Oral hairy leukoplakia and oral candidiasis as predictors of HIV viral load

Oral hairy leukoplakia (OHL) and oral candidiasis (OC) increase in prevalence in immune suppressed cohorts[¹], and have been associated with immune dysfunction and reduced quantities of CD4 lymphocytes[²]. They are markers of clinical progression to AIDS independent of the CD4 cell count[³], although an exact explanation for this observation is not known. This investigation sought to determine the association of OHL and OC with quantitative plasma viral load levels in a clinical setting and to examine the influence of demographic, CD4 cell count, and antiretroviral therapy on the association.

A convenience sample of 250 HIV-infected adults with HIV RNA values obtained by reverse transcriptase- polymerase chain reaction (Monitor; Roche Molecular Systems, Branchburg, NJ, USA) within 1 month of oral examination, who receive medical care at the University of North Carolina, were examined for clinical evidence of OHL and OC. Sociodemographic, risk behavior, clinical characteristics, medication use, and laboratory data were obtained by interview or medical record abstraction. Plasma HIV RNA dichotomized at a clinical cutpoint of 20000 copies/ml (4.3 log₁₀) was used as the outcome variable in bivariate analysis and a multivariable model.

Subjects had a mean age of 37 years (range 18-63), were 73% male and 68% Black. In bivariate analyses, persons with either OHL (24%) or OC (24%) were 1.84 [odds ratio (OR), 95% confidence interval (CI) 1.02, 3.35] times more likely to have a viral load of 20000 copies/ml or greater than persons without oral lesions. A clinical AIDS diagnosis (OR 2.61, 95% CI 1.42, 4.78) and a current CD4 cell count of less than 200 cells/mm³ (OR 3.78, 95% CI 2.19, 6.54) were significant predictors of a higher viral load. Of the patients, 47% were on highly active antiretroviral therapy (combination antiretroviral drug therapy including one or more protease inhibitors), 33% were on combination antiretroviral therapy with nucleoside analogs or non-nucleoside reverse transcriptase inhibitors, whereas 20% were on no antiretroviral therapy. Antiretroviral therapy had a strong association with lower viral loads (Chi square, P=0.001). Antifungal therapy, but not antimycobacterial or herpes-directed antiviral therapy, was significantly (OR 2.19, 95% CI 1.14, 4.19) more common among individuals with viral loads of 20000 copies/ml or greater. Sex, race, age, HIV transmission risk behavior, education, and income were not predictive of higher viral load levels in bivariate or multivariable analyses.

In a multivariable unconditional logistic regression model (Table 1), individuals with OHL were 2.08 times more likely to have a viral load of 20,000 copies/ml or greater than individuals without OHL, independent of the CD4 cell count and antiretroviral therapy. In addition, the association of OC with viral load was dependent upon the presence or absence of antifungal therapy. As expected from the bivariate relationship, in the multivariable model among individuals not being treated with antifungal agents, those with clinical OC were at a slightly increased risk of having a higher viral load (OR, 1.47; 0.66, 3.27) over individuals without oral lesions.

Table 1

Although the strongest predictors of high viral load in our multivariable model were the absence of antiretroviral therapy and having a CD4 cell count of less than 200 cells/mm³, our results suggest that when an HIV-infected individual presents with OHL there is a modest twofold increased likelihood of his or her viral load being 20000 copies/ml or greater, independent of the CD4 cell count. In addition, among individuals not on antifungal agents, those with OC were 1.5-fold more likely to have a viral load of 20000 or greater, although small patient numbers limited the ability of this association to reach statistical significance. These associations of OC and OHL with higher viral loads may explain, at least partly, the previous observation [³] of more rapid HIV disease progression in individuals with these oral lesions, independent of the CD4 cell count.

In conclusion, the relatively easy to identify oral lesions, OHL and OC, serve as markers of the underlying HIV disease process and may reflect the incremental evolution of viral infection. Therefore, consistent with the US Agency for Health Care Policy and Research‚s clinical practice guidelines for the evaluation and management of early HIV infection[⁴], oral mucosal evaluation is important to include as a physical examination tool for the assessment of both immune status and viral replication. In developed countries, identification of these clinical disease markers may indicate either a high plasma viraemia in the absence of antiretroviral therapy or renewed high viral replication in the patient with previously effective suppression of HIV by combination antiretroviral therapy. Any OHL in a highly treated patient on antiretroviral medication may thus suggest possible antiretroviral regimen failure. In addition, in undeveloped nations, where viral load determinations are not routine or often even available, the observation of these oral lesions may indicate a high viral load and rapid CD4 cell destruction.