Analysis of reproducibility of subjective grading systems for breast carcinoma

• Published in: Journal of clinical pathology Vol. 32; no. 10; pp. 979 - 985
• Main Authors: Stenkvist, B, Westman-Naeser, S, Vegelius, J, Holmquist, J, Nordin, B, Bengtsson, E, Eriksson, O
• Format: Journal Article
• Language: English
• Published: England BMJ Publishing Group Ltd and Association of Clinical Pathologists 01-10-1979; BMJ Publishing Group LTD
• Subjects: Breast Neoplasms - pathology; Cell Nucleus - pathology; Humans; Methods; Mitosis; World Health Organization
• ISSN: 0021-9746; 1472-4146
• DOI: 10.1136/jcp.32.10.979

The inter- and intra-observer reproducibility of three grading systems for breast cancer has been analysed: those of the World Health Organisation, of Black, and of Hartveit. In addition, the components forming the basis of these grading systems were analysed separately with regard to reproducibility and interrelationship. In this analysis, degree of differentiation was also included as a parameter. The grading systems of WHO and of Black gave a stratification of the material into three categories of tumours, but that of Hartveit did not. All three systems had a low inter- and intra-observer reproducibility. Truncated component analysis indicated that the grading systems of WHO and of Black are closely related to each other and to the 'nuclear lobulation' component of Hartveit's system. The components 'pleomorphism', 'mitotic frequency', and 'hyperchromasia' of the WHO system were closely interrelated. Descriptors such as 'differentiation' and 'tubule formation' were interrelated but acted in a different direction from the others.