Our data demonstrate that HIV-infected children undergo progressive immune reconstitution in response to HAART that could potentially lead to normalization of immune parameters. Previous studies showed immunologic improvement in HIV-infected children and adults during the first year of HAART [31
] and continuous increase of CD4+
T cells over 6 years of therapy in children and adults who maintained undetectable plasma HIV RNA [34
]. Here, we extend these observations by showing that not only CD4+
T cells, but also functional and phenotypic immune measures continue to improve in HIV-infected children over 3 years of effective HAART. A unique feature of immune reconstitution in our study was that after 3 years of HAART, the CD4+
% of HIV-infected children was similar to those of healthy age-matched controls. Furthermore, the reconstitution of CD4+
T cells did not differ appreciably between the two age cohorts of 3-6 and 7-21 years enrolled in this study.
It has been long recognized that HIV infection alters the distribution of T-cell phenotypes, which may be partially reversed by HAART [31
]. However, this study is the first one to demonstrate complete normalization of T-cell subpopulations including naive and activated CD4+
and naive CD8 T cells. The activated and total CD8+
% remained elevated, which may be due to persistent low-level viremia that can be demonstrated even in patients with plasma HIV RNA less than 50 copies/ml [39
]. However, as changes in the total and subpopulations of CD8+
% were still actively occurring at week 144 of HAART, further improvement, leading perhaps to normal CD8+
% after more than 3 years of HAART, could not be ruled out.
The robust reconstitution of T cells and their subpopulations demonstrated in this study may derive from the large thymic reserve associated with the relatively young age of our study participants. We assessed thymic responses to HAART by the TREC content of CD4+
T cells. The TREC content steadily increased in both age cohorts but was consistently higher in the younger group, indicating that the thymus, whose activity increases with younger age, was the main contributor to the TREC rebound. An alternative explanation ascribing the TREC increase to HAART-associated decrease of CD4+
T-cell proliferation [40
] is less likely, because it does not explain the significant increase of TREC content with age (when all other responses to HAART did not differ between age groups) nor the significant correlation between CD4+
TREC content increases. The association of CD4+
TREC increases in response to HAART suggests that HIV infection inhibits thymic activity at early stages of T-cell ontogeny, before CD4/CD8 differentiation [41
While the de-novo generation of CD4+ and CD8+ T cells continuously increased during the first year of HAART, this translated into an increase of CD4+ but not of CD8+ T cells. This finding illustrates that the number of circulating CD4+ and CD8+ T cells is controlled at multiple levels, such that when HAART suppresses viral replication withdrawing the antigenic stimulus for cytotoxic T-cell lymphocyte (CTL) proliferation, HIV-specific CD8+ T cells decrease leading to the contraction of the entire CD8+ T-cell compartment. This is further evidenced by the strong association between the decrease of HIVCD8 ELISPOT values and total CD8+% during the first year of HAART. Ultimately, the CD8+ T-cell compartment contracts during HAART, in spite of increased thymic de-novo CD8+ T-cell production.
Functional immune reconstitution is an important goal of HAART, as ultimately the ability of the immune system to protect the host against opportunistic agents is a critical prognostic factor. Previous studies have shown that the reconstitution of pathogen-specific CD4+
T-cell responses differs with the microbial agent, though the mechanism that underlies this difference is not well understood [12
]. Here, we confirmed rapid reconstitution of Candida
and delayed reconstitution of HIV-specific CD4+
T-cell responses. However, Candida
values positively correlated with each other and with total and memory CD4+
% and negatively correlated with activated CD8+
%, suggesting that functional immune reconstitution has multiple common features.
T-cell-mediated HIV-specific immunity is a hall-mark of long-term non-progression of the infection, but it is unclear whether these responses contribute to the control of viral replication or denote immune preservation [44
]. Our study was not designed to address this question, and the analysis of HIVCD4
ELISPOT was exploratory. However, our findings, together with previous ones [14
], support pursuing this question in future studies. Finding a positive association between control of viral replication and CD4+
T-cell-mediated responses to HIV may change the currently accepted paradigm that HAART has a deleterious effect on HIV-specific immune defenses.
The concept that HAART decreases HIV-specific immunity stems from the observation that HAART is associated with a decrease in HIV-specific CD8+
]. These cells have been shown to limit HIV replication in vitro
]. The role of CTL in the control of in-vivo retroviral infection was demonstrated in animal models [21
]. In humans, this subject remains controversial particularly with respect to the relative importance of the magnitude against breadth of the CTL response [19
]. Our data confirm previous reports that effective HAART is associated with a decrease in the magnitude of HIVCD8
ELISPOT. However, HAART did not affect the breadth of HIVCD8
A goal of this study was to increase our understanding of the differential contributions of HIV replication against immune activation to CD4+ T-cell dysfunction. We found significant negative correlations of plasma HIV RNA with CD4+% and thymic output before and during the first year of HAART. This coincided with rapid changes in the CD4+ T-cell numbers and phenotypes. Activated CD4+% not only negatively correlated with CD4+% and thymic output in the first year of HAART, but also tended to positively correlate with HIV plasma RNA over the same period, complicating the understanding of their role in immune suppression. Activated CD8+% did not correlate with plasma HIV RNA but negatively correlated with lower CD4+% and function at week 144 of HAART, when most patients had undetectable viral replication. The model that emerges is that immune recovery of HIV-infected children is a biphasic process, including an early rapid phase in which the decay of viral replication is associated with recovery of thymic activity and repopulation of the T-cell compartment and with recovery of functional responses to mitogens and some antigens. A second phase of immune recovery begins or becomes evident after 6 months to 1 year of HAART, when viremia is stable, but low. During this phase, gains in CD4+ T-cell numbers and function and redistribution of CD4+ T-cell subpopulations continue at a slower rate and negatively correlate with CD8+ T-cell activation. Further studies are needed to determine whether the mechanism that underlies the maintenance of CD8+ T-cell activation after 3 years of HAART is the viral replication, albeit at low levels, and to identify the mediators of the immune suppression after prolonged HAART.
In conclusion, this study showed a robust immune reconstitution in HIV-infected children in response to 3 years of effective HAART, that could be best predicted by the baseline immunologic characteristics of the patients. Baseline CD4+
% predicted the recovery of CD4+
T-cell numbers and function in response to HAART, which is in accordance with previous reports [36
]. We also showed that thymic output in response to HAART increased with higher thymic function at the initiation of therapy. Although HAART has potential side effects and poses significant adherence problems, its initiation in early stages of HIV infection has clear advantages with respect to immune reconstitution.