To ascertain the correlations between the histology and radiology of microcalcifications revealing subclinical carcinoma.

Most cancers appear to be sporadic. However, 5 to 10% of cancers occur in genetically predisposed individuals. This inherited genetic risk is observed in syndromes such as familial polyendocrinopathies or phacomatosis such as neurofibromatosis, but also in familial aggregations of frequent cancers such as breast or colon cancers. Thanks to studies on molecular genetics, it has been possible over the last ten years to localize and to identify a large number of predisposing genes. These discoveries have permitted to better understand the biological basis of the predisposition and to offer counselling by identifying at-risk individuals in those families. In the case of multiple endocrine neoplasia type 2 associated with an elevated risk for medullary thyroid cancer, the gene involved is a protooncogene named RET and located on chromosome 10. Point mutations affecting specific regions of this gene are the basis of the genetic predisposition. For familial breast cancer, a susceptibility gene named BRCA1 has been located on the long arm of chromosome 17. Mutation of this gene (yet to be identified) led to very elevated risk of breast cancer and also in some families of ovarian cancer.

A familial lympho-epithelial thymoma with constitutional chromosomal translocation t (14;20) (q24;p13) is presented: the thymoma and its particular translocation are present in the mother and the two sons of her offspring. The small number of cases do not allow establishing any relation between thymoma and this particular translocation. Concerning genetic counseling, an annual thoracic radiography is necessary for all the other family members, carriers or not of the translocation.

The age and family history of colorectal cancer are the main risk factors for colorectal cancer. Taking into account the family history of colorectal cancer, the first degree relatives of a patient have a risk that increases to 2 or 3 as compared to general population. The genetic basis for colorectal cancer was recently investigated by complex segregation analysis. They favoured a dominant gene with a gene frequency of 6 :1000. These analyses represent the first step in order to localize the susceptibility genes that are responsible for the familial aggregation of colorectal cancer.

Clinical and histological characteristics of non polypoid colorectal cancers with an hereditary predisposition are presented. The various known genetic syndromes (hereditary non polyposis colorectal cancers, Lynch syndrome type I and type II, Torre-Muir's syndrome, hereditary flat adenoma syndrome) are discussed to find a possible correlation between this nosological classification and their molecular substratum. The main problem today, is to correctly define the population of hereditary predisposed patients to colon cancer, in order to seek and identify the major responsible gene. Any physician, specialist or not, should be encouraged to give the details of the cancer familial context of his patients, in order to aid the oncologist geneticist in his task.

Familial adenomatous polyposis is a rare genetic disease with a dominant mode of inheritance, involved in 1% of colorectal cancer. The APC gene, responsible for the disease, has been localized on the long arm of chromosome 5 and has recently been cloned and sequenced. Mutations predicted to alter the coding property of the APC gene have been reported in large series of adenomatous polyposis patients. Some of them have been correlated with an attenuated phenotype. A genetic test has been developed in affected families. Systematic screening of registered at-risk relatives has allowed a significant reduction of the median age at adenomatous polyposis diagnosis, and thus the incidence of colorectal cancer.

Tumour and peritumoral mucosa were analyzed using flow cytometry in 21 consecutive patients operated on for colorectal carcinoma. Thirteen tumours were DNA aneuploid. No foci of aneuploidy bordering the tumours were detected, regardless of the tumour type. On the other hand, a significant decrease in the proliferative index was observed adjacent to the DNA aneuploid tumours. This study provides an additional argument for segmentary excision in colorectal cancers. Complementary studies are still required to specify the mechanisms of proliferation inhibition in the area of the DNA aneuploid tumours.

The level and molecular composition of laminin, a major basement membrane glycoprotein formed of three chains (A, B1 and B2) have been analyzed in various human colonic cancer cells (Caco-2 and HT29). The synthesis of laminin over a 24 h period, corresponding to cellular and secreted molecules purified by affinity chromatography, was the highest in the more differentiated cells. Immunocytochemical detection of the constituent chains of laminin in permeabilized cells as well as after separation on polyacrylamide gels showed that A, B1 and B2 chains were expressed in Caco-2 cells, whereas A chain was not detected in HT29 cells. When cancer cells were cultured on a monolayer of confluent fibroblastic cells, laminin was deposited at the basement membrane level only in the case of the most differentiated cells. In an attempt to define the role of laminin-A chain, transfection of Caco-2 cells with A chain antisense cDNA was performed; among the clones obtained, 3 were deficient for the target polypeptide. The consequences of the absence of laminin-A chain on basement membrane formation, cellular differentiation and tumor invasion will be currently determined.

The value of prognostic factors can be measured by their influence on the therapeutic decision. Today, with the exception of clinical stage (TNM classification) and to a certain extent estrogen receptors, prognostic factors have little or no application in the clinical practice. The great number of parameters and the discrepancies between various studies may provide some explanations. By the study of the genetic evolution of breast cancers at the chromosomal level, we propose an approach to the understanding of the redundancy of prognostic factors, often noted by the authors. The consequences on the multiparametric statistical methodology are also discussed. Finally, by distinguishing between prognostic factors (predicting survival) and predictive factors (related to treatment response), we hope to provide indications for a better use of the progress in our knowledge on the biology of breast cancers.

Among the great many prognostic factors currently available in breast cancer, three classes of tissue biological parameters appear to be the most reliable in the establishment of a clinical decision flowchart, when they will have been technically and clinically validated: parameters of hormone dependence, tumour aggressiveness and invasion and parameters of proliferation. This article discusses the difficulties encountered in the evaluation of some of these parameters (hormone receptors by various methodological approaches, proteases, enzymes involved in cell proliferation), with emphasis on standardisation of techniques, development of quality controls, clinical validation and objective information of the medical and scientific communities.