Using Big Data to Improve Metastatic Breast Cancer Research
Genomic data may improve the selection of more precise models for metastatic breast cancer research.
Researchers are investigating ways to use big data to improve metastatic breast cancer research, according to a recent study published in Nature Communications.
The availability of large-scale genomic data provides an opportunity to evaluate their use in metastatic cancer research. With very few therapeutic options available for this patient population, cancer metastasis leads to death in approximately 90% of patients.
Currently, cell lines are used as models to study metastatic cancer, but the extent to which cell lines can capture the genetic makeup of tumors is unknown, according to the study authors.
For the study, the researchers used data from genomic databases to draw comparisons between breast cancer cell lines and tumor samples. By analyzing the data, they found substantial genomic differences between MDA-MB-231, a cancer cell line used in nearly all metastatic breast cancer research, and metastatic tumor samples from patients.
Key differences suggested that cell lines poorly recaptured somatic mutation patterns of metastatic breast cancer samples, whereas their copy number variation profiles were more consistent. Additionally, the researchers noted that cell lines carried many specific genomic alternations, possibly due to culture effects.
“I couldn’t believe the result,” senior author Bin Chen, PhD, assistant professor in the College of Human Medicine, said in a press release about the study. “All evidence pointed to large differences between the two. But, on the flip side, we were able to identify other cell lines that closely resembled the tumors and could be considered, along with other criteria, as better options for this research.”
The researchers have previously proposed different computational methods to measure the similarity between cell lines and patient samples. They indicated that gene expression is one of the most informative features to predict drug response and weighting cell lines based on their transcriptome similarity with patient samples can increase predictive power in gene expression-based drug discovery.
The analysis suggested that organoids, which use 3D tissue cultures, resemble the transcriptome of patient samples more closely than cell lines, according to the authors. These organoids are able to capture more of the complexities of how tumors form and grow, the researchers noted. According to the study, the organoids’ high genomic similarity with patient samples warrants further investigation.
“We hope that the recommendations in this study may facilitate improved precision in selecting relevant cell lines for modeling in metastatic breast cancer research, which may accelerate the translational research,” the researchers concluded.
Liu K, Newbury P, Glicksberg BS, Evaluating cell lines as models for metastatic breast cancer through integrative analysis of genomic data. Nature Communications. 2019. https://www.nature.com/articles/s41467-019-10148-6
Big data helps identify better ways to research breast cancer’s spread [news release]. Michigan State University. https://msutoday.msu.edu/news/2019/big-data-helps-identify-better-way-to-research-breast-cancers-spread/. Accessed May 15, 2019.