Background and Aims:
The NAFLD activity score was developed to measure histologic changes in NAFLD during therapeutic trials. Hepatocyte ballooning (HB) is the most specific feature in steatohepatitis diagnosis, yet the impact of variations in HB has not been incorporated.
Approach and Results:
Liver biopsies from patients enrolled in the NASH Clinical Research Network with an initial diagnosis of NASH or NAFL (n=1688) were evaluated to distinguish classic hepatocyte ballooning (cHB) from smaller, nonclassic hepatocyte ballooning (nHB), and also to designate severe ballooning and assign an extended hepatocyte ballooning (eB) score [0 points, no ballooning (NB); 1 point, few or many nHB; 2 points, few cHB; 3 points, many cHB; 4 points, severe cHB] to the biopsy assessment. The eB score was reproducible among NASH CRN liver pathologists (weighted kappa 0.76) and was significantly associated with older age (mean 52.1 y, cHB; 48.5 y, nHB, p<0.001), gender (72.3% female, cHB; 54.5% female, nHB, p<0.001), diabetes (49.8% diabetes, cHB; 28.2% diabetes, nHB, p<0.001), metabolic syndrome (68.5% metabolic syndrome, nHB; 50.2% metabolic syndrome, NB, p<0.001), and body mass index [33.2, 34.2, 35 mean body mass index (kg/m2); NB, nHB, and cHB, respectively, p<0.05]. Finally, fibrosis stage, as a marker of disease severity, was significantly correlated with the eB score (p<0.001).
Conclusions:
The eB score allows for a reproducible and more precise delineation of the range of ballooned hepatocyte morphology and corresponds with both clinical features of NASH and fibrosis stage.
NAFLD is prevalent in the United States1,2 and is comprised of 2 broad histopathologic categories, both of which include fat and may include inflammation: NASH and NAFL. The former additionally includes hepatocellular injury, commonly in the form of hepatocyte ballooning (HB), which is required for diagnosis.3–9 Liver biopsy to distinguish between NAFL and NASH is important since a significant subset of NASH patients will develop fibrosis and progress to cirrhosis and/or HCC,3 and there are no clinical or imaging features that can reliably make the distinction. NASH risk factors include obesity, type 2 diabetes, hyperlipidemia, and metabolic syndrome.3 Ballooned hepatocytes are the most specific single histologic criteria for diagnosis of NASH and are generally recognized as large hepatocytes with rarefied and clumped cytoplasm, with or without Mallory-Denk bodies,10 but the features of HB occur on a morphologic spectrum, and the current categorical scoring systems used to describe HB are relatively broad, with each category including a spectrum of severity. An extended ballooning score may facilitate the detection of meaningful HB changes that are currently not conveyed because they are assigned the same grade.
The NAFLD activity score (NAS) incorporates histologic findings into a scoring system that can be used to follow response to treatment in therapeutic trials11 and is distinct from a NASH diagnosis, which is based on the pathologist’s integrative interpretation of the constellation of histologic features. The NAS is comprised of steatosis (3-point range), lobular inflammation (3-point range), and HB (2-point range). Although higher NAS scores correlate with a diagnosis of NASH, the score is not used to establish a diagnosis of NASH.11 Ballooned hepatocytes are currently classified as none (score 0), or few (score 1) versus many (score 2), for a maximum score of 2 points using the NASH Clinical Research Network (CRN) system.
Over the course of central review of NAFLD biopsies by the NASH CRN Pathology Committee, we recognized a spectrum of HB morphology. Biopsies that caused the most debate were those with “nonclassic” forms of HB, and it was agreed that prospective categorization was needed to determine if such subtle changes correlate with diagnosis and/or histologic or clinical features of disease severity.
Presented here are the results of a prospective qualitative and semiquantitative evaluation of HB in a large cohort of NAFLD patients within the NASH CRN for determination of relevance to final histologic diagnosis and association with clinical, laboratory, and histologic parameters of NASH. Following our initial report of this approach to classification of HB,12 utilization of this terminology has been separately validated, with an inter-rater intraclass correlation coefficient for HB (0.68–0.79).13
EXPERIMENTAL PROCEDURES
Study design and population
The NASH CRN has enrolled patients with NAFLD from multiple centers in the United States, beginning in 2002, as part of a prospective effort to elucidate the etiology, natural history, diagnosis, treatment, and prevention of NAFLD/NASH. Data for this study were extracted from subjects enrolled from August 2004 through December 2016 for cross sectional analysis from NAFLD Adult database 2 (DB2) (ClinicalTrials.gov identifier: NCT01030484), the Pioglitazone versus Vitamin E versus Placebo for Treatment of Non-Diabetic Patients With Nonalcoholic Steatohepatitis (PIVENS)14 (ClinicalTrials.gov identifier: NCT00063622), and the Farnesoid X Receptor (FXR) Ligand Obeticholic Acid in NASH Treatment Trial (FLINT)15 (ClinicalTrials.gov identifier: NCT01265498) studies. In brief, DB2 inclusion criteria specified adult patients (at least 18 y of age), with minimal to no alcohol use and no other liver disease, who had a standard of care liver biopsy within 120 days of enrollment as well as collection of other biosamples within 90 days of enrollment. PIVENS was a randomized, placebo-controlled, double-blind clinical trial in adults with NASH but without diabetes or cirrhosis in which pioglitazone (30 mg daily) or vitamin E (800 IU daily) was administered for 96 weeks to evaluate for improvement in liver disease as measured by decrease in ballooning score by 1 or more points and with no increase in fibrosis, as well as decrease in NAS to 3 or less, or decrease in NAS by 2 points (with 1 point reduction in either lobular inflammation or steatosis score).14 FLINT was a randomized double-blind interventional treatment trial in adults with noncirrhotic NASH, in which obeticholic acid was administered for 72 weeks to adults with NASH and NAS of at least 4, to evaluate for improvement in liver disease as measured by reduction in NAS of at least 2 points without worsening of fibrosis.15 These studies had Institutional Review Board approval at each NASH CRN clinical center and the data coordinating center (Supplemental Appendix, http://links.lww.com/HC9/A88) and written informed consent from each patient conforming to the ethical guidelines of the 2013 Declaration of Helsinki and 2018 Declaration of Istanbul. For this analysis, biopsy data from 1688 biopsies from subjects/unique participants in DB2, PIVENS, and FLINT were included (patients without evidence of NAFLD on biopsy were excluded); 1203 were from DB2, which consisted of 1081 at baseline and 122 from a subsequent biopsy, 240 from PIVENS (all at baseline), and 245 from FLINT (all at baseline).
Liver histology
NASH CRN Pathology Committee met regularly, in person, as a group and reviewed all biopsies via a multihead microscope to establish a consensus diagnosis and to score the histologic parameters as previously described and validated.11,16 Diagnostic categories, established independently from the NAS, included “not NAFLD not NASH,” “NAFLD not NASH,” “Borderline NASH, zone 1 pattern,” “Borderline NASH, zone 3 pattern,” and “Definite NASH.”6,11,16 Ballooned hepatocytes were originally classified as “none,” “few,” or “many.” Starting in 2010, the Pathology Committee executed a plan to prospectively subclassify HB in all future biopsies (Table 1) according to the designations that are now incorporated into the extended ballooning (eB) score, in which HB was designated as nonclassic (nHB) or classic (cHB); cHB was further designated as “severe” when observed at low power, or “not severe” when not seen on immediate low power evaluation; “few” versus “many” descriptors were still recorded (with “many” indicating >4 ballooned hepatocytes in a biopsy in separate foci) and these data allowed for calculation of an eB score. The eB score, with a range of 0–4 points, replaces the previous HB score (range: 0–2) and extends the NAFL activity score to allow for a new extended NAS (ie, eNAS), with a range of 0–10 points. cHB designation required majority NASH CRN pathologist agreement during central review: cHB were characteristically enlarged (generally >1.5× the diameter of adjacent hepatocytes) with irregularly clumped cytoplasm (sometimes with Mallory-Denk bodies) with rarefaction of cytoplasm, which were easily distinguished from most other hepatocytes. In contrast, nHB generally did not show the degree of enlargement of cHB, but showed similar cytoplasmic changes, and lacked well-formed Mallory-Denk bodies (Figure 1). A typical case of severe HB would have multiple cHB found together around the terminal hepatic venule, and several such foci would be visible from low magnification (Figure 1). Other routinely collected histologic data included steatosis grade, lobular inflammation score, portal inflammation amount, and fibrosis stage as previously described,16 as well as presence of glycogenosis, microvesicular steatosis, acidophil bodies, megamitochondria, and Mallory-Denk bodies and these data were extracted from collected data sheets for this analysis.
TABLE 1.
Definitions of the extended hepatocyte ballooning (eB) scorea
| Extended hepatocyte ballooning (eB) score | Assessment of hepatocyte ballooning (grading in original score) | Existence of classical hepatocyte ballooning | Presence of severe hepatocyte ballooning |
|---|---|---|---|
| 0=No ballooning | None (score 0) | No | No |
| 1=Only nonclassic | Few or many (score 1 or 2) | No | No |
| 2=Few classic | Few (score 1) | Yes | No |
| 3=Many classic | Many (score 2) | Yes | No |
| 4=Severe, classicb | Many (score 2) | Yes | Yes |
NASH CRN ballooning score (BS): 0=no hepatocyte ballooning; 1=few hepatocyte ballooned cells; 2=many hepatocyte ballooned cells.
By definition, presence of severe hepatocyte ballooning includes many classic hepatocyte ballooned cells.
Abbreviation: CRN, Clinical Research Network.
FIGURE 1.
Classic and nonclassic hepatocyte ballooning. (A) Example of severe classic hepatocyte ballooning (eg, arrowheads), ×200 magnification, hematoxylin and eosin (H&E) stain. (B) Example of classic (arrowhead) and nonclassic (arrow) hepatocyte ballooning, ×200 magnification, H&E stain. (C) Example of nonclassic hepatocyte ballooning (arrow), ×400 magnification, H&E stain. (D) Another example of nonclassic hepatocyte ballooning (arrow), ×400 magnification, H&E stain.
Clinical and laboratory variables
Demographic, anthropometric, and laboratory data were abstracted for analysis. Clinical data were obtained prospectively from records review, physical exams, interviews, and questionnaires, as previously described,17 and include calculation of body mass index and diagnosis of type 2 diabetes and/or metabolic syndrome. Laboratory data included alanine aminotransferase (U/L), alkaline phosphatase (U/L), aspartate aminotransferase (U/L), fasting glucose (mg/dL), fasting insulin (mU/L), fasting triglycerides (mg/dL), and HDL-cholesterol (mg/dL), as well as the calculated homeostasis model assessment of insulin resistance (glucose × insulin/405).
Statistical analyses
Patient characteristics (demographic, clinical, laboratory, and histological) were compared between those with no hepatocyte ballooning (NB) and those with nHB, between those with and without cHB among those with HB, between those with nonsevere and severe HB among those with HB, and by extended ballooning score. χ2 test and Student t test were performed for hypothesis testing for nominal and continuous variables of the patient characteristics, respectively. In order to assess the difference in a patient characteristic across HB score and NAS, we calculated p values from hypothesis testing (χ2 test for nominal variables and ANOVA for continuous variable). To assess the observer variability of the eB scoring system, a nonrandom sample of biopsies were re-read in a masked fashion by the pathology committee at a separate session. Repeatability was evaluated using the weighted kappa statistic with Cicchetti-Allison weights. Two-sided p values (<0.05) were considered significant. Analyses were performed using Stata statistical software (version 14.2).
RESULTS
Patient characteristics
For the 1688 adult patients’ biopsies included in this study the average age was 50.3 years and most were women (62.8%) and White (73.2%). Mean body mass index was 34.3 (range: 18.7–61.5) kg/m2. Type 2 diabetes and metabolic syndrome were present in 38.1% and 62.6%, respectively.
Demographic and clinical characteristics associated with HB categories and extended HB score
cHB (eB scores 2, 3, and 4) was more commonly noted in biopsies from older subjects, women, and subjects with diabetes (Table 2). nHB was also associated with higher body mass index and metabolic syndrome (Table 2). Older age, female gender, increased body mass index, diabetes, and metabolic syndrome were all significantly associated with increased extended ballooning (eB) score (Table 3). Serum aspartate aminotransferase, alanine aminotransferase, fasting glucose, fasting insulin, alkaline phosphatase, and homeostasis model assessment of insulin resistance were all significantly elevated in patients with both nonsevere and severe cHB (Table 2) and increased eB score (Table 3). Serum levels of HDL-cholesterol were also significantly increased in patients with nonsevere cHB (Table 2) compared to those with only nHB. Aspartate aminotransferase, alanine aminotransferase, alkaline phosphatase, fasting insulin, homeostasis model assessment of insulin resistance, and HDL were also significantly elevated in patients with nHB (Table 2), compared to those in whom biopsies were without HB. The original NASH CRN ballooning score was significantly associated with the same demographic and clinical characteristics, and laboratory measures, as the eB score (Table 3). There was substantial agreement in classification of HB when cases were rescored by NASH CRN pathologist consensus (in which pathologists were blinded as to whether the case had been reviewed previously), which allowed for a reproducible eB score [weighted Kappa=0.76 (95% CI=0.64–0.88)] (Table 4).
TABLE 2.
Demographic and clinical characteristics, and laboratory measures by hepatocyte ballooning categoriesa
| Those with no hepatocyte ballooning versus those with nonclassic hepatocyte ballooning | Presence of classic hepatocyte ballooning among those with hepatocyte ballooning | Presence of severe hepatocyte ballooning among those with hepatocyte ballooning | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Total (N=1688) | N | None | Nonclassical | p b | N | Absent | Present | p b | N | Absent | Present | p b | |
| Extended ballooning score | 0 | 1 | 1 | 2, 3, 4 | 1, 2, 3 | 4 | |||||||
| Number of biopsy samples | 1688 | 858 | 550 | 308 | 1138 | 308 | 830 | 1138 | 897 | 241 | |||
| Demographics | |||||||||||||
| Age (year, mean) | 50.3 (12.2) | 858 | 48.6 | 48.5 | 0.93 | 1138 | 48.5 | 52.1 | <0.001 | 1138 | 50.8 | 52.3 | 0.09 |
| Gender (% female) | 1060 (62.8%) | 858 | 53.1 | 54.5 | 0.68 | 1138 | 54.5 | 72.3 | <0.001 | 1138 | 64.0 | 80.5 | <0.001 |
| Race (%) | 856 | 0.10 | 1135 | 0.56 | 0.15 | ||||||||
| Non-Hispanic White | 1233 (73.1%) | 71.4 | 75.0 | 75.0 | 73.9 | 1135 | 74.3 | 73.6 | |||||
| Non-Hispanic Black | 62 (3.7%) | 3.6 | 2.6 | 2.6 | 4.1 | 1135 | 3.5 | 4.6 | |||||
| Hispanic | 211 (12.5%) | 12.4 | 11.4 | 11.4 | 13.1 | 1135 | 12.6 | 12.6 | |||||
| Asian | 100 (5.9%) | 9.1 | 5.2 | 5.2 | 4.1 | 1135 | 5.0 | 2.1 | |||||
| Otherc | 82 (4.9%) | 3.5 | 5.8 | 5.8 | 4.8 | 1135 | 4.6 | 7.1 | |||||
| Clinical characteristicsd | |||||||||||||
| BMI (kg/m2, mean) | 34.3 (6.4) | 853 | 33.2 | 34.2 | 0.02 | 1135 | 34.2 | 35.0 | 0.05 | 1135 | 34.5 | 35.8 | 0.006 |
| Diabetes (%) | 640 (37.9%) | 858 | 25.5 | 28.2 | 0.37 | 1138 | 28.2 | 49.8 | <0.001 | 1138 | 41.1 | 54.4 | <0.001 |
| Metabolic syndromee (%) | 1046 (62.6%) | 849 | 50.2 | 68.5 | <0.001 | 1126 | 68.5 | 68.7 | 0.96 | 1126 | 67.5 | 72.9 | 0.11 |
| Laboratory measuresd (mean) | |||||||||||||
| AST (U/L) | 50.9 (33.1) | 856 | 37.2 | 42.0 | 0.005 | 1137 | 42.0 | 63.3 | <0.001 | 1137 | 52.8 | 75.2 | <0.001 |
| ALT (U/L) | 70.5 (50.4) | 856 | 56.2 | 64.3 | 0.006 | 1137 | 64.3 | 82.2 | <0.001 | 1137 | 74.6 | 87.9 | <0.001 |
| Alkaline phosphate (U/L) | 81.4 (30.6) | 855 | 75.3 | 79.1 | 0.05 | 1137 | 79.1 | 86.2 | 0.001 | 1137 | 82.3 | 91.7 | <0.001 |
| Glucose (mg/dL) | 109.6 (38.5) | 855 | 102.7 | 103.5 | 0.73 | 1136 | 103.5 | 116.4 | <0.001 | 1136 | 110.5 | 121.8 | <0.001 |
| Insulin (mU/L) | 259.8 (290.8) | 851 | 20.2 | 24.3 | 0.003 | 1129 | 24.3 | 30.5 | 0.005 | 1129 | 27.3 | 34.3 | 0.003 |
| HOMA-IR | 74.8 (101.8) | 851 | 5.3 | 6.6 | 0.008 | 1129 | 6.6 | 9.3 | <0.001 | 1129 | 7.9 | 11.1 | <0.001 |
| HDL-cholesterol (mg/dL) | 43.8 (12.6) | 854 | 44.1 | 42.2 | 0.03 | 1134 | 42.2 | 44.1 | 0.02 | 1134 | 43.6 | 43.5 | 0.90 |
| Triglycerides (mg/dL) | 178.8 (153.4) | 855 | 173.8 | 182.3 | 0.43 | 1136 | 182.3 | 180.9 | 0.89 | 1136 | 183.9 | 171.5 | 0.25 |
| Ferritin (ng/mL) | 213.0 (237.0) | 854 | 209.0 | 197.5 | 0.55 | 854 | 197.8 | 220.9 | 0.22 | 854 | 214.9 | 215.1 | 0.99 |
Among those with NAFLD. A total 1688 biopsies were included in this analysis: 1203 from the observational database study (1081 at baseline and 122 at subsequent visit); 240 from the clinical trial PIVENS study (all at baseline); and 245 from the clinical trial FLINT study (all at baseline). Biopsy samples were taken between August 2004 and December 2016. When more than 1 biopsy was available per subject, the baseline biopsy was from patients in treatment trials.
p values were calculated using the Student t test for difference in means for continuous variables and using χ2 test for difference in proportions for categorical variables.
Other race includes American Indian, Alaska native, Pacific Islanders, and multirace.
Clinical characteristics and laboratory measures obtained within 6 months of the biopsy were used. Due to incomplete measurements, the sample used in their assessment was smaller or equal to the total number of biopsy samples that were assessed in the study.
Metabolic syndrome defined as having the 3 conditions or more of the following 5 criteria: waist circumference >102 cm in male and >88 cm in female; triglycerides ≥150 mg/dL; HDL-cholesterol <40 mg/dL in male and <50 mg/dL in female; systolic blood pressure ≥130 mm Hg or diastolic blood pressure ≥85 mm Hg; fasting glucose ≥110 mg/dL.
Abbreviatoins: ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body mass index; FLINT, farnesoid X Receptor Ligand Obeticholic Acid in NASH Treatment; HOMA-IR, homeostasis model assessment of insulin resistance; PIVENS, pioglitazone versus Vitamin E versus Placebo for the Treatment of Nondiabetic Patients with Nonalcoholic Steatohepatitis.
TABLE 3.
Demographic and clinical characteristics, and laboratory measures by NASH CRN and extended hepatocyte ballooning scorea
| NASH CRN hepatocyte ballooning score (BS) | Extended hepatocyte ballooning (eB) score | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| N | 0 | 1 | 2 | p b | 0 | 1 | 2 | 3 | 4 | p b | |
| N | 1688 | 550 | 503 | 635 | 550 | 308 | 250 | 339 | 241 | ||
| Demographics | |||||||||||
| Age (year, mean) | 1688 | 48.6 | 50.1 | 52.0 | <0.001 | 48.6 | 48.5 | 51.6 | 52.4 | 52.3 | <0.001 |
| Gender (% female) | 1688 | 53.1 | 57.7 | 75.3 | <0.001 | 53.1 | 54.5 | 62.0 | 74.0 | 80.5 | <0.001 |
| Race (%) | 1683 | 0.02 | 0.03 | ||||||||
| Non-Hispanic White | 71.4 | 73.0 | 75.2 | 71.4 | 75.0 | 71.2 | 76.0 | 73.6 | |||
| Non-Hispanic Black | 3.6 | 3.6 | 3.8 | 3.6 | 2.6 | 4.4 | 3.6 | 4.6 | |||
| Hispanic | 12.4 | 13.5 | 11.9 | 12.4 | 11.4 | 15.2 | 11.8 | 12.6 | |||
| Asian | 9.1 | 5.2 | 3.8 | 9.1 | 5.2 | 5.2 | 4.7 | 2.1 | |||
| Otherc | 3.5 | 4.8 | 5.4 | 3.5 | 5.8 | 4.0 | 3.8 | 7.1 | |||
| Clinical characteristicsd | |||||||||||
| BMI (kg/m2, mean) | 1682 | 33.2 | 34.2 | 35.3 | <0.001 | 33.2 | 34.2 | 34.4 | 34.9 | 35.8 | <0.001 |
| Diabetes (%) | 1688 | 25.5 | 37.8 | 48.8 | <0.001 | 25.5 | 28.2 | 44.8 | 50.1 | 54.4 | <0.001 |
| Metabolic syndromee (%) | 1670 | 50.2 | 67.1 | 69.8 | <0.001 | 50.2 | 68.5 | 64.6 | 68.7 | 72.9 | <0.001 |
| Laboratory measuresd (mean) | |||||||||||
| AST (U/L) | 1685 | 37.2 | 44.5 | 67.8 | <0.001 | 37.2 | 42.0 | 49.0 | 65.4 | 75.2 | <0.001 |
| ALT (U/L) | 1685 | 56.2 | 65.8 | 86.5 | <0.001 | 56.2 | 64.3 | 70.0 | 87.2 | 87.9 | <0.001 |
| Alkaline phosphate (U/L) | 1684 | 75.3 | 80.1 | 87.6 | <0.001 | 75.3 | 79.1 | 81.1 | 86.1 | 91.7 | <0.001 |
| Glucose (mg/dL) | 1683 | 102.7 | 107.8 | 116.9 | <0.001 | 102.7 | 103.5 | 111.9 | 115.8 | 121.8 | <0.001 |
| Insulin (mU/L) | 1675 | 20.2 | 24.8 | 32.0 | <0.001 | 20.2 | 24.3 | 25.5 | 31.4 | 34.3 | <0.001 |
| HOMA-IR | 1675 | 5.3 | 7.0 | 9.8 | <0.001 | 5.3 | 6.6 | 7.4 | 9.4 | 11.1 | <0.001 |
| HDL-cholesterol (mg/dL) | 1681 | 44.1 | 43.3 | 43.8 | 0.58 | 44.1 | 42.2 | 44.6 | 44.2 | 43.5 | 0.15 |
| Triglycerides (mg/dL) | 1683 | 173.8 | 184.6 | 178.6 | 0.52 | 173.8 | 182.3 | 183.8 | 185.4 | 171.5 | 0.70 |
Among those with NAFLD. A total 1688 biopsies were included in this analysis. See the footnote of Table 2 for details.
p values were calculated using the Student t test for difference in means for continuous variables and using χ2 test for difference in proportions for categorical variables.
Other race includes American Indian, Alaska native, Pacific Islanders, and multirace.
Clinical characteristics and laboratory measures obtained within 6 months from the biopsy collection were used. Due to incomplete measurements, the sample used in their assessment was smaller or equal to the total number of biopsy samples that were assessed in the study.
For the definition of metabolic syndrome, see the footnote in Table 2.
Abbreviations: AL, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body mass index; CRN, Clinical Research Network; HDL, high-density lipoprotein; HOMA-IR, homeostasis model assessment of insulin resistance.
TABLE 4.
Interobserver agreement in hepatocyte ballooning classification using repeated readingsa
| Second reading | ||||||
|---|---|---|---|---|---|---|
| First reading | 0 (none) | 1 (only nHB) | 2 (few cHB) | 3 (many cHB) | 4 (severe cHB) | Total |
| 0 (none) | 36 | 3 | 0 | 0 | 0 | 39 |
| 1 (only nHB) | 7 | 4 | 1 | 1 | 0 | 13 |
| 2 (few cHB) | 0 | 0 | 3 | 1 | 0 | 4 |
| 3 (many cHB) | 0 | 2 | 1 | 2 | 0 | 5 |
| 4 (severe cHB) | 0 | 0 | 0 | 1 | 5 | 6 |
| Total | 43 | 9 | 5 | 5 | 5 | 67 |
Sixty-seven biopsies (from patients enrolled in DB2 and FLINT) had repeated scoring of hepatocyte ballooning over the course of routine consensus evaluation by the NASH CRN pathology committee. Kappa=0.57 (95% CI=0.40–0.73), Weighted kappa=0.76 (95% CI=0.64–0.88).
Abbreviations: cHB, classic hepatocyte ballooning; CRN, Clinical Research Network; DB2, database 2; FLINT, farnesoid X Receptor Ligand Obeticholic Acid in NASH Treatment; nHB, nonclassic hepatocyte ballooning. Note: Bland-Atman 95% limits of agreement=±1.2 (interpretation: 95% of repeated readings are within 1.2 categories).
Comparison of histologic features in NAFLD biopsies with cHB to biopsies with nHB
cHB was seen in 72.9% of all biopsies with HB. Diagnostic categories were significantly different between biopsies with cHB and those with nHB. Nearly all cases with cHB were given a consensus diagnosis of “definite steatohepatitis,” whereas, even accounting for the small total number of cases, a higher percentage of “borderline steatohepatitis” diagnoses were associated with nHB (28.6% for nHB vs. 2.2% for cHB) (Table 5). Fibrosis stage was significantly higher in biopsies with cHB than in those with nHB biopsies; 45.9% of biopsies with cHB also had advanced fibrosis (stage 3 or 4 fibrosis), while 16.9% of biopsies with nHB had stages 3 or 4 fibrosis (Table 5). Inflammation, both lobular and portal, was more evident in biopsies with cHB than those with nHB (Table 5). Other histologic findings more commonly encountered in the setting of cHB than nHB included foci of microvesicular steatosis, acidophil bodies, and Mallory-Denk bodies. Steatosis grade was not significantly different between biopsies with cHB and those with nHB (Table 5).
TABLE 5.
Histological characteristics by hepatocyte ballooning categoriesa
| No hepatocyte ballooning versus nonclassic hepatocyte ballooning | Classic hepatocyte ballooning among those with hepatocyte ballooning | Severe hepatocyte ballooning among those with hepatocyte ballooning | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| N | None | Nonclassic | p b | N | Absent | Present | p b | N | Absent | Present | p b | |
| Extended ballooning score | 0 | 1 | 1 | 2, 3, 4 | 1, 2, 3 | 4 | ||||||
| Number of biopsy samples | 858 | 550 | 308 | 1138 | 308 | 830 | 1138 | 897 | 241 | |||
| Diagnostic category (%) | 858 | <0.001 | 1138 | <0.001 | 1138 | <0.001 | ||||||
| NAFLD, not NASH | 63.6 | 1.9 | 1.9 | 0.2 | 0.9 | 0.0 | ||||||
| Borderline NASH, Z3 pattern | 33.6 | 28.6 | 28.6 | 2.2 | 11.7 | 0.4 | ||||||
| Borderline NASH, Z1 pattern | 2.5 | 1.9 | 1.9 | 0.2 | 0.8 | 0.4 | ||||||
| Definite NASH | 0.2 | 67.5 | 67.5 | 97.3 | 86.6 | 99.2 | ||||||
| Fibrosis stage (%) | 858 | <0.001 | 1138 | <0.001 | 1138 | <0.001 | ||||||
| 0 | 52.7 | 25.3 | 25.3 | 2.5 | 10.9 | 0.4 | ||||||
| 1A | 16.5 | 20.8 | 20.8 | 8.1 | 14.3 | 1.2 | ||||||
| 1B | 5.8 | 9.1 | 9.1 | 19.5 | 17.1 | 15.4 | ||||||
| 1C | 5.5 | 5.2 | 5.2 | 0.5 | 2.2 | 0.0 | ||||||
| 2 | 12.4 | 22.7 | 22.7 | 23.5 | 24.5 | 18.7 | ||||||
| 3 | 5.8 | 12.7 | 12.7 | 34.2 | 23.2 | 47.7 | ||||||
| 4 | 1.3 | 4.2 | 4.2 | 11.7 | 7.8 | 16.6 | ||||||
| Steatosis grade (%) | 858 | <0.001 | 1138 | 0.30 | 1138 | 0.17 | ||||||
| <5 | 0.0 | 2.3 | 2.3 | 3.4 | 2.9 | 3.7 | ||||||
| 5–33 | 45.3 | 29.9 | 29.9 | 34.6 | 33.1 | 34.0 | ||||||
| 34–66 | 32.5 | 37.7 | 37.7 | 35.2 | 34.8 | 39.8 | ||||||
| >66 | 22.2 | 30.2 | 30.2 | 26.9 | 29.2 | 22.4 | ||||||
| Lobular inflammation (%) | 858 | <0.001 | 1138 | <0.001 | 1138 | <0.001 | ||||||
| <2/hpf | 76.4 | 60.4 | 60.4 | 32.7 | 45.8 | 19.1 | ||||||
| 2–4/hpf | 20.9 | 35.7 | 35.7 | 47.1 | 44.0 | 44.0 | ||||||
| >4/hpf | 2.7 | 3.9 | 3.9 | 20.2 | 10.1 | 36.9 | ||||||
| Portal inflammation (%) | 857 | <0.001 | 1137 | <0.001 | 1137 | <0.001 | ||||||
| None | 19.5 | 12.4 | 12.4 | 7.5 | 9.9 | 4.6 | ||||||
| Mild | 67.8 | 64.5 | 64.5 | 57.3 | 61.7 | 50.2 | ||||||
| More than mild | 12.7 | 23.1 | 23.1 | 35.2 | 28.3 | 45.2 | ||||||
| Glycogenosis (%) | 726 | 0.05 | 949 | 0.86 | 949 | 0.001 | ||||||
| None | 52.5 | 42.7 | 42.7 | 40.9 | 44.0 | 31.9 | ||||||
| Focal | 23.2 | 27.8 | 27.8 | 29.4 | 29.0 | 29.0 | ||||||
| Diffuse | 24.2 | 29.5 | 29.5 | 29.8 | 27.1 | 39.0 | ||||||
| Microvesicular steatosis (%) | 858 | 3.3 | 12.3 | <0.001 | 1138 | 12.3 | 18.8 | 0.01 | 1138 | 15.3 | 23.7 | 0.002 |
| Acidophil bodies (%) | 858 | 31.1 | 44.5 | <0.001 | 1138 | 44.5 | 57.2 | <0.001 | 1138 | 52.1 | 60.2 | 0.03 |
| Megamitochondria (%) | 858 | 18.2 | 28.6 | <0.001 | 1138 | 28.6 | 33.0 | 0.15 | 1138 | 30.8 | 35.7 | 0.15 |
| Mallory-Denk bodies (%) | 858 | 0.0 | 2.6 | <0.001 | 1138 | 2.6 | 65.5 | <0.001 | 1138 | 35.7 | 96.3 | <0.001 |
Among those with NAFLD. A total 1688 biopsies were included in this analysis. See the footnote of Table 2 for details.
p values were calculated using χ2 test for difference in proportions.
Abbreviation: hpf, high power field.
Comparison of histologic features in NAFLD biopsies with nHB to biopsies with NB
nHB accounted for 27.1% of all biopsies with any HB. Compared to biopsies with NB, and similar to biopsies with cHB, diagnostic categories were significantly different between nHB and NB cases; 96.1% of biopsies with nHB were given a consensus diagnosis of either “definite steatohepatitis,” (67.5%) or “borderline steatohepatitis, zone 3 pattern” (28.6%), whereas the biopsies with NB were most commonly given a consensus diagnosis of either “NAFLD, not NASH” (63.6%) or “borderline steatohepatitis, zone 3 pattern” (28.6%) (Table 5). Fibrosis stages were also significantly higher in nHB than in NB biopsies. nHB biopsies had fibrosis (stages 1–4) in 74.7% of biopsies while the majority (52.7%) of NB biopsies showed no fibrosis (stage 0) (Table 5). Steatosis grade was significantly higher in nHB than NB biopsies. Seven biopsies with nHB had steatosis <5% (ie, grade 0 steatosis). Inflammation, both lobular and portal, was also more evident in nHB than NB biopsies; nearly twice as many nHB cases demonstrated more than mild portal inflammation than NB biopsies (23.1% vs. 12.7%, respectively); the lowest lobular inflammation score (<2/high power field) was more common in NB than nHB biopsies (76.4% vs. 60.4%). Other histologic findings more commonly encountered simultaneously with nHB, versus those with NB, included foci of microvesicular steatosis, acidophil bodies, Mallory-Denk bodies, and megamitochondria.
Comparison of histologic features in NAFLD cases with severe versus not severe cHB
Severe cHB was present in 21.1% of biopsies with HB and 29% of biopsies with cHB. Biopsies with severe cHB were almost always categorized as “definite steatohepatitis” (99.2%) whereas this diagnosis was made in biopsies with non-severe HB (ie, all other cases with HB) in 86.6% of biopsies (Table 5). Most biopsies with severe cHB also had advanced fibrosis (64.3% with stage 3 or 4 fibrosis); fibrosis stages were significantly higher in biopsies with severe cHB than in biopsies with non-severe HB (31% with stage 3 or 4 fibrosis) (Table 5). In fact, only 1 biopsy with severe cHB had no fibrosis. Inflammation scores, both lobular and portal, were usually higher in severe cHB (36.9% lobular >4/high power field; 45.2% more than mild portal inflammation) than in biopsies with nonsevere cHB (10.1% lobular >4/high power field; 28.3% more than mild portal inflammation). Other histologic findings more commonly encountered along with severe cHB versus non-severe cHB included glycogenosis, foci of microvesicular steatosis, acidophil bodies, and Mallory-Denk bodies.
Extended ballooning score compared to histologic features
Extended ballooning score was significantly associated with consensus-determined diagnosis. The majority (99.2%) of biopsies with an eB score of 4 had also received a consensus diagnosis of “definite steatohepatitis” (Table 6). All biopsies with an eB score of 3 were diagnosed with either “definite” (98.5%) or “borderline” steatohepatitis (1.5%). Only rare biopsies with NAFL were found to have an eB score of 1 or 2 (1.9% and 0.8%, respectively). No HB was seen in only 1 case with “definite steatohepatitis.” The absence of HB was a common finding in borderline cases (33.6% of biopsies with “borderline steatohepatitis, zone 3 pattern”). Higher fibrosis stage was significantly associated with higher eB score, with an obvious stepwise progression (eg, stage 3 or 4 fibrosis: 7.1% at eB score=0; 16.9% at eB score=1; 29.6% at eB score=2; 44.8% at eB score=3; 64.3% at eB score=4). Steatosis grade and degree of inflammation (lobular and portal) were also increased along with higher eB score (Table 6). Other histologic findings more commonly encountered with higher eB score include glycogenosis, foci of microvesicular steatosis, acidophil bodies, Mallory-Denk bodies, and megamitochondria.
TABLE 6.
Histological characteristics by the extended ballooning scoresa
| Extended ballooning (eB) score | |||||||
|---|---|---|---|---|---|---|---|
| N | 0 | 1 | 2 | 3 | 4 | p b | |
| N | 1688 | 550 | 308 | 250 | 339 | 241 | |
| Original ballooning score (%) | 1688 | <0.001 | |||||
| 0 | 100.0 | 0.0 | 0.0 | 0.0 | 0.0 | ||
| 1 | 0.0 | 82.1 | 100.0 | 0.0 | 0.0 | ||
| 2 | 0.0 | 17.9 | 0.0 | 100.0 | 100.0 | ||
| Diagnostic category (%) | 1688 | <0.001 | |||||
| NAFLD, not NASH | 63.6 | 1.9 | 0.8 | 0.0 | 0.0 | ||
| Borderline NASH, Z3 pattern | 33.6 | 28.6 | 4.8 | 1.5 | 0.4 | ||
| Borderline NASH, Z1 pattern | 2.5 | 1.9 | 0.4 | 0.0 | 0.4 | ||
| Definite NASH | 0.2 | 67.5 | 94.0 | 98.5 | 99.2 | ||
| Fibrosis stage (%) | 1688 | <0.001 | |||||
| 0 | 52.7 | 25.3 | 6.0 | 1.5 | 0.4 | ||
| 1A | 16.5 | 20.8 | 16.0 | 7.1 | 1.2 | ||
| 1B | 5.8 | 9.1 | 20.4 | 21.8 | 15.4 | ||
| 1C | 5.5 | 5.2 | 1.2 | 0.3 | 0.0 | ||
| 2 | 12.4 | 22.7 | 26.8 | 24.5 | 18.7 | ||
| 3 | 5.8 | 12.7 | 23.2 | 32.7 | 47.7 | ||
| 4 | 1.3 | 4.2 | 6.4 | 12.1 | 16.6 | ||
| Steatosis grade (%) | 1688 | <0.001 | |||||
| <5 | 0.0 | 2.3 | 3.2 | 3.2 | 3.7 | ||
| 5–33 | 45.3 | 29.9 | 34.4 | 35.1 | 34.0 | ||
| 34–66 | 32.5 | 37.7 | 35.6 | 31.6 | 39.8 | ||
| >66 | 22.2 | 30.2 | 26.8 | 30.1 | 22.4 | ||
| Lobular inflammation (%) | 1688 | <0.001 | |||||
| <2/hpf | 76.4 | 60.4 | 46.8 | 31.9 | 19.1 | ||
| 2–4/hpf | 20.9 | 35.7 | 46.4 | 49.9 | 44.0 | ||
| >4/hpf | 2.7 | 3.9 | 6.8 | 18.3 | 36.9 | ||
| Portal inflammation (%) | 1687 | <0.001 | |||||
| None | 19.5 | 12.4 | 9.6 | 8.0 | 4.6 | ||
| Mild | 67.8 | 64.5 | 64.8 | 56.9 | 50.2 | ||
| More than mild | 12.7 | 23.1 | 25.6 | 35.1 | 45.2 | ||
| Glycogenosis (%) | 1448 | <0.001 | |||||
| None | 52.5 | 42.7 | 44.1 | 44.9 | 31.9 | ||
| Focal | 23.2 | 27.8 | 26.9 | 31.6 | 29.0 | ||
| Diffuse | 24.2 | 29.5 | 29.1 | 23.5 | 39.0 | ||
| Microvesicular steatosis (%) | 1688 | 3.3 | 12.3 | 15.2 | 18.0 | 23.7 | <0.001 |
| Acidophil bodies (%) | 1688 | 31.1 | 44.5 | 50.0 | 60.5 | 60.2 | <0.001 |
| Megamitochondria (%) | 1688 | 18.2 | 28.6 | 29.2 | 33.9 | 35.7 | <0.001 |
| Mallory-Denk bodies (%) | 1688 | 0.0 | 2.6 | 24.4 | 74.0 | 96.3 | <0.001 |
Among those with NAFLD. A total 1688 biopsies were included in this analysis (see the footnote of Table 2 for more details).
χ2 test.
Abbreviation: hpf, high power field.
DISCUSSION
Steatohepatitis is a histologic diagnosis defined by the presence of steatosis (ie, >5% macrovesicular fat, including both large and small droplet forms), lobular inflammation, and HB in a characteristic pattern. HB may occur in association with other etiologies of liver injury (eg, amiodarone toxicity, HCV infection, drug-induced liver injury), but remains the most specific histologic finding in steatohepatitis. HB was originally recognized as swollen cells with cleared/clumped cytoplasm.7 The clumped cytoplasm likely represents alterations in intermediate filaments of the cytoskeleton, which can be illustrated by loss of cytoplasmic keratin 8/18 immunostaining.18 The molecular and cellular mechanisms of HB are not known, but oxidative stress may play a role.19
The findings in this study demonstrate that nHB is, in fact, associated with a NASH CRN consensus diagnosis of definite steatohepatitis (up to 67.5% of biopsies) and was found in biopsies associated with clinical parameters commonly associated with NASH (eg, obesity, diabetes mellitus type 2, metabolic syndrome). Since participants with cHB had more advanced fibrosis (ie, stage 3 or 4 fibrosis) and more concerning laboratory findings than those with nHB, it is possible that nHB represent the earliest changes of HB type injury in NASH. This concept is further supported by increased frequency of advanced disease (ie, stage 3 or 4 fibrosis) in patients whose biopsies had severe cHB. nHB are subtle, especially when there is sectioning artifact or diffuse glycogenosis, and it is likely that many cases of NAFLD with such cells have not been diagnosed as steatohepatitis, based on current diagnostic criteria.13,20 We propose that such cases should be considered at least borderline for steatohepatitis and that, in practice, such patients should be managed similarly to patients with definite steatohepatitis. Two studies have shown that fibrosis progression in patients with borderline steatohepatitis occurs and is reduced if they improve to NAFL.21,22
The current NAS is critically important for assessment of improvement in follow up biopsies in therapeutic trials, with a reduction in NAS by 2 points used as a primary endpoint in the FLINT trial.15 A criticism of the NAS is that it places more weight on steatosis, which may not have as important a role in disease activity or fibrosis.23,24 One study concluded that variability in histologic interpretation of NAS parameters resulted in decreased study power for detection of treatment effects.25 These observations led to a call for engagement of expert hepatopathologists in study design and use of precisely defined criteria for scoring each histologic feature26; our findings could allow for increased precision in scoring through recognition of a range of ballooned hepatocyte morphology. Furthermore, an extended ballooning score allows for more flexibility in assessing for histologic improvement/worsening and in defining relationships of histologic injury to noninvasive markers of injury. This increased dynamic range can potentially monitor improvements better than the current system, as it better balances the importance of HB against steatosis and lobular inflammation, which in natural history studies, do not carry the same prognostic weight.23,24
The NASH CRN Pathology Committee stated in the initial proposal of the NAS16 that the activity score (the NAS) was not developed as a replacement for the pathologist’s pattern-determined diagnosis, but rather, as a method of semiquantitative evaluation for comparison of lesions in treatment trials. It was not developed to be utilized in clinical practice. In an evaluation of 978 liver biopsies from the NASH CRN collection,11 the close correlation of NAS and diagnosis was demonstrated, but also the differing value of each was shown. Likewise, the eB score is not intended for use in diagnosis. Additional studies could be helpful to confirm low interobserver variability among additional hepatopathologists, which would provide further support for incorporating the eB score into future clinical trial design.
In conclusion, this work has shown that nHB is a meaningful finding, and likely represents early/borderline changes of NASH that can allow for recognition of patients at risk for liver fibrosis, before cHB develops. The newly proposed eB score can be incorporated into future studies to more precisely delineate the morphologic range in HB. This may allow the detection of meaningful improvements in HB that are not captured with the current broader categories of HB, in which HB can change to a discernable degree but is assigned the same HB score.
Supplementary Material
Acknowledgments
AUTHOR CONTRIBUTIONS
Ryan M. Gill: study concept and design, acquisition of data, analysis and interpretation of data, drafting of manuscript, critical revision, statistical analysis. Daniela Allende: acquisition of data, analysis and interpretation of data, critical revision. Oscar W. Cummings: study concept and design, acquisition of data, analysis and interpretation of data, critical revision. Daniela Carpenter: analysis and interpretation of data, critical revision. Patricia H. Belt: study concept and design, acquisition of data, analysis and interpretation of data, critical revision, statistical analysis. Cynthia A. Behling: study concept and design, acquisition of data, analysis and interpretation of data, critical revision. Cynthia D. Guy: study concept and design, acquisition of data, analysis and interpretation of data, critical revision. Brent A. Neuschwander-Tetri: study concept and design, analysis and interpretation of data, critical revision. Arun J. Sanyal: study concept and design, analysis and interpretation of data, critical revision. James Tonascia: study concept and design, analysis and interpretation of data, critical revision, statistical analysis. Mark L. van Natta: study concept and design, analysis and interpretation of data, critical revision, statistical analysis. Laura Wilson: study concept and design, analysis and interpretation of data, critical revision, statistical analysis. Goro Yamada: study concept and design, analysis and interpretation of data, critical revision, statistical analysis. Matthew Yeh: study concept and design, acquisition of data, analysis and interpretation of data, critical revision. David Kleiner: study concept and design, acquisition of data, analysis and interpretation of data, drafting of manuscript, critical revision, statistical analysis, obtained funding, study supervision.
CONFLICT OF INTEREST
Daniela Allende—advisory arrangement with Incyte. Cynthia A. Behling—Dr. Behling’s institution (Pacific Rim Pathology Laboratory/Analytic Pathology Medical Group) currently receives support for biopsy related work through nonexclusive laboratory services and/or consulting agreements with: Akero Therapeutics Inc., Covance, E study site, Genesis Imaging Service, ICON, Medical Research Group Inc., Southern California Research Center. Dr. Behling has provided consulting or participated in seminars/preceptorships for Alimentev, Pfizer and Novo Nordisk. Oscar W. Cummings—Service contracts with NorthSea Therapeutics, Eli Lilly and Company, Galmed Research and Development Ltd, and Novo Nordisk, as well as consultancy aggrements with Akero Therapeutics, Zydus, Janssen, AstraZeneca, and Intercept Pharmaceuticals. Cynthia D. Guy—consulting agreement with Madrigal and 89 Bio. Daniela Carpenter—consulting agreements with Arrowhead Pharmaceuticals and Covance/LabCorp. Brent A. Neuschwander-Tetri—advisor or consultant: Akero, Alimentiv, Allergan, Allysta, Alnylam, Amgen, Arrowhead, Axcella,Boehringer Ingelheim, BMS, Coherus, Cymabay, Durect, Enanta, Fortress, GSK, Genfit, Gilead, Glympse, Hepeon, High Tide, HistoIndex, Innovo, Intercept, Ionis, LG Chem, Lipocine, Madrigal, Medimmune, Merck, Mirum, NGM, Novo Nordisk, Novus Therapeutics, pH-Pharma, Sagimet, Target RWE, Theratechnologies, 89 Bio; Stock options: HepGene; Institutional research grants: Allergan, BMS, Celgene, Cirius, Enanta, Genfit, Gilead, High Tide, Intercept, Inventiva, Madrigal, NGM. Arun J. Sanyal—has relationships with Exhalenz, Genfit, Gilead, Elsevier, Echosense-Sandhill, Malinckrodt, Intercept, Pfizer, Salix, Uptodate, Boehringer Ingelhiem, Novartis, Bristol Myers Squibb, Merck, Sequana, Hemoshear, Lilly, Novo Nordisk, Fractyl, Durect, Indalo, Terns. Albireo, Jannsen, Northsea, PoxeL, Tiziana, 89 Bio, Siemens, Astrazeneca, Ngm Bio, Amgen, Regeneron, Genentech, Alnylam, Roche, Rivus. Madrigal. Inventiva, Covance, Prosciento, Histoindex, Path AI. The remaining authors have no conflicts of interest.
FINANCIAL SUPPORT
The Nonalcoholic Steatohepatitis Clinical Research Network (NASH CRN) is supported by the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) (grants U01DK061713, U01DK061718, U01DK061728, U01DK061731, U01DK061732, U01DK061734, U01DK061737, U01DK061738, U01DK061730, U24DK061730). Additional support is received from the National Center for Advancing Translational Sciences (NCATS) (grants UL1TR000439, UL1TR000436, UL1TR000006, UL1TR000448, UL1TR000100, UL1TR000004, UL1TR000423, UL1TR000058). This research was supported in part by the Intramural Research Program of the NIH, National Cancer Institute. The PIVENS trial was conducted by the NASH CRN and supported in part by Takeda Pharmaceuticals North America through a Cooperative Research and Development Agreement with the NIDDK. The vitamin E and matching placebo for the PIVENS trial were provided by Pharmavite through a Clinical Trial Agreement with the NIH. The FLINT trial was conducted by the NASH CRN and supported in part by a Collaborative Research and Development Agreement (CRADA) between NIDDK and Intercept Pharmaceuticals.
Footnotes
Abbreviations: ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body mass index; BS, ballooning score; cHB, classic hepatocyte ballooning; CRN, clinical research network; DB2, database 2; eBS, extended ballooning score; eNAS, extended nonalcoholic fatty liver disease activity score; FLINT, farnesoid X Receptor Ligand Obeticholic Acid in NASH Treatment; FXR, Farnesoid X Receptor; HB, hepatocyte ballooning; HOMA-IR, homeostasis model assessment of insulin resistance; NAS, nonalcoholic fatty liver disease activity score; NB, no ballooning; nHB, nonclassic hepatocyte ballooning; PIVENS, pioglitazone versus Vitamin E versus Placebo for the Treatment of Nondiabetic Patients with Nonalcoholic Steatohepatitis.
Supplemental Digital Content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal's website, www.hepcommjournal.com.
Contributor Information
Ryan M. Gill, Email: Ryan.gill@ucsf.edu.
Daniela Allende, Email: allendd@ccf.org.
Patricia H. Belt, Email: pbelt1@jhu.edu.
Cynthia A. Behling, Email: cynthiabehlingmd@gmail.com.
Oscar W. Cummings, Email: ocumming@iupui.edu.
Cynthia D. Guy, Email: cynthia.guy@duke.edu.
Daniela Carpenter, Email: danielle.carpenter@health.slu.edu.
Brent A. Neuschwander-Tetri, Email: brent.tetri@health.slu.edu.
Arun J. Sanyal, Email: arun.sanyal@vcuhealth.org.
James Tonascia, Email: jtonasci@jhu.edu.
Mark L. Van Natta, Email: mvnatta@jhu.edu.
Laura A. Wilson, Email: lwilson9@jhu.edu.
Goro Yamada, Email: goro.yamada@drexel.edu.
Matthew Yeh, Email: myeh@uw.edu.
David E. Kleiner, Email: kleinerd@mail.nih.gov.
REFERENCES
- 1.Browning JD, Szczepaniak LS, Dobbins R, Nuremberg P, Horton JD, Cohen JC, et al. Prevalence of hepatic steatosis in an urban population in the United States: impact of ethnicity. Hepatology. 2004;40:1387–95. [DOI] [PubMed] [Google Scholar]
- 2.Williams CD, Stengel J, Asike MI, Torres DM, Shaw J, Contreras M, et al. Prevalence of nonalcoholic fatty liver disease and nonalcoholic steatohepatitis among a largely middle-aged population utilizing ultrasound and liver biopsy: a prospective study. Gastroenterology. 2011;140:124–31. [DOI] [PubMed] [Google Scholar]
- 3.Chalasani N, Younossi Z, Lavine JE, Charlton M, Cusi K, Rinella M, et al. The diagnosis and management of nonalcoholic fatty liver disease: practice guidance from the American Association for the Study of Liver Diseases. Hepatology. 2018;67:328–357. [DOI] [PubMed] [Google Scholar]
- 4.Chalasani N, Younossi Z, Lavine JE, Diehl AM, Brunt EM, Cusi K, et al. The diagnosis and management of nonalcoholic fatty liver disease: practice guideline by the American Gastroenterological Association, American Association for the Study of Liver Diseases, and American College of Gastroenterology. Gastroenterology. 2012;142:1592–609. [DOI] [PubMed] [Google Scholar]
- 5.Brunt EM. Nonalcoholic steatohepatitis: definition and pathology. Semin Liver Dis. 2001;21:3–16. [DOI] [PubMed] [Google Scholar]
- 6.Kleiner DE, Brunt EM. Nonalcoholic fatty liver disease: pathologic patterns and biopsy evaluation in clinical research. Semin Liver Dis. 2012;32:3–13. [DOI] [PubMed] [Google Scholar]
- 7.Ludwig J, Viggiano TR, McGill DB, Oh BJ. Nonalcoholic steatohepatitis: Mayo Clinic experiences with a hitherto unnamed disease. Mayo Clin Proc. 1980;55:434–8. [PubMed] [Google Scholar]
- 8.Sanyal AJ, Brunt EM, Kleiner DE, Kowdley KV, Chalasani N, Lavine JE, et al. Endpoints and clinical trial design for nonalcoholic steatohepatitis. Hepatology. 2011;54:344–53. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Gill RM, Kakar S. Nonalcoholic steatohepatitis: diagnostic challenges. Surg Pathol Clin. 2013;6:227–57. [DOI] [PubMed] [Google Scholar]
- 10.Caldwell S, Ikura Y, Dias D, Isomoto K, Yabu A, Moskaluk C, et al. Hepatocellular ballooning in NASH. J Hepatol. 2010;53:719–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Brunt EM, Kleiner DE, Wilson LA, Belt P, Neuschwander-Tetri BA. Nonalcoholic fatty liver disease (NAFLD) activity score and the histopathologic diagnosis in NAFLD: distinct clinicopathologic meanings. Hepatology. 2011;53:810–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Kleiner DE, Brunt EM, Belt PH, Behling CA, Gill RM, Guy CD, et al. Extending the ballooning score beyond 2: a proposal for a new balloon score. Hepatology. 2015;62:288A–289A. [Google Scholar]
- 13.Pai RK, Jairath V, Hogan M, Zou G, Adeyi OA, Anstee QM, et al. Reliability of histologic assessment for NAFLD and development of an expanded NAFLD activity score. Hepatology. 2022;76:1150–1163. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Sanyal AJ, Chalasani N, Kowdley KV, McCullough A, Diehl AM, Bass NM, et al. Pioglitazone, vitamin E, or placebo for nonalcoholic steatohepatitis. N Engl J Med. 2010;362:1675–85. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Neuschwander-Tetri BA, Loomba R, Sanyal AJ, Lavine JE, Van Natta ML, Abdelmalek MF, et al. Farnesoid X nuclear receptor ligand obeticholic acid for non-cirrhotic, nonalcoholic steatohepatitis (FLINT): a multicentre, randomised, placebo-controlled trial. Lancet. 2015;385:956–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Kleiner DE, Brunt EM, Van Natta M, Behling C, Contos MJ, Cummings OW, et al. Design and validation of a histological scoring system for nonalcoholic fatty liver disease. Hepatology. 2005;41:1313–21. [DOI] [PubMed] [Google Scholar]
- 17.Brunt EM, Kleiner DE, Wilson LA, Unalp A, Behling CE, Lavine JE, et al. Portal chronic inflammation in nonalcoholic fatty liver disease (NAFLD): a histologic marker of advanced NAFLD—clinicopathologic correlations from the nonalcoholic steatohepatitis clinical research network. Hepatology. 2009;49:809–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Lackner C, Gogg-Kamerer M, Zatloukal K, Stumptner C, Brunt EM, Denk H. Ballooned hepatocytes in steatohepatitis: the value of keratin immunohistochemistry for diagnosis. J Hepatol. 2008;48:821–28. [DOI] [PubMed] [Google Scholar]
- 19.Lackner C. Hepatocellular ballooning in nonalcoholic steatohepatitis: the pathologist’s perspective. Expert Rev Gastroenterol Hepatol. 2011;5:223–31. [DOI] [PubMed] [Google Scholar]
- 20.Brunt EM, Clouston AD, Goodman Z, Guy C, Kleiner DE, Lackner C, et al. Complexity of ballooned hepatocyte feature recognition: defining a training atlas for artificial intelligence-based imaging in NAFLD. J Hepatol. 2022;76:1030–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Brunt EM, Kleiner DE, Wilson LA, Sanyal AJ, Neuschwander-Tetri BA. for the Nonalcoholic Steatohepatitis Clinical Research Network. Improvements in histologic features and diagnosis associated with improvement in fibrosis in nonalcoholic steatohepatitis: Results from the Nonalcoholic Steatohepatitis Clinical Research Network treatment trials. Hepatology. 2019;70:522–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Kleiner DE, Brunt EM, Wilson LA, Behling C, Guy C, Contos M, et al. Association of histologic disease activity with progression of nonalcoholic fatty liver disease. JAMA Netw Open. 2019;2:ee1912565. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Bedossa P. Current histological classification of NAFLD: strength and limitations. Hepatol Int. 2013;7(suppl 2):765–70. [DOI] [PubMed] [Google Scholar]
- 24.Bedossa P, Poitou C, Veyrie N, Bouillot JL, Basdevant A, Paradis V, et al. Histopathological algorithm and scoring system for evaluation of liver lesions in morbidly obese patients. Hepatology. 2012;56:1751–9. [DOI] [PubMed] [Google Scholar]
- 25.Davison BA, Harrison SA, Cotter G, Alkhouri N, Sanyal A, Edwards C, et al. Suboptimal reliability of liver biopsy evaluation has implications for randomized clinical trials. J Hepatol. 2020;73:1322–32. [DOI] [PubMed] [Google Scholar]
- 26.Longerich T, Schirmacher P. Determining the reliability of liver biopsies in NASH clinical studies. Nat Rev Gastroenterol Hepatol. 2020;17:653–4. [DOI] [PubMed] [Google Scholar]