In HIV-infected patients, some differential alterations of CD4 and CD8 T cell homeostasis may not be restored by >=7 years of highly active antiretroviral therapy, in spite of good CD4 T cell repopulation

Sir, We read with interest the recent manuscript by Méndez-Lagares et al. describing significant alterations in T cell homeostasis in patients on highly active antiretroviral therapy (HAART) showing low-level CD4 T cell repopulation, despite long-lasting and persistent viral replication control. In particular, notable reductions of circulating naive CD4 T cells with increased expression of markers for activation, senescence and proliferation were found in comparison with patients with satisfactory CD4 T cell repopulation. These results agree with their proposed model of immune impairment in patients with low-level CD4 T cell repopulation, suggesting intrinsic thymus failure and specific features of premature immune senescence. Our laboratory is actively working on a custom eight-colour flow-cytometry assay (BD LyoTube 8-color CD4 and CD8 bundle, BD Biosciences, San Jose, CA, USA), which is able to analyse differentiation, activation and senescence in CD4 and CD8 T cells (CD4 Lyotube: CD95 FITC/CCR7 PE/CD3 PerCP-Cy 5.5/CD25 PE-Cy7/CD127 Alexa Fluor 647/CD45 APC-H7/CD4 AmCyan/CD45RA V450; clones DX2/150503/SK7/2A3/HIL-7R-M21/2D1/SK3/HI100, respectively; and CD8 Lyotube: CD38 FITC/CCR7 PE/CD3 PerCP-Cy 5.5/CD69 PE-Cy7/CD127 Alexa Fluor 647/CD45 APC-H7/CD8 AmCyan/CD45RA V450; clones HB7/150503/SK7/L78/HIL-7R-M-21/2D1/SK1/HI100, respectively). Stained and lysed whole blood was analysed on an FACS-Canto II flow cytometer using FACS-Diva v.6.1.3 software (BD Biosciences). This test, performed on whole blood on a semiroutine scale, proved useful to analyse T cell homeostasis in HIV-infected persons in a protocol approved by the Institute’s Ethics Committee. In particular, this approach allowed us to verify that naive CD4 T cell frequency was lower in patients unable to reach CD4 reconstitution than in HIV patients reaching CD4 reconstitution [7.4% (IQR: 1.2–20.0) versus 33.1% (IQR: 27.9–43.9), P,0.0001], confirming the results of Méndez-Lagares et al. In order to evaluate how HAART impacts on CD4 and CD8 T cell homeostasis, residual samples from 10 untreated HIV-infected persons were analysed by this assay before starting treatment and after 6 months of treatment. After 6 months of treatment, all patients fulfilled immune reconstitution criteria [CD4 cells/ mm: 530 (IQR: 467–834)] and obtained viral load control [plasma viral load: undetectable (,40 copies/mL)]. Thirty-two seronegative healthy donors (HDs) were used as controls. Table 1 summarizes the immunological parameters that are found to be affected by HIV infection: these include CD95 expression on CD4 T cells, Treg frequency and CD8 activation markers. Many of these factors are positively modulated by 6 months of successful HAART, gradually approaching, but not reaching, the level found in HDs: CD4+, CD4CM+CD95+, Treg, CD8EM+ CD127+ and CD8+CD69+. Others are completely restored, as in HDs: CD8+CD38+, CD8EM+CD38+, CD8TEMRA+CD38+, CD8TEMRA+CD127+ and CD8CM+CD69+. Notably, two CD8 T cell subsets (CD8CM+ and CD8TEMRA+) seem negatively affected by HAART, increasing the difference with respect to HDs. Finally, many immunological markers are not affected at all by HAART (Table 1, last column). Thus, the short-term control of HIV replication seems able to allow a reduction of CD8 T cell activation and of CD4 apoptosis, but fails to completely restore CD4 and CD8 T cell homeostasis. Moreover, in order to ascertain if long-term HAART may be effective in fully restoring CD4 and CD8 T cell homeostasis, residual samples from 18 HIV-infected patients treated with HAART for 7–10 years, with recovered CD4 counts [CD4 cells/ mm: 910 (IQR: 685–1074)] and with a controlled viral load [plasma viral load: undetectable (,40 copies/mL)], were also analysed. Notably, many parameters that were not restored after 6 months of HAART were found fully normalized after longterm HAART. Nevertheless, as shown in Table 2, some markers (CD4+CD95+, CD4CM+CD95+, CD8+, CD8NA+, CD8+CD127+, CD8NA+CD127+ and CD8EM+CD127+) remained significantly altered, despite long-term immunological and virological treatment response. These observations suggest immune homeostasis damage inflicted by HIV infection in the active replication phase that cannot be recovered by (apparently) successful HAART. As markers of treatment efficacy, HIV-RNA plasma levels and CD4 T cell count are routinely used; in particular, the CD4 count itself plays a major role, since counts below the 200 cells/mm threshold have been associated with a worse prognosis, and therefore has been adopted in international guidelines to guide and evaluate antiretroviral treatment decisions. Despite its Research letters