Serious Non-AIDS events: Immunopathogenesis and interventional strategies

Uncontrolled HIV replication causes immune activation and progressive decline in CD4 T cell count [29]. In addition, HIV can also contribute directly to organ dysfunction and SNAEs. HIV can infect human hepatic stellate cells and induce collagen expression and pro-inflammatory cytokines secretion in vitro[30, 31]. HIV can also mediate dysregulation of glomerular podocytes in HIV-transgenic mouse models [32], as well as apoptosis of human renal tubular cell lines [33]. Therefore, direct effect of HIV may contribute to decline in renal function and increase risk of chronic kidney disease [34]. HIV may contribute directly to non-AIDS malignancy [35], as it can be oncogenic by activating proto-oncogenes [36] or by blocking tumour suppressor genes [37] in cell lines. Increased rates of microsatellite instability in tumours of HIV-infected patients have also been found [38].

Lower nadir or pre-ART CD4 T cell count is associated with increased risk of SNAEs [11, 14, 16, 39]. In addition, the degree of CD4 T cell recovery after ART initiation also influences the incidence of SNAEs [14, 18, 21, 39]. A 100 cells/μL lower latest CD4 T cell count in ART-treated patients is associated with a 30% increase in the risk of SNAEs, even after adjusting for smoking status, diabetes mellitus, hyperlipidaemia, HCV and HBV co-infection and alcohol abuse [18]. Lower latest CD4 T cell count in ART-treated patients is also associated with an increase in mortality [40, 41].

Suboptimal restoration of CD4 T cells may be secondary to a number of factors including decreased thymic function [42, 43] and impaired homeostatic responses and survival of T cells [44]. HIV replication and immune activation stimulate the secretion of transforming growth factor (TGF-β), mainly by regulatory T cells, with macrophages also contributing. TGF-β triggers collagen production by fibroblasts [45‐47]. The resultant structural damage and fibrosis of the lymphoid tissues restricts T cell access to interleukin-7 (IL-7) on the fibroblastic reticular cell (FRC) network [45] thus limiting naïve CD4 T cell survival [48, 49]. Furthermore, ongoing immune activation leads to rapid CD4 T cell turnover, overwhelming the already impaired renewal mechanisms [44], resulting in suboptimal CD4 recovery [50, 51].

Due to common routes of transmission, HIV-infected patients also have increased risk of exposure to other infections that can cause organ dysfunction.

About 6-14% of HIV-infected patients have HBV and about 25-30% have HCV infection, with the rates varying depending on endemicity of HBV in the population and the prevalence of injecting drug use [52]. HBV and/or HCV co-infection greatly increases the risk of SNAEs despite ART [9, 11, 14, 18, 39, 53]. Co-infection with HCV is associated with increased risk of renal disease (1.5-fold) [54, 55], cardiovascular disease (1.5-fold) [56], cirrhosis (19-fold) and hepatocellular carcinoma (5-fold) [57] when compared with HIV mono-infected patients in the ART era.

HIV-infected patients also have increased risk of exposure to Human papillomavirus (HPV) [58], that is implicated in the pathogenesis of cervical and anal cancer [59, 60].

HIV-infected patients also have higher prevalence of traditional cardiovascular risk factors such as smoking [61‐63], elevated total cholesterol/HDL ratio [61‐64] and substance abuse [65]. Some studies also found higher rates of hypertension and diabetes [64, 66], but these findings have not been confirmed by others [62, 63, 67].

Smoking is a major cause of increased mortality in ART-treated patients, accounting for a loss of >12 life years, and is associated with >4-fold increase in all-cause mortality, >5-fold increase in non-AIDS mortality, >4-fold increase in cardiovascular-related mortality and >3-fold increase in cancer-related mortality [68].

Though uncontrolled HIV replication, immunodeficiency, co-infection and co-morbid conditions are important contributors to the pathogenesis of SNAEs, these factors only partially account for the increased risk of SNAEs in HIV-infected patients. HIV elite controllers have increased coronary plaques and carotid intima media thickness when compared to uninfected controls even after adjustment for traditional cardiovascular risk factors [69, 70], demonstrating that factors other than uncontrolled viral replication, immunodeficiency, traditional risk factors, and ART toxicities contribute to increased cardiovascular risks in HIV-infected patients.

Inflammation is central to the process of atherosclerosis [71, 72], tumour progression [73, 74] and liver fibrosis [75, 76] in the general population. High levels of biomarkers associated with inflammation (C-reactive protein, CRP, and IL-6) and coagulopathy (D-dimer and fibrinogen) are associated with increased risk of cardiovascular disease [77‐80], increased cancer risks [81‐83] and mortality [80, 84‐86] in the general population.

In HIV-infected patients, CRP, IL-6 and D-dimer levels [61] as well as markers of T cell activation [50, 87] remain higher than uninfected controls despite suppressive ART. Furthermore, higher CRP, IL-6 and D-dimer [88‐90], soluble CD14 (sCD14, a marker of LPS-induced monocyte activation) [91] and lymphocyte activation markers [92] are associated with higher mortality.

In ART-treated patients, CRP, IL-6 and D-dimer levels are also associated with increased risk of CVD, independent of other CVD risk factors [93, 94]. These markers are also associated with increased risks of both infection-related and infection unrelated cancers even after adjusting for demographics and CD4 T cell count [95]. Recently, higher levels of tumour necrosis factor (TNF) were also found to be significantly associated with increased risk of SNAEs [96]. Therefore these data suggest that immune activation plays a very important role in SNAEs pathogenesis.

Finally, adverse effects of ART may also contribute to SNAEs. A detail review of ART-related toxicities is beyond the scope of this review. Protease inhibitors as a class, as well as indinavir, lopinavir and abacavir may be associated with increased risk of CVD [25, 114, 115].

Some studies found that the rate of liver-related deaths is increased per year of ART [116, 117]. The cohort described by Weber et al have high rates of HBV and HCV infection, 17% and 66%, respectively [116]. Though patients with HBV or HCV infection have about 5-fold greater risk of hepatotoxicity after ART initiation [118‐120], ART is essential as it slows fibrosis progression [121] and reduces liver-related mortality by about 10-fold [122]. In patients without chronic viral hepatitis, ART toxicity rarely causes liver-related mortality, at a rate of 0.04 per 1000 person years [123].

ART has been associated with improved renal function and a slower rate of eGFR decline in HIV-infected patients [124, 125]. Though tenofovir use is associated with higher risk of acute renal injury and greater loss of renal function than other ART regimen, the overall risk of serious renal event is not high, in 0.5% of patients [126].

In the D.A.D study, patients who stopped smoking had about 30% reduction in the risk of CVD [145]. Surprisingly, a reduction in mortality was not seen. This may be because patients ceased smoking after the diagnosis of a serious illness and succumb before the benefit of smoking cessation on mortality can be seen [145]. In the Danish HIV cohort, previous smokers had a >1.5-fold reduction in mortality when compared with current smokers. In addition, though previous smokers have higher rates of AIDS-related deaths when compared with never smokers, the incidence of non-AIDS-related death was not different between previous and never smokers [68]. These data suggest that smoking cessation alone would result in significant benefits and should be encouraged.

Modification of other cardiovascular risk factors e.g. treatment of hypertension, dyslipidaemia and optimal glycaemic control in diabetic patients is critical. Suggested target levels have been published [146, 147]. Each 10 mmHg reduction in systolic blood pressure and each 38 mg/dL reduction in total cholesterol is associated with a 5 and 20% reduction in risk of CVD, respectively [148]. However, in patients with known hypertension, diabetes or dyslipidaemia meeting indication for treatment, over 40% were not on treatment [149]. Given that a significant proportion of SNAEs are cardiovascular events, more aggressive detection and management of cardiovascular risk factors will likely reduce SNAEs.

ART modification is a potential strategy to reduce cardiovascular risk [150]. A recent review of switch studies have been published [151]. Switching from stavudine to tenofovir was associated with a reduction in total cholesterol and triglycerides and an increase in limb fat [152]. Switching from protease inhibitors to efavirenz or nevirapine was associated with reduction in total cholesterol [153]. However, this switch is not possible in patients with non-nucleoside reverse transcriptase inhibitor resistance. Switching from lopinavir/ritonavir to atazanavir (both boosted with ritonavir 100 mg or unboosted) was associated with reduction of total cholesterol and triglycerides, though greater reductions were seen with unboosted atazanavir [154‐156]. Switching to atazanavir was also associated with a reduction in cardiovascular risk score [157]. Though switching from lopinavir/ritonavir to raltegravir was associated with improvement in lipid profile [158, 159], no change in endothelial function was detected [160]. The importance of having fully active backbone antiretrovirals was highlighted in the SWITCHMRK study where patients switched to raltegravir had higher rates of virologic failure [158].

Statins are 3-hydroxy-3-methyl-glutaryl-coenzyme A reductase inhibitors. Not only do statins reduce serum cholesterol [180], they may also have anti-inflammatory properties [181]. Statin use is associated with reduced monocyte activation (unpublished data McComsey et al), decline in CRP levels [182] as well as reduced T cell activation [183] in ART-treated and in untreated, HIV-infected patients [184]. A retrospective observational study of ART-treated patients showed that statin use is associated with a 3-fold reduction in mortality [185]. Though not statistically significant, a trend for reduction in SNAEs [186] and mortality [187] has also been seen in other retrospective observational studies.

Hydroxychloroquine (HCQ) and its analogue chloroquine (CQ) have immunomodulatory, anti-inflammatory and anti-HIV properties [188‐190]. In patients with uncontrolled viral replication, the use of CQ was associated with reduced CD8 T cell activation [191] whereas the same effect was not seen with HCQ [192]. However a non-randomized study of HCQ in 20 ART-treated patients showed decline in plasma LPS, IL-6 and reduced T cell and monocyte activation [193]. Thus findings are inconclusive.

COX-2 inhibitors inhibit cyclooxygenase type 2, reducing Prostaglandin E2 production, thereby reducing activation of T cells through the cyclic adenosine monophosphate (cAMP) pathway [194]. Studies on COX-2 inhibitors have been small and reduction in T cell activation tended to occur in viraemic patients [195, 196]. However, it is important to bear in mind that COX-2 inhibitors are associated with increased cardiovascular risk, via a direct pharmacologic consequence of inhibition of COX-2 [197]. Therefore assessing the utility of COX-2 inhibitors without using clinical outcome measures may be insufficient.

Aspirin is a cornerstone in the secondary prevention of vascular disease [198]. In a pilot study, aspirin use was associated with reduced platelet activation, a decrease in sCD14 in monocytes and reductions in CD38 and HLA-DR on CD4 and CD8 T cells. However, there was no change in IL-6, D-dimer and CRP [199]. An aspirin study with larger number of participants is in development with the AIDS Clinical Trials Group.

Leflunomide is an immunomodulatory agent that is used in the treatment of rheumatoid arthritis. The administration of leflunomide in untreated, HIV-infected patients for 28 days was associated with a decrease in CD8 T cell activation [200].

The use of prednisone in patients with untreated chronic HIV infection was associated with less CD4 T cell depletion, a decline in CD4 T cell activation and stable HIV viral load [201, 202]. Prednisone at 0.5 mg/kg/day in ART-treated patients was also associated with a reduction in CD8 T cell activation and TNF levels as well as a transient decrease in IL-6 [203]. However, in another study using prednisone at 40 mg/day, no reduction in CD4 or CD8 T cell activation, plasma IL-6 or TNF levels was found [204]. Furthermore, long-term prednisone use, especially at doses >7.5 mg/day is associated with significant adverse effects such as osteoporosis, impaired glucose tolerance, dyslipidaemia, weight gain, cataract formation and increased risk of infections [205]. Even short courses have been associated with an increased risk of osteonecrosis in HIV-infected patients [206].

Given that HIV infection has been associated with depressed levels of beneficial gut microbiota and elevated levels of pathogenic microbiota [207], a range of prebiotics (selectively fermented ingredients that changes the growth and/or activity of certain gut microflora, resulting in health benefits [208]), probiotics (live microorganisms that when consumed, confer a health benefit [209]) and synbiotics (combinations of pre and probiotics) are under investigation.

A prebiotic oligosaccharide mixture has been associated with improvement in microbiota composition and reduction in sCD14 in untreated HIV-infected patients [210]. A retrospective cohort study on both ART-treated and untreated HIV-infected patients found that probiotic yogurt consumption was associated with a greater increase in CD4 T cell count even after adjustment for ART [211]. A double-blind randomized placebo-controlled trial in 20 untreated HIV-infected patients found reductions in plasma bacterial DNA and IL-6 levels in patients receiving synbiotics [212]. However, a synbiotic agent in ART-treated women found no change in microbial translocation nor immune activation status despite improvement in the levels of probiotic species [213]. Therefore, more randomized controlled clinical trial data are needed to clarify the effects of pre and probiotics in reducing immune activation.

Bovine colostrum contains oligosaccharides, growth factors, immunoglobulins and antimicrobial peptides and has some activity in alleviating HIV-associated diarrhoea in single arm studies [214‐216]. However a randomized controlled trial on the addition of bovine colostrum to suppressive ART found no change in CD4 T cell count, markers of microbial translocation nor T cell activation [167].

A number of new agents that target microbial translocation are under evaluation. Rifaximin is a minimally absorbed oral rifamycin antibiotic that has activity against both gram-positive and gram-negative enteric bacteria [217]. It is effective in the treatment of hepatic encephalopathy, by reducing ammonia-producing enteric bacteria [218, 219]. There are currently 3 clinical trials of rifaximin in HIV-infected patients (ClinicalTrials.gov identifier: NCT01654939, NCT01866826 and NCT01466595). Lubiprostone is a chloride channel activator that is used in the treatment of constipation [220]. It has been found to enhance recovery of mucosal barrier function in ischaemic porcine colon [221]. A pilot study of lubiprostone in ART treated, virologically suppressed patients with CD4 T cell count <350 cells/μL (NCT01839734) is currently recruiting. Sevelamer is a phosphate binder that is used in patients with end-stage renal failure [222]. It can also bind to endotoxins and reduce CRP, IL-6 and sCD14 in patients on haemodialysis [223, 224]. The trial in untreated, HIV-infected patients (NCT01543958) was completed in June 2013 and results are pending. Mesalamine (5-aminosalicylic acid) is an anti-inflammatory agent used in the management of inflammatory bowel disease [225]. A trial using mesalamine in ART treated, virologically suppressed patients with CD4 T cell count <350 cells/μL (NCT01090102) is currently enrolling.