MRI screening of women with hereditary predisposition to breast cancer: diagnostic performance and survival analysis.

Abstract

MRI screening has been shown to allow for an earlier diagnosis of breast cancer in asymptomatic women with proven or suspected mutations in breast cancer susceptibility genes. Major efforts are presently addressed to assess to which extent this improves high-risk women survival. We here discuss the article by Gareth et al (Breast Cancer Res Treat 145(3):663-672, 2014). Gareth and colleagues compared MRI sensitivity and specificity and clinical characteristics of breast cancers detected in their study with those reported in six similar prospective cohort screening studies. We here extended this analysis to a total of nine published cohort studies, including the High Breast Cancer Risk Italian Study 1 (HIBCRIT-1) for which we considered the final results published in 2011, instead of those of our interim report (2007) utilized by Gareth et al. Our updated analysis shows that in a total of 392 diagnosed breast cancers, 77 % (95 % confidence interval [CI] 73-81 %) were invasive and 52 % (95 % CI 46-58 %) were invasive grade 3. Only 23 % (95 % CI 18-28 %) of MRI-detected invasive breast cancers had metastatic lymph nodal involvement and 45 % (95 % CI 39-51 %) had a size ≤ 10 mm. The capability of MRI to detect invasive breast cancers at early stages could at least partly explain the significantly higher 10-year survival estimated by Gareth et al for asymptomatic highrisk women screened using MRI in the period 1997-2013 (95 %) compared with unscreened high-risk women diagnosed for breast cancer after 1990 and identified as BRCA1/BRCA2 mutation carriers in the years following diagnosis (74 %). It appears however worth noting that the evolution of therapeutic protocols applied to high-risk patients after the discovery of BRCA mutations in 1995-1997 could also have contributed to the observed difference in the survival of these two groups. On the other hand, we agree with Gareth et al that larger datasets are needed to evaluate to which extent improvements in the cancer detection impact on disease-free and overall survival of MRI-screened compared with mammography-alone-screened high-risk women.