Background
The PAM50 assay is used routinely in clinical practice to determine breast cancer prognosis and management; however, research assessing how technical variation and intratumoral heterogeneity contribute to misclassification and reproducibility of these tests is limited.
Conclusion
The PAM50 assay achieved very high technical reproducibility for breast cancer subtyping and ROR-P, but intratumoral heterogeneity is revealed by the assay in a small proportion of cases.
Methods
We evaluated the impact of intratumoral heterogeneity on the reproducibility of
Results
Technical replicates (N = 144) achieved 93% agreement for ROR-P group and 90% agreement on PAM50 subtype. For spatially distinct biological replicates (N = 40 intratumoral replicates), agreement was lower (81% for ROR-P and 76% for PAM50 subtype). The Euclidean distances between discordant technical replicates were bimodal, with discordant samples showing higher Euclidian distance and biologic heterogeneity.