Sclerosing cholangitis (SC) is a heterogeneous disease entity. Different etiologies such as choledocholithiasis, biliary tumor, or pericholangitis can manifest as SC. Hepatic inflammatory pseudotumor (IP) is rarely associated with SC (sclerosing cholangitis associated with hepatic inflammatory pseudotumor; SC-hepatic IP), but sclerosing pancreatitis (SP) is not infrequently associated with bile duct lesions (sclerosing pancreatitis-associated sclerosing cholangitis; SP-SC). In this study, we compared the histologic changes of hepatic hilar and extrahepatic bile duct lesions of SC (7 cases), SC-hepatic IP (5 cases), SP-SC (5 cases), and typical primary sclerosing cholangitis (PSC) (5 cases). Histologically, all SP-SC cases showed extensive and dense fibrosis with marked lymphoplasmacytic infiltration, many eosinophils, and obliterative phlebitis. Four cases of SC showed bile duct lesions similar to those of SP-SC, whereas other three cases of SC showed milder lymphoplasmacytic infiltration, scant eosinophilic cell infiltration, and no obliterative phlebitis. All SC-hepatic IP cases showed bile duct lesions identical to those of SP-SC. Immunohistochemically, many IgG4-positive plasma cells were found in the bile duct lesions of all SP-SC cases, 4 SC cases with marked lymphoplasmacytic infiltration, and all SC-hepatic IP cases. By contrast, IgG4-positive plasma cells were scarce or hardly found in the remaining 3 SC cases and all PSC cases. In conclusion, 4 SC cases and all SC-hepatic IP cases showed bile duct lesions identical to those of SP-SC, suggesting that these three conditions may be a single disease entity. Their pathogenesis may be similar or closely related to that of SP, and in that respect they may represent an IgG4-related biliary disease. They may respond to steroid therapy as SP does.

Chronic sclerosing sialadenitis (CSS) is a cryptogenic tumor-like condition of the salivary gland(s). While immune-mediated processes are suspected in its pathogenesis, and CSS is occasionally reported to be associated with sclerosing pancreatitis, an IgG4-related disease, the exact immunopathologic processes of CSS remain speculative. In this study, we examined the clinicopathologic findings of CSS (12 cases) in comparison with sialolithiasis (8 cases) and Sjogren's syndrome (13 cases), and tried to clarify whether CSS is an IgG4-related disease or not. Submandibular gland(s) were affected in all cases of CSS. CSS cases could be divided into two types: 5 cases were associated with sclerosing lesions in extrasalivary glandular tissue (systemic type), while only salivary gland(s) were affected in the remaining 7 cases (localized type). In the former type, which showed male predominance, bilateral salivary glands were frequently affected, and eosinophilia and elevations of gamma-globulin and IgG in serum were frequently found. Histologically, all cases of CSS showed marked lymphoplasmacytic infiltration admixed with fibrosis and the destruction of glandular lobules. Obliterative phlebitis was found in the affected salivary glands in all cases of CSS. Immunohistochemically, the proportion of IgG4/IgG-positive plasma cells was more than 45% in CSS, while it was less than 5% in controls. The resemblance of the clinicopathologic features of CSS with those of sclerosing pancreatitis suggests the participation of a similar immunopathologic process with IgG4 disturbance in CSS. The abundance of IgG4-positive plasma cells in the lesions would be useful for distinguishing CSS from other forms of sialadenitis.

Recently the authors proposed a new staging and grading system for primary biliary cirrhosis (PBC) that takes into account necroinflammatory activity and histological heterogeneity. Herein is proposed a convenient version of this system. Scores for fibrosis, bile duct loss, and chronic cholestasis were combined for staging: stage 1, total score of 0; stage 2, score 1-3; stage 3, score 4-6; and stage 4, score 7-9. Cholangitis activity (CA) and hepatitis activity (HA) were graded as CA0-3, and HA0-3, respectively. Analysis of interobserver agreement was then conducted. Digital images of 62 needle liver biopsy specimens of PBC were recorded as virtual slides on DVDs that were sent to 28 pathologists, including five located overseas. All participants were able to apply this version in all 62 cases. For staging, kappa was 0.385 (fair agreement) and the concordance rate was 63.9%. For necroinflammatory activity, the kappa and concordance rate were 0.110 (slight agreement) and 36.9% for CA, and 0.197 (slight agreement) and 47% for HA, respectively. In conclusion, this new staging and grading system for PBC seems to be more convenient and practical than those used at present, but more instruction and guidance are recommended for the grading of necroinflammatory activity in practice.

No abstract

Biliary lining epithelia of the bile ducts in biliary diseases are known to have intraepithelial atypical/proliferative lesions related to the development of cholangiocarcinoma. The purpose of the present study was to determine the histological criteria for these lesions based on interobserver agreement. Digital images of 30 intraepithelial atypical/proliferative lesions in the stone-containing intrahepatic bile ducts of hepatolithiasis (30 cases) were sent to 10 pathologists. At first, 10 pathologists made a diagnosis (either of reactive/regenerative change, low-grade or high-grade biliary intraepithelial neoplasia (BilIN-1 and BilIN-2), or in situ carcinoma (BilIN-3)) based on their own criteria. The histological criteria for these four lesions were then determined, and the digital images of the same lesions with proposed criteria were re-distributed. Interobserver agreement on these four lesions was slightly improved (kappa = 0.44, first diagnosis; 0.49, second diagnosis) and intraobserver agreement was 'almost perfect' (kappa = 0.82 at both first and second diagnosis). Interobserver agreement between BilIN-1 and BilIN-2 and that between BilIN-2 and BilIN-3 were 'moderate', although the agreement between regenerative/reactive change and BilIN-1 was 'fair'. In this report, we propose histological criteria for reactive/regenerative change, BilIN-1, BilIN-2 and BilIN-3. Improvement of interobserver agreement suggests their applicability in diagnostic and research fields.

Six cases of primary hepatic carcinomas with a significant amount of sarcomatoid elements were examined by using immunohistochemical stainings. Four of the six cases were associated with ordinary hepatocellular carcinoma (HCC), one with cholangiocellular carcinoma (CCC), and one with mixed HCC and CCC. Alphafetoprotein and alpha-1-antitrypsin were negative in sarcomatoid cells of all cases; vimentin stained positively in sarcomatoid tumor cells in two of the six cases; and cytokeratin (CK8) was detected in five cases. The CK8 was not detected in tumor cells of two cases of hepatic angiosarcoma, two of metastatic leiomyosarcomas, and one of metastatic fibrosarcoma, although vimentin stained positively in all these true sarcomas. It was concluded that sarcomatoid dedifferentiation of liver carcinomas might derive from both HCC and CCC. In addition CK8 might be an excellent marker to make a differential diagnosis of sarcomatoid cancers from true metastatic or primary sarcomas of the liver. Cancer 68:93-97,1991.NE OF THE INTERESTING histopathologic features of 0 carcinomas is sarcomatoid dedifferentiation of cancer cells. Differential diagnosis of sarcomatoid carcinomas from true sarcomas is an important consideration in every organ, although it is sometimes difficult. Most hepatic tumors with sarcomatous features are sarcomatoid carcinoma, and true hepatic sarcomas are rare.' By histogenesis, most investigators indicated that sarcomatoid elements in liver cancers were derived from a dedifferentiation of hepatocellular carcinomas (HCC),2-4 but few reports clarified whether or not sarcomatoid dedifferentiation occurred in cancer cells of cholangiocellular (intrahepatic bile duct) carcinomas (CCC).In this study, six cases of sarcomatoid carcinomas of the liver were examined, and immunohistochemical properties of sarcomatoid cancer cells were investigated from the viewpoint of histogenesis and differential diagnosis. A review of pertinent literature is also given. Accepted for publication November IS, 1990. Materials and MethodsAmong 134 autopsied and 136 resected cases of primary hepatic cancers, six cases with significant sarcomatoid features were examined. Only autopsy materials were available in five cases, and both resected and autopsied livers were examined in one case. Two cases of hepatic angiosarcomas, two of metastatic leiomyosarcomas, one of metastatic fibrosarcoma of the liver, five of ordinary HCC of Edmondson's Grade 11, and three of ordinary CCC (intrahepatic bile duct carcinoma) were also investigated for an immunohistochemical control study.All materials were fixed in 10% neutral formalin and embedded in paraffin. Thin sections were stained by hematoxylin and eosin, Masson trichrome, Shikata's orcein stain, and periodic acid-Schiff/alcian blue for mucin. Alpha-fetoprotein (AFP) (Dako, Santa Fe, CA), alpha-I-antitrypsin (AAT) (Dako), carcinoembryonic antigen (CEA) (Dako), epithelial membrane antigen (EMA) (Dako), vimentin (Dako), desmin (Dako), and low molecular weight (54 kilodaltons) of cytokeratin (C...

Pathologic lesions caused by lanthanum carbonate (LC), a recently developed phosphate-binding agent, have not been recorded. A peculiar gastroduodenal histiocytic lesion associated with a mucosal lanthanum overload was reported. Our routine gastrointestinal biopsy series included 6 cases with heavy lanthanum burden in the gastroduodenal mucosa. In addition to routine histopathologic examinations, a series of immunohistochemical analysis and electron microscopic examinations associated with x-ray diffraction and elemental analysis were performed. Six cases, 3 of male and 3 of female individuals with ages from 59 to 69 years, were all patients of end-stage renal diseases managed under dialysis and treated with LC for >21 months. Endoscopic examinations demonstrated gastric erosions in 3, gastric polyps in 2, and duodenal ulcer in 1. In the mucosal layer, there were numerous non-Langerhans cell histiocytes, stained with CD68 but not S100 protein, engulfing a large amount of mineral-like materials. An electron microscopic and elemental analysis revealed a similar distribution of lanthanum and phosphorus in the histiocytes. Long-standing LC administration can cause massive mucosal accumulation of lanthanum in the tissue histiocytes associated with several forms of gastroduodenal lesions. A long-standing outcome is not clear at present; hence, careful follow-up studies of these patients may be needed.

Serum antinuclear antibodies (ANA) are occasionally noted in the patients with non-alcoholic steatohepatitis (NASH). We examined the significance of ANA in NASH by comparing the clinicopathological features in the patients with ANA-positive NASH (n=35) and ANA-negative NASH (n=36). Inflammatory cell profiles and the distribution of oxidative stress markers were also examined immunohistochemically. The ANA-positive NASH was significantly associated with female gender (p= .005), high degree of portal inflammation (p= .039), interface activity (p= .036) and hepatocellular ballooning (p= .0008). In addition, The ANA of high-titer (320-fold or more) was significantly associated with the histological grade and stage of NASH (p= .02). The degree of steatosis is rather mild in high-titer ANA group (p= .01). The analysis of inflammatory cell profiles revealed that CD3-positive T cells were predominant and plasma cells were rather few in portal area and hepatic lobules in both ANA-positive and ANA-negative groups. There was no difference in the distribution of oxidative stress markers between ANA-positive and ANA-negative groups. These findings suggest that the presence of ANA may be related to the progression of NASH and that a different type of autoimmune mechanisms may be involved in the pathogenesis of NASH with ANA, compared to the pathogenesis of autoimmune hepatitis.