Quantitative DNA Fingerprinting May Distinguish New Primary Breast Cancer From Disease Recurrence

Publisher Website
Google Scholar
Abstract
Purpose Approximately 10% of women with breast cancer develop a second breast tumor, either a new primary or a recurrence. Differentiating between these entities using standard clinical and pathologic criteria remains challenging. Ambiguous cases arise, and misclassifications may occur. We investigated whether quantitative DNA fingerprinting, based on allele imbalance (AI) or loss of heterozygosity (LOH), could evaluate clonality and distinguish second primary breast cancer from recurrence. Methods We developed a scoring system based on the AI/LOH fingerprints of 20 independent breast tumors and generated a decision rule to classify any breast tumor pair as related or unrelated. We validated this approach on eight related tumors (cancers and synchronous positive lymph nodes). Finally, we analyzed paired tumors from 13 women (bilateral cancers, primary tumors and contralateral positive axillary lymph nodes, or two ipsilateral tumors). Each pair's genetic classification was compared with their clinical diagnosis and outcome. Results Each independent cancer had a unique fingerprint. Every tumor pair's relationship was quantifiable. Six of eight related tumor pairs were genetically classified correctly, two were indeterminate, and none were misclassified. Among the 13 women with two cancers, four of five clinically indeterminate pairs could be classified genetically. In three of 13 women, the pair's classification contradicted the clinical diagnosis. These women had bilateral cancers genetically classified as related and disease progression. This challenges the paradigm that bilateral cancers represent independent tumors. Overall, women with tumors genetically classified as related had poorer outcomes. Conclusion Quantitative AI/LOH fingerprinting is a potentially valuable tool to improve diagnosis and optimize treatment for the growing number of second breast malignancies.