Background Non-invasive tests that can identify patients with non-alcoholic steatohepatitis (NASH) at higher risk of disease progression are lacking. We report the development and validation of a blood-based diagnostic test to no-ninvasively rule in and rule out at-risk NASH (defined as non-alcoholic fatty liver disease [NAFLD] activity score [NAS] >= 4 and fibrosis stage >= 2). Methods In this prospective derivation and global validation study, blood samples, clinical data, and liver biopsy results from three independent cohorts with suspected NAFLD were used to develop and validate a non-invasive blood-based diagnostic test, called NIS4. Derivation was done in the discovery cohort, which comprised 239 prospectively recruited patients with biopsy-confirmed NASH (NAFLD NAS >= 3; fibrosis stage 0-3) from the international GOLDEN-505 phase 2b clinical trial. A complete matrix based on 23 variables selected for univariate association with the presence of at-risk NASH and avoiding high multi-collinearity was used to derive the model in a bootstrap-based process that minimised the Akaike information criterion. The overall diagnostic performance of NIS4 was externally validated in two independent cohorts: RESOLVE-IT diag and Angers. The RESOLVE-IT diag cohort comprised the first 475 patients screened for potential inclusion into the RESOLVE-IT phase 3 clinical trial. Angers was a retrospective cohort of 227 prospectively recruited patients with suspected NAFLD and clinical risk factors for NASH or fibrosis stage 2 or more according to abnormal elastography results or abnormal liver biochemistry. Both external validation cohorts were independently analysed and were combined into a pooled validation cohort (n=702) to assess clinical performance of NIS4 and other non-invasive tests. Findings The derived NIS4 algorithm comprised four independent NASH-associated biomarkers (miR-34a-5p, alpha-2 macroglobulin, YKL-40, and glycated haemoglobin; area under the receiver operating characteristics curve [AUROC] 0.80, 95% CI 0.73-0.85), and did not require adjustment for age, sex, body-mass index (BMI), or aminotransferase concentrations. Clinical cutoffs were established within the discovery cohort to optimise both rule out and rule in clinical performance while minimising indeterminate results. NIS4 was validated in the RESOLVE-IT diag cohort (AUROC 0.83, 95% CI 0.79-0.86) and the Angers cohort (0.76, 0.69-0.82). In the pooled validation cohort, patients with a NIS4 value less than 0.36 were classified as not having at-risk NASH (ruled out) with 81.5% (95% CI 76.9-85.3) sensitivity, 63.0% (57.8-68.0) specificity, and a negative predictive value of 77.9% (72.5-82.4), whereas those with a NIS4 value of more than 0.63 were classified as having at-risk NASH (ruled in) with 87.1% (83.1-90.3) specificity, 50.7% (45.3-56.1) sensitivity, and a positive predictive value of 79.2% (73.1-84.2). The diagnostic performance of NIS4 within the external validation cohorts was not influenced by age, sex, BMI, or aminotransferase concentrations. Interpretation NIS4 is a novel blood-based diagnostic that provides an effective way to non-invasively rule in or rule out at-risk NASH in patients with metabolic risk factors and suspected disease. Use of NIS4 in clinical trials or in the clinic has the potential to greatly reduce unnecessary liver biopsies in patients with lower risk of disease progression.

Background and Aim: T2D increases the prevalence of steatosis and steatohepatitis (NASH). The aim was to assess the relationship between T2D and two histological features of liver injury defining the need for therapy: disease activity grade (NAS) and fibrotic scar (F). Methods: 2363 patients with metabolic steatosis were screened for the phase 3 elafibranor trial RESOLVE-IT (NTC02704403) with a liver biopsy centrally scored for NAS and F. 71% had obesity, 47% dyslipidemia, 81% MetS. 35% had T2D, 88% of them managed by ≥1 OAD or insulin. Prevalence of active disease (NAS≥4), significant fibrosis (F≥2) and progressive NASH (NAS≥4 with F≥2, i.e., "at-risk" of hepatic outcomes) and clinical and biochemical NASH phenotype was assessed according to the T2D status. Results: The overall prevalence of "at-risk" NASH was 53%, 65% for T2Ds (n=835), and 45% for non-T2Ds (n=1528). T2D was a significant risk factor (p<0.001) for “at-risk” NASH (OR=2.20, 95% CI [1.85;2.62]), NAS≥4 (OR=1.74, 95% CI [1.40;2.16]) and F≥2 (OR=2.50, 95% CI [2.08;3.01]). The association between T2D and “at risk” NASH was independent of age, gender or BMI. T2D patients had higher mean NAS scores (4.96 vs. non-T2Ds: 4.47; p<0.001) and higher mean fibrosis stage (2.17 vs. non-T2D: 1.67; p<0.001). Liver injury markers ALT, AST, GGT or CK18 were higher (p<0.001) in “at-risk” NASH patients but were similar in T2Ds vs. non-T2Ds. Fibrosis serum markers such as TIMP-1, HA, A2M, or P3NP were elevated in “at-risk” NASH patients (p<0.001) and were higher in T2Ds vs. non-T2Ds (p<0.001). In both non-T2D and T2D, HbA1c and glucose homeostasis markers remained higher in “at-risk” NASH (p<0.001). Conclusion: In patients with metabolic steatosis, T2D is associated with necroinflammation and fibrosis and independently increases the risk of developing “at-risk” NASH, highlighting the need for active surveillance of liver injury in T2D patients to identify those in need for clinical intervention. Disclosure B. Staels: None. V. Ratziu: Other Relationship; Self; Gilead Sciences, Inc. A.J. Sanyal: Consultant; Self; Ardelyx, Boehringer Ingelheim Pharmaceuticals, Inc., Exhalenz, Gilead Sciences, Inc., Hemoshear, Intercept Pharmaceuticals, Inc., Lilly, Mallinckrodt Pharmaceuticals, Nimbus Therapeutics, Nitto Denko, Novartis Pharmaceuticals Corporation, Pfizer Inc. Employee; Self; Sanyal Bio. Research Support; Self; Bristol-Myers Squibb Company, Conatus Pharmaceuticals Inc., Echosens, Galectin Therapeutics Inc., Immuron Ltd, Merck & Co., Inc., Salix Pharmaceuticals, Sequanna. Stock/Shareholder; Self; Akarna Therapeutics, GENFIT, Natural Shield, NewCo LLC, Tiziana. Other Relationship; Self; Elsevier, UpToDate. S. Francque: Consultant; Self; GENFIT. S.A. Harrison: Advisory Panel; Self; Gilead Sciences, Inc. A. Roudot: Employee; Self; GENFIT. J. Brozek: Employee; Self; GENFIT. Y. Hajji: Employee; Self; GENFIT. S. Megnien: Employee; Self; GENFIT. D.W. Hum: Employee; Self; GENFIT. P. Birman: Employee; Self; GENFIT. R. Hanf: Employee; Self; GENFIT. B. Cariou: Board Member; Self; Novo Nordisk A/S, Regeneron Pharmaceuticals. Consultant; Self; GENFIT, Sanofi-Aventis. Research Support; Self; Amgen Inc., Pfizer Inc. Speaker’s Bureau; Self; Abbott, Akcea Therapeutics, Merck Sharp & Dohme Corp.

Background and Aim: T2D is an independent risk factor for the development of nonalcoholic steatohepatitis (NASH). Amongst NASH patients, those with elevated disease activity (NAS≥4) and fibrosis (F≥2) upon scoring of a liver biopsy are at higher risk of progressing to negative clinical outcomes. There is a lack of diagnostic tools to identify these “at-risk𠇍 NASH patients that need clinical intervention. Here we compare the ability of NIS4, a novel blood test, to other available tests for detection of “at-risk” NASH in a T2D population. Methods: The blood and liver biopsy samples from 714 patients (446 non-T2D + 268 T2D) with metabolic risk factors for NASH were used. Disease activity (NAS) and fibrosis stage (F) were evaluated by a single expert pathologist. “At-risk” NASH was defined as NASH with NAS≥4 and F≥2. The diagnostic performance of NIS4 to identify patients with “at-risk” NASH in T2D subpopulation was assessed through ROC analysis and compared versus existing tests and statistical comparisons were done by DeLong testing. Results: In the total cohort (T2D prevalence = 38%), AUROC of NIS4 was 0.83 [95% CI: 0.80 - 0.86] for identifying “at-risk” NASH. The prevalence of “at-risk” NASH was higher in T2D (61%) vs. non-T2D (45%). AUROC was comparable in T2D and non-T2D: 0.80 [0.75 - 0.85] vs. 0.83 [0.80 - 0.87]. In T2D, a head to head comparison (AUROC; 95% CI) showed that NIS4 [0.80; 0.75-0.85] significantly outperformed all other tests for the identification of “at-risk” NASH: APRI [0.74; 0.68-0.80], FIB4 [0.70; 0.64-0.77], ELF [0.70; 0.64-0.77], FibroTest [0.68; 0.61-0.74], and NFS [0.60; 0.52-0.67]. Conclusion: Amongst T2D patients with suspected NASH, the prevalence of “at-risk” NASH is high. This highlights the need for active surveillance amongst T2D patients. For that purpose, NIS4 outperforms other tests and is a promising non-invasive tool to identify and potentially track T2D patients who need clinical intervention to manage their disease. Disclosure B. Staels: None. V. Ratziu: Other Relationship; Self; Gilead Sciences, Inc. S. Francque: Consultant; Self; GENFIT. S.A. Harrison: Advisory Panel; Self; Gilead Sciences, Inc. P. Bedossa: Consultant; Self; GENFIT. A. Roudot: Employee; Self; GENFIT. Z. Majd: Employee; Self; GENFIT. J. Brozek: Employee; Self; GENFIT. F. Ben-Sudrik: Employee; Self; GENFIT. P. Birman: Employee; Self; GENFIT. D.W. Hum: Employee; Self; GENFIT. S. Hosmane: Employee; Self; GENFIT. P. Chaumat: Employee; Self; GENFIT. R. Hanf: Employee; Self; GENFIT. B. Cariou: Board Member; Self; Novo Nordisk A/S, Regeneron Pharmaceuticals. Consultant; Self; GENFIT, Sanofi-Aventis. Research Support; Self; Amgen Inc., Pfizer Inc. Speaker’s Bureau; Self; Abbott, Akcea Therapeutics, Merck Sharp & Dohme Corp. A. Sanyal: Consultant; Self; Ardelyx, Boehringer Ingelheim Pharmaceuticals, Inc., Exhalenz, Gilead Sciences, Inc., Hemoshear, Intercept Pharmaceuticals, Inc., Lilly, Mallinckrodt Pharmaceuticals, Nimbus Therapeutics, Nitto Denko, Novartis Pharmaceuticals Corporation, Pfizer Inc. Employee; Self; Sanyal Bio. Research Support; Self; Bristol-Myers Squibb Company, Conatus Pharmaceuticals Inc., Echosens, Galectin Therapeutics Inc., Immuron Ltd, Merck & Co., Inc., Salix Pharmaceuticals, Sequanna. Stock/Shareholder; Self; Akarna Therapeutics, GENFIT, Natural Shield, NewCo LLC, Tiziana. Other Relationship; Self; Elsevier, UpToDate.