BACKGROUND: HIV late presenters are at high risk of cytomegalovirus (CMV) reactivation and end-organ disease. CMV viraemia has been associated with poor survival but the effect of anti-CMV treatment has not been studied in this setting. METHODS: HIV-positive patients were included in a retrospective study if presenting with <350 CD4+ T-cells/μl and starting an antiretroviral treatment within 3 months of the diagnosis. Primary end point was 5-year survival according to the presence of CMV viraemia, CMV end-organ disease and anti-CMV treatment. RESULTS: 302 patients were included. 157 patients (52%) presented CMV viraemia (CMV-V) and 44 (14.6%) CMV end-organ disease (CMV-EOD). 5-year mortality was higher in CMV-EOD and CMV-V patients than in CMV-negative patients (11.4 versus 9.6 versus 0%; P=0.002). In patients with CMV-V, 5-year mortality was numerically higher in untreated patients (12.9% versus 6.9%; P=0.257) without reaching statistical significance. At univariate analysis the diagnosis of serious opportunistic infections (cryptococcosis, progressive multifocal leukoencephalopathy, lymphoma; P=0.001) and the absence of a negative CMV DNA in the follow-up (P<0.001) were associated with poor outcome. At multivariate analysis HCV coinfection (P=0.016; aOR 6.98, 95% CI 1.50, 32.59), the absence of a negative CMV DNA in the follow-up (P<0.001; aOR 19.40, 95% CI 3.70, 101.64) and marginally the absence of anti-CMV treatment (P=0.052; aOR 4.944, 95% CI 0.99, 24.73) were independent predictors of poor outcome. CONCLUSIONS: CMV reactivation in HIV-positive patients with poor immunity is associated with worse prognosis: the pre-emptive use of anti-CMV therapy was associated with a better outcome in patients with CMV-V.
BACKGROUND: HIV late presenters are at high risk of cytomegalovirus (CMV) reactivation and end-organ disease. CMV viraemia has been associated with poor survival but the effect of anti-CMV treatment has not been studied in this setting. METHODS: HIV-positive patients were included in a retrospective study if presenting with <350 CD4+ T-cells/μl and starting an antiretroviral treatment within 3 months of the diagnosis. Primary end point was 5-year survival according to the presence of CMV viraemia, CMV end-organ disease and anti-CMV treatment. RESULTS: 302 patients were included. 157 patients (52%) presented CMV viraemia (CMV-V) and 44 (14.6%) CMV end-organ disease (CMV-EOD). 5-year mortality was higher in CMV-EOD and CMV-V patients than in CMV-negative patients (11.4 versus 9.6 versus 0%; P=0.002). In patients with CMV-V, 5-year mortality was numerically higher in untreated patients (12.9% versus 6.9%; P=0.257) without reaching statistical significance. At univariate analysis the diagnosis of serious opportunistic infections (cryptococcosis, progressive multifocal leukoencephalopathy, lymphoma; P=0.001) and the absence of a negative CMV DNA in the follow-up (P<0.001) were associated with poor outcome. At multivariate analysis HCV coinfection (P=0.016; aOR 6.98, 95% CI 1.50, 32.59), the absence of a negative CMV DNA in the follow-up (P<0.001; aOR 19.40, 95% CI 3.70, 101.64) and marginally the absence of anti-CMV treatment (P=0.052; aOR 4.944, 95% CI 0.99, 24.73) were independent predictors of poor outcome. CONCLUSIONS: CMV reactivation in HIV-positive patients with poor immunity is associated with worse prognosis: the pre-emptive use of anti-CMV therapy was associated with a better outcome in patients with CMV-V.