Objective measurement of basement membrane abnormalities in human neoplasms of colorectum and of breast.

Abstract

The basement membrane (lamina densa) was evaluated as normal, absent, multilayered, or otherwise abnormal at intercepts of an array of parallcl lines superimposed nearly perpendicularly on basement membrane regions. Six to eight consecutive photographs of the basement membrane region of three to four areas of three non‐neoplastic lesions and eight neoplastic lesions of colorectum, and two non‐neoplastic and six neoplastic lesions of the female breast were thus examined. Statistical analyses showed that there was no significant difference between three regions of three non‐neoplastic lesions, nor between the three nonneoplastic lesions of colorectum. The analyses were by χ2 tests performed comparing counts of absent basement membrane with those of all other types of basement membrane lumped together. Significant differences were observed between the sum of the non‐neoplastic lesions and the individual neoplasms. Differences between regions of the same lesion were observed only in malignant neoplasms. In breast lesions similarly, where the observations were pooled into three groups, i.e. counts of normal, absent, or other type of basement membrane, no difference between regions of the same lesion were observed in a duct hyperplasia and a fibroadenoma, but such differences were seen amongst adenocarcinomas and other lesions. Comparing the sum of observations of the above two lesions with the other lesions individually, there was generally an increasingly high χ2 value with increasing malignancy. Thus this approach to evaluation of neoplasms may provide an objective measure which correlates with prognosis. It is adaptable to routine use in hospitals with electron microscopy facilities.