Diagnosis of usual ductal hyperplasia, atypical ductal hyperplasia and low grade ductal carcinoma in situ using copy number alterations


Session type:


Tanjina Kader1,Prue Hill2,Richard Young3,G.bruce Mann4,Ian Campbell1,Kylie Gorringe3
1Cancer Genetics Laboratory, Peter MacCallum Cancer Centre, Melbourne,The Sir Peter MacCallum Department of Oncology, University of Melbourne,2Department of Anatomical Pathology, St Vincent's Hospital,3The Sir Peter MacCallum Cancer Centre,4The Breast Service, The Royal Women's Hospital



Atypical ductal hyperplasia (ADH) is a challenging lesion to diagnose and manage. Due to the similar morphological features, ADH and LG ductal carcinoma in situ (DCIS) are difficult to reproducibly distinguish and therefore, an objective biomarker distinguishing these two lesions is a highly desirable tool in the clinic.


We optimised a robust, cost effective low-coverage whole genome sequencing (LCWGS) method for CNA detection, using 5 ng of FFPE tissue derived DNA. We applied this novel method to pure ADH (not associated with cancer) (n=20). We also applied this method or Molecular Inversion Probe (MIP) arrays for cases with UDH (n=6) and LG DCIS (n=14). Cases were reviewed independently by two pathologists, followed by micro-dissection and DNA extraction. Additional cases of UDH will be validated by fluorescence in situ hybridisation (FISH) and final analysis will be presented.   


Genetic analysis revealed that UDH had no copy number events, in contrast to pure ADH of which 70% had at least one copy number event. Analysis of pure LG DCIS showed that they had a higher level of CN observed compared to ADH, with a median fraction genome altered of 9% (2-33%) and 5% (0-19%), respectively (p=0.04). Moreover, the presence of a CN event of any of chromosomes 11, 13, 17, 21, 22 or X was observed significantly more frequently in pure LG DCIS cases (13/14 cases) compared to pure ADH cases (5/20) (p=0.0001).


Here we report the most detailed molecular taxonomy of this high risk pre-malignant breast lesion with the largest cohort undertaken worldwide. Validation will be required with a larger cohort; however, any CN events of six chromosomes could be used as a potential differential biomarker between ADH and LG DCIS, a highly desirable tool for accurate diagnosis and more precise treatment.