Background: Very few studies have investigated risk factors for cytomegalovirus (CMV) retinitis. Identifying these risk factors will have many benefits, including helping establish screening regimens, examination frequency regimens, and targeted prophylaxis with oral therapy with valganciclovir or other anti-CMV agents. The purpose of this study was to determine the laboratory-based risk factors for CMV retinitis in patients with AIDS.
Methods: We conducted a case-control study involving 120 patients in whom CMV retinitis had been diagnosed in 1990-99 and 159 patients without CMV retinitis from the same period. The sampling was from a primary study base in eastern Ontario and western Quebec of patients with AIDS and CD4 counts less than 50 cells/microL at the time of diagnosis of retinitis in the case subjects or an analogous date for the control subjects. There were two components to the study. In the first component (n = 279) we examined standard-of-care laboratory tests (hematologic and nutrition variables) done during the study period. In the second component (n = 57), which was a subset of the first, we examined laboratory tests (HLA type, qualitative and quantitative CMV polymerase chain reaction [PCR] and HIV load) on stored blood samples from the eastern Ontario site. Multivariate logistic regression was used to model the data and control for confounding. We developed a systematic model-building strategy, from assumption testing to model building to model checking.
Results: A low hemoglobin concentration was a statistically significant predictor of CMV retinitis (odds ratio [OR] 0.96, 95% confidence interval [CI] 0.94-0.98). Both qualitative CMV PCR (OR 21.71, 95% CI 1.80-261.67) and quantitative CMV PCR (OR 33.03,95% CI 2.32-469.39) were strong predictors of CMV retinitis. Among the 80 HLA types tested, HLA-Bw4 (OR 11.68, 95% CI 1.29-105.82) and HLA-DRB115 (OR 9.34, 95% CI 1.14-76.41) were significant predictors of CMV retinitis, whereas HLA-Cw7 was protective against CMV retinitis (OR 0.09, 95% CI 0.01-0.67).
Interpretation: We have identified laboratory variables that elevate (or decrease) the risk of CMV retinitis. These findings may be useful to clinicians and health policy experts in developing rational guidelines for screening, examination frequency and targeted prophylaxis for patients with AIDS.