years (8.5/1000 person-years) and aged 71 years and older (10.6/1000 person-years); individuals for whom the shingles vaccine is indicated. Furthermore, age-adjusted and gender-adjusted standardized incidence rates for HZ in SLE approach 20/1000 person-years; nearly 4-fold higher than healthy individuals (5.5/1000 person-years) or those with diabetes mellitus (7.5/1000 person-years). In juvenile-onset SLE, estimates approach 58.7/1000 person-years.

Studies are mixed regarding relative roles of SLE disease activity and immunosuppressive medications in increasing risk of HZ. Some studies have found that certain immunosuppressive medications, such as glucocorticoids, azathioprine, and cyclophosphamide, increase the risk of HZ, as would be expected based on the degree of immunosuppression caused by these agents. Among patients with SLE, use of prednisone (hazard ratio [HR], 2.29; 95% confidence interval [CI], 1.24–4.23) and mycophenolate mofetil (HR, 5.0; 95% CI, 1.4–17.6) within the past 6 months placed patients at higher risk for developing HZ. Another case-control study of HZ in SLE performed in a nationwide sample from Taiwan found that use of any immunomodulatory or immunosuppressant medication, including hydroxychloroquine, increased the risk of HZ in a dose-dependent manner. Studies are generally mixed, with some showing HZ incidence increasing with immunosuppression, whereas others found a high rate of HZ among patients with SLE with quiescent disease on little to no immunosuppression. In another study of large administrative claims data, the crude incidence of HZ among all patients with SLE was estimated to be 15.2/1000 person-years compared with 4.8/1000 person-years in the general population. Although the use of immunosuppressant therapy among patients with SLE increased the incidence of HZ further, even those patients with SLE not receiving immunosuppressants had increased HZ compared with the general population (12.2/1000 person-years), suggesting that immunosuppressants play a smaller role in the increased risk of HZ imposed by SLE.

Very few studies have evaluated the role of VZV-specific immunity in SLE in relation to risk of HZ reactivation. One study of 24 Korean patients with SLE, all anti-VZV IgG positive, found statistically significantly lower VZV-specific IFNγ-releasing CD4 cells (0.043% ± 0.009%) than in 12 healthy subjects (0.126% ± 0.025%; P = .007). Cell-mediated responses to cytomegalovirus among patients with SLE were higher, suggesting that the lower response to VZV may be more specific than simply a general reduction in surveillance against viral infections. However, rates of HZ in this cohort were not reported.

One group published the results of a pilot study of CMI response to the live-attenuated HZ vaccine (Zostavax) in 10 female patients with SLE with quiescent disease and on minimal immunosuppressant medications (following CDC guidelines) compared with responses in healthy women. Although not designed for statistical power, the study found that patients with SLE mounted a demonstrable, but lower-magnitude, VZV-specific response than healthy women at 2, 6, and 12 weeks after vaccination. The proportion of subjects in each group with a greater than 50% increase in VZV-specific CMI compared with baseline was similar at all time points following vaccination, with no VZV-related adverse events or flares of underlying SLE. Patients with SLE seemed to have a blunted CMI response to VZV exposure, and this may, in part, contribute to the increased rates of HZ seen in this population. However, reasons for the blunted response have yet to be explored.