Quantitative pp65-antigenemia assay for the prediction of human cytomegalovirus disease in HIV-infected patients

Objective: To evaluate the ability of a quantified pp65-antigenemia assay t o predict the development of human cytomegalovirus (HCMV) disease in patien ts with an advanced HIV infection. Design: A prospective longitudinal study between March 1993 and December 19 96. Blood samples for the pp65-antigenemia assay were drawn at 2-3 month in tervals. Setting: AIDS department of an institutional tertiary care centre. Patients: A total of 101 HIV-infected patients with CD4 lymphocyte counts o f 100/mm(3) or less were enrolled. Ninety-seven patients were eligible for analysis. All patients gave informed consent. Main outcome measures: The development of HCMV disease. Results: Of the 97 patients, 24 developed HCMV disease after a median follo w-up of 10.6 months. Three months before the development of HCMV disease, a n increase in the median number of pp65-antigen-positive leukocytes was obs erved. The highest combination of sensitivity (45%) and specificity (94%) f or the development of HCMV disease within the next 3 months was found when an assay cut-off level of 48/10(5) pp65-antigen-positive leukocytes was app lied, with a positive predictive value(PPV) for the development of HCMV dis ease of 75%. The Kaplan-Meier estimate of HCMV disease-free survival after patients reached 48/10(5) or more antigen-positive leukocytes on longitudin al follow-up was a median 3.7 months [95% confidence interval (CI), 2.5-8.5 ]. The hazard ratio (HR) of this threshold level for the development of HCM V disease was 9.6 (95% CI, 4.2-21.8). Conclusion: Longitudinal follow-up using the pp65-antigenemia assay of HIV- infected patients with a low CD4 lymphocyte count improves the identificati on of patients who will develop HCMV disease in the foreseeable future, and should be considered for the selection of patients who may benefit from pr e-emptive HCMV treatment. (C) 1999 Lippincott Williams & Wilkins.