Journal Watch May-June 2016

Progressive familial intrahepatic cholestasis type 3 (PIFIC 3) is caused by variations in the ABCB4 gene which encodes an ATP binding cassette (ABC) transporter protein (also called MDR3). ABCB4 is expressed on the hepatocytic bile canalicular membrane where it mediates secretion of phosphatidylcholine (PC). ABCB4 defects result in low levels of biliary PC with resultant biliary epithelial damage. These investigators evaluated the biologic consequences of 12 ABCB4 variants in 9 patients with PIFIC3. Using site-directed mutagenesis and cell line expression models, the cellular localization of expression and biologic activity of the ABCB4 variants was evaluated. The authors classified the ABCB4 variations into 5 categories based on severity of disease, impact on protein maturity, activity, or stability (i.e., clinical and biological relevance). This proof-of-principle study provides clues for genetically based therapeutic strategies; each category of protein defect suggests a specific therapeutic strategy. This study also demonstrates the possibility of restoring normal cellular function by pharmacological methods.

This is a good review of the advances made and remaining challenges in pediatric NAFLD.

This is a great review which covers aspects ranging from hepatic lipid metabolism, the impacts of the genetic background, diet and physical activity on fatty liver disease.

Bile salt export pump (BSEP) and multidrug-resistance protein 3 (MDR3), both members of the ATP-binding cassette transporter family involved in bile acid homeostasis, are specifically expressed in hepatocytes.  The purpose of this study was to further evaluate the utility of these immunohistochemical markers in retrospective cohorts of select malignancies.  The first cohort consisted of hepatocellular carcinomas (HCCs, n = 54), intrahepatic cholangiocarcinomas (ICCs, n = 34), combined hepatocellular-cholangiocarcinomas (CHCs, n = 23), and hepatoid carcinomas predominantly of gastric origin (n = 27).  For HCCs, BSEP and MDR3 showed similar sensitivities to ARG1 and HepPar-1 in well to moderately differentiated malignancies (all ≥ 94%), but lower sensitivities in poorly differentiated malignancies (e.g. MDR3 = 67% vs. ARG1 = 100%).  No ICCs or hepatoid carcinomas stained for BSEP or MDR3 whereas 9% of ICCs were positive for ARG1 and 44% of hepatoid carcinomas were immunoreactive for HepPar-1.  In CHCs where cholangiocellular areas were sometimes positive for ARG1 (26%) or HepPar-1 (52%), none of these same areas stained for BSEP or MDR3.  The second cohort consisted of primary liver cases originally diagnosed as poorly differentiated carcinoma, NOS (n = 8) due to inconclusive morphologic and immunohistochemical evidence to categorize as HCC or ICC.  In 2 of these cases, either BSEP or MDR3 was positive, suggesting the diagnoses as HCCs.  In conclusion, BSEP and MDR3 appear to be promising adjunctive markers for differentiating HCCs from ICCs and HCCs from hepatoid carcinomas originating from other organs.  

These authors question the validity of grouping HCC containing “stem/progenitor-like” cells into the category of combined hepatocellular-cholangiocarcinoma (CHC) as presented in the WHO Classification of Tumours of the Digestive System (4th ed).  This retrospective case series from a single institution compares clinicopathologic characteristics and prognosis among the typical subtype of CHC with stem cell features [referred to as HCC with reactive ductule-like components (HCC-RD) in this manuscript], classical CHC, and conventional HCC.  Tumors were classified solely on H&E morphology: HCC-RD (n = 46), classical CHC (n = 38), and conventional HCC (n = 483).  Compared to HCC, HCC-RD patients were more often younger (60.3 vs. 64.5 yrs age), female (52.2% vs. 20.5%) and had higher serum AFP (106 vs. 19 ng/mL) and CA19-9 (40.0% vs. 16.9%) levels.  HCC-RD frequently showed clear cell features, interstitial fibrosis, and intratumoral neutrophils compared with HCC.  Both disease free survival and overall survival were similar between HCC-RD and HCC, but worse for CHC compared with HCC (p ≤ 0.032, up to 60 month follow up).  Therefore, despite differences in morphology and some clinical features, the authors propose that based on prognosis, HCC-RD should be considered as an HCC variant rather than a subtype of CHC.  Meanwhile, differences in morphology and prognosis were confirmed for classical CHC.

NAFLD is the most common liver disorder in Western countries and includes two pathologically distinct entities: non-alcoholic fatty liver (NAFL) and non-alcoholic steatohepatitis (NASH). The evidence based Clinical Practice Guidelines is developed from data retrieved by an extensive PubMed search up to April 2015 and include guidelines for NAFLD prevalence, pathogenesis, diagnostic algorithm, and management including pediatric NAFLD. Recent advances in genetic association, non-invasive assessment, and clinical management are also included. The document is intended for both practical use and advancing research.  It should also help improve patient care and awareness of the importance of NAFLD.

Using exosome sequencing and next generation sequencing, recent studies have revealed frequent alterations in several new genes including the three chromatin-remodeling genes (BAP1, ARID1A, and PBRM) in cholangiocarcinomas. The authors investigated the inactivating mutation of ARID1A gene by immunohistochemistry in 13 cHC-CCs, 49 ICCs, 17 IPNBs, 72 EHCC, and 43 GBCs, and correlated with clinical characteristics, KRAS mutation status and survival. Loss of ARID1A expression was found in 14% of biliary carcinomas but 0% of IPNBs. Biliary carcinomas with loss of ARID1A expression were mass forming or flat infiltrating types and histologically tubular adenocarcinomas with abound fibrous stroma. Loss of ARID1A expression was not correlated with staging and patient survival. Interestingly, no biliary carcinomas harbored both ARID1A and KRAS mutations. The study suggests inactivating mutation of ARID1A may be involved in a novel pathway of biliary carcinogenesis, which is different from KRAS pathway in IPNB and BilIN.

Hepatic arterial buffer response is defined as increased hepatic arterial flow secondary to reduced portal flow which has been demonstrated experimentally and surgically. In this report, the pathologic change of hepatic arterial buffer response in 21 patients (12 male and 9 female) with extrahepatic portal vein thrombosis was evaluated. In addition, CD34 staining, outer diameters, luminal diameters and wall thickness of hepatic arteries cut in cross-section and outer diameters of cross-sectioned paired bile ducts were compared with age- and gender-matched controls. Measurements of 280 and 193 arteries from patients and controls demonstrated statistically significant arterial dilatation and arterial wall thinning in patients with extra-hepatic portal vein thrombosis (P<0.05). Diffuse and obvious changes, and portal vein absence or attenuation were seen only in the patient group, although subtle and/or focal dilatation of central veins, portal veins and sinusoids, focal trabecular thinning/thickening and mild ductular reaction were common findings in both groups. Capillarization of sinusoids was not seen on CD34 stain. Two patients showed significant ductular reaction, one of who developed biliary strictures on follow-up. In summary, hepatic arterial dilatation and wall thinning in non-cirrhotic patients with portal vein thrombosis provide histopathologic evidence of hepatic arterial buffer response in the human liver. Diffuse sinusoidal dilatation or absence or attenuation of portal veins is highly suggestive of extrahepatic portal vein thrombosis. Periportal shunt vessels, hypervascular portal tracts, muscularized portal veins, large thick-walled or dilated arteries are rare, but can aid diagnosis. Normal or near-normal biopsies do not rule out portal vein thrombosis.

Rare hepatic adenoma associated with synchronous or metachronous fibrolamellar carcinoma has not been well studied. Four hepatic adenomas co-occurring with or preceding a diagnosis of fibrolamellar carcinoma from three patients were included in this study. Histologically, three of the adenomas showed the typical morphology of HNF1-α inactivated adenomas, and one showed a myxoid phenomenon. Fluorescence in situ Hybridization (FISH) analysis showed negative PRKACA rearrangements in all adenomas. All adenomas displayed complete loss or significant reduction of liver fatty acid binding protein (LFABP) expression immunohistochemically. Interestingly, the fibrolamellar carcinomas in each case also showed loss of LFABP expression. One of the fibrolamellar carcinomas was negative for PRKACA rearrangements by FISH, whereas the others were positive. An additional cohort of 19 fibrolamellar carcinomas was studied (n=19) to investigate if LFBAP loss is typical of fibrolamellar carcinomas in general. All 19 fibrolamellar carcinomas showed the expected PRKACA rearrangements and the loss of LFABP immunohistologically. To validate this observation, mass spectrometry-based proteomics was performed on tumor-normal pairs of six fibrolamellar carcinomas and showed an average 10-fold reduction in LFABP protein levels in tumors. In conclusion, hepatic adenomas co-occurring with fibrolamellar carcinomas show LFABP loss and are negative for PRKACA rearrangements. LFABP loss, a unique feature of fibrolamellar carcinoma, leads to HNF1-α inactivation that is an important event in pathogenesis.

Diabetes at time of liver transplantation is well known to be associated with reduced post-transplant survival. In this study, additional metabolic conditions (obesity and hypertension) were accessed for prognostic impact on post-transplantation survival. A multi-center cohort included 617 adult subjects who underwent liver transplantation between 2003 and 2009. After a median follow-up of 5.8 years (range 0–10.5), 112 (18.2%) patients died. Diabetes was associated with reduced post-transplant survival (hazard ratio 1.89, 95% confidence interval [CI] 1.25–2.86, P = 0.003), whereas obesity, hypertension, dyslipidemia, and the metabolic syndrome itself were not (P>0.3 for all). Patients with concomitant diabetes and obesity had lower survival (adjusted Hazard Ratio [aHR] 2.40, 95%CI 1.32–4.38, P = 0.004), whereas obese non-diabetic or diabetic non-obese patients had similar survival compared with non-diabetic, non-obese individuals. The presence of hypertension or dyslipidemia did not impact on survival in patients with diabetes (P>0.1 for both). Obese diabetic patients had longer intensive care and hospital stays than those of non-obese diabetic or obese, non-diabetic (P<0.05). The impact of concomitant obesity and diabetes on survival was greater in subjects aged 50+ years (52.6% 5-year survival, aHR 3.04, 95% CI 1.54–5.98) or those transplanted with hepatocellular carcinoma (34.1% 5-year survival, aHR 3.35, 95% CI 1.31–5.57). Diabetes without obesity was not associated with an increased mortality rate in these sub-groups. In summary, concomitant diabetes and obesity, but not each condition in the absence of the other, is associated with reduced post-liver transplant survival. The impact of diabetes and obesity is greater in older patients and those with hepatocellular carcinoma. 

This is an editorial and accompanying study published in the same issue reporting the efficacy of GFT505 in NASH in a phase II study.  GFT505 is a dual peroxisome proliferator-activated receptor (PPAR)-α and -δ agonist that has been reported to lower liver lipids and lower expression of proinflammatory and profibrotic genes in NASH. The antifibrotic effect of GFT505 may be mediated by PPAR-δ agonism. The study involved 274 noncirrhotic biopsy-proven NASH cases who received oral elafibranor (GFT505) or placebo for one year. The primary end point of the study was to reduce any component of NAS to 0 without increase in fibrosis. This end point was not met in the study.

However, the efficacy of elafibranor vs. placebo for cases with NAS >4 was demonstrated in post hoc analysis. Resolution of NASH defined by disappearance of ballooning without progression of fibrosis was achieved in higher number of patients compared to placebo.  The efficacy in NAS ≥ 4 and F1-3 fibrosis (22% resolution vs 13%; p=0.026) was higher than cases with NAS ≥ 4 and F2/3 (15% resolution vs 9% in placebo; p=0.002), suggesting lower response with advanced fibrosis. There were no cardiovascular events or deaths with elafibranor, but decrease in renal function was observed in 7 patients and will need monitoring in future studies.

Yeh MM, et al; Nonalcoholic Steatohepatitis Clinical Research Network. Hum Pathol. 201 ;52:28-37.

This study from the NASH Clinical Research Network Database comprised 157 cases: 94 definite NASH, 40 borderline NASH, 23 not NASH. Acidophil bodies (AB) were counted, and expressed as an acidophil body index (ABI) defined as mean AB count per mm2 was calculated. ABI was 0.04 in definite NASH and 0.02 in borderline as well as not NASH cases (p=0.02). Higher ABI was associated with greater lobular inflammation and hepatocellular ballooning, while no association was observed with steatosis or fibrosis stage. 

The Milan criteria are commonly used to determine liver transplantation in patients with hepatocellular carcinoma.  These stringent criteria have been challenged by alternative expanded criteria including the Valencia, UCSF, University Clinic of Navarra, and Hangzhou criteria.  In this study, the authors compare the outcomes of 6012 patients with hepatocellular carcinoma using these various criteria.  The authors showed that those patients meeting Milan criteria had better outcomes than those exceeding Milan criteria.  However, of the 3286 patients exceeding Milan criteria, 2255 patients did not have tumor recurrence during the 5-year follow-up.  The modified Hangzhou criteria (tumor burden ≤8 cm regardless of AFP and differentiation, or tumor burden >8 cm but AFP ≤100 ng/mL and well-moderate differentiation) had 1-, 3-, and 5-year tumor free survival rates comparable to the Milan criteria. 

The authors study 190 adults with biopsy proven NAFLD.  Vitamin D deficiency was present in 55% of patients was associated with a definitive histologic diagnosis of NASH, increased lobular inflammation, more ballooning degeneration, and presence of fibrosis. RNA sequencing identified differentially expressed genes involving in the MAPK and NF-kB pathways between vitamin D deficient and non-deficient subjects.  The authors suggest that dietary and lifestyle modifications to increase vitamin D levels in deficient NAFLD patients may be beneficial.

http://www.ncbi.nlm.nih.gov/pubmed/27002799

Histopathology May-June 2016

Zen Y, Britton D, Mitra V, Pike I, Heaton N & Quaglia A.  Histopathology 2016;68:796–809.

Using a conventional liquid chromatography-tandem mass spectrometry (LC-MS/MS) protocol, proteomics evaluations frozen large bile ducts inpatients with IgG4 sclerosing cholangitis (ISC) and Primary sclerosing cholangitis (PSC) were found to have differential expressions of proteins that appear to relate to the pathogeneses of these diseases. Of 11 activated pathways in ISC > PSC, 3 were found to include B cell- or immunoglobulin- related pathways, while the 2 activated pathways in PSC > ISC related to extracellular matrix remodelling.

The authors in this paper present a semi-quantitative gardening of G1-G3 for basement membrane thickening and its diagnostic accuracy for Primary sclerosing cholangitis (PSC). They found that G3 BMT was 95% specific for PSC but with low sensitivity of 16% among 128 patients studied.

Clinicopathological concepts on acute and chronic liver disease have evolved rapidly during the last few years, with advances in general and specific treatment options and improved patient outcomes. The old paradigm of ‘irreversibility’ of cirrhosis had been challenged in major ways, and the validity of the usage of the term ‘cirrhosis’ has come into question. This paper addresses aetiology-based clinicopathological concepts and features that may deserve attention because they may determine disease outcome and, specifically, patterns of regression and remodelling. A variety of therapeutic interventions may influence remaining disease features after elimination of damaging agents (virus, alcohol, etc.), and determine the final clinical outcome including the risk of hepatocellular carcinoma (HCC). New concepts create new responsibilities and opportunities for the pathologist to contribute to the understanding of liver pathology and communicate this with clinical colleagues and researchers.

The aim of this study was to evaluate the frequency of terminal hepatic venules (THV) injury resembling sinusoidal obstruction syndrome (SOS) in the setting of haematopoietic stem cell transplant (HSCT) recipients. A total of sixty-three consecutive biopsies from allogeneic HSCT recipients were scored for injured THVs. Forty-nine (78%) biopsies had injured THVs, and 10 (16%) were diagnosed with SOS (mean ± standard deviation of injured THVs/biopsy: 90 ± 9%). Biopsies diagnosed with other diseases also had injured THVs (36 ± 33%). Biopsies from patients with cyclophosphamide plus fractionated total body irradiation conditioning and biopsies taken within 100 days post-HSCT had significantly more occluded THVs (respectively: 40 ± 38%, P = 0.0188; and 35 ± 35%, P = 0.0076) than those with other conditioning regimens or in biopsies taken >100 days post-HSCT. All biopsies taken at any time in the 6-year post-HSCT period had similar amounts of THV phlebosclerosis (23 ± 25%). The authors found a high incidence of THV injuries resembling SOS in post-HSCT liver biopsies. THV injuries were detectable for several years post-HSCT, and were concurrent with other diagnoses. Our results also suggest that SOS may be underdiagnosed.

The authors performed anti-smooth muscle actin (ASMA) immunostain and picrosirius red stain on 162 liver transplant biopsies of 54 patients performed at 6 months, 3 years, and 7 years. They demonstrate by computer morphometric analysis that biopsies with ASMA expression (surrogate for activated stellate cells) of 8% of biopsy surface area or greater at 6 months, was an independent risk factor for significant fibrosis at 7-years.

Alcoholic liver disease (ALD) is a major indication for liver transplantation (LT), but up to 20% of patients experience severe alcoholic relapse. The aims of this study were to evaluate the impact of severe alcoholic relapse on the graft (based on histological examination) and to identify predictive factors associated with recurrent alcoholic cirrhosis (RAC). From 1990 to 2010, 369 patients underwent LT for ALD at Edouard Herriot Hospital (Lyon, France) and survived more than 1 year. All patients who presented severe alcoholic relapse and histological follow-up were included. Liver biopsies were performed at 1 and 5 years and at every 5 years after LT, and when clinically indicated. The median follow-up after LT was 11 years (range, 3-18 years). Severe alcoholic relapse was observed in 73 (20%) of the 369 patients, from whom 56 patients with histological evaluation were included. RAC was diagnosed in 18 (32%) of the 56 patients included, which represents 5% of the 369 patients transplanted for ALD. The median delay between LT and RAC was 6 years (range, 3-10 years) and 4.5 years (range, 2-8 years) after severe alcoholic relapse. The median cumulated years of alcohol use before RAC was 3.5 years (range, 2-7 years). The cumulative risk for F4 fibrosis was 15% at 3 years, 32% at 5 years, and 54% at 10 years after severe alcoholic relapse. A young age at LT (≤50 years old) and an early onset of heavy drinking (within the first 3 years after LT) were associated with RAC. In conclusion, severe alcoholic relapse usually occurs in the first years after LT and is responsible for accelerated severe graft injury. 

A review of 4,500 explant pathology findings documented online using the OPTN explant pathology form confirmed the usefulness of this form as information such as Extrahepatic spread, poor tumor differentiation, microvascular invasion, macrovascular invasion, and confirmation of pre-transplant Milan staging were found to have significant prediction for tumor recurrence in the post-transplant period.

Steatosis occurs frequently after liver transplantation (LT). We aimed to determine the prevalence of steatosis in adult LT recipients, to determine the effects of significant (>33%; grades 2-3) steatosis on patient survival, and to identify risk factors for the development of significant steatosis and its effect on fibrosis progression. We retrospectively examined 2360 posttransplant biopsies of 548 LT recipients. Survival was compared between patients with significant steatosis and those with grades 0-1 steatosis. Patients with significant steatosis were compared to controls without steatosis (grade 0) for clinical and laboratory factors and fibrosis progression. Steatosis was found in 309 (56.4%) patients, including 93 (17.0%) patients with significant steatosis. Steatohepatitis (nonalcoholic fatty liver disease activity score ≥ 5) was diagnosed in 57 (10.4%) patients. The prevalence of steatosis increased from 30.3% at 1 year to 47.6% at 10 years after LT (P < 0.001). Survival times did not differ between groups (P = 0.29). On multivariate analysis of pretransplant factors and initial immunosuppression (IS), alcohol-induced cirrhosis (P < 0.001) and high body mass index (BMI; P = 0.002) were associated with the development of significant steatosis, whereas increased levels of alkaline phosphatase (P = 0.01) and mycophenolate mofetil given initially (P = 0.009) appeared to protect against significant steatosis. On multivariate analysis of posttransplant factors, high BMI (P < 0.001), serum triglycerides (P < 0.001), alcohol consumption (P = 0.005), and type 2 diabetes mellitus (P = 0.048) were associated with significant steatosis, whereas high creatinine (P = 0.02) appeared to protect against significant steatosis. Significant steatosis was not associated with a higher fibrosis stage (P = 0.62). Posttransplant steatosis affects 56.4% of LT recipients, and the prevalence increases with time after LT. Recipient factors and types of IS affect the risk for significant steatosis, which is not associated with a higher fibrosis stage or worse patient survival.

Prepared by: