Prevalence and risk factors of nonalcoholic steatohepatitis with significant fibrosis in people with HIV

Introduction

HIV is one the most common infectious diseases worldwide. Although associated with a poor prognosis during earlier decades, significant improvements have been achieved with the introduction of long-term antiretroviral therapy (ART) [1]. However, despite major accomplishments in the life expectancy of people with HIV (PWH), liver-related comorbidities are the second leading cause of mortality [2]. Although the impact of chronic hepatitis B virus (HBV) and hepatitis C virus (HCV) coinfection as the underlying cause of chronic liver diseases is well established, the prevalence of metabolic risk factors and the number of PWH developing nonalcoholic fatty liver disease (NAFLD) has been increasing in recent years [3–5]. In the general population, NAFLD has become one of the most common chronic liver diseases, with a prevalence of approximately 25% globally [6] bearing a high economic impact [7]. NAFLD can progress to a more inflammatory stage, termed nonalcoholic steatohepatitis (NASH), with a significant risk of liver fibrosis, cirrhosis, and even liver cancer [8]. In this context, liver fibrosis has become the most important predictor of mortality [9].

Overall, the metabolic syndrome, a major risk factor for NAFLD, has a higher prevalence in PWH compared with the general population [10]. Both HIV infection itself and ART have been implicated as steatogenic factors [11]. In this context, a recent study suggested that tenofovir alafenamide (TAF) was associated with more weight gain and less favourable metabolic outcomes compared with tenofovir disoproxil fumarate (TDF), independent of age [12]. A meta-analysis, however, could not detect a higher risk of NAFLD on ART [3]. Overall, only few studies have investigated the burden of hepatic steatosis and fibrosis in PWH and data on NASH with significant fibrosis is limited [13,14]. One recent analysis exploring hepatic steatosis in HIV-monoinfected and HCV-coinfected patients in Germany did not report data on hepatic fibrosis [15].

Liver biopsy is not feasible as a screening tool and will be replaced by noninvasive tests (NITs) and biomarkers [16]. Importantly, the histological disease stage is considered as an indication for antisteatotic, antifibrotic, and anti-inflammatory drugs in NASH that are currently being developed. Vibration-controlled transient elastography (VCTE) has been established and validated as an ultrasound-based test to screen for hepatic steatosis and fibrosis [17,18]. More recently, the Fibroscan-AST (FAST) score has been introduced to screen patients at risk of nonalcoholic steatohepatitis with significant fibrosis [19]. For the utility of a diagnostic test, the prevalence of the condition is of importance for the positive-predictive value (PPV). Therefore, the aim of this study was first to analyse the prevalence, risk factors, and independent predictors of NAFLD and significant fibrosis and secondly investigate the prevalence of steatohepatitis by means of VCTE and the FAST score in PWH.

Methods Study design and population

A total of 282 individuals with an HIV infection were enrolled between 2018 and 2021 in this noninterventional, cross-sectional and prospectively enrolling monocentric cohort study (FLASH, Prevalence of Advanced Fibrosis in Patients Living With HIV, NCT04066608) after informed consent was obtained at the outpatient clinic of the Metabolic Liver Research Program at the University Medical Centre Mainz in Germany. Participants with HIV 18 years of age or older were included into this study. HBV or HCV coinfection status was assessed. Patients with an active malignancy were excluded. Clinical assessment was used to categorize patients according to the amount of alcohol consumption. NAFLD (alcohol intake: male patient <20 g/day, female patient <10 g/day) or alcohol-related liver disease (ALD) were defined according to current practice guidelines [20]. The metabolic syndrome and its associated risk factors including waist circumference (men ≥94 cm, women ≥80 cm), diabetes mellitus (previously diagnosed type 2 diabetes), raised triglycerides (≥150 mg/dl, or treatment of this condition), low HDL cholesterol (men <40 mg/dl; women <50 mg/dl) and arterial hypertension (systolic ≥130 mmHg or diastolic ≥85 mmHg) were defined according to the criteria of the international diabetes federation (IDF) [21]. In this context, the metabolic syndrome was evident if central obesity according to a waist circumference or a BMI greater than 30 kg/m2 and any two of the previously mentioned factors were present. BMI [kg/m2; weight (kg)/height2 (m2)] and waist circumference (cm) were assessed as well as laboratory values were obtained at study inclusion. Assessment of medical and treatment history were retrieved from the electronic healthcare records.

Assessment of hepatic steatosis, fibrosis and steatohepatitis

Hepatic steatosis (CAP, dB/m) and fibrosis (LSM, kPa) were noninvasively assessed using VCTE (FibroScan 430 mini; SMART Exam was introduced in 2020; Echosens, Paris, France) [22]. In the majority of patients, the M probe (91.1%; n = 257) was used. In cases of severe obesity, the XL probe (8.9%; n = 25) was used. The success rate was 93.4%. A total of 20 participants had to be excluded from the study because of invalid measurements using VCTE. The recently suggested cut-off above 275 dB/m in the EASL guidelines on noninvasive tests was used to diagnose hepatic steatosis [23]. For the assessment of liver fibrosis, a cut-off value of at least 8.2 kPa was considered a significant fibrosis (≥F2) [17]. Measurement of LSM was considered reliable if the interquartile range (IQR) was less than 30% and the success rate greater than 70% [24]. To identify patients at risk of more progressive and inflammatory fatty liver disease, that is, NASH, the FAST score with cut-off values of greater than 0.35 and at least 0.67, respectively, were applied [19]. Briefly, the FAST score combines LSM, CAP and AST blood values into a specific equation to rule in NASH and has been validated in non-HIV-infected individuals [19]. Due to missing AST blood values, the FAST score was available in 93.3% (n = 263). Additional surrogate scores of advanced fibrosis included the following: NAFLD fibrosis score (NFS) and Fibrosis-4 (FIB-4) score with published cut-off values to rule in or rule out advanced fibrosis [25–27].

Ethics

All patients provided written informed consent. The study was conducted according to the ethical guidelines of the 1975 Declaration of Helsinki (sixth revision, 2008). The study protocol was approved by the ethics committee of the Landesärztekammer Rhineland-Palatinate [Nr. 873.199.10 (7208)].

Statistical analysis

Descriptive analysis of data is expressed as median values with interquartile ranges (IQR 25th–75th). The Mann–Whitney U rank test was used to compare groups and to calculate differences between two groups with continuous variables. Categorical variables are presented as frequencies and percentages. For the comparison of two or more patient-groups, a chi-squared test was applied. All tests were two-tailed, statistically significant values were defined as P less than 0.05. All variables with P less than 0.05 and the clinical parameters age, sex and alcohol intake were then included into a multivariable logistic regression model to examine associations with NAFLD, fibrosis (LSM ≥8.2 kPa) and a FAST score greater than 0.35. Due to the large number of tests, P values should be interpreted with caution and in connection with effect estimates. For all data analysis and statistical tests, IBM SPSS Statistic Version 23.0 (IBM Corp., Armonk, New York, USA) was used. For all figures, Microsoft Excel 2016 (Microsoft Corp., Redmond, Washington, USA) was used.

Results Demographic and clinical characteristics

A total of 282 PWH fulfilling the inclusion and exclusion criteria were analysed. The majority of patients were men (n = 198, 70.2%). The median age was 51 years (IQR 42–58) and the median duration of established HIV infection was 12 years (IQR 6–20). The median BMI (kg/m2) was 25 (IQR 22.3–28.1), and 15.6% (n = 44) were obese according to a BMI greater than 30 kg/m2. The median waist circumference (cm) was 96 (IQR 86.8–104), and a total of 40.8% of male and 25.2% of female participants showed a waist circumference above 94 and 80 cm, respectively. The criteria of the metabolic syndrome were fulfilled in 72 (25.5%) individuals at study inclusion. Alcohol consumption in most individuals was less than 20 g/day (male) and less than 10 g/day (female). Most PWH received NRTI as part of their ART with a majority receiving TAF-containing combinations. The majority showed a controlled HIV disease as indicated by HIV RNA below the threshold and CD4+ cells above the threshold of 500 cells/μl. Baseline characteristics and laboratory results are summarized in Table 1.

Table 1 - Baseline characteristics and metabolic profile of people with HIV. Variable Total cohort (n = 282) Age in years 51 (42; 58) Time since diagnosis (years) (n = 268) 12 (6; 20) Male 198 (70.2) Female 84 (29.8) VCTE  CAP (dB/m) 248 (214.8; 300)  LSM (kPa) 4.6 (3.8; 5.7) Metabolic comorbidities  BMI (kg/m2) (n = 272) 25 (22.3; 28.1)   Underweight (<18.5 kg/m2) 9 (3.2)   Normal weight (18.5 to <25 kg/m2) 126 (44.7)   Overweight (25 to <30 kg/m2) 93 (33)   Obese (>30 kg/m2) 44 (15.6)  Waist circumference (cm) (n = 270) 96 (86.8; 104)   Male >94 cm 115 (40.8)   Female >80 cm 71 (25.2)  Type 2 diabetes (n = 261) 30 (10.6)  Total cholesterol >200 mg/dl (n = 181) 91 (32.3)  Triglycerides >150 mg/dl (n = 175) 71 (25.2)  HDL-cholesterol: male <40 mg/dl 39 (13.8)  female <50 mg/dl (n = 149) 13 (4.6)  Arterial hypertension (n = 268) 85 (30.1)  Metabolic syndrome 72 (25.5) Alcohol consumption (n = 282)  Male >20 (g/day) No: 165 (90.2); yes: 22 (9.8)  Female >10 (g/day) No: 68 (90.7); yes: 7 (9.3) Laboratory values  ALT (U/l) (n = 263) 24 (18; 32)  AST (U/l) (n = 263) 26 (23; 32)  Triglycerides (mg/dl) (n = 175) 131 (91; 190)  Cholesterol (mg/dl) (n = 181) 200 (174; 226)  HDL (mg/dl) (n = 133) 48 (39; 57.5)  LDL (mg/dl) (n = 133) 120 (102; 144)  HbA1c (%) (n = 133) 5.4 (5.1; 5.7)  Uric acid (mg/dl) (n = 146) 5.5 (4.8; 6.4) Hepatitis serology  Anti-HCV positive (n = 134) 8 (2.8)  HBsAg positive (n = 129) 4 (1.4) HIV-related parameters and medication (ART)  CDC stage (n = 185) A: 80 (43.2); B: 47 (16.7); C: 58 (20.6)  HIV RNA (n = 274)   Above threshold 102 (36.2)   Below threshold 172 (61)  CD4+ (cells/μl) (n = 266) 723.5 (515.8; 910.8)   >500 cells/μl 207 (73.4)  NRTI 253 (89.7)   TAF as part of ART 179 (63.5)   TDF as part of ART 31 (11)  NNRTI 62 (22)  PI 41 (14.5)  INSTI 188 (66.7)   DTG 68 (36.2)  TAF and INSTI 124 (44)

Data are expressed as numbers, median, percentage (%) or interquartile ranges (IQR 25th–75th). ALT, alanine-aminotransaminase; ART, antiretroviral therapy; AST, aspartate-aminotransaminase; CDC, Centres for Disease Control and Prevention; DTG, dolutegravir; HDL, high-density lipoprotein; INSTI, integrase inhibitors; LDL, low-density lipoprotein; TAF, tenofovir alafenamide; TDF, tenofovir disoproxil fumarate; NRTI, nucleoside reverse-transcriptase inhibitors; NNRTI, nonnucleotide reverse transcriptase inhibitors; PI, protease inhibitors.


Prevalence of hepatic steatosis and fibrosis in people with HIV

The median CAP (dB/m) was 248 (IQR 214.8–300). According to a cut-off value of at least 275 dB/m, a total of 100 PWH (35.5%) presented with hepatic steatosis (Supplementary Figure 1a, https://links.lww.com/QAD/C567). On the basis of the presence of hepatic steatosis according to a CAP greater than 275 dB/m and alcohol intake (g/day), 76 (27%) PWH were considered to have NAFLD, whereas 13 (4.6%) qualified for alcohol-related liver disease (ALD). Because of missing data on alcohol intake, a total of 11 PWH remained undefined.

The median LSM (kPa) in the study cohort was 4.6 (IQR 3.8–5.7). A total of 263 (93.3%) PWH had a LSM below 8.2 kPa excluding relevant fibrosis, whereas 19 (6.7%) participants presented with significant fibrosis (F2; cut-off ≥ 8.2 kPa) (Supplementary Fig. 1b, https://links.lww.com/QAD/C567).

Comparison of people with HIV characteristics with nonalcoholic fatty liver disease and fibrosis

A high prevalence (n = 38, 38.2%) of the metabolic syndrome and its associated risk factors was seen in PWH with NAFLD. The median BMI (kg/m2) was significantly higher if NAFLD was present [27.7 (IQR 25.4–31.6)]. Obesity (BMI >30) was more prevalent in PWH with NAFLD. Median waist circumference (cm) was higher in NAFLD [102 (IQR 97–113)], and a higher prevalence of a waist circumference (cm) greater than 94 (men; n = 49) and greater than 80 (women; n = 16), respectively, was seen. The blood levels of ALT, triglycerides and uric acid showed higher median values in NAFLD. PWH with significant fibrosis (LSM ≥8.2 kPa) were older [55 (IQR 50–63)] and showed a longer disease duration (time since diagnosis). Furthermore, a higher median BMI (kg/m2) [27.1 (IQR 24.5–34.7)] and waist circumference (cm) [104 (95–117.3)] were observed. Liver enzymes were significantly higher in PWH presenting with fibrosis [ALT: 36 (IQR 24–59); AST: 35 (IQR 28–40.5)]. In contrast, HIV-related parameters were not significantly different between these groups. Numerically, a higher proportion of PWH and NAFLD were treated with TAF compared with TDF. A comparison of patients based on the presence of NAFLD and significant fibrosis is shown in Table 2. Furthermore, anti-HBc and anti-HCV were positive in n = 5 and n = 2 in the subgroup with a LSM at least 8.2 kPa, respectively. The NFS identified a higher number of PWH at risk of fibrosis than the FIB-4. The surrogate scores of advanced fibrosis are shown in Supplementary Table 1, https://links.lww.com/QAD/C567.

Table 2 - Clinical characteristics of people with HIV with and without NAFLD and fibrosis. No steatosis (n = 182; 64.5%) NAFLD (n = 76; 27%) No fibrosis (n = 263; 93.3%) Fibrosis (n = 19; 6.7%) Variable n (% or IQR) n (% or IQR) P value n (% or IQR) n (% or IQR) P value Age (years) 50 (41–57) 54 (49–60) 0.011 51 (41–58) 55 (50–63) 0.061 Time since diagnosis (years) 12 (6–19) 14 (6–23) 0.475 12 (6–19) 21 (7–27) 0.061 Sex 0.025 0.389  Male 118 (64.8) 60 (78.9) 183 (69.6) 15 (78.9)  Female 64 (35.2) 16 (21.1) 80 (30.4) 4 (21.1) Metabolic comorbidities  BMI (kg/m2) 23.4 (21.5; 26.4) 27.7 (25.4–31.6) <0.001 24.9 (22.3–27.9) 27.1 (24.5–34.7) 0.039  Obese (>30 kg/m2) 12 (6.8) 26 (35.1) <0.001 37 (14.6) 7 (38.8) 0.007  Waist circumference (cm) 91 (84–100) 102 (97–113) <0.001 96 (86–103) 104 (95–117.3) 0.001  Male >94 52 (44.1) 49 (81.6) <0.001 102 (55.7) 13 (86.6) 0.020  Female >80 51 (79.7) 16 (100) 0.049 68 (85) 3 (75) 0.589  Type 2 diabetes 15 (8.7) 11 (16.9) 0.071 24 (9.8) 6 (40) <0.001  High triglycerides 31 (29.2) 33 (61.1) <0.001 66 (40) 5 (50) 0.532  High cholesterol 51 (45.9) 40 (61.1) 0.067 87 (50.6) 4 (44.4) 0.720  Arterial hypertension 47 (26.5) 38 (39.1) 0.053 77 (30.5) 8 (55) 0.105  Metabolic syndrome 34 (18.6) 38 (38.2) 0.001 64 (24.3) 8 (42.1) 0.086 Laboratory values  ALT (U/l) 22 (17–30) 28 (18.3–38) 0.001 23 (17.8–31) 36 (24–59) 0.001  AST (U/l) 26 (22–30) 26 (23–32.8) 0.507 26 (22–31) 35 (28–40.5) 0.001  Triglycerides (mg/dl) 108.5 (82.8–157.3) 183 (123.8–246.3) <0.001 131 (86.5–188) 154.5 (106–242.5) 0.337  Cholesterol (mg/dl) 196 (173–222) 207 (182.8–228) 0.147 200 (173.3–226) 186 (178–217) 0.669  HDL (mg/dl) 48 (41–59.5) 43.5 (38–52) 0.071 48 (39–58.3) 45 (40–53) 0.614  LDL (mg/dl) 120 (102.5–144.5) 122.5 (103.5–143.5) 0.589 120 (102–144) 127 (114–148) 0.789  HbA1c (%) 5.4 (5.1–5.6) 5.5 (5.2–6.1) 0.101 5.4 (5.1–5.7) 5.7 (5.5–7.0) 0.024  Uric acid (mg/dl) 5.3 (4.6–6.3) 6.0 (5.1–6.9) 0.013 5.5 (4.8–6.3) 6.9 (4.7–8.3) 0.230 HIV-related parameter  HIV RNA at inclusion 0.684 0.297  Above threshold 67 (37.9) 26 (35.1) 98 (37.9) 4 (25)  Below threshold 110 (62.1) 48 (64.8) 160 (62.1) 12 (75)  CD4+ (cells/μl) 695 (515.8–881.5) 802.5 (533.5–1019) 0.114 723 (515.8–897) 769.5 (436.8–1105.5) 0.685  >500 CD4+ cells/μl 135 (74.2) 72 (72.4) 0.764 195 (74.1) 12 (63.2) 0.295  CDC C 36 (31) 22 (25) 0.426 53 (31.3) 5 (31.3) 0.993  NRTI (TAF vs. TDF) 0.369 0.303  TAF 112 (83.6) 52 (88.1) 170 (85.9) 9 (75)  TDF 22 (16.4) 7 (11.9) 28 (14.1) 3 (25)  INSTI 120 (66.6) 52 (70.3) 0.577 175 (67.6) 13 (72.2) 0.683  TAF and INSTI 73 (76) 39 (78) 0.790 119 (76.3) 5 (71.4) 0.768

Data are expressed as numbers, median, percentage (%) or interquartile ranges (IQR 25th–75th). ALT, alanine-aminotransaminase; AST, aspartate-aminotransaminase; CAP, controlled attenuation parameter; HDL, high-density lipoprotein; INSTI, integrase inhibitors; LDL, low-density lipoprotein; LSM, liver stiffness measurement; NAFLD, nonalcoholic fatty liver disease; NRTI, nucleoside reverse transcriptase inhibitors; TAF, tenofovir alafenamide; TDF, tenofovir disoproxil fumarate. Mann–Whitney U test and chi-square test were used to compare continuous and categorical values, respectively. Boldface indicates statistical significance. A P value less than 0.05 was considered statistically significant.


Clinical predictors of nonalcoholic fatty liver disease and fibrosis in people with HIV

A multivariable logistic regression model was built including all clinical variables with a P less than 0.05 in the univariable analysis as well as the clinical parameters age and sex to assess predictive factors of NAFLD and fibrosis in this cohort. Waist circumference (OR = 1.1, 95% CI 1.067–1.135, P < 0.001) remained the only independent predictor of NAFLD. The variable type 2 diabetes (OR = 5.056, 95% CI 1.386–18.44, P = 0.014) remained the only independent predictor of significant fibrosis (Table 3). In turn, HIV-related parameters showed no association with NAFLD and fibrosis in the univariable analysis (data not shown).

Table 3 - Multivariable analyses of predictors for NAFLD and fibrosis in people with HIV. NAFLD Fibrosis Multivariable logistic regressiona Multivariable logistic regressionb Variable OR CI P value OR CI P value Waist circumference (cm) 1.1 1.067–1.135 <0.001 Type 2 diabetes 5.056 1.386–18.44 0.014

Multivariable logistic regression analysis of data is shown. With all factors showing a P value less than 0.05 in the univariable analysis and the variables, age and sex, a multivariable regression model was built. Age, sex, BMI and ALT were not predictive of NAFLD. Likewise age, sex, BMI, ALT, AST and alcohol intake were not predictive of fibrosis. Metabolic syndrome was excluded from the multivariable analysis to avoid multicollinearity. The variable triglycerides was excluded because of limited numbers available. Odds ratio (OR) and 95% confidence interval (CI) are shown. Boldface indicates statistical significance. A P value less than 0.05 was considered statistically significant. NAFLD, nonalcoholic fatty liver disease.

aMultivariable logistic regression: age, sex, BMI, waist circumference, ALT (n = 230).

bMultivariable logistic regression: age, sex, BMI, waist circumference, ALT, AST, type 2 diabetes, alcohol intake (n = 212).


Prevalence of steatohepatitis with significant fibrosis in people with HIV

To identify PWH with steatohepatitis and at least significant fibrosis, the FAST score was explored (n = 263). In total, the FAST score was above the cut-off more than 0.35 in 32 (12.1%) of the patients. Of these, 9.1% (n = 24) of the cohort were in the intermediate range between greater than 0.35 and less than 0.67, whereas 3% (n = 9) ruled-in applying the cut-off of at least 0.67 (Supplementary Figure 2, https://links.lww.com/QAD/C567). In turn, 231 (87.8%) of PWH were FAST score negative. In the 32 individuals with a FAST score greater than 0.35, n = 28 were classified as NASH, whereas n = 2 qualified for alcoholic steatohepatitis (ASH). Due to missing data on alcohol intake, a total of two PWH remained undefined.

Comparison of people with HIV characteristics with a FAST score greater than 0.35 and at least 0.67

The majority of PWH with a FAST score greater than 0.35 were of male sex (84.4%). Metabolic comorbidities were generally more prevalent in PWH with a FAST score greater than 0.35. The BMI (kg/m2) was significantly higher in both groups (>0.35 and ≥0.67) in comparison to less than 0.35, and more patients were obese (>30 kg/m2). Median waist circumference was also higher in both groups, and particularly more men showed a waist circumference of more than 94 cm. Likewise, a diagnosis of type 2 diabetes at study inclusion was more prevalent in both groups. Liver-related blood values (ALT, AST) as well as triglycerides and HbA1c (%) were significantly higher in PWH and a FAST score greater than 0.35. No major differences were seen in the analysis of HIV-related parameters within these groups. A comparison of these characteristics is shown in Table 4. Additionally, in the group with a FAST score greater than 0.35, n = 7 were anti-HBc-positive and n = 2 anti-HCV-positive.

Table 4 - Comparison of people with HIV characteristics with a FAST-score greater than 0.35 and at least 0.67. FAST score <0.35 (n = 231; 87.8%) FAST score >0.35 (n = 32; 12.2) FAST score <0.67 (n = 255; 97%) FAST score ≥0.67 (n = 8; 3%) Variable n (% or IQR) n (% or IQR) P value n (% or IQR) n (% or IQR) P value Age in years 52 (42–58) 51.5 (44.3–59.5) 0.512

留言 (0)

沒有登入
gif