MRI has been increasingly used for preoperative evaluation of lesions in breast cancer patients. Although MRI could provide a clear imaging with high sensitivity especially for non-mass lesions in breast, its value in surgery decision-making and prognosis remains controversial [16–18]. In addition, there has been limited evidence on the application of MRI in DCIS. In this study, we demonstrated that younger age, family history of breast cancer, manifestation at diagnosis, and lesion patterns were associated with the choice of preoperative MRI evaluation of DCIS. MRI examination was related with a higher rate of BCS, but had limited impact on disease outcome.
Mammography played a pivotal role in breast disease screening. However, it is less sensitive to non-calcified lesions, which is commonly seen in DCIS patients [19, 20]. With the MRI technical development, it shows a higher sensitivity in detecting DCIS lesions than mammography and ultrasound [8, 10], especially for high-level DCIS [21]. Bae et al. found that preoperative MRI in DCIS patients could detect additional lesions in both ipsilateral and contralateral breast [22]. This advantage was more distinct in patients over 50 years old or with lesions larger than 2.5cm. Bijker et al. and Tunonde Lara et al. found that DCIS patients under 40 years old were more likely to be presented with non-mass symptoms including nipple discharge or breast pain [23, 24]. According to our results, MRI was more likely to be applied in patients younger than 40 years, with a family history of breast cancer, presenting with non-mass clinical manifestation, multifocal lesions, and high-grade disease, which was consistent with previous studies. The size of tumor and other pathological factors including hormone-receptor status or Ki-67 showed no significant impact on the decision of preoperative MRI examination.
Patients with DCIS are eligible for BCS with equivalent safety and survival benefit compared with mastectomy [25]. Preoperative breast MRI has been used in breast cancer patients to help decision-making on surgical management, thus to reduce the rate of BCS positive margin and the possibility of BCS failure [26]. Lehman et al. demonstrated that MRI showed superior ability in detecting the extent of DCIS lesions compared with mammography or ultrasound, with acceptable specificity [21]. Renata Faermann et al. found that the volume ratio measurement of MRI could assist patients in choosing the best type of surgery and can convert a proportion of patients from mastectomy to BCS [27]. Our study used PSM to balance clinicopathological factors between groups and found that DCIS patients received preoperative MRI examination had a higher BCS rate compared with patients received no MRI. The possible explanations included providing a more accurate assessment of lesions and excluding the existence of multifocal lesions by MRI, thereby increasing surgeon’s willing and confidence to make a reasonable surgical plan of BCS.
Among well-established risk factors for recurrence in DCIS patients, surgical margin status was described as the most important one [28, 29]. Sufficient evidence had shown that complete resection of tumor with negative surgical margins followed by radiotherapy could reduce the risk of ipsilateral recurrence for DCIS patients [30]. Therefore, a secondary mastectomy should be performed in patients who planned BCS if negative surgical margins were not available. Although preoperative MRI is important for determining BCS, its role in achieving negative surgical margin remain unanswered. Previous study by Lehman et al. found that MRI showed an advantage in tumor detection, size measurement, and to reduce the chance of positive surgical margin in DCIS patients [21]. However, a large meta-analysis demonstrated no statistical difference in positive surgical margins for DCIS patients regardless of receiving preoperative MRI [31]. In our study, we found that MRI examination was not associated with a lower rate of positive surgical margin among planned BCS patients. The rate of a secondary mastectomy was also similar between MRI and non-MRI cohorts, which was consistent with previous studies. One possible explanation could be the choice of candidates to received planned BCS. For patients in the non-MRI cohort, surgeon may choose patients with conservative strategy to receive BCS. Patients with uncertainty of achieving negative surgical margin would receive mastectomy directly.
Currently, the treatment pattern for DCIS includes surgery, postoperative radiotherapy, and endocrine therapy. An excellent clinical outcome could be achieved following standard therapy, with over 80% DCIS patients would not experience local recurrence or distant metastasis during 15-year follow-up [32]. In our cohort, all patients receiving standard treatment and regular follow-up showed superior RFS and BCFI. No significant difference in prognosis between MRI and non-MRI cohorts was observed in our study.
Our study still has several limitations. Firstly, this is a retrospective study. There was no consensus across breast centers on preoperative evaluation standard. The decision of receiving preoperative MRI examination or not largely depends on surgeon’s experience. Moreover, the follow-up period in our study is limited, considering the long natural history of DCIS. More large-scaled prospective studies are needed to further explore the role of preoperative MRI in DCIS patients’ management.
In summary, our study evaluated the role of preoperative MRI in determining local management of DCIS patients and found that preoperative MRI evaluation were associated with a higher rate of BCS, but had no impact on disease outcome.