Analysis of the progression of intraductal proliferative lesions in the breast by PCR-based clonal assay

Multiclonal tumor origin: Evidence and implications

An accurate understanding of the clonal origins of tumors is critical for designing effective strategies to treat or prevent cancer and for guiding the field of cancer risk assessment. The intent of this review is to summarize evidence of multiclonal tumor origin and, thereby, contest the commonly held assumption of monoclonal tumor origin. This review describes relevant studies of X chromosome inactivation, analyses of tumor heterogeneity using other markers, single cell sequencing, and lineage tracing studies in aggregation chimeras and engineered rodent models. Methods for investigating tumor clonality have an inherent bias against detecting multiclonality. Despite this, multiclonality has been observed within all tumor stages and within 53 different types of tumors. For myeloid tumors, monoclonal tumor origin may be the predominant path to cancer and a monoclonal tumor origin cannot be ruled out for a fraction of other cancer types. Nevertheless, a large body of evidence supports the conclusion that most cancers are multiclonal in origin. Cooperation between different cell types and between clones of cells carrying different genetic and/or epigenetic lesions is discussed, along with how polyclonal tumor origin can be integrated with current perspectives on the genesis of tumors. In order to develop biologically sound and useful approaches to cancer risk assessment and precision medicine, mathematical models of carcinogenesis are needed, which incorporate multiclonal tumor origin and the contributions of spontaneous mutations in conjunction with the selective advantages conferred by particular mutations and combinations of mutations. Adherence to the idea that a growth must develop from a single progenitor cell to be considered neoplastic has outlived its usefulness. Moving forward, explicit examination of tumor clonality, using advanced tools, like lineage tracing models, will provide a strong foundation for future advances in clinical oncology and better training for the next generation of oncologists and pathologists.

Multicolor immunofluorescence reveals that p63- and/or K5-positive progenitor cells contribute to normal breast epithelium and usual ductal hyperplasia but not to low-grade intraepithelial neoplasia of the breast

We contend that knowledge about the cellular composition of normal breast epithelium is a prerequisite for understanding proliferative breast disease. Against this background, we used multicolor immunofluorescence to study normal breast epithelium and two types of intraepithelial proliferative breast lesion for expression of the p63, basal keratin K5, glandular keratin K8/18, SMA, ER-alpha, and Ki67. We studied eight normal breast epithelium samples, 12 cases of usual ductal hyperplasia, and 33 cases of low-grade intraepithelial neoplasia (9 flat epithelial atypia, 14 low-grade ductal carcinoma in situ and 10 cases of lobular neoplasia). Usual ductal hyperplasia showed striking similarity to normal luminal breast epithelium including p63+ and/or K5+ luminal progenitor cells and the full spectrum of luminal progeny cells. In normal breast epithelium and usual ductal hyperplasia, expression of ER-alpha was associated with lack of expression of the proliferation antigen Ki67. In contrast, we found in both types of low-grade intraepithelial neoplasia robust expression of keratin K8/18 and a positive association between ER-alpha and Ki67 expression. However, these lesions were consistently negative for p63 and/or K5. Our observational study supports the view that usual ductal hyperplasia and low-grade intraepithelial neoplasia are different entities rather than part of a spectrum of the same disease. We propose a new operational model of cell differentiation that may serve to better understand correlations between normal breast epithelium and proliferative breast diseases. From our data we conclude that p63+ and/or K5+ progenitor cells contribute to maintenance of normal epithelium and usual ductal hyperplasia, but not to low-grade intraepithelial neoplasia of the breast.

High prevalence of PIK3CA/AKT pathway mutations in papillary neoplasms of the breast

Papillary lesions of the breast have an uncertain relationship to the histogenesis of breast carcinoma, and are thus diagnostically and managerially challenging. Molecular genetic studies have provided evidence that ductal carcinoma in situ and even atypical ductal hyperplasia are precursors of invasive carcinoma. However, papillary lesions have been seldom studied. We screened papillary breast neoplasms for activating point mutations in PIK3CA, AKT1, and RAS protein-family members, which are common in invasive ductal carcinomas. DNA extracts were prepared from sections of 89 papillary lesions, including 61 benign papillomas (28 without significant hyperplasia; 33 with moderate to florid hyperplasia), 11 papillomas with atypical ductal hyperplasia, 7 papillomas with carcinoma in situ, and 10 papillary carcinomas. Extracts were screened for PIK3CA and AKT1 mutations using mass spectrometry; cases that were negative were further screened for mutations in AKT2, BRAF, CDK, EGFR, ERBB2, KRAS, NRAS, and HRAS. Mutations were confirmed by sequencing or HPLC assay. A total of 55 of 89 papillary neoplasms harbored mutations (62%), predominantly in AKT1 (E17K, 27 cases) and PIK3CA (exon 20 >exon 9, 27 cases). Papillomas had more mutations in AKT1 (54%) than in PIK3CA (21%), whereas papillomas with hyperplasia had more PIK3CA (42%) than AKT1 (15%) mutations, as did papillomas with atypical ductal hyperplasia (PIK3CA 45%, AKT1 27%, and NRAS 9%). Among seven papillomas with carcinoma in situ, three had AKT1 mutations. The 10 papillary carcinomas showed an overall lower frequency of mutations, including 1 with an AKT1 mutation (in a tumor arising from a papilloma), 1 with an NRAS gene mutation (Q61H), and 2 with PIK3CA mutations (1 overlapping with the NRAS Q61H). These findings indicate that approximately two-thirds of papillomas are driven by mutations in the PI3CA/AKT pathway. Some papillary carcinomas may arise from these lesions, but others may have different molecular origins.

Changes in CpG islands promoter methylation patterns during ductal breast carcinoma progression

Aberrant promoter methylation of several known or putative tumor suppressor genes occurs frequently during carcinogenesis, and this epigenetic change has been considered as a potential molecular marker for cancer. We examined the methylation status of nine genes (APC, CDH1, CTNNB1, TIMP3, ESR1, GSTP1, MGMT, THBS1, and TMS1), by quantitative methylation specific PCR. Synchronous preinvasive lesions (atypical ductal hyperplasia and/or ductal carcinoma in situ) and invasive ductal breast carcinoma from 52 patients, together with pure lesions from 24 patients and 12 normal tissues paired to tumor and 20 normal breast distant from tumor were analyzed. Aberrant promoter methylation was detected in both preinvasive and invasive lesions for genes APC, CDH1, CTNNB1, TIMP3, ESR1, and GSTP1. However, hierarchical mixed model and Generalized Estimating Equations model analyses showed that only APC, CDH1, and CTNNB1 promoter regions showed a higher frequency and methylation levels in pathologic samples when compared with normal breast. Whereas APC and CTNNB1 did not show differences in methylation levels or frequencies, CDH1 showed higher methylation levels in invasive tumors as compared with preinvasive lesions (P < 0.04, Mann-Whitney test with permutation correction). The analysis of APC, CDH1, and CTNNB1 methylation status was able to distinguish between normal and pathologic samples with a sensitivity of 67% (95% confidence interval, 60-71%) and a specificity of 75% (95% confidence interval, 69-81%). Our data point to the direct involvement of APC, CDH1, and CTNNB1 promoter methylation in the early stages of breast cancer progression and suggest that they may represent a useful tool for the detection of tumor cells in clinical specimens.