The Burden of Nonalcoholic Steatohepatitis: A Systematic Review of Epidemiology Studies

Background and Aims Nonalcoholic steatohepatitis (NASH) is associated with increased mortality and risk of complications but is often asymptomatic and under-recognized. A systematic review of NASH epidemiology was conducted to provide information on the burden of NASH and highlight important evidence gaps for future research. Methods Medline, EMBASE, and Cochrane Library databases were searched for English-language publications published from 2010 to January 2022 that reported on natural history, risk factors, comorbidities, and complications of a NASH population or subpopulation. Results Overall, 173 publications were included. NASH was shown to have a variable disease course and high prevalence of comorbid disease. Although many patients progressed to cirrhosis and end-stage liver disease, disease regression or resolution was reported in up to half of patients in some studies. Reported risk factors for disease progression or resolution included levels of (or changes in) serum fibrosis markers, liver enzymes, and platelets. The presence of NASH increased the risk of liver cirrhosis and other serious diseases such as hepatocellular carcinoma, colorectal cancer, chronic kidney disease, and cardiovascular disease. In 2017, NASH was responsible for ∼118,000 cirrhosis deaths globally, and an increasing proportion of patients are receiving liver transplantation for NASH in Europe and the United States. Consolidation of data was hampered by heterogeneity across the studies in terms of patient populations, follow-up time, and outcomes measured. Conclusion NASH is associated with significant morbidity and mortality, an increased risk of comorbidities, and imposes an increasing burden among liver transplantation recipients. Longer studies with harmonized study criteria are required to better understand the impact of NASH on patient outcomes.


Introduction
N onalcoholic fatty liver disease (NAFLD) is one of the most common causes of chronic liver disease worldwide, with an estimated global prevalence of 25%. 1 NAFLD represents a spectrum of liver disease ranging from simple steatosis to nonalcoholic steatohepatitis (NASH), which can then progress to cirrhosis. 2NASH, the most severe form of NAFLD, occurs when excess fat has accumulated in the liver, and lobular inflammation and hepatocellular ballooning are present. 3Despite this, NASH is often perceived as asymptomatic or unrecognized and may not be diagnosed until late stages. 4,5ASH occurs in almost one-quarter of patients with NAFLD. 6It has an estimated prevalence of 1.50%-6.45% in the general population, 1 although disease prevalence and incidence estimates vary considerably due to differing patient populations and disease criteria. 2NASH is associated with obesity and type 2 diabetes (T2D), hepatic complications including cirrhosis, hepatocellular carcinoma (HCC), and extrahepatic conditions including cardiovascular (CV) disease (CVD) and cancer. 3,7NASH cirrhosis is predicted to become the leading cause of liver transplantation (LT) in the United States between 2020 and 2025. 8his study aimed to summarize and critically appraise studies on the natural history of NASH, risk factors for NASH, hepatic and extrahepatic complications associated with NASH, related comorbidities, and LT outcomes, and to highlight important evidence gaps for future research.

Methods
A systematic literature review was conducted using Medline, EMBASE, and Cochrane Library databases via the Ovid platform using predefined search strategies (Table A1).The review was conducted in line with Cochrane guidelines 9 and is reported in line with Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. 10he primary systematic literature review was conducted in February 2020, with updates in January 2021 and January 2022 (last searches ran on January 17, 2022).Relevant conference proceedings from 2017 to February 10, 2020, were handsearched for unpublished studies (repeated on January 6, 2021, and January 17-18, 2022), and submission documents from Health Technology Assessment agencies were reviewed for relevant data (Table A2).Conference abstracts were identified and screened, but data were not extracted as it was anticipated that the large number of full-text publications would provide an adequate evidence base of relevant information.Eligible publications were full-text English-language epidemiology studies that included outcomes in NASH populations (with or without comorbidities), or a NASH subgroup within a NAFLD study, published from 2010 to January 2022.Data from NAFLD or heterogeneous NAFLD/NASH populations were excluded.Full eligibility criteria are described in Table A3.
First-round screening of titles and abstracts was followed by second-round full-text screening of short-listed articles and data extraction from articles meeting the eligibility criteria.Both firstand second-round screenings were performed by 2 independent researchers, and final inclusion was verified by the project lead.Data extraction was performed using predesigned data extraction tables in Microsoft ® Excel by an analyst with a 100% quality check by an independent senior analyst.Data extracted from each study included, but were not limited to, reference, country, study design, publication and study characteristics, baseline characteristics, intervention(s), and outcomes reported.Disputes regarding eligibility or data extraction were referred to a third party.
The quality of included observational studies was assessed using the Quality Assessment Tool for Quantitative Studies produced as part of the Effective Public Health Practice Project. 11This assesses the quality of 6 components to assign a global rating for each study of "strong" (no weak ratings for any of the listed criteria), "moderate" (one weak rating), or "weak" (2 or more weak ratings).The components are selection bias (likelihood of representativeness of target population and percentage participation), design (study type), confounders (percentage of relevant confounders controlled for), blinding (of outcome assessors and study participants), data collection methods (validity and reliability), and withdrawals and dropouts.Any component that did not apply to the study was marked "not applicable."Quality assessment (risk of bias) of randomized controlled trials was conducted using the 7-criteria checklist provided in Section 4.6 of the UK National Institute for Health and Care Excellence single technology appraisal user guide to assess the likelihood of selection, performance, attrition, and detection bias. 12Quality assessments were not used to exclude studies.Two reviewers independently assessed the likelihood of bias, and any disagreements were resolved by discussion and/or additional referees.

Results
The electronic database searches identified 12,105 citations, of which 2022 were identified as duplicates and excluded.A further 9926 were excluded after screening by title and abstract or full-text screening.Sixteen studies that met the eligibility criteria were identified from handsearching, resulting in inclusion of 173 publications covering 167 unique studies (Figure A1 and Table A4).An overview of the study and patient characteristics across the identified studies is included in Tables 1-7 and Tables A5-A7.

Disease Trajectory
Only 6 studies reported on aspects of disease progression or regression that occurred naturally over time without pharmacological intervention (Table 1).Study populations and outcomes measured were heterogeneous, and it was therefore not possible to make direct comparisons across studies.][15][16] Rates of progression to cirrhosis were shown to increase with increasing fibrosis severity in a Spanish database analysis, 16 and a 22% rate of progression from F3 fibrosis to cirrhosis over a median 29 months was reported in a pooled analysis of the phase 2b simtuzumab trials (in which simtuzumab is assumed to have no biological activity). 15nterestingly, patient age was linked to disease progression, but not regression, in a study of first-time adult LT registrants; in this study, removal from the LT wait list because of death or being too sick increased with increasing age, whereas removal from the LT wait list due to condition improving was similar regardless of patient age. 14 b Nondrinkers were defined as current "never" drinkers who have not, over the course of their lifetime, had !1 alcoholic drink/month during a 12-month period or had !3 drinks per day for !3 consecutive days, whereas modest drinkers were defined as those drinking 2 drinks on a typical day, in the absence of monthly (or more frequent) heavy drinking.c Simtuzumab-and placebo-treated patients were pooled based on the assumption that simtuzumab has no biological activity, supported by its failure to improve liver histology, portal pressure, liver biochemistry, and serum fibrosis markers in RCTs.d NCT01672866 and NCT01672879.e Excluding patients who underwent liver biopsy during bariatric surgery (n ¼ 539) or were enrolled in clinical trials (n ¼ 120).f The activity part of the SAF scoring system that incorporates scores for ballooning and inflammation.
g Only 2 patients receiving placebo were included in the biopsy-evaluable set; therefore, meaningful comparison with placebo is not possible.
h Contributed to by at least 2 of: steatosis, inflammation, ballooning.
Nineteen interventional studies (over 21 publications) reported on the reversibility of histologic changes associated with NASH and fibrosis stage in response to therapy, lifestyle intervention, or placebo (summarized in Table 2).Histologic improvement in NASH (NAFLD Activity Score [NAS] improvement of !1 or !2 stages or improvement in Steatosis, Activity, and Fibrosis score) were reported in up to 92% of patients receiving pharmacologic therapies, with the highest rates of improvement seen with efruxifermin and semaglutide. 33,377][38][39] Significantly greater proportions of patients achieving resolution of NASH (or resolution of NASH without worsening of fibrosis) compared with placebo were reported for few pharmacologic interventions.These included vitamin E (36% vs 21%, P ¼ .05),pioglitazone (47% vs 21%, P < .001and 51% vs 19%, P < .001),elafibranor (19% vs 12%, P ¼ .045),and semaglutide 0.4 mg (59% vs 17%, P < .001). 21,26,27,3737,39 Bariatric surgery was the most effective intervention in terms of NASH improvement, with NASH disappearing in 85% of recipients. 23In some studies, improvements in NASH were accompanied by improvements in fibrosis, 23,24,27,30,33 but improvements in NASH without substantial improvement in fibrosis were also reported in several studies, [19][20][21]34,37 and one study reported improvements in fibrosis without improvements in NAS. 28 Reolution of NASH may result in improvement in metabolic parameters.Two post-hoc analyses of the Pioglitazone vs Vitamin E vs Placebo for the Treatment of Non-Diabetic Patients with Nonalcoholic Steatohepatitis (PIVENS) study, which pooled data from patients receiving pioglitazone, vitamin E, and placebo in nondiabetic patients with NASH, found that NASH resolution was associated with favorable changes in triglycerides, high-density lipoprotein levels, and lipoprotein fractions, compared with patients without NASH resolution.40,41

Risk Factors for Progression or Resolution of NASH and/or Fibrosis
Five publications reported on factors associated with regression or resolution of NASH or fibrosis (Table 3). 15,17,31,38,42In the PIVENS trial, factors associated with higher odds of NASH resolution included reductions in alanine aminotransferase (ALT) and aspartate aminotransferase (AST; vitamin E and pioglitazone arms), lower baseline metabolic syndrome (MetS) severity Z-score (MetS-Z; vitamin E and pioglitazone arms), and reductions in MetS-Z (placebo arm). 31In patients not receiving pharmacologic intervention, lower odds of NASH resolution were also reported in consistent moderate drinkers compared with consistent nondrinkers. 17Factors that predicted fibrosis regression included higher baseline fibrosis stage, lower baseline levels of fibrosis markers, and reduced/normal liver enzyme levels. 15,38everal publications reported on factors associated with disease progression or more advanced disease; these were predominantly cross-sectional studies that evaluated markers in patients with different stages of fibrosis (Table 3).One study in Japanese patients reported that absence of ALT response to pharmacologic treatment was associated with worsening NAS or fibrosis. 43Two studies noted factors associated with liver-related clinical events in patients with compensated cirrhosis 15 or NASH 44 ; these included baseline (or lack of improvement in) fibrosis stage; higher baseline levels of hepatic collagen, lysyl oxidase like 2, bilirubin, international normalized ratio, and Model for End-Stage Liver Disease; lower baseline platelets; higher noninvasive tests (enhanced liver fibrosis, NAFLD fibrosis score [NFS], fibrosis-4, AST-to-platelet ratio index, Fibro-Sure/FibroTest); and higher (or worsening/lack of improvement in) hepatic venous pressure gradient.The most prevalent factors associated with fibrosis progression or advanced fibrosis stage across the studies were higher NFS/change in NFS, 15,18,42,45 higher AST:ALT ratio, 18,46 ALT normalization/nonresponse, 42,43 higher AST-to-platelet ratio index, 15,45 higher fibrosis-4, 15,18 lower platelets/change in platelet count, 15,42,47 and T2D. 18,42The predictive value of some factors was not consistent.Serum ALT level and AST:ALT ratio, for example, were each reported in one study as not being significantly associated with patients' fibrosis stage. 45,46Clinical and sociodemographic predictors of fibrosis progression included longer time since NASH diagnosis, male sex, current smoking, obesity, and lack of fulltime employment. 48

Hepatic Complications of NASH
One study reporting data from the US National Inpatient Sample found that NASH accounted for 9.3% of all cirrhosisrelated hospitalizations from 2006 to 2014, with the proportion of hospital admissions due to NASH-related cirrhosis doubling from 2006-2008 to 2012-2014. 49welve studies reported the incidence (or an increased risk) of HCC in individuals with NASH, and 6 studies reported on the prevalence of NASH in HCC (Table 4 and  Table A5).A study in diabetic patients with NASH and bridging fibrosis or cryptogenic cirrhosis reported that 14.7% of the study population developed HCC over the study period (2004-2016). 50Studies in patients with NASH and cirrhosis reported development of HCC in 13% and 15% of patients, after a median 5.1 years and mean 6.8 years follow-up, respectively, 51,52 and in LT wait list patients with NASH cirrhosis, the annual incidence of HCC was 2.7%. 134][55][56] HCC was more common in patients with advanced (stage 3/4) fibrosis compared with early-stage (0-2) fibrosis 53,57 and in patients with diabetes vs without. 52In studies that assessed the prevalence of     b Nondrinkers were defined as current "never" drinkers who had not, over the course of their lifetime, had !1 alcoholic drink per month during a 12-month period or had !3 drinks per day for !3 consecutive days, whereas modest drinkers were defined as those drinking 2 drinks on a typical day in the absence of monthly (or more frequent) heavy drinking.Consistent drinkers and nondrinkers were those whose drinking status did not change between baseline and follow-up.c Adjusted for age at biopsy, sex, race, and ever smoker.
d Simtuzumab-and placebo-treated patients were pooled based on the assumption that simtuzumab has no biological activity, supported by its failure to improve liver histology, portal pressure, liver biochemistry, and serum fibrosis markers in RCTs.e NCT01672866 and NCT01672879.
f Presence of hyperuricemia defined as !7 mg/dL for men and !6 mg/dL for women.g Clinical outcomes were death, liver transplantation, and end-stage hepatic complications (HCC, ascites, spontaneous bacterial peritonitis, hepatic encephalopathy, de novo varices, or significant worsening of varices).
h ALT response was defined as a decrease of 30% or more from baseline.NASH in HCC, NASH was reported in 4%-20% of HCC patients.[68]

Extrahepatic Comorbidities and Complications
This review identified 16 studies reporting extrahepatic complications of NASH and 21 studies reporting CV complications.Key findings are summarized below, with further details in Table 5.
Three studies have reported an increased risk of developing colorectal cancer in patients with NASH/NAFLD (Table 5) [69][70][71] ; these suggest that NAFLD or NASH approximately doubles the risk of developing colorectal neoplasms.Three studies reported that NASH is an independent risk factor for chronic kidney disease (CKD) or is associated with an increased risk for severe CKD post-LT (Table 5).When compared against nonsteatotic controls matched for age, gender, and body mass index (BMI), NASH was associated with an 8.3-fold increased risk of CKD. 72In individuals undergoing LT, the increased risk was approximately 3-fold higher in those with NASH vs those without 73 and 3.5-fold higher in patients with NASH vs non-NASH NAFLD. 74The presence of severe CKD (stage 3 or 4/5) in patients with NASH was associated with significantly increased risk of 1-year LT wait list mortality. 75ssociations between NASH and CVD were investigated in 15 studies (Table 5).Several studies reported increased risk of CV conditions with NASH compared with nonalcoholic fatty liver, [76][77][78][79][80] and accordingly, NASH patients were found to have a more atherogenic lipid profile in NASH compared with non-NASH populations. 81One database study of more than 9000 patients with NASH/NAFLD reported a higher prevalence of atrial fibrillation in individuals aged <65 years with biopsy-proven NASH (4.0%) vs a non-NASH/NAFLD tertiary care cohort (2.7%) and vs a general population estimate from published literature (2.0%). 77Furthermore, the presence of atrial fibrillation increased the risk of other CV complications, and led to more frequent and longer hospitalizations, compared with NASH patients without atrial fibrillation. 77Coronary artery disease was shown to occur more frequently in patients with NASH compared with hepatitis C virus (HCV) and alcoholic liver disease, and NASH is a significant risk factor for both coronary artery disease and major adverse CV events. 82In addition, one study found that hepatic steatosis with fibrosis is independently associated with the progression of carotid atherosclerosis in patients with T2D. 78omorbid conditions such as T2D, obesity, and hepatitis increase the risk of NASH/NAFLD and/or are associated with worse outcomes.In a single-center Belgian study, NASH was one of the most frequently reported comorbidities, occurring in 72% of participants. 83Another study of 1069 patients with NAFLD, 30% of whom had diabetes, showed that family history of diabetes is strongly associated with risk of NASH, fibrosis, and advanced fibrosis. 84T2D in NASH patients is associated with poor outcomes; in an international cohort of 299 patients with NASH, the presence of T2D was associated with a 4.59-fold increased risk of mortality, 2.46-fold increased risk of hepatic decompensation, and 4.2-fold increased risk of HCC vs patients without T2D. 526][87][88][89][90] However, individuals with NAFLD without obesity may also be at risk of progression to NASH.One Italian study of patients with biopsy-proven NAFLD showed that increasing waist circumference was associated with increasing prevalence of severe hepatosteatosis and NASH (but not fibrosis), but in a multivariate analysis, neither NASH nor fibrosis were significantly associated with waist circumference, implying that patients with a normal waist circumference may also be at risk of NASH and fibrosis. 91nother study showed that the presence of NASH in patients with hepatitis was associated with significantly worse long-term outcomes, including reduced event-free survival, increased risk of HCC, HCC-free survival, transplant-free survival, and LT/death (all P < .01). 92

LT and Post-transplant Outcomes
As NASH progresses, many patients require LT, and consequently, NASH is one of the main indications for LT in many countries.4][95][96][97] Both US and European studies have shown that NASH patients as a proportion of all patients undergoing LT have increased over time. 93,94In 2 large US studies, NASH was the second most common indication for LT. 95,97Additional detail on the proportion of NASH patients undergoing LT across the studies is provided in Table A6.

Mortality
Key non-LT mortality data is presented in Table 7, including the Global Burden of Disease, Injuries, and Risk Factors study.5][106] Individual studies reported that NASHrelated cirrhosis was responsible for w118,000 deaths globally in 2017, 107 and there was a 3.7% inpatient mortality rate among patients hospitalized for cirrhosis in the United States between 2006 and 2014. 49There was no consistent position on the impact of NASH etiology (vs other underlying etiologies) on mortality in HCC populations. 66,67,108,109

Quality of Included Studies
Information on the quality assessment of included studies is provided in Tables A8 and A9.

Discussion
The findings of this study provide a detailed picture of the course of NASH and the associated complications, as well as risk factors for the development of NASH and worsening fibrosis.Overall, 173 publications were included in this review, but considerable heterogeneity in study designs and methodologies, population sizes, study duration, outcomes, and study quality meant that direct comparison across studies was challenging.
Few studies evaluated the natural history of NASH in the absence of pharmacologic or bariatric surgery intervention, and no firm conclusions could be drawn due to study heterogeneity.[24][25][26][27] The reasons for these spontaneous regressions are not clear, and although variance within the limitations of histologic assessments may play a role, they likely reflect behavioral changes.34,37 The reasons for this discordancy are unclear but could include differences between studies in patient baseline characteristics (eg, fibrosis stage, BMI) or study design (eg, dosing, treatment duration). 34It is also possible that assessment of fibrosis using continuous variables, such as the degree of liver stiffness, may have shown improvements that could not be detected using categorical outcome measures. 37Overall, the balance of disease  e High-risk CAD patients underwent percutaneous balloon angioplasty along with coronary artery stent insertion before LT.
g RISK includes patients with risk profile of NASH using the following formula: age >50 years, ALT >30 U/L, and (BMI >30 kg/m 2 or T2D).trajectory data suggests that development of steatohepatitis and fibrosis are mostly reversible changes that may respond to pharmacologic interventions but also to changes in lifestyle (ie, diet and physical activity) mediated largely by weight loss.In the absence of long-term follow-up for many of the identified studies, the durability and impact of these regressions or resolutions remain unclear.
Liver biopsy remains the only validated tool for NASH diagnosis and staging, and as such, repetitive biopsies are a typical regulatory requirement for interventional studies. 128,129However, liver biopsies are inherently invasive, can be a burden to patients, and present logistical challenges for use in clinical trials (eg, costs, availability of sufficiently experienced pathologists). 128,129Therefore, development of noninvasive markers, such as blood-based biomarkers, is an active area of research, 130,131 and indeed, both the U.S. Food and Drug Administration and the European Medicines Agency encourage their development to potentially replace liver biopsy in future clinical trials. 128,129The present review includes several studies that identified blood-based risk factors for NASH disease progression or regression; however, many risk factors were identified by single studies only, and the more frequently occurring factors were not always consistent across studies.Further research is needed to test and validate such biomarkers to determine their reliability and potential for use in clinical practice. 130,131he burden of NASH in terms of comorbidities and complications is high and growing.More than 80% of patients with NASH are estimated to have one or more comorbidity, with notably high global prevalence rates for obesity (82%), hyperlipidemia/dyslipidemia (72%), MetS (71%), hypertension (68%), and diabetes (44%). 1 Furthermore, increasing incidence rates of NAFLD-related cirrhosis, NAFLD cirrhosis-related mortality, and NAFLDrelated liver cancer have been reported. 1328,113 Surprisingly, no studies were identified in this review that reported an association between NASH with other solid tumor types.This may arise from the constraints of this study's eligibility criteria to include outcomes reported in NASH populations or subpopulations only because associations between mixed severity NAFLD populations with other extrahepatic tumor types have been reported. 7,133For example, a large longitudinal cohort study in patients with NAFLD, including patients with NASH and nonalcohol-related cirrhosis, has reported increased risk of uterine, stomach, and pancreatic cancers in addition to liver and colorectal cancer. 7Lack of universal noninvasive diagnostic methods and unreliability of liver enzymes as serum biomarkers makes it difficult to attribute cancer risk to NASH vs simple steatosis in large population-based studies, 7 which may contribute to the limited reports identified in this NASH-specific review.There was no difference in overall survival among NASH-HCC (1-, 3-, and 5-y overall survival of 82.8%, 75.7%, and 71.3%), hepatitis-HCC (95.7%, 77.6%, and 60.4%), and ALD-HCC (92.0%, 79.9%, and 79.9%), P ¼ .952NASH-HCC was associated with longer disease-specific survival than hepatitis-HCC (1-, 3-, and 5-y survival of 93%, 93%, and 87.5% vs 95.7%, 81.9%, and 63.7%, P ¼ .048)Recurrence-free survival was comparable between NASH-HCC, hepatitis-HCC, and ALD-HCC Presence of T2D had no impact on outcomes in either liver disease Alongside the rising prevalence of NASH and its adverse complications, the proportion of LTs resulting from NASH has increased over time in the United States, Europe, Australia, and New Zealand. 93,94,138Although NASH is not yet one of the most common causes of LT in Europe, 93 it was reported to be the second most common cause for LT in 2 large US studies, behind HCV. 95,97The magnitude of the impact of NASH on LT in the United States may serve as an indication of the future burden in Europe, considering the projected rise in obesity across the continent. 93In addition, HCV is likely to decline as a cause of LT in coming years considering improvements in pharmacologic treatment options, and transplantation for NASH will therefore increase in both relative and absolute terms. 82Crucially, LT for NASH can be precluded by the presence of comorbidities.For example, the cause of death in individuals with NASH or cryptogenic cirrhosis is often CVD-related (whereas HCV patients are more likely to die of liver-related complications).Indicators of comorbid conditions (eg, diabetes, obesity, lipid disorders, and hypertension) should be treated aggressively to limit the risk of conditions that may be fatal or make patients ineligible for transplant. 13

Strengths and Limitations
This review synthesized a large body of evidence (173 publications), with most (116/151) of the quantitative studies rated as "strong" or "moderate" quality to provide a comprehensive picture of the available data.
However, we acknowledge several limitations.First, the search criteria for the present study focused on patients with NASH, and although NAFLD was included in the search strategy, publications were not included if outcome data for a NASH subgroup was absent.As NASH is the progressive phenotype of NAFLD, it is likely that some NAFLD publications lacking explicit mention of NASH were excluded.However, as all publications that described NASH in their title/abstract were included, it is unlikely that excluded publications would have focused on NASH, and they may not have presented any discrete results outside of the overall NAFLD population.Second, although the review identified an extensive body of evidence regarding the course of NASH and common complications, the evidence may not address differences in the characteristics or course of disease in different ethnic groups or different geographic regions.Furthermore, more than half of the publications reported data from the United States (88/173), and consequently, most of the evidence, especially relating to LT, is based on studies performed in the United States.Third, there was considerable heterogeneity across the included publications with regard to the sample size, study design, follow-up time, and outcomes.Furthermore, conference abstracts were excluded from any further analysis, and only full-text publications were included.Finally, despite most publications receiving "strong" or "moderate" ratings in the quality assessment, 35 publications were rated as "weak," primarily because of missing data regarding withdrawals and dropouts, as most of these studies were observational, and inadequate controlling of confounders.

Evidence Gaps and Recommendations
This systematic literature review included studies from 2010 to 2022, and almost half of the publications (n ¼ 78) were published in 2019-2021.Although this shows a growing interest in research on the NASH population, data on NASH incidence and prevalence are limited and extremely variable. 2There is also a paucity of studies evaluating the natural history of NASH in the absence of pharmacologic or surgical intervention.
Evidence identified by this review clearly demonstrates that a significant number of interventional trials have measured histologic improvement and regression of NASH after pharmacological treatment; however, the durability and impact of regression are unclear in studies with only a few years' follow-up.Longitudinal studies demonstrate effectively how a slowly progressing disease such as NASH occurs, but their follow-up is also not always sufficient to accurately capture NASH progression or regression.For example, Sanyal et al 15 noted that their follow-up of 2 years is short for a slowly progressing disease such as NASH, and Ampuero et al 16 commented that a mean follow-up time of 4.7 years may not be enough to observe a high number of prognostic outcomes (eg, HCC), whereas Hirose et al 57 lost w40% of participants to follow-up over a median of 19.5 years.More longitudinal studies or real-world evidence studies over longer time frames are needed to accurately assess NASH progression/regression, especially given the expected increase in future prevalence and impact of NASH with the obesity epidemic.Harmonized study durations and patient populations and appropriate control of confounders are also needed to fully assess the prevalence and impact of NASH, which in turn could support an evidencebased focused approach to its disease burden.
Despite a wealth of studies assessing risk factors for progression of NASH and/or fibrosis, validated noninvasive biomarkers that predict patients' disease trajectory or response to interventions are lacking.Testing and validation of such biomarkers should be prioritized, both for use in clinical practice and clinical trials, and to support our future understanding of NASH epidemiology and disease progression.Future studies should also focus on improving our understanding of NASH progression and its association with comorbidities, morbidity, and mortality, as well as associated risk factors.Except for colorectal cancer, there are limited data on the increased risk of solid tumors specifically in NASH populations.It would be beneficial for studies assessing the risk of solid tumors to stratify patients by the degree of steatohepatitis to determine if increased risks are incurred once patients' disease progresses from nonalcoholic fatty liver to NASH.

Conclusions
NASH is associated with significant morbidity and mortality, an increased risk of comorbidities that include HCC, CKD, and CVD, and imposes an increasing burden among LT patients.
Although it is difficult to predict, NASH/NAFLD prevalence is expected to increase in parallel with the obesity and diabetes epidemics, exacerbated by a lack of general awareness, pharmacologic treatments, or a clear care pathway.Longitudinal studies with longer follow-up are required to truly understand the impact of NASH on patient outcomes.

a
Participants were drawn from 3 groups: (1) the adult NAFLD Database study, (2) adults on placebo in the Pioglitazone vs Vitamin E vs Placebo for the Treatment of Non-Diabetic Patients with Nonalcoholic Steatohepatitis (PIVENS) trial, and (3) adults on placebo in the Farnesoid Â Receptor Ligand Obeticholic Acid in NASH Treatment (FLINT) trial.

Table 1 .
Progression or Regression of NASH and Fibrosis Over Time B Consistent nondrinkers: 22% B Modest drinker to nondrinker: 17% B Nondrinker to modest drinker: 13% B Consistent modest drinker: 11% In logistic regression models adjusted for sex, age, race, and smoking history, consistent modest drinkers were significantly less likely to have NASH resolution than consistent nondrinkers (adjusted OR 0.32, 95% CI 0.11-0.92,P ¼ .04)Median 3.8 y between paired biopsies for mixed population At follow-up, 7 patients with NAFL progressed to NASH and 5 patients with NASH no longer met NASH criteria (NAS !5)There was a 50% fibrosis progression in patients with NASH: B At baseline, 80% of patients had a fibrosis score of 2, and 20% had a fibrosis score of 3/4

Table 1 .
Continued AE 3.8 y follow-up for full NAFLD cohort e Despite having more patients with advanced liver disease, survival and rate of first cirrhosis decompensation were similar for patients with NASH vs indeterminate NASH or NAFL, respectively: B Survival: 2.7% vs 1.4% vs 4.6% B Rate of first cirrhosis decompensation: 2% vs 1.7% vs 2.1% Progression to cirrhosis was significantly higher for NASH and indeterminate NASH vs NAFL (9.5% vs 7.9% vs 5%, respectively); this translated to a higher annual incidence of progression to cirrhosis with NASH and indeterminate NASH vs NAFL Annual incidence of progression to cirrhosis increased with increasing fibrosis severity ALD, alcoholic liver disease; CC, cryptogenic cirrhosis; CI, confidence interval; CRN, Clinical Research Network; HCV, hepatitis C virus; LT, liver transplantation; MELD, Model for End-Stage Liver Disease; NAFL, nonalcoholic fatty liver; NAFLD, nonalcoholic fatty liver disease; NAS, NAFLD Activity Score; NASH, nonalcoholic steatohepatitis; OR, odds ratio; RCT, randomized controlled trial; UNOS, United Network for Organ Sharing.

Table 2 .
Changes in NASH and Fibrosis in Response to Interventions

Table 2 .
Continued Resolution of NASH as disappearance of ballooning (score 0), together with either disappearance of lobular inflammation or the persistence of mild lobular inflammation only (score 0 or 1), and resulting in an overall pathologic diagnosis of either steatosis alone or steatosis with mild inflammation.
cChange by !1 point.d e The cencricriviroc program has been canceled after early termination of AURORA (NCT03028740) due to lack of efficacy.

Table 3 .
Predictors of NASH/Fibrosis Progression or Resolution Change in HbA 1c (OR 0.39; P < .01),HOMA-IR (OR 0.88; P ¼ .01),platelets count (OR 1.22; P < .01),NFS (OR 0.27; P < .01)Change in liver enzymes (ALT, AST) and other fibrosis models (APRI, FIB-4) were not associated with fibrosis improvement even after adjustments by improvements in other histologic outcomes BAge at baseline (OR 0.94; P ¼ .04)BNormal levels of ALT at EOT (<19 U/L for women and <30 U/L for men; OR 9.7; P < .01)BInpatients with confirmed NASH, consistent modest drinkers had significantly lower odds of NASH resolution vs consistent nondrinkers (22% vs 11%; adjusted OR 0.32, 95% CI 0.11-0.92;P ¼ .04)b,c Modest drinkers who reported nondrinking at follow-up had the second highest proportion with NASH resolution (17%) Changes in lobular inflammation, hepatocyte ballooning, portal inflammation, fibrosis stage, and NAS were not significantly different between consistent drinkers and nondrinkers In patients with bridging fibrosis at baseline, fibrosis regression was associated with: B Baseline Ishak fibrosis stage (4 vs 3; P ¼ .040)B Lower baseline hepatic collagen content (P ¼ .044)B Greater reduction in hepatic collagen over time (P < .0001)B Lower a-SMA expression at baseline (P ¼ .007)B a-SMA reduction at wk 96 (P < .001)B Lower baseline ELF, NFS, FIB-4, and APRI (all P .013)B Smaller increase in LOXL2 over time (P ¼ .001)In patients with compensated cirrhosis at baseline, cirrhosis regression was associated with: B Baseline Ishak fibrosis stage (6 vs 5; P ¼ .034)B Lower levels of a-SMA expression at baseline (P ¼ .029)B Smaller increases in a-SMA expression or hepatic collagen content over time (both P < .001)

Table 3 .
Continued In the overall NAFLD population, independent predictors worsening fibrosis from baseline to follow-up were: B Intralobular fibronectin >1 (OR 14.1; P < .001)B Hypertension (OR 4.8; P ¼ .028)B HOMA-IR score >10 (OR 1.9; P ¼ .004)These factors also correlated with more advanced fibrosis In the overall NAFLD population, APRI score increased with fibrosis stage (no/mild fibrosis: 0.7; advanced fibrosis: 1.54 Lower NFS was also associated with more advanced fibrosis (P < .01),but AST:ALT ratio was not significantly associated with fibrosis stage (P ¼ .06)APRI was superior to AST:ALT ratio and comparable to NFS

Table 3 .
Continued By multivariate analysis, j the odds of fibrosis progression were: B 17% higher with each year since NASH diagnosis (P < .001)B 41% lower for women than men (P ¼ .016)B 131% higher for smokers vs nonsmokers (P ¼ .004)B 89% higher for those with obesity (P ¼ .002)B 395% higher when a liver transplant was proposed at diagnosis vs not proposed (P ¼ .002)Compared with those in full-time employment, the odds of fibrosis progression were: B 75% higher for those with part-time employment (P ¼ .046)B 106% higher for patients who were retired (P ¼ .045)B 108% higher for those who are unemployed (P ¼ .039)B 2563% higher for those who are physically unable to work due to NASH or related complications (P ¼ .002)k a-SMA, alpha-smooth muscle actin; ACE2, angiotensin-converting enzyme 2; ALT, alanine aminotransferase; APRI, AST-to-platelet ratio index; AST, aspartate aminotransferase; AUC, area under the curve; BMI, body mass index; CI, confidence interval; CRN, Clinical Research Network; ELF, enhanced liver fibrosis; EOT, end of treatment; FGF21, fibroblast growth factor 21; FIB-4, fibrosis-4; GA, glycated albumin; GGT, gamma-glutamyl transferase; HbA 1c , glycated hemoglobin; HCC, hepatocellular carcinoma; HCV, hepatitis C virus; HDL-C, high-density lipoprotein cholesterol; HOMA-IR, homeostasis model assessment of insulin resistance; HR, hazard ratio; HVPG, hepatic venous pressure gradient; INR, international normalized ratio; LOXL2, lysyl oxidase like 2; MELD, Model for End-Stage Liver Disease; MetS-Z, metabolic syndrome severity Z-score; NAFL, nonalcoholic fatty liver; NAFLD, nonalcoholic fatty liver disease; NAS, NAFLD Activity Score; NASH, nonalcoholic steatohepatitis; NFS, NASH fibrosis score; OR, odds ratio; PIVENS, Pioglitazone vs Vitamin E vs Placebo for the Treatment of Non-Diabetic Patients with Nonalcoholic Steatohepatitis; r, correlation coefficient; RCT, randomized controlled trial; T2D, type 2 diabetes; VCAM-1, vascular cell adhesion molecule-1; WHVP, wedge hepatic venous pressure.a Participants were drawn from 3 groups: (1) the adult NAFLD Database study, (2) adults on placebo in the Pioglitazone vs Vitamin E vs Placebo for the Treatment of Non-Diabetic Patients with Nonalcoholic Steatohepatitis (PIVENS) trial, and (3) adults on placebo in the Farnesoid Â Receptor Ligand Obeticholic Acid in NASH Treatment (FLINT) trial.

Table 4 .
Incidence of HCC in Patients With NASH

Table 6 .
Overall Survival in Patients With NASH Undergoing Transplantation

Table 6 .
a Population includes patients for whom NASH is primary (n ¼ 1840) and secondary (n ¼ 119) indication for LT.b Survival data were pooled for patients with NASH (n ¼ 46) and CC (diagnosis based on exclusion of all other forms of liver disease; n ¼ 37).c Final follow-up was January 2011, approximately 4 years after study initiation.d

Table 7 .
NASH Mortality Outcomes in Non-LT Patients