Possible role of plasma Galectin-9 levels as a surrogate marker of viremia in HIV infected patients on antiretroviral therapy in resource-limited settings

Background Early detection of viremia in HIV infected patients on anti-retroviral therapy (ART) is important to prevent disease progression as well as accumulation of drug resistance mutations. This makes HIV viral load (VL) monitoring indispensable in HIV infected patients on ART. However VL, being an expensive test, results in heavy financial burden on health services. Hence, cheaper surrogate markers of viremia are desired to reduce overall cost of management of HIV infected patients. Methods We enrolled aviremic (n = 63, M:F = 31:32) and viremic (n = 43, M:F = 21:22) HIV infected patients at 1 year after ART initiation. Viremic individuals were identified as those having a plasma VL of more than 1000 copies/mu l and aviremic individuals as less than 40 copies/mu l. The study participants also included immuno-virologically discordant patients as they demonstrate differential degrees of immune-reconstitution and are likely to harbour concomitant infections influencing levels of immune-activation markers screened as the surrogate markers. Immune activation markers viz. plasma hs-CRP, soluble-CD14 and Galectin-9 levels were estimated by ELISA, IL-6 by luminex assay and percentages of CD38+ CD8+ cells were determined by flow cytometry. The levels were compared between viremic and aviremic patients and correlated with plasma viral load. Receiver operated curve (ROC) analysis was done for plasma Galectin-9 levels. Results Viremic patients had significantly higher levels of Galectin-9 and %CD38+ CD8+ cells (p values < 0.0001) than aviremic patients. Levels of the other activation markers did not differ between viremic and aviremic individuals. Galectin-9 levels (r = 0.76) and %CD38+ CD8+ cells (r = 0.39) correlated positively with VL. Area under curve for Galectin-9 levels for distinguishing between viremic and aviremic individuals was 0.98. Youden index, sensitivity, specificity, positive predictive value and negative predictive value for Galectin-9 levels were 0.87, 0.97, 0.90, 0.87 and 0.98, respectively, at the cut-off value of 5.79 ng/ml. Conclusions Plasma Galectin-9 levels could identify viremic individuals with sensitivity and specificity of more than 90%. Thus, they showed a potential to serve as a surrogate marker of viremia in HIV infected patients on ART and would have cost implications on HIV management especially in resource-limited settings. However, the findings need to be confirmed in the patients on ART for different durations of time.