Prevention strategies for herpes zoster and post-herpetic neuralgia

SUMMARY

1. Introduction

Like herpes simplex virus, varicella zoster virus (VZV) establishes and maintains a latent infection in neurons in sensory ganglia.^1,2 Herpes zoster (HZ) (shingles) results from the reactivation of this latent VZV. Clinically significant reactivation occurs when VZV-specific cell-mediated immunity (CMI) declines below a critical threshold. This occurs commonly with aging, human immunodeficiency virus (HIV) infection, lymphoproliferative malignancies and immunosuppressive therapies, all of which are associated with an increased risk of developing HZ.^3,4 HZ develops in these individuals despite the high levels of circulating anti-VZV antibodies, underscoring the crucial role of VZV-specific CMI (VZV-CMI) responses.⁵ Thus, a vaccine that boosted VZV-CMI in high-risk individuals was predicted to prevent or attenuate the development of HZ. Furthermore, drugs are being tested to prevent post-herpetic neuralgia (PHN) in patients who develop HZ. These include early (pre-emptive) treatment with tricyclic antidepressants, anticonvulsants and analgesics. Additional strategies, such as the use of antivirals and nerve blocks, are considered elsewhere in this supplement.⁶

2. Vaccination for the prevention of HZ

Early studies demonstrated that the immunization of immunocompetent older adults with a live, attenuated Oka strain VZV vaccine safely increased VZV-CMI responses.^7,8 In 2005, the results of the Shingles Prevention Study (SPS) – a large, randomized, placebo-controlled trial – were published, demonstrating that immunization of older adults with a high-dose Oka strain VZV vaccine reduced the incidence and severity of HZ and PHN.⁹ This led to the licensure of the first therapeutic vaccine for the control of a persistent viral infection.

The SPS investigated the efficacy of a VZV vaccine for preventing both HZ and PHN. The HZ vaccine contains the same virus used in the vaccine licensed to prevent varicella, but the number of VZV plaque-forming units (PFU) is 14-fold higher. The key questions were whether immunization could reduce any and all of the following:¹⁰ burden of illness (a measurement of HZ incidence, HZ severity and the duration of the associated pain); incidence of PHN (defined as significant pain [≥3 on a 0–10 Likert scale]) at 90 days after rash onset; and incidence of HZ.

The SPS enrolled 38,456 immunocompetent adults aged ≥60 years, who were randomized to receive vaccine or placebo and were followed for a mean of 3.13 years. Overall, the study found that, compared with placebo, the vaccine significantly reduced the HZ burden of illness by 61.1% (95% CI, 51–69; p < 0.001), lowered the incidence of PHN by 66.5% (95% CI, 48–79; p < 0.001), and reduced the incidence of HZ by 51.3% (95% CI, 42–58; p < 0.001) (Figure 1).⁹ Although some vaccinees developed HZ, the mean duration of their pain was less than in non-vaccinated subjects (21 vs. 24 days; p = 0.03), as was the mean severity of illness score (141.2 vs. 180.5; p = 0.008).¹⁰

Open in a separate window

Fig. 1

Kaplan–Meier estimates of the effect of HZ vaccine on the cumulative incidence of (A) PHN and (B) HZ.¹⁰

The efficacy for reducing the three endpoints was the same for men and women. In a comparison between subjects aged <70 years and those aged ≥70 years, the vaccine was equally efficacious in reducing the incidence of PHN but was somewhat less effective in reducing the HZ burden of illness and preventing HZ in the older subjects.⁹

Analysis of VZV-specific immune responses to vaccination showed that, at 6 weeks, and persisting for at least 3 years, the responses were significantly increased in the vaccinees compared with baseline (p = 0.006) for three assays: two assays for VZV-CMI (responder cell frequency [RCF] and interferon-γ enzyme-linked immunosorbent spot-forming cells [ELISPOT]), and an assay for VZV antibody (glycoprotein-based enzyme-linked immunosorbent assay [gpELISA]). Post-vaccination VZV-CMI responses (RCF and ELISPOT), however, were greater for those aged <70 years compared with vaccinees aged ≥70 years (p < 0.001) (Figure 2),¹¹ which concurred with the clinical outcome. The persistence of vaccine-induced responses beyond 3 years is currently being studied in a cohort of 7,000 participants from the SPS.

Open in a separate window

Fig. 2

RCF value, number of responding cells per 10⁵ peripheral blood mononuclear cells (PBMCs); ELISPOT counts, number of spot-forming cells per million PBMCs; gpELISA titre, gpELISA units/mL.¹¹ Reproduced with permission of The University of Chicago Press from Levin MJ et al, J Infect Dis 2008;197:825–35. © The Infectious Diseases Society of America.

3. Issues for a HZ vaccine

The incidence and severity of HZ and PHN increase with age, reflecting the age-related decline in VZV-CMI that characterizes aging.¹¹ The efficacy of the vaccine and the level of VZV-CMI conferred also decline with increasing age. A large literature study describes multiple qualitative and quantitative age-related changes in both the innate and adaptive immune systems,^12,13 resulting in reduced humoral and CMI responses following vaccination.^11,14,15 Although increasing the dose of a vaccine often increases the immune response, this approach to immunosenescence is not feasible with the HZ vaccine because of the limitations in the growth of the VZV in cell culture systems.¹⁶ Additional doses of HZ vaccine did not produce significantly greater responses in elderly vaccinees. Studies with influenza vaccines have shown that the inclusion of adjuvants can enhance vaccine-induced immune responses, including those in elderly subjects,^17,18 but this strategy is not well studied with live vaccines. One approach to improving active immunization against HZ in the elderly is a new adjuvant-containing VZV subunit vaccine that is in development.^19,20

4. Effects of varicella vaccination on HZ epidemiology

The USA adopted a universal childhood single-dose varicella vaccination programme in 1995. By 2005, the incidence of varicella had declined by 90–95% in children aged 1–9 years (Figure 3).²¹ In 2006, the programme was changed to a two-dose regimen with the goal of preventing cases of breakthrough varicella and further reducing endemic disease. However, a potential downside of this stems from the observation that re-exposure to varicella boosts VZV-CMI in people who have previously had the virus.²² It is hypothesized that this ‘environmental boost’ contributes to preventing reactivation of latent VZV. The varicella vaccine clearly benefits infants and children (very significant reduction in complications, emergency room visits, hospitalizations and death), but there is a concern that its widespread use will secondarily lead to a more rapid decline in VZV-CMI with age and earlier onset of HZ.^23,24 Studies to date have not clarified this issue. Long-term longitudinal cohort and additional epidemiological studies will be required in order to do so.²⁵

Open in a separate window

Fig. 3

Varicella incidence, hospitalization and mortality rates: USA, 1990–2005. (A) Varicella incidence: incidence of reported varicella cases from four states (Illinois, Michigan, Texas and West Virginia) that maintained consistent and adequate surveillance during 1990–1995. (B) Hospitalization: varicella was the primary diagnosis code; data are for individuals aged <50 years. (C) Mortality rates: varicella was the underlying cause of death.²¹ Reproduced with permission from Pediatrics, Vol. 122, Page e744–e751, © 2008 by the AAP.

More encouraging data from varicella vaccination suggest that the vaccine substantially decreases the risk of HZ among vaccinated children, making it likely that its widespread use will reduce the overall HZ burden. Active surveillance in the USA has shown that the incidence of HZ among children aged <10 years declined by 55% between 2000 and 2006,²⁶ and that their risk for developing HZ was 4–12 times lower than the risk of matched controls with a history of varicella disease.

As VZV is exclusively a human pathogen, it could theoretically be eradicated by the global use of a highly effective vaccine. However, given the ability of VZV to establish latency and to reactivate and spread many years after the primary infection, it would take many decades of maintaining extremely high, unrealistic levels of vaccine-induced immunity to eradicate the virus.^23,27

5. Current indications and exclusions for the HZ vaccine

The live, attenuated HZ vaccine (Zostavax^®; Merck & Co., Whitehouse Station, NJ, USA) is approved for the prevention of HZ in adults with a healthy immune system. In the USA and Canada, the vaccine is recommended for people aged ≥60 years, whereas in Europe and Australia the vaccine is approved for adults aged ≥50 years.^28–30 The recommendation extends to those who have previously had an episode of HZ and those with chronic medical conditions, such as chronic renal failure, diabetes mellitus, rheumatoid arthritis, and chronic pulmonary disease.³⁰ It is likely that the HZ vaccine can be safely administered to patients without a history of chickenpox, because it has been safely administered to VZV-naïve individuals.³¹ The vaccine is not indicated for the treatment of acute HZ or PHN, or to prevent PHN in patients who develop HZ.³⁰ The vaccine should not be administered to anyone receiving an antiviral drug that can interfere with VZV replication, such as acyclovir, famciclovir or valacyclovir; however, it can be administered if these drugs are discontinued at least 24 hours before receiving the HZ vaccine and are not reinitiated for at least 14 days. Contraindications to using the vaccine include allergy to the vaccine components. Live viral vaccines are generally contraindicated for use in pregnancy and in severely immunocompromised patients in case the attenuated vaccine virus continues replicating in the absence of an adequate specific immune response.^32,33

The Advisory Committee on Immunization Practices (ACIP) of the US Centers for Disease Control and Prevention has made specific recommendations regarding the use of the HZ vaccine in immunocompromised patients, and these are presented in Table 1.³⁰ The HZ vaccine can be co-administered with the trivalent inactivated influenza vaccine³⁴ and several other inactivated vaccines such as Td and Tdap,³⁵ but each vaccine must be administered separately at a different body site. As simultaneous administration of the HZ vaccine and the pneumococcal polysaccharide vaccine resulted in some depression of the VZV antibody response,³⁶ it is now recommended that these two vaccines be separated by at least a 4-week interval, similar to the ACIP recommendation that the HZ vaccine and other live, attenuated vaccines be separated by a period of 4 weeks.³⁷

There is a need to improve upon the current HZ vaccine; namely, to enhance the immune response and efficacy, particularly in the very elderly, and to develop a strategy for active immunization of high-risk immunocompromised patients. An inactivated or subunit vaccine formulated with a potent adjuvant might address these two issues.^20,28

A different problem is presented by the varicella vaccine. This vaccine in children can establish latency and reactivate to cause HZ,^38,39 although Oka strain VZV vaccine-related HZ is less frequent and less severe than HZ after varicella.²⁶ Latency with Oka strain VZV does not occur when the HZ vaccine is administered to elderly people, probably because the recipients have pre-existing immunity that prevents latency from being established.⁴⁰

6. Remaining areas of uncertainty with vaccine strategies⁴¹

The age of the target population for the HZ vaccine remains to be clarified. Vaccination of people aged 50–59 years is not currently undertaken in some countries, including the USA, even though the burden of HZ is substantial in this age group. Such people are likely to be working regularly, which increases the economic burden of HZ at this age. Moreover, the vaccine may be more immunogenic in younger people. Ongoing analysis of the cost-effectiveness of the HZ vaccine as a function of age is likely to inform this decision. In view of the concern that universal vaccination against varicella in childhood could result in an earlier age of onset of HZ, the administration of the HZ vaccine at age 50 may assume added importance. The impact of vaccination on select immunocompromised populations also remains to be determined.

7. Ongoing work

The clinical development of HZ vaccines continues. Merck & Co. has ongoing studies examining the safety and efficacy of Zostavax^® in subjects aged 50–59 years,⁴² and its safety, tolerability and immunogenicity in patients on chronic/maintenance doses of corticosteroids.⁴³ Concomitant use trials confirmed the safety and tolerability of Zostavax^® administered concomitantly with Pneumovax^® 23 (Merck & Co.) in subjects aged ≥60 years, but a reduction in the VZV-specific antibody has led to the recommendation that these two vaccines be administered separately with a 4-week interval.³⁷ The safety, tolerability and immunogenicity of a refrigerator-stable formulation of Zostavax^® have also been established.⁴⁴ Merck & Co. is also proceeding with the development of a heat-inactivated VZV vaccine, which is being evaluated in both healthy adults and immunocompromised patients.^45–47 GlaxoSmithKline is evaluating a recombinant subunit HZ vaccine (GSK1437173A) in adults, including older vaccinees, and immunocompromised patients.^19,20,48

8. Other strategies for the prevention of PHN

Despite developments in understanding the pathogenesis of HZ and PHN, and major advances in the treatments available, as many as half of those who develop PHN will be at least partly refractory to the best therapies available. It is therefore imperative that early, aggressive antiviral treatment is initiated.

When considering the prevention of PHN, it is useful to consider the risk factors that can be used as guides in developing preventive strategies. Although many of the risk factors for PHN have been examined, by far the most important are older age, more severe acute pain and greater rash severity. Many studies have confirmed the relationship between greater acute pain and the subsequent development of PHN. A study by Whitley et al.⁴⁹ showed that individuals with HZ who had severe pain during the acute phase were at all subsequent time points more likely to continue to be in pain (Figure 4).⁴⁹ Post-mortem and skin punch biopsy studies have shown that VZV causes greater damage to the central and peripheral nervous systems in patients with PHN than in those with a history of HZ who did not have persistent pain.^50,51

Open in a separate window

Fig. 4

Duration of zoster-associated pain according to pain severity at presentation in patients who were aged (A) ≥50 or (B) <50 years. (A) For very mild vs. severe pain, hazard ratio (HR) = 3.00 (CI, 2.26–3.99; p = 0.0001); for mild vs. severe pain, HR = 2.23 (CI, 1.69–2.95; p = 0.0001); for moderate vs. severe pain, HR = 1.58 (CI, 1.21–2.06; p = 0.0007). (B) For mild vs. severe pain, HR = 1.69 (CI, 1.34–2.13; p = 0.0001).⁴⁹ Reproduced with permission of The University of Chicago Press from Whitley RJ, J Infect Dis 1998;178(Suppl 1):S71–5. © The Infectious Diseases Society of America.

It may be possible to prevent the development of PHN in some patients by treating them pre-emptively with tricyclic antidepressants. In a study of pre-emptive amitriptyline vs. placebo, the incidence of PHN was halved by amitriptyline.⁵² Post-hoc analyses of a subgroup of patients showed that amitriptyline in combination with acyclovir reduced PHN to a greater extent than acyclovir alone (Figure 5).⁵³ These observations led to the development of a model for preventing PHN and nerve damage during HZ (Figure 6).⁵⁴ This algorithm recognizes that treatment with an antiviral is essential, but it proposes that concomitant analgesic treatment might further reduce acute neuropathic and nociceptive pain and thus decrease the risk of developing PHN more than with antiviral therapy alone.

Open in a separate window

Fig. 5

Combination of antivirals with tricyclic antidepressants reduces the incidence of PHN at 6 months.⁵³ Reprinted from The Lancet Volume 353. Dworkin RH. Prevention of postherpetic neuralgia. Pages 1636–7, © 1999, with permission from Elsevier.

Open in a separate window

Fig. 6

A hypothesized model for attenuating nerve damage and acute pain during HZ and thereby preventing the development of PHN. This model proposes that combined antiviral and analgesic treatment in patients with HZ might decrease the risk of developing PHN by both inhibiting viral replication and reducing the acute afferent barrage of nociceptive and neuropathic pain.⁵⁴ Reprinted with permission from Dworkin RH, Perkins FM, Nagasako EM. Prospects for the prevention of postherpetic neuralgia in herpes zoster patients. Clin J Pain 2000;16(2 Suppl):S90–100.

A proof-of-concept study has shown that treatment with famciclovir plus oxycodone reduced acute pain in patients with HZ more than famciclovir plus placebo.⁵⁵ In the same study, the anticonvulsant gabapentin in combination with famciclovir was not more efficacious compared with famciclovir plus placebo, but it is questionable whether the dosages of gabapentin used were sufficient. Data from an animal model of chronic herpes simplex pain⁵⁶ show that the administration of high doses of gabapentin reduced acute pain and markedly lowered the incidence of delayed post-herpetic pain. Furthermore, a single-dose study of gabapentin in adults with HZ showed that oral gabapentin (900 mg) decreased the pain and allodynia associated with HZ in subjects with onset of unilateral HZ rash within 45 days of enrolment, and it reported average daily pain ≥40 on a 100 mm visual analogue scale (VAS).⁵⁷ Pain severity decreased by 66% with gabapentin compared with 33% with placebo; reductions in allodynia area and severity, and overall pain relief, were also greater with gabapentin.

Medications that could be considered candidates for administration during the initial stage of HZ for the prevention of PHN include: opioid analgesics and tramadol; tricyclic antidepressants and other dual serotonin and norepinephrine re-uptake inhibitors; pregabalin; and oral corticosteroids.⁵⁸ There is some evidence from animal models to suggest that non-steroidal anti-inflammatory drugs (NSAIDs) may be effective in certain neuropathic pain conditions, and their efficacy for reducing acute pain in HZ should be examined.⁵⁹

9. Summary

The successful commercial development of a vaccine to prevent HZ is the first demonstration that a ‘therapeutic’ vaccine can be used to control a latent viral infection. This important proof of concept should encourage the development of other therapeutic vaccines for equally challenging persistent viral infections such as herpes simplex virus, hepatitis viruses B and C, and HIV. The HZ vaccine is efficacious in reducing the incidence of and morbidity associated with HZ and PHN in older adults. The mechanism of action involves boosting VZV-CMI, which prevents and/or controls the reactivation of latent VZV. Despite the success of the vaccine, there is room for improvement. The vaccine is less immunogenic and efficacious in very elderly subjects, probably as a consequence of immunosenescence. Also, the vaccine is contraindicated in some immunosuppressive states. The development of inactivated or subunit vaccines that include potent adjuvants or immunostimulants might be expected to provide enhanced protection and permit use in immunocompromised patients. As well as vaccination, new ways of employing drugs and psychosocial interventions may also improve the prognosis of HZ.

Acknowledgments

Abbreviations

References