Comprehensive, up-to-date information on HIV/AIDS treatment, prevention, and policy from the University of California San Francisco

Infection with human cytomegalovirus (CMV) is common and usually inapparent; CMV can be acquired during infancy, early childhood, or adolescence. Transmission can occur vertically from an infected woman to her offspring or horizontally by contact with virus-containing breast milk, saliva, urine, or sexual fluid or through transfusion of infected blood or transplantation of infected organs. During infancy and early childhood, infection usually occurs secondary to ingestion of infected breast milk or exposure to infected saliva or urine. Infection occurs at younger ages in locations where sanitation is less than optimal. Among adolescents, sexual transmission is the major mode of CMV acquisition.

Age-related prevalence of infection varies widely, depending on living circumstances and social customs. Breast-feeding, child-rearing practices, crowding, sanitation, and sexual behavior most likely influence age-related variations in CMV prevalence. Where rates of maternal seropositivity are high and breast-feeding is common, more than half of infants acquire CMV during the first year of life (575Stagno S, Reynolds DW, Pass RF, et al. Breast milk and the risk of cytomegalovirus infection. N Engl J Med 1980;302:1073-6.). Group care of children facilitates spread of CMV, especially in toddlers, and leads to higher prevalence of infection in children who attend child care centers and in their caregivers (576Pass RF, Hutto C, Reynolds DW, et al. Increased frequency of cytomegalovirus infection in children in group day care. Pediatrics 1984;74:121-6., 577Adler SP. Cytomegalovirus and child day care. Evidence for an increased infection rate among day-care workers. N Engl J Med 1989;321:1290-6.). In Africa, Asia, and Latin America, most children are infected by CMV before adolescence. In the United States and western Europe, the prevalence of antibody to CMV in adults from middle and upper socioeconomic strata is 40%-60%, whereas the prevalence in low-income adults is ≥80% (578Pass RF. Epidemiology and transmission of cytomegalovirus infection. J Infect Dis 1985;152:243-8.). Overall, among U.S. women of childbearing age, the prevalence of CMV infection is 50%-80%, with the highest prevalence in women in lower socioeconomic strata (579Stagno S, Pass RF, Cloud G, et al. Primary cytomegalovirus infection in pregnancy. Incidence, transmission to fetus, and clinical outcome. JAMA 1986;256:1904-8., 580Yow MD, Williamson DW, Leeds LJ, et al. Epidemiologic characteristics of cytomegalovirus infection in mothers and their infants. Am J Obstet Gynecol 1998;158:1189-95.). The prevalence of CMV infection among HIV-infected pregnant women is higher than in the general population, with approximately 90% of HIV-infected pregnant women coinfected with CMV (581Mussi-Pinhata MM, Yamamoto AY, Figueiredo LT, et al. Congenital and perinatal cytomegalovirus infection in infants born to mothers infected with human immunodeficiency virus. J Pediatr 1998;132:285-90., 582Quinn TC, Piot P, McCormick JB, et al. Serologic and immunologic studies in patients with AIDS in North America and Africa. The potential role of infectious agents as cofactors in human immunodeficiency virus infection. JAMA 1987;257:2617-21.).

CMV is the most common congenitally transmitted infection, occurring in 0.2%-2.2% of live-born infants in the United States (583Prober CG, Enright AM. Congenital cytomegalovirus (CMV) infections: hats off to Alabama J Pediatr 2003;143:4-6.). Congenital (in utero) CMV infection occurs most commonly among infants born to women who have primary CMV infection during pregnancy. Following primary infection during pregnancy, the rate of transmission to the fetus is approximately 30%-40% (579Stagno S, Pass RF, Cloud G, et al. Primary cytomegalovirus infection in pregnancy. Incidence, transmission to fetus, and clinical outcome. JAMA 1986;256:1904-8., 584Revello MG, Zavattoni M, Furione M, et al. Diagnosis and outcome of preconceptional and periconceptional primary human cytomegalovirus infections. J Infect Dis 2002;186:553-7.). In comparison, the rate of congenital infection after nonprimary CMV infection usually is believed to be significantly lower (range: 0.15%-1.0%) (583Prober CG, Enright AM. Congenital cytomegalovirus (CMV) infections: hats off to Alabama J Pediatr 2003;143:4-6., 585Fowler KB, Stagno S, Pass RF. Maternal immunity and prevention of congenital cytomegalovirus infection. JAMA 2003;289:1008-11., 586Azam AZ, Vial Y, Fawer CL, et al. Prenatal diagnosis of congenital cytomegalovirus infection. Obstet Gynecol 2001;97:443-8.). More recent studies demonstrate that in utero transmission of nonprimary maternal infection can occur because of reactivation of infection among women infected before pregnancy or reinfection with a different CMV strain among CMV-seropositive women (587Boppana SB, Fowler KB, Britt WJ, et al. Symptomatic congenital cytomegalovirus infection in infants born to mothers with preexisting immunity to cytomegalovirus. Pediatrics 1999;104:55-60., 588Boppana SB, Rivera LB, Fowler KB, et al. Intrauterine transmission of cytomegalovirus to infants of women with preconceptional immunity. N Engl J Med 2001;344:1366-71.); these studies challenge the traditional understanding of transmission risk and clinical outcomes following nonprimary maternal CMV infection.

CMV also can be transmitted during the intrapartum or postpartum periods from mother to infant. Up to 57% of infants whose mothers shed CMV at or around delivery become infected with CMV, and up to 53% of children who are breast-fed milk containing infectious virus can become CMV infected. Symptomatic CMV disease in the infant is much less common when CMV is acquired intrapartum or through breast-feeding than when acquired antenatally and occurs primarily in premature neonates.

HIV-infected women with CMV infection have a higher rate of CMV shedding from the cervix than do women without HIV infection (52%-59% and 14%-35%, respectively) (589Mostad SB, Kreiss JK, Ryncarz A, et al. Cervical shedding of herpes simplex virus and cytomegalovirus throughout the menstrual cycle in women infected with human immunodeficiency virus type 1. Am J Obstet Gynecol 2000;183:948-55.). The risk for mother-to-infant transmission of CMV may be higher among infants born to women dually infected with CMV and HIV. In one study of 440 infants born to HIV-infected U.S. women, the overall rate of in utero infection was 4.5% (11Kovacs A, Schluchter M, Easley K, et al. Cytomegalovirus infection and HIV-1 disease progression in infants born to HIV-1-infected women. Pediatric Pulmonary and Cardiovascular Complications of Vertically Transmitted HIV Infection Study Group. N Engl J Med 1999;341:77-84.), higher than the <2% rate of in utero infection in the general U.S. population.

HIV-infected children appear to be at higher risk for CMV infection during early childhood than do HIV-uninfected children (11Kovacs A, Schluchter M, Easley K, et al. Cytomegalovirus infection and HIV-1 disease progression in infants born to HIV-1-infected women. Pediatric Pulmonary and Cardiovascular Complications of Vertically Transmitted HIV Infection Study Group. N Engl J Med 1999;341:77-84.). The rate of CMV acquisition in HIV-infected children appears to be particularly high during the first 12 months of life but remains higher among HIV-infected than HIV-uninfected children through age 4 years when they are exposed to CMV infection in other children in child care or school.

CMV disease occurs less frequently among HIV-infected children than HIV-infected adults but contributes substantially to morbidity and mortality. In the pre-antiretroviral era, CMV caused 8%-10% of pediatric AIDS-defining illness (590Kitchen BJ, Engler HD, Gill VJ, et al. Cytomegalovirus infection in children with human immunodeficiency virus infection. Pediatr Infect Dis J 1997;16:358-63.). Data in HIV-infected adults have shown a 75%-80% decrease in the incidence of new cases of CMV end-organ disease with the advent of antiretroviral therapy, with an incidence now estimated to be <6 cases per 100 person-years (591Jabs DA, Van Natta ML, Holbrook JT, et al. Longitudinal study of the ocular complications of AIDS: 1. Ocular diagnoses at enrollment. Ophthalmology 2007;114:780-6.). In a study of OIs in approximately 3000 children followed in PACTG studies during the pre-HAART era, the frequency of CMV retinitis was 0.5 cases per 100 child-years and, of other CMV disease, 0.2 cases per 100 child-years (1Dankner WM, Lindsey JC, Levin MJ, et al. Correlates of opportunistic infections in children infected with the human immunodeficiency virus managed before highly active antiretroviral therapy. Pediatr Infect Dis J 2001;20:40-8.). The rate varied significantly by CD4 percentage; the incidence of CMV retinitis was 1.1 cases per 100 child-years in children with CD4 <15%, compared with 0.1 case per 100 child-years in children with CD4 >25%. In data from the same cohort during the HAART era, the overall rate of CMV retinitis was <0.5 per 100 child-years (3). In the Perinatal AIDS Collaborative Transmission Study, the incidence of nonocular CMV before and after January 1997 (pre- and post-HAART eras) was 1.4 per 100 child-years and 0.1 per 100 child-years, respectively (4Nesheim SR, Kapogiannis BG, Soe MM, et al. Trends in opportunistic infections in the pre- and post-highly active antiretroviral therapy eras among HIV-infected children in the Perinatal AIDS Collaborative Transmission Study, 1986-2004. Pediatrics 2007;120:00 100–9.). Similarly, CMV retinitis declined from 0.7 to 0.0 per 100 child-years during the same period.

Symptomatic HIV-infected children coinfected with CMV have a higher rate of CMV viruria than do asymptomatic HIV-infected or HIV-exposed children. Overall, up to 60% of children with AIDS shed CMV. This compares with one third of HIV-infected children shedding CMV; 15%-20% of CMVinfected, HIV-exposed but uninfected children; and <15% of CMV-infected infants not exposed to HIV (592Chandwani S, Kaul A, Bebenroth D, et al. Cytomegalovirus infection in human immunodeficiency virus type 1-infected children. Pediatr Infect Dis J 1996;15:310-4.).

Approximately 10% of infants with in utero CMV infection are symptomatic at birth with congenital CMV syndrome (CMV inclusion disease); mortality of children with symptomatic disease is as high as 30%. About half of neonates with symptomatic congenital CMV disease are small for gestational age; additional findings may include petechiae (76%), jaundice (67%), hepatosplenomegaly (60%), chorioretinitis (20%), microcephaly (53%), intracranial calcifications (55%), and hearing impairment (≤65%) (593Boppana SB, Pass RF, Britt WJ, et al. Symptomatic congenital cytomegalovirus infection: neonatal morbidity and mortality. Pediatr Infect Dis J 1992;11:93-9., 594Fowler KB, Boppana SB. Congenital cytomegalovirus (CMV) infection and hearing deficit. J Clin Virol 2006;35:226-31.). Approximately 90% of infants with symptomatic disease at birth who survive have late complications, including substantial hearing loss, mental retardation, chorioretinitis, optic atrophy, seizures, or learning disabilities (579Stagno S, Pass RF, Cloud G, et al. Primary cytomegalovirus infection in pregnancy. Incidence, transmission to fetus, and clinical outcome. JAMA 1986;256:1904-8.). Although most children with in utero CMV infection do not have symptoms at birth, 10%-15% are at risk for later developmental abnormalities, sensorineural hearing loss, chorioretinitis, or neurologic defects.

HIV-infected children coinfected with CMV appear to have faster progression of HIV disease than do those without CMV infection (11Kovacs A, Schluchter M, Easley K, et al. Cytomegalovirus infection and HIV-1 disease progression in infants born to HIV-1-infected women. Pediatric Pulmonary and Cardiovascular Complications of Vertically Transmitted HIV Infection Study Group. N Engl J Med 1999;341:77-84., 590Kitchen BJ, Engler HD, Gill VJ, et al. Cytomegalovirus infection in children with human immunodeficiency virus infection. Pediatr Infect Dis J 1997;16:358-63., 595Doyle M, Atkins JT, Rivera-Matos IR. Congenital cytomegalovirus infection in infants infected with human immunodeficiency virus type 1. Pediatr Infect Dis J 1996;15:1102-6.). In one study from the pre-HAART era, 53% of infants coinfected with HIV and CMV had progression to AIDS or had died by age 18 months, compared with 22% of HIV-infected children without CMV infection; those with HIV/CMV coinfection also were more likely to have CNS manifestations (36% versus 9%). The relative risk for HIV disease progression in children coinfected with CMV compared with children without CMV was 2.6 (95% CI: 1.1-6.0) (11Kovacs A, Schluchter M, Easley K, et al. Cytomegalovirus infection and HIV-1 disease progression in infants born to HIV-1-infected women. Pediatric Pulmonary and Cardiovascular Complications of Vertically Transmitted HIV Infection Study Group. N Engl J Med 1999;341:77-84.).

CMV retinitis is the most frequent severe manifestation of CMV disease among HIV-infected children, accounting for approximately 25% of CMV AIDS-defining illnesses. CMV retinitis among young HIV-infected children is frequently asymptomatic and discovered on routine examination. Older children with CMV retinitis present similarly to adults, with floaters, loss of peripheral vision, or reduction in central vision. Diagnosis of CMV retinitis is based on clinical appearance with white and yellow retinal infiltrates and associated retinal hemorrhages. A more indolent, granular retinitis also can occur. HIV-infected children with CD4 counts <100 cells/mm³ are more likely than those with higher CD4 counts to develop CMV retinitis; however, CD4 count is less predictive of risk for CMV disease in young infants, and systemic and localized CMV disease can occur in HIV-infected infants with higher, age-adjusted CD4 counts (592Chandwani S, Kaul A, Bebenroth D, et al. Cytomegalovirus infection in human immunodeficiency virus type 1-infected children. Pediatr Infect Dis J 1996;15:310-4., 596Zaknun D, Zangerle R, Kapelari K, et al. Concurrent ganciclovir and foscarnet treatment for cytomegalovirus encephalitis and retinitis in an infant with acquired immunodeficiency syndrome: case report and review. Pediatr Infect Dis J 1997;16:807-11.).

End-organ CMV disease has been reported in the lung, liver, GI tract, pancreas, kidney, sinuses, and CNS (596Zaknun D, Zangerle R, Kapelari K, et al. Concurrent ganciclovir and foscarnet treatment for cytomegalovirus encephalitis and retinitis in an infant with acquired immunodeficiency syndrome: case report and review. Pediatr Infect Dis J 1997;16:807-11., 597Mueller BU, MacKay K, Cheshire LB, et al. Cytomegalovirus ureteritis as a cause of renal failure in a child infected with the human immunodeficiency virus. Clin Infect Dis 1995;20:1040-3., 598Olivero MT, Nelson RP Jr, Andrews T, et al. Cytomegalovirus sinus disease in a human immunodeficiency virus-infected child. Pediatr Infect Dis J 1995;14:629-31., 599Marriage SC, Booy R, Hermione Lyall EG, et al. Cytomegalovirus myelitis in a child infected with human immunodeficiency virus type 1. Pediatr Infect Dis J 1996;15:549-51.). In children with extraocular CMV disease, predominantly nonspecific symptoms (e.g., fever, poor weight gain, and loss of developmental milestones with laboratory abnormalities of anemia, thrombocytopenia, and elevated lactic dehydrogenase) are initially observed, although the extent to which CMV or HIV infection themselves contribute to these findings is unclear (592Chandwani S, Kaul A, Bebenroth D, et al. Cytomegalovirus infection in human immunodeficiency virus type 1-infected children. Pediatr Infect Dis J 1996;15:310-4.). GI manifestations among HIV-infected children include CMV colitis (the most common GI manifestation), oral and esophageal ulcers, hepatic involvement, ascending cholangiopathy, or gastritis. Odynophagia is a common presentation of CMV esophagitis, whereas abdominal pain and hematochezia frequently occur with CMV colitis. Sigmoidoscopy in CMV colitis is nonspecific, demonstrating diffuse erythema, submucosal hemorrhage, and diffuse mucosal ulcerations. Esophageal or colonic ulcerations may cause perforation or hemorrhage. The role of CMV in pulmonary disease among HIV-infected children is difficult to assess because it often is isolated with other organisms (e.g., P. jirovecii). Histologic evidence of CMV disease is needed to determine whether active disease is present. CMV pneumonia is an interstitial process with gradual onset of shortness of breath and dry, nonproductive cough; auscultatory findings may be minimal.

CNS manifestations of CMV include subacute encephalopathy, myelitis, and polyradiculopathy (primarily observed in adults but rarely reported in children). The subacute or chronic encephalopathy of CMV can be difficult to differentiate clinically from HIV dementia, with symptoms of confusion and disorientation attributable to cortical involvement. Focal signs can be attributed to lesions in the brainstem. CSF findings are nonspecific and might indicate a polymorphonuclear predominance (>50% of patients), elevated protein (75%), and low glucose (30%). However, up to 20% of children with CMV CNS involvement have completely normal CSF indices. CMV also can cause a rapidly progressive, often fatal, CNS disease with defects in cranial nerves, nystagmus, and increasing ventricular size (600Kalayjian RC, Cohen ML BR, Flanigan TP. Cytomegalovirus ventriculoencephalitis in AIDS. A syndrome with distinct clinical and pathologic features. Medicine (Baltimore) 1993;72:67-77.).

CMV infection (versus disease) can be difficult to differentiate in HIV-infected children. Because of transplacental transfer of antibody from mother to child, a positive CMV antibody assay in an infant aged <12 months can indicate maternal infection but not necessarily infection of the infant. In an infant aged >12 months, a positive CMV antibody assay indicates previous infection with CMV but not necessarily active disease.In children of any age, a positive CMV culture or PCR indicates infection but not necessarily disease.

CMV can be isolated in cell culture from peripheral blood leukocytes, body fluids (e.g., urine), or tissues. Using centrifugation- assisted shell vial culture amplification techniques, CMV can be detected within 16-40 hours of culture inoculation. A positive blood buffy-coat culture establishes CMV infection and increases the likelihood that disease or symptoms were caused by CMV because children with positive blood cultures are at higher risk for end-organ disease.

Different methods have been used to detect viral antigen or DNA directly and to identify patients at risk for CMV disease, including detection of pp65 antigenemia, qualitative and quantitative PCR, and DNA hybridization. The DNA assays are more sensitive than buffy-coat or urine cultures for detecting CMV and can be used to identify patients at higher risk for clinically recognizable disease. CMV DNA detection in CSF by DNA PCR is highly sensitive for CMV CNS disease. Quantitative DNA PCR can be used as a marker of risk for disease and to monitor response to therapy (601Nigro G, Krzysztofiak A, Gattinara GC, et al. Rapid progression of HIV disease in children with cytomegalovirus DNAemia. AIDS 1996;10:1127-33.).

Recovery of virus from tissues (e.g., endoscopically guided biopsies of GI or pulmonary tissue) provides evidence of disease in symptomatic patients. The limitation of this method is that detection of visible cytopathic effects in cell culture takes 1-6 weeks. Staining of shell vial culture with CMV monoclonal antibodies or immunostaining for CMV antigens can allow earlier diagnosis of infection. Histopathology demonstrates characteristic “owl's eye” intranuclear and smaller intracytoplasmic inclusion bodies in biopsy specimens; staining with CMV monoclonal antibodies or immunostaining for CMV antigens also can be done. The same procedures can be used on cells obtained from bronchoalveolar lavage.

Because primary prophylaxis with antiviral agents in persons coinfected with CMV and HIV is not recommended (as discussed above), no consideration of discontinuing primary prophylaxis is necessary.

CMV retinitis should be managed in concert with an experienced ophthalmologist. Recommendations for HIV-infected adults include indirect ophthalmoscopy through a dilated pupil performed at diagnosis of CMV retinitis, after completion of induction therapy, 1 month after initiation of therapy, and monthly thereafter while the patient is on anti-CMV treatment; recommendations should be similar for HIV-infected children with CMV retinitis (AIII). Monthly fundus photographs, using a standardized photographic technique that documents the appearance of the retina, provide the optimum method for following patients and detecting early relapse (AIII). For patients who have experienced immune recovery (Table 3), the frequency of ophthalmologic follow-up can be decreased to every 3 months. However, because relapse of the retinitis occurs among patients with immune recovery, regular ophthalmologic follow-up still is needed.

The major side effects of ganciclovir and valganciclovir are myelosuppression (i.e., anemia, neutropenia, and thrombocytopenia) and renal toxicity. Dose reduction or interruption because of hematologic toxicity might be necessary in up to 40% of patients receiving IV ganciclovir; granulocyte colony-stimulating factor can be used to ameliorate marrow suppression. The main toxicities of foscarnet are decreased renal function and metabolic derangements. Renal toxicity and foscarnet binding to divalent metal ions, such as calcium, lead to metabolic abnormalities in approximately one third of patients, and serious electrolyte imbalances (including abnormalities in calcium, phosphorus, magnesium, and potassium levels) and secondary seizures, cardiac dysrhythmias, abnormal liver transaminases, and CNS symptoms can occur. Metabolic disturbances can be minimized if foscarnet is administered by slow infusion, with rates not exceeding 1 mg/kg/minute. Concomitant use of other nephrotoxic drugs increases the likelihood of renal dysfunction associated with foscarnet therapy. For patients receiving ganciclovir or foscarnet, complete blood counts and serum electrolytes and renal function should be monitored twice weekly during induction therapy and once weekly thereafter (AIII).

The major side effect of cidofovir is potentially irreversible nephrotoxicity; the drug produces proximal tubular dysfunction including Fanconi syndrome and acute renal failure. Renal toxicity manifests as proteinuria and glycosuria. To minimize nephrotoxicity, probenicid should be administered before each infusion, and IV hydration with normal saline should be administered before and after each cidofovir infusion. For patients receiving IV cidofovir, blood urea nitrogen, creatinine, and urinalysis should be performed before each infusion; administration of the drug is contraindicated if renal dysfunction or proteinuria is detected. Other reported adverse events include anterior uveitis and ocular hypotony; serial ophthalmologic monitoring for anterior segment inflammation and intraocular pressure is needed while receiving the drug systemically. Cidofovir should not be administered concomitantly with other nephrotoxic agents. Cidofovir therapy must be discontinued if serum creatinine increases ≥0.5 mg/dL above baseline.

Immune recovery uveitis after initiation of effective HAART is an immunologic reaction to CMV associated with inflammation in the anterior chamber and/or the vitreous (614Nguyen QD, Kempen JH, Bolton SG, et al. Immune recovery uveitis in patients with AIDS and cytomegalovirus retinitis after highly active antiretroviral therapy. Am J Ophthalmol 2000;129:634-9.). Ocular complications of uveitis include macular edema and development of epiretinal membranes, which can cause loss of vision. Immune recovery uveitis may respond to periocular corticosteroids or a short course of systemic steroids. Oral valganciclovir was beneficial in one small uncontrolled study (615Kosobucki BR, Goldberg DE, Bessho K, et al. Valganciclovir therapy for immune recovery uveitis complicated by macular edema. Am J Ophthalmol 2004;137:636-8.).

Resistant strains of CMV should be suspected when progressive disease and continued recovery of virus occurs despite ganciclovir therapy. Foscarnet is the drug of choice when ganciclovir resistance is suspected (AI).

In patients with CMV retinitis, although drug resistance occurs among patients receiving long-term therapy, early relapse might be caused by the limited intraocular penetration of systemically administered drugs; in HIV-infected adults, the placement of a ganciclovir implant in a patient whose retinitis has relapsed during systemic treatment is recommended because it achieves greater drug levels in the eye and often will control the retinitis for 6-8 months until the implant requires replacement (BIII). Because of the size requirements of the implants, this option would be limited to older children with CMV retinitis. Many experts would initially treat early first relapse of retinitis with reinduction with the same drug, followed by reinstitution of maintenance therapy (AII). However, if drug resistance is suspected or if side effects or toxicities interfere with optimal courses of the initial agent, change to an alternative drug is reasonable (AIII). Combination ganciclovir and foscarnet can be considered but is accompanied by greater toxicity.

Courses of antiviral agents (e.g., ganciclovir, foscarnet, or cidofovir) do not cure CMV disease. After induction therapy, the standard recommendation has been to provide secondary prophylaxis (chronic maintenance therapy) for the remainder of the patient's life (AI). Regimens that can be considered for chronic suppression in adults and adolescents include IV ganciclovir, oral valganciclovir, IV foscarnet, combined IV ganciclovir and foscarnet, parenteral cidofovir, and (for retinitis only) ganciclovir administration by intraocular implant (AI) (616Martin DF, Kuppermann BD, Wolitz RA, et al. Oral ganciclovir for patients with cytomegalovirus retinitis treated with a ganciclovir implant. Roche Ganciclovir Study Group. N Engl J Med 1999;340:1063-70., 617Drew WL, Ives D, Lalezari JP, et al. Oral ganciclovir as maintenance treatment for cytomegalovirus retinitis in patients with AIDS. Syntex Cooperative Oral Ganciclovir Study Group. N Engl J Med 1995;333:615-20., 618Parenteral cidofovir for cytomegalovirus retinitis in patients with AIDS: the HPMPC peripheral cytomegalovirus retinitis trial. A randomized, controlled trial. Studies of Ocular complications of AIDS Research Group in Collaboration with the AIDS Clinical Trials Group. Ann Intern Med 1997;126:264-74., 619Palestine AG, Polis MA, De Smet MD, et al. A randomized, controlled trial of foscarnet in the treatment of cytomegalovirus retinitis in patients with AIDS. Ann Intern Med 1991;115:665-73., 620Spector SA, Weingeist T, Pollard RB, et al. A randomized, controlled study of intravenous ganciclovir therapy for cytomegalovirus peripheral retinitis in patients with AIDS. AIDS Clinical Trials Group and Cytomegalovirus Cooperative Study Group. J Infect Dis 1993;168:557-63., 621The Studies of the Ocular Complications of AIDS Research Group, Group icwtACT. Combination foscarnet and ganciclovir therapy vs monotherapy for the treatment of relapsed cytomegalovirus retinitis in patients with AIDS. The cytomegalovirus retreatment trial. Arch Ophthalmol 1996;114:23-33., 622Diaz-Llopis M, España E, Muñoz G, et al. High dose intravitreal foscarnet in the treatment of cytomegalovirus retinitis in AIDS. Br J Ophthalmol 1994;78:120-4., 623de Smet MD, Meenken CJ, van den Horn GJ. Fomivirsen- a phosphorothioate oligonucleotide for the treatment of CMV retinitis. Ocul Immunol Inflamm 1999;7:189-98.). Because of limited data on drug pharmacokinetics and dosing in children, IV ganciclovir or foscarnet are the preferred secondary prophylaxis regimens for children; oral valganciclovir can be considered for older children able to receive adult dosing. Repetitive intravitreous injections of ganciclovir, foscarnet, and cidofovir reportedly are effective for secondary prophylaxis of CMV retinitis (624Kirsch LS, Arevalo JF, Chavez de la Paz E, et al. Intravitreal cidofovir (HPMPC) treatment of cytomegalovirus retinitis in patients with acquired immune deficiency syndrome. Ophthalmology 1995;102: 533-42; discussion 542-3., 625Young S, Morlet N, Besen G, et al. High-dose (2000-microgram) intravitreous ganciclovir in the treatment of cytomegalovirus retinitis. Ophthalmology 1998;105:1404-10.), although intraocular therapy alone does not protect the contralateral eye or other organ systems and therefore typically is combined with systemic treatment (616Martin DF, Kuppermann BD, Wolitz RA, et al. Oral ganciclovir for patients with cytomegalovirus retinitis treated with a ganciclovir implant. Roche Ganciclovir Study Group. N Engl J Med 1999;340:1063-70.). Additionally, frequent intravitreous injections are impractical for use in most children (DIII).

A chronic maintenance regimen for patients treated for CMV disease should be chosen in consultation with a specialist. Chronic maintenance therapy is not routinely recommended for GI disease but should be considered if relapses occur (BII). A role for maintenance therapy for CMV pneumonitis has not been established (CIII). For patients with retinitis, decisions should be made in consultation with an ophthalmologist and should take into consideration the anatomic location of the retinal lesion, vision in the contralateral eye, and the immunologic and virologic status of the patient (BIII). Intraocular implants should not be used in children aged <3 years because of the small size of their eyes (EIII).

Multiple case series have reported that maintenance therapy can be discontinued safely in adults and adolescents with CMV retinitis whose CD4 counts have increased substantially in response to HAART(626Tural C, Romeu J, Sirera G, et al. Long-lasting remission of cytomegalovirus retinitis without maintenance therapy in human immunodeficiency virus-infected patients. J Infect Dis 1998;177: 1080-3., 627Vrabec TR, Baldassano VF, Whitcup SM. Discontinuation of maintenance therapy in patients with quiescent cytomegalovirus retinitis and elevated CD4+ counts. Ophthalmology 1998;105:1259-64., 628Macdonald JC, Torriani FJ, Morse LS, et al. Lack of reactivation of cytomegalovirus (CMV) retinitis after stopping CMV maintenance therapy in AIDS patients with sustained elevations in CD4 T cells in response to highly active antiretroviral therapy. J Infect Dis 1998;177: 1182-7., 629Whitcup SM, Fortin E, Lindblad AS, et al. Discontinuation of anticytomegalovirus therapy in patients with HIV infection and cytomegalovirus retinitis. JAMA 1999;282:1633-7., 630Jabs DA, Bolton SG, Dunn JP, et al. Discontinuing anticytomegalovirus therapy in patients with immune reconstitution after combination antiretroviral therapy. Am J Ophthalmol 1998;126:817-22., 631Jouan M, Saves M, Tubiana R, et al. Discontinuation of maintenance therapy for cytomegalovirus retinitis in HIV-infected patients receiving highly active antiretroviral therapy. AIDS 2001;15:23-31.). These patients have remained disease free for <30 to 95 weeks, whereas during the pre-HAART era, retinitis typically reactivated in <6 to 8 weeks after stopping CMV therapy. Plasma HIV RNA levels varied among these patients, supporting the hypothesis that the CD4 count is the primary determinant of immune recovery to CMV. CMV retinitis can occur in HAART-treated adults with high CD4 counts, however (632Torriani FJ, Freeman WR, Macdonald JC, et al. CMV retinitis recurs after stopping treatment in virological and immunological failures of potent antiretroviral therapy. AIDS 2000;14:173-80. Vol. 58 / RR-11 Recommendations and Reports 117), suggesting that CMV-specific cellular immunity may be important in controlling CMV in immunereconstituted HIV-infected adults (633Lilleri D, Piccinini G, Genini E, et al. Monitoring of human cytomegalovirus (HCMV)-specific CD4+ T cell frequency by cytokine flow cytometry as a possible indicator for discontinuation of HCMV secondary prophylaxis in HAART-treated AIDS patients. J Clin Virol 2004;29:297-307., 634Tamarit A, Alberola J, Mira JV, et al. Assessment of human cytomegalovirus specific T cell immunity in human immunodeficiency virus infected patients in different disease stages following HAART and in long-term non-progressors. J Med Virol 2004;74:382-9.). In HIV-infected adults with CMV retinitis, discontinuing secondary prophylaxis is considered for patients with a sustained increase in CD4 count to >100 cells/mm³ in response to treatment.

The safety of discontinuing secondary prophylaxis after immune reconstitution with HAART in HIV-infected children has not been as well studied. Low or undetectable HIV replication in children is the strongest correlate with CMV immune reconstitution, being associated with a higher frequency of CMV-specific CD4 cells (635Weingberg A, Wiznia AA, Lafleur BJ, et al. Cytomegalovirus-specific cell-mediated immunity in HIV-infected children on HAART. AIDS Res Hum Retroviruses 2006;22:283-8.). Early institution of HAART may help control CMV infection by maintaining normal CD4 count and cytotoxic T-lymphocyte responses in HIV-infected children (636Saitoh A, Viani RM, Schrier RD, et al. Treatment of infants coinfected with HIV-1 and cytomegalovirus with combination antiretrovirals and ganciclovir. J Allergy Clin Immunol 2004;114:983-5.). In deciding whether to discontinue secondary prophylaxis, the significant toxicities associated with antiviral drugs active against CMV, including those in in vitro and animal models, need to be considered.

Recognizing the limitations of the data in children but drawing on the growing experience in adults, discontinuing prophylaxis may be considered for children aged 1-5 years who are receiving HAART and have a sustained (e.g., >6 months) increase in CD4 count to >500 cells/mm³ or CD4 percentage to >15%, and for children aged ≥6 years, an increase in CD4 count to >100 cells/mm³ as for adults (CIII). Such decisions should be made in close consultation with an ophthalmologist and should account for such factors as magnitude and duration of CD4 cell increase, anatomic location of the retinal lesion, vision in the contralateral eye, and the feasibility of regular ophthalmologic monitoring (CIII).

All patients in whom anti-CMV maintenance therapy has been discontinued should continue to undergo regular ophthalmologic monitoring at at least 3- to 6-month intervals for early detection of CMV relapse and for immune reconstitution uveitis (AII). CMV viral load or other markers of CMV infection (e.g., antigenemia or viral DNA tests) are not well standardized; their role in predicting relapse remains to be defined, and they are not recommended for routine monitoring (DIII) (637Spector SA, Wong R, Hsia K, et al. Plasma cytomegalovirus (CMV) DNA load predicts CMV disease and survival in AIDS patients. J Clin Invest 1998;101:497-502., 638Salmon-Ceron D, Mazeron MC, Chaput S, et al. Plasma cytomegalovirus DNA, pp65 antigenaemia and a low CD4 cell count remain risk factors for cytomegalovirus disease in patients receiving highly active antiretroviral therapy. AIDS 2000;14:1041-9.).

CMV retinitis relapses among adults whose anti-CMV maintenance therapies have been discontinued and whose CD4 counts have decreased to <50 cells/mm³ (624Kirsch LS, Arevalo JF, Chavez de la Paz E, et al. Intravitreal cidofovir (HPMPC) treatment of cytomegalovirus retinitis in patients with acquired immune deficiency syndrome. Ophthalmology 1995;102: 533-42; discussion 542-3.); reinstitution of secondary prophylaxis is recommended for HIV-infected adults when CD4 count falls to <100 cells/mm³. For HIV-infected children in whom secondary prophylaxis has been discontinued because of immune reconstitution, secondary prophylaxis should be reinstituted in children aged 1-5 years when the CD4 count decreases to <500 cells/mm³ or CD4 percentage to <15%, and for children aged ≥6 years when the CD4 count decreases to <100 cells/mm³ or CD4 percentage to <15% (BIII).