Objective To investigate the correlation of genotypes with clinicopathological and prognostic characteristics of breast cancer patients.

Methods Breast cancer microarray datasets were downloaded from ArrayExpress, and integrated by Affy program and YuGene packages, named as AE-meta dataset. Breast cancer RNA sequencing data was downloaded from the cancer genome atlas (TCGA), named as TCGA breast cancer dataset. The absolute intrinsic molecular subtyping (AIMS) package was used to classify breast cancer subtypes. The correlation of genotypes with clinicopathological and prognostic data of breast cancer patients was analyzed.

Results The LumA patients were more common in older women with smaller-sized tumors, better differentiation and less lymph node invasion. The Basal patients were more common in young women with bigger-sized tumors, poor differentiation and less lymph node invasion. Conversely, the HER2 patients were no specific age of onset, with significant lymph node metastasis and poor differentiation. Compared with other types, LumA patients had the longest survival time and rarely suffered from recurrence or metastasis. HER2 and Basal patients had the shortest survival time, with relapse and metastasis. All AIMS had high risk of metastasis to the bone and lung. In patients with brain metastasis, Basal patients had the worst prognosis. In patients with liver metastasis, Basal and HER2 patients had worse prognosis than other genotypes.

Conclusion Breast cancer genotyping has a good clinical application value on reflecting the clinicopathological and prognostic features.