Background: Limited data exist on factors predicting the development of opportunistic infections (OIs) at higher-than-expected CD4(+) cell counts in human immunodeficiency virus (HIV) type 1-infected adults.
Methods: Multivariate Poisson regression models were used to determine factors related to the development of groups of OIs above their respective traditional upper CD4(+) cell count thresholds: group 1 (>or=100 cells/ microL), OIs caused by cytomegalovirus, Mycobacterium avium complex, and Toxoplasma gondii; group 2 (>or=200 cells/ microL), Pneumocystis pneumonia and esophageal candidiasis; and group 3 (>or=300 cells/ microL), pulmonary and extrapulmonary tuberculosis.
Results: In groups 1, 2, and 3, 71 of 9,219, 125 of 7,934, and 36 of 7,838 patients, respectively, developed >or=1 intragroup OI. The strongest predictor of an OI in groups 1 and 2 was current CD4(+) cell count (for group 1, incidence rate ratio [IRR] per 50% lower CD4(+) cell count, 5.37 [95% confidence interval {CI}, 3.71-7.77]; for group 2, 4.28 [95% CI, 2.98-6.14]). Injection drug use but not current CD4(+) cell count predicted risk in group 3. Use of antiretroviral treatment was associated with a lower incidence of OIs in all groups, likely by reducing HIV-1 RNA levels (IRR per 1-log(10) copies/mL higher HIV-1 RNA levels for group 1, 1.50 [95% CI, 1.15-1.95]; for group 2, 1.68 [95% CI, 1.40-2.02]; and for group 3, 1.89 [95% CI, 1.40-2.54]).
Conclusion: Although the absolute incidence is low, the current CD4(+) cell count and HIV-1 RNA level are strong predictors of most OIs in patients with high CD4(+) cell counts.