Abstract
Introduction
Globally, 15-30% of patients with chronic hepatitis C develop compensated advanced chronic liver disease (cACLD). cACLD may further advance to decompensated ACLD (dACLD), which is associated with higher liver-related mortality, even after successful antiviral treatment. This progression is accelerated by HIV coinfection, yet its underlying molecular mechanisms remain poorly understood.
Methods
In this cross-sectional study, we characterized the plasma lipidomic profiles of 58 HIV/HCV-coinfected patients using untargeted liquid chromatography-mass spectrometry. Multivariate (OPLS-DA) and univariate (GLM) statistical models were employed to identify lipid species associated with disease severity.
Results
We identified a signature of 28 lipids—predominantly phosphatidylcholines, phosphatidylethanolamines, and triglycerides— that were significantly depleted in patients with dACLD (17.2% of the cohort). This systemic lipid depletion showed relevant correlations with the pro-inflammatory chemokine IP-10 and soluble immune checkpoint proteins.
Discussion
Our findings indicate that dACLD in HIV/HCV-coinfection is defined by a profound collapse in plasma lipids. This metabolic failure correlates with markers of inflammation and immune activation, suggesting that lipid dysregulation plays a critical role in the pathogenesis of liver decompensation.
Keywords: biomarkers, hepatitis C, HIV, lipidomics, liver decompensation
Graphical Abstract
1. Introduction
Globally, a substantial proportion of individuals with chronic hepatitis C will progress to compensated advanced chronic liver disease (cACLD), a critical clinical stage defined by advanced fibrosis or cirrhosis (1). While often asymptomatic, cACLD places patients at high risk of transitioning to a decompensated state (dACLD), which is associated with a substantially increased risk of liver-related mortality (1). A central clinical challenge is that this risk persists even after achieving a virological cure for HCV, as the underlying drivers of liver injury often remain active (2). This issue is particularly exacerbated in the context of HIV coinfection, which is known to accelerate the progression to dACLD and its associated complications (3). Despite extensive characterization of the clinical predictors of liver decompensation, the underlying molecular mechanisms that trigger this shift remain poorly understood.
In advanced liver disease, the liver’s central role in lipid synthesis and transport becomes progressively impaired, leading to profound systemic lipid dysregulation (4, 5). These metabolic derangements may not simply reflect liver injury but could be linked to broader metabolic and immune disturbances. Lipid species, particularly glycerophospholipids (GP), are integral to cell membrane structure, and their alteration can compromise hepatocyte integrity and intracellular signalling (4, 6). Crucially, this lipid dysregulation is mechanistically linked to the immune dysfunction that characterizes advanced liver disease (7). Lipids can act as signalling molecules that directly modulate inflammatory pathways; for instance, changes in the lipid composition of T-cell membranes can impair their activation and contribute to the state of immune exhaustion frequently observed in chronic viral infections (7, 8).
Predicting the transition to dACLD remains clinically challenging, as it often occurs without clear preceding signs (1). Consequently, there is an urgent need for molecular biomarkers that reflect the underlying pathophysiology and can identify patients at the highest risk before irreversible deterioration occurs (2). Lipidomics has emerged as a powerful systems-level approach to capture these metabolic perturbations and their link to disease progression. Although prior metabolomic studies in HIV/HCV coinfection have reported associations between lipid−related pathways and advanced liver disease (9), to our knowledge none have specifically characterized the plasma lipidomic profile associated with liver decompensation in this high-risk population using a comprehensive untargeted lipidomics workflow.
This study aimed to characterize the plasma lipidomic profile that differentiates dACLD from cACLD in a high-risk cohort of patients with HIV/HCV-coinfection, and to investigate its links to systemic inflammation and immune dysregulation.
2. Materials and methods
2.1. Study design and participants
We conducted a cross-sectional study nested within the GeSIDA 10318 cohort (Supplementary Data 1). We included HIV/HCV-coinfected patients with advanced chronic liver disease (ACLD), defined as liver stiffness measurement (LSM) ≥10 kPa), obtained using transient elastography (FibroScan). Inclusion criteria required stable antiretroviral therapy (ART) for >6 months with undetectable HIV viral load and available frozen plasma samples. Exclusion criteria were hepatitis B coinfection, acute hepatitis C, or hepatocellular carcinoma.
2.2. Outcome variable
The outcome of interest was dACLD, defined by the presence of a Child-Turcotte-Pugh (CTP) score ≥7 or a history of clinical decompensation (ascites, bleeding esophageal varices, or hepatic encephalopathy). Patients not meeting these criteria were classified as compensated ACLD (cACLD). This combined definition integrates both functional impairment and clinical events, allowing a comprehensive characterization of advanced liver disease severity within the study cohort.
2.3. Non-targeted lipidomics analysis
Protocols regarding the reagents and standards, lipid extraction, quality management assurance, analytical conditions, lipid annotation, and data reprocessing and normalization can be found in Supplementary Data 2
2.4. Multiplex immunoassays and ELISA
Plasma markers were measured using Luminex 200™ technology (Luminex Corp., Austin, TX). We quantified inflammatory markers (IL-8, IL-18, IL-1RA, IP-10, MCP-1, TNF-RI) using ProcartaPlex™ assays. Additionally, soluble immune checkpoint proteins (ICPs) were measured using Immuno-Oncology Checkpoint Panels 1 and 3 (Invitrogen™), listed in Supplementary Data 2.
2.5. Statistical analysis
For the descriptive analysis, continuous variables were compared using the Mann-Whitney U test and categorical variables using the Chi-square test.
For lipidomic analysis, outliers were removed using the standard 1.5×IQR rule, and data were log-transformed (log10) and auto-scaled. Orthogonal partial least squares discriminant analysis (OPLS-DA) was performed to identify influential lipids (VIP >1), which was strictly used as an exploratory, heuristic feature-screening step to reduce dimensionality due to subgroup unbalance. The final 28-lipid signature was selected based on a sequential combination of criteria: only lipids with an OPLS-DA VIP score > 1.0 that subsequently achieved both statistical significance (adjusted GLM p-value < 0.05) and an exploratory false discovery rate threshold (FDR q-value < 0.20) in the covariate-adjusted GLMs were included. Significant lipids were subsequently determined using Generalized Linear Models (GLM) with a gamma distribution (log-link), adjusted for clinically relevant covariates such as age, gender, BMI, HCV viral load, and alcohol intake, previously selected by a stepwise method (forward), according to the specific model’s lowest Akaike information criterion (AIC). This stepwise selection was performed to identify the most parsimonious model for each lipid, preserving statistical power and preventing overfitting by adjusting only for covariates that genuinely contributed to model fit. P-values were corrected for multiple testing (False Discovery Rate (FDR)). Statistical significance was defined as a p-value<0.05 (two-tailed) and q-value <0.20, consistent with the exploratory nature of the analysis. An FDR threshold of 20% was chosen to balance the discovery of biologically meaningful lipidomic signatures with the control of false-positive rates in a discovery-phase clinical cohort, where tight physiological correlation among lipids can inflate false-negative rates under overly conservative corrections. This multi-step framework (OPLS-DA feature pre-screening, AIC-based covariate selection, and FDR correction) was specifically implemented to minimize multiple-testing inflation and reduce the risk of overfitting in the identification of the final lipid signature.
Associations between significant dACLD-related lipids and inflammatory biomarkers/immune checkpoint proteins (ICPs) were assessed using Spearman correlation. Correlations with r >0.25 or r <-0.25 were considered relevant. All correlation p-values were corrected for multiple testing using the FDR correction, and statistical relevance was defined as meeting both an unadjusted p-value < 0.05 and an FDR q-value < 0.20.
Analyses were performed using MetaboAnalyst 5.0 and R v4.3.1.
3. Results
3.1. Individuals’ characteristics
Baseline characteristics are detailed in Table 1. The cohort included 58 HIV/HCV-coinfected patients, of whom 10 (17.2%) had dACLD. Patients in the dACLD group had a history of decompensation events (ascites, variceal bleeding, or encephalopathy) and/or CTP ≥ 7, and also exhibited higher liver stiffness measurements. Other demographic and virological variables, including age, gender, and HCV genotype, were similar between groups. Biochemical markers are shown in Supplementary Table 1.
Table 1.
Clinical, epidemiological, and virological characteristics of HIV/HCV-coinfected patients according to decompensated and compensated Advanced chronic liver disease.
| Variable | All patients | dACLD | cACLD | p |
|---|---|---|---|---|
| No. | 58 | 10 (17.2%) | 48 (82.8%) | |
| Age (years) | 51 (48–54) | 53 (49–54) | 51 (48–53) | 0.635 |
| Gender (male) | 44 (79.5%) | 8 (80.0%) | 36 (75.0%) | 0.999 |
| BMI (kg/m2) (n =57) | 24.4 (21.9–26.9) | 24.3 (23.6–25.0) | 24.4 (21.6–27.0) | 0.906 |
| Smoker (n = 56) | 0.808 | |||
| Never | 6 (10.7%) | 1 (11.1%) | 5 (10.6%) | |
| Previous (>6 months) | 14 (25.0%) | 3 (33.3%) | 11 (23.4%) | |
| Current | 36 (64.3%) | 5 (55.6%) | 31 (66.0%) | |
| Alcohol intake (>50g/day) (n = 57) | 0.071 | |||
| Never | 27 (47.4%) | 2 (20.0%) | 25 (53.2%) | |
| Previous (>6 months) | 27 (47.4%) | 8 (80.0%) | 19 (40.4%) | |
| Current | 3 (5.3%) | 0 (0.0%) | 3 (6.4%) | |
| Intravenous drug user | 0.537 | |||
| Never | 13 (22.4%) | 1 (10.0%) | 12 (25.0%) | |
| Previous (>6 months) | 45 (77.6%) | 9 (90.0%) | 36 (75.0%) | |
| Current | 0 (0%) | 0 (0%) | 0 (0%) | |
| Previous HCV therapy | 36 (62.1%) | 7 (70.0%) | 29 (60.4%) | 0.834 |
| Liver markers | ||||
| HSI (n = 56) | 33.8 (29.3–37.2) | 32.1 (29.9–37.1) | 33.8 (29.2–37.2) | 0.983 |
| LSM (kPa) | 26.0 (17.3–35.8) | 34.7 (18.6–42.3) | 25.7 (17.2–35.1) | 0.308 |
| APRI | 1.9 (1.0–3.4) | 1.3 (1.0–3.3) | 1.9 (1.0–3.5) | 0.536 |
| TyG (n = 57) | 8.7 (8.4–9.0) | 8.6 (8.3–8.9) | 8.7 (8.4–9.0) | 0.796 |
| TGHDL (n = 55) | 2.8 (1.9–5.5) | 2.8 (2.2–2.9) | 2.8 (1.8–5.8) | 0.838 |
| METS-IR (n =53) | 2.4 (2.3–2.7) | 2.4 (2.4–2.5) | 2.4 (2.3–2.7) | 0.753 |
| dACLD | ||||
| CTP≥7 | 1 (1.7%) | 1 (10.0%) | 0 (0.0%) | NA |
| Ascites (n = 57) | 5 (8.8%) | 5 (50.0%) | 0 (0.0%) | |
| Bleeding esophageal varices | 3 (5.2%) | 3 (30.0%) | 0 (0.0%) | |
| Hepatic encephalopathy | 1 (1.7%) | 1 (10.0%) | 0 (0.0%) | |
| HCV markers | ||||
| HCV genotype | 0.816 | |||
| 1 | 43 (74.1%) | 7 (70.0%) | 36 (75.0%) | |
| 3 | 7 (12.1%) | 1 (10.0%) | 6 (12.5%) | |
| 4 | 6 (10.3%) | 1 (10.0%) | 5 (10.4%) | |
| Others | 2 (3.4%) | 1 (10.0%) | 1 (2.1%) | |
| Log10 HCV-RNA (IU/mL) (n = 57) | 6.2 (5.7–6.7) | 6.0 (5.2–6.1) | 6.3 (5.8–6.7) | 0.064 |
| HCV-RNA > 850.000 IU/mL) | 38 (65.5%) | 5 (50.0%) | 33 (68.8%) | 0.442 |
| HIV markers | ||||
| Previous AIDS (n = 57) | 1 (1.8%) | 0 (0.0%) | 1 (2.1%) | 0.999 |
| Nadir CD4+/mm3 (n = 55) | 135 (84–244) | 158 (95–201) | 130 (80–245) | 0.735 |
| Nadir < 200 CD4+/mm3 (n = 57) | 39 (68.4%) | 7 (70.0%) | 32 (68.1%) | 0.999 |
| CD4+ T-cells/mm3 | 470 (303–700) | 453 (334–604) | 487 (314–714) | 0.742 |
| < 500 CD4+/mm3 | 31 (53.4%) | 6 (60.0%) | 25 (52.1%) | 0.914 |
| HIV antiretroviral therapy (n = 53) | 0.137 | |||
| NRTI + NNRTI | 18 (34.0%) | 0 (0.0%) | 18 (40.9%) | |
| NRTI + II | 24 (45.3%) | 7 (77.8%) | 17 (38.6%) | |
| NRTI + PI | 6 (11.3%) | 1 (11.1%) | 5 (11.4%) | |
| Others | 5 (9.4%) | 1 (11.1%) | 4 (9.1%) |
Statistics: The values are expressed as the absolute number (percentage) and median (interquartile range). P-values were calculated by the Chi-square test and the Mann-Whitney U test. dACLD, decompensated advanced chronic liver disease; cACLD, compensated advanced chronic liver disease; HSI, hepatic steatosis index; BMI, body mass index; HCV, hepatitis C virus; LSM, liver stiffness measurement; kPa, kilopascal; CTP, Child-Turcotte-Pugh score; APRI, AST to platelet ratio index; TyG, Triglycerides and glucose index; TGHDL, triglyceride/HDL-cholesterol ratio; METS-IR, Metabolic score for insulin resistance; NA, not applicable; HCV-RNA, viral load of hepatitis C; AIDS, acquired immune deficiency syndrome; NRTI, nucleoside analogue HIV reverse transcriptase inhibitor; NNRTI, non-nucleoside analogue HIV reverse transcriptase inhibitor; II, HIV integrase inhibitor; PI, HIV protease inhibitor.
3.2. Lipidomic signature of liver decompensation
We identified 566 distinct plasma lipid species, classified according to the LIPID MAPS structure database (LMSD) (Figures 1A, B).
Figure 1.
Lipidomic analysis in patients with HIV/HCV according to decompensated advanced chronic liver disease (dACLD). (A) Proportion of main classes of identified lipids. (B) Proportion of lipid subclasses in each main class. (C) Proportion of main classes of differential lipids according to dACLD. (D) Counts of up- or down-regulated differential lipids in dACLD vs cACLD patients with HIV/HCV. Glycerolipids (GL) are referred to monoglycerides (MG), diglycerides (DG) and triglyceride (TG); glycerolphospholipids (GP) to lysophosphatidylcholines (LPC), lysophosphatidylethanolamines (LPE), lysophosphatidylinositols (LPI), phosphatidylcholines (PC), phosphatidylethanolamines (PE), phosphatidylinositols (PI), phosphatidylserines (PS) and phosphatidic acids (PA); sphingolipids (SP) to sphingomyelins (SM) and ceramides (Cer); and fatty acyls to fatty acids (FA), fatty acyl ester of hydroxy fatty acid (FAHFA), N-acyl amides and Acyl carnitines.
Multivariate OPLS-DA models were utilized to explore the separation between dACLD and cACLD patients (Supplementary Figure 1A, B). For the positive ionization mode (LC-MS ESI+), the model exhibited a cumulative R2Yof 0.734 (p = 0.216 under 1, 000 permutations) and a predictive Q2 of −0.458 (p = 0.690 under 1, 000 permutations) (Supplementary Figure 1C). Similarly, the negative ionization mode (LC-MS ESI-) model showed a cumulative R2Y of 0.423 (p = 0.374 under 1, 000 permutations) and a predictive Q2 of −0.274 (p = 0.491 under 1, 000 permutations) (Supplementary Figure 1D). These parameters (negative Q2 values and non-significant permutation p-values) indicate global multivariate model overfitting due to subgroup unbalance (n=10 dACLD vs. n=48 cACLD); therefore, the OPLS-DA was strictly employed as a heuristic feature-screening step to select lipids with a VIP score > 1.0. Subsequently, adjusted GLM analysis of these VIP-filtered lipids identified a specific signature of 28 lipids significantly associated with dACLD. Notably, all 28 lipids were depleted in patients with dACLD compared to cACLD (Figure 1C, D; Supplementary Table 2).
This signature was dominated by glycerophospholipids (67.9%), particularly phosphatidylcholines (PC) and phosphatidylethanolamines (PE), followed by glycerolipids (21.4%), primarily triglycerides (TG). Structurally, these depleted lipids were exclusively long-chain species, predominantly polyunsaturated (82.1%), and included a substantial subset of plasmalogens (Figure 2).
Figure 2.
Plasma lipids associated with decompensated advanced chronic liver disease (dACLD) in patients with HIV/HCV. Statistics: Data were calculated by Generalized Linear Models (GLM) with a gamma distribution (log-link), adjusted by age, gender, body mass index (BMI), HCV viral load, and alcohol intake, previously selected by a stepwise method (forward). The q-values represent p-values corrected for multiple testing using the False Discovery Rate (FDR). aAMR, adjusted arithmetic mean ratio; DG, diglyceride; TG, triglyceride; LPC, lysophosphatidylcholine; LPE, lysophosphatidylethanolamine; LPI, lysophosphatidylinositol; PC, phosphatidylcholine; PC O, plasmalogen PC; PE, phosphatidylethanolamine; PE O, plasmalogen PE; PI, phosphatidylinositol; PS; phosphatidylserine; PA, phosphatidic acid; SM, sphingomyelin; FA, fatty acid.
3.3. Correlation between significant lipids and immune biomarkers
Correlation analysis across the entire cohort (N = 58) revealed distinct associations between the lipidomic signature and immune markers (Figure 3). Depleted structural lipids (PC and PE) showed widespread negative correlations with the pro-inflammatory chemokine IP-10 and the immune checkpoints HVEM and CD134. In contrast, specific depleted TG species (e.g., TG 42:1, TG 58:7) exhibited positive correlations with the inhibitory checkpoint BTLA and the inflammatory marker S100A8/A9.
Figure 3.
Spearman correlation plot between significant plasma lipids and plasma (A) inflammatory-related biomarkers and (B) immune checkpoint proteins (ICPs) across the entire study cohort (N = 58). The size of the circles is proportional to the strength of the correlation. The color represents the direction (colour legends are shown on the right), whereas large dark red and blue represent strong positive and negative correlations, respectively. Lipids are on the vertical axis, and inflammatory-related biomarkers and immune checkpoint proteins are on the horizontal axis. Those correlations with rho>0.25 o rho<-0.25, p-value<0.05, and q-value<0.2 are shown with an asterisk. DG, diglyceride; TG, triglyceride; LPC, lysophosphatidylcholine; LPE, lysophosphatidylethanolamine; LPI, lysophosphatidylinositol; PC, phosphatidylcholine; PC O, plasmalogen PC; PE, phosphatidylethanolamine; PE O, plasmalogen PE; PI, phosphatidylinositol; PS; phosphatidylserine; PA, phosphatidic acid; SM, sphingomyelin; FA, fatty acid; IL, interleukin; IP-10, human interferon-inducible protein 10; MCP-1, monocyte chemoattractant protein-1; IL-1RA, IL-1 receptor antagonist; TNF-RI, tumor necrosis factor receptor-1; BTLA, B, and T lymphocyte attenuator; CD, cluster of differentiation; GITR, glucocorticoid-induced TNFR-related; HVEM, herpesvirus entry mediator; IDO, indoleamine 2, 3-dioxygenase; LAG-3, lymphocyte activation gene-3; PD-1, programmed cell death protein 1; PD-L1, programmed death-ligand 1; PD-L2, programmed death-ligand 2; TIM-3, T-cell immunoglobulin and mucin-domain containing-3; TIMD-4, T-cell immunoglobulin and mucin domain-containing protein 4; VISTA(B7-H5), V-domain Ig suppressor of T cell activation.
4. Discussion
This study characterizes the plasma lipidomic profile associated with dACLD in patients with HIV/HCV-coinfection. Our findings reveal a distinct plasma lipid profile in dACLD, primarily defined by a significant reduction of GP and GL. Furthermore, the consistent correlations observed between these depleted lipids and key markers of inflammation and immune dysregulation, point to a close interplay between impaired lipid metabolism and the systemic immune activation that fuels the advancement of liver disease. This specific signature of lipid depletion is likely driven by the complex interplay between HIV/HCV coinfection and progressive loss of hepatic reserve.
The significant depletion of PC and PE observed in dACLD patients highlights a critical disruption in hepatic phospholipid homeostasis. Reduced plasma PC levels are consistent with previous reports linking lipid alterations to fibrosis progression (4, 10) and likely reflect the impaired hepatic synthesis and secretion of lipoproteins characteristic of advanced liver failure (6). Similarly, the decrease in PE could reflect potential alterations consistent with mitochondrial dysfunction, given PE’s essential role in maintaining mitochondrial inner membrane structure and the electron transport chain (11). Since disrupted PE homeostasis has been previously linked to ER stress and apoptosis (5), this combined depletion of structural lipids may point to an altered hepatocyte integrity potentially contributing to the severity of liver decompensation.
Regarding GL, while routine clinical measurements (total triglycerides and cholesterol) failed to distinguish between cACLD and dACLD, lipidomic profiling revealed a specific and profound depletion of long-chain TG in the decompensated group. This discrepancy underscores the limitations of standard biochemical panels, which only measure total lipid pools and lack the molecular resolution to detect the selective remodeling of lipid species. Similarly, conventional functional markers (such as glucose, albumin, or bilirubin) often remain homeostatically buffered or clinically managed, failing to reflect early subcellular perturbations. In contrast, untargeted lipidomics bypasses these limitations by directly capturing early, specific alterations, which signal systemic metabolic collapse before standard clinical markers deviate significantly. This systemic hypolipidemia likely reflects the progressive failure of hepatic very-low-density lipoproteins (VLDL) synthesis and secretion, a characteristic feature of advanced liver disease (6, 12). Mechanistically, impaired VLDL assembly—dependent on apolipoprotein B and microsomal triglyceride transfer protein (MTTP)—leads to a paradox where intracellular TG retention coexists with low plasma TG levels (13). In the context of chronic hepatitis C, this reduction in serum VLDL-TG fractions parallels fibrosis progression and diminished hepatic export capacity (14). Consequently, the observed depletion of long−chain TG could reflect disrupted hepatic lipid handling in advanced liver disease.
A detailed examination of the 28 depleted lipid species reveals specific structural characteristics that provide deeper insight into the pathophysiology of dACLD. Notably, these lipids were predominantly polyunsaturated (PUFAs) and included a significant subset of plasmalogens. The selective depletion of these long-chain polyunsaturated species reflects specific pathological vulnerabilities. First, the high density of double bonds in PUFAs renders them highly susceptible to lipid peroxidation under the heightened oxidative stress conditions that characterize HIV/HCV coinfection and hepatic decompensation (15), leading to their preferential degradation. Second, the synthesis of long-chain PUFAs is highly dependent on liver-specific elongases and desaturases, whose activity is markedly impaired during the progressive loss of functional hepatocyte volume in dACLD (8). Third, the selective impairment of hepatic lipoprotein packaging (VLDL) prevents the systemic distribution of these newly synthesized PUFA-containing lipids (14). Together, these mechanisms explain why the loss of these complex long-chain species serves as a highly sensitive biomarker of systemic metabolic failure, whereas standard cumulative lipid measurements remain unaltered.
This selective depletion suggests a dual pathological impact. First, the loss of PUFAs implies a structural deficit leading to increased membrane rigidity and a compromised capacity to generate anti-inflammatory lipid mediators (16). Second, the reduction in plasmalogens—key endogenous antioxidants—strongly indicates a state of heightened oxidative stress where protective lipids are consumed faster than they are synthesized (17). This loss of antioxidant defense likely accelerates liver damage, contributing to mitochondrial dysfunction and cell death. Collectively, this generalized loss of long-chain lipids, which are primarily synthesized and packaged by the liver, underscores the progression from a localized hepatic pathology to a systemic metabolic collapse, reflecting the liver’s inability to supply the body with the essential structural and energetic building blocks required for systemic homeostasis (6).
Our findings suggest these metabolic derangements are intricately linked to the immune dysregulation characteristic of advanced liver disease. Notably, the correlational patterns differ markedly between structural (PC and PE) and energy-storage lipids (TG), revealing distinct and potentially complementary roles in the metabolic-immune interactions that contribute to liver decompensation.
Structural phospholipids, such as PC and PE, were consistently and strongly negatively associated with the pro-inflammatory chemokine IP-10 and the co-stimulatory ICPs HVEM and CD134. IP-10 is a well-known marker of inflammation and liver injury (18), often elevated in chronic HCV and HIV infections (19), while HVEM and CD134 are critical for T-cell activation (20). This inverse relationship suggests that as the liver’s capacity to maintain phospholipid homeostasis declines, a pro-inflammatory and immune-activated state intensifies. The depletion of these structural lipids may disrupt cell membrane integrity and signaling, potentially exacerbating immune dysregulation and contributing to T-cell exhaustion, factors that could play a role in accelerating disease progression. This observation is consistent with established evidence that lipid dysregulation contributes to hepatic inflammation and oxidative stress (4, 5).
In contrast, energy-storage TG lipids showed positive correlations with the inhibitory checkpoint BTLA and the inflammatory marker S100A8/A9. BTLA suppresses T-cell activation (21), while S100A8/A9 modulates inflammatory responses (22). In the context of dACLD, where circulating TG are low due to hepatic failure, these positive correlations may indicate that this profound metabolic collapse is linked to a dysfunctional immune state unable to mount effective responses (7). These findings highlight that different lipid classes are associated with divergent immune pathways that collectively contribute to liver decompensation.
Regarding clinical translation, while untargeted LC-MS serves as a powerful discovery-phase tool, its high complexity and cost currently limit its direct bedside application. However, for clinical implementation, this 28-lipid signature, or a reduced subset of these biomarkers, could be adapted into targeted LC-MS/MS assays, which are more suitable for routine clinical workflows. Although further validation and standardization are required, such approaches may facilitate early risk stratification and preventive management of patients at high risk of decompensation, potentially reducing the burden of advanced liver disease.
Our study is limited by its cross-sectional design, which precludes causal inference. Additionally, although the sample size was statistically adequate for metabolomic analysis based on MetSizeR, these findings require validation in larger, independent cohorts, as the relatively small number of patients with dACLD underscores their exploratory nature. We also acknowledge that our global multivariate OPLS-DA models exhibited overfitting (reflected by negative Q2 values and non-significant permutation tests), which is a common limitation in heterogeneous clinical cohorts with unbalanced subgroups. To mitigate this multidimensional overfitting, OPLS-DA was strictly employed as an initial heuristic screening tool to reduce dataset dimensionality (VIP > 1.0). Crucially, the final 28-lipid signature was determined and validated using covariate-adjusted univariate GLMs coupled with Benjamini-Hochberg FDR correction, supporting the robustness of the identified association while accounting for key clinical confounders. Nevertheless, given the exploratory nature of this study, both the identified 28-lipid signature and its correlation patterns with immune biomarkers must be interpreted with caution as hypothesis-generating trends that require validation in larger, independent prospective cohorts before any clinical translation.
In conclusion, we identified a profound depletion of plasma GP and GL as a lipid signature associated with ACLD in HIV/HCV coinfection. This metabolic pattern correlates with systemic inflammation and immune checkpoint activation, suggesting a critical interplay between lipid dysregulation and viral-related immune dysregulation. These lipids represent potential exploratory biomarkers associated with liver decompensation that require prospective validation in larger, independent cohorts before clinical implementation.
Acknowledgments
This research was made possible thanks to the collaboration of patients, medical and nursing staff, and data managers. We extend our gratitude to the HIV BioBank integrated into the Spanish AIDS Research Network and collaboration centers for providing clinical samples. The HIV BioBank is partially funded by the RD16/0025/0019 project, under the Plan Nacional R+D+I, and co-funded by ISCIII-FEDER.
Funding Statement
The author(s) declared that financial support was received for this work and/or its publication. This study was supported by grants from Instituto de Salud Carlos III (ISCIII; grant numbers PI24CIII/00003 (SR and RM-E) and PI21CIII/00033 (MJ-S)). The study was also funded by the CIBER -Consorcio Centro de Investigación Biomédica en Red-(CB 2021), Instituto de Salud Carlos III, Ministerio de Ciencia e Innovación, and Unión Europea -NextGenerationEU (CB21/13/00044). RM-E is César Nombela researcher supported and funded by Comunidad de Madrid (grant number 2023-T1/SAL-GL-28980).
Footnotes
Edited by: Alpo Juhani Vuorio, University of Helsinki, Finland
Reviewed by: Masahiko Ito, Hamamatsu University School of Medicine, Japan
Joanna Gdula-Argasinska, Jagiellonian University Medical College, Poland
Data availability statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation. Requests to access these datasets should be directed to jimenezsousa@isciii.es.
Ethics statement
The studies involving humans were approved by the Research Ethics Committee of the Institute of Health Carlos 241 III (CEI PI 72_2021) and was conducted following the Declaration of Helsinki. The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.
Author contributions
RB-L: Investigation, Writing – review & editing. BF: Investigation, Writing – review & editing. JB: Data curation, Investigation, Writing – review & editing. AV-B: Investigation, Writing – review & editing. JG-G: Investigation, Writing – review & editing. CG-R: Investigation, Writing – review & editing. CD: Investigation, Writing – review & editing. VH: Investigation, Writing – review & editing. AV-R: Investigation, Writing – review & editing. CB: Writing – review & editing. SR: Conceptualization, Funding acquisition, Methodology, Writing – review & editing. RM-E: Conceptualization, Data curation, Formal analysis, Funding acquisition, Methodology, Supervision, Visualization, Writing – original draft. MJ-S: Conceptualization, Funding acquisition, Methodology, Project administration, Supervision, Visualization, Writing – review & editing.
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declared that generative AI was not used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Supplementary material
The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fimmu.2026.1878207/full#supplementary-material
Members of the GeSIDA 10318 Study Group.
Additional description of methods section.
Biochemical markers of patients with HIV/HCV according to decompensated and compensated Advanced chronic liver disease.
Significant associations of lipids with decompensated advanced chronic liver disease (dACLD) in patients with HIV/HCV.
References
- 1. de Franchis R, Baveno V. Expanding consensus in portal hypertension: Report of the Baveno VI Consensus Workshop: Stratifying risk and individualizing care for portal hypertension. J Hepatol. (2015) 63:743–52. doi: 10.1016/j.jhep.2015.05.022 [DOI] [PubMed] [Google Scholar]
- 2. Mandorfer M, Simbrunner B. Prevention of first decompensation in advanced chronic liver disease. Clin Liver Dis. (2021) 25:291–310. doi: 10.1016/j.cld.2021.01.003 [DOI] [PubMed] [Google Scholar]
- 3. Manns MP, Buti M, Gane E, Pawlotsky JM, Razavi H, Terrault N, et al. Hepatitis C virus infection. Nat Rev Dis Primers. (2017) 3:17006. doi: 10.1002/9781118637272.ch20 [DOI] [PubMed] [Google Scholar]
- 4. Heymann CJF, Mak AL, Holleboom AG, Verheij J, Shiri-Sverdlov R, van Mil SWC, et al. The plasma lipidome varies with the severity of metabolic dysfunction-associated steatotic liver disease. Lipids Health Dis. (2024) 23:402. doi: 10.1186/s12944-024-02380-x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. van der Veen JN, Kennelly JP, Wan S, Vance JE, Vance DE, Jacobs RL. The critical role of phosphatidylcholine and phosphatidylethanolamine metabolism in health and disease. Biochim Biophys Acta Biomembr. (2017) 1859:1558–72. doi: 10.1016/j.bbamem.2017.04.006 [DOI] [PubMed] [Google Scholar]
- 6. Acuñña M, Cohen DE, Feingold KR, Adler RA, Ahmed SF, Anawalt B, et al. Lipid and lipoprotein metabolism in liver disease. South Dartmouth (MA): Endotext. (2000). [Google Scholar]
- 7. Irvine KM, Ratnasekera I, Powell EE, Hume DA. Causes and consequences of innate immune dysfunction in cirrhosis. Front Immunol. (2019) 10. doi: 10.3389/fimmu.2019.00293 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Yang LJ, Tang M, Yang L. Lipid metabolism, viral infection, and antiviral immunity: a new host-pathogen interface. Front Cell Infect Microbiol. (2026) 16:1765502. doi: 10.3389/fcimb.2026.1765502. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Naggie S, Lusk S, Thompson JW, Mock M, Moylan C, Lucas JE, et al. Metabolomic signature as a predictor of liver disease events in patients with HIV/HCV coinfection. J Infect Dis. (2020) 222:2012–20. doi: 10.1093/infdis/jiaa316 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Virseda-Berdices A, Requena B, Berenguer J, Gonzalez-Garcia J, Gonzalez-Riano C, Diez C, et al. Long-term effects of HCV eradication on lipid profiles associated with MASLD among people with HIV with advanced fibrosis or cirrhosis. J Infect Public Health. (2025) 18:102981. doi: 10.1016/j.jiph.2025.102981 [DOI] [PubMed] [Google Scholar]
- 11. Decker ST, Funai K. Mitochondrial membrane lipids in the regulation of bioenergetic flux. Cell Metab. (2024) 36:1963–78. doi: 10.1016/j.cmet.2024.07.024 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Sundaram M, Yao Z. Recent progress in understanding protein and lipid factors affecting hepatic VLDL assembly and secretion. Nutr Metab (Lond). (2010) 7:35. doi: 10.1186/1743-7075-7-35 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Kawano Y, Cohen DE. Mechanisms of hepatic triglyceride accumulation in non-alcoholic fatty liver disease. J Gastroenterol. (2013) 48:434–41. doi: 10.1007/s00535-013-0758-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Nishimura M, Yamamoto H, Yoshida T, Seimiya M, Sawabe Y, Matsushita K, et al. Decreases in the serum VLDL-TG/non-VLDL-TG ratio from early stages of chronic hepatitis C: alterations in TG-rich lipoprotein levels. PloS One. (2011) 6:e17309. doi: 10.1371/journal.pone.0017309 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Gravier-Hernandez R, Gil-Del Valle L, Valdes-Alonso L, Hernandez-Ayala N, Bermudez-Alfonso Y, Hernandez-Requejo D, et al. Oxidative stress in hepatitis C virus-human immunodeficiency virus co-infected patients. Ann Hepatol. (2020) 19:92–8. doi: 10.1016/j.aohep.2019.05.009 [DOI] [PubMed] [Google Scholar]
- 16. Calder PC. Polyunsaturated fatty acids, inflammation, and immunity. Lipids. (2001) 36:1007–24. doi: 10.1007/s11745-001-0812-7 [DOI] [PubMed] [Google Scholar]
- 17. Wallner S, Schmitz G. Plasmalogens the neglected regulatory and scavenging lipid species. Chem Phys Lipids. (2011) 164:573–89. doi: 10.1016/j.chemphyslip.2011.06.008 [DOI] [PubMed] [Google Scholar]
- 18. Wada N, Takaki A, Ikeda F, Yasunaka T, Onji M, Nouso K, et al. Serum-inducible protein (IP)-10 is a disease progression-related marker for non-alcoholic fatty liver disease. Hepatol Int. (2017) 11:115–24. doi: 10.1007/s12072-016-9773-y [DOI] [PubMed] [Google Scholar]
- 19. Roe B, Coughlan S, Hassan J, Grogan A, Farrell G, Norris S, et al. Elevated serum levels of interferon- gamma -inducible protein-10 in patients coinfected with hepatitis C virus and HIV. J Infect Dis. (2007) 196:1053–7. doi: 10.1086/520935 [DOI] [PubMed] [Google Scholar]
- 20. Watts TH. TNF/TNFR family members in costimulation of T cell responses. Annu Rev Immunol. (2005) 23:23–68. doi: 10.1146/annurev.immunol.23.021704.115839 [DOI] [PubMed] [Google Scholar]
- 21. Ning Z, Liu K, Xiong H. Roles of BTLA in immunity and immune disorders. Front Immunol. (2021) 12. doi: 10.3389/fimmu.2021.654960 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Wang S, Song R, Wang Z, Jing Z, Wang S, Ma J. S100A8/A9 in inflammation. Front Immunol. (2018) 9. doi: 10.3389/fimmu.2018.01298 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Members of the GeSIDA 10318 Study Group.
Additional description of methods section.
Biochemical markers of patients with HIV/HCV according to decompensated and compensated Advanced chronic liver disease.
Significant associations of lipids with decompensated advanced chronic liver disease (dACLD) in patients with HIV/HCV.
Data Availability Statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation. Requests to access these datasets should be directed to jimenezsousa@isciii.es.