Chapter 65Antiviral therapy of varicella-zoster virus infections

Brivudin

Brivudin (bromovinyl deoxyuridine) is a highly potent thymidine nucleoside analogue with selective activity against HSV-1 and VZV (Keam et al., 2004). The mechanism of action of brivudin appears to be inhibition of the viral DNA polymerase. The drug is well-absorbed after oral administration and has a favorable pharmacokinetic profile which permits once-daily dosing. Brivudin is generally well-tolerated; nausea is the most frequently reported adverse event. Because of concerns about the safety profile of the drug, commercial development of brivudin was halted in the United States. The drug is available in several countries as a 125 mg tablet and as a 0.1% ointment for ophthalmologic use.

Foscarnet

Foscarnet (phosphonoformic acid) is a pyrophosphate analogue that functions as an inhibitor of viral DNA polymerase by blocking the pyrophosphate binding site (Wagstaff and Bryson, 1994). Unlike the nucleoside analogues discussed above, foscarnet does not require intracellular activation by TK, therefore, TK-deficient HSV and VZV isolates that are resistant to acyclovir and related drugs remain susceptible to foscarnet. Foscarnet is administered only by the intravenous route and 80%–90% of an administrated dose is excreted unchanged in the urine. The appropriate dose of foscarnet for treatment of acyclovir-resistant VZV infections has not been assessed systematically, but doses ranging from 40 mg/kg every 8 hours to 100 mg/kg every 12 hours have been used successfully. The most important adverse effect associated with foscarnet therapy is nephrotoxicity. Dose limiting renal toxicity was noted in 15%–20% of patients treated with foscarnet for CMV retinitis. Loading the patient with intravenous saline prior to foscarnet infusion can help reduce the risk of nephrotoxicity. Foscarnet can also induce a variety of electrolyte and metabolic abnormalities, most notably hypocalcemia. Foscarnet-induced electrolyte disturbances can predispose the patient to cardiac arrhythmias, tetany, altered mental status, or seizures. To avoid serious adverse effects that can result from bolus infusion, foscarnet must be administered with an infusion pump over a duration of at least one hour. Serum creatinine levels should be checked at least three times weekly in patients receiving foscarnet and the dosage adjusted according to the manufacturer’s guidelines.

Vidarabine

Vidarabine (adenine arabinoside) was the first intravenous antiviral drug accepted for widespread clinical use and was shown to be effective for VZV infections in immunocompromised patients. Vidarabine has now been replaced by more effective and less toxic antiviral drugs.

Interferon

Administration of alpha-interferon to immunocompromised patients with herpes zoster reduces the risk of viral dissemination, but has little impact on dermatomal rash healing or pain. Interferon therapy was associated with significant adverse events and has been supplanted by more specific antiviral drugs.

Children

In healthy children, varicella is associated with low rates of morbidity and mortality. For most children, supportive care alone is sufficient. Astringent soaks, antipruritics, and antipyretics (preferably acetaminophen) improve comfort. Trimming the fingernails closely helps prevent bacterial superinfections caused by scratching. If bacterial cellulitis (especially caused by group A streptococcus) develops, antibiotics may be required.

Oral acyclovir has been evaluated for treatment of uncomplicated varicella in immunocompetent children (Balfour et al., 1990; Dunkle et al., 1991). Acyclovir therapy, initiated within 24 hours of the onset of rash, resulted in shorter duration of fever, fewer skin lesions, and accelerated lesion healing. Overall, oral acyclovir was well tolerated and reduced the duration of symptomatic illness by about 24 hours. The populations studied in these clinical trials were not significantly large to assess the impact of acyclovir therapy on the incidence of varicella complications. Unlike acyclovir, valacyclovir and famciclovir are not available as suspension formulations and have not been evaluated extensively for treatment of varicella in small children. Some pediatricians still view antiviral therapy as optional for otherwise healthy children with chickenpox. Since the introduction of the varicella vaccine in the United States in 1995, the incidence of chickenpox has declined dramatically, reducing the need for antiviral options in this population.

Adults

Immunocompetent adolescents and adults with varicella can be seriously ill, with high fever, hundreds of cutaneous lesions, incapacitating constitutional symptoms, and a higher risk of complications (especially pneumonitis). Since they are likely to miss at least seven days of school or work, interventions that will reduce the duration of the acute illness are warranted. In a placebo-controlled trial of therapy for 148 adults with varicella, acyclovir (800 mg orally five times daily) was shown to reduce the duration of new lesion formation, reduce the maximum number of lesions, accelerate cutaneous healing, and shorten the duration of fever (Wallace et al., 1992). Similarly, a study of acyclovir treatment in otherwise healthy adolescents demonstrated shorter duration of new lesion formation and of constitutional symptoms, including fever (Balfour et al., 1992). In these studies, the benefit of acyclovir therapy was minimal when treatment was initiated later than 24 hours after rash onset. Overall, acyclovir reduced the duration of illness by about two days. Valacyclovir and famciclovir are also likely to be effective in this setting, but data from controlled clinical trials are lacking. While antiviral therapy is considered optional for healthy children with varicella, the higher potential for morbidity clearly favors treatment in adolescents and adults (Table 65.1). Available data are insufficient to determine whether acyclovir therapy reduces the risk of complications such as pneumonitis or encephalitis.

In immunocompetent patients, visceral dissemination of varicella most often involves the CNS (presenting as cerebellar ataxia, encephalitis, transverse myelitis, or stroke syndromes) or the lungs (viral pneumonitis) (Gnann, 2002). No controlled studies of antiviral therapy for these complications of varicella have been performed. However, information derived from clinical experience and case reports suggests that intravenous acyclovir (10–15 mg/kg every 8 hours) may be beneficial (Haake et al., 1990; Wilkins et al., 1998) (Table 65.1).

The decision whether to initiate antiviral therapy in a patient with chickenpox will hinge on the patients age, underlying medical conditions, and the risk of complications (Arvin, 2002). In general, young children (under age 12 years) are at lower risk for complications than are adolescents or adults. An exception may be secondary pediatric cases in a household, who tend to have more severe disease than the index case. Benefits of antiviral therapy are minimal for healthy children presenting with greater than 24 hours of illness. Because of the greater risk of complications, antiviral therapy is appropriate for adolescents and adults with chickenpox, probably even for those presenting 48–72 hours into the course of illness. Immunocompromised patients with varicella are at significant risk for viral dissemination and visceral involvement and should always receive antiviral therapy.

Pregnant women

Although based more on case reports than on prospectively acquired data, the evidence that varicella in pregnancy is associated with enhanced morbidity is compelling (Nathwani et al., 1998). Women who contract varicella while pregnant have an estimated 10% risk for developing severe VZV pneumonitis. Aggressive antiviral therapy is recommended for a pregnant woman with varicella who develops any evidence of pulmonary involvement, including cough, shortness of breath, or abnormal chest radiograph. Data from clinical trials are lacking, but several case series have reported clinical improvement in pregnant women with varicella pneumonia who were treated with intravenous acyclovir. Although acyclovir is not approved for use during pregnancy for any indication, no fetal toxicity attributable to acyclovir has been demonstrated and the risk-benefit ratio clearly supports the use of acyclovir in the setting of maternal varicella pneumonia (Reiff-Eldridge et al., 2000). Many experts favor antiviral therapy (with acyclovir or valacyclovir) for all pregnant women with chickenpox in an effort to reduce maternal morbidity. No data are available to indicate whether treating the mother will alter the risk of the rare fetal varicella syndrome (Harger et al., 2002).

Immunocompromised patients

The availability of safe and effective antiviral drugs has greatly reduced the high mortality rate previously associated with varicella in immunocompromised patients. Populations at high risk include organ transplant recipients, patients with cancer (especially hematologic malignancies), and other patients receiving immunosuppressive medications (including corticosteroids). Because of the high frequency of visceral involvement in immunocompromised children (or adults) with chickenpox, antiviral therapy is mandatory (Nyerges et al., 1988). A small placebo-controlled trial of intravenous acyclovir in immunocompromised children with varicella demonstrated a dramatic reduction in the frequency of VZV pneumonitis from 27% to 0% (Prober et al., 1982). Therapy with intravenous acyclovir (10 mg/kg or 500 mg/m² every 8 hours for 7–10 days) should be initiated at the first sign of infection. A switch to oral antiviral therapy (acyclovir, valacyclovir, or famciclovir) can be considered when the patient is afebrile and new lesion formation has ceased. When feasible, the dosage of immunosuppressive medications should be temporarily reduced in immunosuppressed patients with varicella. Despite the lack of data from large-scale controlled trials, the safety and efficacy of intravenous acyclovir have led to its acceptance as the drug of choice for varicella in severely immunocompromised patients (Table 65.1). Oral antiviral therapy may be efficacious in modestly immunocompromised patients (e.g., those with solid tumor malignancies or a low-dose corticosteroids), but prospectively acquired data are limited. In a retrospective review of 14 pediatric heart transplant recipients with varicella, half received intravenous acyclovir and half received oral valacyclovir; all patients recovered without serious complications (Dodd et al., 2001).

Patients with HIV infection

Varicella does not appear to be unusually severe in most HIV-seropositive children, although some investigators have reported a longer duration of new lesion formation and higher median lesion counts. A variety of varicella complications in HIV-infected children have been reported (including DIC, pneumonitis, hepatitis, and encephalitis), although reliable incidence figures are not available. Deaths attributable to chickenpox in children with HIV infection are rare and are usually due to pneumonitis. No controlled prospective studies of antiviral therapy for chickenpox in HIV-infected children have been reported, so recommendations must be derived from anecdotal experience. Most clinicians prescribe oral antiviral therapy, reserving intravenous acyclovir for patients with unusually severe or complicated infections (Gershon et al., 1997).

Immunocompetent adults

The goals of therapy for herpes zoster in immunocompetent adults are to accelerate the events of cutaneous healing, reduce the severity of acute neuritis, and most importantly, to reduce the incidence, severity, and duration of chronic pain (Gnann and Whitley, 2002). Even without antiviral therapy, the cutaneous lesions of herpes zoster almost always resolve within a month. However, chronic pain (postherpetic neuralgia) can persist for months or even years and is the most significant manifestation of herpes zoster in the normal host (Johnson, 2002). Three oral antiviral drugs are currently approved in the United States for treatment of herpes zoster. Acyclovir, valacyclovir, and famciclovir have been demonstrated to reduce the duration of viral shedding, promote resolution of skin lesions, and limit the duration of pain when antiviral therapy is initiated within 72 hours of lesion onset (Table 65.1).

In placebo-controlled trials, oral acyclovir (800 mg five times daily for 7 days) was shown to accelerate cutaneous healing and to reduce the severity of acute neuritis in immunocompetent adults with herpes zoster (McKendrick et al., 1986; Huff et al., 1988; Wood et al., 1988; Morton and Thomson, 1989). Overall, acyclovir therapy reduced the duration of new vesicle formation by about 1.5 days and the time to 50% lesion healing by about 2.5 days. These clinical trials with acyclovir showed variable benefit for reduction of the frequency and duration of postherpetic neuralgia (PHN), partially due to limitations in study design and population size. Data from these studies were reexamined in another analysis which demonstrated that acyclovir was significantly superior to placebo for reducing the duration of “zoster associated pain,” defined as the continuum of pain measured from initial onset until final resolution (Wood et al., 1996). Among patients ≥50 years of age, the median time to resolution of pain was 41 days and 101 days and the proportion with persistent pain at 6 months was 15% and 35% in the acyclovir and placebo treatment groups, respectively. Intravenous acyclovir is also effective in this setting, but is impractical for outpatient management of most patients with shingles. Extending oral acyclovir therapy beyond 7 days does not produce any additional benefit (Wood et al., 1994).

Valacyclovir (1000 mg three times daily for 7 days) was compared with oral acyclovir in a study of 1141 immunocompromised patients over 50 years of age with herpes zoster (Beutner et al., 1995). When initiated within 72 hours of lesion onset, the two drugs were equivalent for accelerating the events of cutaneous healing, but valacyclovir was superior to acyclovir in shortening the median time to resolution of zoster associated pain (38 days vs. 51 days; P = 0.001). The proportion of patients still experiencing pain at six months was 25.7% in the acyclovir treatment group and 19.3% in the valacyclovir groups (P = 0.02). Extending valacyclovir therapy to 14 days did not result in any additional benefit.

In a controlled trial conducted in 419 immunocompetent patients presenting within 24 hours of lesion onset, famciclovir (500 mg three times daily) was significantly superior to placebo in reducing the duration of viral shedding, limiting the duration of lesion formation, and accelerating the events of cutaenous healing (Tyring et al., 1995). In a subset of shingles patients ≥50 years of age who had persistent pain after skin healing (n = 170), the median duration of PHN was reduced from 163 days to 63 days (P = 0.004) in the placebo and famciclovir treatment Groups, respectively. In a study comparing famciclovir and acyclovir, the two drugs were shown to have similar efficacy for herpes zoster (Degreef and the Famciclovir Herpes Zoster Study Group, 1994).

Valacyclovir and famciclovir were compared for treatment of herpes zoster in immunocompetent patients in a randomized clinical trial. In this population of 597 patients ≥50 years of age enrolled within 72 hours of rash onset, the two drugs were shown to be therapeutically equivalent, both in terms of cutaneous healing and pain resolution (Tyring et al., 2000). At six months after onset of shingles, 19% of patients in each treatment group still reported pain. Acyclovir, valacyclovir, and famciclovir are all well tolerated and appear to be approximately comparable in clinical efficacy for managing herpes zoster in the immunocompetent host. Because their improved pharmacokinetic properties allow simpler dosing regimens, valacyclovir and famciclovir are preferred over acyclovir for this indication. Comparative drug cost is also a legitimate variable in selecting an antiviral drug for treatment of herpes zoster.

Brivudin (125 mg once daily × 7 days) was compared with acyclovir (800 mg 5 times daily × 7 days) in a study of 1227 immunocompetent adults with herpes zoster. Brivudin was judged to be superior to acyclovir for reducing the time to cessation of new vesicle formation and equivalent to acyclovir in terms of cutaneous healing and acute pain alleviation (Wassilew and Wutzler, 2003a,b). In a follow-up survey of subjects ≥50 years old, the incidence of PHN was lower in brivudin recipients (32.7%) than in acyclovir recipients (43.5%) (Wassilew and Wutzler, 2003a,b). Brivudin is commercially available in several European Union countries, but has not been approved in the United Kingdom or the United States because of concerns about potential drug-related toxicities (Gross et al., 2003).

Certain characteristics have been defined which identify immunocompetent patients at highest risk for complications of shingles and thus most likely to benefit from antiviral therapy. Careful studies have clearly showed that older age, greater skin surface area involved with herpes zoster, and severity of pain at time of clinical presentation are all predictors of more severe and long lasting pain (Wood et al., 1996; Dworkin et al., 1998; Harrison et al., 1999; Whitley et al., 1999; Nagasako et al., 2002). Patients meeting these criteria should be targeted for therapy with antiviral drugs and potent analgesics. Conversely, patients under 50 years of age are at lower risk for severe or prolonged pain and an argument could be made that antiviral therapy in this group is optional. Available efficacy data from published studies relate to patients who present within 72 hours of lesion onset, although patients frequently present for medical care beyond that window (Wood et al., 1998). The presence of new vesicles correlates with recent viral replication and may be a marker for patients who would benefit from antiviral therapy, even beyond 72 hours. In addition, patients presenting with the high-risk characteristics cited above should be considered for antiviral treatment, even when presenting beyond 72 hours after lesion onset. However, patients whose lesions that have all begun to crust are unlikely to derive benefit from antiviral therapy.

Adding corticosteroids to antiviral therapy in patients with acute herpes zoster has been suggested as a way to reduce pain. A study conducted in the United Kingdom compared acyclovir with and without prednisolone in 400 immunocompetent patients over 18 years of age (Wood et al., 1994). Another clinical trial, conducted in the United States, enrolled 208 patients over 50 years of age into a four-armed study (acyclovir plus placebo, prednisone plus placebo, acyclovir plus prednisone, placebo plus placebo) (Whitley et al., 1996). Both studies targeted patients within 72 hours of the appearance of lesions. Both of these studies demonstrated that corticosteroid therapy led to a reduction of pain during the acute phase of herpes zoster, but neither showed any reduction in the risk of postherpetic neuralgia (Wood et al., 1994; Whitley et al., 1996). Addition of corticosteroids to antiviral therapy for treatment of herpes zoster in selected older adults may result in improvements in quality of life measurements such as reduction in time to uninterrupted sleep, reduction in time to return to usual activities, and reduction in analgesic use (Whitley et al., 1996). In the American trial cited above, prednisone was given for three weeks (60 mg daily for 7 days, 30 mg daily for 7 days, and 15 mg daily for 7 days), although it is possible that shorter courses of prednisone are also effective. Corticosteroid therapy can have significant adverse effects and should not be used in patients at risk for steroid toxicity (e.g., patients with diabetes mellitus, gastritis, etc.). Although only the combination of corticosteroids plus acyclovir has been studied, combination therapy using valacyclovir or famciclovir is assumed to be equally effective. Use of corticosteroids for herpes zoster without concomitant antiviral therapy is not recommended. Furthermore, use of corticosteroids in immunocompromised patients with herpes zoster has not been evaluated and is not recommended.

Symptomatic measures should be suggested to keep the patient with herpes zoster more comfortable. Patients should keep the cutaneous lesions clean and dry to reduce the risk of bacterial superinfection. Patients may wash the skin lesions with soap and water in the shower and then carefully pat the skin dry with a clean towel. Some patients find warm or cool astringent soaks (e.g., Domeboro^® solution) to be soothing. A sterile non-occlusive, non-adherent dressing placed over the involved skin will protect the lesions from contact with clothing, which may be especially helpful for patients with increased skin sensitivity (i.e., allodynia). There is no role for topical creams or ointments (including topical acyclovir or penciclovir) in management of herpes zoster. The acute pain of shingles can be very severe and should not be underestimated by the clinician. The pain may be disproportionate to the rash; that is, patients with limited skin involvement can still have severe neuralgic pain. Pain is the most important symptom of herpes zoster and should be aggressively managed. In patients with severe neuralgic pain, sympathetic nerve blocks can provide rapid, but temporary relief (Opstelten et al., 2004). Short-acting narcotic analgesics given on a scheduled (rather than as-needed) basis should be prescribed. Some models used to explain the pathogenesis of PHN suggests that early attenuation of acute pain will reduce the degree of nociceptive input that reaches the spinal cord neurons and prevent the initiation of central mechanisms of chronic pain, thereby reducing the risk of PHN (Dworkin et al., 2000).

Medical management of established PHN is complex and often requires a multifaceted approach (Dworkin and Schmader, 2003; Johnson and Dworkin, 2003). Opioid analgesics are the mainstay of therapy during the acute phases of neuralgic pain (Table 65.2). A clinical trial with controlled-release oxycodone for patients with PHN demonstrated a significant level of pain reduction (67% of those receiving oxycodone versus 11% receiving placebo) as measured by visual analogue scale (Watson and Babul, 1998). Long-acting opioid preparations (oral or transdermal) are preferable to short-acting analgesics for management of chronic PHN. Several randomized, controlled clinical trials have shown tricyclic antidepressants (including amitriptyline, nortriptyline and desipramine) to be effective in reducing the pain of PHN, either as a single agent or in combination with other drugs (Raja et al., 2002; Bowsher, 2003). Because tricyclic antidepressants are frequently associated with sedation and anticholinergic side effects, treatment should begin with a relatively low dose at bedtime, with a gradual increase in dosage as required and tolerated. Nortriptyline is as efficacious as amitriptyline for PHN, but nortriptyline is associated with fewer adverse effects in elderly patients (Watson et al., 1998). In two clinical trial, the anticonvulsant gabapentin was shown to significantly reduce established PHN when used alone or in combination with other modalities (Rowbotham et al., 1998; Rice and Maton, 2001). For treatment of PHN, physicians should initiate gabapentin at a low dose of 100 mg three times daily and escalate (in increments of 100 mg t.i.d.) as required, watching for adverse effects such as somnolence, dizziness, and ataxia. Total daily doses of 1800–3600 mg may be required (Stacey and Glanzman, 2003). Pregabalin has also been shown to be effective and well-tolerated in studies of patients with PHN and is likely to replace gabapentin for this indication (Dworkin et al., 2003; Sabatowski et al., 2004). The adverse effects of these medications can be additive (such as sedation due to opioid analgesics, tricyclic antidepressants, and gabapentin), especially in elderly patients (Schmader, 2001). Local transdermal administration of lidocaine via patches has been shown to significantly reduce PHN in two controlled trials (Davies and Galer, 2004). Topical treatments should only be used on intact healed skin. Topical application of capsaicin cream can provide relief of PHN for some patients, but the local stinging and burning associated with capsaicin may be intolerable for many individuals. In a controlled clinical trial of 277 patients with intractable PHN, intrathecal injection of 60 mg of methylprednisolone acetate once weekly for 4 weeks resulted in significant pain reduction, but these results require confirmation (Kotani et al., 2000). There is no evidence that prolonged administration of antiviral drugs has any benefit for treatment of established PHN (Acosta and Balfour, 2001).

Herpes zoster ophthalmicus

Special emphasis should be given to patients presenting with herpes zoster involving the first division of the trigeminal nerve because of the potential for sight-threatening ocular complications. The ophthalmic division of the trigeminal nerve is the cranial nerve most frequently affected by herpes zoster. Without antiviral therapy, 50% of patients with herpes zoster ophthalmicus (HZO) will develop significant ocular complications (which can include neurotrophic keratopathy, episcleritis, iritis, epithelial or stromal keratitis, etc.) (Liesegang, 1999). Controlled prospective clinical trials clearly demonstrated that oral acyclovir therapy reduced the frequency of serious late ocular inflammatory complications of HZO from about 50%–60% to 20%–30% (Cobo et al., 1986; Harding and Porter, 1991; Herbort et al., 1991; Hoang-Xuan et al., 1992; Beutner et al., 1995). A clinical trial comparing the efficacy of valacyclovir and acyclovir for HZO demonstrated the two drugs to be comparable (Colin et al., 2000). Similarly, a controlled study comparing acyclovir and famciclovir in 454 patients with HZO found that the prevalence of severe and non-severe ocular manifestations (58%) was the same for both treatment groups (Tyring et al., 2001a,b). Some experts favor intravenous acyclovir as initial therapy for patients (especially immunocompromised patients) with severe HZO. Systemic antiviral therapy has largely replaced topical antiviral preparations for treatment of the ocular complications of HZO. Systemic or topical corticosteroids may be indicated for some of the ocular inflammatory phenomena that accompany HZO (e.g., uveitis), but should only be administered under the supervision of an experienced ophthalmologist (Liesegang, 1999). Available data strongly support the routine and early use of systemic antiviral therapy in all patients with HZO in an effort to reduce the risk of ocular complications (Severson et al., 2003; Zaal et al., 2003).

Immunocompromised patients

Patients with disorders of cell-mediated immunity are at increased risk for development of herpes zoster. In this population, those patients with the greatest degree of immunosuppression (such as hematopoietic stem-cell transplant (HSCT) recipients or patients with lymphoproliferative malignancies) are at highest risk for VZV dissemination and visceral organ involvement. Clinical trials with intravenous acyclovir for localized or disseminated herpes zoster in immunocompromised patients clearly demonstrated that treatment resulted in more rapid virus clearance and halted disease progression (Serota et al., 1982; Balfour et al., 1983). Subsequent studies in HSCT recipients have demonstrated that acyclovir, in addition to promoting faster disease resolution, is highly effective at preventing VZV dissemination (Meyers et al., 1984; Shepp et al., 1986). Because most VZV-related fatalities result from disseminated infection, the ability to prevent dissemination has markedly reduced the herpes zoster mortality rate in immunocompromised patients. In addition, intravenous acyclovir is considered the drug of choice for treating dissemination when it occurs, although efficacy data from prospective studies are limited (Balfour et al., 1983; Whitley et al., 1992). The recommended dose of intravenous acyclovir for herpes zoster in severely immunocompromised patients is 10–15 mg/kg (or 500 mg/m²) every 8 hours (Table 65.1). When the infection is under control, therapy can be switched from intravenous acyclovir to an oral antiviral drug for the remainder of the course of treatment. Patients should be treated until healing is complete or for a minimum of 10–14 days (whichever is longer) to reduce the risk of relapsing disease.

Treating shingles in immunocompromised patients on an outpatient basis with oral antiviral drugs is an attractive approach, although supporting data are limited. One small study randomized 27 allogenic HSCT recipients with herpes zoster to either oral or intravenous acyclovir. No VZV dissemination occurred in either group, and no differences in healing or clinical outcome were apparent (Ljungman et al., 1989). Published data from clinical trials with famciclovir and valacyclovir for herpes zoster in immunocompromised patients remain limited, but a growing body of clinical experience suggests that these drugs are safe and effective in this setting (Tyring et al., 2001a,b). For less severely immunosuppressed patients, oral therapy with acyclovir (800 mg five times daily), valacyclovir (1000 mg three times daily), or famciclovir (500 mg three times daily), coupled with close clinical observation, is a reasonable option. Because of the risk of ocular involvement, intravenous acyclovir plus evaluation by an ophthalmologist are recommended for highly immunocompromised patients who present with HZO.

HIV-seropositive patients

The incidence of herpes zoster is about 15-fold higher in HIV-seropositive men than in age-matched controls. Shingles in this population is associated with higher rates of CNS complications, necrotizing retinitis, and recurrent episodes. Prospectively acquired data to guide clinicians when selecting antiviral therapy for herpes zoster in HIV-seropositive patients are currently limited. Nearly 300 HIV-infected patients with shingles were enrolled in controlled studies comparing orally administered acyclovir with the investigational antiviral drug sorivudine. Overall, the time to cessation of new vesicle formation, total crusting, and resolution of zoster-associated pain were 3–4 days, 7–8 days, and about 60 days, respectively (Bodsworth et al., 1997; Gnann et al., 1998). These studies confirm the efficacy and safety of oral antiviral therapy for herpes zoster in patients with HIV infection. Valacyclovir and famciclovir have not been systematically evaluated as treatments for herpes zoster in HIV-infected patients, although anecdotal clinical experience suggests therapeutic benefit. Long term administration of antiherpes virus drugs to prevent recurrences of herpes zoster in patients with AIDS is not routinely recommended. Because of the documented risk of relapsing infection, VZV disease in HIV-seropositive patients should be treated until all lesions are completely resolved, which is often longer than the standard 7–10-day course. What impact anti-VZV therapy may have on the risk of subsequent complications such as CNS infection or retinitis is unknown. Adjunctive therapy of herpes zoster with corticosteroids has not been evaluated in HIV-infected patients and is not currently recommended.

On the basis of clinical experience, most physicians select intravenous acyclovir as the drug of choice to treat severe or complicated herpes zoster in HIV-infected patients. The literature contains numerous case reports documenting successful therapy of neurologic complications with intravenous acyclovir (Poscher, 1994; Lionnet et al., 1996). Some experts have recommended intravenous acyclovir for initial therapy of HZO in HIV-infected patients, although oral therapy appears adequate in most cases.

A syndrome of herpetic retinal necrosis can occur as a late complication of herpes zoster in either immunocompetent or immunocompromised patients, but is seen with the greatest frequency in patients with AIDS. Responses to intravenous acyclovir or ganciclovir have been inconsistent and disappointing. Several case reports have documented preservation of vision in patients treated with a combination of intravenous ganciclovir plus foscarnet, with or without intravitreal ganciclovir (Galindez et al., 1996). The optimal duration of induction therapy and options for long-term maintenance therapy for acute retinal necrosis in HIV-seropositive patients have not been established (Ormerod et al., 1998). When VZV retinitis occurs in immunocompetent patients, the clinical outcome is clearly improved by acyclovir therapy and the prognosis is better. In this population, a suggested treatment regimen based on clinical experience is intravenous acyclovir (10–15 mg/kg every 8 hours) for 10–14 days, followed by oral valacyclovir 1 gram po three times a day for 4–6 weeks (Palay et al., 1991).

Immunocompetent patients

Administration of varicella vaccine within the first few days after exposure to VZV will produce a protective (or partially protective) immune response in VZV seronegative individuals (Watson et al., 2000). About half of patients receiving post-exposure immunization may still develop some signs and symptoms of chickenpox, but the disease manifestations are usually very mild. Postexposure vaccination appears to be more effective and less expensive than preemptive therapy with antiviral drugs. This approach may be useful for managing VZV exposures that occur in a family, in the workplace, or in a medical care setting.

Pregnant women

Advisory committees have recommended administration of varicella-zoster immune globulin (VZIG) to VZV-susceptible pregnant women who have been exposed to varicella (Centers for Disease Control and Prevention, 1996). For maximal efficacy, VZIG must be administered as soon as possible after exposure (within 96 hours). VZIG (as available in the United States) is administered by deep intramuscular injection at a dose of 125 units/10 kg of body weight, to a maximum of 625 units. Intravenous immunoglobulin also contains substantial titers of VZV-specific IgG and may be substituted if VZIG is not available. Unfortunately, in this time-critical scenario, the true VZV serologic status of a pregnant woman with a negative history of varicella is often not known. The clinician may be faced with a decision to initiate passive immunoprophylaxis empirically or to wait for the results of serologic testing. The ideal time to determine VZV serologic status is before pregnancy, when vaccination can be offered to women who are confirmed to be seronegative (Glantz and Mushlin, 1998). Varicella vaccination of pregnant women is not currently recommended because of the theoretical risk of the live virus vaccine for both the fetus and the mother. Prophylactic (or pre-emptive) therapy with acyclovir for a pregnant woman after VZV exposure may be effective, but is an unproven approach.

Immunocompromised (including HIV-seropositive) patients

VZV-seronegative immunocompromised patients with a defined close exposure to either chickenpox or herpes zoster should receive VZIG to provide passive immunity (Zaia et al., 1983). In most cases, VZIG administration will not prevent infection in the susceptible host, but it will significantly reduce the severity of the resultant illness. Placebo-controlled trials in immunocompromised children have demonstrated that VZIG ameliorates the severity of chickenpox and that it significantly reduces the risk of disseminated infection. A single treatment reduces the risk of disseminated infection by about 75% and provides four weeks of passive immunity. VZIG must be administered within 96 hours of exposure at the dose described above. VZIG is not useful for the treatment of established varicella or herpes zoster. The efficacy of VZIG prophylaxis in HIV-seropositive children or adults has not been evaluated prospectively.

Prophylactic administration of acyclovir following VZV exposure has been studied to a limited extent in susceptible immunocompetent patients, but not in immunocompromised individuals. In studies of healthy children conducted in Japan, varicella developed in 16% of the children prophylactically treated with acyclovir and in 100% of children in the control group (Asano et al., 1993) About 80% of children prophylactically treated with acyclovir subsequently seroconverted, indicating VZV infection without significant disease (Suga et al., 1993). However, additional data are required before this approach of preemptive antiviral chemotherapy can be routinely recommended in either immunocompetent or immunocompromised populations. A suggested (but unvalidated) regimen is acyclovir 200 mg orally four or five times daily for 21 days beginning five days after VZV exposure.

Concerns about the use of the live, attenuated VZV_oka vaccine in immunocompromised patients have focused on the potential for the vaccine virus to cause disease and on the possibility that immunocompromised patients will fail to mount a protective immune response. Limited experience with the vaccine in leukemic children and renal transplant recipients have demonstrated that it can be used safely in highly selected populations (Arbeter et al., 1990; Furth and Fivush, 2002).

Immunocompetent patients

There are no circumstances that warrant antiviral chemotherapy to try to prevent herpes zoster in immunocompetent individuals. A live-virus vaccine has proven to be effective for preventing herpes zoster and reducing PHN (Oxman et al., 2005). A randomized, double-blind, placebo-controlled clinical trial enrolling 38,546 adults (age 60 and over) was conducted to evaluate the live attenuated Oka/Merck VZV vaccine. The primary endpoint was “herpes zoster burden of illness,” a composite score capturing zoster incidence, duration, and severity of total pain and discomfort. Compared with placebo, the vaccine reduced the zoster burden of illness by 61.1%, reduced the incidence of herpes zoster by 51.3%, and reduced the incidence of PHN by 66.5% (P < 0.001 for all comparisons). The vaccine was associated with mild reactogenicity (local erythema or tenderness) in 48.3% of recipients, but was otherwise well tolerated. The herpes zoster vaccine was approved for use in the United States in 2006 for immunocompetent adults 60 years of age and over.

Immunocompromised patients

Drug regimens designed to prevent HSV recurrences in immunocompromised patients undergoing cancer chemotherapy or organ transplantation will also effectively prevent herpes zoster (Ljungman, 2001). Combined results from two placebo-controlled trials of long-term (6 months) acyclovir prophylaxis in HSCT recipients demonstrated herpes zoster in 11 (18%) of 62 placebo recipients and in none of the 62 acyclovir treated patients (Lundgren et al., 1985; Perren et al., 1988). Interestingly, the incidence of zoster increased dramatically after the discontinuation of prophylaxis such that, 12 months after transplantation, the cumulative number of herpes zoster cases was virtually identical between the acyclovir and placebo groups. Nonetheless, acyclovir prophylaxis effectively prevents herpes zoster during the early post-transplant period when patients are most severely immunosuppressed and thus have the highest risk for VZV-related complications. Although transplant specialists almost universally recommend 3–6 months of acyclovir prophylaxis, no consensus currently exists regarding the relative merits of longer term prophylaxis. Development of a heat-inactivated VZV vaccine for use in immunocompromised patients is an area of active investigation (Hata et al., 2002).

HIV-seropositive patients

Antiviral chemoprophylaxis for prevention of herpes zoster in patients with AIDS is not routinely recommended. A significant number of HIV-seropositive patients take suppressive antiviral drugs to prevent genital HSV reactivations, which may also prevent herpes zoster. In patients with multiple recurrent episodes of herpes zoster, chemoprophylaxis could be considered (e.g., valacyclovir 1 gram orally twice a day or famciclovir 500 mg orally twice a day), although this approach is unvalidated.