Systematic Bias in Genomic Classification Due to Contaminating Non-neoplastic Tissue in Breast Tumor Samples
Author Information
Author(s): Elloumi Fathi, Hu Zhiyuan, Li Yan, Parker Joel S, Gulley Margaret L, Amos Keith D, Troester Melissa A
Primary Institution: University of North Carolina at Chapel Hill
Hypothesis
How does normal tissue contamination in tumor samples affect genomic predictions for breast cancer?
Conclusion
Normal tissue contamination in tumor samples is a significant source of bias in genomic predictors, affecting their accuracy.
Supporting Evidence
- Normal tissue contamination caused misclassification of tumors in all genomic predictors evaluated.
- The PAM50 assay showed predictable bias direction, allowing for effective correction.
- Adjustments for normal tissue contamination improved sensitivity and negative predictive value in genomic predictions.
- Different genomic predictors exhibited varying vulnerabilities to normal tissue bias.
- Quality control standards can enhance the reliability of genomic assays.
Takeaway
When doctors test for breast cancer, they sometimes mix normal tissue with tumor tissue, which can lead to wrong results. This study shows that this mix-up can change how doctors predict the cancer's behavior.
Methodology
The study analyzed 55 tumor samples and paired adjacent normal tissue to evaluate how normal tissue contamination affected genomic predictions.
Potential Biases
Normal tissue contamination can lead to misclassification of tumor subtypes and risk assessments.
Limitations
The study relies on paired samples, which are not always available in clinical settings, and the results may not generalize to all breast cancer types.
Participant Demographics
The study involved patients undergoing surgery for invasive breast cancer at UNC Hospitals.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website