In the pre-HAART era, many studies demonstrated that SI and/or X4 viruses emerged in approximately 40%–50% of patients progressing to AIDS, suggesting that these variants were associated with a worse clinical and viro-immunological outcome . More recently, this percentage was found to range from 10% to 20% in naïve patient cohorts [31–33] and from 40% to 50% in multi-failed heavily-treated patients from maraviroc registrative trials [17–19, 34], including a small proportion (1%-4%) of pure X4 and a large majority of R5/X4 (dual mixed, DM) strains.
Herein, we focused on patients surviving chronic HIV infection for at least 15 years with a long history of suppressed viremia under antiretroviral therapy in the majority, high CD4 median level, and no opportunistic infection at time of analysis. The results indicate that, also in this population, >40% of patients presented an archived X4 virus in proviral DNA, a percentage which is similar, or even higher, when compared to other recent reports [35, 36] in suppressed patients with a shorter history of viremia suppression. Therefore, examining DNA for CRT determination should be recommended for all patients before prescription of chemokine inhibitors. Although the proportion of X4-carrying patients might be overestimated in DNA with respect to plasma , recent studies confirm that tropism switches under ART suppression are rare. Therefore, the current proviral DNA can be considered a good mirror of the patient viral quasispecies, including both recent or former replication variants [28, 36].
Not only was a correlation with CD4 nadir evidenced, in agreement with previous studies , but X4 variants were significantly associated with exposure to higher cumulative viral loads and poorer immunological control over time, rather than with the viro-immunological state at time of testing. From the design of the study, it is not possible to resolve the pending question of whether X4 viruses emerged (and were archived in proviral DNA) as a consequence of the continuous replication and immunological impairment or, on the contrary, were responsible for disease progression in these patients. Even if several successive regimens might have masked the actual impact of therapy on CRT, according to other publications , the duration and type of antiretroviral therapy did not seem to influence CRT in our series. This finding also confirms our previous observation that, in patients treated by HAART and sequentially tested for CRT, the co-receptor shift is independent of virological success .
None of the baseline demographic and clinical characteristics, apart from CD4 and viral load trends over time, seem to influence the predominant strain archived in proviral DNA. In particular, even if the risk factor for HIV acquisition might be a determinant for CRT selection at time of transmission , it seems to lose its importance over time. A lower proportion of X4 strains has been observed among some non-B-subtypes, in particular subtype C , suggesting a possible association with an improved virological outcome for non-B patients when excluding racial and social variables, as recently described by some authors [40, 41]; in the present study, however, non B-subtypes were rare and no definite conclusion can be drawn concerning the relationship between CRT and viral subtype. Moreover, the co-infection with other viruses, in particular with HBV and HCV, failed to demonstrate any association with CRT. In HIV-HCV co-infection, the env gp120 of HIV was supposed to interact with HCV E2 in triggering the apoptosis mechanisms  to modulate the biology of human hepatic stellate cells which play a key role in the fibrosis pathogenesis , and to enhance the replication of HCV through the engagement of extracellular coreceptors on hepatocytes . In this study, a slightly higher proportion of R5 patients was observed among the HCV-positive patients when compared to negatives, but without statistical significance. If R5 and X4 strains can be differentially involved in determining viral persistence and fibrosis progression in HCV positive patients should be object of further research. No differences in AIDS-related and serious non-AIDS events were observed in the two groups, but, as only long surviving patients were included in the study, pathological events leading to death were not considered and therefore we cannot exclude that severe AIDS occurrences might be variously distributed in X4 and R5 patients.
Many reports indicate that HIV-infected patients develop an inflammation and hypercoagulation state which is involved in the aging mechanisms. Surrogate serological markers such as IL-6, D-dimers and hsPCR have been demonstrated to be good prognostic markers in HIV-infected patients and are generally associated with elevated HIV-1 viral loads [23–26]. It has been hypothesized that X4-patients present an accelerated rate of disease progression compared to R5. In fact, also in our patient cohort aging with HIV, the presence of X4 strains was associated with a poorer immunological and virological control over time; therefore, we evaluated the hypothesis that CRT might be correlated with a worsened inflammation state as measured by the above mentioned surrogate markers. However, HsPCR, D-dimer and IL-6 s levels did not differ between the two groups of R5 and X4 patients. Therefore, in our experience, CRT did not seem to affect the inflammation state in patients aging with HIV, at least when measured by means of surrogate markers. None of the other patient characteristics was predictive of higher levels of IL-6, HsPCR and D-dimers in our cohort, except for a lower current and cumulative CD4 cell count for IL-6, which would seem to suggest, according to other studies , that immune-depression is an essential driving force for inflamm-aging in HIV positive patients. A limitation of our findings is represented by the fact that surrogate markers do not show any difference according to age or, more importantly, between patients with active HIV replication compared to those with undetectable viraemia. This is most likely due to the particular characteristics of our population which included patients with a narrow age range (median 50, IQR 47–53 years) and only a limited number of subjects with detectable HIV-RNA (25%). Moreover, although IL-7 was demonstrated to induce viral evolution of X4 viruses in vitro, no association between current IL-7 levels and the presence of archived X4 variants in vivo was observed in the present study; however, it cannot be excluded that lymphopenia-induced IL-7 production might have previously favored the R5 to X4 switch at previous timepoints during the patient’s history, but this association was not apparent at time of analysis.
Lastly, the issue concerning the clinically relevant FPR cut-off for CRT attribution based on geno2pheno is still controversial; herein, according to the 2011 European Guidelines , the 20% FPR cut-off was chosen since interpretation was based on a single DNA amplification and sequencing; however, similar results were also obtained when data were analyzed using a 10% and a 5% cut-off, excluding the potential impact of this choice on the final outcome of the analysis.
This is one of the few studies regarding CRT in patients aging with HIV (all patients living at the time of study design have been included) and the first, to our knowledge, which investigates the relationship between the CRT and state of inflammation. However, the study presents several limitations. All the above reported lack of associations might be due to a sample of insufficient size to detect small effects, particularly in subgroup analyses, interactions and multivariable adjustments. Furthermore, due to its retrospective design, some relevant baseline characteristics might not have been measured: for instance, for some of the older patients, baseline viral load was not available (before 1995); in fact, cumulative viral load is probably underestimated for all patients. Also, repeated and prospectively collected measurements of inflammation markers were not performed, and we can only estimate the influence of point determination of these parameters. Finally, as already stated, patients have been “naturally” selected on the basis of their vital status at time of CRT testing, and it must be kept in mind that patients with a shorter HIV history might differ from this highly selected population.