We don t need a liver biopsy. We have non-invasive tests

Transcription

1 We don t need a liver biopsy We have non-invasive tests Laurent CASTERA, MD PhD Department of Hepatology, Hôpital Beaujon, Clichy Université Paris Diderot, France PHC

2 Liver biopsy: an unrealistic and risky procedure given the high prevalence of NAFLD

3 Diagnosing NAFLD with liver biopsy: a challenge! Worldwide prevalence of NAFLD 25% of general = more than one billion population liver biopsies!!! still more than One hundred million liver biopsies!!! Younossi Zetetal. al.hepatology Hepatology 2016: Younossi 2016;64; 64: 73-84

4 Diagnosing NAFLD with liver biopsy: a challenge! Patients 3-6 million! Hepatologists # 400

5 Diagnosing NAFLD with liver biopsy: a challenge! Patients 3-6 million! Hepatologists # 400

6 Diagnosing NAFLD with liver biopsy: a challenge! Patients 3-6 million! Pathologists # 40

7 Inter-observer variability Pathologist Experience (Specialization, duration, Specimen Characteristics Size, fibrosis stage academic practice) Rousselet et al. Hepatology 2005; 41:

8 Limitations of liver biopsy Invasive Sampling error Interobserver variability Nondynamic evaluation of fibrosis Regev et al. Am J Gastroenterol 2002; 97: Bedossa et al. Hepatology 2003;38: Rousselet et al. Hepatology 2005; 41:

9 The patient perspective!

10 Trends in liver biopsy practice: HCV vs. HBV The Beaujon Experience HCV HBV Courtesy of Pierre Bedossa

11 Non Invasive Tests: An exponential increase in publications! Source PubMed

12 Non-invasive tests are ready for Prime Time

13 Non-Invasive Tests: recommended by international guidelines

14 Available non-invasive methods 2 different but complementary approaches «Biological» approach Serum Biomarkers «Physical» approach CAP / TE PDFF / MRE

15 Serum Biomarkers vs. FibroScan Summary Serum biomarkers Advantages FibroScan Advantages Good reproducibility High applicability (95%) Low cost & wide availability (non-patented) Disadvantages Genuine property of the liver High performance for cirrhosis User-friendly Disadvantages Non-specific for the liver Performance for cirrhosis Cost & availability (patented) Low applicability (80%) False positive (inflammation) Requires a dedicated device Castera L. Gastroenterology 2012;142:

16 Liver fibrosis is the main prognostic factor Loomba R et al. Gastroenterology 2015; 149:

17 Liver fibrosis is the main prognostic factor Loomba R et al. Gastroenterology 2015; 149:

18 Mortality is related to cirrhosis P<0,001 P<0.001 F1-F2 F3-F4 N=229 NAFLD patients ; f-up 26.4 yrs Ekstedt M et al. Hepatology 2015: 61;

19 TE has high diagnostic accuracy for viral cirrhosis Talwalkar et al. CGH 2007 Friedrich-Rust et al. Gastroenterology 2008 Stebbing et al. APT 2010 Tsochatzis et al. J Hepatol 2011 Chon et al. PLoS ONE 2012

20 TE has high diagnostic accuracy for NAFLD cirrhosis Meta-analysis Sum AUROCS: 0.94 Sum Se: 87% Sum Sp: 86% 13 studies; n= 1780 patients. Xiao et al. Hepatology 2017; 66:

21 Recommendations Cirrhosis diagnosis EASL-ALEH Clinical practice Guidelines. J Hepatol 2015; 63:

22 Use in clinical practice EASL-EASD- EASO CPG. J Hepatol 2016; 64:

23 Non-invasive tests have prognostic value in NAFLD Serum biomarkers Liver stiffness NFS < NFS >0.676 N= 320 NAFLD patients ; F-up 105 mo Angulo et al. Gastroenterology 2013; 145: N= 360 NAFLD patients; f-up 6 yrs Boursier et al. J Hepatol 2016; 65:

24 Screening the general population for NAFLD using non-invasive tests

25 Screening general population for NAFLD

26 The typical NAFLD patient Asymptomatic Low awareness Low risk-perception No approved treatment Tru MP et al. J Fake News 2017; in press

27 Screening General population FibroScan a p value < final model, Charact eri compared ac Fisher exact t hen st rat ifie Age 58 ans and $ 8.0 kpa H 60 % t o t he cat ego BMI 26.5 kg/m2 covariance a Age-/ sex-adju furt her perf regression m All st at ist i soft w are vers St at ist ical sig t ests). RESULTS Subjects F4 0.7% F>2 7.5% Figure 1 Flow chart of the study. After exclusion of subjects with missing data, failure of liver stiffness measurement (LSM) or unreliable A t ot al of 13 liver disease (1.7%) w ere dat a (figure Roulot et al. Gut 2011; 60:

28 Screening General population FibroTest Age 57 ans H 45 % BMI? kg/m2 Fibrosis ( 2) 2.8 % Cirrhosis 0.3% Poynard et al. BMC Gastroenterology 2010; 10: 40

29 ly 12 subject s w it h fat t y liver had significant mpt ion bet w een 140 and 350 g per w eek. Aft er se subjects, t he populat ion prevalence of NAFLD 5% CI 24.5% t o 30.2%). Among subject s w it h fat t y 3%) had an IHTG cont ent of 5e 10.9%, and 110 an IHTG cont ent $ 11.0%, a level suggest ive of c st eat osis by hist ology.2 1 hand, t he percent age of subject s w it h or wit hout a fat t y liver w it h valid examinat ion w as similar (82.2% vs 82.4%; p¼0.95). The liver st iffness w as significant ly higher in subject s w it h fat t y liver t han in cont rols ( kpa vs kpa; p< 0.001) (t able 1). Eight (3.7%) pat ient s w it h fat t y liver and 7 (1.3%) cont rols had liver st iffness $ 9.6 kpa, a level suggest ive of advanced fibrosis (p¼0.032). Screening General population MRI-Spectroscopy/ FibroScan dy recruitment and. Age 48 ans H 42 % BMI 23 kg/m2 NAFLD 29% F3-F4 1% doi: /gutjnl Wong V et al. Gut 2012; 61:

30 Screening Diabetics CAP/ FibroScan N= 1918 Chinese diabetics patients NAFLD (CAP >222 db/m) 73% F3-F4 (LSM >9.6 kpa) 18%. Kwok et al. Gut 2016; 65:

31 Screening Diabetics CAP/ FibroScan N= 435 French diabetics patients NAFLD (CAP >236 db/m) 75% F3-F4 (LSM >9.6 kpa) 7.3%. Roulot et al. Liver Int 2017; 37:

32 Summary. Gines P... Castera L. Lancet Gastroenterol Hepatol 2016; 1:

33 Take Home Message NIT Liver biopsy Triage in large unselected populations Assessment in selected populations or in clinical trials Best used as an integrated system to allow more efficient evaluation of patients with NAFLD