Viral Encephalitis and Meningitis

Herpes simplex encephalitis

HSV-1 encephalitis is the most common identified cause of sporadic fatal encephalitis in the United States, accounting for approximately 10% of all cases of encephalitis and occurring with a frequency of about 2–4 cases per 1,000,000 population per year ( ). Early recognition is important because the efficacy of the antiviral drug acyclovir (ACV) in reducing morbidity and mortality decreases as neurological disease progresses ( ). HSV-1 strains cause over 90% of cases of herpes simplex encephalitis (HSE) in adults, with the remainder due to HSV-2. Conversely, HSV-2 is a more common cause of meningitis and neonatal meningoencephalitis than HSV-1. Both HSV types 1 and 2 have been associated with myelitis.

Fever is present in about 75%–80% and headache in about 60%–70% of patients with brain biopsy or cerebral spinal fluid (CSF) polymerase chain reaction (PCR) proven HSE ( ). Other common features include disorientation or altered consciousness (70%–90%), personality change (60%), focal or generalized seizures (55%), memory disturbance (35%), motor deficit (40%), and aphasia (40%–60%) ( ). Immunocompromised patients may have atypical clinical and MRI presentations including the absence of prodromal symptoms, fewer focal findings, and MRI abnormalities outside the “classical” frontotemporal areas ( ). Atypical presentations may also occur in young children ( ).

It is important to recognize that no sets of signs or symptoms are pathognomonic of HSE, and multiple infectious and inflammatory processes can mimic HSE ( ). Definitive diagnosis of HSE is based on amplification of HSV DNA in CSF using an HSV-specific PCR assay or, less commonly, isolation of virus, or detection of viral antigen or viral nucleic acid from brain tissue at biopsy or autopsy (see later discussion).

Examination of CSF is a critical diagnostic test in suspected cases of HSE. CSF is usually under increased pressure, with a median lymphocytic pleocytosis of 25–150 white blood cells (WBC) per μL (range 0–1000). In early studies, over 95% of PCR or biopsy-proven cases of HSE have a CSF pleocytosis ( ), although some more recent reports suggest normocellular CSF can occur in up to 26% of cases (Saraya et al., 2016), and this phenomenon may be more common in patients with HIV-infection or otherwise immunocompromised ( ). A recent retrospective multicenter review of pediatric HSE found that 16% of children with HSE had normal CSF glucose, protein, and cell counts ( ). HSE is often hemorrhagic, and both red blood cells and xanthochromia can be detected in CSF, although neither feature occurs with significantly greater frequency in HSE compared to other causes of focal encephalitis ( ). The presence of red cells in CSF may be associated with a worse prognosis ( ). CSF protein concentration is moderately elevated (65–85 mg/dL) in about 60%–70% of patients and glucose concentration is normal in the majority (75%–90%) patients ( ). Autoantibodies, notably those against the NMDA receptor, may develop as a consequence of HSE in 30% of patients ( ). A retrospective study found that 30% of patients with PCR confirmed HSE had detectable serum NMDAR antibodies ( ). In adult HSE cases the presence of NMDAR antibodies has been associated with delayed recovery and resurgent symptoms suggesting disease relapse, that can respond to appropriate immunomodulatory therapies. In children, NMDAR antibodies may be associated with choreiform and other movement disorders ( ).

CSF PCR is the diagnostic assay of choice in HSE, as virus is cultured from CSF in less than 5% of cases. Amplification of HSV DNA from CSF by PCR testing has a sensitivity of 98% and specificity of 94%–99% for the diagnosis of HSE compared to brain biopsy ( ) ( Table 78.6 ). Surprisingly, CSF HSV genome copy number (“HSV viral load”) is not a reliable predictor of outcome ( ). HSV CSF PCR results must be interpreted considering the pre-test probability that a patient has HSE (Bayesian decision analysis). For example, a negative HSV CSF PCR in a patient with a low (∼5%) prior probability of HSE reduces the post-test likelihood of HSE to approximately 0.2%, whereas a negative PCR in a patient with a high (∼60%) pre-test likelihood of HSE only reduces the post-test likelihood of HSE to around 6% ( ). CSF PCR remains a sensitive technique for the detection of HSE, even in patients who have received up to a week of ACV therapy ( ). Despite the high overall sensitivity of CSF HSV PCR, false-negative results have been reported, most notably in patients in whom CSF was obtained within the first 72 hours of illness onset ( ). For this reason, caution should be used in stopping ACV therapy in patients with suspected HSE on the sole basis of a single negative CSF PCR test obtained within 72 hours of symptom onset, unless a suitable alternative diagnosis has been established.

When definitive diagnosis of HSE was dependent on brain biopsy, patients often had significantly depressed levels of consciousness and advanced neurological disease at the time of diagnosis. With routine use of HSV CSF PCR instead of brain biopsy for definitive diagnosis of HSE, patients are often identified at an earlier stage of illness, exhibit less depression of consciousness, and have improved response to therapy. Over 75% of cases with PCR-proven HSE had a Glasgow Coma Score (GCS) above 12 ( ), whereas only 30% of cases with biopsy-proven HSE had a GCS above 10 in the Collaborative Antiviral Study Group (CASG) trial of vidarabine versus ACV therapy ( ).

Magnetic resonance imaging (MRI) is significantly more sensitive than computed tomography (CT) and is the neuroimaging procedure of choice in patients with suspected HSE. Approximately 90% of patients with PCR-proven HSE will have MRI abnormalities involving the temporal lobes ( ) ( Fig. 78.1 ). The electroencephalogram (EEG) may be abnormal early in the course of disease, demonstrating diffuse slowing, focal abnormalities in the temporal regions, or periodic lateralizing epileptiform discharges (PLEDs). EEG abnormalities involving the temporal lobes are seen in approximately 75% of patients with PCR-proven HSE ( ). Brain biopsy is now only rarely performed for diagnosis of HSE. Biopsy is reserved for atypical cases in which the diagnosis remains in question and for those who respond poorly to treatment. Biopsy specimens from patients with HSE show hemorrhagic necrosis, HSV antigen in infected cells, and accumulations of viral particles forming acidophilic intranuclear inclusion bodies in neurons (Cowdry type A inclusions; Fig. 78.2 ).

Empirical therapy with ACV should be started immediately in acute cases of focal encephalitis of suspected viral etiology ( Table 78.7 ). Delay in the institution of ACV therapy is surprisingly common and occurs more frequently in patients with other severe underlying diseases, chronic alcohol abuse, pleocytosis of less than 10 cells/μL, and delayed initial neuroimaging studies, as these factors may make clinicians less likely to consider HSE in the differential diagnosis ( ). Mortality in untreated cases of HSE is around 70%, but this is reduced to 19%–28% in patients treated with ACV ( ), and to less than 10% in more recent studies ( ). Morbidity due to HSV-1 encephalitis remains high even in patients receiving ACV, with only 37.5% of all patients surviving with no or only mild deficits ( ). In a more recent study 23% had no sequelae and 32% mild sequelae ( ). However, in specific subpopulations of ACV-treated HSE patients, prognosis can be considerably better. For example, 50% (12/24) of patients in whom treatment with ACV was initiated when their GCS exceeded 6 survived with no or only minor sequelae, and more than 60% who were younger than age 30 and had a GCS above 6 survived with no or only minor sequelae ( ). Outcomes have been generally similar in cases of PCR-proven HSE treated with ACV, with 35%–56% of patients returning to normal functional status or having only mild disability or residual sequelae at 6 months follow-up ( ). Survival and prognosis are influenced by several factors including level of consciousness at initiation of therapy (e.g., GCS), patient age, and duration of disease before therapy ( ). In the CASG trial, all (9/9) patients treated with ACV within 4 days of onset of fever, headache, and focal neurological deficits survived, whereas the mortality was 35% in those in whom ACV treatment was initiated when disease was more than 4 days old ( ). In another study, the recovery rate was 50% in patients treated within 5 days of illness onset ( ). In a third study of PCR-proven HSE, 75% of patients with an ultimately favorable outcome had received ACV therapy within 2 days of hospital admission compared to only 30% of those with a poor outcome ( ).

The standard adult dose of ACV for HSE is 10 mg/kg, given intravenously (IV) every 8 hours (20 mg/kg every 8 hours in neonates and children) for 14–21 days. Renal insufficiency is an infrequent, usually reversible side effect of ACV therapy and the risk is reduced by appropriate hydration. Neither higher doses of ACV nor prolonged therapy with oral valacyclovir after a standard intravenous course improved outcomes in adults ( ). ACV dosing should be adjusted appropriately in patients with renal insufficiency. Periodic shortages of intravenous ACV and the cost of intravenous preparations in resource-poor countries have led some to suggest that oral valacyclovir may be an alternative when intravenous ACV is unavailable ( ), but it is critical to emphasize that there are only isolated case reports and no comprehensive studies of the efficacy of oral drugs in HSE and this alternative should only be considered in extraordinary circumstances or settings. When a specific shortage of intravenous formulations of ACV exists, ganciclovir can be used as an alternative. Retrospective studies have suggested that there may be some benefit in adding corticosteroids to ACV treatment ( ), although controlled clinical trials are needed before the role of steroid therapy as an adjunct to ACV can be definitively established. Cases of HSE due to ACV-resistant HSV strains have been reported predominantly in immunocompromised patients and in extraordinarily rare instances in immunocompetent individuals (<1%) who have never been exposed to antiviral drugs ( ). Foscarnet (180 mg/kg IV daily given in two or three divided doses) is an alternative therapy for patients with suspected or proven ACV-resistant strains or with allergy to ACV ( ).

Neonatal herpes simplex virus meningoencephalitis

In contradistinction to adults, in whom HSE is usually caused by HSV-1, HSV-2 is the most common causal agent of meningoencephalitis in neonates (although HSV-1 disease may also occur). Neonates who acquire HSV from the birth canal develop infection of the CNS in 50% of cases ( ). CNS disease occurs as either a component of an overwhelming sepsis-like disseminated disease with multiorgan involvement (in the first week of life) or as isolated CNS disease that usually presents later (weeks 2–10 of life), with or without accompanying vesicular skin, mucous membrane, or conjunctival lesions (skin, eye, mouth disease). The presence of vesicular skin or mucosal lesions in an infant of this age, even in the absence of fever or systemic symptoms, warrants immediate evaluation of CSF for HSV infection, because up to 30% of infants with presumed isolated skin, eye, and mouth disease are subsequently identified as having CNS involvement. Neonates with possible HSV disease should be treated empirically with IV ACV, 20 mg/kg, every 8 hours ( ). Treatment should be continued for 14 days in HSV-infected infants with isolated skin, eye, and mouth disease and 21 days in infants with sepsis or CNS involvement. Relapses of skin, eye, and mouth disease (with the potential for CNS involvement and subsequent neurological deficits) are common in the first year of life following neonatal HSV disease. A recent double-blind, placebo-controlled study found that a 6-month course of oral ACV (300 mg/meter sq. of body surface area 3× daily), after completion of the standard 14- to 21-day course of intravenous ACV, resulted in improved neurodevelopmental outcomes at 12 months after treatment compared to placebo and this has become standard of care ( ).

Herpes simplex virus meningitis

HSV meningitis accounts for approximately 10% of acute viral meningitis in adults. In most series greater than 75% of those affected are women ( ). At the time of their first episode of genital herpes (generally due to HSV-2), approximately 36% of women and 11% of men have symptoms of meningitis including fever, headache, and nuchal rigidity. In a review of the Mayo Clinic experience with HSV-2 meningitis all patients had headache and about half had fever, photophobia, and meningismus, and just over a quarter had nausea or vomiting ( ). Only ∼13%–25% of patients have a history of genital herpes or active lesions ( ). The CSF shows a lymphocytic pleocytosis with an average of ∼500 cells/μL, an elevated protein (average ∼150 mg/dL) and a normal glucose. Neuroimaging is normal in nearly 85% but may show meningeal enhancement in rare cases ( ). Following an initial episode of HSV-2 meningitis, 20%–30% go on to develop recurrent episodes ( ). A randomized double-blind controlled study failed to show any effect of long-term valacyclovir suppressive therapy (500 mg 2×/day) in reducing the frequency of recurrent episodes of meningitis ( ). CSF viral cultures are invariably negative during recurrent episodes of meningitis, although the virus may be isolated during the first (primary) episode. Other neurological complications such as paresthesias, urinary retention, and transverse myelitis have been described with HSV-2 infections.

CSF PCR has identified HSV-2 as the primary etiological agent in patients with benign recurrent lymphocytic meningitis (Mollaret meningitis) ( ). This syndrome may be associated with the presence in the CSF of large cells of monocyte-macrophage lineage (Mollaret cells).

Treatment of HSV-2 meningitis varies considerably across published series and no high-quality randomized controlled trials have been performed ( ). A typical regimen is to use oral valacyclovir (500 mg 3×/day) for 7–10 days. Patients with more severe symptoms may benefit from an initial short course of intravenous ACV (e.g., 500 mg q8h ×3 days) followed by oral valacyclovir.

Primary infection

At least 95% of the adult population has been infected with VZV, and the lifetime risk of developing shingles is 3%–5%. VZV can involve virtually every part of the CNS and PNS ( ). Primary VZV may produce meningitis or encephalitis in immunocompromised patients. An acute self-limited cerebellar ataxia occurs in about 1 in 4000 children (<15 years of age) during or immediately following primary VZV infection (chickenpox). Postinfectious encephalomyelitis follows an estimated 1 in 2500 cases of primary VZV infection.

Reactivation of herpes zoster can produce meningitis, myelitis, and encephalitis in older children and adults. Zoster “encephalitis” can occur in both immunocompetent and immunocompromised hosts and in adults typically results from reactivation of herpes zoster decades after an initial infection with varicella (chickenpox). It has been suggested that VZV encephalitis may be second only to that caused by HSV as a cause of sporadic encephalitis in adults, and may be the leading cause of sporadic encephalitis in children ( ). VZV encephalitis is in fact a vasculopathy rather than a true encephalitis ( ). In older (>60 years) immunocompetent adults, VZV vasculopathy (previously called granulomatous arteritis ) typically presents with acute focal stroke-like deficits due to inflammatory involvement of large cerebral arteries, typically occurring in the trigeminal distribution. Most cases are monophasic, although rare cases follow a chronic or relapsing-remitting course ( ). CSF shows a lymphocytic pleocytosis. MRI shows a large single infarct, most commonly in the carotid, middle, or anterior cerebral territory. At autopsy, viral particles, antigen, and DNA can be found in the involved artery. Diagnosis can be made by demonstration of VZV DNA in CSF by PCR or by demonstration of VZV immunoglobulin (Ig)M or intrathecal synthesis of VZV IgG in CSF.

In immunocompromised individuals, VZV reactivation produces a multifocal vasculopathy predominantly involving small and medium-sized arteries, resulting in a clinical syndrome of mental status changes, focal deficits, and a CSF mononuclear pleocytosis ( ). The typical rash of zoster may be absent. Neuroimaging shows multifocal hemorrhagic and ischemic cortical and subcortical infarcts. Diagnosis can be established by VZV CSF PCR or by demonstration of CSF VZV IgM or intrathecal VZV-specific IgG antibody synthesis. While most cases of VZV encephalitis present as a vasculopathy on imaging, a recent case series of VZV encephalitis cases found that 14 of 20 patients had a nonvascular, nonspecific or normal neuroimaging result ( ).

VZV can also cause an aseptic meningitis in immunocompetent hosts even in the absence of a characteristic varicella (chickenpox) or zoster (shingles) rash ( ). VZV accounts for up to 8% of viral meningitis cases, making it third in importance behind EVs and HSV-2 ( ).

VZV has recently been associated with giant cell arteritis (GCA). VZV antigen can be detected in up to 75% of temporal artery biopsy specimens from patients with associated evidence of GCA ( ). VZV antigen is also detected in a high percentage (64%) of biopsy specimens of temporal arteries taken for diagnosis of GCA even when histological evidence of GCA is not found. The incidence of VZV in “normal” temporal arteries is reported at 8%–22%. The significance of these findings and their implications for pathogenesis and potentially treatment remain to be established. Although the suggestion that GCA patients be treated with valacyclovir in combination with prednisone has been made ( ), no clinical trials of this strategy have been reported.

There are no controlled randomized clinical trials of antiviral therapy of CNS VZV infection. Intravenous ACV (10 mg/kg every 8 hours) for 7–10 days is generally recommended for treatment of immunocompromised children and adults with chickenpox. Patients with VZV CNS disease including vasculopathy receive IV ACV (10–20 mg/kg or 500 mg/m ² every 8 hours) for a minimum of 7 ( ) to 14 ( ) days, combined with a steroid pulse for 3–5 days (e.g., prednisone 60–80 mg/day) ( ). HIV-infected individuals with localized (non-disseminated) dermatomal zoster can be treated with oral famciclovir (500 mg 3× daily) or valacyclovir (1000 mg 3× daily) for 7–10 days.

Herpes zoster

Following primary infection, VZV becomes latent in cells of the dorsal root ganglia. Reactivation of endogenous latent virus produces herpes zoster (shingles). The virus can reactivate after injury or trauma to the spine or nerve roots or in response to waning cell-mediated immunity to VZV caused by age or immunosuppression related to HIV infection, cancer, cytotoxic drugs, or systemic illness. Herpes zoster is frequently the first clinical presentation of underlying HIV infection and may present as protracted or multidermatomal disease. The incidence of shingles is up to 25-fold greater in HIV-infected individuals than in the general population. Following universal varicella vaccination for US children, concerns arose that decreases in exposure and boosting of immunity in older adults would result in increased rates of zoster. However, recent work shows that widespread childhood immunization against zoster did not seem to affect the incidence of herpes zoster from 1992 to 2010 ( ).

Herpes zoster typically begins with pain and paresthesias in one or two adjacent spinal or cranial dermatomes ( ). Pain is followed in 3–4 days by a painful pruritic vesicular eruption in the area supplied by the affected root. The eruption typically lasts 10–14 days. Eruption most commonly occurs in the lower thoracic dermatomes but also commonly involves the trigeminal distribution and the cervical or lumbosacral dermatomes. Involvement of the first division of the trigeminal ganglion produces ophthalmic zoster and may be associated with conjunctivitis, keratitis, anterior uveitis, or iridocyclitis. Fortunately, vision loss following herpes zoster ophthalmicus is rare. Involvement of the geniculate ganglion produces otic zoster, or the Ramsay Hunt syndrome—painful facial paresis accompanied by tympanic membrane and external auditory canal vesicular rash. Herpes zoster involving cervical and thoracic levels may be associated with myelitis, and in the lumbosacral region may be accompanied by bladder dysfunction or ileus. Complications of zoster include postherpetic neuralgia, segmental motor atrophy in the affected dermatome, meningitis, myelitis, large-vessel vasculitis (usually involving the carotid or its branches on the side of zoster ophthalmicus), and multifocal leukoencephalitis or encephalitis with generalized cerebral vasculopathy ( ).

Risk factors for postherpetic neuralgia in patients with shingles include age older than 50 years and prodromal sensory symptoms. Treatments are directed toward lessening pain, reducing virus shedding, and shortening healing time. ACV (800 mg orally, 5× daily for 7 days), famciclovir (200 mg orally, 3× daily for 7 days), or valacyclovir (1 g orally, 4× daily for 7 days) accelerates cutaneous healing and decreases acute zoster pain if begun within 72 hours of onset of rash. Whether these agents significantly decrease the incidence, duration, or severity of postherpetic neuralgia is uncertain. In patients without contraindications, a short course of corticosteroids (e.g., 40 mg prednisolone/day, tapered over 3 weeks) may be added to antiviral therapy. Compared with ACV therapy alone, addition of corticosteroids has been shown to improve comfort levels (pain reduction during the acute phase) following herpes zoster, although its efficacy in reducing subsequent risk of postherpetic neuralgia remains uncertain. Currently, zoster vaccination is the primary approach to decreasing zoster incidence and the associated risk of developing post-herpetic neuralgia. A new vaccine (“Shingrix”) containing VZV glycoprotein E and adjuvant has recently been licensed. In phase III trials the vaccine (IM ×2 doses 2 months apart) reduced the incidence rate of shingles from 9.1/1000 person-years to 0.3 per 1000 person years in individuals over the age of 50 followed for a median of ∼3 years ( ). Vaccine efficacy was estimated at between 96.6% and 97.9% for all age groups. The safety profile was excellent, although local injection site and systemic reactions (grade 3) were common (17%). This vaccine is now recommended for all adults over the age of 50 regardless of their prior VZV vaccination history or history of shingles. Because this is not a live attenuated vaccine it can also be used safely in immunocompromised individuals.

Congenital human cytomegalovirus

HCMV infection is the most common human congenital infection and can cause severe injury to the infected fetus. Up to 75% of congenital HCMV infections are due to nonprimary maternal infection, implying that better screening procedures are needed to detect and prevent HCMV reactivation and fetal infection ( ). A review of 15 studies published from 1970 to 2004 that screened newborns for HCMV reported that 12.7% of 117,986 newborns with congenital HCMV infection were symptomatic at birth ( ). The mortality for newborns with symptomatic disease is 10%–30%, and up to 90% of survivors will have neurological sequelae. The rate of sequelae in newborns with asymptomatic primary infection has been estimated at roughly 15% ( ), most commonly including microcephaly, poor feeding, lethargy, hypotonia, and seizures. In addition to congenital cases, there are cases that arise in the perinatal period as a consequence of passage through an infected birth canal or following breastfeeding. Persistent high levels of viral replication in the eye and brain of the developing fetus produce encephalitis, ependymitis, and retinitis, a pattern similar to that seen in patients with opportunistic HCMV infection in the setting of HIV infection. Pathologically, encephalitis occurs in a periventricular pattern and may cause polymicrogyria and hydrocephalus ( Fig. 78.3 ). CT scans show characteristic periventricular calcifications in 20%–30% of children with symptomatic HCMV infection. Retinitis with optic atrophy is seen in about 15% of those affected and results in characteristic hyperpigmented retinal scars. As noted, 90% of survivors of congenital HCMV infection have residual neurological sequelae including psychomotor retardation, learning delays, mental retardation, seizures, optic atrophy and retinitis, and hearing loss. Mild or subclinical congenital infections may also manifest later in childhood as sensorineural deafness or developmental delay. Congenital human HCMV infection is the most common, nonheritable cause of hearing loss in the United States.

Diagnosis of congenital HCMV is made by identification of HCMV DNA using PCR testing of urine, saliva, or CSF during the immediate postnatal period. Because urinary excretion of HCMV can often persist during the first year of life, isolation of virus or viral DNA from urine may be useful in later diagnosis. HCMV inclusion-bearing cells may also be found in affected organs ( Fig. 78.4 ) and in stained preparations of urinary sediment and saliva. Serological studies may be difficult to interpret owing to transplacental transfer of antibody from the mother. Although detection of virus-specific IgM antibodies has been used to diagnose congenital HCMV, direct detection of viral markers is more accurate and preferred.

In a study evaluating the efficacy of IV ganciclovir therapy (4–6 mg/kg every 12 hours for 6 weeks) in neonates with symptomatic congenital HCMV infection, 69% of treated children, compared to only 39% of untreated children, showed improvement in brainstem auditory evoked potentials. More significantly, none of the ganciclovir-treated children, compared to 42% of those not treated, showed worsening hearing loss over the initial 6 months post infection ( ). In another study, 6 weeks of ganciclovir therapy in neonates with symptomatic congenital CMV infection involving the CMS led to fewer developmental delays at 6 and 12 months as compared to untreated infants ( ). While 6 weeks of therapy was beneficial, it was not known how long infants with congenital CNS HCMV disease should be treated. In a randomized, placebo-controlled study of 96 neonates, infants treated with valganciclovir (16 mg/kg, orally, twice daily) for 6 months were more likely to have normal hearing (73% vs. 57%) at 12- and 24-month follow-up and had improved neurodevelopmental scores when compared to infants treated with valganciclovir for 6 weeks and placebo for 4.5 months ( ). Thus, treating infants diagnosed with congenital HCMV infection for 6 months with valganciclovir may decrease the risk of hearing loss or neurodevelopmental abnormalities.

Cytomegalovirus in immunocompromised adults

HCMV can cause neuroinvasive infections in immunocompromised patients with hematopoietic stem cell transplants, solid organ transplants (SOTs), or in patients with acquired immunodeficiency syndrome (AIDS) ( ).

HCMV is an important pathogen following SOT and hematopoietic stem cell transplantation (HSCT). Risk factors for CNS disease in this group include delayed T-cell recovery, umbilical cord blood transplantation, graft versus host disease, and a history of recurrent HCMV viremia ( ). Antiviral prophylaxis for prevention of primary infection or recurrence is indicated when either or both donor (D+) and recipient (R+) are seropositive for HCMV. A phase 3, double-blind, randomized, placebo-controlled clinical trial recently showed that HSCT patients treated with letermovir prophylaxis (a new antiviral drug that inhibits the CMV-terminase complex) showed reductions in clinically significant CMV disease (37.5%) compared to the placebo group (60.6%) ( ). Use of valganciclovir or ganciclovir is often limited in the HSCT patient population due to the associated myelosuppressive side effects of treatment, so the availability of a less myelosuppressive alternative (letermovir) that can be more widely employed may decrease the risk of CNS disease in this patient population. Prior to the availability of letermovir, a preemptive treatment strategy was used in HSCT patients instead, defined as prevention of development of disease when reactivation has occurred. Patients were monitored for CMV viremia and started on therapy with antiviral treatment to prevent clinical CMV disease and minimize the toxic effects of antiviral drugs to engrafted bone marrow.

Despite recent advances, HCMV infection remains one of the most common complications affecting SOT patients. All patients undergoing SOT who are at risk for CMV disease receive universal prophylaxis, most commonly with valganciclovir starting within 10 days after transplantation and continuing for a period of 3–6 months.

HCMV encephalitis typically occurs in HIV+ patients with a CD4+ cell count of less than 50 cells/mm ³ , although patients with counts below 100/mm ³ are at increased risk of developing HCMV viremia ( ). In HIV patients with low CD4 counts in the era before highly active antiretroviral therapy (HAART), HCMV most commonly caused retinitis, esophagitis, and colitis and less commonly caused encephalitis, peripheral neuropathy, or polyradiculoneuritis. The increased use of HAART for the treatment of HIV has dramatically reduced the occurrence of HCMV viremia and disease.

HCMV encephalitis presents either as a microglial nodular encephalitis with acute onset of confusion and delirium, or as a more slowly progressive ventriculoencephalitis characterized by confusion and cranial nerve palsies. There is a broad pathological spectrum of HCMV infection of the brain, ranging from scattered microglial nodules to widespread necrotizing leukoencephalopathy or focal necrosis deep in the parenchyma.

Diagnosis of HCMV is largely dependent on quantitation of viral load in the blood using PCR and evidence of end organ disease or involvement consistent with HCMV infection. CSF HCMV PCR has a reported sensitivity of 82% and specificity of 99% in AIDS patients with CNS disease due to HCMV ( ). In immunosuppressed patients, a wide spectrum of neuroimaging results are reported, ranging from normal findings to detection of generalized atrophy, periventricular abnormalities, and focal discrete white-matter lesions. The most characteristic MRI finding is periventricular increased signal on T2-weighted images and ependymal enhancement following gadolinium administration on T1-weighted images. In advanced HIV disease, HCMV can manifest as polyradiculitis with diffuse enhancement of cauda equina nerve roots and is often associated with concurrent HCMV infection elsewhere in the body ( ). Many patients with this syndrome have an almost pathognomonic CSF profile of neutrophilic pleocytosis with a low glucose concentration. The diagnosis is confirmed by PCR amplification of HCMV DNA in CSF.

Ganciclovir, foscarnet, and cidofovir all have efficacy against HCMV in vitro and in some clinical settings in vivo . In all immune suppressed patients, immune reconstitution following introduction of HAART or decreasing immune suppressing medications may help control HCMV replication and disease. General recommendations for immunocompromised patients, including those with AIDS, are for initial antiviral therapy with IV ganciclovir (5 mg/kg every 12 hours for 2–3 weeks), followed by maintenance dosing with either IV ganciclovir (5 mg/kg/day, 5 days/wk) or oral valganciclovir (900 mg daily) for at least an additional 4 weeks. For SOT patients, oral valganciclovir or intravenous ganciclovir treatment are associated with similar long-term outcomes based on the VICTOR study conducted in adult renal, liver, heart, and lung transplant recipients ( ). However, patients with severe or life-threatening HCMV disease were excluded from this study, so intravenous ganciclovir is still recommended for CNS disease in this population. A full discussion of CMV resistance and treatment is beyond the scope of this chapter. Briefly, foscarnet should be reserved for treatment of ganciclovir-resistant HCMV because of its nephrotoxicity and IV administration ( ). Foscarnet dosage is 60 mg/kg IV, 3 times daily for 2–3 weeks for initial therapy, then 90–120 mg/kg IV daily for maintenance. There are limited clinical data on the use of cidofovir for treatment of HCMV CNS disease ( Table 78.8 ). In transplant cases, CNS disease is treated with IV ganciclovir, 5–7.5 mg/kg per dose 2 or 3 times daily with the addition of IV foscarnet 90 mg/kg twice daily for refractory cases ( ).