| |
Association of immune-activation and senescence markers with non-AIDS-defining comorbidities in HIV-suppressed patients - Activation and Senescence Markers in HIV Patients With Chronic Kidney Disease: "A higher CIADIS score (Higher IA and IS levels) [immune activation, senescence markers] was independently associated with advanced CKD.
|
| |
| |
Download the PDF here
Download the PDF here
Download the PDF here
"We studied the link between T-cell activation, differentiation and senescence phenotypes and non-AIDS-related comorbidities in HIV-suppressed patients.......In this cross-sectional study, we aimed at studying the link between immune activation, and immunosenescence phenotypes and non-AIDS-related comorbidities in HIV-infected patients with undetectable viral load......
The association of immune activation and immunosenescence with non-AIDS-related comorbidities has been widely studied in the past few years [20-23]. However, one of the limits of these studies is that immune markers defining activation, differentiation, and senescence were not studied simultaneously
We summarized immune markers [CD4+ and CD8+ activation (DR+), differentiation (naive and terminally differentiated memory T cells), and senescence (CD57+CD28-)] in a weighted immune score by principal component analysis called CIADIS.
Higher CD4+ and CD8+ T-cell activation were significantly associated with the presence of at least three comorbidities (respectively, P = 0.0004, P = 0.0001). Higher percentage of senescent and terminally differentiated T cells, and lower percentage of naïve T cells were also significantly correlated with a high comorbidity score (Supplementary Material Table 3, http://links.lww.com/QAD/A751)."
----------------------------
CROI 2016: Activation and Senescence Markers in HIV Patients With Chronic Kidney Disease: "A higher CIADIS score (Higher IA and IS levels) [immune activation, senescence markers] was independently associated with advanced CKD. Although other factors may contribute to kidney damage, persisting T-cells activation and senescence could induce inflammation and thus play a direct role even in successfully treated patients."
eGFR was ≥90, 60-89 and <60 in 68 %, 28% and 4% of patients, respectively......The proportion of all comorbidities was higher among patients with an eGFR <60 (p<0.01), except for hypertension and degenerative CNS disorders. Patients with an eGFR <60 were also more likely to present two or more comorbidities (57% vs. 29%, p<0.01).....The CIADIS score was determined with CD4 and CD8 activation (DR+), maturation (T naive, memory) and senescence (CD57+CD28-) markers....In univariable analysis: an increase of the CIADIS score was significantly associated with an eGFR<60, (OR=1.2; 95% CI 1.0-1.5; P=0.04) and eGFR<90 (OR=1.1; 95% CI 1.0-1.2; P<0.01). After adjustment, the CIADIS score remained significantly associated with an eGFR<60 (OR=1.3; 95% CI 1.0-1.6; P=0.03) only. Among 221 patients with an eGFR ≥90 and available uPCr, 9% had an early kidney dysfunction.
Looking at the different immune markers used to construct the CIADIS score separately, patients with an eGFR <60 compared to those with and eGFR >=60 had higher percentages of CD4 T cells expressing CD4+ DR (16% vs. 14%, p=0.03) and lower percentages of CD8+ naïve T cells (30% vs. 39%, p=0.03), respectively.
Patients with an eGFR <90 compared to those with an eGFR >=90 had higher percentages of CD4 T cells expressing CD4+ CD57+CD28- (4% vs. 2%, p<0.01), CD8 T cells expressing CD8+ CD57+CD28- (28% vs. 26%, p<0.01) and lower percentages of CD8+ naïve T cells (37% vs. 40%, p<0.01), respectively (Data not shown).
The distribution of the CIADIS score (combining the above described markers) according to eGFR stages is depicted in Figure 1: the lower the eGFR, the higher the CIADIS score.
Poster Pdf attached
-----------------------------
Association of immune-activation and senescence markers with non-AIDS-defining comorbidities in HIV-suppressed patients
AIDS 2015
Duffau, Pierrea,b,c, ; Wittkop, Lindad,e,f, ; Lazaro, Estibalizb,c,g; le Marec, Fabiend,e; Cognet, Célineh; Blanco, Patrickb,c,h; Moreau, Jean-Françoisb,c,h; Dauchy, Frédéric-Antoined,e,i; Cazanave, Charlesi; Vandenhende, Marie-Anned,e,g; Bonnet, Fabriced,e,g; Thiebaut, Rodolphed,e,f,j; Pellegrin, Isabelled,e,h; for the ANRS CO3 Aquitaine Cohort Study Group
Abstract
Objectives: We studied the link between T-cell activation, differentiation and senescence phenotypes and non-AIDS-related comorbidities in HIV-suppressed patients.
Design: Patients included in the ANRS CO3 Aquitaine Cohort were consecutively enrolled in this cross-sectional study between October 2011 and May 2013 called Chronic Immune Activation and Senescence (CIADIS) study.
Methods: We summarized immune markers [CD4+ and CD8+ activation (DR+), differentiation (naive and terminally differentiated memory T cells), and senescence (CD57+CD28-)] in a weighted immune score by principal component analysis called CIADIS. Previously described Veterans Aging Cohort Study (VACS) index and immune risk profile (IRP) scores were calculated. We used adjusted logistic regression to assess the association between the CIADIS score and the presence of at least three non-AIDS-defining comorbidities.
Results: Of 876 patients with an undetectable viral load, 73.4% were men and median age was 50.5 years [interquartile range (IQR) 44.7-56.7 years]. Median CD4+ T-cell count was 579/μl (IQR 429-759 cells/μl), and median duration of HIV viral suppression was 5.3 years (IQR 2.3-8.7). The weighted CIADIS score was associated with at least three comorbidities (odds ratio 1.3 for 1 SD more, 95% confidence interval 1.0, 1.6) independently of age, sex, AIDS stage, and the Veterans Aging Cohort Study score. The CIADIS and the immune risk profile scores were significantly associated with at least three comorbidities in adjusted models restricted to patients younger than 60 years. None of the tested scores were associated with at least three comorbidities in patients older than 60 years.
Conclusions: The weighted CIADIS score based on activation, senescence, and differentiation markers might help physicians identifying patients at a higher risk for non-AIDS-related comorbidities.
Introduction
The widespread introduction of combination antiretroviral therapy (cART) has improved the virological and immunological outcomes, whereas overall mortality of HIV-infected patients has been dramatically reduced [1,2]. AIDS-related illnesses are no longer the primary causes of death, but a new set of non-AIDS-related comorbidities has emerged including cardiovascular disease, kidney disease, malignancy, as well as neurological and bone disorders, resulting in a novel systemic chronic disease [3,4].
Immune dysfunction, inflammation, and coagulation abnormalities persist and strongly predict risk of non-AIDS morbidity [5,6], even in successfully treated patients with restored peripheral CD4+ T-cell count and complete or near-complete suppression of HIV replication [7-9]. However, reasons for immune activation in long-term HIV-RNA-suppressed patients remain yet to be fully understood [3,10]. Comorbidities referred to as non-AIDS-related comorbidities are typically those associated with ageing in the general population, suggesting that HIV-infected individuals might suffer from accelerated ageing as they are found at a younger age in HIV-infected individuals [11-13]. Cardiovascular risk factors and direct toxic effects of antiretroviral drugs contribute, but do not fully explain, the higher risk for non-AIDS-related comorbidities [14-16]. An emerging body of evidence suggests that persistent innate and adaptative immune dysfunction and/or activation are major risk factors for non-AIDS-related comorbidities [17,18]. Many of the immunological abnormalities that persist during cART are similar to those observed in the elderly, raising the hypothesis that age-related decline in immune function contributes to disease progression and adverse outcomes. For example, markers of T-cell activation [human leukocyte antigen-D related (HLA-DR) and CD38 co-expression] are elevated despite effective cART and associated with subsequent morbidity and mortality, even after adjustment for CD4+ T-cell count [8,19]. The association of immune activation and immunosenescence with non-AIDS-related comorbidities has been widely studied in the past few years [20-23]. However, one of the limits of these studies is that immune markers defining activation, differentiation, and senescence were not studied simultaneously [23]. Moreover, studies were often based on small sample sizes limiting statistical power, favoring selection, and focused on a specific non-AIDS-related comorbidity. In this cross-sectional study, we aimed at studying the link between immune activation, and immunosenescence phenotypes and non-AIDS-related comorbidities in HIV-infected patients with undetectable viral load.
Discussion
In a large assessment of T-cell activation markers in patients under routine care, we found a significant association between the CIADIS score and a high comorbidity score, whereas the VACS index as well as the IRP score were not associated with a high comorbidity score in multivariable analysis.
Our results are in contrast with the recently published data from the study by Tenorio et al.[23] where T-cell activation was not significantly related to the occurrence of non-AIDS-defining events. In this study, all comorbidities usually associated to HIV infection were not studied simultaneously. Immune activation was assessed 1 year after cART initiation and at the time-point immediately preceding the event, which occurred at a median of 2.9 years after initiation. The different findings might be explained by the differences in patients' characteristics as in our study population, the median duration since diagnosis was 17.3 years, with a median duration of viral load suppression of 5.8 years in patients under cART since a median of 14.1 years. In a recent study by Guaraldi et al.[28], duration since diagnosis was a major factor associated with comorbidities. Even after adjustment for duration since HIV diagnosis, we confirmed a statistically signification association between the CIADIS score and the presence of more than two comorbidities (Supplementary Material Table 6, http://links.lww.com/QAD/A751).
Evidence is accumulating that the risk of non-AIDS-related comorbidities is elevated in HIV-positive individuals, they may be occurring at a younger age than what is typically observed among comparable HIV-negative populations [4,29]. Guaraldi et al.[4] demonstrated that the prevalence of polypathology (defined as the concurrent presence of at least two comorbidities) in HIV-infected persons anticipated polypathology prevalence observed in the general population among persons who were 10 years older. These data call for an earlier screening of noninfectious comorbidities in HIV-infected patients. Among a broad array of factors that may be contributing to the emergence of comorbidities at younger ages in the HIV population, HIV-specific cofactors have been already described [30].
However, HIV-related characteristics do not fully explain the occurrence of comorbidities that could also be linked to immune dysfunctions that persist even after long-term viral suppression and CD4+ reconstitution, especially T-cell activation [31,32]. To study immune activation and immunosenescence, we used a PCA approach to evaluate the association of activation (as measured by the percentage of HLA-DR+ T cells among CD4+ and CD8+ cells), senescence (as measured by the percentage of CD57+CD28- cells among CD4+ and CD8+ cells), and differentiation (with a focus on naive and terminally differentiated CD4+ and CD8+ cells). Activation and senescence markers were inversely associated with naive markers. It led us to identify two patterns of patients: a cluster of patients characterized by high expression of activation and senescence markers and low levels of naive T cells, and a cluster of patients exhibiting high levels of naive T cells and low expression of activation and senescence markers. If cellular activation is associated with AIDS-defining outcomes in untreated and treated HIV infection [14,18], the impact of cellular immune activation on comorbidities is less well studied. Most of the studies on this topic examined the relationship between immune activation and morbidity before cART initiation, but do not address whether elevated immune activation after a defined and uniform period of cART-induced viral suppression is associated with outcomes. As HCV has the ability to stimulate the adaptative immune system, it was interesting to look at the role of HCV coinfection in T-cell activation and comorbidities. In our study, we found that HCV coinfection seemed to be protective against T-cell activation and senescence. It perhaps explains why we did not find any correlation between HCV coinfection and a high score of comorbidities. These data suggest that HIV-associated T-cell activation and senescence are involved in comorbidities in suppressed HIV recipients and points to immune activation being involved in comorbidities independently of a unique cardiovascular pathway. T-cell activation in the context of HIV infection is usually assessed by HLA-DR and CD38 coexpression on CD4+ and CD8+ cells, whereas HLA-DR alone is regularly studied as a marker of T-cell activation in non-HIV populations, based upon the fact that CD38 is also expressed by naive T cells. In patients who achieved viral suppression, Cockerham et al.[33] found a strong association between the frequencies of CD4+ or CD8+ cells expressing HLA-DR and HIV-DNA levels in resting CD4+ cells. Similar correlations were also observed when activation was defined by the percentage of CD4+ or CD8+ cells coexpressing CD38 and HLA-DR; however, the strength of these heavily relied on HLA-DR expression by T cells, as no correlation was observed with CD38 alone [33]. We therefore studied association between comorbidities and T-cell activation assessed by coexpression of HLA-DR and CD38, and found the same pattern of data we found with HLA-DR expression (data not shown).
As occurrence of comorbidities at younger ages has been part of the definition of accelerated and premature aging of HIV patients, we examined the putative association between a high comorbidity score and the CIADIS score separately in patients younger than 60 and those older than 60 years of age. We found that the CIADIS score, IRP score and VACS index were statistically associated with at least three comorbidities in younger patients. However, the association between the CIADIS score (and IRP) and the comorbidity score may be biased, especially in the stratum over 60 years, and underestimated due to a higher probability for older patients having died before the inclusion period in the CIADIS study.
The development and validation of a multivariable risk index for people with HIV has been a priority. The VACS has raised the challenge with the rollout of the points-based VACS index [26]. Because the impact of HIV infection on the immune system mirrors that of aging in many respects, IRP defined as a cluster of immunological factors that are strongly predictive of earlier mortality among a non-HIV elderly cohort should be interesting to study in the context of HIV infection. The CIADIS score was significantly associated with a high comorbidity score, especially in younger patients. However, the added discriminant power was little and an external validation of our findings should be done using a completely independent sample. Furthermore, the CIADIS cross-sectional study cannot show that immune activation, as well as immunosenescence, is involved in the occurrence of non-AIDS-related comorbidities. Other markers such as soluble inflammation markers may also be relevant for predicting non-HIV-related comorbidities. Thus, future analyses using longitudinal data from patients included in the CIADIS study are planned.
In conclusion, the weighted CIADIS score based on activation, senescence, and maturation markers might be a tool to help physicians identify patients at a higher risk for non-AIDS-related comorbidities in the future.
Results
Of the 3500 patients routinely followed in the ANRS CO3 Aquitaine Cohort, 1010 patients were consecutively included in the CIADIS study, of whom 876 (86.7%) patients had an undetectable viral load and were analyzed. Their baseline characteristics are summarized in Table 1. Among them, 73.4% were men with a median age of 50.5 years (IQR 44.7-56.7 years). Median CD4+ T-cell count was 579 cells/μl (IQR 429-759 cells/μl), median CD4+/CD8+ ratio was 0.9 (IQR 0.6-1.2), 61.4% of patients had a CD4+/CD8+ ratio below 1, 99.4% of patients were under cART, and median duration of HIV viral suppression was of 5.3 years (IQR 2.3-8.7). HCV coinfection was noted in 23.5% of patients and 89.4% of patients had a positive CMV serology.
Median percentages of naive T cells among CD4+ and CD8+ T cells were, respectively, 40% (IQR 28-51) and 38% (IQR 29-48), whereas terminally differentiated memory T cells represented 26% (IQR 16-36) among CD8+ T cells and only 1% (IQR 0-3) among CD4+ T cells. CD8+ T cells were more activated and senescent compared to CD4+ T cells (P < 0.0001) (Supplementary Material Table 1a, http://links.lww.com/QAD/A751). Activated CD4+ and CD8+ T cells expressing HLA-DR+ were positively correlated (Spearman's rank correlation coefficient r = 0.63, P < 0.0001) and were also positively correlated to T CD57+CD28- senescent T cells, as well as TEMRA subsets (P < 0.0001) (Supplementary Material Table 1b, http://links.lww.com/QAD/A751).
The first two principal components accounted for 68.7% of the variability among immunological markers (Fig. 1a). The first principal component, accounting for 51% of the marker variability, separated naive T cells (CD45RA+CD27+) from activated T cells (HLA-DR+ on CD4+ and CD8+ grouped together), as well as senescent T cells (CD4+ and CD8+ CD57+CD28- and CD4+ and CD8+ CD27-CD45RA+ terminally differentiated T cells grouped together). The second principal component, accounting for 17% of the variability, opposed CD4+ and CD8+ T cells expressing HLA-DR to senescent CD4+ and CD8+ T cells (CD4+ and CD8+ CD57+CD28− and CD4+ and CD8+ CD27−CD45RA+ terminally differentiated T cells grouped together).
Comorbidities were also projected onto the plan spanned by the first two principal components (Fig. 1b) that needs to be read in conjunction with Fig. 1a. Patients with comorbidities are found on the right side and thus were more often found in patients presenting an immune phenotype with high T-cell activation, high expression of terminal differentiation, and senescence markers.s
Quantitative variables related to HIV (CD4+ T-cell nadir, last CD4+ T-cell count, CD4+ T-cell count at treatment start, duration of CD4+ T cell <350 cells/μl, duration of infection, duration of cART) were mostly located at the center of the circle and were not associated with activation or senescence phenotypes (Fig. 2a). The projection of qualitative HIV and non-HIV-related variables onto the plan spanned by the first two components are represented on Fig. 2b that needs to be read in conjunction with Fig. 2a. These variables were not associated with a particular T-cell phenotype in most of the cases. HCV coinfection, plotted on the left bottom corner, was not associated with either T-cell activation or senescence.
The distribution of comorbidities is represented in Table 1. Dyslipidemia and hypertension were the most frequent comorbidities accounting for 36.2 and 35.4% of patients, respectively. Severe kidney function impairment (eGFR <60 ml/min) and CNS degenerative disorders only concerned, respectively, 3.5 and 2.9% of patients. Also, 37.4% of patients had no comorbidity and 13.6% of patients suffered from at least three comorbidities.
In univariable analysis, we found that an older age, sex (male), and AIDS stage were associated with at least three comorbidities (respectively, P < 0.0001, P = 0.0014, P < 0.0001). Last CD4+ T-cell count or HCV coinfection was not associated with a high number of comorbidities. Patients with a higher CD4+ nadir had a lower risk for presenting at least three comorbidities (Supplementary Material Table 2, http://links.lww.com/QAD/A751). Higher CD4+ and CD8+ T-cell activation were significantly associated with the presence of at least three comorbidities (respectively, P = 0.0004, P = 0.0001). Higher percentage of senescent and terminally differentiated T cells, and lower percentage of naïve T cells were also significantly correlated with a high comorbidity score (Supplementary Material Table 3, http://links.lww.com/QAD/A751).
Variables used for the calculation of the VACS index and the IRP score are summarized in Supplementary Material Tables 4 and 5 (http://links.lww.com/QAD/A751), respectively. Among the patients included in the CIADIS study, 14.2% had a VACS index at least 35 and 50.3% had a high IRP (≥3). A high VACS index as well as a high IRP score were significantly associated with a high number of comorbidities in our study (respectively, P < 0.0001, P = 0.0006). Furthermore, the CIADIS score was also significantly associated with the presence of at least three comorbidities (P < 0.0001) (Supplementary Material Table 3, http://links.lww.com/QAD/A751).
In multivariable analysis (Table 2), after adjustment for age, sex, and AIDS stage, a higher CIADIS score was significantly associated with the presence of at least three comorbidities [model 1: odds ratio (OR) 1.3, 95% confidence interval (CI) 1.0-1.7, P = 0.02] and was also significantly associated after additional adjustment with the VACS index (model 2: OR 1.3, 95% CI 1.0-1.6, P = 0.02). However, the discriminant power measured by AUROC increased little for the model including the CIADIS score (AUROC = 0.787) compared to a model without the CIADIS score (AUROC = 0.782), respectively.
Neither the VACS index (model 5: OR 1.2, 95% CI 0.7-2.1, P = 0.59) nor the IRP score (model 3: OR 1.5, 95% CI 0.9-2.5, P = 0.09) was statistically associated with the presence of at least three comorbidities. An older age as well as B and C AIDS stages were associated with a high comorbidity score, irrespective of the adjustment model. As comorbidities might occur in younger ages in HIV patients compared to the general population, we did separate analysis in younger patients (<60 years old) versus older patients (≥60 years old) (Table 3). We found that the CIADIS score as well as IRP were significantly associated with a high comorbidity score in younger patients (OR 1.4, 95% CI 1.1-1.9, P = 0.02; and OR 1.8, 95% CI 1.0-4.4, P = 0.047, respectively). The VACS index was significantly associated in the model including the IRP and CIADIS scores only (Table 3). We did not find any significant association between the CIADIS score, the VACS index, or the IRP score with a high comorbidity score in patients older than 60 years. However, there was no statistically significant interaction between age and the different scores (P = 0.57, P = 0.63; P = 0.21, respectively). In sensitivity analyses including patients with detectable viral load (n = 1010), we found similar results regarding the association between the CIADIS score and comorbidities (data not shown).
|
|
| |
| |
|
|
|
