Radiological peer review of interval cancers in the East Anglian breast screening programme: What are we missing?

Abstract

Objective - To determine whether association between classification of interval cancers (false negative, minimal sign, true interval, occult) and tumour characteristics can indicate areas for improvement in effectiveness of screening. Setting - East Anglian regional Breast Screening Programme. Method - The consensus classification of interval cases, obtained through peer review, was compared with tumour characteristics: size; grade; and nodal status, and with radiological signs. A set of 923 cancers with known tumour characteristics, presenting in unscreened women, was available for comparison. Result - In total, 385 interval cases were analysed. Prognosis of interval cancers was at least as good as that of cancers in unscreened women, with fewer node positive cases. No significant relation was found between classification of cases and age of patients. True intervals tended to present later after a negative screen, although false negatives commonly presented earlier. Occult cancers were significantly smaller than false negatives. Analysis of histological type yielded little of relevance, and mammographic appearance was similar in all classifications. At diagnosis, cancers in the low risk group were more likely to be seen as spiculate masses, whereas high risk cases more often appeared as opacities. A higher proportion of interval cases had been assessed than in the normal screened population. Conclusion - Sensitivity may be improved by reducing false negative and minimal sign cases (21% of interval cancers in East Anglia). These will include tumours in all risk groups, and with all radiological features, thus no particular age group or radiological feature may be targeted for improvement. Misdiagnosis at assessment reduces sensitivity and therefore merits particular study.