The study population included HIV-infected children and adolescents that had been comprehensively studied by CD38 expression on CD8 T cell

and LPR to mycotic antigens along with traditional VL and CD4. The aim of this study was to evaluate the discriminatory potential of CD38 expression and antigen-specific lymphocyte proliferation to differentiate non-responders and a mixed population of responders with full and partial virus suppression on HAART and two NRTIs suppressive regimens. According to guidelines [4–6], two NRTIs backbone CHIR-99021 solubility dmso is not longer considered preferred, although at the time of the study was still in use and at present continues to be used in developing countries where the cost of antiretroviral agent drives the antiretroviral therapy.

We found CD38 expression on CD8 T cell accurately discriminates responders versus non-responders. CD38 ABC has long been recommended as a more accurate measure of CD38 staining than %CD38/CD8, due to the unimodal heterogeneous CD38 expression [20, 22]. However, in our study, CD38 ABC and %CD38/CD8, showed a good correlation, a high concordance, resulting their cutoff points in the same responder and non-responder frequencies and in identical sensitivity and specificity. However they did not classify all patients in the same way. For this reason the combination of the two assays in alternative way, ‘either CD38 ABC or %CD38/CD8’ improved sensitivity to 83.3%. Conversely, the combination ‘CD38 ABC and %CD38/CD8’ decreases sensitivity to 66.7%. Studies in adults and paediatric patients [9, 26, 27] have looked at the correlation of VL and CD38 expression finding LY294002 solubility dmso that as a VL decreases so does activation, supporting the

use of CD38 expression as a marker of viral replication to monitor response to therapy. In adults a direct association ID-8 between CD38 expression and viral replication was observed only in patients with >400 copies HIV-1 RNA/ml [28]. The low level of activation observed in subjects with full virus suppression (<50 copies/ml) may be due to factors other than plasma viraemia, such as proinflammatory cytokines, microbial products, residual HIV replication in lymph nodes. Steel et al. [(29] found the sensitivity and specificity of CD8 CD38high percentage to detect HIV-1 viraemia was 85% and 81% respectively at a viral load of 10,000 HIV-1 copies/ml. Accordingly we found 75% sensitivity and 93.8% specificity for both CD38 ABC and %CD38/CD8, and sensitivity improved to 83.3% when the two assays for CD38 expression were combined in alternative way. CI intervals included values reported by Steel et al., although our patients were distinguished in responders and non-responders and not stratified by viraemia. In particular, a high CD38 expression level seems to be satisfactory at identify non-responders, while low CD38 expression level, especially in combination with good LPR, identify responders.