Reply letter to Leong LT

Dear Editor,

We read with interest the letter on our article [1] by Dr. Leong, and agree on the need for balancing the likelihood of false-positive versus false-negative cases in the screening scenario [2]. On the other hand, the fact that women would accept the likelihood of false-positive results [3,4] reflects an emotional attitude rather than a goal of a screening program, in which the recall rate should be minimized (e.g., <5% according to the European guidelines for quality assurance in breast cancer screening and diagnosis [5]). While acknowledging that women’s well being is of paramount importance, we believe that the above argument should be used with caution when discussing on the balance between screening sensitivity and specificity.

We do not agree with the corresponding colleague on the potential influence of our results on screening guidelines, since our work covered a completely different clinical scenario, i.e. the preoperative assessment of breast cancer using contrast-enhanced magnetic resonance imaging (CEMRI) versus the combination of unenhanced magnetic resonance imaging (UMRI) with digital breast tomosynthesis (DBT) (UMRI + DBT) [1]. In this setting, reducing the false-positives of preoperative CEMRI is of paramount importance, as its disappointing specificity is still a matter of debate as a potential source of surgical overtreatment [6]. We confirm that using UMRI + DBT might be helpful at least in decreasing the number of “second-look” imaging procedures or biopsies prompted by CEMRI. On the other hand, a reasonably limited increase in small false-negatives such as the one we observed with UMRI + DBT (Tab. 3 and Tab. 4 in the article) is of questionable importance in the preoperative setting, i.e. in a context in which the index lesion prompting imaging is already known and has been already biopsied. We believe that the observations by Dr. Leong cannot reasonably refer to the results of our study.

Finally, we did not propose to extract UMRI images from a full CEMRI protocol, which would be of questionable significance once contrast medium has been administered. Rather, we proposed performing UMRI only, in line with current research on abbreviated protocols as a mean to save costs of breast magnetic resonance imaging (MRI), and make the examination more rapid and accessible to patients [1]. Sparse MRI methods proposed by the corresponding colleague sound interesting as a technical mean to accelerate individual MRI sequences. However, regardless of technical characteristic of each sequence, we need something different, i.e. an MRI examination balancing fast and safe acquisition with diagnostic efficacy in a certain clinical scenario.