Search

The presence of a high number of infiltrating CD1a+ cells in malignant neoplasms has been reported to be associated with an improved prognosis, reduced tumour recurrence and fewer metastases. This study identified a population of CD1a+ cells within the lymphoid cell infiltrate in human ductal breast carcinoma (n = 52), which was significantly different from normal breast tissue, in which only two out of nine cases expressed CD1a+ cells (P = 0.0192). In the majority of cases, the infiltrate was low compared with the number of macrophages and T cells present (results not shown). There was no correlation between the number of CD1a+ cells and tumour grade, with all tumour grades expressing similar numbers of infiltrating CD1a+ cells. There was clear evidence, however, that the CD1a+ cells were closely associated with tumour cells. It is likely that CD1a+ cells have a role in antigen capture and presentation in human tumours, and this study documents the density of CD1a+ cells in a large sample of all histological grades of human breast carcinomas.

In order to determine the mechanisms of relapse following response to endocrine therapy, we have measured the oestrogen receptor (RE) content of biopsies of breast cancer in patients receiving various types of endocrine treatment. RE content fell in responding (means of 260.2 to 12 fmol/mg protein) and in nonresponding (means of 155.1 to 31.8 fmol/mg protein) patients who had measurable receptor at the start of treatment. Some of these patients, and a further group of responders to endocrine therapy, were monitored until relapse. Tumour biopsies at the time of relapse showed that 10/14 tumour samples contained significant RE (mean of 86.7 fmol/mg protein; range less than 10-271 fmol/mg protein) after successful endocrine therapy. No relationship could be found between RE content and plasma gonadotrophin or steroid-hormone concentration, but the fall in RE content correlated with reduced numbers of tumour cells in the biopsy. These results indicate that relapse following successful endocrine therapy in breast cancer does not appear to be due to the emergence of RE-negative tumour cells. The fall in RE content during response to endocrine therapy may be due to reduced tumour-cell content of the biopsy.

Gonadotrophins, oestradiol, androstenedione, testosterone and dehydroepiandrosterone sulphate (DHAS) were measured sequentially in 72 patients with advanced breast cancer receiving endocrine therapy of various types. Tamoxifen significantly reduced gonadotrophins but did not effect other hormones. Danazol also reduced gonadotrophins. Aminoglutethimide (AGT) reduced oestradiol and DHAS but had not effect on gonadotrophins. The effects of administering tamoxifen, AGT and danazol together (TAD) together were therefore examined. This combination reduced gonadotrophins, oestradiol and DHAS, but no further than tamoxifen and AGT alone. The degree and duration of hormone suppression were similar in both responders and non-responders to tamoxifen, AGT or TAD, though patients responding to AGT showed more complete suppression at the end of the course of treatment, perhaps because they were treated longer. On relapse, adequate gonadotrophin and steroid suppression was demonstrated in patients receiving tamoxifen and AGT respectively. We conclude that (a) response to endocrine therapy is unlikely to be related to the degree of endocrine suppression produced by the therapy; (b) combination endocrine therapy does not further reduce serum-hormone concentrations and (c) relapse is unlikely to be due to escape from the hormone-inhibitory effects of endocrine agents.

Immune-suppressed mice carrying xenografts of several different types of human germ-cell tumours were injected with a radiolabelled monoclonal antibody (LICR LON/HT13) raised against membrane components of a human germ-cell tumour (HX39). Subsequent assessment of radioactivity in excised organs and tumours showed a selective accretion of antibody in the tumour. Quantitative autoradiography supported the results of radiolocalization observed in vivo in different tumours, and also showed that the antibody localized to viable tumour cells and in close association with their cell membrane. The vascular architecture of tumours was found to be an important factor governing antibody distribution. No localization occurred with radiolabelled normal mouse IgG.

The presence of a tumour-associated immune response in 37 patients with ovarian cancer as assessed by blastogenesis (lymphocyte transformation) evoked by ovarian cancer cell extracts, has been correlated with survival following the test. The difference in these responses is unlikely to be accounted for on the basis of general impairment of cell-mediated immuno-competence. Serum carcinoembryonic antigen (CEA) was also determined in 27 ovarian cancer patients to assess its prognostic significance. Raised CEA levels and absence of blastogenic response to tumour cell extract during relapse are associated with a worse prognosis but neither of these parameters are significant in remission. Possible applications of these findings to the clinical management of ovarian cancer patients are discussed. Serum alpha feto-protein levels measured by radioimmunoassay were not found to be raised in any of the 32 ovarian cancer patients in whom it was measured.

The carcinoembryonic antigen (CEA) was produced and released by human colorectal carcinomata and also the normal attached mucosa when maintained in both monolayer and organ culture. Immunoperoxidase cytochemical methods were employed for the cellular localization of CEA which was demonstrable only on the neoplastic cells. Gel filtration and immunological methods confirmed that CEA, produced by normal and neoplastic cells, had properties similar to "authentic" CEA derived from metastatic colorectal carcinomata. In addition, two other CEA cross-reacting macromolecules, neither of which was CCEA-2, were produced by these tumours in culture.

A series of xenografts of human breast carcinomas has been established and serially transplanted in immune-suppressed mice. Certain structural and functional features of the original human tumours, including carcinoembryonic antigen and epithelial membrane antigen, continue to be expressed by the resulting xenografts. Stromal responses such as elastosis and oestrogen-receptor activity were lost by the xenografts. No metastases were detected in tumour-bearing mice. This study suggests that xenografts may have some value in experimental pathology as one type of model of human breast carcinoma.

Twenty-eight mammary carcinomata were maintained in organ culture in the presence of various hormones. The effects of the hormones have been assessed histologically by estimation of total dehydrogenases activity of the pentose glycolytic pathway and by the incorporation of tritiated thymidine or uridine into DNA or RNA. No significant effects on tumour cell activity due to hormones have been observed.

The evolution of metastatic colorectal cancer in patients who have had surgical treatment for a primary lesion was studied in relation the progressive changes in the blood levels of carcinembryonic antigen (CEA), to gamma glutamyl transpeptidase (GGT) and routine liver function tests (LFTs). Involvement of the liver could ofter be reliably predicted many weeks in advance of clinical diagnosis while metastases to other sites were less likely to be detected early by this test. The association of the extent of disease with the patterns of biochemical changes is discussed with reference to several illustrative examples.

A prospective study of 232 patients with primary invasive breast cancer (UICC Stages I, II and III) and histologically confirmed axillary node status was carried out to assess the prognostic significance of several readily available clinical and pathological characteristics. In addition to the recognised utility of tumour size and axillary lymph node status, the presence or absence of cohesive clumps of malignant cells in peritumoral vascular spaces (both lymphatic and blood vessels) was found to be prognostically important.

The profiles of 4 acute-phase reactant proteins (APRPs) (haptoglobin (HPT), alpha1 antitrypsin (AAT), alpha1 acid glycoprotein (AGP) and prealbumin (PALB)) have been studied during the evolution of bowel cancer. Serial measurements of these APRPs can add to the information obtained from measurements of the level of CEA and hepatic enzymes during the monitoring of postoperative patients. There is considerable stability in the profile in a given individual in health, Rises of AAT and AGP are associated with metastases. High levels of HPT may suggest involvement of the bowel wall by recurrent cancer. PALB levels tend to reflect the nutritional status. A discriminant function based on the log CEA, AAT and AGP preoperative blood levels can considerably improve on the predictive value attained using CEA levels alone.

In 1974, the National Institute for Biological Standards and Control (NIBSC) established the first British Standard for carcinoembryonic antigen (CEA) for use in comparative quantitative assays. The Standard, which was prepared for material processed by the Chester Beatty Research Institute, is in the form of a freeze-dried powder, sealed in all glass ampoules code labelled 73/601 and containing pure dry nitrogen. For practical purposes, each ampoule contains 100 units of CEA activity.

Monolayer cultures have been established from a poorly differentiated carcinoma of the lung. Homogeneous cell growth and morphology have been maintained for over 18 months through more than 80 subculture passages, and the cells have been found to produce both immunoreactive calcitonin and an immunoreactive carcinoembryonic antigen-like material.

Radiolabelled affinity-purified antibody to carcinoembryonic antigen (CEA) was injected i.v. into immune-suppressed mice carrying xenografts of human breast carcinoma. Its distribution in the tumours was examined by a combination of immunocytochemistry and autoradiography. The antibody interacted predominantly with the CEA in the extracellular tumour space, rather than on the cell membrane or cytoplasm.

--alpha1-Foetoprotein (AFP) levels have been measured by radioimmunoassay in the serum of 153 male patients with gonadal and extragonadal germ cell tumours. Thirty-five patients with pure seminoma, and 34 patients with teratoma but without any postoperative evidence of residual or recurrent tumour, consistently had normal serum AFP levels (less than 25 ng/ml). Of 84 patients with active teratomas, 56 (67%) had serological evidence of AFP production. Ten patients with histological evidence of pure yolk sac (endodermal sinus) tumours all had raised levels. Teratomas containing yolk sac (elements may or may not be associated with raised serum levels. Trophoblastic (choriocarcinomatous) elements in a teratoma were not normally associated with high values. Fourteen patients with teratomas had elevated levels in the absence of histologically detectable yolk sac elements. Serum AFP levels often became elevated before clinical evidence of recurrence, so that AFP can act as an effective marker of the course of the disease and its response to therapy in many patients, but recurrent or progressive disease may be present in the absence of raised levels.

A total of 190 patients being treated or followed up for urothelial carcinoma have been studied by the serial estimation of their urinary and plasma CEA levels. Only 46% of patients with a urothelial neoplasm present have a raised urinary CEA level. Infection or ileal conduit urine vitiate the result as they produce high CEA levels in the urine in the absence of any neoplastic disease. The accuracy of urinary CEA estimations is compared with that of cytology. Plasma CEA levels do not serve as a useful guide to the presence of extra-urinary tract tumour spread if taken as isolated readings. However, serial plasma CEA estimations may indicate that metastatic disease is present several months before its detection by the more usual clinical methods in a minority of patients.

We have developed a technique for the immunocytochemical staining of marrow smears using antiserum to epithelial membrane antigen (EMA). This membrane component is confined to, but widely distributed in, epithelial tissues and tumours derived from them, and is strongly expressed by infiltrating breast carcinoma cells. Marrow aspirates from patients with both early and metastatic breast cancer have been examined, and the results of immunocytochemical staining compared with conventional cytology and histology. Staining with antiserum to EMA enabled us to detect small numbers of carcinoma cells, and increased the yield of positive samples. Furthermore, using this technique, we found malignant cells in the marrow of patients with primary breast cancer with no other evidence of metastatic disease. Thus immunocytochemical staining for EMA may be of value in the detection of micrometastases in patients with primary breast carcinoma.

Affinity-purified antibodies to carcinoembryonic antigen (CEA) have been injected into immune-suppressed mice bearing xenografts of human breast tumours. It has been shown that the antibodies localized in the tumours but not in normal tissues. The degree of tumour localization correlates with the amount of tumour CEA, and is unaffected by levels of circulating CEA or CEA/anti-CEA immune complexes.

A mouse monoclonal antibody (LICR-LON/HT13) has been developed to a cell-surface antigen carried on a human germ-cell tumour xenograft (HX39). After radioiodination, the antibody localized in vivo preferentially in xenografted tumours as opposed to normal mouse tissue, whereas tumor uptake did not occur with normal mouse IgG or nonspecific monoclonal IgG. This selective localization could be abolished by simultaneous injection of an excess of the unlabelled LICR-LON/HT13. The kinetics of and factors influencing localization have been examined. Tumour weight was important in that the smaller the tumour the better the localization. LICR-LON/HT13 was found to localize also in other xenografted germ-cell tumours, but not in non-germ-cell tumour xenografts. Thus monoclonal antibodies are capable of selective in vivo localization of human tumours in an animal model, and their clinical value should now be assessed.

Walker carcinosarcoma cells cause in vitro osteolysis which may be inhibited by aspirin. In the rat, this tumour produces osteolytic bone deposits and hypercalcaemia, both of which can be prevented by aspirin and indomethacin, whereas soft tissue tumour deposits are unaffected by these drugs. Some human breast tumours cause in vitro osteolysis which may be inhibited by aspirin.