Concordance of HER2 expression in paired primary and metastatic sites of gastric and gastro-oesophageal junction cancers.

HER2 is amplified/overexpressed in a subset of gastric and gastro-oesophageal junction cancers. Addition of anti-HER2 therapy has been shown to provide survival benefit in this setting. However, there are limited data assessing the concordance of HER2 status between primary and metastatic sites.A total of 113 samples from 43 paired primary and metastatic tumours were tested for HER2 status, by immunohistochemistry (IHC) for protein expression and silver in situ hybridisation (SISH) for gene amplification.Primary sites tested included endoscopic biopsies (n = 30) and resections (n = 24). Metastatic samples included lymph nodes (n = 29), peritoneal effusions (n = 21) and miscellaneous sites (n = 9). The overall HER2+ rate was 11%. Of 41 (95%; 95% CI 88.5-100%) concordant cases, 38 were HER2- and three were HER2+. There were two (5%) discordant cases, one of which showed heterogeneity of HER2 expression.This series confirms a high concordance rate of 95%, supporting that testing of primary tumours and metastases is equally valid and providing clinical rationale for the addition of anti-HER2 therapy in HER2+ disseminated disease.

HER2 testing in malignant effusions of metastatic gastric carcinoma: is it feasible?

HER2 testing on effusions of metastatic gastric carcinoma may provide a valuable alternative to testing on primary biopsies. This study assessed the feasibility of this method by immunohistochemistry (IHC) and silver in situ hybridization (SISH). Cell blocks from 46 effusions from metastatic gastric carcinoma were examined. IHC was scored using the modified criteria of Hofmann et al. An average of ≥6 signals per nucleus was considered amplification on SISH. Results were compared with histological specimens to assess HER2 status concordance. Cell blocks showed a poorly cohesive pattern in 30 (65%), aggregates in 7 (15%), and mixed pattern in 9 (20%). Seven (15%) showed an IHC 2+/3+ reaction with a membranous pattern, appearing more granular than in histology. Three (7%) showed HER2 amplification on SISH. In 18 cases (39%), HER2 status was compared with histological specimens, showing 100% concordance. Difficulties were encountered in distinguishing malignant cells from reactive mesothelial cells in 12 (26%). This study supports the feasibility of HER2 assessment on effusions. The incidence of HER2 positivity (7% with SISH+ and IHC 2+/3+) was lower than reported in histological samples. Further refinement and validation will enhance the use of this method in clinical practice and overcome difficulties encountered.

Oncocytic adrenocortical neoplasms--a clinicopathologic study of 13 new cases emphasizing the importance of their recognition.

Oncocytic adrenocortical neoplasms (OANs) are a rare but important subtype of adrenal tumors with unique clinical and morphological features. We present 13 previously unpublished cases, of which 3 were classified as benign, 2 as having borderline malignant potential, and 8 as malignant according to the Lin-Weiss-Bisceglia criteria. Seven tumors (54%) showed evidence of endocrine activity. All were composed of more than 90% oncocytes confirmed immunohistochemically using the antimitochondrial antibody mES-13 and ultrastructurally in 4 cases. Small oncocytes were a frequent finding that challenges the conventional notion of oncocytes as necessarily having abundant cytoplasm. Most cases were immunoreactive for vimentin, synaptophysin, inhibin-α, melan A, and calretinin, the latter being a novel finding in this group of neoplasms. Cytokeratin positivity with AE1/AE3 and CAM5.2 was variable. The literature was comprehensively reviewed to identify all cases of OANs reported to date. Hormone production is not as uncommon as previously believed, occurring in 30%. The Lin-Weiss-Bisceglia criteria were retrospectively applied to all published cases with sufficient information and were shown to effectively separate tumors according to their future risk of recurrence and survival using Kaplan-Meier survival curves (log-rank test, P < .001 for both). The estimated overall median survival for malignant oncocytic neoplasms is 58 months (95% confidence interval = 27.5-88.5 months), providing the first preliminary evidence that the prognosis of malignant OANs is likely to be more favorable than conventional adrenocortical carcinomas, in which the reported median survival is between 14 and 32 months.

Hepatocellular MxA protein expression supports the differentiation of recurrent hepatitis C disease from acute cellular rejection after liver transplantation.

Differentiation of recurrent hepatitis C virus (HCV) disease from acute cellular rejection (ACR) following liver transplantation can be difficult. Previously, we have found that MxA protein, a specific and sensitive marker for type 1 interferon production, is strongly expressed in chronic HCV disease. Here, we investigate MxA expression as a marker for recurrent HCV disease in the livers of 14 adult HCV patients who underwent liver transplantation. Serial liver biopsies available for 12 of these patients were stained for MxA protein and scored using a semi-quantitative approach. Hepatocellular MxA protein levels were significantly up-regulated (p = 0.025) in recurrent HCV disease in comparison to ACR. In biopsies that showed histological changes consistent with recurrent HCV disease, strong hepatocellular MxA staining was present in 14/18 (78%). In the liver biopsies with histological evidence of ACR, strong MxA hepatocellular staining was present in only three of 10 (30%). Thus, assessment of hepatocellular MxA protein expression can contribute to the differential diagnosis of ACR and recurrent HCV disease following liver transplantation. In conclusion, analysis of intrahepatic MxA levels has the potential to reduce the inappropriate use with high-dose pulsing of steroids post-operatively.

Ultra-structural localisation of hepatocellular PKR protein using immuno-gold labelling in chronic hepatitis C virus disease.

The greater resistance of HCV genotype 1 infection to IFN therapy has been partially attributed to functional inhibition of the type 1 interferon induced anti-viral protein PKR in vitro. Whether PKR has antiviral activity against HCV in vivo is unknown. Whilst the ultra-structural localisation of PKR is known in vitro, it is not defined in chronic hepatitis C disease. Using a novel immuno-gold technique we characterised the expression of intrahepatic PKR protein at the ultra-structural level in four patients with chronic HCV disease compared to normal human PBMCs, HepG2 cells and a normal human liver biopsy. All four HCV patients labelled for PKR protein, localising to the nucleus, nucleolus and cytoplasm. Nuclear labelling was confined mainly to the nucleolus and euchromatin. Cytoplasmic labelling was evident within smooth vesicles. Strong immunogold labelling was also evident within the cisternae of the rough endoplasmic reticulum. A similar pattern of ultra-structural nuclear and cytoplasmic PKR protein labelling was seen in PBMCs from healthy donors, HepG2 cells and a normal liver biopsy. The mean nuclear and cytoplasmic count for PKR protein in the HCV group was 21 +/- 4 and 18 +/- 3 gold particles/microm(2), respectively. This represented an increase, though not statistically significant, in nuclear and cytoplasmic labelling for PKR protein in HCV biopsies relative to normal liver tissue.

Perineurioma: a clinicopathological study of eight cases.

AIMS: Perineuriomas (PN) are uncommon, benign neoplasms that mimic a number of benign and malignant soft tissue lesions. There are two main forms: a rare intraneural PN (IPN), and a relatively more common extraneural soft tissue PN (STPN) including a conventional form (STPNc), sclerosing (SPN), reticular and lipomatous variants. Their diagnosis requires immunohistochemical (IHC) and/or ultrastructural (US) confirmation of perineurial cell differentiation. This study aims to review the clinicopathological features of eight PN we encountered recently, to raise awareness of PN as an entity and to highlight the differential diagnoses which include potentially aggressive lesions. METHODS: Clinical, histological, IHC and US features of seven STPN and one IPN were studied. RESULTS: The eight PN arose in the limbs of six females and two males aged 30-58 years. Five STPN occurred in subcutis, one intramuscularly and one intradermally. The STPN were well-circumscribed, multinodular growths. STPNc contained bland spindle cells with long cytoplasmic processes arranged in lamellae, storiform patterns and whorls. Three SPN were acrally located and additionally contained small epithelioid cells in cords and clusters in a myxohyaline stroma with extensive sclerosis. One SPN had giant collagen rosettes of spiral collagen. The IPN showed 'pseudo-onion bulbs' of perineurial cells. All PN were at least focally EMA positive, six of eight were Claudin-1 positive and all were S100 protein negative. Common US features were organelle-poor cell processes, many pinocytotic vesicles, sparse intermediate filaments, and tight junctions and patchy external lamina. There were no recurrences (follow-up 1-49 months). CONCLUSION: PN has a variable morphology and can mimic many benign, borderline and malignant lesions, the differential diagnoses of which are discussed. When confronted with a subcutaneous (in particular) spindle and/or epithelioid cell lesion, EMA/Claudin-1 stains and/or US are essential to identify PN and thereby avoid inappropriately aggressive therapy.

Intrahepatic MxA and PKR protein expression in chronic hepatitis C virus infection.

The therapeutic effect of interferon-alpha and ribavirin in the treatment of chronic hepatitis C viral infection is limited. To identify patient characteristics that may predict responsiveness to treatment, the intrahepatic protein expression of two directly induced IFN-alpha effector proteins, MxA and PKR, were studied. Forty liver biopsy samples from patients with a variety of chronic liver diseases were stained for MxA and PKR protein using immunohistochemical techniques. In a HCV patient cohort, 30 liver biopsies were stained for MxA and PKR protein prior to treatment with IFN-alpha and ribavirin. PKR protein expression was not upregulated in viral liver disease. In contrast, MxA protein expression was significantly upregulated in viral liver disease (P = 0.005). In chronic HCV liver disease, moderate to strong cytoplasmic expression of MxA protein was observed in hepatocytes and monocytes, indicating endogenous hepatocellular IFN-alpha pathway activation. In the HCV patient cohort treated with combination therapy, strong pre-treatment MxA hepatocyte expression was predictive of a non-response to treatment (odds ratio 9.33; P = 0.01; 95% confidence interval 1.63-53.2). This effect was independent of HCV genotype and viral load. It is concluded that pretreatment hepatocellular MxA expression may become a useful predictor of response to combination treatment with IFN-alpha and ribavirin.