Human immunodeficiency virus (HIV) infection remains a major global health issue, affecting an estimated 39.9 million people in 2023 (WHO, HIV and AIDS, Fact sheets, 22 July 2024 (1)). Antiretroviral therapy (ART) has transformed HIV infection into a manageable chronic condition and brought the life expectancy of people living with HIV (PWH) nearly in line with that of uninfected individuals (2, 3). However, ART is not curative and requires life-long adherence; while it effectively suppresses viral replication, it does not eradicate the virus from latent reservoirs or fully restore immune function (4–6). PWH often suffer from chronic immune activation and inflammation, increasing their risk of non-AIDS-related comorbidities, such as cardiovascular diseases, neurocognitive disorders, and metabolic syndrome (7, 8).

This persistent inflammatory state in ART-treated PWH is complex and multifactorial (9). HIV persists in viral reservoirs, especially within CD4+ T cells and monocytes (10, 11), where it can produce low levels of viral RNA, and even viral proteins, potentially triggering antiviral immune responses (12). Additionally, low-level viral replication in sanctuary sites, where ART do not fully penetrate, may contribute to ongoing inflammation (13, 14). Microbial translocation, which is the leakage of bacterial products from the gut into the bloodstream due to gut-associated lymphoid tissue (GALT) disruption by HIV, further stimulates immune responses and increases levels of inflammatory mediators, such as lipopolysaccharide (LPS) and soluble CD14 (sCD14) (15, 16). Microbial translocation is linked to thymic function failure, another crucial factor involved in HIV disease progression and immune dysregulation (17). Immune dysregulation, particularly in CD4+ and CD8+ T cells, adds to this pro-inflammatory environment. CD4+ T cell depletion drives the release of inflammatory cytokines (e.g., IL-6, TNF-α), which are closely associated with poorer clinical outcomes in individuals with chronic HIV infection (18). Moreover, ART-treated individuals often show increased immune exhaustion (PD-1, LAG-3) and activation markers (CD38, HLA-DR) on T cells. These markers correlate with larger HIV reservoirs and reduced T-cell function (19, 20). Monocytes and other immune cells also contribute significantly to the inflammatory milieu (21). Monocytes play a central role in sustaining chronic inflammation, as they release pro-inflammatory cytokines and chemokines, which further activate immune pathways (22).

Lastly, clinical factors like age, gender, nadir CD4 count, co-infections (e.g. CMV, hepatitis B, hepatitis C, EBV), and timing of ART initiation, also influence inflammation levels. For instance, older age is associated with poorer immune responses (23–25), a phenomenon increasingly referred to as “inflammageing”, which has become more evident as ART extends life expectancy in PWH (26). Similarly, gender differences have been noted, with the female gender being linked to higher levels of immune activation (27, 28).

Understanding the relationship between HIV persistence markers and immune activation is crucial, as the inflammatory environment not only drives T-cell proliferation but also perpetuates a vicious cycle that maintains and replenishes viral reservoirs (13).

In this study, we aim to investigate the associations between HIV persistence, inflammatory markers, clinical indicators, and immune cell populations in a cohort of 49 PWH treated with the same dolutegravir-based ART regimen. We analysed over 40 parameters, categorized into four main groups:

This research seeks to deepen our understanding of the complex immune landscape in ART-treated PWH and its implications for managing HIV-associated inflammation and related comorbidities.

In this study, we investigated the associations between HIV persistence markers, inflammatory markers, immune cell populations and clinical indicators in a cohort of 49 PWH on long-term DTG-based ART. The relationship between HIV persistence and immune activation has been a subject of debate, with the literature presenting conflicting findings.

Understanding the origins of persistent inflammation and immune activation in ART-treated PWH holds significant clinical implications. If inflammation is primarily driven by the HIV reservoir, therapeutic strategies should focus on reducing reservoir size or silencing its activity, for example, by inhibiting viral transcription (e.g., Tat inhibitors) or minimizing residual viral replication. Conversely, if the link between reservoirs and inflammation is weak, alternative drivers such as lifestyle factors (e.g., smoking, obesity), microbial translocation, co-infections, or other comorbidities must be prioritized. Identifying these drivers is essential for optimizing therapeutic interventions and improving long-term outcomes.

On the one hand, a significant portion of the literature supports a strong link between viral reservoirs and immune activation, suggesting an interdependent relationship between HIV persistence and inflammation (13). Early evidence from Chomont et al. demonstrated that HIV proviral DNA preferentially resides in CD4+ T cells expressing immune activation (PD-1) and proliferation (Ki67) markers (31). Further analyses across different compartments – blood, rectal tissue and lymph nodes – confirmed a positive correlation between the frequency of PD-1+ CD4+ T cells in blood and total HIV DNA, as well as between PD-1+ CD4+ T cells in rectal tissue and both integrated HIV DNA and CA US HIV RNA (32). Similarly, strong positive associations were observed in rectal tissue and lymph nodes between CD8+ T cells expressing HLA-DR and CD38 and both integrated HIV DNA and CA US HIV RNA. Later, Hatano et al. corroborated these findings, demonstrating significant associations between proviral DNA levels and the frequency of PD-1-expressing CD4+ T cells in blood samples. They further concluded that low CD4+ T cell count (<350 cells/mm³), despite suppressive therapy, was linked to higher CA RNA levels, proviral DNA levels and increased frequencies of CD4+ T cells expressing CD38, HLA-DR, CCR5 and/or PD-1 (33). The same group found a consistent association between CD4+ and CD8+ T cells expressing HLA-DR and the frequency of resting CD4+ T cells containing HIV DNA (34). In a more recent study conducted in 2021 by Olson et al., the focus shifted from total or integrated HIV DNA and CA RNA to intact proviral DNA (35). Interestingly, while intact proviral DNA did not correlate with plasma inflammation markers (e.g. D-dimer), a significant correlation was observed with intracellular HIV RNA. This finding suggests that the viral transcriptional activity, rather than the presence of intact DNA, may play a more important role in driving inflammation in ART-treated individuals. Furthermore, the study linked aging to reduced control over HIV-1 transcription, highlighting how age-related immune changes may exacerbate viral transcription and its inflammatory consequences in long-term ART-treated individuals. In line with these results, Scherpenisse et al. demonstrated that HIV-1 transcription at an early time point during ART correlates with markers of immune activation, exhaustion, and apoptosis (36).

Another critical marker of HIV persistence, low-level residual viremia, has been associated with microbial translocation and systemic inflammation (37–39). Riddler et al. found that residual viremia was associated with higher CD8+ T cell counts and lower CD4/CD8 ratio in ART-treated individuals, suggesting incomplete immune recovery (40). In a study by Falasca et al., residual viremia showed a significant correlation with sCD14 levels, an established marker of microbial translocation and monocyte activation (41). However, this study found no correlations between HIV DNA, HIV RNA, and inflammatory markers, highlighting inconsistencies with previous findings. Finally, several studies on raltegravir intensification have reported reductions in HIV replication markers, such as 2-LTR circles or US HIV RNA, along with decreased immune activation levels (42–49).

On the other hand, some intensification studies have failed to replicate these findings (50–56) and several studies have not consistently found a strong relationship between HIV persistence markers and immune activation (35, 57–63). For example, Gandhi et al. proposed that inflammation and immune activation in ART-treated individuals might predominantly come from immunologic events occurring before ART initiation rather than being perpetuated by ongoing viral replication sustained by persistent low-level viremia during suppressive ART (63). Recently, Bailon et al. demonstrated that a larger reservoir size was initially associated with higher levels of soluble inflammatory biomarkers, as well as increased activation and exhaustion of CD4+ and CD8+ T cells (64). However, these associations disappeared upon achieving viral suppression, regardless of the antiretroviral regimen used (dual or triple therapy). Similarly, previous comparative studies between two-drug regimens (2DRs) and three-drug regimens (3DRs) have not shown consistent differences in inflammatory markers, residual viremia or T-cell activation and exhaustion markers (65–68).

Overall, our findings revealed limited association between HIV persistence markers – including both blood and tissue reservoirs – and systemic inflammation or immune activation. However, some of our observations aligned with previous research. Notably, we observed an association between NK cells and residual viremia, a relationship that has been documented in both adults and children (69–72). This finding underscores the pivotal role of NK cells as key effectors of the innate immune system, highlighting their potential contribution to the control and modulation of HIV reservoir cell dynamics.

Several factors could explain the lack of a strong link between HIV persistence and inflammation in our study. First, compared to earlier studies from Chomont and Hatano, our cohort exhibited relatively low reservoir sizes, with a median total HIV DNA of 131 copies/10⁶ cells in PBMCs and 231 copies/10⁶ cells in rectal tissue (31, 33). However, in comparison to the LoViret cohort described by Galvez et al., which reported exceptionally low reservoir levels in chronically treated HIV-1 infected individuals (<50 HIV-DNA copies/10⁶ PBMC), the reservoir sizes in our cohort were comparatively higher (73). Reservoir activity, as measured by unspliced RNA levels, was also limited (median: 112 copies/µg total RNA). Although direct comparison with earlier studies is challenging due to differences in sampling methods (e.g., CD4-enriched populations vs. PBMCs) and techniques, these values are in line with limited reservoir size and activity in these participants. However, in comparison to a recent study of early treated adolescents living with HIV and receiving long-term suppressive ART, our cohort exhibited higher levels of CA US HIV RNA, as that study reported a median of 19 copies/10⁶ PBMCs (74).

These discrepancies likely reflect key changes in the management of PWH, including the initiation of treatment regardless of CD4+ T cell count and advancements in ART regimens, particularly the widespread adoption of integrase inhibitor-based ART (75). The impact of these modifications in PWH care on HIV reservoir size and activity should not be underestimated. In recent cohorts, these changes have resulted in prolonged viral suppression, such as ours (median duration: 7.62 years), and treatment with more potent regimens such as the DTG-based regimen that was used by all participants of the present study.

Longitudinal studies evaluating the dynamics of HIV reservoir cells and the impact of the host immune system have recently emerged, alongside new technologies like the intact proviral DNA assay (IPDA) and near full-length individual proviral next-generation sequencing. These PCR-based methods have revealed significant differences in the longitudinal dynamics of intact versus defective proviruses. Intact proviruses decline more rapidly, likely due to the active immunological elimination of transcriptionally active reservoirs, which are more easily recognized and targeted by the immune system (76). Defective proviruses that encode proteins also exhibit a shorter half-life compared to those incapable of protein expression (77, 78). Over time, the HIV reservoir is thus progressively reshaped into a more silent and defective state under the persistent pressure of the immune system, which preferentially targets infected cells with higher residual proviral transcriptional activity (79).

We thus hypothesize that, in virally suppressed individuals, the contribution of the HIV reservoir to chronic immune activation and inflammation will diminish over time on suppressive ART.

In our cohort, this is likely reflected by a reservoir dominated by deeply latent or transcriptionally silent clones which contribute minimally to inflammation and immune activation. This aligns with the prolonged duration of viral suppression and the robust immune recovery observed in our participants (median CD4+ counts: 877 cells/mm³ and CD4/CD8 ratio ~ 1). In addition, all participants in this study were receiving an integrase inhibitor-based regimen. Integrase inhibitor-based therapies are known to achieve faster and probably stronger viral suppression compared to protease inhibitor or non-nucleoside reverse transcriptase inhibitor-based regimens (80–82).

These findings reinforce the idea that inflammation and immune activation in ART-treated PWH are influenced by a multifaceted interplay of factors extending beyond HIV persistence alone. Moreover, they suggest that the contribution of the viral reservoir to inflammation may vary substantially between individuals and even in the same individuals over time, depending on the clonal dynamics that dominate the proviral landscape. Such variability highlights the importance of considering both host and viral factors when evaluating the drivers of inflammation in ART-treated populations.

While our findings are informative, several limitations should be acknowledged. First, the modest sample size (n=49) may have limited statistical power to detect subtle correlations. Second, we did not evaluate the impact of common sexually transmitted infections (e.g., chlamydia, gonorrhoea, syphilis) or examine gut-microbiome dysbiosis in detail, even though emerging evidence links microbiome alterations to chronic inflammation and non-communicable diseases (83, 84). Third, cell-associated unspliced HIV RNA could not be measured reliably in rectal tissue. Fourth, key immunological features such as immune-senescence markers and the Th17/Treg balance were not assessed. Taken together, these constraints should be borne in mind when interpreting the study’s conclusions.

In conclusion, the narrative surrounding HIV persistence and its impact on inflammation has shifted. Earlier studies described a scenario where larger, more active reservoirs perpetuated immune activation and inflammation. In contrast, today’s PWH, treated earlier with more potent regimens, have smaller, less active reservoirs that exert a diminished impact on immune activation. This highlights the dynamic nature of reservoir evolution under long-term ART and underscores the importance of considering these shifts when interpreting contemporary findings.