Jules Henry Sumkin DO, Presenter: Scientific Advisory Board, Hologic, Inc

Margarita Louise Zuley MD, Abstract Co-Author: Research Grant, Hologic, Inc

David Gur PhD, Abstract Co-Author: Nothing to Disclose

In a stable clinical practice women undergoing baseline examinations constitute a small fraction of the screening population that is not routinely assessed as a special group. We report results from a prospective clinical study designed to specifically evaluate the impact of using digital breast tomosynthesis (DBT) during baseline screening. 

Under an institutional review board approved protocol, consenting women between 40 and 55 years of age scheduled for their initial/baseline screening mammogram underwent both digital mammography (FFDM) and DBT. The FFDM images (mode 1) and the FFDM+DBT images (mode 2) were interpreted independently in a balanced design by two experienced radiologists (out of 14 participating radiologists) under the standard clinical screening protocol. A woman was recalled for a diagnostic workup if either radiologist recommended recall. This analysis reports the results from the first 957 participants in this single institution prospective study. We assessed the proportion of recommended recalls (BIRADS 0 ratings) from clinical interpretations of FFDM alone and FFDM+DBT, as well as their diagnostic outcome.

Density distribution (BIRADS 1-4) of all participants was 4.6%, 40.3%, 51.3% and 3.8%, respectively. The fractions of women recommended for recall by either or both modalities (FFDM alone and FFDM+DBT) were 366/957 (38%) and 250/957 (26%). This represents is a 32% reduction in recall rate (p<0.01). The vast majority of the reduction was in density BIRADS 2 and 3. Recall rates varied substantially among radiologists (range 21% to 51% for FFDM alone and 11% to 41% for FFDM+DBT) and so did the reduction in recall rates (from -2% to 60%). Our clinical baseline recall rate is 37% range (18% to 55%) and 28% for FFDM+DBT. One DCIS was found on FFDM alone and one IDC on FFDM+DBT alone. PPV3 was substantially higher in women recalled as a result of FFDM+DBT (5/25=0.20 versus 5/16=0.31). There was no “learning” effect or “study knowledge” effect in our study.

Performance during baseline should be assessed carefully. FFDM+DBT resulted in a significant decrease in recall rates during baseline screening of young women. 

Recall rates during baseline screening need special attention and the use of DBT in our practice results in a significant, albeit not sufficient, reduction in recall rates.

Sumkin, J, Zuley, M, Gur, D, Recall Rate Reduction with Tomosynthesis during Baseline Screening Examinations – Assessment from a Prospective Screening Trial.  Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL. http://archive.rsna.org/2014/14005618.html