Showing papers in "Journal of Clinical Pathology in 1995"

IARC Monographs on the Evaluation of Carcinogenic Risks to Humans

Abstract: Viral hepatitis in all its forms is a major public health problem throughout the world, affecting several hundreds of millions of people. Viral hepatitis is a cause of considerable morbidity and mortality both from acute infection and chronic sequelae which include, in the case of hepatitis B, C and D, chronic active hepatitis and cirrhosis. Hepatocellular carcinoma, which is one of the 10 commonest cancers worldwide, is closely associated with hepatitis B and, at least in some regions of the world, with hepatitis C virus. This timely monograph is a distillation of knowledge of hepatitis B, C and D, based on a review of 1000 studies by a small group of scientists. (It is interesting to note in passing that some 5000 papers on viral hepatitis are published annually in the world literature.) The epidemiological, clinical and experimental data on the association between infection with hepatitis B virus and primary liver cancer in humans are reviewed in a readable and succinct format. The available information on hepatitis C and progression to chronic infection is also evaluated and it is concluded (perhaps a little prematurely) that hepatitis C virus is carcinogenic. However, it is concluded that hepatitis D virus, an unusual virus with a number of similarities to certain plant viral satellites and viroids, cannot be classified as a human carcinogen. There are some minor criticisms: there are few illustrations and some complex tabulations (for example, Table 6) and no subject index. A cumulative cross index to IARC Monographs is of little value and occupies nearly 30 pages. This small volume is a useful addition to the overwhelming literature on viral hepatitis, and the presentation is similar to the excellent World Health Organisation Technical Reports series on the subject published in the past. It is strongly recommended as a readable up-to-date summary of a complex subject; and at a cost of 65 Sw.fr (approximately £34) is excellent value. A J ZUCKERMAN

A "quickscore" method for immunohistochemical semiquantitation: validation for oestrogen receptor in breast carcinomas.

Abstract: Immunohistochemistry is increasingly used in the assessment of markers for breast cancer prognosis. Semiquantitation is frequently desirable but, other than by the use of image analysis, the approaches currently in use are cumbersome. The most common method used is the H-score which takes into consideration the staining intensity in conjunction with the percentage of cells staining positively in breast carcinoma tissue. A "quickscore" has been developed which dispenses with the need to count individual cells. The quantitative biochemical Abbott enzyme immunoassay (EIA) and the Dako immunohistochemical assay (IHA) incorporating a semiquantitative H-score, have been used as standards against which the IHA quickscore for the semiquantitation of oestrogen receptor expression was tested. A good correlation was found between the quickscore and the EIA, which was as good as that between the H-score and EIA. The quickscore is a valid approach and there is no advantage in using the more rigorous H-score. A positive cut off quickscore of > or = 3 has been suggested.

Helicobacter pylori induced interleukin-8 expression in gastric epithelial cells is associated with CagA positive phenotype.

Abstract: AIMS--To use a range of natural phenotypically variant strains of Helicobacter pylori with disparate CagA and VacA (vacuolating cytotoxin) expression to determine which bacterial factors are more closely associated with epithelial interleukin-8 (IL-8) induction. METHODS--Gastric epithelial cells (AGS and KATO-3) were co-cultured with five H pylori strains which were variously shown to express the cagA gene/CagA protein, VacA and/or to exhibit biological cytotoxicity. Secreted IL-8 was assayed by enzyme leaked immunosorbent assay (ELISA) and IL-8 messenger RNA (mRNA) was assayed using a reverse transcription polymerase chain reaction based technique (RT-PCR). RESULTS--Strains expressing CagA, including a variant strain (D931) which is non-cytotoxic and does not express the VacA protein, were found to upregulate epithelial IL-8 secretion and gene expression. In contrast, strains with no CagA expression, even in the presence of VacA and/or biological cytotoxicity, (G104, BA142), failed to induce IL-8 protein or mRNA above control values. CONCLUSIONS--These results strongly support a role for H pylori CagA or coexpressed factors other than the cytotoxin in upregulation of gastric epithelial IL-8. Increased epithelial IL-8 secretion and concomitant neutrophil chemotaxis and activation in addition to direct cytotoxicity may be an important factor in tissue damage and ulceration.

Atypical ductal hyperplasia of the breast: clonal proliferation with loss of heterozygosity on chromosomes 16q and 17p.

Abstract: AIMS--To determine if allelic loss on chromosomes 16q and 17p, commonly encountered in in situ and invasive ductal carcinomas, is present in atypical ductal hyperplasia (ADH); to determine whether ADH is a neoplastic (clonal) or hyperplastic (polyclonal) proliferation. METHODS--Fourteen cases of ADH were examined for allele loss at loci on chromosome 16q and 17p using a microdissection technique, polymorphic DNA markers and the polymerase chain reaction (PCR). RESULTS--Loss of heterozygosity (LOH) was detected in five of nine informative cases on chromosome 16q at the microsatellite D16S413 and two of eight informative cases on chromosome 17p at D17S796. CONCLUSIONS--The incidence of LOH at these loci is similar to that previously observed in ductal carcinoma in situ and in invasive ductal carcinoma. Because of the nature of the technique used, our findings also demonstrate that ADH is a monoclonal, and hence, neoplastic proliferation rather than a hyperplastic (polyclonal) condition as its name suggests. There is thus a case for including ADH, as presently defined, within the spectrum of ductal carcinoma in situ.

Induction of interleukin-8 secretion from gastric epithelial cells by a cagA negative isogenic mutant of Helicobacter pylori.

Abstract: The ability of Helicobacter pylori strains to induce interleukin-8 (IL-8) gene expression and protein secretion from gastric epithelial cell lines in vitro is variable. This cellular response is associated with bacterial expression of the CagA protein present in type I H pylori strains. To determine the role of CagA in this host cell response, an isogenic cagA negative mutant, N6.XA3, was constructed. The cagA negative isogenic mutant and the wild-type parental cagA positive strain, N6, were cocultured with AGS, ST-42 and KATO-3 gastric epithelial cell lines and secreted interleukin-8 assayed by enzyme linked immunosorbent assay. In all three cell lines there was no significant difference in the IL-8 secretion induced by the cagA negative isogenic mutant, N6.XA3, and the wild-type parent strain, N6. These studies show that CagA is not the inducer of IL-8 secretion from gastric epithelial cells. As all wild-type CagA positive strains studied to date induce IL-8, the bacterial factor(s) inducing this inflammatory response is closely associated with the expression of CagA.

High prevalence of Helicobacter pylori in saliva demonstrated by a novel PCR assay.

Abstract: AIMS--To investigate the prevalence of Helicobacter pylori in the saliva of patients infected with this bacterium. METHODS--A novel polymerase chain reaction (PCR) assay was developed to detect H pylori in saliva and gastric biopsy specimens from patients undergoing endoscopy. RESULTS--Our PCR assay amplified a 417 base pair fragment of DNA from all 21 DNAs derived from H pylori clinical isolates but did not amplify DNA from 23 non-H pylori strains. Sixty three frozen gastric biopsy and 56 saliva specimens were tested. H pylori specific DNA was detected by PCR in all 39 culture positive biopsy specimens and was also identified from another seven biopsy specimens which were negative by culture but positive by histology. H pylori specific DNA was identified by PCR in saliva specimens from 30 (75%) of 40 patients with H pylori infection demonstrated by culture or histological examination, or both, and in three patients without H pylori infection in the stomach. CONCLUSION--The results indicate that the oral cavity harbours H pylori and may be the source of infection and transmission.

The polymerase chain reaction in the demonstration of monoclonality in T cell lymphomas.

Abstract: AIMS--To evaluate polymerase chain reaction (PCR) amplification of T cell receptor (TCR) beta and gamma chain genes as a means of demonstrating monoclonality in T cell lymphomas using histological samples; to compare the performance of PCR with Southern blot analysis. METHODS--TCR-beta, TCR-gamma and immunoglobulin heavy chain (IGH) genes were analysed using PCR in 55 cases of T cell lymphoma (28 frozen tissue and 27 paraffin wax embedded samples), diagnosed using morphological and immunohistochemical criteria. The 28 frozen samples were subjected to Southern blot analysis using TCR-beta, TCR-gamma and IGH gene probes. Twenty five B cell lymphomas and 21 non-neoplastic lymphoid tissue samples were used as controls. RESULTS--Using TCR-beta PCR, monoclonality was detected in 24 (44%) of 55 T cell lymphomas compared with 43 (78%) of 55 using TCR-gamma PCR and in 82% with both techniques. Five (9%) of 55 T cell lymphomas were IGH PCR positive. None of the non-neoplastic lymphoid control samples were PCR positive. All B cell lymphomas showed a polyclonal pattern with TCR-beta PCR while a single B cell lymphoma was positive using TCR-gamma primers. With TCR-beta PCR, a monoclonal result was seen in 12 (43%) of 28 frozen samples of T cell lymphoma, compared with 23 (82%) of 28 using Southern blot analysis. With TCR-gamma PCR, 19 (68%) of 28 frozen tissue samples were positive, compared with 26 (93%) of 28 using Southern blot analysis. A single case showed IGH rearrangement by Southern blot analysis. CONCLUSION--TCR-gamma PCR should be the method of choice for analysis of clonality in paraffin wax embedded sections of lymphoproliferative lesions, as TCR-beta PCR has a high false negative rate. Southern blot analysis remains the most successful technique when sufficient fresh tissue samples and resources are available.

Increased gastric production of interleukin-8 and tumour necrosis factor in patients with Helicobacter pylori infection.

Abstract: AIMS--To investigate the role of interleukin-8 (IL-8) and tumour necrosis factor (TNF) in patients infected with Helicobacter pylori. METHODS--The study population comprised 52 patients with dyspepsia attending for upper gastrointestinal endoscopy. Of these patients, 35 were infected with H pylori. IL-8 and TNF concentrations in plasma, gastric juice, and gastric biopsy homogenate supernatant fluid were measured by radioimmunoassay and L929 cell bioassay, respectively. RESULTS--The concentrations of IL-8 and TNF in gastric juice and gastric biopsy homogenates were substantially greater in patients infected with H pylori. In H pylori positive patients IL-8 concentrations in gastric juice and gastric biopsy homogenates were higher in those with moderate gastritis than in those with mild gastritis. There was a positive correlation between IL-8 and TNF concentrations in gastric juice and gastric biopsy homogenate supernatant fluid from H pylori positive patients. There were no significant differences between H pylori positive and negative patients with respect to IL-8 and TNF plasma concentrations. CONCLUSION--This study suggests that increased gastric production of IL-8 and TNF may be implicated in the pathogenesis of H pylori associated gastroduodenal disease.

c-erbB-2 overexpression in the dysplasia/carcinoma sequence of Barrett's oesophagus.

Abstract: AIMS--To investigate overexpression of the oncoprotein c-erbB-2 in the dysplasia/carcinoma sequence of Barrett's columnar-lined oesophagus (CLO). METHODS--Immunohistochemical staining was performed using the monoclonal antibody NCL-CB-11 on formalin fixed tissue from 31 cases of Barrett's carcinoma, 20 cases of cancer associated dysplastic CLO, seven cases of dysplastic CLO without cancer, and 20 cases of non-dysplastic CLO. Membranous staining was regarded as positive for c-erbB-2 overexpression; cytoplasmic staining was recorded separately as its significance is uncertain. RESULTS--Membranous c-erbB-2 overexpression was observed in eight of 31 (26%) carcinomas and in none of the cases of dysplastic CLO. Variable cytoplasmic staining was seen in four of 31 (13%) tumours and seven of 27 (26%) cases of dysplastic CLO. No staining was observed in non-dysplastic CLO. CONCLUSIONS--C-erbB-2 overexpression is a relatively late event in the development of some Barrett's carcinomas and is unlikely to be involved in the early stages of neoplastic transformation of CLO.

Sampling variability of percutaneous liver biopsy in primary sclerosing cholangitis

Abstract: AIMS--To study sampling variability of percutaneous liver biopsy in primary sclerosing cholangitis (PSC). METHODS--One hundred and twelve biopsy specimens (that is, 56 pairs) from 44 patients with PSC, confirmed by cholangiography, were evaluated blindly. Six different features, qualitative grading of four other features and staging according to Ludwig were assessed. RESULTS--Quantitative sampling variability was confined mainly to just one grade or stage, although 11% (six of 56) of the biopsy specimen pairs differed by more than one stage (7% (one of 15) in pairs > 2 cm in length). Qualitative sampling variabilities were between 18 and 71%. Advanced disease (stages 3 or 4) was missed in 40% (two of five) of the biopsy specimens while cirrhosis was missed in 37%. CONCLUSION--Paired liver biopsy specimens should be taken in clinical studies of PSC using liver histology for evaluation or prognosis.

Influence of local peritoneal involvement on pelvic recurrence and prognosis in rectal cancer.

Abstract: AIMS--To evaluate the influence of involvement of the peritoneal surface by carcinoma of the rectum on local recurrence and prognosis. METHODS--Prospective analysis of pathological prognostic factors in 209 resections for rectal carcinoma between 1988 and 1993 with meticulous pathological technique particularly to assess the relation of tumour to the peritoneal surface. Comprehensive clinical follow up with cause of death established from all available sources of information (hospital and general practitioner data) with necropsies where necessary. Local recurrence was determined by accepted clinical, radiological and pathological criteria. RESULTS--Local peritoneal involvement was detected in 25.8% (54/209) of cases. It was more common in women and was associated with tumour differentiation, size and site, and lymph node involvement. Local peritoneal involvement showed considerable prognostic disadvantage in all cases and in curative cases alone. Multivariate analysis demonstrated independent prognostic disadvantage for all cases although this was lost in the curative group. With a 30 month median follow up time, comprehensive clinical surveillance detected 25 (12.0%) local recurrences. Thirteen (52%) palliative cases had shown spread to involve the mesorectal (deep, circumferential) resection margin. Of the 12 curative cases, six were upper rectal cancers with local peritoneal involvement suggesting that tumour seeding into the pelvic peritoneal cavity was the cause of local recurrence. Local recurrence of the six other rectal tumours was probably because of intraluminal seeding in two, involvement of the distal margin in one, extensive extramural venous involvement in two, and tumour spread to the bladder in one. CONCLUSIONS--Comprehensive pathological analysis of a resection specimen can identify cases with a high probability of local recurrence which may benefit from early adjuvant therapy. Involvement of the peritoneal surface is a common event in rectal cancer, has adverse prognostic influence and may be an important factor in local recurrence of upper rectal carcinoma.

Improvement of gastric inflammation and resolution of epithelial damage one year after eradication of Helicobacter pylori.

Abstract: AIMS--To investigate the effect of eradication of Helicobacter pylori infection on gastric epithelial damage and gastritis, scored according to the Sydney system. METHODS--Gastritis scores and epithelial damage were assessed in gastric biopsy specimens before, and five weeks and one year after anti-H pylori therapy in 66 patients with H pylori related gastritis. RESULTS--The mean initial levels of activity, inflammation, atrophy, intestinal metaplasia, and H pylori scores were higher in the antrum than in the corpus or fundus. Eradication of H pylori resulted in an improvement in the mean inflammatory score in antral biopsy specimens from 2.23 before treatment to 1.32 and 1.06, respectively, five weeks and one year after treatment. Corresponding values for fundic biopsy specimens were 1.30, 0.36 and 0.35. Activity scores improved from 1.41 before treatment to 0.13 and zero, respectively, five weeks and one year after treatment in antral biopsy specimens and from 0.60 before treatment to zero in fundic biopsy specimens. Before treatment, epithelial damage was present in 51% of biopsy specimens taken from the antrum and 23% of those from the corpus. Five weeks after eradication of H pylori none of the biopsy specimens revealed evidence of epithelial damage. CONCLUSION--Eradication of H pylori is followed by a rapid, significant improvement in the gastritis score and resolution of epithelial damage in antral and fundic mucosa.

Histological discrimination of idiopathic inflammatory bowel disease from other types of colitis.

Abstract: AIMS--To assess the value of histology in diagnosing inflammatory bowel disease (IBD) in colorectal biopsy specimens. METHODS--Retrospective, double blind evaluation of colorectal biopsy specimens from 41 patients with colitis (28 with ischaemic colitis and 13 with acute self-limited colitis) and 84 patients with IBD (42 with Crohn's disease and 42 with ulcerative colitis). RESULTS--The features distinguishing IBD from other forms of colitis included distorted architecture, lymphocyte and plasma cell infiltrate, excess of polymorphonuclear leucocytes, polymorphonuclear cryptitis, crypt abscesses, and basal lymphoid aggregates. The features discriminating between Crohn's disease and ulcerative colitis included an irregular or villous surface, distorted architecture, decrease in mucus content, and polymorphonuclear cryptitis. Using multivariate analysis, 90% of patients with Crohn's disease and 71% of those with ulcerative colitis were correctly classified, the former being strongly defined by epithelioid granulomas, microgranulomas and isolated giant cells, and the latter best defined by an irregular or villous surface, decrease in mucus content and crypt atrophy. CONCLUSIONS--Examination of colorectal biopsy specimens is a reliable method for diagnosing IBD. In the absence of epithelioid granulomas, microgranulomas and isolated giant cells a diagnosis of Crohn's disease is based on the absence of histological criteria favouring ulcerative colitis. The histological spectrum of indeterminate colitis remains to be clarified.

Lymphoepithelial carcinoma of the salivary gland: in situ detection of Epstein-Barr virus.

Abstract: AIM--To examine the role of Epstein-Barr virus (EBV) in lymphoepithelial carcinoma of the salivary gland in Hong Kong Chinese. METHODS--Ten cases of lymphoepithelial carcinoma of the salivary gland (eight parotid and two submandibular) were examined. In situ hybridisation was used to localise EBER RNA, immunohistochemical methods to detect expression of latent membrane protein 1 (LMP-1) in EBV positive tumours, and Southern blot analysis to examine the clonality of EBV in the two cases where frozen tissue was available. RESULTS--None of the cases had a history of Sjogren's syndrome or histological evidence of a benign lymphoepithelial lesion. The IgA antibody titre against EBV viral capsid antigen was elevated in four cases. All cases were EBV positive by in situ hybridisation, with a strong uniform positive signal in the epithelial cells, and all cases expressed LMP-1. Southern blot analysis revealed that the clonal episomal form of the virus was present. Two of the three female patients in this series also developed carcinoma of cervix. One of these carcinomas had histological features of a lymphoepithelioma-like carcinoma but was EBV negative. CONCLUSIONS--A consistent association between EBV and lymphoepithelial carcinoma of the salivary gland was found. The presence of the virus in a clonal episomal form, and the expression of LMP-1 viral oncoprotein is further evidence of the role of EBV in the oncogenesis of this tumour.

Immunological factors and risk of infection in plateau phase myeloma.

Abstract: AIMS--A series of patients with myeloma were investigated to assess whether immunological risk factors predisposing to serious infection could be identified. METHODS--Patients (n = 102) with predominantly plateau phase myeloma were monitored prospectively for infections. Immunological parameters including total non-paraprotein immunoglobulins and specific antibody titres were measured in all patients and compared with a control population of healthy individuals of a similar age; response to immunisation with Pneumovax II, tetanus and diphtheria toxoids and IgG subclasses were measured in a subgroup of 41 patients. Other characteristics investigated for any association with infection included age, sex, paraprotein type, disease stage, and chemotherapy. RESULTS--Specific antibody titres to pneumococcal capsular polysaccharides and tetanus and diphtheria toxoids were significantly reduced compared with the control population. Low antipneumococcal and anti Escherichia coli titres correlated with risk of serious infection and low anti-pneumococcal titres with severity of non-paraprotein immunosuppression. In 41 immunised patients responses to Pneumovax II, tetanus and diphtheria toxoids were poor; IgG subclass levels were significantly reduced and a poor IgG response to Pneumovax II immunisation was associated with an increased risk of septicaemia and low IgG2 levels. The overall serious infection rate was 0.92 infections per patient year and was four times higher during periods of active disease (1.90) compared with plateau phase myeloma (0.49). The predominant site of infection was the respiratory tract. Clinical and laboratory parameters showed only male sex and reduced non-paraprotein IgG and IgA levels to be significantly associated with at least one serious infection. CONCLUSIONS--A subgroup of patients with myeloma with poor IgG responses to exogenous antigens, who are at increased risk of serious infection, can be identified and may benefit from replacement immunoglobulin therapy to reduce the risk of infection.

Endothelin-1 in idiopathic pulmonary fibrosis.

Abstract: AIMS--To evaluate whether endothelin-1 is involved in the pathology of idiopathic pulmonary fibrosis (IPF). METHODS--Plasma endothelin-1 concentrations were evaluated in 37 patients with IPF and 27 normal controls by radioimmunoassay. In addition, expression of endothelin-1 in lung tissue was evaluated in biopsy specimens obtained from four patients with IPF. Three biopsy specimens of normal lung were used as controls. Endothelin-1 immunoreactivity was detected using immunohistochemistry. RESULTS--Elevated endothelin-1 plasma concentrations were found in patients with IPF compared with controls and a positive correlation was found with duration of disease. No significant difference was observed between treated and untreated patients with IPF. Increased endothelin-1 immunoreactivity was found in lungs of three of four patients with IPF. Endothelin-1 positive consisted mainly of small vessel endothelial cells. Some scattered macrophages were also positive. CONCLUSIONS--Elevated plasma concentrations and expression of endothelin-1 in lung tissue are suggestive of increased production of endothelin-1 in at least a proportion of patients with IPF. Consequently, endothelin-1 activity could play a role in the fibrogenic process of the disease.

Malignant mesothelioma of the pleura: interobserver variability.

Abstract: AIMS--To assess the consistency of histopathological diagnosis of pleural malignant mesothelioma by estimating interobserver agreement between five pathologists. METHODS--Eighty eight histological sets pertaining to original diagnoses of pleural malignant mesothelioma were reviewed separately by each pathologist. Diagnostic likelihood was graded as definite (A), probable (B), possible (C), improbable (D), and definitely not (E) malignant mesothelioma. The following indexes were estimated: observed proportion of agreement (Po), kappa statistics and proportion of agreement for "positive" (Ppos) and "negative" (Pneg) diagnoses. RESULTS--Sixty cases (68.2%) were rated by at least three reviewers as A or B and by none of the others as D or E. Five (5.7%) were rated by at least two reviewers as D or E and by none of the others as A or B. Nine (10.2%) showed a serious disagreement, rated A or B and D or E. Agreement for sets obtained at necropsy/surgery (median kappa w = 0.57) was similar to that for endoscopic material (median kappa w = 0.54). Agreement was poor on material obtained by needle biopsy (median kappa w = 0.21). The median value of Ppos varied between 0.94 (necropsy/surgery) and 0.67 (needle biopsy) and that of Pneg between 0.78 (necropsy/surgery) and 0.34 (unspecified biopsy). Interobserver agreement on histological typing was good overall (median kappa = 0.59). CONCLUSIONS--Of the original histological diagnoses, 70% were consistently reproduced through panel review. Most indexes of agreement between pathologists ranged from poor (needle biopsy) to moderate (necropsy/surgery). Agreement in confirming malignant mesothelioma was greater than that regarding exclusion of this diagnosis. Of the cases finally considered to have malignant mesothelioma, the reproducibility of histological typing was relatively satisfactory.

The biology of low grade MALT lymphoma.

Abstract: In the early 1980s Isaacson and Wright described a group of extranodal low grade B cell lymphomas that recapitulated the histological features of mucosa associated lymphoid tissue (MALT) rather than those of lymph nodes.'2 MALT lymphomas arise in a wide variety of anatomical sites of which the stomach is the commonest; gastric MALT lymphomas have therefore been characterised in the greatest detail. MALT lymphomas from different sites are linked not only by their histological features but also by certain aspects oftheir pathogenesis, clinical behaviour, and molecular genetic properties. Clinically, low grade MALT lymphomas are distinguished by their quite remarkable indolence.3 This group of lymphomas tends to remain localised to their site of origin for extremely long periods. Although spread to regional lymph nodes often occurs, distant dissemination to the bone marrow and other tissues is only infrequently present at the time of diagnosis. This is in direct contrast to low grade B cell lymphomas of lymph nodes which characteristically disseminate to the marrow and other lymphoid tissues early in their course. Explanations for the favourable clinical course of low grade MALT lymphomas have included suggestions that they are not truly malignant but are, instead, \"pseudolymphomas\", that their ability to disseminate is limited by lymphocyte homing phenomena, and that their growth is influenced by exposure to antigen. There is good evidence that MALT lymphomas are malignant neoplasms. As judged by both immunohistochemical and molecular techniques they are clearly monoclonal.4 They have also been shown to display a characteristic clonal cytogenetic abnormality in the form of trisomy 3 which can be detected in over 50% of cases.' MALT lymphomas are invasive; they commonly erode gastric mucosa and are capable of invasion through the gastric wall to the peritoneum. Dissemination to regional lymph nodes is common, and more distant spread of MALT lymphomas to the bone marrow and other organs is well documented.6 Moreover, it has been shown that the primary and disseminated tumour are representative of the same neoplastic clone.6 It has been known for many years that large dividing B cells which leave activated gut associated lymphoid tissue and enter the blood are able to home back to the gut, where they differentiate into plasma cells.7 More recently it has been shown that subsets of circulating lymphocytes express homing receptors which bind to site specific endothelial ligands, the so called \"vascular addressins\".' This system is thought to mediate the preferential recirculation of lymphocyte subsets through specific anatomical sites and has been considered as a possible explanation for the tendency oflow grade MALT lymphomas to remain localised. There is some evidence that such mechanisms may influence the behaviour of low grade MALT lymphomas. In immunoproliferative small intestinal disease (IPSID),9 which is a subtype of gastrointestinal MALT lymphoma, the lymphoma usually arises in the proximal jejunum and shows marked plasma cell differentiation. The neoplastic plasma cells are characteristically found in mesenteric lymph nodes and throughout the intestinal lamina propria in the absence of evidence of spread to peripheral lymphoid tissue, including the bone marrow. The distribution of these plasma cells could be considered to be a reflection of the homing proclivities of normal MALT. It is possible that site specific trafficking of lymphocytes could also explain the multifocal nature of gastric MALT lymphoma,'0 the frequent occurrence of bilateral salivary gland\" and conjunctival lymphoma,\"2 and the tendency for MALT lymphomas to involve other extranodal mucosal sites, as exemplified by gastric lymphomas spreading to the intestine and conjunctival lymphomas to the lung. However, there is no evidence as yet to support the existence of such a finely tuned receptor-ligand system which could mediate the movement of lymphocytes into single or small combinations of mucosal sites. It is possible that the pattern ofgrowth oflow grade MALT lymphomas, as outlined above, is due to dependence on help provided by a local immune response to a specific antigen. In the case of gastric lymphoma, for example, an immune response to a local antigen in the gastric mucosa and gastric lymph nodes may promote growth of neoplastic B cells, whereas cells reaching peripheral sites fail to grow in the absence of antigenic stimuli. This could also account for such phenomena as the multifocality of gastric lymphoma and bilaterality of conjunctival lymphoma. This possibility is supported by an experiment in which an antiidiotype was raised against the tumour immunoglobulin in a case of gastric lymphoma.\"' In this case a scattered population of cells Department of Histopathology, University College London Medical School, Rockefeller Building, University Street, London WC1E 6JJ P G Isaacson J Spencer

E-cadherin expression in intestinal epithelium.

Abstract: AIMS--To investigate E-cadherin expression in normal and inflamed intestine, in the colonic adenocarcinoma cell line HT29, in normal fetal intestine, and in a fetal gut organ culture model where a T cell mediated enteropathy can be generated; to determine whether expression of E-cadherin changes in intestinal inflammation. METHODS--Immunohistochemistry was used to determine E-cadherin expression in following tissues: frozen and paraffin wax sections of normal and inflamed intestine; HT29 colonic adenocarcinoma cell line cultured on coverslips in the presence or absence of cytokines; frozen sections of fetal small intestinal tissue (gestational age 11-22 weeks); and frozen sections of cultured fetal gut in which a T cell mediated enteropathy had been induced. RESULTS--E-cadherin was strongly and evenly expressed by the epithelium in all specimens of intestine studied. Although there was no change in inflammation generally, in some cases of Crohn's disease groups of glands with the characteristic morphology of "ulcer associated cell lineage" showed lower expression of E-cadherin. In fetal gut organ cultures epithelial expression of E-cadherin was lower when local T cells were activated with mitogens, compared with control explants. By contrast, the HT29 cell line showed low levels of expression which increased after treatment with conditioned medium from activated tonsil cells. CONCLUSIONS--E-cadherin is strongly and evenly expressed by epithelium in normal and inflamed intestine, although an increase in E-cadherin expression in cytokine treated HT29 cells was observed. E-cadherin expression is reduced in the epithelium adjacent to ulcers (ulcer associated cell lineage), possibly to assist regeneration.

Immunostaining for CD31 and CD34 in Kaposi sarcoma.

Abstract: AIMS--To evaluate antibodies directed against CD31 (JC70/A) and CD34 (QBEND/10 and anti-HPCA-1) more extensively in Kaposi sarcoma; to assess their value in routine diagnosis; and to compare them with the traditional endothelial cell markers Ulex europaeus agglutinin 1 (UEA-1) and factor VIII related antigen. METHODS--Twenty four cases of Kaposi sarcoma were studied retrospectively. All specimens had been fixed in formalin and embedded in paraffin wax. The antibodies were applied using the Streptavidin biotin technique in all cases except for UEA-1, for which an indirect two stage method was used involving peroxidase conjugated anti-ulex as the secondary antibody. RESULTS--Tumours were classified into those showing angiomatoid or lymphangiomatoid elements and spindle cell lesions. Universal labelling of all lesions and virtually all elements within lesions was seen with the anti-CD34 antibodies QBEND/10 and HPCA-1. Labelling of spindle cells was less consistent with JC70/A but both markers were superior to the traditional endothelial cell markers UEA-1 and factor VIII related antigen. CONCLUSIONS--These data confirm that Kaposi sarcoma is a tumour of endothelial cell origin. They shed further light on the histogenesis of this complex tumour and demonstrate that immunostaining for CD34 and CD31 can be used as an aid to diagnosis in routinely processed tissue.

Soluble forms of the adhesion molecule E-cadherin in urine.

Abstract: The adhesion molecule E-cadherin is essential for maintaining epithelial intercellular adhesion. Loss or reduced expression of E-cadherin has been related to invasive behaviour in a wide range of carcinomas. Using immunoblotting techniques, the existence of multiple soluble forms of E-cadherin was demonstrated in urine from healthy volunteers and patients with benign urinary tract disorders or bladder cancer. The existence of soluble forms of E-cadherin in the urine may reflect shedding from the urinary tract epithelium as part of the normal turnover of this molecule. The possibility that enhanced shedding may contribute to the loss of E-cadherin expression/function in malignancy is discussed.

Simple method for quantifying viable bacterial numbers in sputum.

Abstract: AIMS--To establish a simple method of quantitative culture for determining the viable bacterial numbers present in expectorated sputum samples. METHODS--Sputum samples were homogenised with dithiothreitol, sterile saline or glass beads to determine which method recovered the greatest number of viable bacteria. Culture broths were also incubated with dithiothreitol and sampled over time to determine its effect on bacterial viability. Sputum samples homogenised with dithiothreitol were diluted in sterile saline and sampled using either standard bacteriological loops or a precision pipette to determine which method resulted in the least variation. RESULTS--Homogenisation of sputum using dithiothreitol increased the recovery of viable bacteria compared with sterile glass beads and/or saline, with no apparent effect on bacterial viability when incubated with culture broths. By inoculating agar plates with 10(-3), 10(-4) and 10(-5) dilutions of the homogenised sputum sample, all potential pathogens could easily be identified. A 10 microliter sample volume dispensed by precision pipette and spread with a "hockey stick" resulted in the least variation between plates (less than 16%) and an even distribution of bacterial colonies. Numbers of viable bacteria recovered from different aliquots of individual sputum samples were generally of the same order of magnitude. CONCLUSIONS--This method represents a relatively quick and simple technique for accurately quantifying viable bacteria present in sputum samples. The use of a small portion appears to be representative of the sample as a whole.

Human papillomavirus in false negative archival cervical smears: implications for screening for cervical cancer.

Abstract: AIM--To assess the value of detecting human papillomavirus (HPV) DNA in false negative archival cervical smears in population based screening programmes for cervical cancer. METHODS--Cytomorphologically classified false negative archival Pap smears (n = 27) taken from 18 women up to six years before cervical cancer was diagnosed were blindly mixed with 89 smears from hospital patients with a variety of gynaecological complaints and tested for HPV by the polymerase chain reaction (PCR). Corresponding cervical cancer biopsy specimens were also available for HPV analysis. Neither the examining cytopathologist nor the molecular biologist was aware of the study design. RESULTS--HPV DNA was detected in the smears of 16 patients with cervical cancer missed previously by cytology. HPV 16 and 18 were found predominantly in those smears taken up to six years before the diagnosis of cervical cancer. The smears of the two remaining patients were reclassified as inadequate for cytology or contained no suitable DNA for PCR. In 15 patients the same HPV type could be found in the smears and the cervical cancer biopsy specimens. CONCLUSIONS--The results indicate that high risk HPV types can be detected in archival smears classified as false negative on cytology and that cytological screening errors may be reduced if combined with PCR testing for HPV.

Circulating endothelin-1 concentrations in patients with chronic hypoxia.

Abstract: AIMS--To evaluate the behavior of plasma endothelin-1 in patients with chronic hypoxia. METHODS--Fifteen male patients (mean age 52.1 +/- 3.1 years) with mild chronic obstructive pulmonary disease (COPD) were studied. Twelve healthy men (mean age 48.3 +/- 5.4 years) served as controls. Both patients and controls underwent standard pulmonary function tests, echocardiographic evaluation, and arterial blood gas evaluation. Blood samples for endothelin-1 assay were taken from a previously incannulated antecubital vein after 60 minutes of rest in the supine position. Endothelin-1 was measured by radioimmunoassay after extraction from plasma. RESULTS--Patients with chronic hypoxia had lower PaO2 values (66.1 +/- 6.2 mmHg) than controls (83.8 +/- 2.7 mmHg) but PaCO2 values were similar (38.1 +/- 2.5 v 36.7 +/- 3.1 mmHg, respectively). Arterial pulmonary pressure, therefore, was higher in patients (18.1 +/- 3.7 mmHg) than in controls (10.4 +/- 2.7 mmHg) as were circulating endothelin-1 concentrations (1.22 +/- 0.36 v 0.57 +/- 0.1 pg/ml). Furthermore, plasma endothelin-1 concentrations were negatively correlated with PaO2 and directly correlated with pulmonary pressure levels. No significant correlations were found in controls. CONCLUSIONS--These results show a clear relation between chronic hypoxia and circulating endothelin-1 concentrations. Therefore, chronic hypoxia may be regarded as an important stimulus for endothelin-1 release and as one of the main contributors to increased vasoconstriction in the vascular pulmonary bed which often accompanies lung disease.

Pathologist variation in reporting cervical borderline epithelial abnormalities and cervical intraepithelial neoplasia.

Abstract: AIMS--To evaluate the interobserver variation in the diagnosis of cervical intraepithelial lesions, including the new category "borderline abnormalities of uncertain significance" (BAUS) which has not been tested before. METHODS--Biopsy specimens of 122 patients were reviewed by five histopathologists and the diagnoses subjected to kappa statistical analysis. RESULTS--There was poor interobserver agreement in all categories, particularly between BAUS and normal tissue. CONCLUSIONS--The current guidelines for the histological diagnosis of cervical intraepithelial neoplasia and BAUS are poorly reproducible.

Sudden death from coronary artery dissection.

Abstract: Spontaneous dissection of the coronary arteries is an uncommon condition that may lead to sudden, fatal coronary artery occlusion. Three cases of sudden death attributable to coronary artery occlusion are presented. Dissection was associated with Marfan's syndrome in the first case, and occurred three weeks postpartum in the second case. In case 1, dissection occurred within the intima, and was not associated with an inflammatory cell infiltrate. In cases 2 and 3, dissection occurred between the tunica media and the external elastic lamina, and was associated with a mixed inflammatory infiltrate, rich in eosinophils, T lymphocytes, and histiocytes. The spatial limitation of the inflammatory infiltrate to the adventitial compartment, together with the absence of inflammation in case 1, suggests a reactive origin rather than a causative role for the inflammatory cells. Detailed examination of serial blocks of any coronary artery occlusion is essential in young patients.

Routine application of the polymerase chain reaction for detection of Mycobacterium tuberculosis in clinical samples.

Abstract: AIM--To investigate the use of the polymerase chain reaction (PCR) in the routine laboratory for the detection of Mycobacterium tuberculosis in clinical samples. METHODS--Samples were divided and processed separately for the detection of M tuberculosis by microscopy, culture and PCR. After DNA extraction, PCR was performed with primers specific for the insertion element IS6110 and the product was analysed by agarose gel electrophoresis, Southern blotting or dot blotting and hybridisation with a digoxigenin labelled internal probe. Each sample was tested for inhibitors of Taq polymerase with the aid of an internal control. Multiple negative and positive controls were used to monitor each step of the procedure. RESULTS--The data from two laboratories, using the same operating procedures, were combined. Of 1957 specimens, 79 (4%) were culture and PCR positive, while 1839 (93.9%) were negative in both tests. Thirty specimens (1.5%) were PCR positive only and nine (0.5%) were culture positive but PCR negative. CONCLUSION--Using culture and clinical history as the gold standard, sensitivity and specificity for PCR were 92.1% and 99.8%, respectively. With elaborate precautions, PCR is a suitable and reliable method for the detection of M tuberculosis in clinical samples in a routine microbiology laboratory.

Apoptotic and mitotic indices in malignant melanoma and basal cell carcinoma.

Abstract: AIMS--To assess the relative frequency of mitotic and apoptotic cells in malignant melanoma and basal cell carcinoma. METHODS--Retrospective evaluation of haematoxylin and eosin stained slides from 10 basal cell carcinomas, 10 nodular melanomas, and 10 superficial spreading melanomas, with counting of apoptotic and mitotic cells per 1000 cells. Selected cases were studied with in situ end-labelling. RESULTS--The ratio of apoptotic to mitotic cells was higher in basal cell carcinoma than in either form of melanoma because of the presence of a greater number of apoptotic cells in basal cell carcinoma. CONCLUSION--The differing growth rates and biological behaviour of these two tumours is reflected in the apoptotic:mitotic ratio. Further assessment of this ratio is warranted, both between tumour types and between individual tumours of one type, to determine its value as an indicator of biological potential.

The clinical value of urinary cytology: 12 years of experience with 615 patients.

Abstract: AIMS--To analyse the diagnostic value of cytological examination compared with histological findings in a large series of patients (n = 615) with tumours of the urinary tract epithelium. METHODS--Cytological examinations (n = 785) after bladder washing and exfoliative cytology were retrospectively compared and correlated with histological findings. In addition, 1527 bladder washings were obtained during follow up of patients after transurethral resection of bladder tumours. RESULTS--Cytology in bladder washings (overall diagnostic accuracy 66%) provides considerably more information that exfoliative cytology (overall accuracy 49%). Cytological examinations (n = 1125) in patients with bladder tumours receiving intravesical cytostatic drugs (for example, mitomycin C) yielded suspicious or positive results in 28% of patients, without being confirmed by endoscopy during follow up. CONCLUSION--Our results illustrate two major drawbacks of urinary cytology. First, a high rate of false positive results in patients on intravesical chemotherapy. Second, a high rate of false negative results in highly differentiated carcinomas, stressing the need for additional diagnostic tests such as staining with monoclonal antibodies directed against tumour antigens or assessment of ploidy.

Cytokeratin intermediate filament expression in benign and malignant breast disease.

Abstract: AIM--To carry out a comprehensive study of cytokeratin expression in benign and malignant breast epithelium and breast myoepithelial cells; to examine changes in the cytokeratin profile in malignant and benign epithelium and in carcinomas of increasing histological grade. METHODS--Frozen sections from fibroadenomas (19 cases), fibrocystic disease (19 cases), and infiltrating ductal (68 cases), lobular (seven cases), and mucinous carcinomas (three cases) were examined using a panel of monoclonal antibodies. RESULTS--The luminal epithelium in all fibroadenomas and all cases of fibrocystic disease, as well as tumour cells in most carcinomas, reacted with the specific antibodies to cytokeratins 7, 8, 18, and 19 and to antibodies which included these cytokeratins in their specificities (Cam 5.2, AE1, AE3, RCK102, and LP34). In a few ductal carcinomas none of the tumour cells reacted for cytokeratins 7, 8, or 18. Three ductal carcinomas expressed cytokeratin 14. Only occasional cases expressed cytokeratins 3, 4, 10, and 13. Antibodies which included cytokeratins 5 and 14 in their specificities detected myoepithelial cells less efficiently than antiactin antibodies. CONCLUSION--The cytokeratin profiles in the luminal epithelium in benign breast disease and in tumour cells in most carcinomas are similar in most cases. Some carcinomas, however, are negative for cytokeratins 7, 8, or 18. This may provide a means of predicting the biological behaviour of a histologically borderline lesion.