Bcl-2 expression is moderately correlated with long-term variability of CD4 T-cell increase under successful highly active antiretroviral therapy

The objective of highly active antiretroviral therapy (HAART) is to induce a complete suppression of the plasma viral load (pVL), to achieve a CD4 T-cell increase, and to protect against opportunistic infections. However, the cause–effect relationship between the suppression of HIV replication and immune restoration is not completely clear. Indeed, in up to 30% of patients, CD4 T cell counts remain low or do not increase further, despite complete pVL suppression, providing a margin of intervention for immunomodulating approaches. We and others [1,2] have demonstrated that lymphocyte spontaneous apoptosis (LSA) provides a predictive factor for the immunological and clinical evolution of HIV infection, partly independent of the CD4 T-cell count and pVL. Among the cascade of factors that regulate LSA, Bcl-2 plays a key role in exerting antiapoptotic activity [3]. We have also demonstrated that HAART induces Bcl-2 hyperexpression in CD4 T cells (particularly in the naive subset) of patients with complete pVL suppression, in contrast to those who have residual HIV replication [4]. Therefore, it has been suggested that Bcl-2 upregulation and, as consequence, a decrease in LSA might contribute to the CD4 T-cell increase induced by HAART [4,5]. This hypothesis is supported by the findings recently reported by David et al. [6]. In a cross-sectional study, two groups of patients with prolonged pVL suppression (> 9 months) were evaluated, and it was shown that those who had previously failed to achieve a CD4 T-cell increase had lower levels of Bcl-2 expression and more LSA at the time of analysis than those in whom a good immune recovery had occurred. It could therefore be argued that levels of Bcl-2 expression might predict the long-term CD4 T-cell recovery in patients achieving sustained pVL suppression. Here we present the results of a prospective study aimed at evaluating this hypothesis. Twenty-one patients with pVL less than 50 copies/ml for 3 months or longer were selected for this study [15 men; mean age 44 years (SD 8.2); nine at first HAART experience; duration of previous pVL suppression 37.8 weeks (SD 29.8); four were on non-nucleoside reverse transcriptase inhibitor-containing regimens, whereas 17 were on protease inhibitor-containing regimens at baseline]. Bcl-2 expression in the CD4 T-cells was measured at baseline and calculated as the molecular equivalent of soluble fluorescein (MESF) [4]. Patients were followed with routine laboratory testing and clinical evaluation every 3 months until the most recent pVL, still undetectable for a mean of 99.7 weeks (SD 52). A summary measure of CD4 T-cell variation was calculated as the area under the curve of the percentage change (Δ%AUC) [7] during the period of observation, divided into semesters, and correlated with MESF expression of Bcl-2 at baseline. Correlations were analysed using Pearson's coefficient. All tests were two-sided. Analyses were performed using Statistica for Windows software (StatSoft Inc., Tulsa, OK, USA). The mean Bcl-2 expression in CD4 T cells was 81 239 MESF (SD 28 040). During the period of observation, the mean Δ%AUC of CD4 T cells was +26.74% (SD 25.93) per semester. As shown in Fig. 1, a trend for a correlation was found between Bcl-2 expression in CD4 T cells and their increase during the follow-up, although it did not reach statistically conventional significance (r = 0.40, P = 0.07). The time of observation, the duration of pVL suppression preceding baseline, the line of therapy, and the type of regimen at baseline did not significantly influence the correlation between Bcl-2 expression and Δ%AUC of the CD4 T cells (data not shown).Fig. 1.: Correlation between Bcl-2 expression and percentage change in the area under the curve of the CD4 T cells. AUC, Area under the curve; MESF, molecular equivalent of soluble fluorescein.In a subgroup of 12 patients, Bcl-2 expression was measured both in memory (CD4 CD45R0+) and naive (CD4 CD45RA+) T cells, as previously described [4]. No statistically significant correlations were found betbetween Δ%AUC of CD4 T cells and either Bcl-2 expression in the memory T cells (r = 0.24, P = 0.45) or in the naive T cells (r = 0.29, P = 0.36). In conclusion, the prospective study presented here suggests that Bcl-2 expression in CD4 T cells may be a predictor of CD4 cell increase in patients with sustained pVL suppression induced by HAART. However, the correlation reported here is only moderate. Although the prediction of a CD4 T-cell increase by means of Bcl-2 expression would be an attractive tool to individualize the use of immunomodulating agents in patients who really deserve them, Bcl-2 did not explain completely the viro-immunological discordant trends in this study, therefore other factors should be investigated. The correlation between Bcl-2 expression and CD4 T-cell increase was not attributable to a particular subpopulation (either memory or naive), on the basis of our data. It would be of interest to see whether IL-2 or other immunomodulating agents could induce Bcl-2 expression particularly in naive T cells, thus promoting a more pronounced immune-recovery with respect to HAART alone.