La valutazione non invasiva delle Epatopatie Biliari Autoimmuni Marco Carbone, MD, PhD Divisione di Gastroenterologia Università di Milano Bicocca Milano UOC Gastroenterologia Ospedale Universitario San Gerardo ASST Monza
Conflict of interest statement COMPANY NAME Intercept Pharma Gilead Tobira therapeutics Menarini RELATIONSHIP Consultant, Advisory Board, Travel grant Investigator, travel grant Investigator Travel grant
PBC PSC
Disease course of cholestatic liver diseases Intact bile ducts
PBC
Primary Biliary Cirrhosis Cholangitis Definition Chronic, cholestatic liver disease characterized by non suppurative granulomatous cholangitis; duct destruction and ductopenia, and portal fibrosis that progresses slowly to biliary cirrhosis. Putative autoimmune patogenesis Diagnosis AMA and PBC specific ANA in a patient with otherwise unexplained elevation of ALP is diagnostic of PBC. A liver biopsy is not essential for the diagnosis of PBC, except for seronegative PBC. Disease course: The rate of disease progression in PBC is highly variable; majority have indolent disease. Therapy: UDCA first line (20 40% non response) OCA second line (FDA & EMA approval)
Flow chart of management of PBC Grading & Staging 1 year YES Assessment of high- vs low-risk patients 20-40% NO Biopsy? EASL CPG J Hepatol 2017
PBC is an heterogeneous disease Premature ductopenic variant Portal hypertensivetype Fastprogressors UDCA nonresponse PBC AIH overlap Hepatocellular failure type
Unmet needs in PBC CHALLENGES IN PROGNOSIS Biomarkers for high risk disease at diagnosis Predictive markers of response to UDCA CHALLENGES IN CARE DELIVERY Implementation of stratification criteria in practice enabling identification of high risk patients Management location and approach for low risk patients CHALLENGES IN TRIAL DESIGN Consensus stratification protocols to select high risk patients for second line therapy trials Surrogate markers of outcome as endpoints acceptable to regulatory bodies Dyson J et al. J Hepatol 2015
Current needs for non invasive tests in PBC Diagnosis: AMA M2 highly specific Disease activity: Alk Phos well established (enough?) Fibrosis: preliminary data on ELF and elastography Ductopenia: no markers proposed Treatment response: Alk Phos & bilirubin dismal marker Outcome: ALP and bilirubin
Liver biopsy in PBC Features Portal inflammation Ductopenia Copper associated proteins Interface hepatitis Limitations sampling error Inter observer variation
Standard serum liver tests
Prediction of survival Treatment response criteria Criteria Definition No patients Evaluation time point c statistics at 5, 10, or 15 years * Barcelona, 2006 >40% decrease of ALP or normalization 192 1 year 0.56, 0.61, 0.61 Paris I, 2008 ALP<3xULN, AST<2xULN and bilirubin 1mg/dL 292 1 year 0.81, 0.81, 0.80 Rotterdam, 2009 Normalization of abnormal bilirubin and/or albumin 375 1 year NA Toronto, 2010 ALP 1.67xULN 69 2 year 0.65, 0.70, 0.70 Paris II, 2011 ALP 1.5xULN, AST 1.5xULN and bilirubin 1mg/dL 165 1 year 0.75, 0.75, 0.74 Ehim, 2011 70% decrease of GGT 138 6 months NA * c-statistics calculated in the UK-PBC Research Cohort.
Treatment response criteria UDCA response is a strong predictor of long term survival Carbone M, Mells GF et al. Gastroenterology 2013
Continous risk scores Fitted lines derived from the best fitting multivariable fractional polynomial model UK PBC cohort Cubic spline function GLOBE cohort Hazard ratio 0.5 1 2 3 0.5 1 2 3 Hazard ratio Hazard ratio 0.5 1 2 3 0 1 2 3 4 5 ALP x ULN after 12 months of UDCA 0 1 2 3 4 5 Transaminases x ULN after 12 months of UDCA 0 1 2 3 4 5 Bilirubin x ULN after 12 months of UDCA 0.5 1 2 3 Hazard ratio Hazard ratio 0.5 1 2 3.5 1 1.5 2 2.5 Albumin x LLN 0 1 2 3 4 5 Platelet count x LLN Carbone M, et al. Hepatology 2016 Lammers et al. Gastroenterology 2014
The UK PBC Risk Score Algorithm = 1 baseline survival function^exp(.0287854*(alp12 xuln 1.722136304).0422873 * (((altast12xuln/10)^ 1)28.675729006) 11.4199 *(ln(bil12xuln/10) 12.709607778)21.960303*(albxlln 1.17673001).4161954*(pltxlln 1.873564875)) Baseline survivor function: 0. 982 (at 5 years); 0. 941 (at 10 years); 0.893 (at 15 years) Carbone M et al. Hepatology 2016
The PBC Risk Scores Tool for disease management to identify high risk patients for closer monitoring and second line therapies, as well as low risk patients who require infrequent monitoring and might even be followed up in primary care. Stratification of patients in clinical trials Surrogate endpoint measure in clinical trials
Response guided approach Elastography Gp-210 YES UK-PBC risk score Globe score 20-40% NO Biopsy
Response guided approach Elastography Gp-210 YES UK-PBC risk score Globe score 20-40% NO Biopsy
Baseline prediction of treatment response to UDCA AIM Identification of BASELINE variables predicting likelihood of UDCA response and model development METHODS Study population: 3062 (UK) patients who received UDCA within 1 2 years from the diagnosis Endpoints: UDCA response (defined as ALP<1.67 x ULN); Entire cohort used for model derivation & cross validation External validation (Italian cohort=450)
Probability of UDCA response ALP BILIRUBIN Manuscript submitted
Probability of UDCA response AGE Manuscript submitted
Probability of UDCA response ALT Manuscript submitted
Model of UDCA response Variable (variable scale) Parameter Estimate Standard error Wald statistic p-value Intercept 0.774 0.425 ALP (IU/L) * (Log scale) Bilirubin (umol/l) * (Inverse square root scale) TA (IU/L) * (Log scale) Age (years) (Original scale) Interval Diagnosis-UDCA start (years) (Original scale) ALP delta (IU/L) Original scale -2.730 0.138-19.765 <0.0001 0.600 0.165 3.637 0.0003 0.350 0.108 3.236 0.0012 0.028 0.006 5.074 <0.0001-0.154 0.035-4.362 <0.0001-0.557 0.073-7.588 <0.0001 Manuscript submitted
Model of UDCA response Probab ility level Event Correct Incorrect Percentage Event Nonevent Nonevent Correct Sensitiv ity Specific ity False POS False NEG 0.05 364 4 125 1 74.5 99.7 3.1 25.6 20.0 0.10 361 7 122 4 74.5 98.9 5.4 25.3 36.4 0.15 358 13 116 7 75.1 98.1 10.1 24.5 35.0 0.20 353 24 105 12 76.3 96.7 18.6 22.9 33.3 0.25 350 34 95 15 77.7 95.9 26.4 21.3 30.6 0.30 345 40 89 20 77.9 94.5 31.0 20.5 33.3 0.35 339 45 84 26 77.7 92.9 34.9 19.9 36.6 0.40 331 51 78 34 77.3 90.7 39.5 19.1 40.0 0.45 327 59 70 38 78.1 89.6 45.7 17.6 39.2 0.50 317 68 61 48 77.9 86.8 52.7 16.1 41.4 0.55 306 79 50 59 77.9 83.8 61.2 14.0 42.8 0.60 288 84 45 77 75.3 78.9 65.1 13.5 47.8 0.65 269 92 37 96 73.1 73.7 71.3 12.1 51.1 0.70 251 101 28 114 71.3 68.8 78.3 10.0 53.0 0.75 225 105 24 140 66.8 61.6 81.4 9.6 57.1 0.80 200 113 16 165 63.4 54.8 87.6 7.4 59.4 0.85 264 121 8 201 57.7 44.9 93.8 4.7 62.4 0.90 101 127 2 264 46.2 27.7 98.4 1.9 67.5 0.95 39 129 0 326 34.0 10.7 100.0 0.0 71.6 Manuscript submitted
Ductular reaction and intermediate hepatocytes predict treatment failure (n=20) With the permission of G. Carpino, V. Cardinale, Gaudio E, Alvaro D. Sapienza University, Rome Manuscript submitted
Response guided approach Elastography Gp-210 YES 20-40% NO UK-PBC risk score Globe score Biopsy?
Baseline risk guided approach Elastography UDCA treatment Gp-210 failure score YES 20-40% NO UK-PBC risk score Globe score Biopsy?
Serological profile
AMA M2 no prognostic value Carbone M, Mells GF et al. Gastro 2013 Lammers W et al. Gastro 2014 Anti gp210 6 fold risk of progression to liver failure/transplantation Wesierska Gadek & Invernizzi P et al. Hepatology 2006 Nakamura et al. Hepatology 2007 Anti centromere antibodies associate with PH Nakamura et al. Hepatology 2007
Anti gp210 and anti centromere Wesierska Gadek & Invernizzi P et al. Hepatology 2006 Nakamura et al. Hepatology 2007
Disease trajectories based on antibody profile The Japan Society of Hepatology. Hepatol Research 2007
Markers of fibrosis
Transient elastography N=150, single center and UDCA treated Corpechot C. Hepatol 2012
Transient elastography N=150, single center and UDCA treated Corpechot C. Hepatol 2012
Transient elastography Corpechot C. Hepatol 2012
Transient elastography Corpechot C. Hepatol 2012
Transient elastography No external validation yet Unclear whether adds predictive value to biochemical response Neglect presence of cholestasis, ductopenia, interface hepatitis
Enhanced Liver Fibrosis (ELF) Study cohort=161 Multicenter national data extrapolated from a clinical trial of MTX+UDCA AUROC 0.76 AUROC 0.75 Mayo M. Hepatol 2008
Enhanced Liver Fibrosis (ELF) Mayo M. Hepatol 2008
Enhanced Liver Fibrosis (ELF) Unclear whether adds predictive value to biochemical response to UDCA Unclear impact of longitudinal stability vs. fluctuation over time No external validation yet Marker of disease severity or stage?
AST/platelet ratio index Derivation cohort=386 (UK); Validation cohorts: 479 (Germany) and 150 (Canada) Trivedi P. Gut 2016
AST/platelet ratio index Elevated APRI is associated with future risk of adverse events, independently and additively of UDCA response Possible future surrogate endpoint in clinical trials in PBC
PSC
Primary Sclerosing Cholangitis PSC is a challenging illness that is characterised by chronic bile duct destruction and progression to end stage liver disease Incidence varies geographically (1.3 x 100,000 in Northern Europe ) 80% men, median age 30 40 60 80% of cases in northern European populations are associated with inflammatory bowel disease Unknown pathogenesis No effective medical treatments are available Consistent risk of developing colangioca and Colorettal Ca
Proposed pathway of PSC management Trivedi P et al. Hepatology 2015
Current needs for non invasive tests in PSC Diagnosis: MRI cholangiography +/- biopsy Disease activity: no markers accepted, imaging required Cancer screening: no markers accepted Outcome: lacking
Standard serum liver tests
Can Biochemical Surrogates Be Extrapolated to PSC? ALP based biochemical response criteria Does PSC biochemical responders in clinical trials mean improved survival?
Amsterdam prognostic model Derivation cohort (Amsterdam)=692; validation (Oxford)=264 Model: PSC subtype, age at diagnosis, albumin, platelets, aspartate aminotransferase, alkaline phosphatase and bilirubin C statistics=0.64 De Vries E. Gut 2017
Markers of fibrosis
Transient elastography Single center cohort=167 (France) Corpechot et al. Gastro 2014
Transient elastography Neglect impact of cholestasis/cholangitis, IBD activity External validation pending (FICUS ongoing) Corpechot et al. Gastro 2014
Enhanced Liver Fibrosis (ELF) Derivation cohort=167; Validation cohorts: 138 (Norway) Vesterhus M et al. Hepatology 2015
Enhanced Liver Fibrosis (ELF) ELF cannot be benchmarked disease affection is even more patchy than in PBC, and it is probable that the performance of the ELF test may surpass liver biopsy Internal validation but short disease duration in LTfree survivors (<5 years) Impact of longitudinal stability vs. fluctuations not clear Stratifier of disease severity vs. stage? Vesterhus M et al. Hepatology 2015
Serological profile
Autoantibodies ANCA low sensitivity and specificity Positive correlation between serum levels of anticardiolipin antibody and disease severity measured by MRS and histologic stage of disease was shown
Serum IgG4 Between 9% and 15% of PSC patients have raised serum IgG4 Unclear impact on survival IgG4 >1.4 and <2.8 g/l, IgG4/IgG1 ratio with a cutoff at 0.24 is correlate to IgG4 colangiopathy Stratifying properties of serum IgG4 in PSC remain unsubstantiated and require further evaluation Boonstra K et al. Hepatology 2014 Mendes F et al. Am J Gastro 2006 Benito de Valle et al. DLD 2014
Future potential stratifiers
Genome wide time to event analysis (N=1600) SNPs at 6 loci achieved P DISCOVERY <5 10 6 Of these, none identified in GWAS of susceptibility of PBC Deviation in the tail of the distribution is suggestive of true associations CHR SNP A1 EST SE P-value Genes 7p22 rs12666575 G 0.38 0.07 1.32 10-07 MAD1L1,MIR4655,FTSJ2,NUDT1,SNX8 7q31 rs4727713 G 0.53 0.12 8.93 10-06 NRCAM,PNPLA8,RPL7P32,THAP5,DNAJB9 8p22 rs1378029 A 0.34 0.07 2.79 10-06 SGCZ 8q21.3 rs10429299 A 0.39 0.08 7.12 10-07 SOX5P,LOC100419762,DCAF4L2,MMP16 11p15.5 rs1004446 G -0.35 0.08 4.76 10-06 INS-IGF2,IGF2, MIR483,IGF2-AS,INS,TH,MIR4686 12q22-23 rs34580 A 0.45 0.09 1.91 10-06 Intergenic Mells GF & Carbone M Unpublished data
Metabolomics Bell et al. Liver Int 2015
Serum micrornas as novel biomarkers for PSC and CCA PSC CCA Bernuzzi F, Invernizzi et al. Clin Exp Immunol 2016
Summary A stratified management of PBC and PSC represents a major unmet clinical need in hepatology Development/improvement of (non invasive) markers is needed in PBC and PSC for the assessment of: stage of disease treatment response disease progression/outcome The rise of data intensive biology, advances in information technology and the availability of large scale cohort offer a unique opportunity for developing a stratified medicine in the field
Grazie CENTRO MALATTIE AUTOIMMUNI DEL FEGATO Pietro INVERNIZZI Francesca BERNUZZI Federica MALINVERNO Vincenzo RONCA Marta GEMMA Alessio GERUSSI Laura CRISTOFERI Giulia BONATO George F MELLS Richard R SANDFORD Steven FLACK Lynda SMITH Dave E JONES UK PBC Consortium Abenavoli L Almasio P Alvaro D Andreone P Andriulli A Azzaroli F Baiocchi L Battezzati PM Bollani S Calvaruso V Cardinale V Colleredo G Coco B Colombo Crocè L Donato M Fabris L Floreani A Frugiuele Galli A Giannini E Labbadia G Lleo A Marra F Marzioni M Mattalia A Miele L Missale G Morisco F Muratori L Portincasa P Picciotto A Rigamonti C Rosina F Spinzi G Toniutto P Valenti L Zuin M
Active and recently run clinical trials in PBC Carbone M, Mells GF et al. Dig Liv Dis IN PRESS
PBC is an heterogeneous disease Variant syndromes Autoantibody profile Symptom profile PBC AIH overlap syndrome may be found in ~10% and the premature ductopenic variant in ~5% of cases Anti centromere antibodies (ACA) are found in ~30%, anti sp100 antibodies in ~20 30% and anti gp210 antibodies in ~10% of cases Pruritus is present in 40% and fatigue is present in 45% of cases Modes of disease progression Rate of disease progression The biochemical response to UDCA Portal hypertensive type versus hepatocellular failure type progression Ranging from no overt progression at one end of the spectrum, to ESLD occurring within a few years of diagnosis, at the other Variable; it strongly predicts the long term outcome.
Clinical scenarios showing the effect of the transaminasis on the chance to respond to UDCA therapy based on the ALP and total bilirubin levels. Recipient age ALP ratio Bilirubin ratio Ta ratio Delta ALP ratio Diag- UDCA time Estimated probability of responese 95% C.I. 50 1 0.5 0.5 0 0 0.94 (0.92, 0.96) 50 1 0.5 1 0 0 0.95 (0.94, 0.96) 50 1 0.5 2 0 0 0.96 (0.95, 0.97) 50 1 0.5 3 0 0 0.97 (0.95, 0.98) 50 1 0.5 4 0 0 0.97 (0.96, 0.98) 50 2 0.5 0.5 0 0 0.71 (0.65, 0.76) 50 2 0.5 1 0 0 0.76 (0.72, 0.79) 50 2 0.5 2 0 0 0.80 (0.67, 0.82) 50 2 0.5 3 0 0 0.82 (0.79, 0.85) 50 2 0.5 4 0 0 0.83 (0.79, 0.87) 50 2 1 0.5 0 0 0.65 (0.58, 0.73) 50 2 1 1 0 0 0.71 (0.66, 0.75) 50 2 1 2 0 0 0.75 (0.72, 0.79 50 2 1 3 0 0 0.78 (0.74, 0.82) 50 2 1 4 0 0 0.80 (0.75, 0.84) 50 2 2 0.5 0 0 0.61 (0.52, 0.70) 50 2 2 1 0 0 0.67 (0.60, 0.73) 50 2 2 2 0 0 0.72 (0.66, 0.77) 50 2 2 3 0 0 0.75 (0.69, 0.80) 50 2 2 4 0 0 0.77 (0.70, 0.82) 50 2 2 4 0 0 0.78 (0.72, 0.83) Manuscript submitted
Ductopenic variant
Primary Biliary Cholangitis
Challenges in PBC Treatment in high risk patients (beyond UDCA) PRESENT: non responder to UDCA as 1 st line therapy FUTURE: more proactive, don t waste liver & bile duct time, aim at LFTs normalisation, choleretic as backbone Young female with premature ductopenic variant Male patients with HCC (even despite UDCA response) Patients with advanced fibrosis and risk of disease progression Optimize treatment (in non responders) Now several therapeutic stategies will be available Opportunities for combination (FXR + PPAR?) How to bring all of this into clinical reality by efficient trial strategy Classic playground for P5 medicine
External validation of the UK PBC Risk Score China (N=223) Switzerland (N=501) Poster FRI-379
Dynamic risk prediction in PBC Bilirubin 12 (log-scale) Delta ALP after treatment Poster SAT-377
The PBC Risk Scores Algorithm = 1 baseline survival function^exp(.0287854*(alp12 xuln 1.722136304).0422873 * (((altast12xuln/10)^ 1)28.675729006) 11.4199 *(ln(bil12xuln/10) 12.709607778)21.960303*(albxlln 1.17673001).4161954*(pltxlln 1.873564875)) Baseline survivor function: 0. 982 (at 5 years); 0. 941 (at 10 years); 0.893 (at 15 years) Tool for disease management to identify high risk patients for closer monitoring and second line therapies, as well as low risk patients who require infrequent monitoring and might even be followed up in primary care. Stratification of patients in clinical trials Surrogate endpoint measure in clinical trials
Cholangiographic Stratification Annual three dimensional MRC to score liver parenchymal appearances, PH and bile duct lesions predicted radiological progression from baseline with high accuracy (AUROC, >0.8) Ruiz A et al. Hepatology 2014 Dominant strictures natural history data are restricted to specialist centers, with reduced survival Small duct PSC disease progression is relatively infrequent, occurring over a longer time period reflecting difficulties in CCA recognition
EASL CPG J Hepatol 2017
Interaction between ALT * Bilirubin on survival Unpublished data