Abstract 4077: Preserving and contextualizing in situ transcriptomes of hormone receptor-negative breast premalignancy Free

Triple-negative breast cancer (TNBC) is a poor-prognosis molecular subtype of breast cancer that is thought to begin as an estrogen receptor-negative ductal carcinoma in situ (ER- DCIS). ER- DCIS is much less common than other types of DCIS, suggesting that it is a transient premalignant state before invasion begins. To better characterize the premalignant transcriptome, we use laser capture microdissection-guided stochastic 10-cell RNA sequencing to profile the transcriptome of ER- DCIS cases that present to the University of Virginia (n = 10 cases thus far). In early cases, RNA integrity was poor with RNA integrity numbers (RIN) near zero. To improve RNA integrity in future samples, we began transporting surgical excision material on ice to minimize RNase activity, and coordinated the pathological processing to minimize time to cryoembedding. We have reduced time to cryoembedding to 26 minutes on average and obtained usable RNA for 10-cell RNA sequencing, though RIN remains case dependent. We also optimized proteinase K and inhibitor concentrations during RNA extraction to ensure maximal reverse transcriptase activity after cell digestion. 10-cell transcriptomes will be combined with RNA sequencing data from the Human Tumor Atlas Network — a large, multisite patient dataset that includes ER- DCIS profiles acquired from whole lesions isolated by laser capture microdissection (LCM). These external LCM samples have been filtered to remove samples with low read counts across all genes. These additional samples in combination with our own will allow us to distinguish whether differences in handling and sample preservation give rise to tangibly higher quality transcriptomes when compared to existing archival samples.