To evaluate the efficacy of combined breast MRI in the detection of synchronous contralateral breast cancer and the impact of consequent metachronous contralateral breast cancer, we performed a case control comparison study, which enrolled 1468 primary operable breast cancer patients with two different groups of pre-operative breast imaging modalities and received surgery at a single institution, with a mean follow-up of 102 ± 28.6 months. We found that, with the addition of breast MRI to conventional imaging studies, more surgical excisions were performed in contralateral breasts, more synchronous contralateral breast cancer detected, and resulted in numerical less metachronous contralateral breast cancer during follow-up.
In Brennan et al.’s study for evaluation of the contralateral breast cancer detection by pre-operative MRI, the incidence of suspicious MR imaging findings was 9.3%, with a PPV 47.9%, and a false-positive rate of 52%[31]. In our current study, MR imaging detected 70(9.5%) contralateral lesions in 735 patients, and 21 of them were proved to be malignancy. The PPV of the MRI detected contralateral lesions in our study was 48.8%, which is compatible to literature reported series (Table 3).
The detection of additional abnormal enhancing lesions in breast MRI might be “problematic” for subsequent patient management. Sometimes it is confusing whether these lesions should be ignored, followed up, or referred for biopsy. And the image-guided biopsy before surgery is increasing when breast MRI was incorporated in the pre-operative evaluation[1; 2]. As showed in Fig 1, in the conventional imaging cohort, only 10 (1.4%) patients received contralateral breast surgeries. In the combined MRI group, 44(6%) patients received contralateral breast surgeries when primary index breast cancer surgery was performed (P<0.01). However, the “false positive” rate for the additional lesions detected by MRI is not low as observed in our current study, which around 51.2% of the so called “occult contralateral breast lesions” turned out to be benign neoplasms after surgical excisions (Fig 1), and most of them were fibroadenomas (34.5%) and fibrocystic change (43.5%).
Second-look ultrasound and core needle biopsy for suspicious lesion or MRI-guided stereostatic biopsy could obtain tissue diagnosis and decrease unnecessary operations[32]. However, biopsy results of proliferative lesion like papilloma or atypia of the synchronous contralateral lesions were suggested to be excised at the same time when primary index breast cancer operation is performed. As one of our patients was found to have papilloma at core needle biopsy specimen of the contralateral breast lesion, DCIS arising from papilloma was diagnosed in the final lumpectomy pathology.
Controversies existed that MRI was reported to have increase ipsilateral mastectomy rate and bilateral mastectomy rate[33]. In current case control analysis, the mastectomy rate in the combined MRI cohort of patients was not significantly increased when compared with the conventional imaging cohort (52.7% versus 48.6%, P=0.13). The higher number of patients received bilateral mastectomy in combined MRI group were in part due to higher synchronous contralateral breast cancer detection before operation. However, a substantial of patients received bilateral mastectomy due to psychologic stress of the suspicious lesions detected by breast MRI. We also observe a numerical but not statistically significant increase of bilateral mastectomy cases in patients received pre-operative evaluation with MRI than conventional imaging only group (12 out of 735 versus 6 out of 733, P=0.24). A significantly increase of breast reconstructions (39.8% versus 17.1%, P<0.01) was observed in patients received combined MRI evaluation than conventional breast imaging alone. Other study had suggested that pre-operative MRI study may increase the mastectomy rate and therefore increased the breast reconstruction rate[34; 35].
In our total 1468 breast cancer patients, 29 (2%) synchronous contralateral breast cancer was detected. Most (58.6%, 17/29) of them were DCIS lesions. Compared with the pathology of the primary index breast cancer, either in the combined MRI group (61.9% versus 15.9%, P<0.01) or in the conventional breast imaging cohort (50% versus 16.6%), the detected synchronous contralateral breast cancer were associated with higher in situ carcinoma. These findings could explain why synchronous contralateral breast cancer, which usually presented in an earlier stage than the primary breast cancer[13], was a challenge for clinicians for early detection.
During a mean follow-up of 102±28.6 months, we observed 38 patients developed new metachronous contralateral breast cancer, and 22 (3%) of them were in initial conventional pre-operative imaging group, and 16(2.2%) of them in combined MRI group (Fig 3). Contralateral metachronous cancer could be either synchronous occult cancers not detected by pre-operative imaging and diagnosed during follow-up or could be new contralateral breast cancer developed after initial primary index breast cancer operations. The higher rate of contralateral synchronous breast cancer detected and lower number of contralateral metachronous breast cancer found during follow-up in the pre-operative combined with MRI group (Fig 1, 3) supported the hypothesis that combined with pre-operative MRI could detected more synchronous breast cancer than conventional imaging alone, and during follow-up some occult lesions not detected in contralateral breast at time of primary index cancer operations would be diagnosed as metachronous breast cancer.
Our study was limited due to its retrospective nature and possible selection bias. Patients with contralateral occult lesions detected by conventional imaging alone or combined with MRI did not received surgical excision in every case to confirm the nature of the lesion. This might, therefore, underestimate the actual synchronous breast cancer rate. However, we did collected 733 primary operable breast cancer patients with pre-operative conventional breast imaging, and 735 combined MRI patients who received surgical intervention at a single institution with detailed pathologic report and with a mean follow-up of 102±28.6 months, which enable us to complete this case control comparison study with the occurrence of new metachronous contralateral breast cancer. We provide solid evidence that adding breast MRI increase the contralateral occult synchronous breast cancer detection rate from 1.1% to 2.9%, and adding MRI increased about 1.8% of contralateral occult breast cancer detection rate (Table 2).