Second-look studies should be considered for all women with suspicious lesions on MRI. If the lesion is delineated with mammography or ultrasound, biopsy under imaging guidance using these modalities is preferred. MRI-guided biopsy is more expensive and difficult for the patient. However, the reported correlation rate of second-look ultrasound ranges from 23–57%; MRI-only lesions require an MRI-guided biopsy (15161718).

Informed consent should be obtained before the biopsy. Patients with a known contraindication for MRI or gadolinium administration should not have an MRI-guided biopsy. Patients with allergy to gadolinium or local anesthetics are also not suitable. Bleeding risk due to use of aspirin or anticoagulants or an underlying coagulation disorder is a relative contraindication for biopsy, requiring careful consideration of the risks and benefits. The patient should be able to remain prone during the biopsy for a minimum of 60 minutes. The possibility of a nonvisualized target lesion should be discussed with the patient. Even with successful visualization, the cancellation of a biopsy is sometimes necessary because of unforeseen safety issues. Complications such as hematoma, infection, and skin injury should also be discussed with the patient. For patients with breast implants, rupture is a possible complication of the biopsy.

The patient lies in a prone position, and a dedicated interventional breast coil is used. Either a grid or a pillar and post system is generally used. We focus on the grid system, which is most widely used (Fig. 1). The breast within the coil is compressed by the grid. Compression pressure is adjusted for stable immobilization while preserving blood flow. The approach to a posterior mass is often difficult, requiring that sufficient breast tissue be contained in the grid to maintain distance from the chest wall. Fixing a fiducial marker to the grid enhances target localization.

MRI-guided vacuum assisted biopsy (VAB) has advantages over standard core biopsy. The larger core size results in decreased sampling error and compensates for tissue shifting during needle placement (19). The European consensus group recommend a minimum 11-gauge probe size (20). The MRI-guided biopsy kit consists of an introducer stylet, obturator, introducer sheath, and needle guide. The needle guide, a cube-shaped plastic block with multiple holes for the biopsy device (Fig. 2), maintains the VAB device perpendicular to the grid.

Fig. 3 shows the MRI-guided biopsy sequence. A precontrast T1-weighted fat-saturation sequence is obtained to determine whether the target lesion is within the grid. If the target lesion is too small for detection on the precontrast image, anatomic landmarks are helpful for identification. If the precontrast image indicates that the lesion is inaccessible, the patient's position should be adjusted. After planning the proper approach route, postcontrast T1-weighted fat-saturation images are obtained. The thickness and in-plane resolution, which are similar to that of diagnostic imaging for accurate needle positioning, make localization easier. Sagittal and axial plane images are obtained. Alternatively, axial or sagittal plane images with perpendicular reconstruction can be used for lesion localization depending on the physician's preference. The entry site on the grid and depth from the skin to the lesion are measured using the fiducial marker as a reference. A worksheet provided by the manufacturer is helpful in manual localization (Fig. 4). The thickness of the needle guide should be included in the calculation. A computer-assisted diagnostic system that improves accuracy and speed of the procedure is also commercially available. About 8–13% of MRIguided biopsy target lesions are not visible at the time of biopsy (212223). If the target is not visible on the first postcontrast image, it is sometimes identified on a delayed image. Overpressure by the grid may interfere with breast perfusion and should be checked. A subtraction image also aids in lesion identification (Fig. 5). If the target lesion is still not visible, short-term followup is recommended at about 6 months.

After standard skin preparation, local anesthesia is administered. A small nick in the skin facilitates smooth entry by the VAB device. The introducer stylet within its sheath is inserted through the needle guide to the measured depth. A twisting motion is helpful to avoid skin tenting and tissue displacement. The stylet is removed and replaced by the obturator. T1-weighted fat-saturation images are then obtained to confirm the depth and position of the introducer (2425). In the case of insufficient depth, the introducer sheath can be advanced after reinsertion of the stylet; if advanced past the target lesion, the introducer is gradually withdrawn. In general, the optimal position is in the center of the target. Directional sampling can be performed with a peripheral location of the needle. When the introducer position is verified as correct, the obturator is exchanged with a VAB device. The European consensus group recommends no less than 24 samples for an 11-gauge or equivalent volume if a larger probe is used (20). Liberman (26) reported that an 11-gauge VAB device collects 100 mg, and a 9-gauge VAB device collects 200 mg. T1-weighted fat-saturation images are obtained immediately after the biopsy to evaluate adequacy. Image assessment can be difficult due to contrast washout, background enhancement, hemorrhage, and air, but careful review using anatomic landmarks improves the evaluation accuracy. An additional biopsy can be performed if the target sample is insufficient. Once the target sampling is acceptable, marker clip should be placed through the introducer sheath. A post-clip-insertion image may be obtained with MRI or mammography. The importance of marker clip insertion should never be underestimated. The marker clip facilitates mammographic or ultrasound-guided localization in place of MRI-guided localization for subsequent excisional biopsy. Moreover, if the entire target lesion is removed by VAB, the marker clip is the only way to identify the biopsy site. The breast is compressed for at least 15 minutes after the biopsy.

Major complications requiring surgical intervention seldom occur with MRI-guided biopsy. Complication rates are less than 5%. Bleeding and hematoma formation are most common, and can be controlled by compression. Other rare complications are skin laceration, vasovagal syncope, and infection. Termination of the procedure due to a complication is rare (1927282930313233343536).