The complement regulatory proteins CD46 and CD59, but not CD55, are highly expressed by glandular epithelium of human breast and colorectal tumour tissues.

Three of the proteins protecting cells from autologous lysis by complement are: membrane cofactor protein (MCP; CD46), an inhibitor of the membrane attack complex formation (CD59), and decay accelerating factor (DAF; CD55). We have investigated the expression of these proteins in breast and colorectal carcinoma by immunohistochemistry and immunoblotting of breast tissue for CD46. CD46 was consistently and strongly expressed in the epithelial compartment in 26/28 ductal carcinomas of the breast, 9/9 fibroadenomas, and 9/11 cases of control non-neoplastic breast tissue. CD59 showed a similar degree of expression in the fibroadenomas (9/9), but was less strongly expressed in carcinomatous (22/28) and control (5/11) tissues. In marked contrast, no CD55 expression was detected in tissue from 15 ductal carcinomas. Immunoblotting of breast tissue for CD46 showed the same size of the molecule as for lymphocytes. It had however considerably stronger expression in tumour tissue than in non-neoplastic tissue. CD46 and CD59 were either lacking or only weakly expressed in the epithelial component of control colorectal mucosa: 2/15 and 5/15, respectively. In contrast, tissue samples from colorectal adenocarcinomas showed clear staining for both CD59 (10/18) and, more markedly, CD46 (15/18). There was no association between the pattern or intensity of CD46 and CD59 expression and tumour differentiation. As the complement regulatory proteins CD46 and CD59 are also strongly expressed by trophoblast at the feto-maternal tissue interface, these results support the concept that similar mechanisms are employed both by the genetically dissimilar fetus and certain tumours to evade immune attack by their host.

Differential excretion of leucocyte granule components in inflammatory bowel disease: implications for pathogenesis.

1. Faecal excretion of the leucocyte primary granule component, myeloperoxidase, and of the secondary granule component, lactoferrin, were compared in inflammatory bowel disease and infective diarrhoea. 2. Faecal lactoferrin correlated with faecal myeloperoxidase in both inflammatory bowel disease (P = 0.0018; n = 32) and infective diarrhoea (P = 0.00013; n = 37), but inflammatory bowel disease was associated with a much higher faecal excretion of lactoferrin but lower excretion of myeloperoxidase than infective diarrhoea. As a consequence, the median ratio of lactoferrin/myeloperoxidase excretion (both expressed as ng/mg of protein) for inflammatory bowel disease was 7.5 (range 3.5-21.3) with similar values for ulcerative colitis (n = 18) and Crohn's disease (n = 14) compared with only 0.9 (range 0.4-2.3; P < 0.0001) for infective diarrhoea. In inflammatory bowel disease faecal lactoferrin and myeloperoxidase excretion remained increased even in clinical remission. 3. In subsequent immunohistochemical studies to assess the possible explanation for these findings, lactoferrin and myeloperoxidase were demonstrated within crypt abscesses and surface mucus, both in inflammatory bowel and in infective diarrhoea mucosal samples. There was a slight increase in the number of lactoferrin-containing cells in the mucosal samples from ulcerative colitis and in the submucosa of samples from Crohn's disease compared with infective diarrhoea, but these changes were not sufficient to account for the marked increase in faecal lactoferrin excretion in inflammatory bowel disease. 4. In all mucosal samples, including those from normal mucosa, lactoferrin was also shown to be contained within mast cells. 5. These results could best be explained by a different mechanism for leucocyte activation in inflammatory bowel disease compared with infective diarrhoea, and are compatible with selective secretion of secondary granule components, which include lactoferrin but not myeloperoxidase, as a result of leucocyte activation by N-formylated bacterial peptides in inflammatory bowel disease.

Comparison of the molecular genetics of c-erb-B2 and p53 expression in stomach cancer in Britain and Japan.

BACKGROUND: Differences in the epidemiology and treatment outcome of stomach cancer have led to the suggestion that in Japan, this disease may be biologically less aggressive than that found in the West. The authors compared p53b and c-erb-B2 expression, trying to identify genetic differences in Japanese compared with Western stomach cancers. METHODS: Paraffin embedded formalin fixed tissues from 89 British and 89 matched Japanese patients were examined by immunohistochemistry after microwave treatment. Cases were matched for T-stage, year of surgery, and histopathologic grade. RESULTS: Tumors from 48 British and 46 Japanese patients expressed p53, whereas those of 27 British and 28 Japanese patients expressed c-erb-B2. No significant difference in the density or distribution of protein expression was found between the two populations. The distribution of expression between diffuse and intestinal types and the proportion of cases expressing both antigens were similar in the two groups. CONCLUSIONS: p53 and c-erb-B2 are expressed in the same way in stomach carcinomas from Japanese and British patients. This study found no evidence of genetic differences in the cancers from the two countries.

The prognostic significance of immunohistochemically detected lymph node micrometastases in colorectal carcinoma.

Micrometastases have been detected by immunocytochemical means in the lymph nodes of patients with otherwise node-negative cancer of the colon and rectum. This study examines the incidence and prognostic significance of nodal micrometastases in Dukes' B carcinoma. Five hundred and fifty-nine lymph nodes from 77 cases of Dukes' B carcinoma were examined for lymph node micrometastases by immunocytochemical staining for cytokeratin AE1:AE3. Micrometastases were detected in 19 cases (25 per cent). Cell clusters were present in ten cases, the remaining nine cases displaying only single cells. The presence of micrometastases was unrelated to age (P = 0.06), sex (P = 0.32), tumour site (P = 0.37), tumour size (P = 0.67), or tumour differentiation (P = 0.66). Ten-year survival estimates by the Kaplan-Meier lifetable method was 47 per cent in patients with and without micrometastases (chi 2 = 0.35 and 1 df, P = ns). The presence of nodal micrometastases detectable only by immunocytochemistry in patients with Dukes' B colorectal cancer does not justify reassignment to a more advanced disease stage.