Setting a lower risk threshold for surveillance within breast cancer family services

Background

Counselling, risk assessment and surveillance are provided for women with a significant family history of breast cancer through a network of clinical centres across the United Kingdom. Before 2004, the recommended minimum 'threshold' for significant familial risk was set by a number of guidelines issued, which broadly required one first-degree relative diagnosed with breast cancer before age 40 or two close relatives both diagnosed before age 60. In 2004, NICE issued detailed guidelines in which the age requirement for two affected relatives was removed. However, it is widely recognised that the evidence base for any specific minimum threshold is limited and that there is a need for empirical studies to validate current and future recommendations. That is the object of the present study.

Methods

Records of the four Scottish Breast Cancer Family clinics were scrutinised for the period January 1994–December 2003 to identify any women referred but discharged because the level of familial risk was judged to fall below the (pre-NICE) threshold. From dates of birth and dates of discharge, the number of women-years of observation (to December 2003) within each 5-year age group (35–39 years, 40–44 years, and so on) was calculated. With permission from the Privacy Committee, the list was then checked against Scottish Cancer Registry records and any breast cancers recorded were rechecked from hospital notes. Expected cancer rates for an age-matched Scottish population were derived from Cancer Registry Statistics.

Results

A total of 2,074 'low risk' women were identified, giving over 8,000 woman-years of observation. Twenty-eight invasive breast cancers were recorded while 14.4 would have been expected (relative risk = 1.9 assuming complete ascertainment). A further eight invasive breast cancers have been recorded since 2003 (records incomplete). One-third of the cancers were in women who would have met the new NICE criteria for surveillance, whereas only some 10% of the total cohort had 'NICE moderate' family histories. The great majority of the cancers occurred in women between age 45 and 56. For them the relative risk approached 2 even when 'NICE moderate' women were excluded.

Conclusion

The new NICE family history guidelines are more accurate than previous ones in identifying women who should be included in breast surveillance programmes, but consideration should be given to making some provision particularly for women between age 45 and 56 with 'limited' family histories of breast cancer. The cohort we have identified should continue to be followed up since cancers are continuing to accrue and each year provides a further 2,000+ woman-years of observation.