In vitro, beta-herpesviruses can stimulate or inhibit HIV replication under particular circumstances. In order to investigate the effects of beta-herpesvirus infection on HIV replication and vice versa at an organ level, we determined the quantitative relationships between cytomegalovirus (CMV), human herpesviruses (HHV) 6 and 7, and HIV-1 proviral DNA using quantitative competitive PCR methods in 141 organs collected at autopsy from 11 AIDS patients. The presence of HHV-6 DNA in an organ was significantly associated with elevated HIV-1 proviral DNA (difference in HIV median loads, 1.3 log10 genomes; P = 0.004). Consistent with this, there was a trend for the presence of HIV-1 proviral DNA to be associated with an elevated HHV-6 load (0.44 log10 difference; P = 0.07). In contrast, there were no significant differences between viral loads in the combinations of either CMV or HHV-7 with HIV-1 proviral DNA load. Pairwise combinations of the beta-herpesviruses revealed that the quantity of HHV-7 was increased in the presence of HHV-6 (difference in median loads, 1.3 log10; P = 0.001) and the quantity of HHV-6 was increased in the presence of HHV-7 (difference in median loads, 0.7 log10; P=0.002). These results demonstrate that the presence of HHV-6 in an organ is significantly associated with an elevated HIV-1 proviral load and have implications for understanding HIV pathogenesis in the human host and the role that beta-herpesviruses, especially HHV-6, might play as cofactors in the HIV disease process.

Five AIDS patients with cytomegalovirus (CMV) retinitis who had received ganciclovir (GCV) therapy were followed with serial blood sampling to detectchanges both in CMV load and in the genetic composition of genes UL97 and UL54 whilst receiving cidofovir (CDV) therapy. CDV neither reduced CMV load in blood nor prevented its quantitative resurgence during therapy. These effects were not explained by the initial presence or development of CDV-associated drug resistance mutations in UL54. In two patients, UL97 genotypic resistance to GCV involving either a L595S mutation or a deletion of amino acids 590-603 were present at the initiation of CDV and, in both patients, repopulation of CMV strains with wild-type UL97 sequences occurred during CDV therapy. These data are consistent with GCV-resistant strains containing UL97 mutations being less fit than their wild-type counterparts and so being able to persist only with the selective pressure of GCV.

Reactivation of cytomegalovirus (CMV) following immunosuppression may result in the development of CMV disease and is associated with an increased risk of death. CMV viraemia detected by the polymerase chain reaction (PCR) precedes CMV disease in HIV-infected patients and identifies individuals at high risk of disease. Pre-emptive ganciclovir (GCV) therapy in patients who have evidence of CMV viraemia is effective in preventing disease. An open study was conducted to assess the response of CMV viraemia to oral GCV at a dose of 3 or 6 g/day for 28 days. HIV RNA was measured to determine if CMV inhibition affected HIV viral load. Fourteen patients were studied, three of whom entered both phases of the study. None of the patients had evidence of CMV disease at the time of entry into the trial; two patients developed CMV retinitis after completion of the trial. Oral GCV at both 3 and 6 g/day caused a decrease in CMV viral load in individual patients. However, a rebound in CMV viral load occurred in patients receiving the 3-g/day dose. None of the patients receiving oral GCV 3 g/day became PCR negative after 21 days compared with six of eight patients receiving 6 g/day. Five of eight patients (63%) receiving GCV 6 g/day were concurrently taking protease inhibitors compared with two of nine (22%) receiving 3 g/day. Ten patients remained PCR negative throughout follow up. No change was found in HIV viral load during receipt of GCV at either dose. Thus, oral GCV is effective in reducing CMV viral load, but a dose of 3 g/day is insufficiently potent for pre-emptive therapy.

Cytomegalovirus (CMV) is generally described as a slowly replicating virus. During studies of immunocompromised patients, we observed rapid changes in the quantity of CMV DNA present in serial blood samples by quantitative-competitive polymerase chain reaction commensurate with a doubling time of <2 d. To further investigate the dynamics of replication in vivo, patients in three distinct situations were studied in detail: (a) those receiving intravenous ganciclovir; (b) those in whom ganciclovir-resistant strains appeared during long-term therapy; and (c) those in whom ganciclovir-resistant strains disappeared with alternative drug therapy. In all cases, it was possible to provide accurate estimates of the doubling time of CMV and its half-life of disappearance after antiviral chemotherapy. The results from all three approaches demonstrated that the doubling time/half-life of CMV in blood is ∼1 d when frequent samples are collected. These results show that CMV DNA replication in vivo is a highly dynamic process. We conclude that the reputation of CMV as a slowly replicating virus based on the time taken to produce cytopathic effects in vitro is unwarranted. These findings have implications for the potency, dose, and duration of antiviral chemotherapy needed for the effective treatment of this important human pathogen.

Objectives: To determine whether recurrence of polymerase chain reaction (PCR) viraemia during maintenance ganciclovir for cytomegalovirus (CMV) retinitis correlates with (i) CMV disease at a new anatomical site, (ii) progression of the presenting retinitis, or (iii) acquisition of genetic changes in gene UL97 associated with resistance to ganciclovir. Design: A previously described cohort of 45 patients presenting with first episode retinitis was followed clinically using ophthalmoscopy and serial tests for PCR viraemia for a median of 7 months. CMV viral load and genetic markers of ganciclovir resistance were measured in PCR-positive samples. Methods: PCR amplification of the glycoprotein B region of CMV and quantitative competitive PCR assays were employed. Genetic changes in UL97 were identified by sequencing/point mutation assay. Results: PCR viraemia correlated significantly with new episodes of CMV disease (P = 0.011) and a trend was seen for the association with progression of retinitis (P = 0.07). Amongst the 14 patients PCR-positive during maintenance ganciclovir, 10 (71%) had genetic markers of resistance. None of these patients became PCR-negative in blood after reinduction ganciclovir therapy compared with three out of four without markers of resistance (P = 0.022). Conclusions: CMV PCR viraemia correlated strongly with the development of new episodes of CMV disease. Most patients with progression of retinitis remained PCR-negative in blood, consistent with therapeutic failure due to poor intraocular penetration of ganciclovir. However, the minority who were PCR-positive in blood may have reinfected their eye, and frequently had markers of ganciclovir resistance. The implications of these findings for the management of patients with CMV disease are discussed. © 1998 Lippincott-Raven Publishers

Background: Cytomegalovirus (CMV) disease is a major cause of morbidity in patients with HIV infection. Despite treatment, CMV retinitis causes substantial visual loss, especially in patients with CD4 cell counts below 50 x 10(6)/l. Although routine ophthalmological screening of these patients has been recommended, no controlled trials have evaluated how frequently it should be performed. The aim of this study was to assess whether CMV polymerase chain reaction (PCR) results could direct ophthalmological screening to patients at high risk of CMV retinitis.Methods: In a prospective study of HIV-positive patients with CD4 cell counts below 50 x 10(6)/l, CMV viraemia was detected by qualitative PCR of whole blood. Patients who were CMV PCR-viraemic were allocated to monthly virological and ophthalmological follow-up; patients who were PCR-negative received 3-monthly virological and ophthalmological follow-up. CMV viral load was determined in all CMV-positive samples using a quantitative competitive PCR.Results: Nineteen out of 97 patients developed CMV disease over the first 12 months of the study. Sixteen (59%) out of 27 patients who were CMV-positive developed disease compared with three (4%) out of 70 of patients who were PCR-negative (P = 0.0001). A positive CMV PCR result was significantly associated with the development of disease (P = 0.0001), with a relative hazard of 20.15 [95% confidence interval (CI), 5.80-69.98]. Median CMV viral toad was significantly higher in those individuals who went on to develop CMV disease (P = 0.02). In PCR-positive patients, each 0.25 log,, increase in viral load increased the risk of disease (relative hazard, 1.37; 95% CI, 1.15-1.63; P = 0.0004).Conclusions: CMV PCR predicts the development of CMV disease and can be used to target ophthalmological resources to those patients at highest risk of retinitis. Asymptomatic patients who are PCR-positive represent a high-risk group in whom controlled trials of pre-emptive therapy could be conducted.

Objectives: Despite life‐long maintenance therapy, cytomegalovirus (CMV) retinitis frequently progresses in patients with AIDS. Virological markers that could clarify pathogenesis and identify risk factors for progression are required. Design and methods: We prospectively recruited 45 patients with CMV retinitis. Blood and urine samples were collected before and after induction therapy, and on a monthly basis thereafter during routine medical and ophthalmological assessment, and at any time retinitis progressed. CMV load was measured by quantitative‐ competitive polymerase chain reaction (PCR). Results: The median time to first progression of retinitis was 78 days and to death was 8.7 months. Eighty‐five per cent of patients who were PCR‐positive at diagnosis of retinitis became PCR‐negative after 21 days of ganciclovir induction therapy. Six patients who remained PCR‐positive after 21 days of treatment had a significantly higher CMV load at presentation (P = 0.005), and a shorter time to first progression of retinitis of 40 days. High CMV loads in blood at presentation were associated with a shorter time to progression (P = 0.16; relative hazard, 1.57) and a significantly shorter time to death (P = 0.004; relative hazard, 1.76). This significant relationship with survival remained after adjustment for potential confounding variables (CD4 count, age, method of drug administration). Conclusions: We conclude that CMV load in the blood of AIDS patients is an important factor in the pathogenesis of retinitis, and quantification of CMV could be used to both select patients for controlled clinical trials and to optimize individual anti‐CMV induction therapy.

BACKGROUND:More than 90% of patients with HIV have been infected at some time with cytomegalovirus (CMV) and up to 40% of those with advanced HIV will develop CMV disease. The incidence of CMV disease is increasing but the prognosis for the patient remains poor.MONITORING FOR CMV:It is therefore important to monitor patients with low CD4+ counts in order to identify those most at risk of developing CMV disease and to treat them before the disease becomes established. Polymerase chain reaction (PCR) is probably the most effective and sensitive method of detecting CMV and a positive result is predictive for development of CMV disease; more than 80% of patients with CMV retinitis are CMV PCR-positive at the time of diagnosis. PCR can also detect the presence of CMV up to 14 months before the development of retinitis.TREATMENT OF CMV RETINITIS:In patients with detectable CMV, but no evidence of active infection, pre-emptive treatment with ganciclovir or valaciclovir has been shown to reduce the risk of developing retinitis in these high-risk patients. Such oral therapy, which is generally better tolerated than intravenous therapy and results in a better quality of life for the patient, is likely to be more effective at this stage whilst viral loads are low.CONCLUSIONS:CMV PCR can be used to prospectively monitor patients in order to identify those most at risk of developing CMV retinitis. If CMV infection is diagnosed early, while viral loads are still low, pre-emptive oral therapy can be instituted which will reduce the chances of developing retinitis in those patients most at risk.

Cytomegalovirus infection is common in patients with AIDS, and often causes retinitis. Treatment is rarely curative, but the progression of retinitis is delayed. The untreated course of cytomegalovirus retinitis in AIDS is unknown. We report a 35-year-old man with retinitis who refused treatment. Retinitis resulted in blindness within 6 months. Measurement of cytomegalovirus genomes showed an increasing viral load in blood and urine.