Astronomical algorithms for automated analysis of tissue protein expression in breast cancer

• Published in: British journal of cancer Vol. 108; no. 3; pp. 602 - 612
• Main Authors: Ali, H R, Irwin, M, Morris, L, Dawson, S-J, Blows, F M, Provenzano, E, Mahler-Araujo, B, Pharoah, P D, Walton, N A, Brenton, J D, Caldas, C
• Format: Journal Article
• Language: English
• Published: London Nature Publishing Group UK 19.02.2013; Nature Publishing Group
• Subjects: 631/114; 692/699/67/1347; 692/700/139/422; Adult; Aged; Algorithms; Automation; Biological and medical sciences; Biomarkers, Tumor - metabolism; Biomedical and Life Sciences; Biomedicine; Breast cancer; Breast Neoplasms - metabolism; Breast Neoplasms - mortality; Breast Neoplasms - pathology; Cancer Research; Cell Membrane - metabolism; Cell Nucleus - metabolism; Cohort Studies; Cytoplasm - metabolism; Drug Resistance; Epidemiology; Female; Gynecology. Andrology. Obstetrics; Humans; Image Processing, Computer-Assisted; Mammary gland diseases; Medical sciences; Middle Aged; Molecular Diagnostics; Molecular Medicine; Observer Variation; Oncology; Prognosis; Proto-Oncogene Proteins c-bcl-2 - metabolism; Receptor, ErbB-2 - metabolism; Receptors, Estrogen - metabolism; Reproducibility of Results; Survival Rate; Tissue Array Analysis; Tumors; Young Adult; breast cancer; immunohistochemistry; digital pathology; image analysis; systems pathology; Mammary gland diseases; Breast disease; Cancerology; Breast cancer; Malignant tumor; Algorithm; Protein; Cancer
• ISSN: 0007-0920; 1532-1827; 1532-1827
• DOI: 10.1038/bjc.2012.558

Background: High-throughput evaluation of tissue biomarkers in oncology has been greatly accelerated by the widespread use of tissue microarrays (TMAs) and immunohistochemistry. Although TMAs have the potential to facilitate protein expression profiling on a scale to rival experiments of tumour transcriptomes, the bottleneck and imprecision of manually scoring TMAs has impeded progress. Methods: We report image analysis algorithms adapted from astronomy for the precise automated analysis of IHC in all subcellular compartments. The power of this technique is demonstrated using over 2000 breast tumours and comparing quantitative automated scores against manual assessment by pathologists. Results: All continuous automated scores showed good correlation with their corresponding ordinal manual scores. For oestrogen receptor (ER), the correlation was 0.82, P <0.0001, for BCL2 0.72, P <0.0001 and for HER2 0.62, P <0.0001. Automated scores showed excellent concordance with manual scores for the unsupervised assignment of cases to ‘positive’ or ‘negative’ categories with agreement rates of up to 96%. Conclusion: The adaptation of astronomical algorithms coupled with their application to large annotated study cohorts, constitutes a powerful tool for the realisation of the enormous potential of digital pathology.