Interval breast cancer: is it a different type of breast cancer?

We have compared tumour type, tumour size, tumour grade and axillary lymph node status in three groups of women, 230 interval breast cancers (IC) in the West Sussex Breast Screening programme and 625 screen detected (SD) cancers and 916 symptomatic (S) cancers treated at Worthing Hospital between July 1989 to April 1996. Our true interval cancer detection rates were 5.28, 11.28 and 15.3 per 10,000 screened women for the 1st, 2nd and 3rd year after screening. The proportionate incidences of true interval cancer were 29%, 61% and 82% for the 1st, 2nd and 3rd year, similar to others' programmes in UK. In our programme a large proportion (42%) of IC and more than half of the true IC presented in the 3rd year after screening. Out of 230 interval cancers, 40% (90) were unclassifiable, the remaining 60% (140) were classified as: True interval cancers (T) 54% (76), False Negative Subtle (FNS) 12% (16), Occult (O) 12% (17), and 22% (31) as False Negative (FN). Analysis of interval cancers according to their classification did not demonstrate any significant difference with respect to tumour size (chi2 5.59, df 4, P=0.22), tumour grade (chi2 5.29, df 4, P=0.25) and axillary node status (chi2 3.16, df 4, P=0.53) thus establishing interval cancers as a single group. Invasive ductal carcinoma of no specific type was the main tumour type in all three groups. Analysis of variance (ANOVA) showed significant differences in size between the groups (df 2, F=71.36, p<0.0001). Symptomatic cancers were 1.19 times the size of IC while SD were 0.83 times the size of IC. The difference in groups in terms of tumour grade was significant (Kruskal-Wallis test chi2 33.31, df 2, P<0.0001). The incidence of grade 2 tumours was similar in the three groups while a third of the IC and S were grade 3 tumours. Comparison of axillary node status showed a significant difference between the three groups (chi2 26.59, df 2, P<0.0001). When means and 75th percentiles were compared IC had the greatest number of positive nodes while SD had the smallest number of positive nodes. Interval cancers are the middle spectrum between symptomatic and screen detected breast cancers and represent small cancers (<10 mm) not detected at the time of screening and de novo cancers developing in the screening interval. The need for improving the sensitivity of current screening methods and identifying newer methods of breast cancer detection is highlighted by our study.