Robert P. Eckstein

Aberrant p16 INK4A and DPC4 /Smad4 expression in intraductal papillary mucinous tumours of the pancreas is associated with invasive ductal adenocarcinoma

Background and aims: Intraductal papillary mucinous tumours (IPMT) of the pancreas constitute a unique pathological entity with an overall incidence of associated invasive malignancy of 20%. The malignant potential of an individual IPMT cannot be accurately predicted. Preoperative estimation of the risk of associated invasive malignancy with IPMT would be of significant clinical benefit. As aberrations in cell cycle regulatory genes are associated with the progression of precursor pancreatic ductal lesions to invasive adenocarcinoma, we examined expression of key cell cycle regulatory genes in the cyclin D1/retinoblastoma pathway and the transforming growth factor β/Smad4 signalling pathway in a cohort of patients with surgically resected IPMT. Methods: Sections of formalin fixed paraffin embedded pancreatic tissue from a cohort of 18 patients with IPMT were examined using immunohistochemistry for protein expression of cell cycle regulatory genes p16INK4A, p21CIP1, p27KIP1, cyclin D1, pRb, and p53, as well as the cell signalling molecule Smad4. A comparison of expression levels was made between adenoma/borderline IPMT (10 patients) and intraductal papillary mucinous carcinoma (IPMC) (eight patients, four of whom harboured invasive carcinoma). Statistical analysis was performed using the χ2 and Fishers exact tests. Results: Aberrant expression of the proteins examined increased in frequency from adenoma/borderline IPMT to IPMC. Specifically, there was a significantly greater incidence of loss of p16INK4A expression in IPMC: 8/8 lesions (100%) compared with 1/10 (10%) adenoma/borderline IPMT (p<0.001). Similarly, loss of Smad4 expression was associated with IPMC: 3/8 (38%) versus adenoma/borderline IPMT 0/10 (p<0.03). Loss of Smad4 expression within the IPMT was the best marker for the presence of invasive carcinoma (p<0.001). Conclusions: These data indicate that loss of p16INK4A and Smad4 expression occur more frequently in IPMC alone, or with associated invasive carcinoma, compared with adenoma/borderline IPMT. Aberrant protein expression of these cell cycle regulatory genes in IPMT and pancreatic intraepithelial neoplasia in the current model of pancreatic cancer progression suggest similarities in their development and may also represent the subsequent risk of invasive carcinoma.

Immunohistochemistry for SDHB divides gastrointestinal stromal tumors (GISTs) into 2 distinct types.

The Carney triad (CT) is gastrointestinal stromal tumor (GIST), paraganglioma, and pulmonary chondroma. The GISTs of CT show different clinical, molecular, and morphologic features to usual adult GISTs but are similar to the majority of pediatric GISTs. We postulated that these GISTs would show negative staining for succinate dehydrogenase B (SDHB). We performed SDHB immunohistochemistry on GISTs arising in 5 individuals with CT, 1 child, 7 individuals with GIST in young adulthood including 2 with germline KIT mutations, 3 individuals with neurofibromatosis 1, one 63-year-old female with multifocal gastric epithelioid GIST with lymph node metastases, and 104 consecutive unselected individuals with apparently sporadic GIST. The GISTs and paragangliomas arising in CT, the pediatric GIST, and the multifocal gastric GIST from the 63-year-old showed negative SDHB staining. GISTs from the 7 young adults and 3 with neurofibromatosis were SDHB positive. Of the unselected GISTs, 101 (97%) were positive. One of the negative GISTs arose in a 48-year-old female with previous recurrent multifocal gastric GISTs and the other 2 arose in females also in their 40s with gastric GISTs with epithelioid morphology. We conclude that negative staining for SDHB is characteristic of the GISTs of CT and the subgroup of pediatric GISTs which it resembles. Furthermore, when negative staining occurs in apparently sporadic GISTs in adults, the GISTs show morphologic and clinical features similar to pediatric and CT type GISTs. GISTs may therefore be divided into type 1 (SDHB positive) and type 2 (SDHB negative) subtypes.

Peritoneal Morphology on Maintenance Dialysis

Thirty-eight peritoneal biopsies from 37 patients with normal renal function or with end-stage renal failure without replacement therapy or utilizing continuous ambulatory peritoneal dialysis (CAPD), haemodialysis (HD) or a functioning transplant were examined histologically. No abnormality in peritoneal membrane morphology was observed in uraemia in the absence of dialysis. Significant abnormalities of peritoneal membrane morphology were observed in association with CAPD, the predominant finding being the development of peritoneal fibrosis which had a deleterious effect on membrane function. Abnormal peritoneal morphology was less commonly observed in patients on maintenance HD and with functioning transplants but may have implications regarding the future use of CAPD in these patients.

Synoptic reporting improves histopathological assessment of pancreatic resection specimens

Aim: We examined whether introduction of a standardised pancreatic cancer minimum data set improved the reporting of key pathological features across multiple institutions. Methods: From seven different pathology departments that are members of the New South Wales Pancreatic Cancer Network, 109 free text reports and 68 synoptic reports were compared. Results: AJCC stage could not be inferred from 44% of free text reports, whereas stage was reported in all 68 synoptic reports. In the free text reports 28 different names were used to designate margins. All margins were reported in only 12 (11%) of the free text reports compared with 64 (94%) of the synoptic reports (p = 0.0011). The presence or absence of lymphovascular or perineural invasion was reported in 72 (66%) and 92 (84%) of free text reports, respectively. In contrast, lymphovascular space and perineural invasion were reported in all synoptic reports (p = 0.0011 and p = 0.0058). Conclusion: We conclude that synoptic reporting of pancreatic resections without any other intervention increases the information contained within histopathology reports. Therefore, the introduction of minimal data set synoptic reports is a simple and feasible mechanism to immediately improve reporting for pancreatectomy specimens.

Immunohistochemistry for PMS2 and MSH6 alone can replace a four antibody panel for mismatch repair deficiency screening in colorectal adenocarcinoma.

Aims: Currently, testing for mismatch repair deficiency in colorectal cancers is initiated by performing immunohistochemistry with four antibodies (MLH1, PMS2, MSH2 and MSH6). If any one of these stains is negative the tumour is considered microsatellite unstable and, if clinical circumstances warrant it, the patient is offered genetic testing for Lynchs syndrome. Due to the binding properties of the mismatch repair heterodimer complexes, gene mutation and loss of MLH1 and MSH2 invariably result in the degradation of PMS2 and MSH6, respectively, but the converse is not true. We propose that staining for PMS2 and MSH6 alone will be sufficient to detect all cases of mismatch repair deficiency and should replace routine screening with all four antibodies. Methods: The electronic database of the department of Anatomical Pathology, Royal North Shore Hospital, Sydney, Australia, was searched for all colorectal carcinomas on which a four panel immunohistochemical microsatellite instability screen was performed. An audit of the slides for concordant loss of MLH1‐PMS2 and MSH2‐MSH6 was then undertaken. Unusual or discordant cases were reviewed and, in some cases, re‐stained to confirm the staining pattern. Results: Of 344 cases of colorectal cancer which underwent four antibody immunohistochemistry, 104 displayed loss of at least one mismatch repair protein. Of these, 100 showed concordant mismatch repair loss (i.e., loss of MLH1 and PMS2 or loss of MSH2 and MSH6). The four discordant cases comprised two single negative cases (1 MSH6 negative/MSH2 positive case, 1 PMS2 negative/MLH1 positive) and two triple negative (both MLH1/PMS2/MSH6 negative). The microsatellite instability (MSI) group showed a relatively high median age (69.3 years) due to the departmental policy of testing all cases with possible MSI morphology regardless of age. Conclusions: The sensitivity and specificity of a two panel test comprised of PMS2 and MSH6, compared to a four panel test, is 100%. No false negatives or positives were identified. We conclude that the two panel test should replace a four panel protocol for immunohistochemical screening for mismatch repair deficiency.

Idiopathic Subglottic Stenosis: Diagnosis and Endoscopic Laser Treatment

The records of 15 patients with idiopathic subglottic stenosis treated at Royal North Shore Hospital, Sydney, between 1980 and 1994 were reviewed. All were female and had similar characteristic clinical and histopathologic features. Endoscopic laser vaporization was the primary treatment in 12 patients and was successful in maintaining the airway of 8 of these 12; this outcome indicates that the disease can be managed, at least initially, by endoscopic laser treatment in most cases.

Iron tablets cause histopathologically distinctive lesions in mucosal biopsies of the stomach and esophagus

&NA; Iron tablets are widely used in the community. Severe, sometimes lethal, damage to the gastrointestinal tract following overdose is well known, but there is less appreciation of the damage that can be caused by therapeutic dosage. In this histological study of three esophageal lesions and six gastric lesions (including one autopsy case), heavy iron accumulation was demonstrated within ulcer granulation tissue, in connective tissue and blood vessels of the mucosal lamina propria, and within glandular and squamous epithelium. The appearance was distinctive and was similar to that seen following overdosage, although more localized. Five of the patients studied had evidence of delayed esophageal or gastric emptying which could have contributed to the damaging effect of the iron tablets. In most of the patients ulceration appeared to have preceded the commencement of iron therapy, so that the iron probably exacerbated rather than initiated ulceration and stricture formation. Histopathological reporting of iron‐induced changes in endoscopic biopsies will alert clinicians to a correctable pathological process.

Mutations associated with microsatellite unstable colorectal carcinomas exhibit widespread intratumoral heterogeneity

Although microsatellite instability (MSI) has been shown to be present in 15% of sporadic colorectal carcinomas, the genetic events underlying the development of these tumors have not been well described. By investigating intratumoral heterogeneity, this study attempts to elucidate whether MSI‐positive colorectal carcinomas develop as the result of a random accumulation of mutations or as an ordered, stepwise sequence of genetic alterations. Eighty‐six regions from 16 MSI‐positive sporadic colorectal carcinomas were examined for mutations in repeat nucleotide sequences of the tumour suppressor genes transforming growth factor β type II receptor (TGFBRII), insulin‐like growth factor II receptor (IGFIIR), and BAX, and the mismatch repair genes MSH3 and MSH6. At least 2 and up to 5 of these genes were mutated in each tumour, and widespread intratumoral heterogeneity was observed for each gene. Regions of tumour with TGFBRII mutations were correlated with a poorly differentiated histology. Unlike the situation in microsatellite stable colorectal carcinomas, the findings of the present study did not suggest that a particular sequence of tumour suppressor and mismatch repair genes are mutated during colorectal tumorigenesis. It seems likely that a random accumulation of mutations, as a result of a defect in the mismatch repair pathway, drives tumour progression in this type of colorectal carcinoma.

FLUCLOXACILLIN INDUCED LIVER DISEASE: HISTOPATHOLOGICAL FINDINGS AT BIOPSY AND AUTOPSY

&NA; The histological appearances of liver biopsies of 13 patients who developed cholestasis following courses of flucloxacillin are presented. In most of the cases jaundice and pruritus were protracted and in nearly all cases liver function tests are yet to return to normal after mean follow‐up of 18 mths. One patient died after 7 mths of jaundice and another shows clinical evidence of secondary biliary cirrhosis. Biopsies typically showed hepatocellular and canalicular bile stasis with minimal or no hepatitis. Mild portal fibrosis and a patchy portal lymphocytic infiltrate were usually present. In 4 cases bile ducts were reduced in number and in 6 cases reduced in size. Bile duct epithelium showed degenerative changes but only occasional infiltration by inflammatory cells. Ductular proliferation was quite variable and in some cases — most noticeably the fatal case — was inconspicuous despite depletion of bile ducts. The appearances suggested damage not only of hepatocytes but also of bile ducts and proliferating ductules. This may explain the prolonged and occasionally irreversible hepatic disease associated with the use of flucloxacillin. Flucloxacillin should be included amongst the causes of vanishing bile duct syndrome.

Epithelioid hemangioendothelioma of the liver. Report of two cases histologically mimicking veno-occlusive disease

&NA; Summary This report describes two patients with epithelioid hemangioendothelioma presenting with multiple tumour nodules in the liver, but without evidence of tumour elsewhere. One patient died in liver failure twelve months after presentation. The other patient has been well for over two years. Histologically the tumours were remarkably similar, being composed of a proliferation of primitive endothelial cells embedded in a myxohyaline stroma, infiltrating central veins, hepatic sinusoids and, to a lesser extent, terminal portal venules. In each case an initial mistaken histological diagnosis of veno‐occlusive disease was made.