Second-look US after breast MRI.

Abstract

Two recent articles in UiM addressed second-look ultrasound (SL-US) after breast magnetic resonance imaging (MRI) 1 2 . You might be wondering about the usefulness of a second look. After all, isn't breast MRI the imaging method with the highest sensitivity for diagnosing breast cancer 3 ? That is true, but the specificity of modern MRI (with dynamic examination after contrast administration) is significantly lower than the sensitivity of up to 100 % in numerous studies. As a result of the high sensitivity, breast lesions are detected with breast MRI than with mammography and US. This includes lesions suspicious for malignancy and lesions classified as suspicious according to MRI-BIRADS that are only visible on MRI and not on mammography and real-time US. The German Radiological Society recently updated the recommendations for performing breast MRI 4 . Increasing numbers of breast MRI examinations and improved image quality result in an increase in the number of breast lesions that can only be detected with MRI 5 . A recent thorough overview 5 describes the diagnostic work-up that is necessary for these breast Re-assessment with mammography (...the retrospectoscope, the re-assessment of image data with greater knowledge, in this case knowledge of a breast lesion on MRI, is an important instrument for the quality control of imaging methods...) and the implementation of targeted second-look US are the first and most important additional diagnostic steps here 5 . Most authors therefore understand SL-US of the breast as a targeted real-time US examination of a breast lesion detected on breast MRI but not be detected on mammography and ultrasound of the breast performed prior to MRI 6 . However, others understand SL-US as the targeted sonographic clarification of a suspicious lesion detected on breast MRI in general even if there is no prior mammography and/or US information 7 . Retrospective mammography evaluation and SL-US can be used to determine whether the MRI-suspicious lesion can be histologically clarified via mammography-guided or US-guided biopsy in a manner that is simple, cost-effective, and minimally stressful for the patient. Both techniques have numerous advantages compared to the significantly less available MRI-guided biopsy technique that is reserved for lesions that can only be detected on MRI but not on SL-US (or in a retrospective mammography evaluation) 5 6 . So how good is SL-US, how high is the detection rate, what must be taken into consideration, how is the MRI-suspicious lesion located, and how is the US morphology interpreted? Ideally, SL-US should be performed by a single person with significant experience, knowledge and skill with respect to both breast US and breast MRI 5 . The MRI image data must be evaluated at the same time as the SL-US examination. Searching for a lesion described in an MRI report without image data is not productive. This is because not only the different image morphologies of lesions in MRI and US, but also in particular differences in lesion location between US and MRI must be taken into consideration. While the patient is in a prone position with the arms next to the body and the breast is subjected to no or only minimal compression in the dedicated breast coil during breast MRI, US of the breast is performed in a supine position or an inclined position often with the arms over the head. This results in significant differences in breast shape and can result in dramatic misinterpretation of the location of a lesion which affects the distance of the lesion from the breast wall (fascia), the distance of the lesion from the nipple, and the position of the lesion based on the clock 8 . Park et al. provide a very good overview of all aspects of SL-US 8 . It further complicates matters that the shape and size of lesions differ greatly between MRI and US as a result of the different patient positioning and US transducer compression. Many lesions are therefore measured smaller on US than on MRI 8 . The distance to known lesions (e. g. cysts and fibroadenomas) also differs significantly between MRI and US. However, observation of other lesions is important for detecting the target lesion on US. So how good is SL-US and how high is the detection rate? The study published in UiM 2 provides a detailed BIRADS classification of lesions on MRI and US. 132 lesions in 101 patients were classified as MRI-BIRADS 0 (50.8 %), BIRADS 3 (25 %), and BIRADS 4 (23.5 %) 2 . There were 19 malignant and 113 benign The MRI-US correlation was successful in 78 of 132 lesions (59 %) but was different between malignant lesions (correlation in 95 %) and benign lesions (correlation in 53 %). 1 of 6 BIRADS 4 lesions on MRI that could not be detected with US was a carcinoma. The second study examined whether the US characteristics of breast lesions primarily found with US differ from lesions found with second-look US 1 . The answer is yes since second-look lesions are smaller (7 mm vs. 9 mm), more frequently have a blurred margin, and more frequently have an immediate retroareolar position or a position in the breast periphery compared to first-look detected breast lesions 1 . A recently published meta-analysis 6 summarizes the results of 17 included studies regarding SL-US in breast The detection rate has a broad range between 22.6 % and 82.1 % (pooled 57.5 %). Focal breast lesions (mass lesions) had a higher detection rate than non-mass lesions. It is very important that the detection is higher for malignant lesions than for benign However, it is even more important that the lack of detection of a lesion with SL-US does not make it possible to rule out a malignant lesion 6 . Therefore, the most important basic points regarding the clarification of suspicious breast lesions primarily detected only on MRI (BIRADS 4 and 5) are summarized 5 6 : The US technique has brought significant improvements in recent years particularly in breast ultrasound. US detection of microcalcifications seems possible 9 10 , new US criteria for the diagnosis of malignant breast lesions have been introduced 11 , and elastography has become part of breast US diagnostics 12 13 14 15 16 . Can SL-US results be improved by these new techniques? There is only minimal literature in this regard. In a prospective study regarding the use of shear wave elastography in SL-US (73 patients, 96 BIRADS 4 and 5 lesions on breast MRI), 1 of 29 lesions could only be detected with elastography and the technique was helpful in determining the exact location of 5 additional lesions 17 . An initial study shows that SL-US guided by real-time MRI navigation (the MRI data were acquired with the patient in a supine position) improves the detection of MRI breast lesions 18 . Therefore, second-look US following breast MRI is useful. 57 % of MRI lesions can be located by SL-US and can be histologically clarified by US-guided biopsy in a manner that is cost-effective and minimally stressful for the patient. It remains to be seen whether modern US techniques can improve the detection rate.