| |
A gathering storm: HIV infection and nonalcoholic fatty liver
disease in low and middle-income countries
|
| |
| |
Download the PDF here
• HIV-associated risk factors for NAFLD are related to ART and lipodystrophy
• Reported prevalence of NAFLD and NASH in HIV ranged from 26-65 and 31-72%, respectively.....Approximately 25-30% of HIV-negative individuals with NAFLD develop NASH [8]. In HIV-positive individuals, however, the severity both at initial presentation and that of progression is higher [30]. Although there are limited data available on histopathological correlation, in a cohort of HIV-positive individuals (91% on ART) with biopsy-proven NAFLD, the prevalence of NASH was 65% and the majority had fibrosis [30].
• they looked at 14 studies- Only two studies reported statistically significant associations between HIV-related factors and NAFLD or NASH after adjustment, which included duration of HIV infection [30] and longer exposure to NRTIs [86]. All other significant associations were non-HIV related, including traditional risk factors that have been identified in the general population: waist circumference [86,87], age [88], lower high-density lipoprotein (HDL) [87], higher triglycerides [87,89], alanine aminotransferase (ALT)/aspartate aminotransferase (AST) ratio and/or ALT [86,90,91], homeostatic model for assessment for insulin resistance [24,88,92,93], higher fasting glucose and insulin [24], male sex [86,88], black ethnicity [46], lower albumin [46], BMI [90,91], visceral adipose tissue (VAT) [92], dyslipidaemia [90], higher APRI and fibrosis-4 scores [30] and γ: glutamyl transferase (GGT) [88,93].
• A recent systematic review and meta-analysis examined 10 of these studies and reported a prevalence of NAFLD and NASH in the setting of HIV of 35% (95% CI 29-42) and 42% (95% CI 22-64), respectively [31]. Although the key risk for factors for NAFLD were related to metS (BMI, waist circumference, T2DM, hypertension and triglycerides), they did report an association between NAFLD and higher CD4+ cell count [mean difference of 54.83 (95% CI 11.55-98.11, P = 0.0)], based on data from four studies.
• Although NAFLD is more prevalent in HIV than in the general population, studies in HIV monoinfection cohorts to date have not consistently found statistically significant associations with HIV-related factors. This may likely be due to the limited numbers of studies that have investigated NAFLD in HIV and excluded viral hepatitis/excess alcohol, and the diversity in the cohorts. There is an obvious need for prospective longitudinal studies sufficiently powered to robust outcomes.
• There are no treatments for NAFLD or NASH that are specific to people living with HIV. Currently, lifestyle modification and dietary changes are the main interventions for NAFLD, in conjunction with the specific treatment of associated metabolic disorders......Weight loss of about 3-5% of body weight is recommended to improve steatosis and about 10% to improve histological features of NASH, including portal inflammation and fibrosis.
• There are several new agents in development in the drug pipeline. Of interest in HIV-NAFLD is cenicriviroc (CVC), a chemokine receptor type 5 (CCR5) and chemokine receptor type 2 (CCR2) antagonist that has anti-HIV activity both in vitro and in vivo[101,102]. CCR2 plays a critical role in recruiting monocytes with an inflammatory phenotype into the liver, while CCR5 attracts lymphocytes into the liver and is directly involved in fibrogenic activation of hepatic stellate cells [103]. In addition, CVC has been shown to reduce circulating markers of monocyte inflammation in HIV-infected individuals on ART, including soluble CD14 [102]. In a large randomized phase 2b study of CVC in HIV-negative individuals with NASH (CENTAUR), CVC resulted in a reduction in multiple markers of inflammation, including IL1b, high-sensitivity CCRP, IL6 and fibrinogen
• Management and monitoring of NAFLD in HIV-infected individuals, particularly in LMIC where there is increased visceral fat in the healthy weight range, increasing levels of obesity and prediabetes combined with large numbers of people living with HIV, could be a very significant health burden that requires management and prevention strategies.
--------------------------------------
Kapoor, Nitina,b,*; Audsley, Jenniferc,*; Rupali, Priscillad; Sasadeusz, Joee; Paul, Thomas V.a; Thomas, Nihala; Lewin, Sharon R.c,f
AIDS: June 1 2019
ABSTRACT
Despite the decreasing total incidence of liver-related deaths, liver disease remains one of the major non-AIDS causes of morbidity and mortality amongst people living with HIV, and a significant proportion of liver disease in these individuals can be attributed to nonalcoholic fatty liver disease (NAFLD). NAFLD in HIV infection is a growing problem in view of increasing life expectancy associated with the use of effective antiretroviral therapy (ART), wider uptake of ART and increasing rates of obesity in many Asian as well as western countries. The problem may be more pronounced in developing countries where there are limited resources available for mass screening and diagnosis of NAFLD. There is a small but growing body of literature examining NAFLD in the setting of HIV, with data from low and middle-income countries (LMICs) particularly lacking. Here, we review the cohort data on NAFLD in HIV, and discuss the risk factors, pathogenesis of hepatic steatosis, NAFLD and nonalcoholic steatohepatitis (NASH), diagnostic approaches and therapeutic options available for NAFLD in the setting of HIV, and the specific challenges of NAFLD in HIV for LMICs.
Conclusion
Although there is a small but growing body of literature on HIV-associated NAFLD, there are still large gaps regarding the optimal guidelines for NAFLD screening and management in the setting of HIV, particularly in LMIC. Guidelines in those who are HIV-infected may need to differ from the HIV-negative population, as NAFLD prevalence in the setting of HIV appears higher, progression to NASH more frequent and a role for HIV-related factors has not been fully elucidated. Importantly, there are no data to date prospectively examining NAFLD progression in HIV-positive individuals and very little published from LMIC and in HIV-infected women. Further research is also required to develop noninvasive markers for diagnosis and monitoring of NAFLD and its progression to NASH to identify those at highest risk for advanced liver disease. An increased role of genetic testing for predicting progression and new therapeutic interventions could potentially play a role in the future.
Introduction
Since the introduction of widespread effective HIV antiretroviral therapy (ART), there has been a reduction in the rates of all-cause death, AIDS-related deaths and most non-AIDS related causes of death in people who are HIV-positive [1]. The exception to this trend has been non-AIDS defining cancers in people with viral suppression [1]. Despite decreasing total incidence of liver-related deaths, liver disease remains one of the major non-AIDS causes of morbidity and mortality amongst people living with HIV [1-4]. A significant proportion of the liver disease in these individuals can be attributed to nonalcoholic fatty liver disease (NAFLD), which has been defined as hepatic steatosis without significant alcohol intake (less than 30 g/day for men and 20 g/day for women) or coinfection with such as hepatitis B (HBV) and hepatitis C (HCV) [5-7].
The disease spectrum of NAFLD (Fig. 1) [8-12] can range from mild steatosis to an inflammatory state known as nonalcoholic steatohepatitis (NASH), which may progress to cirrhosis, and hepatocellular carcinoma (HCC). NAFLD in HIV infection is a growing problem in view of increasing life expectancy associated with the use of effective ART, wider uptake of ART and increasing rates of obesity in the general population of many Asian as well as western countries. Recent WHO global estimates indicate that many low and middle-income countries (LMICs) are experiencing a rapid surge in obesity and overweight, particularly in urban settings [13]. Although the standard economic term LMIC includes countries with a wide range of ethnic, dietary and cultural differences, cross-sectional data have identified significant proportions of obesity in HIV-positive populations in a range of African countries, India and China [14-16]. Increasing rates of obesity have also been observed in people living with HIV in the United States, Nigeria and the Democratic Republic of Congo [17-19]. The problem may be more pronounced in developing countries where there are limited resources available for mass screening and diagnosis of NAFLD [20]. There is a small but growing body of literature examining NAFLD in the setting of HIV, with data from LMIC particularly lacking.
Pathophysiology: risk factors and natural history
The risk factors that predispose individuals with HIV to NAFLD include traditional risk factors that are also observed in the general population and others, which are more specific to HIV infection. Traditional risk factors include most components of metabolic syndrome (metS) such as obesity, Type 2 diabetes mellitus (T2DM) associated with underlying insulin resistance due to reduced physical activity and overnutrition [21-23], dyslipidaemia and hypertension. When compared with HIV-negative individuals, the risk of having metS is almost double in those with HIV and a prevalence of 15-60% has been reported in some studies. This increased incidence has been observed in both ART-naive and HIV-infected individuals on ART [24,25].
Certain drugs, other than ART, are also responsible for the development of steatohepatitis. Commonly used medications such as amiodarone, methotrexate, tamoxifen, sodium valproate and glucocorticoids are associated with developing fatty liver; the exact pathophysiologic mechanisms, which are multifactorial, are not fully elucidated [26]. It is postulated that not only do these drugs predispose to metS and thereby increase the chances of developing fatty liver, but may also escalate hepatocyte lipogenesis, reduce the secretion of fatty acids and interrupt mitochondrial beta oxidation [26]. The concurrent use of these medications with ART may result in an added risk of developing NAFLD.
HIV-associated risk factors for NAFLD are related to ART and lipodystrophy.
Antiretrovirals that precipitate an adverse lipid profile include the nucleotide reverse transcriptase inhibitors (NRTIs); these may indirectly increase the risk of NAFLD via lipid profile alterations. Didanosine (ddI) and stavudine (d4T) are the most commonly implicated NRTI associated with NAFLD; however, their use has been phased out in nearly all treatment guidelines and national programmes [27]. Switching to NRTI-sparing regimens has been associated with an increase in subcutaneous fat and there is some evidence that those on efavirenz (EFV)-based regimens had higher loss of limb fat compared with those on protease inhibitor containing regimens [28]. ART-associated lipodystrophy has primarily occurred with the antiretrovirals ddI, d4T, zidovudine (ZDV) and to a lesser extent EFV [28]; is associated with insulin resistance and dyslipidaemia and is also a major factor driving NAFLD in HIV-infected individuals. Trunk to lower limb fat ratio, as measured using dual-energy X-ray absorptiometry (DEXA), can be used as a surrogate measure of lipodystrophy. In a South Asian HIV cohort receiving ART, a trunk to lower limb fat ratio cut-off more than 2.28 identified lipodystrophy and had the highest odds ratio (OR) for predicting metS, an additional cardiovascular risk [29].
HIV-associated NAFLD seems to be more severe in clinical and biochemical severity.
Compared with HIV-uninfected individuals, in the setting of HIV, there is significantly higher liver enzyme and serum triglyceride levels and higher rates of steatohepatitis with more features of liver injury, such as the presence of acidophil bodies and lobular inflammation [30]. Risk factors associated with progression of fibrosis included advancing age, a higher HIV viral load, increased liver enzymes and the use of ddI/d4T [30]. As ddI/d4T are now rarely used, the role of contemporary NRTIs merits further investigation and there is currently insufficient published data for metanalysis of exposure by drug class [31]. Meta-analyses have shown that in the general NAFLD population, those with fibrosis have an increased risk for all-cause mortality and that risk increases with stage of fibrosis [32]. HIV can infect hepatic stellate cells, hepatocytes and Kupffer cells [33,34], and given that significant fibrosis has been observed in treated HIV monoinfection [33,35,36], prospective fibrosis assessment is very important in HIV-NAFLD. In HIV monoinfection, duration of ddI and zidovudine has been associated with cirrhosis, undetectable HIV (<40 copies/ml) with the absence of significant of fibrosis and duration of boosted protease inhibitors had a negative correlation with abnormal liver-stiffness values [35,36].
The pathogenesis of hepatic steatosis, nonalcoholic fatty liver disease and nonalcoholic steatohepatitis
Insulin resistance mediates the increased flux of free fatty acids to hepatocytes from adipose tissue. This induces oxidative stress at the level of the endoplasmic reticulum (ER), promoting steatosis and inflammation. This, coupled with cellular dysfunction, is now referred to as a ‘multiple parallel hit model’ to explain the development of NASH, which precedes NAFLD [37]. This cellular dysfunction is impacted by the adipocytokines (a range of biologically active mediators generated in adipose tissue) and gut microflora [38]. In obesity, adipocytokines such as interleukin (IL) 6 and tumour necrosis factor-alpha (TNF-α) are elevated, but paradoxically others such as adiponectin are reduced. These alterations have been linked to the development of fatty liver in HIV-negative populations [39]. Recent data have shown that increased adiponectin level is associated with a lower risk of fatty liver regardless of HIV status [39].
The relationship between NAFLD, insulin resistance and visceral fat is complex and not fully understood; however, pro-inflammatory cytokines such as IL-6, C-reactive protein (CRP) and pro-fibrogenics such as fibrinogen may play an important role in this complex interplay [9].
In the setting of HIV infection, increased levels of adiponectin have been associated at the univariate level with lower levels of liver fibrosis, but not when adjusted for MetS [40]. NASH is defined by steatosis, inflammation and hepatocyte damage, and increased bacterial translocation associated with altered gut permeability has been proposed as a trigger for liver inflammation [41]. As HIV infection leads to altered gut permeability and increased bacterial translocation, together with mitochondrial toxicity secondary to NRTIs, this may further add to inflammation in the setting of HIV.
Natural history of nonalcoholic fatty liver disease and nonalcoholic steatohepatitis in HIV infection
The early stages of NAFLD usually begin with hepatic steatosis, which is largely benign; however, this may be a forerunner in some individuals to the development of NASH. NASH, when coupled with fibrosis, may in turn lead to more serious consequences such as cirrhosis and, in some, even HCC which can occur in the absence of cirrhosis [21]. Approximately 25-30% of HIV-negative individuals with NAFLD develop NASH [8]. In HIV-positive individuals, however, the severity both at initial presentation and that of progression is higher [30]. Although there are limited data available on histopathological correlation, in a cohort of HIV-positive individuals (91% on ART) with biopsy-proven NAFLD, the prevalence of NASH was 65% and the majority had fibrosis [30].
Coinfection with HCV genotype 3 is also associated with insulin resistance and a higher prevalence of hepatic steatosis, and it is unclear how effectively this regresses after achieving a sustained virological response [42]. However, with effective treatment now available for HCV, the risk of ongoing inflammation and contribution to development of NAFLD from HCV may reduce over time, although this is dependent on access to directly acting antivirals for HCV which is variable in LMIC [43,44]. In addition, fibrosis associated with HCV does not always resolve, with up to approximately 10% of patients who achieve a sustained virological response after treatment showing persistent or progressive fibrosis [45].
Diagnostic modalities
The diagnostic criteria and therapeutic approach to NAFLD in HIV-positive individuals is the same as for HIV-negative individuals. A range of biochemical and imaging tools have been used both to diagnose NAFLD and to determine the severity, especially in those HIV-positive individuals with metS, lipodystrophy and/or abnormal liver function tests (LFTs). Of the currently available biomarkers/panels, there is no single ideal biomarker that can be used to diagnose NAFLD; however, there are certain noninvasive models utilizing routine clinical parameters that have been proposed to assist diagnosis of steatosis. These are summarised in Table 1 (serum biomarkers) [46-58] and Table 2 (diagnostic/imaging modalities) [59-84].
Current noninvasive diagnostic and imaging techniques have limited success in identifying those individuals with NAFLD who are likely to progress to NASH. The development and validation of reliable, cost-effective and widely available noninvasive biomarkers and imaging techniques for use in clinical care are essential to screen and closely follow individuals with NAFLD to diagnose NASH earlier in the course of the disease and to enable the initiation of life style measures to reduce the risk factors for progression [85]. A combined approach, using the various advantages of different biomarkers/modalities to individual requirements, may provide an efficient way forward.
Cohort studies examining nonalcoholic fatty liver disease in the setting of HIV
To date, there have been more than 30 studies examining NAFLD in the setting of HIV. We have reviewed the 14 studies that all excluded coinfection with HCV and/or HBV as well as excessive alcohol intake (Table 3) [24,30,86-96]. All but one were cross-sectional [46] and cohort size ranged from 14 to 796, with nine studies having cohorts of more than 60 participants. None of these studies were performed in LMICs indicating the significant gap of our understanding of NAFLD in this setting.
Modalities used for the diagnosis of NAFLD/NASH/steatosis varied, including ultrasound, liver biopsy, computed tomography (CT), magnetic resonance spectroscopy (MRS), hepatic steatosis index (HSI) and controlled attenuation parameter (CAP). Reported prevalence of NAFLD and NASH in HIV ranged from 26-65 and 31-72%, respectively. Only two studies reported statistically significant associations between HIV-related factors and NAFLD or NASH after adjustment, which included duration of HIV infection [30] and longer exposure to NRTIs [86]. All other significant associations were non-HIV related, including traditional risk factors that have been identified in the general population: waist circumference [86,87], age [88], lower high-density lipoprotein (HDL) [87], higher triglycerides [87,89], alanine aminotransferase (ALT)/aspartate aminotransferase (AST) ratio and/or ALT [86,90,91], homeostatic model for assessment for insulin resistance [24,88,92,93], higher fasting glucose and insulin [24], male sex [86,88], black ethnicity [46], lower albumin [46], BMI [90,91], visceral adipose tissue (VAT) [92], dyslipidaemia [90], higher APRI and fibrosis-4 scores [30] and γ: glutamyl transferase (GGT) [88,93].
Two studies identified associations with metabolic genetic factors. The overexpression of sterol regulatory element binding protein 1 (SREBP-1), involved in the regulation of lipid metabolism, was associated with steatosis [94] and the presence of single-nucleotide polymorphisms in the gene coding for patatin-like phospholipase domain-containing 3 protein (PNPLA3) was associated with NASH [95]. However, these polymorphisms are general risk factors and not HIV-specific.
Although all 14 study cohorts had a lower proportion of women participants (0-29%) than men, male sex was reported as a risk factor in two cohorts (24 and 8% women [86,88]). One cohort (35% women) presented major analyses by sex [92]. They reported that although HIV-infected women (73% on ART) had a higher liver fat fraction (LFF) than HIV-negative women, LFF was similar in HIV-positive (97% on ART) and HIV-negative men. LFF was defined as the ratio of total lipids to total lipids and unsuppressed water, as measured by MRS and expressed as a percentage. There was less VAT in women than men, and HIV-infected women had 25% less LFF than HIV-infected men.
Only one published study to date has examined hepatic steatosis longitudinally; however, although the cohort was large (n = 796), the study was retrospective and diagnosed steatosis using the HSI [46]. Median follow-up was 4.9 years and they found an incidence of steatosis of 6.9/100 person-years [95% confidence interval (95% CI) 5.9-7.9]. They reported associations between development of hepatic steatosis with black ethnicity and lower albumin, in adjusted models.
A recent systematic review and meta-analysis examined 10 of these studies and reported a prevalence of NAFLD and NASH in the setting of HIV of 35% (95% CI 29-42) and 42% (95% CI 22-64), respectively [31]. Although the key risk for factors for NAFLD were related to metS (BMI, waist circumference, T2DM, hypertension and triglycerides), they did report an association between NAFLD and higher CD4+ cell count [mean difference of 54.83 (95% CI 11.55-98.11, P = 0.0)], based on data from four studies [86,87,89,90].
Although NAFLD is more prevalent in HIV than in the general population, studies in HIV monoinfection cohorts to date have not consistently found statistically significant associations with HIV-related factors. This may likely be due to the limited numbers of studies that have investigated NAFLD in HIV and excluded viral hepatitis/excess alcohol, and the diversity in the cohorts. There is an obvious need for prospective longitudinal studies sufficiently powered to robust outcomes.
Treatment for nonalcoholic fatty liver disease and nonalcoholic steatohepatitis
Currently available options
There are no treatments for NAFLD or NASH that are specific to people living with HIV. Currently, lifestyle modification and dietary changes are the main interventions for NAFLD, in conjunction with the specific treatment of associated metabolic disorders. A multidisciplinary team approach is beneficial to manage such cases [97].
Lifestyle changes including weight loss achieved by both dietary changes and enhanced physical exercise remain the first line of treatment [98]. Weight loss of about 3-5% of body weight is recommended to improve steatosis and about 10% to improve histological features of NASH, including portal inflammation and fibrosis [98].
Pharmacologic therapy is currently offered only to those with NASH, and in particular those with NASH and significant fibrosis. Current European guidelines suggest the use of glitazones or Vitamin E, or in combination [99]. There are no drugs to date that have completed phase III trials and none approved for NASH by regulatory authorities [99]. The underlying risk factors should be managed as per the standard of care for individual disorders [100].
Approaches in development
There are several new agents in development in the drug pipeline. Of interest in HIV-NAFLD is cenicriviroc (CVC), a chemokine receptor type 5 (CCR5) and chemokine receptor type 2 (CCR2) antagonist that has anti-HIV activity both in vitro and in vivo[101,102]. CCR2 plays a critical role in recruiting monocytes with an inflammatory phenotype into the liver, while CCR5 attracts lymphocytes into the liver and is directly involved in fibrogenic activation of hepatic stellate cells [103]. In addition, CVC has been shown to reduce circulating markers of monocyte inflammation in HIV-infected individuals on ART, including soluble CD14 [102]. In a large randomized phase 2b study of CVC in HIV-negative individuals with NASH (CENTAUR), CVC resulted in a reduction in multiple markers of inflammation, including IL1b, high-sensitivity CCRP, IL6 and fibrinogen [104]. In addition, a reduction in soluble (s)CD14 and increase in the ligands for CCR2 (CCL2) and CCR5 (CCL3 and 4) was observed. These changes in chemokine activity reduce inflammation and fibrotic activity, which was, most importantly, observed clinically in this study with a significant decline in hepatic fibrosis [104]. CVC is now in Phase 3 clinical trials and is not yet approved.
Specific challenges in low and middle-income countries
Most LMIC ART programmes routinely offer TDF-based therapy as first-line treatment due to toxicity associated with ddI, d4T and zidovudine. However, the availability of newer NRTIs, such as TAF and abacavir-based regimens, is still out of reach for many in LMIC due to cost. In addition, most government programmes do not currently have any defined guidelines to detect ongoing liver inflammation. Most liver dysfunction is diagnosed either incidentally by assessment of LFTs (often in the stage of steatohepatitis) or on presentation with decompensated liver disease. Cost-effective screening algorithms for LMICs based on sound scientific data are needed to screen for NAFLD-related disorders.
In India specifically, as South Asians have higher total body fat for a given BMI than western populations, including at normal range BMI [105,106], healthy BMI may not be indicative of total body fat in this population. Rates of metS and insulin resistance are high in Asian populations, and in particular in South Asia [107]. There are limited published studies with data from HIV-infected cohorts, including Japan [90], China [89] and Thailand [108]. Data in HIV-infection from Japan and China have shown an incidence of NALFD among HIV-infected individuals of Asian origin similar to that observed in western countries, and an association with conventional NAFLD-related risk factors. However, the Thai data (retrospective study of HIV-positive forensic autopsies from one hospital) reported that only 7.5% showed fatty change of the liver [108]. Data from the Indian subcontinent are sparse and mainly confined to autopsy in the general population, where in one study 9.2% of the cohort was HIV positive [109]. Data from African countries are not uniform, with both a lower prevalence than in whites reported in HIV-infected individuals on ART (range 6-42 months) in Nigeria [110], while a similar prevalence to that in the West was reported in HIV-infected individuals on ART (minimum 6 months) in Cameroon [111]. There is a great need for prospective cohort studies generating data in LMICs, which would then lead to specific policies and guidelines, tailored to these countries [112,113].
Management and monitoring of NAFLD in HIV-infected individuals, particularly in LMIC where there is increased visceral fat in the healthy weight range, increasing levels of obesity and prediabetes combined with large numbers of people living with HIV, could be a very significant health burden that requires management and prevention strategies. Screening individuals newly diagnosed with HIV using a cost-effective and widely available modality such as liver ultrasound would be beneficial, although effective treatment options for NAFLD beyond modification of lifestyle factors remain elusive. An initial focus on monitoring those diagnosed with NAFLD in the setting of HIV in LMICs would be useful, particularly tracking fibrosis progression with T/E or CAP to identify those at risk of NASH and/or HCC.
| |
| |
| |
|
|
|
