Minimally Invasive Image-Guided Breast Biopsy and Ablation

Introduction

Screening mammography has been shown to decrease the incidence of mortality associated with breast cancer.¹ Since the initial implementation of widespread screening mammography programs, breast imaging has evolved to include full-field digital mammography (FFDM), ultrasound (US), magnetic resonance imaging (MRI), and more recently, molecular imaging techniques including breast-specific gamma imaging (BSGI) and positron emission mammography (PEM) to aid in the evaluation of breast pathology. The ultimate goal of these technologic advances is to allow for the detection of small and early-stage breast cancer, thus providing patients with the best chance of cure and the least amount of treatment-associated morbidity. However, with the detection of smaller lesions comes the challenge of developing methods to arrive at a pathologic diagnosis in an accurate, cost-effective, and safe manner. Wire-guided surgical breast biopsy was the gold standard for the diagnosis of nonpalpable radiographically detected breast lesions, but given that 80% of indeterminate breast lesions that undergo biopsy are subsequently found to be benign, development of less invasive, more cost-effective techniques was needed.²

Breast imaging and intervention has undergone great changes over the past several decades because of the development of many image-guided minimally invasive technologies allowing for accurate and cost-effective diagnosis of breast cancer and high-risk lesions. Such techniques are now available utilizing all traditional forms of breast imaging, including mammography (stereotactic), US, MRI, and now most recently, more novel approaches such as nuclear medicine guidance for molecular breast imaging techniques. As methods for pathologic diagnosis of breast cancer have trended toward minimally invasive techniques, so have potential options for treatment. Various methods of image-guided ablation of small focal breast cancers are now under investigation, with promising initial results. This chapter will review available methods for minimally invasive breast biopsy, including the various modalities of imaging guidance and equipment, and provide an algorithm for clinicians to best tailor their biopsy approach. A brief review of new ablative treatment techniques, specifically radiofrequency ablation (RFA) and cryoablation, will be provided.

Stereotactic Interventions

Targeting mammographically detected lesions for minimally invasive biopsy employs stereotactic guidance. The most common type of suspicious mammographic abnormality targeted for biopsy via this method are microcalcifications, although any suspicious abnormality seen solely or best with mammography, including masses or asymmetries, can be targeted and biopsied via stereotaxis. In a large multicenter trial performed in 2003, investigators found that 70% of stereotactic vacuum-assisted biopsy (VAB) was performed for lesions containing suspicious microcalcifications.³

For a stereotactic breast biopsy, either a dedicated table is needed or an attachment to a digital mammography unit and a special chair to allow for proper patient positioning. If a dedicated table is used, the patient is prone with the breast positioned dependently through an aperture in the table (Fig. 155-1). Attached to the underside of the table is a mammography unit on an articulating arm, which can be moved such that +15-degree and −15-degree images are obtained. These angles are used by the computer to determine the coordinates of the targeted lesions to within 1 mm of accuracy. If a standard mammography unit with an attachment allowing stereotactic guidance is used, a chair can be positioned to optimally target the lesion. The advantages of a dedicated stereotactic table include greater comfort for the patient and physician and no risk of a vasovagal reaction. However, it requires a dedicated room and space. A mammography unit allows for more efficient use of space and equipment because it can be used for routine imaging when not being used for biopsy. This latter solution may be best suited for practices with lower volumes where use of a mammography unit for biopsy will not impact workflow of screening and diagnostic mammography.

FIGURE 155-1 Patient lying in prone position on a dedicated stereotactic biopsy table.

The position used to approach the lesion is decided pre-procedurally by examining the patient’s diagnostic mammogram. The decision as to approach should be based on the projection in which the lesion is best visualized, not necessarily the shortest distance, as well as considerations for ease of patient positioning. Images of the lesion are obtained at stereotactic angles, typically 15 or 20 degrees, and computer-generated x, y, and z coordinates are obtained. These coordinates place the targeted lesion in the center of the biopsy chamber of the probe. The biopsy probe is then loaded onto a base that is mobilized to the determined x and y coordinates. To minimize bleeding during the procedure, care should to be taken to ensure there is no (or minimal) blood vessel overlying the lesion and that the depth coordinate (z) is neither too superficial nor too deep. A too-superficial depth runs the risk of including skin within the biopsy specimen, and a depth too deep may make it impossible to sample the lesion without compromising the equipment if the biopsy probe should contact the image detector, potentially resulting in misalignment of the equipment. However, there are certainly strategies to overcome these challenges. For a superficial lesion, the probe can be advanced to a position where the entire biopsy chamber is just beyond the skin. This will allow for inclusion of the targeted lesion, even though the z coordinate will be deeper than determined. For a depth that is too deep, the patient can be repositioned.

The overlying skin is prepped in a sterile fashion, followed by injection of topical anesthetic into the skin and subcutaneous and parenchymal tissue.⁴ A 3-mm incision is made with a scalpel, and the biopsy probe is advanced through this incision to the predetermined depth. Typically, 6 to 12 samples are obtained. Following biopsy, the device is removed, and a titanium marker is placed through the biopsy probe to denote the area of biopsy. A postprocedure mammogram is critical to confirm the targeted lesion was indeed biopsied, determine clip placement, and have a new baseline; an image of the entire breast rather than the small field of view of the stereotactic device is paramount (Fig. 155-2).

FIGURE 155-2 Images acquired stepwise through a stereotactic biopsy, including scout image, stereo pairs, and pre- and postfire images.

Ultrasound-Guided Interventions

US is an indispensable tool in the breast imager’s armamentarium. This technology not only allows for differentiation of cystic versus solid lesions but can also provide more detailed characterization of solid lesions into probably benign, indeterminate, or suspicious categories.5–14 This detailed evaluation has largely been made possible by the introduction of high-frequency 7.5- to 10-MHz linear array transducers in the mid-1990s. The ability to carefully evaluate a mass sonographically has proven extremely useful clinically and can improve the specificity of mammographically detected abnormalities. Findings at sonography can also serve to help tailor interventional procedures.

US allows for accurate diagnosis of a mammographic or palpable mass as a simple cyst, a benign lesion. Specific criteria have been developed to allow for classification of a breast lesion as a cyst. When all criteria for a simple cyst are strictly adhered to, the accuracy of US is 96% to 100%.¹⁵ These criteria are demonstration of an imperceptible wall, through transmission of sound, and lack of internal echoes (anechoic). Cysts are typically round or oval in shape and can vary in size from several millimeters to several centimeters. Once a lesion has been diagnosed as a simple cyst, intervention may occur for one of several reasons. Most commonly, the patient may be complaining of a palpable lump or pain and tenderness related to the cyst, and aspiration may alleviate these symptoms.¹⁶ However, it is important to relate to the patient that recurrence of a cyst after aspiration is very common. A second reason to aspirate a benign cyst is very large size, which may obscure a significant portion of the breast and thereby limit mammographic interpretation. In this case, the cyst is aspirated and the mammogram is repeated to allow adequate evaluation.

A cyst is typically aspirated under sonographic guidance (Fig. 155-3). The lesion is first localized, and a skin entry site is chosen. The skin is then prepped with alcohol, and intradermal and subcutaneous local anesthesia is injected. Under direct sonographic visualization, a 22- to 25-gauge needle is advanced into the lesion, and the fluid contained within the cyst is aspirated through sterile tubing into a syringe. At many facilities, the aspirated fluid may be discarded if it is clear, straw-colored, green, milky or “motor-oil” colored, since these are compatible with benign cystic fluid. If the cyst contains bloody fluid, even if the sonographic criteria of a simple cyst were met, the fluid should be sent for cytologic evaluation.

FIGURE 155-3 A, Pre– and, B, post–fine-needle aspiration sonographic images, with resolution of lesion.

Once the possibility that a mass is a simple cyst has been excluded sonographically, the possibility of a solid or complex cystic/solid lesion must be entertained. These are lesions for which biopsy is considered. Options include fine-needle aspiration (FNA), core needle biopsy (CNB), or vacuum-assisted biopsy (VAB).

Regardless of the type of needle chosen, all US-guided interventional procedures follow a similar protocol. The patient lies supine on the sonography table, with the ipsilateral arm raised up and over the head. For a lesion in the lateral portion of the breast, particularly in women with pendulous breasts, positioning the patient at an oblique angle with the shoulder on a bolster or pillow may aid in the biopsy. First the area is prepped and draped using sterile technique. The lesion is then localized sonographically. Lidocaine is used to anesthetize the skin and subcutaneous and intraparenchymal tissues. This is also done with sonographic guidance. Following this, a small (3-mm) skin incision is made 1 to 2 cm from the edge of the US probe. The distance of the skin insertion site to the lesion should include considerations of the amount of purchase needed as well as needle angle. Deeper lesions generally require more purchase, and the incision should be farther from the lesion. Additionally, a very steep needle angle hinders visualization of the needle with US. Through this incision, the biopsy device is advanced to the edge of the lesion. The needle is then either advanced into the lesion manually (e.g., with a 22- to 25-gauge needle) or advanced through the lesion via a spring-loaded component of the biopsy device. It may be optimal to include the transition of the breast parenchyma and the lesion in at least one of the samples, because this may facilitate radiologic/pathologic correlation of a mass lesion. In general, a minimum of three samples are collected, and a biopsy clip may be deployed at the site of biopsy. Care should be taken that the biopsy needle remains parallel to the chest wall throughout the procedure to eliminate the possibility of a pneumothorax or injury to the pectoralis muscle.

Biopsy Devices

Stereotactic Biopsy

Through the years, the stereotactic biopsy needle has undergone many developments. Initial stereotactic biopsy efforts were focused on FNA or biopsy with a spring loaded-device, which required removal from the breast following every sample. However, high insufficiency rates (10%-38%)17,18 plus the inability to determine invasion from ductal carcinoma in situ (DCIS) were significant limitations to FNA, and substantial undersampling, particularly of calcifications with a spring-loaded device, resulted in important modifications to this original technique. Gradually the gauge of biopsy needles increased to the now used 11-, 9-, 8-, and even 7-gauge probes. Further development followed the introduction of VAB to actively draw tissue into the biopsy chamber, as opposed to sampling that tissue which was passively in the specimen chamber. This resulted in larger tissue samples with decreased sampling error and improved accuracy of diagnosis.^19–24 Because of the potential for calcifications to be histologically heterogeneous (e.g., regions of both atypical ductal hyperplasia [ADH] and DCIS), biopsy performed with a larger-gauge needle, along with directional vacuum assistance, has proved very helpful in obtaining larger samples and decreasing underestimation rates.^25–28 Overall, in comparison to CNB, underestimation drops about 1.8-fold when using vacuum-assisted devices.²⁹ Therefore, directional VAB is now used to biopsy lesions with mammographically detected microcalcifications.

An additional advantage of the use of a VAB probe is that it does not have to be removed and reinserted. Instead, the vacuum draws the tissue into the specimen chamber, cuts it, and transports the specimen into a collection chamber at the other end of the device. The result is that multiple samples can be obtained while the needle remains in the breast. The vacuum device has been shown to produce tissue samples larger in size and heavier in weight.29–32 Acquisition of larger tissue samples significantly increases the accuracy of diagnosis of ADH and DCIS.³³ The reported rates of upgrade for ADH lesions average 44% for 14-gauge spring-loaded stereotactic core biopsy, 24% for 14-gauge directional vacuum-assisted breast biopsy, and 19% for 11-gauge vacuum-assisted breast biopsy needles (range, 0%-58%). The diagnosis of ADH following percutaneous biopsy is an indication for surgical excision to exclude the diagnosis of breast carcinoma. A diagnosis of lobular atypia at minimally invasive biopsy remains controversial, with some institutions not recommending surgical excision at all. Increasingly, the recommendations are for surgical excision of all lobular neoplasia diagnosed at minimally invasive biopsy, or the excision of pleomorphic lobular carcinoma in situ.^34,35

Ultrasound-Guided Biopsy

FNA may be performed on cystic, solid, or mixed solid lesions. A 22- to 25-gauge needle is used, and the technique is similar to the aforementioned description of cyst aspiration. Typically, several passes are made with the needle through different parts of the lesion with a back and forth motion while aspirating with a syringe (Fig. 155-4). Benefits of FNA include decreased cost of procedure and decreased likelihood of hematoma. However, as stated earlier, there are significant and well-established limitations to FNA. The success of FNA is operator dependent, with the best results achieved when performed by an experienced operator. Insufficiency rates have been found to improve with the presence of a cytopathologist during the procedure.³⁶ Another disadvantage of FNA is its inability to differentiate between DCIS and invasive cancer when malignant ductal cells are identified due to the lack of tissue architecture. Therefore, lesions chosen for biopsy with FNA should be selected based on the sonographic appearance, with less suspicious lesions favored for biopsy with this method, to lessen the likelihood of re-biopsy with a larger gauge needle to establish a diagnosis, or in clinical situations where a core biopsy is more challenging, such as when a lesion is immediately abutting an implant or large vessel. Radiologic-pathologic concordance is paramount when performing FNA.

FIGURE 155-4 Sonographic image depicting fine-needle aspiration with needle in persistent lesion.

US-guided CNB has been shown to be a reliable alternative to surgical excision.37,38 The procedure is cost-effective and well tolerated by patients and has become a mainstay in clinical practice. US-guided CNB enables accurate diagnosis of malignancy and allows for improved surgical planning for treatment. Even lesions that are considered highly suspicious for malignancy should undergo biopsy prior to definitive surgical procedure; this will alleviate the need for a second surgical procedure for nodal sampling. Estrogen and progesterone receptor status, as well as Her-2 status, can be ascertained on a CNB and are generally performed on the biopsy obtained at minimally invasive biopsy. Such assessment can assist with patient counseling and treatment options prior to surgery and can result in a less traumatic and more complete discussion of surgical options. Studies have shown that when a diagnosis of cancer is known prior to lumpectomy/partial mastectomy, the likelihood of negative margins is higher as the surgical excision is more extensive.³⁹ Additionally, it can reduce the need for repeat surgeries to evaluate and/or treat the axilla in the case of a diagnosis of invasive versus in situ disease. These procedures can be performed utilizing automated or vacuum-assisted devices.

Automated CNB devices are typically spring loaded. Once the skin has been prepped and anesthetized and the CNB device has been advanced under US guidance, it should be placed 1 to 3 cm proximal to the edge of the lesion. The device is fired into the lesion, resulting in a core biopsy specimen that maintains the architecture of the lesion, thereby allowing for a surgical-quality specimen (Fig. 155-5). Real-time documentation of the needle within the lesion is critical to confirm sampling of the targeted lesion. The “throw” of the needle varies with the manufacturer, as does the gauge of the needle and chamber size. It has been noted in the literature that one disadvantage of this US-guided CNB using a spring-loaded device is the need for multiple reinsertions of the needle through the skin incision. Typically, three to five passes are made through various aspects of the lesion, thereby sampling all quadrants of the lesion and decreasing sampling error. There has been debate in the literature as to the number of specimens needed to achieve a reliable histologic diagnosis, with some groups reporting the greatest accuracy achieved with three passes,⁴⁰ and over 90% accuracy in the first pass.⁴¹

FIGURE 155-5 A, Pre- and, B, postfire sonographic images obtained during ultrasound-guided core needle biopsy of a hypoechoic mass.

A more recent development for US-guided percutaneous breast biopsy is the availability of VAB probes. These are larger-gauge needles 9,11 and allow for procurement of larger cores under real-time imaging (Fig. 155-6). These devices obtain a larger amount of tissue to further increase accuracy of diagnosis, similar to stereotactic-guided vacuum-assisted devices. Additionally, these are typically single-insertion devices, eliminating the need for multiple reentries into the breast. These VAB probes may be useful when a lesion is sonographically visible but also contains microcalcifications, and therefore inherently contain the issues of heterogeneity and underestimation of disease associated with microcalcifications, as previously discussed. The use of sonographically guided VAB is increasingly being used to excise biopsy-proven benign masses, such as fibroadenomas, when excision is desired but surgery is not. This requires additional time, but with the acquisition of sequential samples, the sonographically visible lesion can be excised. Of course, this does not ensure complete pathologic excision.

FIGURE 155-6 A, Pre-, B, post- (aperture open), and, C, post- (aperture closed) sonographic images obtained during ultrasound-guided vacuum-assisted biopsy of an intraductal papilloma.

Most lesions biopsied under sonographic guidance are masses.⁴² If the mass is sufficiently small that there is concern that the entire lesion is removed, or if there is a possibility of neoadjuvant chemotherapy with subsequent disappearance of the lesion, a metallic marker should be placed immediately following biopsy to identify the lesion’s precise location. In our practice, all lesions that undergo biopsy have a metallic clip placed to ensure accurate localization of the lesion that underwent biopsy.

Most masses are adequately sampled with a spring-loaded device, usually 12 to 14 gauge. However, in certain clinical situations, such as with intraluminal lesions where there is the possibility of a papillary lesion, use of sonographically guided VAB may be helpful. In the specific case of papillomas, it was found in a recent study that the sensitivity and specificity at automated CNB was 28% and 100%, respectively. For vacuum-assisted CNB, both sensitivity and specificity were found to be 100%.⁴³ Therefore, use of sonographically guided VAB if the diagnosis of a papilloma is favored may be preferential.