There are numerous breast cancer prediction models, including IBIS/Tyrer-Cuzick (ITC), Gail model, BRCAPRO and, BOADICEA/CanRisk. Of these, ITC and BOADICEA/CanRisk appear to show superior discrimination. The issue for all these models is that they were developed primarily from non-Hispanic White (NHW) women. This issue is important as the weight of these risk factors likely differs between racial and ethnic groups. This paper by Kurian et al. evaluates the performance of the ITC model in over 90,000 women enrolled in the Women’s Health Initiative (WHI), with particular emphasis on Hispanic women.
The WHI is a prospective study of morbidity and mortality in postmenopausal women who were randomized into 1 of 3 clinical trials or an observational study. The study enrolled participants from 1993 to 1998 and followed for at least two decades. Hispanic women had a significantly lower risk than NH white women after accounting for all ITC risk factors. The researchers saw a slightly lower risk in non-Hispanic Black (NHB) compared to NHW. The authors also examined the US vs. foreign-born status in over 45,000 women for whom the information was present. The US vs. foreign-born was not associated with time to breast cancer diagnosis in any racial/ethnic group.
Until a new risk model becomes available, it is essential to understand how these risk models perform in different ethnic/racial groups. The ITC model has performed as well as or better than other models accounting for its widespread adoption. This is not to say that the ITC model is perfect, as it has been reported to overestimate breast cancer risk in women with atypical hyperplasia (ADH) and lobular carcinoma in situ (LCIS). However, this is the largest study evaluating the ITC model and the development of breast cancer. Taken as a whole, the ITC model performed exceptionally well. However, the model did overestimate the risk in Hispanic women. Hispanic women are known to have a lower breast cancer risk than NH White women, with foreign-born Hispanic women having a lower risk than US-born. The results of this study would suggest that some adjustment is required for the use in Hispanic women, particularly those born outside the US.