Orthogonal Radiographic Guidance (X-Ray–Guided Needle Localization)

One guidance method uses an upright mammographic unit with a compression plate that has an open aperture with an alphanumeric grid or that contains a series of holes. To perform needle localization, the radiologist reviews the original orthogonal mammograms to identify the shortest distance to the lesion from the skin surface. The technologist places the aperture over the skin closest to the lesion, places permanent ink marks at the edges of the aperture at its contact with skin to make sure the patient has not moved, takes a single mammogram image, and leaves the breast in compression. The mammogram should show the lesion within the open aperture. The radiologist determines the coordinates of the lesion on the mammogram and, with the patient still in compression, marks this location in ink on the patient’s skin, cleans the skin, and injects a local anesthetic on the ink mark.

The radiologist then passes a needle parallel to the chest wall into and through the lesion. To ensure that the needle path is straight, the radiologist should check that the shadow from the needle hub lies directly over the needle shaft during insertion. After the radiologist passes the needle deep enough into the breast to pass through the lesion, the technologist takes a mammogram to ensure that the needle shaft projects over the lesion.

Once the radiologist confirms that the needle is through the lesion, he or she holds the needle deep in the breast; the technologist releases compression and takes an orthogonal mammogram with the needle still in place. The radiologist reviews the orthogonal mammogram and adjusts the needle depth so that the needle tip is just through the lesion. Blue dye, if used, and a hookwire are inserted.

Stereotactic Guidance

With this technique, the radiologist localizes the lesion under stereotactic guidance, as described for stereotactic core biopsy later in this chapter. The radiologist places a needle into the breast, obtains stereotactic views to make sure the needle is in the middle of the lesion, may inject blue dye, and inserts the hookwire. The patient is removed from the stereotactic device and undergoes standard orthogonal mammograms. The usual problem with stereotactic wire placements is adjusting the depth of the needle in the z-axis.

Ultrasound Guidance

Real-time hand-held ultrasound units with a small transducer provide guidance for preoperative needle localization for ultrasonographically detected breast lesions (Fig. 6-3). To do the localization, the patient is placed in the supine position and the radiologist plans the needle path to the lesion. The radiologist rolls or angles the patient on the table until the needle path is directed safely away from the chest wall to prevent pneumothorax. Using sterile technique and under direct ultrasound visualization, the radiologist anesthetizes the skin and inserts a longer needle for deep anesthesia, keeping the entire shaft of the needle, the needle tip, and the target in the same plane. The anesthesia needle can be used as a “trial run” to judge the safety of the needle path and the difficulty of needle insertion. Then the radiologist inserts the preoperative localization needle into the lesion under real-time ultrasound guidance. Blue dye, if used, and a hookwire are inserted.

Figure 6-3 Ultrasound-guided needle localization. Craniocaudal (CC) (A) and spot compression mediolateral (ML) (B) mammograms show a mass in the medial portion of the breast that was solid on ultrasound, and excisional biopsy was requested by the patient. C, Ultrasound was used to guide a needle into the mass for preoperative localization; the image shows the needle tip in the middle of the mass. After hookwire deployment, CC (D) and ML (E) mammograms show the hookwire tip in the mass. The films were marked with a white grease pen to show the skin outline and skin entry, and an X was placed on the skin over the mass marked with two BBs. F, The specimen radiograph shows inclusion of the mass and the hookwire. Histologic examination revealed a fibroadenoma.

At this point, some facilities place skin BBs before the postwire localization mammogram is obtained. A skin BB may be placed at the wire skin entry site. In addition, the radiologist may place two skin BBs and an indelible ink X over the skin where the lesion lies for the surgeon to see when the patient arrives in the operating room. The technologist then takes a mammogram with the ultrasound-placed wire within the breast.

In some facilities, radiologists or surgeons perform intraoperative ultrasound to direct the breast biopsy.

Specimen Radiography

The needle localization procedure is not over until the specimen radiograph is taken by the technologist and reviewed by the radiologist. The radiologist reports whether the specimen contains the entire lesion, how far the lesion is away from the specimen edge, if the lesion was transected, and whether the hookwire, hookwire tip, and any markers are included (Box 6-3). The radiologist then calls these findings to the surgeon in the operating room. If the lesion is not in the specimen, the radiologist directs the surgeon to the expected location by using landmarks in the excised tissue and on the mammogram and waits for a second specimen (Fig. 6-4). If subsequent specimen radiographs still do not contain the lesion, the surgeon may close the breast and obtain a mammogram to determine whether the targeted lesion is still in the breast. The mammogram is usually done a few weeks after the biopsy.

Figure 6-4 Importance of specimen radiography. The hookwire films from a freehand localization show the tip of the hookwire in microcalcifications on the craniocaudal (A) and mediolateral (B) views. C, The first specimen shows the hookwire but no calcifications. These findings were reported to the surgeon in the operating room. D, Calcifications are seen in the second specimen.

Tissue excised at ultrasound-guided preoperative localizations also undergoes specimen radiography, even if the finding cannot be seen on mammogram. The specimen radiograph may or may not show the ultrasound-localized finding, but will show if the entire hookwire or its tip, as well as any metallic markers, was excised. If the specimen radiograph does not show the ultrasound-localized finding, the radiologist can perform specimen ultrasound to see if the tissue contains the mass (Figs. 6-5 and 6-6).

Figure 6-5 Ultrasound-guided preoperative needle localization with specimen radiography. In an 18-year-old woman, ultrasound-guided core biopsy of an oval mass in the 6-o’clock position showed pathology suggestive of fibroadenoma versus phyllodes tumor. A marker was placed in the mass. The surgeon requested ultrasound-guided preoperative needle localization. A, The ultrasound shows the bright speckle inside the mass (arrow), representing the marker placed at the time of the core biopsy. This patient then underwent ultrasound-guided preoperative needle localization for excisional biopsy. B, Ultrasound shows the needle within the mass just before wire placement for ultrasound-guided preoperative needle localization. C, Ultrasound shows the wire tip (arrow) in the mass after wire placement through the needle. D, Mammogram after ultrasound-guided needle localization shows the hookwire near the marker in the mass, which is obscured by surrounding dense breast tissue. E, After excisional biopsy, the specimen radiograph shows inclusion of the hookwire, dense tissue, and the marker. Just as on the mammogram, the mass is invisible. F, Ultrasound of the breast specimen shows inclusion of the mass previously seen within the patient. The surgeon was called in the operating room to confirm that he had removed the mass, marker, hookwire, and hookwire tip. Pathology showed fibroadenoma.

Figure 6-6 Ultrasound-guided bracket needle localization. A, Positron emission tomography/computed tomography (PET/CT) shows a mass in the right breast on the CT scan (arrow). B, The mass demontrates increased uptake of fluorodeoxyglucose on the PET scan. C, Ultrasound shows a palpable lobulated, hypoechoic, suspicious mass in the 12-o’clock position on the right breast in transverse plane, corresponding to the mass seen on PET/CT. D, The mass is slightly wider in its caudal aspect. Core biopsy showed cancer. Because the mass was difficult to see against the dense breast tissue on the mammogram, the surgeon requested ultrasound bracket localization. E, Ultrasound bracket localization shows the needle at the medial aspect of the mass (arrow). F, A second wire can be seen in the lateral aspect of the mass (arrow). G, Lateral-medial mammogram after localization shows the two wires at either end of the mass (the mass is difficult to see against the dense tissue). Use of the open aperture of the alphanumeric plate avoids pushing the wires further into the breast. Annotations show clusters of three and four BBs marking the skin entry sites. Single and double BBs show the medial and lateral aspects of the mass, marked by an X written on the skin to help the surgeon find the mass in the operating room. H, Annotated magnified craniocandal view shows the wires next to the mass. Annotations show clusters of three and four BBs marking the skin entry sites. Single and double BBs show the medial and lateral aspects of the mass, marked by an X written on the skin to help the surgeon find the mass in the operating room. I, A specimen radiograph shows inclusion of the mass and the two hookwires. Note that one of the wires no longer inside the specimen was placed within the container to show that it and its tip were removed. J, The same specimen photographed at lighter contrast shows the letters and numbers on the container. The mass is located at coordinates F.0/4.0. K, An ultrasound of the breast specimen shows inclusion of the entire mass scanned at coordinates F.0/4.0, showing that the entire mass and its margins were removed.

Pathology Correlation

Later, the radiologist reviews the pathology report to see if the pathology reflects what the radiologist expected, based on the lesion’s imaging characteristics. Radiologic–pathologic correlation ensures that the targeted lesion analyzed at pathologic evaluation is concordant with the imaging finding and, specifically, that the pathology report describes a histologic finding that is known to correlate with the imaging findings. For example, if the targeted lesion shows fine pleomorphic calcifications, a diagnosis of malignancy, high-risk lesion, or benign lesion would all be concordant if the targeted calcifications were definitely seen in the specimen radiograph and preferably also on the pathology slides (Fig. 6-7). If the pathology report showed an uncalcified fibroadenoma when the targeted radiographic finding was fine pleomorphic calcifications, the pathologic–radiologic correlation would be discordant, and the case would warrant additional investigation.

Figure 6-7 A, Magnification mammogram shows invasive ductal cancer (IDC) and ductal carcinoma in situ (DCIS) as a mass and calcifications previously sampled by stereotactic biopsy. The magnified craniocaudal mammogram shows a spiculated mass, residual calcifications, air, and a metallic marker. This area was localized for surgical excisional biopsy. B, After preoperative localization and excisional biopsy, magnified, cropped specimen radiograph shows the hookwire, hookwire tip, marker, the mass, and calcifications. The uncropped specimen radiograph (not shown) showed that the entire mass and calcifications were removed. Because IDC and DCIS had been shown by stereotactic core biopsy, the radiologist expects IDC, DCIS, and calcifications in the pathology report. Pathology showed IDC, DCIS, calcifications, and postbiopsy change, which was concordant with imaging.

Pathologic–radiologic correlation of targeted calcifications is a special subset of breast biopsy correlation. Calcifications from targeted calcifications must be seen on the surgical specimen radiograph for the biopsy to be concordant. Calcifications seen just on histologic slides and not on the specimen radiograph do not represent the calcified lesion being targeted and are not concordant. Calcifications seen on the specimen radiograph are usually seen on pathology slides, but the pathologist may not see them for several reasons.

First, the calcifications may be calcium oxalate and are seen on the slides. Unlike calcium phosphate calcifications, which are easily seen on hematoxylin and eosin (H&E) staining, calcium oxalate is not visualized with H&E staining and requires a special polarized light to show the calcifications.

Second, the calcifications may be in the paraffin blocks. During specimen processing, thin breast tissue samples are embedded in paraffin blocks, which are then sliced and placed on slides for staining. Each block is several millimeters thick, but each slide contains only micromillimeters of paraffin and tissue. The calcifications may still be in the block and may never have been placed on a slide for review. A radiograph of the blocks may show the calcifications, and re-sectioning of that particular block will show the calcifications (Fig. 6-8).

Figure 6-8 Tissue specimen radiography in pathology department. A, Radiograph of six tissue specimens sectioned from an excisional breast biopsy performed for calcifications. Calcifications are found in all six specimens. B, Magnified view of two tissue specimens containing calcifications (arrows). C, Pathology technologists place the tissue pieces in paraffin in a plastic tissue cassette. This paraffin block radiograph shows the calcifications (arrows) from the two tissue specimens shown in part B. Pathology showed fibrocystic change and calcifications. If the slides from this cassette did not show calcifications, the pathologists would have taken additional samples from this cassette.

(Images courtesy of Dr. Gerald Berry, Stanford University, Palo Alto, CA.)

Third, other calcifications may be removed from the specimen if the microtome cutting device that slices the tissue/paraffin block for slides pushes large calcifications out of the specimen at the time of sectioning.

If the targeted calcifications seemed to be present in the specimen radiograph but no calcifications are found in the pathology slides or in the paraffin blocks, a repeat mammogram can determine whether the calcifications are still in the breast and were not removed at surgery. Rarely, the calcifications seen in the specimen radiograph can be incidental calcifications and not the ones that were targeted.

Needle Types

The types of biopsy needles used for specific breast lesions and guidance methods vary around the world. A trend toward progressively larger needles and more tissue samples per biopsy site has been noted, especially in the United States. Three main types of needles are used for percutaneous biopsies (Table 6-1). Fine-needle aspiration (FNA) needles, usually 25- to 20-gauge, are used for cyst aspirations and for solid breast masses. The aspirated material requires interpretation by expert cytopathologists. FNA is usually done with ultrasound or palpation guidance with at least four needle passes. FNA is less commonly done in the United States compared to Europe and Asia.

Table 6-1 Needles Used for Percutaneous Breast Biopsies

Automated large-core (core) needles in 18- to 14-gauge (Fig. 6-9A and B) commonly are used to biopsy masses with ultrasound or palpation guidance. In some facilities, especially outside the United States, core needles are used with stereotactic guidance to biopsy masses or calcifications. An automated large-core biopsy needle obtains a single specimen with each pass of the needle, and 2 to 12 specimens are obtained by firing the needle multiple times. Pathologists who are comfortable interpreting surgically excised breast biopsy tissue can interpret the histologic material obtained.

Figure 6-9 Needle types and coaxial guide for core needle biopsy. A, Schematic of core biopsy needle parts showing the needle, coaxial sheath, and inner stylet. B, Schematic of how to use a multifire core biopsy needle for breast biopsies. The outer cutting cannula shoots over the trough and cuts the mass. The entire needle is removed each time. C, Schematic of a vacuum-assisted probe for needle core biopsy. The outer cutting cannula shoots over the trough and cuts the mass. The vacuum transports the specimen to the needle end for removal. There may be one or multiple insertions, depending on the vendor.

Directional vacuum-assisted (vacuum) needles (see Fig. 6-9C) are available in 7- to 14-gauge and are used for stereotactic, ultrasound-guided, and MRI-guided biopsies. Depending on the manufacturer, vacuum biopsy can be done with just one needle pass, and multiple specimens are obtained by rotating the collection aperture of the needle to obtain between 6 and 18 specimens. Other directional vacuum-assisted needles obtain single vacuum specimens with each pass, requiring multiple insertions. In some facilities, vacuum biopsies are used to excise benign lesions such as fibroadenomas to avoid the need for surgical excision or imaging follow-up, once the fibroadenoma has been diagnosed by core needle biopsy and adequate sampling.

Both single-insertion and multi-insertion needles can be used with or without a coaxial guide (Fig. 6-10A). The coaxial guides are usually used with ultrasound or MRI guidance. The purpose of the coaxial guide is to provide a path to the target that the radiologist can use again and again without retraumatizing the breast tissue. The coaxial device consists of an inner sharp stylet and an outer sheath. The coaxial device is placed through the tissue so that the stylet tip/sheath edge is at or in the lesion. Then the radiologist removes the stylet, leaving a sheath that provides a “tunnel” through the breast tissue directly to the lesion. The radiologist then places the biopsy needle through the sheath into the lesion and takes samples. The radiologist can repeatedly place the biopsy needle through the sheath without having to disturb the surrounding breast tissue. Coaxial biopsies can be done with the sheath near the mass or through the mass (see Fig. 6-10B).

Figure 6-10 A, Schematic of how to use a coaxial sheath next to a mass for ultrasound-guided core biopsies. The radiologist places a coaxial containing a sharp, inner stylet through the breast tissue next to a mass. The radiologist removes the stylet, places a core biopsy needle through the coaxial sheath, fires the needle through the mass, and withdraws the needle. The coaxial sheath is left adjacent to the mass, providing a tunnel through the breast tissue. The radiologist removes the specimen from the needle and can replace the needle through the coaxial sheath for the next core. B, Schematic of how to use a coaxial sheath inside a mass for ultrasound-guided core biopsies. The radiologist places a coaxial containing a sharp, inner stylet into a mass. The radiologist removes the stylet, places an open core biopsy needle through the coaxial sheath, withdraws the coaxial to expose the biopsy trough inside the mass, and fires the needle to take the sample. Before withdrawing the biopsy needle, the radiologist threads the coaxial sheath over the fired needle and leaves the sheath inside the mass. The core biopsy needle can now be replaced into the sheath to take the next sample.

Cyst Aspiration

Masses on mammograms often prompt requests for breast ultrasound and cyst aspiration. To do a cyst aspiration the radiologist advances a fine needle into the cyst by palpation or image guidance. If the cyst is tense, fluid wells up into the needle hub. To aspirate the cyst, the radiologist attaches a syringe to the needle and draws fluid into the syringe until no more fluid can be obtained. Cyst aspiration can be done by ultrasound (Fig. 6-11) or, less commonly, by x-ray guidance using a fenestrated compression plate and mammography. If cyst aspiration is done under ultrasound, the radiologist should be able to watch the cyst disappear in real time.

Figure 6-11 Cyst aspiration. A, Ultrasound shows a needle in a cyst near an implant. B, The follow-up ultrasound shows that the cyst is gone.

Aspirated fluid is sent for cytologic evaluation only if an intracystic mass is present or the fluid is bloody. A large series of cyst aspirations by Tabar and colleagues showed that cyst fluid cytology is often falsely negative, even in the presence of an intracystic mass. In these cases, the pneumocystogram was enough to diagnose an intracystic mass and prompt biopsy of the rare intracystic cancer.

Pneumocystograms are mammograms obtained after the radiologist injects air into a cyst cavity. The pneumocystogram shows the air-filled cyst cavity on the mammogram, enabling the radiologist to make sure that a mass prompting biopsy on the mammogram corresponds to the aspirated cyst and to exclude an intracystic mass. Air is thought to be therapeutic in preventing cyst recurrence (Fig. 6-12). To do a pneumocystogram, the radiologist aspirates the cyst first. Once the fluid has been aspirated completely, the radiologist disengages the syringe while carefully holding the needle tip in the decompressed, flattened cyst cavity. The radiologist attaches an air-filled syringe to the needle, injects a small amount of air into the cyst cavity, takes the needle out, and obtains CC and mediolateral mammograms immediately. A normal pneumocystogram should show an air-filled, thin-walled, round or oval cavity without intracystic solid masses or mural nodules.

Figure 6-12 Pneumocystogram. A, A craniocaudal (CC) mammogram shows an oval breast mass in the inner portion of the breast. B, Ultrasound shows a complicated cyst versus a solid oval mass. After cyst aspiration under ultrasound guidance, air was placed in the cyst cavity. CC (C) and lateral (D) pneumocystogram mammograms show air (arrows) replacing fluid in the mass, which confirms that the finding on the mammogram represents a cyst that was aspirated.

Although radiologists can usually tell if a cyst on ultrasound corresponds to a specific mammographic mass, this correlation can be tricky. When the correlation is unclear and the radiologist has chosen not to do a pneumocystogram, the radiologist orders a postaspiration mammogram to see if the “cyst” disappears. The mass should be gone on the postaspiration mammogram if the aspirated cyst is the mammographic mass. If the mass still shows on the postaspiration mammogram, the mammographic finding is separate from the cyst and needs further investigation (Fig. 6-13).

Figure 6-13 Correlating mammography and ultrasound after cyst aspiration. Right mediolateral oblique (MLO) (A) and craniocaudal (CC) (B) views show a possible mass on screening mammography (circles). C, Ultrasound shows a probable multiseptated, complicated cyst in the expected location of the mammographic mass. D, Fine-needle aspiration under ultrasound shows the needle in the mass. E, Cyst fluid was aspirated and the mass disappeared. Postaspiration MLO (F) and CC (G) views show that the mammographic mass disappeared, indicating that the ultrasound and mammographic findings represented the same simple cyst.

Palpation Guidance

FNA or core biopsy can be performed on palpable masses. With this method, CC and mediolateral mammograms and other imaging studies are reviewed. Palpation-guided procedures are similar to ultrasound-guided procedures, but there is no visualization of the lesion or needle during the procedure. The lesion must be discretely palpable and well away from the chest wall for the biopsy to be done with accuracy and safety. This procedure is usually reserved for cysts and solid masses that are almost definitely malignant or benign by imaging and palpation criteria.

Ultrasound Guidance

When compared with stereotactic biopsy, ultrasound-guided biopsy has the advantage of using readily available equipment and is fast and cost-effective. The first step in ultrasound-guided biopsy is to find the questioned lesion for biopsy. This commonly occurs when a mass on the mammogram prompts an ultrasound to further characterize the mass and localize it for biopsy. When correlating the mammogram to the ultrasound, the mass can be far away from the chest wall on the mammogram and lie next to the pectoralis muscle on the ultrasound. This occurs because the breast tissue is compressed far away from the chest wall when the patient stands up for the mammogram. On ultrasound, the breast falls dependently onto the chest wall when the patient lies down (Fig. 6-14A).

Figure 6-14 Schematics of how masses appearing to be far away from the chest wall on mammography may be found near the chest wall on ultrasound and how to keep the needle tip away from the chest wall on ultrasound-guided core biopsy A, Schematic mammograms show a mass in the upper inner quadrant that appears to be far from the chest wall. However, mammograms are obtained with compression; when the patient lies supine, the mass falls dependently against the chest wall. On ultrasound, the mass may be closer to the chest wall than expected from the mammogram. B, With superficial lesions, the needle tip and throw are usually far away from the chest wall. C, With deeper lesions, the needle angle is steeper, and the radiologist judges whether the needle “throw” will penetrate the lung. D, When the patient is flat and the lesion is even deeper, the needle trajectory can point toward the chest wall. E, To change the needle trajectory, the patient can be angled so that her chest wall parallels the needle track/throw. F, Injecting anesthetic underneath the lesion can lift it away from the chest wall for biopsy. G, Inserting the needle tip into the lesion and redirecting the throw of the needle avoids placing the needle tip into the chest wall.

(B to G, Courtesy of Dr. Sunita Pal, Stanford Radiology, Stanford, CA.)

When planning ultrasound-guided needle biopsies, it is important to keep the needle tip away from the chest wall to prevent pneumothorax. Unlike upright preoperative x-ray–guided needle localization or prone stereotactic localization, the ultrasound-guided biopsy is done supine and the needle is not necessarily parallel to the chest wall. Further complicating matters, some core biopsy needles “throw” the cutting trough 2.5 cm further into the tissue beyond the needle tip. Thus, planning a safe ultrasound-guided needle biopsy trajectory must take into account both the needle tip and the needle “throw” trajectory. To plan a safe procedure, the radiologist rolls the patient on the table so that the needle trajectory is as parallel to the chest wall as possible and not at a steep angle aiming toward the lungs. Patient positioning can take some time, but it is worth the few minutes to position the patient accurately to avoid an untoward complication. Another way to keep the needle away from the chest wall is to inject anesthetic underneath the targeted mass to lift it away from the pectoralis muscle. Alternatively, in some cases, the radiologist can stick the biopsy needle tip into the mass and lift it into a safer trajectory before firing the needle (see Fig. 6-14B to G).

For ultrasound-guided FNA, the radiologist introduces a needle in the plane of the transducer axis to show the entire shaft of the needle, its tip, and the lesion. Once the needle is within the lesion, the radiologist aspirates the mass with a vigorous to-and-fro movement to obtain material for cytologic evaluation and then withdraws the needle. At least four passes should be performed; optimally, the material should be analyzed immediately to ensure that adequate cellular material has been obtained for diagnosis. After aspiration, direct pressure is applied to the site (Fig. 6-15).

Figure 6-15 Fine-needle aspiration of axillary lymphadenopathy in a patient with invasive ductal cancer. A, Initial ultrasound shows a large invasive ductal cancer in the upper left breast undergoing ultrasound-guided core biopsy. B, Ultrasound of the left axilla in this patient shows a lymph node with an enlarged, thickened cortex, suspicious for metastatic disease. C, Fine-needle aspiration under ultrasound guidance shows the needle tip in the lymph node cortex (arrow). Cytology showed lymph node metastases. D, In another patient, the right axillary lymph node has an irregular, thickened cortex and a compressed fatty hilum. E, Ultrasound shows the fine-needle tip in the lymph node. Cytology showed metastatic disease from invasive ductal cancer. F, In another patient, a lymph node has an irregular cortex. Fine-needle aspiration showed metastatic disease. G, To make sure this irregular lymph node was removed at surgery, the surgeon requested preoperative needle localization. The ultrasound shows a hookwire traversing the lymph node at ultrasound-guided needle localization. Pathology showed metastatic disease.

To perform a core biopsy under ultrasound guidance, the radiologist localizes the lesion by ultrasound and chooses the course of needle insertion that offers the most accuracy and safety. While anesthetizing the core biopsy track under direct ultrasound guidance, the radiologist uses the anesthesia needle to get an idea of how dense the breast feels and to see the needle trajectory. The radiologist also calculates the core needle “throw” to determine where to place the core needle tip “pre-fire” so the core trough will be in the middle of the lesion “postfire.”

Then, under direct ultrasound visualization, the radiologist introduces the core biopsy needle into the breast. If the lesion is large enough, the radiologist introduces the needle into the edge of the lesion to hold it in place. Otherwise, the radiologist may choose to fire the needle through the mass with or without a coaxial system. In any case, the radiologist fires the biopsy core needle under direct visualization and harvests the cores. Optimally, at least three to five tissue specimens are obtained from different parts of the mass. After sampling, the radiologist places a metallic marker into the mass, and the technologist holds direct pressure on the breast to establish hemostasis. After hemostasis is established, the technologist bandages the wound and takes orthogonal mammograms to show the marker and any residual mass (Fig. 6-16).

Figure 6-16 Use of the coaxial to core a mass adjacent to an implant. A, Ultrasound shows an oval mass on top of a breast implant, for which core biopsy was recommended. B, Ultrasound-guided core biopsy shows the trough of the needle (double arrows) traversing an oval mass (single arrow) on top of an implant (triple arrows). C, Postfire ultrasound shows the needle traversing the mass. At this point, a coaxial was placed over the needle, securing the needle’s original position within the mass. D, The needle was replaced in the coaxial sheath and the coaxial (triple arrows) was withdrawn to expose the trough (double arrows) and allow the mass (single arrow) to fall within it. E, Postfire ultrasound with the coaxial sheath extended over the needle shows both the needle and coaxial traversing the mass. F, Ultrasound showing a needle threaded through the coaxial sheath (double arrows) placing a marker (single arrow) in the mass (triple arrows). G, Cropped implant-displaced digital lateral mammogram shows dense tissue and the marker. Histology showed fibroadenoma.

A vacuum biopsy is similar to an automated multifire core biopsy, but the vacuum needle is usually placed under or occasionally inside the lesion. The probe “vacuums” tissue into the trough to be sampled. The vacuum technique carries a special caveat regarding the skin. If the probe is too close to the skin, the skin can be “vacuumed” into the trough and sampled, causing skin injury, requiring a suture or, in extreme cases, a skin graft. During a needle biopsy using vacuum technique, the radiologist obtains several samples, concentrating on aiming the trough at the mass (Fig. 6-17). Afterward, the radiologist can place a marker in the mass either through the probe or (depending on the manufacturer) by using a marker that has its own separate needle.

Figure 6-17 Vacuum-assisted core biopsy under ultrasound. A young patient has a new palpable mass in the left breast; biopsy was recommended. Transverse (A) and longitudinal (B) ultrasounds show an oval solid mass near the chest wall. C, To perform the ultrasound-guided vacuum-assisted biopsy, the radiologist injected lidocaine between the mass and the chest wall to lift it away from the pectoralis muscle. The radiologist then inserted the vacuum-assisted probe, shown here traversing the mass. Note the amount of mass (arrow) above the probe (double arrow). D, After vacuum-assisted sampling, in which the vacuum draws the mass into the probe and cuts off the tissue in the trough, there is less mass (arrow) above the probe after tissue removal. E, After biopsy, the ultrasound shows the needle placing a marker (arrow) within the mass. F, Magnified digital mammogram shows the marker within dense tissue, and the mass is not seen. Histology showed fibroadenoma.

If a mass previously cored under ultrasound guidance must be removed, the radiologist localizes the mass or marker under ultrasound, places a wire, then takes orthogonal mammograms to show the wire and marker, and waits for the excised tissue specimen (Fig. 6-18).

Figure 6-18 Ultrasound-guided core biopsy with marker placement and subsequent needle localization. Mediolateral (A) and craniocaudal (CC) (B) mammograms show a spiculated mass in the midportion of the left breast at the 9-o’clock position. C, Ultrasound shows a hypoechoic, spiculated shadowing mass. D, Ultrasound-guided needle placement shows the tip of the core biopsy needle before biopsy just proximal to the mass. Note that the trajectory of the needle is well away from the chest wall and will not produce a pneumothorax. E, A postfire ultrasound scan shows the needle traversing the mass and the expected location of the sampling trough inside the mass. All postfire films are imaged. F, A marker is placed in the mass and is displayed as a bright echo in the middle of the mass. Follow-up CC (G) and mediolateral (H) views show the marker in the mass. A single marker shows the skin entry site for the core needle, and two BBs show the location of the mass on the skin. I, Subsequent ultrasound-guided needle localization shows a wire in the mass. CC (J) and lateral (K) mammograms taken after ultrasound-guided needle localization show the wire tip in the mass. L, The specimen radiograph shows the mass, hookwire tip, and the marker. Pathologic examination revealed invasive ductal cancer and lobular carcinoma in situ; one of three sentinel lymph nodes was positive.

Stereotactic Guidance

This method uses a compression device with a small aperture and an x-ray tube that has the ability to take two stereotactic views about 15 degrees off perpendicular (Fig. 6-19A to I). The patient is in a prone, upright, or decubitus position with the breast compressed by a fenestrated compression paddle for stereotactic needle biopsy. The radiologist reviews prebiopsy CC and mediolateral mammograms to determine the lesion’s location on orthogonal views. The breast is then firmly compressed with the compression paddle aperture placed on the skin surface closest to the breast lesion. After taking a straight-on scout view that visualizes the lesion, the stereotactic technologist takes two stereo views of the lesion. The radiologist locates the lesion on the stereo views and passes a needle into the breast to a calculated depth. Prefire stereotactic images, which should show the tip of the biopsy needle at the edge of the lesion, are obtained. The radiologist then fires the needle deeper into the breast and reviews postfire stereotactic images to ensure that the trough of the needle is within the breast lesion. After the radiologist collects multiple specimens, he or she reviews the core specimen radiograph to make sure that the calcifications or mass has been sampled. The tissue specimens are labeled and sent to the pathology laboratory. At this point the radiologist decides whether to deploy a metallic marker into the biopsy cavity. If the radiologist deploys a marker, the technologist takes additional stereotactic images to confirm marker deployment before releasing the patient from compression. The technologist maintains direct pressure on the biopsy site after release of the compression paddle to achieve hemostasis, places a bandage, and obtains immediate postbiopsy upright CC and mediolateral mammograms. These show the biopsy cavity, confirm removal of all or a portion of the calcifications or mass, and show the location of the marker and its position relative to the targeted findings.

Figure 6-19 Eleven-gauge stereotactic vacuum-assisted core biopsy and marker placement. A, A mammogram shows suspicious branching microcalcifications. B, A stereotactic straight-on scout view shows the suspicious microcalcifications in digital format. C, A 15-degree stereotactic view shows that the calcifications are within the aperture of the compression plate. D, Prefire films demonstrate that the stereotactic needle is directed toward the calcifications. E, Postfire stereotactic views show that the needle is traversing the suspicious microcalcifications. F, A specimen radiograph shows inclusion of the suspicious microcalcifications corresponding to the mammographic finding. G, A stereotactic view at marker placement demonstrates that the marker is near the biopsy cavity on the stereotactic view. Immediate postbiopsy craniocaudal (CC) (H) and lateral (I) mammograms show that the marker is near the biopsy site and that air is present in the biopsy site. J, Cropped digital magnification CC mammogram shows fine pleomorphic calcifications marked for stereotactic needle biopsy. K, Cropped magnified digital specimen radiograph shows inclusion of the suspicious calcifications. Pathology showed invasive ductal cancer and DCIS with calcifications. L to R, Ultrasound-guided core biopsy of calcifications with core specimen radiography. L, CC mammogram shows calcifications in the outer left breast. Note the radiologist has annotated the mammogram with instructions on how to manage this patient. Ultrasound later showed a mass and calcifications in this location. Transverse (M) and longitudinal (N) ultrasounds show a 1.1-cm hypoechoic mass containing calcifications. O, The ultrasound shows the vacuum-assisted core needle trough below the mass and calcifications. P, Nine samples were obtained by vacuum-assisted core biopsy, and the ultrasound shows less of the mass and calcifications above the trough. Q, Core specimen radiographs of samples show the calcifications first seen on the mammogram. R, Postbiopsy mammogram shows the marker on the outer breast, absence of the calcifications, and air near the chest wall from the biopsy. Biopsy showed DCIS.

Burbank in 1996 and Jackman and Marzoni in 2003 discussed various techniques used to successfully biopsy lesions in technically challenging situations.

Core Specimen Radiography

Magnification specimen radiography is mandatory after stereotactic biopsy of calcifications and optional after biopsy of a mass to ensure that the lesion that prompted biopsy has been adequately sampled or removed (see Fig. 6-19J and K). If the targeted lesion is not present in the specimen, the radiologist obtains more specimens. Specimen radiography is not usually done after ultrasound-guided core biopsy unless the biopsy is targeting radiographically detectable calcifications (see Fig. 6-19L to R).

At times, specimen radiography may be equivocal in determining if the specimens include the lesion prompting biopsy, especially for masses. In such cases, the radiologist compares the breast specimen pathology report with the specimen radiograph. If there is a discrepancy between the appearance of the mammographic finding and the pathology report, the radiologist reviews the patient’s immediate postbiopsy mammograms to see whether the lesion that prompted biopsy has been sampled or removed.

Calcifications from a biopsied calcific cluster must be evident on the core specimen radiograph for the imaging and histologic findings to be concordant. Most, but not all, calcifications seen on a specimen radiograph are also seen on histologic slides. If the radiologist was targeting calcifications and no calcifications are seen on the specimen radiograph, the ensuing pathology report will not be representative of the calcifications prompting biopsy (which are probably still inside the breast). However, the pathology report may still report calcifications even if they are absent on the core specimen radiograph because pathologists can see tiny calcifications on breast specimen slides that cannot be seen by core specimen radiography. These calcifications are usually smaller than 100 microns and are seen on the slides by serendipity. Thus, patients who undergo biopsy of calcifications and have specimen radiographs that show no calcifications need rebiopsy, even if the pathology report describes calcifications.

Marker Placement, Movement, and Compatibility with Ultrasound and MRI

Immediate postbiopsy insertion of a metallic marker at the biopsy site is usually indicated in core or vacuum needle biopsies guided by stereotactic technique. If a residual mass or calcifications is still left on postbiopsy stereotactic images after adequate sampling, a marker may not be needed. The marker is needed in case a cancer, high-risk lesion, or discordant benign lesion is diagnosed and the patient needs subsequent surgical excision of the needle biopsy site. With ultrasound- and MRI-guided biopsies, a metallic marker is often placed after biopsy even if some of the biopsied lesion is still evident. The marker helps to correlate the ultrasound, MRI, and mammogram findings.

After stereotactic biopsy and marker placement, the radiologist reviews stereotactic images before releasing breast compression to be sure the marker has deployed and is at or near the biopsy site. As a rule of thumb, our facility deploys a second marker if the first is more than 7 mm in depth (either deeper or more superficial) away from the initial biopsy site. For ultrasound-guided biopsies, ultrasound-visible markers are used to be sure the marker has deployed accurately.

However, there is no practical way after MRI-guided biopsy to be sure the marker has deployed accurately by using MRI only. Even though an MRI scan done immediately after marker placement can show the signal void from the metallic marker, air introduced by the biopsy can also cause a signal void that simulates metallic markers. A post-MRI procedure mammogram will determine if the marker was deployed but not if it is accurately placed.

For stereotactic core biopsies, upright orthogonal CC and lateral mammograms obtained immediately after stereotactic biopsy show the location of the marker in relation to the biopsy site. Usually, the marker is located in or near the biopsy site. If the marker is some distance away from the site (i.e., inaccurate initial deployment), there is no practical way to insert a second marker. If the patient with an inaccurately placed marker needs surgical excision of the biopsy site, it may help to proceed with surgery as quickly as possible, hoping that a postbiopsy hematoma can be visualized by ultrasound or mammography to help guide the needle localization.

For ultrasound- or MRI-guided biopsies, upright orthogonal mammograms are obtained immediately after biopsy to correlate with prebiopsy mammogram images and to see the metallic marker’s location in the breast.

If biopsy is performed on two sites in the same breast, markers with two unique shapes can be used to differentiate the two different sites. Rarely, patients request removal of a biopsy marker, which can be performed percutaneously with a vacuum-assisted biopsy device.

Various types of markers are available, including those containing stainless steel or titanium alone and those with metal embedded in plugs of various types. Terms for these markers include both clips and markers. The initial marking devices, which truly clipped to the edge of the biopsy cavity, are correctly said to be clips. Later, marking devices were developed to fall into the cavity without clipping to tissue; these are more correctly called markers. In scientific literature, the terms markers or clip/markers refer to both types of devices but are used inconsistently. The marker plugs are composed of Gelfoam, bovine or porcine collagen, suture-type material, or other materials. If markers containing bovine or porcine collagen are used, the patient should be asked about allergies to either beef or pork before deploying the markers.

A variety of problems are associated with the markers (Box 6-4). The first potential problem, nondeployment, is uncommon. It was anecdotally noted that some plugs may get stuck during deployment, making it impossible to push the plunger in or difficult to withdraw or close the needle. This problem is possibly due to the plugs filling with fluid, expanding in the deployment device, and getting stuck in the trough. For vacuum-assisted needles, the deployment device can get stuck on a retained fragment in the vacuum needle.

The second and most common potential problem is inaccurate initial deployment of the marker. An even rarer problem is delayed migration of the marker (i.e., it moves from the initial site to a different site in the breast). This can occur whether the initial deployment was accurate or inaccurate. Both inaccurate initial deployment and delayed migration are primarily along the axis of needle biopsy insertion (i.e., the z-axis).

Because there is no way to predict delayed migration, we advise upright orthogonal mammogram views immediately before x-ray–guided needle localization. The radiologist compares the marker on those images with the lesion position on prebiopsy mammograms and the marker (and postbiopsy changes) on the immediate poststereotactic biopsy mammograms. This will determine if the marker has moved.

If the marker was inaccurately deployed or has later migrated away from an original biopsy site, the radiologist determines the location of the original targeted lesion by using breast architecture and landmarks. The goal of subsequent localization is to remove the targeted biopsy site, including any residual cancer (Fig. 6-20). Whether it is necessary to also localize and remove the inaccurately positioned marker is controversial, but it should be considered if the needle biopsy revealed cancer. If the needle biopsy revealed a high-risk lesion or a discordant benign lesion, the inaccurately positioned marker presumably does not need to be removed. When the marker is in an inaccurate position, some facilities use presurgical or intraoperative ultrasound to try to identify the needle biopsy site.

Figure 6-20 Value of two different types of markers for two stereotactic biopsy sites and marker migration. Because this patient had two suspicious microcalcification clusters, two different markers were placed at stereotactic biopsy. Craniocaudal (CC) (A) and mediolateral (B) mammograms show the difference in marker configuration, which clearly identifies the biopsy site. If the same type of marker had been placed in both sites, it would be difficult to determine the location of each biopsy site on subsequent mammograms. Cancer was found in the upper biopsy site, and the lower biopsy site was benign. Two months later, preoperative CC (C) and lateral (D) views show that the upper marker has migrated to the inferior portion of the breast. Because cancer was taken from the upper biopsy site, breast architecture was used to target the original upper biopsy site for excision because the marker is now in a different quadrant. This case shows the importance of correlation of the prestereotactic biopsy scout mammogram, the immediate postmarker placement mammogram, which documents orientation of the marker to the biopsy site, and scout mammograms before preoperative needle localization.

Some markers placed by stereotaxis are embedded in plugs visible by ultrasound. Facilities may use ultrasound to localize these plugs for subsequent needle localization after stereotactic biopsy. After an ultrasound-guided biopsy, facilities may use markers visible by ultrasound. This allows the physician placing the marker to see whether the marker has deployed.

MRI is increasingly being used to stage the breast for cancer and to plan surgical management. Because biopsy site markers placed by stereotaxis, ultrasound, or MRI may be imaged by subsequent MRI studies, understanding of marker MRI compatibility and safety, and of marker artifacts, is becoming increasingly important. Accordingly, all facilities should use MRI-compatible metallic markers when markers are placed with any modality because MRI might be performed later. Metallic markers cause a signal void on MRI, and the size of the signal void varies according to the marker type and pulse sequence (Fig. 6-21). There is a difference between MRI marker compatibility and safety. MRI compatibility means that the marker can be used in the MRI magnet and will cause little artifact. Marker safety means that the marker will produce no harm to the patient in the magnet. Some markers are MRI compatible but still cause large artifacts of up to 2 cm, thus rendering the MRI less readable than when using other markers. MRI testing of markers for artifact by using phantoms on the facility’s pulse sequences is a simple way to determine the marker artifact and the size of the signal void. This should be done before inserting metallic markers for marking tumors or biopsy sites. However, correlation between MRI, ultrasound, and mammography can be challenging, when markers or even wires are placed for preoperative localization (Fig. 6-22).

Figure 6-21 Signal void on breast magnetic resonance imaging (MRI) due to metallic markers placed from stereotactic core biopsy. A, Postcontrast 3-D spectral-spatial excitation magnetization transfer (3DSSMT) breast MRI shows signal voids as dark areas (arrows) in a patient who had four stereotactic core biopsies with metallic markers placed. B and C, Postcontrast 3DSSMT breast MRI slices showing signal voids (arrows) in the same patient from the other metallic markers.

Figure 6-22 Correlation of mammogram, ultrasound, and magnetic resonance imaging (MRI) on wire localization. A, Right mediolateral oblique (MLO) digital mammogram shows architectural distortion in the upper right breast (arrows) and a skin marker on the distortion because there was a palpable mass. B, Slightly lighter technique MLO shows the distortion to greater advantage. C, Craniocaudal (CC) mammogram shows the architectural distortion after core biopsy with a marker in it (arrows). D, Ultrasound in the upper outer quadrant shows an irregular, hypoechoic, shadowing mass corresponding to the mammographic finding. Core biopsy showed invasive lobular cancer. E, Radial scan of the invasive lobular cancer on ultrasound. F, Two other satellite masses representing invasive lobular cancer in the same quadrant on ultrasound. G, Contrast-enhanced 3-D spectral-spatial excitation magnetization transfer (3DSSMT) MRI shows an irregular spiculated mass (arrows) representing the invasive lobular cancer seen on the MLO mammogram. The kinetic curve showed a rapid wash-in and late washout. H, Postcontrast 3DSSMT axial reformat shows the irregular enhancing cancer in the outer breast, similar to its appearance on the CC mammogram (arrows). I, Sagittal 3DSSMT postcontrast MRI shows two unexpected incidental spiculated enhancing lesions (IELs) in the midbreast on another slice, far away from the upper outer quadrant invasive lobular cancer (arrows). These were not evident on the mammogram, nor found with certainty by ultrasound. J, Postcontrast 3DSSMT axial reformat MRI shows one of the two spiculated IEL masses suspicious for cancer (arrow). K, Axial noncontrast T1-weighted MRI on the day of the MRI-guided needle localization shows the low signal architectural distortion representing invasive lobular cancer (triple arrows), the unenhanced IELs (single arrow) and the fiducial on the skin marking the planned needle entry site (double arrows). L, Postcontrast Dixon technique axial MRI shows the enhancing palpable spiculated invasive lobular cancer (triple arrows), the signal void from the needle traversing the satellite lesion (double arrows), and the enhancing IEL (arrow). M, Specimen radiograph shows inclusion of the hookwire and hookwire tip in the possible dense mass and architectural distortion in the dense tissue with the marker within it. Pathology showed invasive lobular cancer near the marker and at the hookwire tip.

Carbon Marking

This method uses activated charcoal USP (Mallinckrodt, Phillipsburg, NJ) sterilized and suspended as a 4% weight/weight aqueous suspension. It is mixed with 0.3 mL sterile saline or water and injected in a to-and-fro motion after core biopsy along the stereotactic or ultrasound needle track to yield a dark line of carbon particles in the breast. This line of carbon particles can be used as a guide for excisional biopsy days to weeks after the percutaneous biopsy. It is used as an alternative to a postbiopsy metallic marker, mainly in Europe.

Patient Safety and Comfort after Biopsy

After needle biopsy, hemostasis is achieved by direct pressure. In some institutions, after hemostasis is confirmed, the patient is taught how to “hold pressure” on the biopsy site so that she knows what to do if subsequent untoward bleeding occurs. After adequate hemostasis is achieved, some institutions close the wound with Steri-Strips and cover the Steri-Strips with Opsite, a self-adhesive polyurethane film sterile material used to cover operative wounds. Opsite prevents the Steri-Strips from getting wet and keeps the wound clean and dry. The patient can take a shower with the Opsite on but is instructed not to “scrub” the Opsite, take a bath, swim, or engage in other activities that might immerse the wound site. The patient is told to expect a quarter-sized spot of blood on the Steri-Strips and a bruise at the biopsy site that may travel to dependent sites; the patient is told to put direct pressure on the biopsy site if oozing or unexpected bleeding occurs. The patient is instructed to remove the Opsite and Steri-Strips after 4 to 7 days. Some facilities also bind the breast with wraparound bandages or commercially available binders (commonly used after mastectomy) to hold the breasts tight to the chest wall. These bandages restrict breast motion while the patient is awake and limit breast motion during the night. Patients are also told that the biopsy site may feel bigger because of a blood clot in the biopsy site.

Part of recovery is pain control. If the patient is not allergic to acetaminophen and has no liver problems, she may take acetaminophen initially and then every 6 hours as needed, up to 4 g/day. Rarely, stronger medication such as Tylenol No. 3 (acetaminophen with codeine) or Vicodin (acetaminophen with hydrocodone), may be prescribed for pain. Any pain medications (usually aspirin or nonsteroidal anti-inflammatory drugs [NSAIDs]) withheld for 7 days before biopsy are to be avoided for 3 days after biopsy to decrease the risk of bleeding. After the technologist puts an ice pack on the biopsy site, the patient is told to leave the ice on for 60 minutes initially and then for 10 minutes every hour until bedtime. She is advised not to keep it on longer because of the possibility of frostbite. The ice helps decrease postbiopsy discomfort and bleeding. To keep the ice pack in place, the patient may put the ice pack inside her brassiere or use a commercially available breast binder. Afterward, patients are given verbal and written postbiopsy wound care instructions and a phone number to call for problems. Each time the 10-minute ice pack is removed, the patient is told to look at the bandage (which may require using a mirror). If the amount of blood has increased since the last inspection, the patient is told to again firmly compress the biopsy site for 10 more minutes. Patients are instructed on where and how to obtain their biopsy result. Most patients do well with these instructions. Some facilities call the patient later in the afternoon, early evening, or the next day as a courtesy call to see how the patient is doing and to answer any questions.

Complete Lesion Removal

Core biopsy samples showing cancer routinely undergo subsequent excisional biopsy to remove residual cancer at the biopsy location. Even if the entire radiographic finding has been removed by percutaneous biopsy, occult residual cancer has been demonstrated in the needle biopsy site in 73% (15/23) of cases at subsequent surgery.

How is the core biopsy site localized when the entire lesion has been removed and no marker was placed in the biopsy site, or a marker was placed but is not at the biopsy site (either from inaccurate initial placement or delayed migration)? Because air can move along the biopsy track, air alone is not enough to identify the biopsy site. Although a hematoma may sometimes be mammographically seen in the immediate postbiopsy period, it usually persists for just a few days or weeks. Studies have shown no mammographically detectable findings some months after a stereotactic core or vacuum needle biopsy.

Brenner suggests using landmarks within the breast to localize the approximate stereotactic biopsy site. Others have suggested localizing the fluid collection in the biopsy site by ultrasound. Although this may work in the immediate postbiopsy period, fluid can be reabsorbed if the excisional biopsy is delayed by several weeks, and using the fluid alone may not be reliable.

Calcification and Epithelial Displacement

Percutaneous biopsy needles rarely displace calcifications to locations distant from the original biopsy site. It is controversial whether displaced metallic markers or displaced calcifications far from the biopsy site require surgical excision when the percutaneous biopsy specimen shows cancer.

Epithelial displacement of breast cancer cells into benign tissue along the needle track may occur with any gauge biopsy needle. These displaced tumor cells may simulate breast cancer invasion or a second focus of tumor unless the pathologist knows there was a prior core biopsy. Pathologists should be informed that a needle biopsy has been performed so that they do not mistakenly diagnose displaced epithelium as invasive cancer in a ductal carcinoma in situ (DCIS) lesion or erroneously stage a tumor as multifocal when only one cancerous site is present. Epithelial displacement is more common with FNA or core biopsy (where the needle is removed from the breast with each needle pass) than with vacuum biopsy (where the external part of the needle stays in the breast until the rotational biopsy is finished). To our knowledge, epithelial displacement has not been evaluated with the type of vacuum biopsy needle that is inserted several times, but we would assume that displacement would be similar to that seen with a core biopsy needle that is inserted several times. The displaced epithelial cancer cells are thought to rarely, if ever, acquire a blood supply and grow as metastatic disease along the needle track, in draining lymph nodes, or systemically elsewhere in the body.

Pathology Correlation

Routine correlation between the mammographic appearance of the breast lesion and the pathology report is an essential part of quality assurance to reduce the number of false-negative results and to excise cancers. Biopsies that are initially benign (exclusive of high-risk lesions) and later proven to be carcinoma at that same site are called false-negative biopsies. Liberman and colleagues and Jackman and colleagues have separately defined the false-negative rate as all false-negative lesions in a study divided by all breast cancers found at any time in the study. Immediate false-negatives occur if an initial benign biopsy is discordant and immediate repeat biopsy reveals cancer. Delayed false-negatives occur if the benign biopsy is considered concordant and lesion growth at follow-up imaging leads to later rebiopsy and discovery of the missed cancer.

False-negative biopsies occur with any kind of biopsy. They are most thoroughly reported for stereotactic biopsies done in prone position with automated large-core multiple-insertion needles and with single-insertion directional vacuum-assisted needles. False-negatives have also been thoroughly studied for image-guided diagnostic surgical biopsies. In separate literature reviews, Jackman and colleagues found the mean false-negative rates to be 1% for stereotactic 11-gauge vacuum-assisted needle biopsy, 4% for stereotactic 14-gauge automated large-core needle biopsy, and 2% for image-guided needle-localized diagnostic surgical excision.

False-negative rates are less well-defined for other image-guided percutaneous biopsies, but are approximately 2% to 2.5% for ultrasound-guided needle biopsies (with both vacuum-assisted and automated large-core needles).

Liberman and colleagues emphasized in 2000 that benign histologic diagnoses that do not explain the imaging findings must be considered discordant and should lead to repeat biopsy done by surgical excision or a more aggressive needle core biopsy (Box 6-5). American College of Radiology Breast Imaging Reporting and Data System (BI-RADS®) category 5 lesions with a benign histologic diagnosis are discordant. Specific benign diagnoses for mass lesions include fibroadenomas, lymph nodes, and benign cysts. There are no specific diagnoses for nonmass lesions whether the nonmass lesions are calcified or uncalcified. With nonspecific benign diagnoses, one relies on the quality of the needle biopsy to determine if the diagnosis is concordant. As discussed, biopsy of calcifications must include adequate sampling or removal of the calcifications as determined on specimen radiographs of the core or vacuum biopsy samples (looking for calcifications) and postbiopsy stereotactic or upright images (looking for reduction or absence of calcifications seen before biopsy compared to after biopsy).

Discordance is hardest to determine with BI-RADS® category 4 noncalcified mass and nonmass lesions. When needle pathology is benign on stereotactic, ultrasound, or MRI-guided needle biopsy, a subjective combination of confidence with the accuracy of the needle biopsy and decreased size of the lesion after biopsy are used to make that decision.

Surgical excision is advised for all DCIS lesions, the same as for invasive cancer. Women with invasive cancer also undergo axillary node sampling or dissection at initial therapeutic surgery, but women with DCIS usually do not undergo axillary node biopsy. A moderate percentage of patients with DCIS diagnosed at needle biopsy may later need axillary node dissection because invasive cancer is diagnosed at subsequent excisional biopsy. Biopsy specimens showing DCIS at initial biopsy and invasive cancer at subsequent surgical excision (called DCIS underestimates) occurred at prone stereotactic biopsy in 20% of DCIS lesions diagnosed with 14-gauge core biopsy and in 11% of DCIS lesions diagnosed with 11-gauge vacuum biopsy.

DCIS underestimates with ultrasound-guided biopsies have been less thoroughly studied, but those underestimates are also decreased with 11-gauge vacuum biopsy compared to 14-gauge core biopsy. DCIS underestimates also occur with diagnostic surgical biopsies, but we do not know the overall rate. When just those with positive or close histologic margins for DCIS have a repeat operation (meaning there is a large selection bias) the surgical DCIS underestimate rate is 11%.

High-Risk Lesions, Including Controversies

There are no universally accepted criteria to decide which lesions diagnosed at core or vacuum needle biopsy are high-risk and should undergo surgical excision to determine the presence of an associated cancer underestimated by the biopsy. Sickles defines a “probably benign” BI-RADS® category 3 mammographic lesion as a finding with a less than 2% chance of malignancy that can be followed with imaging. In 2002, Jackman and colleagues suggested that the less than 2% chance of malignancy could be used to decide if specific histologic lesions diagnosed at needle biopsy could be followed or needed excision. This was initially applied to atypical ductal hyperplasia (ADH) lesions and has subsequently been accepted by many as a practical way to decide what to do with both ADH and non-ADH lesions. Those with a greater than 2% chance of malignancy at follow-up would be considered high-risk and would be sent for surgical excision. Those with a less than 2% chance of malignancy at follow-up would be considered benign and would be carefully followed with imaging.

Lesions that can underestimate the associated presence of cancer are called high-risk lesions, and they occur in roughly 10% of percutaneous biopsies. Half of the high-risk lesions are ADH. Biopsies showing ADH at initial biopsy and cancer at subsequent surgical excision (termed ADH underestimates) occurred at prone stereotactic biopsy in 44% of ADH lesions diagnosed with 14-gauge large-core needle biopsy and 19% of ADH lesions diagnosed with 11-gauge vacuum-assisted biopsy. Although pathologists have trouble distinguishing between ADH and low-grade DCIS, the ADH underestimates are a more significant problem. The cancers found with the ADH underestimates were invasive carcinoma in 25% of cases and any grade DCIS in the other 75%. Jackman and colleagues showed that there are no patient, lesion, or biopsy risk factors that might obviate excision after stereotactic core biopsy showing ADH.

More controversy exists about the need for surgical excision of non-ADH lesions. Most authors in the scientific literature report that lobular carcinoma in situ (LCIS), atypical lobular hyperplasia (ALH), and any other lesions with atypia (such as flat epithelial atypia, papillary lesions with atypia, and radial scar with atypia) need excision because cancer is often found at surgical excision (Box 6-6).

Large studies by Brenner and colleagues and Becker and colleagues strongly suggest that radial scars without atypia do not need excision if the biopsy was done with a vacuum-assisted device (as opposed to a large-core device) with removal of at least 12 specimens. It is important that the patient comply with imaging follow-up. Imaging follow-up should be done, as it is for benign concordant percutaneous biopsies, 6, 12, 24, and perhaps 36 months after biopsy to be sure a cancer was not missed. Radial scars without atypia that are diagnosed with a core needle instead of vacuum needle or removal of less than 12 specimens should prompt surgical excision.

Papillary lesions without atypia might be safe to follow without excision if the biopsy criteria outlined for radial scars are met, but there are no large studies to prove that. Currently individual decisions are made to either excise or follow papillary lesions, mucocele-like lesions, and pseudoangiomatous stromal hyperplasia (PASH), if they have no atypia (see Box 6-6). However, PASH or hemangiomas that have possible features of angiosarcoma need surgical excision.

Columnar lesions without atypia, including columnar alterations with prominent apical snouts and secretions, are usually considered benign and do not need excision.

Phyllodes tumor, although generally benign, has a small percentage of malignant forms that are diagnosed only by complete histologic examination. Phyllodes tumor also tends to recur in the biopsy site and should be completely excised by surgery. This means all phyllodes tumors should be excised.

ADH underestimates are less thoroughly studied with ultrasound-guided biopsies, but those underestimates are also decreased with 11-gauge vacuum biopsy compared to 14-gauge core biopsy. ADH underestimates also occur with diagnostic surgical biopsies, but we do not know the overall rate. When just those with positive or close histologic margins for ADH have a repeat operation (meaning there is a large selection bias), Arora and colleagues found a surgical ADH underestimate rate of 27%.

Follow-Up of Benign Lesions

Long-term studies of benign concordant lesions diagnosed at core or vacuum biopsy can occasionally be falsely negative and emphasize the importance of a good follow-up program to minimize the potential for missing cancer. Postbiopsy follow-up imaging, using the same imaging modality that guided the needle biopsy, should be done at 6, 12, 24, and perhaps 36 months postbiopsy for all benign concordant lesions. Specific concordant lesions diagnosed as fibroadenoma or lymph node can have the initial follow-up at 12 months rather than 6 months. If the lesion increases in size at follow-up imaging, the lesion should undergo repeat biopsy by needle or surgical excision.

Complications

Complications are discussed in the “Informed Consent” section of this chapter and are also listed in Box 6-2.

Vasovagal reactions occur more frequently with presurgical needle localization because most are done upright with the patient fasting and often dehydrated. Thus, all personnel in the procedure room must be able to recognize and treat a vasovagal reaction and be able to release the breast from compression for mammographic procedures. A stretcher and resuscitation cart should be in close vicinity to the procedure room, and the patient should never be unaccompanied in the room during the procedure. The patient must be able to respond, be alert, be able to remain motionless during the procedure, and be able to cooperate with the radiologist during the procedure.

Methods to decrease untoward bleeding include familiarity with the patient’s current prescribed medications and over-the-counter self-prescribed drugs, herbs, and vitamins, as well as when the patient should stop taking them. The radiologist works with the referring physician to determine whether administration of Coumadin (warfarin), heparin, or Plavix (clopidogrel) can be safely curtailed. At many facilities patients are instructed to stop taking all pain medications except for acetaminophen for 1 week before the biopsy because aspirin, NSAIDs, and other medications can inactivate platelets. Some institutions instruct patients to also stop taking all herbal medications (particularly Ginkgo biloba, which potentiates anticoagulants), vitamin E, and fish oils for 1 week before the biopsy.

At the other extreme of not stopping medications before needle biopsy, Melotti and Berg reported needle biopsies in 18 patients undergoing anticoagulation therapy. The patients were taking warfarin (n = 11), heparin (n = 1), or aspirin (n = 6). Hematomas measuring 13 to 40 mm occurred in 3 of 8 anticoagulated patients undergoing stereotactic 11-gauge vacuum biopsy. A 10-mm hematoma occurred in 1 of 10 anticoagulated patients undergoing ultrasound-guided 14-gauge core biopsy. Their study suggests that needle biopsy can be performed in anticoagulated patients if the need for biopsy is urgent, but that hematomas may occur after the biopsy.

For ultrasound-guided biopsies, pneumothorax is an unusual but reported complication. The risk of pneumothorax increases if the patient is unable to hold still or is coughing, if the angle needed to biopsy the lesion is very steep, if the lesion is on the chest wall, and particularly if the lesion lies between ribs. Pneumothorax has been reported as a complication of both fine-needle breast aspiration and large-core biopsy. It is imperative that the radiologist identifies the chest wall and pleura before the biopsy to evaluate the trajectory of the needle throw. Taking the extra time to roll the patient into the perfect position so that the needle trajectory is parallel to the chest wall is especially important. When there is a possibility of pneumothorax during the biopsy, the radiologist should obtain informed consent from the patient specifically for the possibility that the “needle could puncture the lung and result in the need for an emergency room visit and possible stay in the hospital, which is very unusual.” Knowledge of pneumothorax and its consequences, as well as strict instructions to the patient to remain immobile during the biopsy, are important for informed consent. If there are serious concerns about pneumothorax during a procedure, one should consider using a needle that has the needle tip inserted to the deepest point without firing. The external cutting part of the needle then fires from the “pre-fire” position to shear off the tissue specimen. The cutting part of the needle stops at the needle tip and has no “throw” beyond the tip. Another alternative is to not do a core needle biopsy and to proceed with needle localization and surgical excision.

Differences between Core and Vacuum Needle Biopsies

In two seminal articles, Parker and colleagues introduced 14-gauge automated large-core needle biopsies guided by prone stereotactic technique in 1991 and by ultrasound in 1993 as practical alternatives to image-guided needle localization and diagnostic surgical excision of nonpalpable lesions. Parker and colleagues then reported successful use of those biopsy methods in multiple institutions in 1994. Burbank introduced directional vacuum-assisted needle biopsies in 1996 as a way to remove more biopsy tissue more rapidly, more accurately, and with less bleeding. Vacuum needles were initially 14-gauge and are now available in 7- to 14-gauge sizes.

The weight of the individual specimens is about 17 mg with 14-gauge core, 37 mg with 14-gauge vacuum, 95 mg with 11-gauge vacuum, and 120 mg with 9-gauge vacuum biopsy devices. The core biopsy needles and some vacuum needles are removed from the breast after acquisition of each tissue sample and re-fired into the breast to obtain each new sample. The most frequently used vacuum biopsy needles are inserted once into the breast, unless a very large lesion requires multiple skin entry sites for accurate sampling.

For stereotactic biopsy, the collection trough of the vacuum needle is rotated “around the clock” to acquire tissue from different parts of the lesion and surrounding tissue, with the external part of the needle staying in the breast and the internal cutting part of the needle extracting tissue out of the breast. For ultrasound-guided biopsy, the collection trough is aimed at the lesion and rotated though just a few clock positions to acquire tissue from different parts of the lesion. The vacuum portion of the biopsy device both pulls in breast tissue to be biopsied (which is a major factor in a vacuum biopsy specimen weighing about twice as much as a core biopsy specimen from the same gauge needle) and vacuums blood out of the breast away from the biopsy site (allowing one to extract more tissue and less blood). In addition, the vacuum biopsy specimens are extracted contiguously. All of these factors make the vacuum biopsy faster and more accurate if initial targeting is correct.

The increased accuracy of vacuum needle biopsy is most dramatic in reducing the calcification miss rate, as judged on specimen radiographs from stereotactic biopsy of calcification lesions. The negative radiograph specimen rate for calcifications using prone stereotactic guidance is 14% with 14-gauge core needle biopsy and 1% with 11-gauge vacuum needle biopsy. These compare with a miss rate of 4% for x-ray–guided needle localization and surgical excision calcification.

As discussed, prone stereotactic biopsies performed with 11-gauge vacuum needles rather than 14-gauge core needles have decreased false-negative rates and decreased underestimation rates for both DCIS and ADH. Ultrasound-guided needle biopsies performed with 11-gauge vacuum needles rather than 14-gauge core needles also have decreased underestimation rate for both DCIS and ADH. There is, however, no proven decrease in the false-negative rate with vacuum biopsies compared to core biopsies done with ultrasound guidance. There is so much selection bias in the literature, however, that valid comparison of the false-negative rate with ultrasound guidance is impossible. Also, wide variation exists in how often different institutions use vacuum needles for ultrasound-guided needle biopsies.

Ultrasound-guided biopsies are usually done with core needles, but biopsy with vacuum needles is being done with increased frequency in the United States. This is most common for lesions that can be difficult to see after extraction of one or two samples (e.g., small lesions, obscure masses, and complex cystic masses [i.e., those with cystic and solid components that may be difficult to see if the fluid is drained from the mass during the course of the biopsy]). They are also used in some institutions for lesions perceived to be in a dangerous location (e.g., close to the chest wall or a prominent vessel) because of a desire to insert the biopsy needle only once into the breast and aim the collection aperture away from the anatomic structure of concern while acquiring tissue.

MRI-guided needle biopsies are done almost exclusively with vacuum needles, but can be done with core biopsy needles. MRI-guided biopsies are covered in Chapter 7.

Patient Audits, Follow-Up, and Noncompliance

Lesions diagnosed as high-risk (most commonly ADH but also the non-ADH lesions discussed above) at core or vacuum needle biopsy and cancer at excision are considered to be high-risk underestimates. They are neither true positives nor false negatives and thus preclude measuring true sensitivity and specificity. Burbank and Parker published an auditing solution to this dilemma in 1998.

It is important to audit the results from one’s own needle biopsy practice to compare with the literature. Most importantly one should determine the false-negative rate with each imaging modality, which requires long-term follow-up. One can more quickly and easily determine the calcification lesion miss rate (from specimen radiographs), the DCIS underestimation rate, and the ADH underestimation rate.

In addition, standard imaging surveillance after the needle biopsy is essential to diagnose missed cancers. Follow-up details are defined in the section “Follow-up of Benign Lesions” in this chapter. The initial informed consent should indicate that imaging follow-up is expected and that the patient is to return for all three or four visits after the procedure.

The problem of follow-up and compliance with follow-up is a difficult one even in the best of hands. Pal and colleagues showed that as many as 40% of women do not return for all their follow-up mammograms after benign results and 15% do not complete the recommended surgery after an abnormal needle biopsy. Jackman and colleagues, however, managed to have the patient return for at least one postbiopsy mammogram in 99% (295/298) of benign concordant lesions having stereotactic core needle biopsy that did not undergo surgical excision. This was achieved with a vigorous, time-consuming follow-up protocol that is not practical for routine use.

As a result of the Mammography Quality Standards Act of 1992, U.S. federal law mandates follow-up on all abnormal mammograms. However, tracking in clinical practice is complicated, time-consuming, and expensive, with multiple costs for personnel, computer updates, and mailing. This requirement is a cause of frustration for many radiologists despite computerized follow-up programs. Goodman and colleagues showed that women’s outcomes may be difficult to track as a result of relocation, changing of insurance, decisions by their referring physicians contrary to recommended follow-up, or other reasons. Those same tracking problems occur after benign concordant needle biopsies. Accordingly, informed consent before biopsy assumes even more importance so that proper patient management can be implemented.