Enhancement on MDCT or MRI is due to angiogenesis and increased capillary permeability [6,7]. Enhancement of a lesion is an indication of proliferation regardless of whether it is malignant or benign. These lesions are not palpable, therefore imaging-guided biopsy is required for definitive diagnosis. There are three available ways to perform a biopsy under imaging guidance: stereo-guided, sonographically guided and MRI-guided procedures.

National Comprehensive Cancer Network (NCCN)’s guidelines have recommended the performance of MRI-guided needle sampling and/or wire localization of MRI-detected findings since 2007 [8]. Consequently, the MRI-guided biopsy is widely available in imaging centers in the United States. Although the MRI-guided biopsy is often used for MRI-detected lesions [9,10], MRI has limitations such as its low specificity, the fact that it is a time-consuming procedure, and the variation in its technical accuracy among institutions [9-12]. The availability of MRI-guided biopsy is limited in Asia, including in Japan [9-12]. On the other hand, CT-guided biopsy for MDCT detected lesions is not practical due to radiation exposure concerns [13-18].

The rationale for further modalities needed to evaluate MDCT- or MRI-detected lesions is latent formation of malignant tumors [19-25]. Since ultrasonography involves no radiation exposure, its repeated use is feasible. A second ultrasonography is performed in MDCT- or MRI-detected lesions. This use of ultrasonography is different from the initial ultrasonography involving the screening of the whole breast, and is known as targeted sonography or second-look sonography. When a lesion is detected on targeted sonography, other options in addition to MRI guidance are available, including sonographically guided biopsy, surgical excision with sonographic marking and follow-up with ultrasonography. Targeted sonography is an important tool as a breakthrough to the further examination; however, the detection rate with targeted sonography has varied in previous reports [19-24].

As for the 53 patients, the age and menopausal status of the pathologically confirmed group and the follow-up group are shown in Table ​Table1.1. Patients in the pathologically confirmed group were significantly older and more likely to be postmenopausal than those in the follow-up group. There were no significant differences in the follow-up period between the two groups (Table ​(Table11).

Example images of MDCT and targeted sonography are shown in Figures ​Figures2,2, ​,33 and ​and4.4. Figure ​Figure22 shows a suspicious lesion in the ipsilateral breast, and it is more than 3cm distant from the main lesion. MDCT shows linear enhancement surrounded by fat tissues, while targeted sonography depicts a hypoechoic lesion in the atrophic thin breast with a size of 14 x 2mm and abundant fat tissues surrounding the lesion. Ductal carcinoma in situ (DCIS) was suspected on the MDCT and ultrasonography image findings. It is often difficult to keep the lesion visible during an ultrasound-guided biopsy. We made an excision after marking on the skin of the lesion under ultrasound. The pathologic finding of this lesion was DCIS.

Figure ​Figure33 shows a clustered enhanced lesion in the ipsilateral breast on MDCT. The lesion was 1.5cm distant from the nipple lower outer quadrant in the breast on the MDCT. Targeted sonography showed a faintly hypoechoic lesion as an enhanced area. We excised it simultaneously with performing a lumpectomy of the known breast cancer after sonographically guided marking because it was located in a difficult place to perform sonographically guided biopsy. The pathologic finding of the lesion was columnar cell hyperplasia.

Figure ​Figure44 showed a well-demarcated enhanced tumor in the thin breast tissue which was 1.5cm distant from the nipple on the inner side. The tumor size was 5 × 2mm. Targeted sonography depicted a hypoechoic mass in the atrophic breast and sonographically guided biopsy was performed under ultrasound guidance. The pathologic finding was hyperplasia.

Sonography-guided fine needle aspiration was performed in 22 lesions, and a pathologic diagnosis was made in 45 lesions. Of the 22 lesions, fine needle aspiration cytology was benign or normal in 7, inadequate in 7, indeterminate in 2, suspicious for malignancy in 1, and malignant in 5. For a pathological diagnosis, 30 lesions underwent sonographically guided biopsy, while 9 lesions underwent surgical biopsy during and before the operation, 5 lesions had extended excisional ranges, and one underwent core needle biopsy.

Pathologic examinations revealed benign in 20 (44%) lesions and malignant in 25 lesions (56%), including one lesion diagnosed as cancer at another hospital. Of the 20 benign tumors pathologically diagnosed in our hospital, 7 were hyperplasia, 5 were intraductal papillomatosis, 3 were fibrocystic change, 2 were fibroadenoma and one was adenoma. Proliferation was not found in 2 lesions with discharge in the dilated duct. Of 25 malignant tumors, DCIS was found in 13 and invasive ductal carcinoma in 12 (Table ​(Table2).2). The mean follow-up period (SD) of 56 patients was 940.2 (553.1) days.

The maximum diameter of the detected lesions as determined by targeted sonography is shown in Table ​Table3.3. The maximum diameter was under 5.0mm in 26 lesions, 5.1-10.0mm in 25 lesions, 10.1-15.0mm in 6 lesions, and over 15.1mm in one lesion. The mean (SD) of the maximum tumor size was 6.5 (3.8) mm. As the lesion over 15.1mm was suspected to be present in the image of one patient, we measured the size of the hypoechoic lesion. There were no significant differences in size among the malignant, benign and follow-up groups (Table ​(Table33).

As shown in Table ​Table4,4, the numbers of lesions with depth:width ratios under and over 0.7 were 45 and 13, respectively. There were no significant differences in the depth:width ratio in the three groups (Table ​(Table44).

Based on our experience, we developed a diagnostic imaging strategy for MDCT- or MRI-detected lesions in breast cancer (Figure ​(Figure55).

Pathologic diagnosis is required for MDCT- or MRI-detected lesions because the imaging finding is not the definitive diagnosis. The MDCT- or MRI-detected lesions are not palpable; therefore biopsy is performed under imaging guidance. Regardless of the availability of MRI-guided biopsy, it is practical that targeted sonography is the first procedure to be considered for MDCT- or MRI-detected lesions. The sonographically guided technique provides an advantage. Ultrasonography is the most frequently used and easy-to-use diagnostic tool in daily practice. Once the lesion is depicted on sonography, targeted sonography provides more options for further evaluation including sonographically guided biopsy, surgical excision after marking under sonography and follow-up with ultrasonography.

The guiding principle in diagnosis depends on whether the lesion is detected with targeted sonography. Because these lesions are difficult to depict by initial mammography or ultrasonography, the detection rate of lesions with targeted sonography varies among institutions [19-24]. There are tips for detecting MDCT- or MRI-detected lesions with targeted sonography. First, the location difference due to the position should be considered. MRI is performed in the prone position, while MDCT is performed in the supine or prone position, depending on the institutions. We performed MDCT in the prone position. The positional difference is larger in the prone position than in the supine position. Attention should be paid to the deviation caused by the arm position even in the supine position. Second, the anatomical positional relation is important. The distance from the nipple, the thickness of the breast gland and the depth of the tumor in the gland should all be referenced. As the nipple is a useful milestone, we use it as a center with the medial or lateral and caudal or cranial directions used to describe locations in reference to it. Breast thickness is also helpful. These factors are not affected by positional changes. Third, tumor size, the morphology of the lesion and the structure of the breast gland surrounding the lesion are also useful information for predicting the positional difference. To determine whether the enhanced lesion is nodular, linear or segmental, and whether the breast structure surrounding the lesion is fat or abundant in the breast gland, it is helpful to perform targeted sonography. Also, these attributes are not affected by positional changes. These techniques are helpful to improve the detection rate.

There were no significant differences in the size or the depth:width ratio in the benign, malignant, and follow-up groups in our study. The mean tumor size was 6.5mm and the mean depth:width ratio was 0.56, suggesting a relatively small and flat lesion. The MDCT or MRI detected lesion, however, has few distinct features; thus it may be found falsely negative on the initial imaging findings. We also performed cytology in 22 lesions in order to obtain a diagnosis. It is often difficult to obtain a large enough sample with fine needle aspiration, and the reliability of this technique is poor if we fail to obtain a sufficient sample. Also, cytology is not sensitive to low-grade tumors. Suspicious lesions warrant pathologic evaluation.

In this study, more than half of the pathologic findings were revealed to be malignant, and half of them were DCIS. The higher probability of malignancy requires preoperative assessment of the extent of the disease because it may alter the surgical management. We reported that MDCT contributed to the detection of occult breast cancer in 2.6% of contralateral breasts [26]. MDCT or MRI in conjunction with targeted sonography is a useful technique for detecting early carcinoma in women who are at increased breast cancer risk.

The pathologic findings revealed that there was no proliferation in 2 out of 45 lesions with discharge in the dilated duct. Because no proven proliferating lesion was found, concern remains that the lesion observed might not be the MDCT- or MRI-detected lesion. La Trenta suggests that it is not unlikely that undetected lesions are malignant [21]. It is difficult to perform core biopsy or small excisional biopsy when the lesion is not detected. At the same time, wide excisional biopsy is rather excessive treatment. A careful follow-up is always necessary to avoid delayed or missed diagnoses. The mean follow-up period for lesions unidentified by targeted sonography was 916days in this study. So far there are no newly developed lesions or tumor growth.

It has recently been reported that real-time virtual sonography (RVS) can synchronize a sonography image and an MRI or CT image of the same section in real time [12,27]. Accurate comparison of individual positions is a useful technique for confirming a lesion. Ultrasonography greatly depends on the technique, knowledge and experience of the operator. RVS reduces the differences among operators and increases the physician’s confidence about the MDCT- or MRI-detected lesion. When the lesion is not detected with targeted sonography or RVS, MRI-guided biopsy is indicated; this option or, at the very least, follow-up is required.

We found that significantly more older patients had pathologically confirmed tumors, and they were more likely to be postmenopausal than those in the follow-up group. It is reported that younger patients have higher background parenchymal enhancement and a higher proportion of nodular parenchymal enhancement patterns [28-30]. Since younger and premenopausal groups were suspected to have fibroglandular changes, these patients require follow-up.