Non-focal liver biopsies are generally performed using sonography. Focal liver biopsies may be guided using ultrasound, CT, or rarely, MRI.

Ultrasound is the preferred modality for non-focal (Fig. 13.1) and focal (Fig. 13.2) percutaneous liver biopsy for several reasons. US offers real-time visualization and multiplanar capability, is relatively inexpensive, and is readily accessible. The lack of radiation is another advantage, particularly in younger patients. However, sonography is limited in its ability to identify the pleural space, which may increase the risk of post-procedure pneumothorax. Additionally, collapsed bowel may be difficult to discern.

Preliminary sonography is performed to evaluate the liver as well as the position of the gallbladder, if present. Doppler sonography is useful for the identification of intervening vessels [10, 18]. The most common practice is use of freehand technique. In general, one hand controls the transducer, whereas the other controls the biopsy needle. Some physicians prefer to use a biopsy needle guide, which attaches to the transducer. In either case, the needle can be visualized as an echogenic complex; visualization of the needle may be improved with a jiggling “in-and-out” motion [18]. Echogenic needle-tip enhancers are also commercially available.

CT guidance is associated with longer procedure times, higher cost, and radiation compared to US. In addition, the needle trajectory is generally restricted to the axial plane. CT guidance is indicated in cases where the target lesion is not well demonstrated on sonography (e.g., cirrhosis) or if there is a concern for intervening structures such as bowel based on pre-procedure imaging. CT guidance may also be helpful in obese patients or patients who had a nondiagnostic result from a prior US-guided biopsy. New practitioners may also find that CT-guided biopsy is generally easier to perform than US guidance, with a shorter learning curve.

Pre-procedure images are reviewed, and a grid with radiopaque lines is placed on the patient’s skin in the area of the expected needle entry site. A preliminary CT is performed and the biopsy trajectory is confirmed (Fig. 13.3). The needle entry position on the grid is selected; the angle of entry and distance to the lesion are also assessed on the images. The skin is then marked. If an entry site is chosen that is outside the field demarcated by the grid, the skin entry site may be marked by measuring the transverse and/or craniocaudal distances of the entry site from a selected point on the grid. For example, in an anterior approach, if the planned entry site is located medial to the most medial radiopaque grid line, the distance between that line and the entry site can be measured on the images and translated on the skin. A craniocaudal component can also be assessed by counting the number of axial images above or below the CT image in which the grid is last seen and multiplying that number by the slice thickness. Alternatively, the grid position can be modified and repeat CT images obtained. After a sterile preparation and the application of local anesthetic, a needle is inserted at the marked entry site and advanced in step-by-step fashion, with intermittent CT imaging to confirm needle position. If a coaxial needle is used, the stylet may be removed prior to each CT scan to reduce needle artifact that may obscure the lesion (Fig. 13.4).

Although many practitioners use a stepwise approach as outlined above, CT fluoroscopy (CTF) may be useful as it may reduce procedure duration [19, 20]. Although reduction in needle placement time has been documented, CTF has not proven to provide a statistically significant increase in procedure success [19]. In CTF, images may be reconstructed at a rate of up to 8 images per second and may be utilized in either continuous fluoroscopy mode or as a quick-check technique [19, 21]. With the former, the use of CT is analogous to conventional fluoroscopy, as needle manipulation is visualized in real time. The quick-check technique involves the use of single CT fluoroscopic spot images to intermittently check needle position. The quick-check technique is analogous to the conventional step-by-step CT technique but improves upon it in that the operator has use of a “floating table,” enabling quick movement of the table and the aid of a laser light to demarcate the axial slice, as well as hand and foot controls. Although the quick-check technique does not fully utilize the real-time capability of CTF, it improves on procedure time compared to conventional CT technique and significantly lowers radiation to patient and operator compared with the continuous fluoroscopy technique [19].

The success of any image-guided procedure is predicated on visibility of the lesion. CT-guided biopsy may be challenging when the targeted lesion is isodense to surrounding liver parenchyma on non-contrast images. In such cases, prior cross-sectional imaging that best demonstrates the lesion should be reviewed and the position of the lesion on preliminary CT images located based on anatomic landmarks. A reference needle (e.g., 20-gauge Chiba) may be advanced to this position using CT guidance (Fig. 13.4). Contrast-enhanced images can then be obtained and the location of the lesion relative to the reference needle assessed. A second needle entry site may be selected and the biopsy needle advanced to the lesion; although the lesion may no longer be enhancing while the second needle is inserted, the first needle continues to serve as a reference. Lesions identified on MRI images are occasionally not visible on US and contrast-enhanced CT. If MR guidance is unavailable, then CT guidance may be considered with the sole use of anatomic landmarks.

Another limitation of CT guidance is the lack of multiplanar capability. Hepatic dome lesions, for instance, may be better suited for US guidance if visible sonographically or MRI guidance, if available. If CT is to be used, gantry angulation may be helpful to avoid pleural or pulmonary transgression, although this may be time-consuming. A transpleural or transpulmonary trajectory may be unavoidable (22). The need for such a trajectory may be anticipated on prior imaging, in which case informed consent of the patient may include a discussion of pneumothorax and chest tube placement. The decision to pursue a transpleural or transpulmonary route depends on the clinical scenario; it may be performed safely although the operator should be prepared for a pneumothorax if one is detected on CT images (Fig. 13.5). A transpleural or transpulmonary route may be a relative contraindication in patients with tenuous clinical status who may not be able to tolerate a pneumothorax with chest tube placement. The risk of pneumothorax increases with the number of pleural transgressions; hence, the risk of pneumothorax with a transpulmonary approach is generally considered higher than with a transpleural approach as two pleural surfaces need to be traversed [22–25].

MR guidance offers multiplanar imaging capability and excellent soft tissue contrast and lacks radiation [26]. However, its use is limited by procedure duration, cost, and limited availability compared to US and CT [27]. MR guidance is rarely indicated and may be helpful when encountering lesions within the hepatic dome, lesions which are poorly visible on US or CT, or when a nondiagnostic result is obtained after an US- or CT-guided biopsy.

The technique varies with the available equipment, and as such, a complete description of the available MR-guided techniques is beyond the scope of this chapter; the reader is referred to several excellent papers [26, 27]. In general, open- and closed-bore systems are available for interventional procedures. Open-bore designs offer improved patient access compared to closed-bore systems. C-shaped open-bore designs enable improved horizontal but poor vertical access; hence, some patients may need to be positioned in a decubitus position for anterior or posterior approaches [21]. Double-doughnut configuration open-bore systems enable excellent vertical and horizontal access [21]. These systems typically operate in the 0.2-T–0.5-T range, and hence, the quality of imaging is poorer compared to closed-bore systems that operate in the 1-T–1.5-T range but nevertheless may be sufficient for biopsy [26, 28]. Although a closed-bore design may offer better signal-to-noise and contrast-to-noise ratio, patient access is restricted, particularly in obese patients; instrument manipulations are thus often performed with the patient outside of the magnet as with conventional CT guidance. However, the availability of closed-, short-, wide-bore 1.5-T systems may offer improved access for interventions allowing needle manipulations when the patient is in the magnet while maintaining excellent imaging quality [27].

A step-by-step technique similar to CT has been described with both systems, although MR fluoroscopy may also be performed [27]. In general, preliminary non-contrast MRI images are obtained with a three-plane localizer sequence as well as fast T2-weighted spin-echo sequences; gradient-echo sequences can also be employed before and after gadolinium administration [26]. If a step-by-step technique is employed, as is commonly with conventional closed-bore systems, an MR-compatible grid visible on either T1- or T2-weighted sequences is applied to the skin. The technique is similar to CT, with the exception that the needle trajectory is not restricted to the axial plane [26]. T1- or T2-weighted sequences are obtained depending on the type of grid used. Software specific to the MR scanner is used to select the needle entry site, plan the angle of the needle trajectory, and measure the distance to the lesion. The needle is inserted at the entry site and is advanced using a stepwise technique, with each manipulation performed while the patient is outside the magnet.

Open-bore as well as closed-, short-, wide-bore systems may enable real-time guidance. By repeating fast imaging sequences, the operator’s palpating finger rather than a grid may be use to mark the needle entry site [26, 27]. Confirmatory images in two orthogonal planes may be obtained when the needle tip is at the edge of the lesion to detect any significant deviation of the needle [27]. If there is difficulty correcting needle deviation after several attempts using MR fluoroscopy (e.g., firm liver due to cirrhosis that restricts manipulation of soft MR-compatible needles), the stylet can be inserted into the needle while the patient is outside the magnet and the needle redirected, with intermittent imaging used as in a conventional step-by-step approach [27].

In general, the needle appears hypointense with a hyperintense rim on MR [26]. Many variables affect the degree of artifact induced by the needle. Distortion is worse with smaller needles, increased needle angle relative to B0, higher-field MR scanners, and gradient-echo versus spin-echo sequences [26].