Multicentric mammary carcinoma

Evidence of monoclonal proliferation


We read with interest the article by Middleton et al.1 in which, by comparing histologic and immunohistochemical features, the authors attempted to evaluate whether multiple, ipsilateral breast tumors arise through intramammary spreading of a single clonal proliferation or are independent primaries arising in the same breast. Based on the phenotypic similarities they found in most cases, the authors suggested that synchronous breast tumors arise through spreading of a single primary breast carcinoma. However, they acknowledge that “future molecular genetic analysis of these multifocal separate tumors ultimately will aid in ascertaining their pathogenesis and biology.”

However, such data, acquired both by cytogenetic and molecular genetic means, have already been reported and are already available in the scientific literature. Our group has studied the evolutionary relationship among multiple ipsilateral breast carcinomas by cytogenetic analysis of 37 tumorous lesions from 17 patients.2–4 At least two macroscopically distinct carcinoma foci were karyotypically abnormal in each of 12 patients, making them informative with regard to the clonal relationship among the breast tumors. Nine of these cases (75%) had an evolutionarily related, cytogenetically abnormal clone in the different tumor lesions from the same breast. This is strong evidence that the dominant mechanism for the origin of multiple ipsilateral breast tumors is intramammary spreading from a single primary neoplasm. In the remaining three informative patients, however, no one clonal chromosome aberration was common to the foci. The disparate nature of the observed karyotypic changes in each focus indicates that the synchronous lesions arose by pathogenetically independent processes within the same breast in these three cases, although we cannot completely disregard the possibility that the karyotypically unique foci could have been related by a shared submicroscopic mutation.

The study of the pathogenetic relationship among multiple tumorous breast lesions has also made good use of molecular genetic techniques. Based on an analysis of restriction fragment length polymorphisms of the X-linked phosphoglycerokinase gene and on the random inactivation by methylation early in embryogenesis of one X chromosome in females, Noguchi et al.5 showed that the same X chromosome was inactivated in all foci from each of three patients with unilateral breast carcinoma, indicating that the foci were clonally related. Conversely, Tsuda et al.6 compared the pattern of loss of heterozygosity from 16q among multiple synchronous breast tumors and concluded that in some cases the disparate tumors were clonally related, whereas in other they were independent.

The data now available make it possible to use the terms multicentric and multifocal breast carcinoma in a more biologically precise manner. We suggest that the term multicentric breast carcinoma be used only about multiple, clonally independent breast carcinomas, whereas the term multifocal breast carcinoma be reserved for cases where multiple breast tumors have arisen through intramammary dissemination of a single carcinomatous process. The title chosen by Middleton et al,1 would be misleading by this classification, as it links multicentricity with monoclonal proliferation.

Because the majority of malignant breast tumors are ductal carcinomas of no special type, the tumor histology in most instances cannot discriminate between multifocal and multicentric carcinomas. Other parameters, such as the location of the foci in different quadrants of the breast or the presence of a carcinoma in situ (CIS) component in each lesion, have been considered indicators of multicentricity.7 It is nevertheless becoming clear that these morphologic features are insufficient to distinguish between the two pathogenetically different entities. Some cases with carcinoma foci located far apart (6 cm) were nevertheless classified unequivocally as multifocal carcinomas because the same rare karyotypic changes were found in both lesions.4 On the other hand, our cytogenetic findings do not support the equally common notion that the presence of CIS in each lesion is evidence of multicentricity,7 since four of the cases classified as multifocal by the karyotypic data had a CIS component in both foci.4 The data presented by Middleton et al.1 concur with our earlier suggestion, since 72% of their series had CIS in both lesions in spite of the fact that the authors consider them to be biologically related (multicentric in their terminology, multifocal in ours).

To explain these findings, Middleton et al.1 proposed that neoplastic cells can spread along the ducts and eventually develop into two invasive carcinoma foci, even in separate quadrants of the breast. This interpretation is supported by the earlier findings of Ohtake et al.,8 who performed a detailed study using stereomicroscopic techniques to generate three-dimensional graphic reconstructions of the mammary ductal systems in 20 patients with breast carcinoma. Intraductal tumor extension was found in 80% of the cases, all of which had continuous and segmental involvement, and in some cases also involving adjacent duct systems.

In conclusion, we consider that both multifocality (often) and multicentricity (rarely) exists in breast carcinogenesis and that they call for nomenclature standardization. Morphologic criteria, including the distance between foci and the presence or absence of CIS in each lesion, are inadequate to discriminate between these two biologic entities, something that is better done using the genetic changes of the tumor lesions as clonality markers.

Manuel R. Teixeira M.D.* †, Nikos Pandis Ph.D.‡, Sverre Heim M.D.?, * Department of Genetics, Portuguese Oncology Institute, Porto, Portugal, † Department of Cancer Genetics, The Norwegian Radium Hospital, Oslo, Norway, ‡ Department of Genetics, Saint Savas Hospital, Athens, Greece, ? Department of Cancer Genetics, The Norwegian Radium Hospital, Oslo, Norway