Prevalence of Liver Steatosis and Fibrosis in Adults With Primary Hypobetaliproteinemia: Results From the HYPOCHOL Study

Primary hypobetalipoproteinemia (HBL) is characterized by sustained very low levels (<5th percentile for age and sex) of low-density lipoprotein cholesterol (LDL-C).1Welty F.K. Curr Opin Lipidol. 2020; 31: 49-55Crossref PubMed Scopus (47) Google Scholar Primary monogenic HBLs are referred as familial hypobetalipoproteinemia (FHBL) disorders.2Bredefeld C. et al.Gastroenterology. 2021; 160: 1912-1916Abstract Full Text Full Text PDF PubMed Scopus (22) Google Scholar In adults, FHBL most frequently arises from heterozygous variants in the apolipoprotein B (APOB) gene, which encodes a truncated form of apoB, leading to a defect in hepatic very low-density lipoprotein (VLDL) secretion.3Bredefeld C. et al.J Clin Lipidol. 2022; 16: 797-812Abstract Full Text Full Text PDF PubMed Scopus (24) Google Scholar Although people with HBL are usually asymptomatic, some of them can develop hepatic steatosis due to an impaired secretion of VLDL.4Schonfeld G. et al.J Lipid Res. 2003; 44: 470-478Abstract Full Text Full Text PDF PubMed Scopus (132) Google Scholar,5Sankatsing R.R. et al.Arterioscler Thromb Vasc Biol. 2005; 25: 1979-1984Crossref PubMed Scopus (98) Google Scholar Consistently, we recently demonstrated that primary HBL was associated with an increased risk of liver complications, including cirrhosis and primary liver cancer, in the general population.6Wargny M. et al.J Hepatol. 2024; 80: 846-857Abstract Full Text Full Text PDF PubMed Scopus (3) Google Scholar However, only few prospective data address the natural history of liver injury in HBL. Here, we sought to determine the prevalence of liver steatosis and fibrosis as evaluated with the FibroScan device in primary HBL. Among 201 adults with primary HBL included in the HYPOCHOL and GENELIP studies, 104 who had undergone a FibroScan examination were included in this analysis (mean LDL-C, 47.7 ± 15.7 mg/dL). Liver steatosis was defined by a controlled attenuation parameter (CAP) >275 dB/m and significant and advanced liver fibrosis by a liver stiffness of ≥8 kPa and ≥12 kPa, respectively. Multivariable logistic regressions were used to assess the association with liver steatosis and fibrosis, as described in the Supplementary Methods. Liver steatosis was detected with FibroScan in 33 of 104 patients (31.7%; 95% confidence interval [CI], 23%–42%) and with liver ultrasound in 27 of 84 (32.1%; 95% CI, 22%–43%). The prevalence of liver steatosis increased with age, reaching 68% (95% CI, 43%–87%) after the age of 50, and it was detected at an earlier age in individuals with FHBL with pathogenic APOB variant (Figure 1A–B). Clinical and biological characteristics according to the presence of liver steatosis are detailed in Supplementary Table 1. Patients with HBL with liver steatosis were more frequently men (63.6% vs 36.6%; P = .011), older (45.9 ± 18.4 vs 30.5 ± 12.9 years; P < .001), with a higher body mass index (BMI) (28.1 ± 5.5 vs 21.8 ± 3.7 kg/m2; P < .001) and a higher frequency of type 2 diabetes (27.3% vs none in the absence of steatosis; P < .001) and arterial hypertension (39.4% vs 2.9%; P < .001). They were also more frequently carriers of a pathogenic APOB mutation (48.5% vs 19.7%; P = .005). In univariable logistic regression analyses, advancing age, higher BMI, diabetes, elevated plasma triglycerides concentrations, and the presence of pathogenic APOB variant were significantly and positively associated with liver steatosis. Conversely, female sex was inversely associated with the risk of liver steatosis (Supplementary Table 2). Applying multiple adjustments based on common confounding factors, only increasing BMI (odds ratio [OR], 3.68; 95% CI, 1.90–8.06; P = .0003) and pathogenic APOB variant (OR, 5.56; 95% CI, 1.64–21.5; P = .0078) remained associated with liver steatosis. Significant liver fibrosis was detected in 14.4% (95% CI, 8.3%–22.7%) of the study population. Analysis of fibrosis distribution across age categories revealed that significant liver fibrosis was diagnosed in only 4% (95% CI, 0.5%–13.0%) in the 18- to 29-year-old age group, but up to 53% (95% CI, 29%–76%) in the oldest 50- to 85-year-old age group. Moreover, 26% (95% CI, 9%–51%) of subjects with HBL in the oldest group were suspected of having advanced fibrosis, and this figure was as high as 42% (95% CI, 10%–82%) for those with FHBL (Figure 1C–D). The majority of phenotypic traits observed in individuals with significant liver fibrosis were comparable to those in subjects with liver steatosis. However, although pathogenic APOB genetic variants were numerically more frequent in subjects with fibrosis, this difference did not reach statistical significance (40% vs 27%; P = .36). To identify the factors associated with significant liver fibrosis, the same regression analyses as for liver steatosis was performed (Supplementary Table 2). Apart from the presence of a pathogenic APOB variant and male sex, the factors were the same as those identified for liver steatosis: older age, higher BMI, diabetes, and higher triglyceride levels. In multivariable analysis, only increasing age remained significantly associated with liver fibrosis (OR [+1 standard deviation], 3.22; 95% CI, 1.60–7.14; P = .0017). This study shows that primary HBL is associated with a high prevalence of liver fibrosis, representing a major risk factor for liver complications. Above all, it clearly demonstrates that the risk of liver fibrosis increases significantly with age, affecting over one-half of individuals with HBL over 50 years of age. This is certainly explained by the transition from steatosis to fibrosis, which takes several years. Even more strikingly, 26% of individuals with HBL over 50 years of age have a risk of advanced liver fibrosis (liver stiffness ≥12 kPa) requiring specific hepatological management, in accordance with the latest guidelines.7Rinella M.E. et al.Hepatology. 2023; 77: 1797-1835Crossref PubMed Scopus (563) Google Scholar In line with a recent study,8Zheng M. et al.JHEP Reports. 2023; 5100692Abstract Full Text Full Text PDF Scopus (9) Google Scholar our data also show that metabolic factors, such as weight gain, contribute independently to the development of liver steatosis in HBL. These findings hold considerable clinical interest, advocating for more widespread screening for HBL whithin the general population, especially in people with steatotic liver disease. These data also resonate with recent observations indicating that primary HBL was associated with a significant increase in the risk of liver complications in the general population. In the UK Biobank cohort, primary HBL was associated with over 3-fold increase in the risk of liver complications, including decompensated cirrhosis and hepatocellular carcinoma, compared with a control population.6Wargny M. et al.J Hepatol. 2024; 80: 846-857Abstract Full Text Full Text PDF PubMed Scopus (3) Google Scholar Additionally, the 14.4% prevalence of liver fibrosis in our HBL cohort is much higher than that the 2.6% observed among people with steatosis in the French cohort CONSTANCES.9Nabi O. et al.Gastroenterology. 2020; 159: 791-793.e2Abstract Full Text Full Text PDF PubMed Scopus (40) Google Scholar As previously reported,10Rimbert A. et al.Arterioscler Thromb Vasc Biol. 2021; 41: e63-e71Crossref PubMed Scopus (19) Google Scholar the presence of FHBL (due to APOB truncating variants) was significantly and independently associated with steatosis, but not with fibrosis, possibly due to insufficient statistical power. Furthermore, patients with FHBL presented earlier and with more severe hepatic impairment. However, liver monitoring in primary HBL should not be limited to patients with FHBL, as 60% of patients with liver fibrosis do not have an identified pathogenic genetic variant. Further work with histological confirmation of liver damage will be required to support these results. In conclusion, primary HBL should be considered as an independent risk factor for both liver steatosis and fibrosis. Early diagnosis of HBL, in the early stages of steatosis, could facilitate the implementation of appropriate measures to detect and prevent liver fibrosis occurence. The authors thank Marie Marrec and Pierre Morcel for data collection and management, Gilles Degraef for coordinating the recruitment of patients in medical analysis laboratories, and Samy Hadjadj, Cédric Le May, and Antoine Rimbert for their careful review of the article. Supplementary Table 1Clinical and Biological Characteristics in Patients With Hypocholesterolemia According to Liver SteatosisAvailable dataAll (N = 104)No liver steatosis (n = 71)Liver steatosis (n = 33)P-valueClinical characteristics Female sex10457 (54.8%)45 (63.4%)12 (36.4%).012 Age, years10435.4 ± 16.430.5 ± 12.945.9 ± 18.4< .001 BMI, kg/m210423.8 ± 5.221.8 ± 3.728.1 ± 5.5< .001 Hypertension10315 (14.6%)2 (2.9%)13 (39.4%)< .001 Current smoker10322 (21.4%)17 (24.3%)5 (15.2%).44 Type 1 diabetes1045 (4.8%)5 (7%)0 (0%).18 Type 2 diabetes1049 (8.7%)0 (0%)9 (27.3%)< .001 Cardiovascular diseases history (any)aCardiovascular disease history: one stroke.1041 (1%)0 (0%)1 (3%).32 Excessive alcohol consumption1044 (3.8%)1 (1.4%)3 (9.1%).093 Other substance abuse10411 (10.6%)8 (11.3%)3 (9.1%)> .99 Personal cancer historybCancer history: adrenocortical cancer (1); tongue parakeratosis (1); mediastinal paraganglioma (1); prostate (1).1044 (3.8%)1 (1.4%)3 (9.1%).093Treatments Estroprogestin5714 (24.6%)11 (24.4%)3 (25%)> .99 Psychotropic-any10411 (10.6%)6 (8.5%)5 (15.2%).32 Psychotropic-neuroleptic1042 (1.9%)2 (2.8%)0 (0%)> .99 Psychotropic-antidepressant1048 (7.7%)4 (5.6%)4 (12.1%).26 Psychotropic-anxiolytic / hypnotics1044 (3.8%)2 (2.8%)2 (6.1%).59 Antidiabetic-any10416 (15.4%)5 (7%)11 (33.3%).001 Antidiabetic-insulin1049 (8.7%)5 (7%)4 (12.1%).46 Antidiabetic-other than insulin10411 (10.6%)0 (0%)11 (33.3%)< .001Genetic variants (ACMG IV and/or V) APOB10430 (28.8%)14 (19.7%)16 (48.5%).005 PCSK91042 (1.9%)2 (2.8%)0 (0%)> .99 Any mutation: APOB and/or PCSK910432 (30.8%)16 (22.5%)16 (48.5%).012Liver investigations US: steatosis8427 (32.1%)8 (13.8%)19 (73.1%)< .001 FibroScan: median CAP104246.9 ± 70.9204.9 ± 34.2337.3 ± 36.1< .001 FibroScan: median elastometry, kPa1045.9 ± 3.75 ± 2.17.9 ± 5.4.006 FibroScan elastometry ≥8.0 kPa10415 (14.4%)5 (7%)10 (30.3%).005 FibroScan elastometry ≥12.0 kPa1048 (7.7%)2 (2.8%)6 (18.2%).012Biology Total cholesterol, mg/dL104123.3 ± 23.5125.9 ± 20.9117.5 ± 27.9.13 LDL-c, mg/dL10447.7 ± 15.748.3 ± 14.646.5 ± 17.9.61 ApoB, mg/dL9848.7 ± 23.448.1 ± 15.749.7 ± 35.1.81 HDL-c, mg/dL10463.2 ± 16.566.9 ± 15.455.4 ± 16.2.001 Triglycerides, mg/dL10451 [38; 69]48 [37.5; 65.5]59 [43; 83].094 Lipoprotein(a), mg/dL556 [3; 14]6.5 [3; 14.5]5 [3; 11].84 Fasting blood glucose, mg/dL10397.5 ± 33.592.8 ± 29.5107.9 ± 39.5.059 HbA1c, mmol/mol9137.3 ± 1134.8 ± 7.443.3 ± 15.2.01 HbA1c, %915.6 ± 1.05.3 ± 0.76.1 ± 1.4.01 HOMA-IR712.44 ± 1.881.90 ± 1.013.94 ± 2.77.005 AST (ULN)990.62 ± 0.40.55 ± 0.190.81 ± 0.65.041 ALT (ULN)990.59 ± 0.330.54 ± 0.190.73 ± 0.51.066 GGT (ULN)990.5 ± 0.330.46 ± 0.180.6 ± 0.54.17 AST >ULN997 (7.1%)2 (2.9%)5 (17.2%).022 ALT >ULN995 (5.1%)2 (2.9%)3 (10.3%).15 GGT >ULN994 (4%)1 (1.4%)3 (10.3%).074 Total bilirubinemia, μmol/L9611.5 ± 7.711.9 ± 8.610.5 ± 5.32 Albumin, g/L9344.5 ± 3.744.9 ± 3.543.6 ± 4.1.13 Fibrosis-4 index940.60 [0.44; 0.94]0.57 [0.42; 0.84]0.71 [0.53; 1.34].019 Platelet count, G/L96243.2 ± 53.2245.3 ± 51.1238.6 ± 58.1.59 Hemoglobin, g/dL9614 ± 1.313.8 ± 1.314.2 ± 1.4.18 Red cells, Tera/L964.8 ± 0.54.8 ± 0.54.9 ± 0.5.41 MCV, fL9687 ± 5.787 ± 5.987.2 ± 5.2.88 Creatinine, μmol/L8970.4 ± 15.571.1 ± 15.969 ± 14.8.54 eGFR, mL/min78108.5 ± 18.4108.9 ± 19.1107.4 ± 16.7.74Note: Categorical parameters are described using number (%). Quantitative parameters are presented using mean ± standard deviation or median (interquartile range), according to distribution.Note: Liver steatosis is considered when CAP >275 dB/m measured by FibroScan. Polygenic hypocholesterolemia is defined by a polygenic risk score (PRS) <10th percentile.ACGM, Amplified consensus genetic marker; ALT, alanine transaminase; AST, aspartate aminotransferase; BMI, body mass index; CAP, controlled attenuation parameter; eGFR, estimated glomular filtration rate; GGT, gamma-glutamyl transferase; HbA1C, hemoglobin A1C; HDL-c, high-density lipoprotein cholesterol; HOMA-IR, homeostasis model assessment of insulin resistance; LDL-c, low-density lipoprotein cholesterol; MCV, mean corpuscular volume; NR, not relevant; ULN, upper limit of normal; US, ultrasound.a Cardiovascular disease history: one stroke.b Cancer history: adrenocortical cancer (1); tongue parakeratosis (1); mediastinal paraganglioma (1); prostate (1). Open table in a new tab Supplementary Table 2Characteristics Associated With Steatosis and Fibrosis, Assessed Using FibroScanLogistic regression - UnivariableLogistic regression - Multivariable (no selection)Logistic regression - Multivariable (selection)Event: steatosis status (33/104)VariablesOR (95% CI)POR (95% CI)POR (95% CI)PAge (+ 1 SD)2.67 (1.70–4.44)< .00011.32 (0.68–2.54).40––Sex (women/men)0.33 (0.14–0.77).0110.69 (0.21–2.33).54––BMI (+1 SD)4.54 (2.54–9.34)< .00013.68 (1.90–8.06).00034.71 (2.63–9.65)< .0001Excessive alcohol consumption (yes/no)7.00 (0.86–145).0987.86 (0.36–289).20––Diabetes mellitus (any/none)6.60 (2.15–23.0).00151.56 (0.22–9.92).64––Triglyceride levels (+1 SD)1.86 (1.20–3.19).0121.37 (0.74–2.96).34––ACMG IV or V mutation on APOB (yes/no)3.83 (1.57–9.59).00345.56 (1.64–21.5).00785.26 (1.81–16.6).0030Event: fibrosis status (15/104)VariablesOR (95% CI)POR (95% CI)POR (95% CI)PAge (+ 1 SD)4.00 (2.23–8.01)< .00013.22 (1.60–7.14).00174.00 (2.23–8.01)< .0001Sex (women/men)0.36 (0.10–1.09).0791.01 (0.22–5.09).99––BMI (+1 SD)2.23 (1.36–3.83)< .00011.19 (0.58–2.43).63––Excessive alcohol consumption (yes/no)6.69 (0.75–59.9).0683.31 (0.28–40.7).32––Diabetes mellitus (any/none)7.78 (2.27–27.2).00112.65 (0.43–15.0).28––Triglyceride levels (+1 SD)1.70 (1.09–2.76).0201.07 (0.56–1.94).83––ACMG IV or V mutation on APOB (yes/no)1.81 (0.55–5.56).311.03 (0.20–4.60).97––Note: Univariable and multivariable logistic regression models.Note: P-values were calculated using Wald test.Note: Quantitative parameters were standardized and expressed as standard deviation (SD). Note: One SD = 16.4 years for age, 5.2 kg/m2 for BMI, 41 mg/dL for triglycerides.Note: Regarding multivariable models, first, all variables were chosen based on background knowledge and put in a single multivariable logistic regression model. Second, a stepwsise backward/forward selection was applied using a P-value selection threshold at 0.05.BMI, Body mass index; CI, confidence interval; OR, odds ratio; SD, standard deviation. Open table in a new tab Note: Categorical parameters are described using number (%). Quantitative parameters are presented using mean ± standard deviation or median (interquartile range), according to distribution. Note: Liver steatosis is considered when CAP >275 dB/m measured by FibroScan. Polygenic hypocholesterolemia is defined by a polygenic risk score (PRS) <10th percentile. ACGM, Amplified consensus genetic marker; ALT, alanine transaminase; AST, aspartate aminotransferase; BMI, body mass index; CAP, controlled attenuation parameter; eGFR, estimated glomular filtration rate; GGT, gamma-glutamyl transferase; HbA1C, hemoglobin A1C; HDL-c, high-density lipoprotein cholesterol; HOMA-IR, homeostasis model assessment of insulin resistance; LDL-c, low-density lipoprotein cholesterol; MCV, mean corpuscular volume; NR, not relevant; ULN, upper limit of normal; US, ultrasound. Note: Univariable and multivariable logistic regression models. Note: P-values were calculated using Wald test. Note: Quantitative parameters were standardized and expressed as standard deviation (SD). Note: One SD = 16.4 years for age, 5.2 kg/m2 for BMI, 41 mg/dL for triglycerides. Note: Regarding multivariable models, first, all variables were chosen based on background knowledge and put in a single multivariable logistic regression model. Second, a stepwsise backward/forward selection was applied using a P-value selection threshold at 0.05. BMI, Body mass index; CI, confidence interval; OR, odds ratio; SD, standard deviation. Two hundred and one adults with primary hypobetalipoproteinemia (HBL), defined by an low-density lipoprotein cholesterol (LDL-C) value <5th percentile for age and sex, without any known secondary cause of hypocholesterolemia (hyperthyroidism, severe undernutrition, vegan diet, severe liver failure with hepatic insufficiency, and hemoglobinopathies) were recruited in the University Hospital Centers of Nantes (HYPOCHOL study, NCT02354079) and Lyon (GENELIP study, NCT03939039). Written informed consent was obtained from each participant in accordance with the principles of the Declaration of Helsinki, the French bioethics law, and protocols approved by the University Hospital Centers of Nantes and Lyon. Two patients were excluded due to lipid-lowering therapy, 86 patients were excluded due to the lack of FibroScan evaluation, and 9 patients were excluded because the FibroScan evaluation was performed more than 5 years after the blood tests. Ultimately, 104 patients were included in the study analysis. All co-authors had access to the study data and had reviewed and approved the final manuscript. During the inclusion visit, clinical research associates and physicians collected data including anthropometric data, medical history of diabetes, hypertension, smoking, excessive alcohol consumption or other substance abuse, or cancer. Excessive alcohol consumption was defined as daily alcohol intake ≥20 g/day in women and ≥30 g/day in men. Fasting blood samples were drawn from all participants for genetic testing and biochemical measurements, including: lipid profile with total cholesterol, triglycerides, high-density lipoprotein cholesterol (HDL-C), calculated LDL-C according to Friedewald formulae, apoB, Lp(a); liver enzymes: aspartate aminotransferase (AST), alanine aminotransferase (ALT), gamma-glutamyl transferase (GGT), PAL, bilirubin; fasting plasma glucose, and hemoglobin A1c (HbA1c) were determined using dedicated commercial kits either on an Architect C16000 autoanalyzer (Abbott Diagnostics) in Lyon or on a Roche Cobas 6000 ce analyzer (Roche) in Nantes. Fibrosis-4 index (FIB-4) was calculated on the basis of age, AST, and ALT levels, and platelet count. Homeostatic Model Assessment for Insulin Resistance (HOMA-IR) was calculated with the formula: fasting insulin (μU/mL) × fasting glucose (mmol/L)/22.5. A liver ultrasound was conducted in an out-of-hospital radiology practice. FibroScan (Echosens) examination was performed by staff with certified training from the manufacturer, using M or XL probes according to the probe detection tool, and following an overnight fast. Only examinations with at least 10 valid measurements and interquartile range (IQR)/median ≤30% for liver stiffness were considered reliable. Categorical parameters are presented using size (%), with 95% confidence intervals calculated using the Clopper-Pearson exact method. Quantitative parameters are presented as either mean ± standard deviation or median (interquartile range), depending on their distribution. P-values were calculated using the Fisher exact test, Student t-test, or Mann-Whitney ranksum test, as deemed appropriate. Logistic regression models were used to calculate the odds ratios (ORs) associated with the liver steatosis and fibrosis status based on FibroScan. For quantitative parameters, ORs were expressed for one SD increase in the respective parameter. Multiple logistic regression analyses were performed, incorporating parameters identified as clinically relevant based on background knowledge. Two multivariable models are presented, one including all the variables, and one after backward/forward stepwise selection, applying a P-value selection threshold of .05. All analyses were performed on available data, without imputation. Two-sided P-values were deemed significant if < .05, without accounting for multiple testing. Statistical analyses were performed using statistical software R.