A practical guide for planning pelvic bone percutaneous interventions (biopsy, tumour ablation and cementoplasty)

CT has been found to provide the most convenient and safest guidance modality [17]. Gantry angulation can be used to facilitate needle placement. Unless using the highly radiating CT fluoroscopy, its main drawback is the absence of real time imaging and control in the Z-axis, which are important in some procedures, such as cementoplasty. This disadvantage can be solved by combining the CT scanner with a C-arm fluoroscopy [10, 12, 18]. In this setting, CT would be typically used for guiding needle placement, while the fluoroscopy allows real-time evaluation of cement distribution during the injection. CT scanner can also be used to guide biopsies by identifying anatomical landmarks when lesions are only visible with other imaging techniques such as magnetic resonance imaging (MRI) or positron emission tomography (PET) scan [16, 19].

The complex anatomy of the pelvis is rather difficult to appreciate with fluoroscopy. In some cases, bi-plane fluoroscopy provides enough guiding information as it allows multiplane imaging [20] (Fig. 1). The development of flat panel detectors and cone beam CT also allows real-time imaging and capability for three-dimensional reformations with lower radiation doses than CT [20‐22]. Although the image quality of the multiplane reformations is inferior to a conventional CT scanner, it is usually appropriate for guiding bone procedures safely [21, 23].

MRI guidance could be especially valuable when tumour lesions are not seen with other modalities [16]. The absence of ionising radiation makes it an option for procedures in particular situations such as pregnancy. Because of its high soft tissue contrast, it can show chemical and thermal variations during tumour ablation procedures [5, 20]. However, its main disadvantages remain its higher cost, longer procedure times and limited availability. In addition, the material used during MR-guided interventions should be ideally compatible, which makes it more expensive, or used with special precautions [5, 20, 24].

On occasions, ultrasound can be used to guide pelvic bone biopsies, particularly in the presence of tumour extension to the soft-tissues (Fig. 2) [16]. Among its advantages are its availability, low cost, absence of ionising radiation, and real-time and multiplane imaging [16, 20, 24].

Electromagnetic navigation allows real-time device tracking [22]. Needle position information in the magnetic field is processed and placed on a preprocedural imaging (CT or MRI), which is used as a map. Generally, fluoroscopy or CT scan images are acquired to confirm the final needle placement in the target, as the main pitfall is the potential mismatch with the preprocedural images [22].

Other techniques, such as laser guidance, can facilitate needle placement [21, 22]. After CT or cone beam CT images are obtained, the target point is defined and a straight path from the skin is selected [21, 22]. A laser beam indicates the chosen entry site on the skin and the needle orientation [21, 22].

At this level, the bone landmarks are the iliac wings with the superior anterior iliac spine and the posterior iliac tuberosity, the sacrum and the sacroiliac joints. The important structures to notice and avoid when performing a bone procedure are the iliac vessels and femoral nerve, the lumbosacral trunk, the sacral nerves and all the pelvic visceral structures (Fig. 3). Five different approaches are routinely performed:

At this level, the bone landmarks are the acetabular roof, sacrum and coccyx. Attention should be paid to the pelvic visceral structures, the femoral nerve and external iliac vessels anteriorly, and the internal iliac vessels and gluteal branches, sacral plexus and structures exiting the pelvis through the greater sciatic notch posteriorly (Fig. 10).

We recommend to avoid the anterior approach as the femoral nerve and vessels are lying just anterior and the risk of traversing the peritoneum exists. A posterior approach must be planned very carefully because many vessels and nerves transit through the greater sciatic foramen.

At this level, the bony landmarks are the anterior and posterior walls of the acetabulum and the femoral head. Special attention must be paid to the femoral nerve and vessels anteriorly and to the sciatic nerve posteriorly (Fig. 14).

At this level, the bony landmarks are the ischial tuberosities, pubic symphysis and femoral neck. Femoral nerves and vessels anteriorly and sciatic nerve posteriorly, between the ischial tuberosity and the femur, should be avoided (Fig. 18).

A well-planned approach for percutaneous bone biopsies is essential to obtain a representative sample of the lesion and to minimise procedural risks [4, 16, 24‐26].

If a primary bone lesion is biopsied, gluteal muscles and rectus femoris should be avoided, as they are essential in a limb-sparing procedure and the needle tract would have to be resected [16, 26‐28].

When a pelvic bone metastasis is favoured, intramuscular needle path is less determinant, albeit cases of tumour seeding along the needle path after core biopsy have been reported [25]. A direct access through gluteal muscles can be used provided that no major vessels or nerves lie in the needle path. However, reducing the length of the path in the soft tissues decreases the risk of bleeding. Moreover, iliac lesions tend to have a longer diameter in the wing axis and more material can be sampled by accessing the lesion along its greater axis.

The choice of the needle depends on the mineralisation of the lesion and the presence of a cortical breach [16]. A 14- to 16-G soft tissue cutting biopsy needle is the favoured choice whenever possible. A 10- to 16-G bone biopsy needle is indicated for dense lesions and when the cortex is intact. A coaxial technique is recommended because it allows keeping bone access to take several samples and occasionally perform percutaneous embolisation, and also protects the needle path from tumour dissemination [24, 25].

The number of samples required depends on the pathology department of each institution [3, 16, 24]. Usually two or three samples in formalin are enough [3, 24‐26]. More samples may be needed when using smaller-gauge needles, mainly in paediatric patients [26]. If a lymphoma is suspected, a sample in saline serum should be sent to allow the realisation of flow cytometry [16]. Because infection is an occasional mimicker of bone tumours, systematically sending one or two samples for microbiological analysis is a good practice in uncertain cases [29].

Percutaneous tumour ablation can be curative or palliative [5‐7, 30]. There are several techniques available with different indications: ethanol, laser, radiofrequency, microwave and cryoablation [5‐7, 30].

For a safe procedure, it is essential to carefully plan the approach in order to obtain a good coverage of the target lesion and avoid neighbouring critical structures [6, 7]. Frequently, insertion of several applicators is needed to obtain adequate lesion coverage (Fig. 21). MRI has the unique ability to monitor chemical and thermal variations in the treated area. With cryoablation, CT can also demonstrate the formation of an “ice ball” as a hypodensity in the treated area, although this is much more evident in the soft tissues than within the bone (Fig. 22) [5, 7, 16, 20, 30].

Different techniques have been described to prevent thermal damage to adjacent critical structures (mainly nerves and visceral structures), such as temperature monitoring to prevent overheating or overcooling, carbon dioxide gas or liquid dissection to increase the distance between the target area and critical structures and counteract temperature changes [30‐32]. Covering the skin with sterile gloves filled with warm saline can also be helpful to prevent frost bites during cryoablation of superficial lesions (Fig. 22) [30].

Pelvic cementoplasty is used for pain management and bone reinforcement in certain cases of pelvic bone fractures and metastasis [8‐16]. A well-planned approach will be determinant for optimising bone filling with acrylic surgical cement, while reducing the risk of extraosseous leakage (Figs. 16, 19 and 23). Treatment of extensive lesions may require the insertion of several needles to optimise bone filling [9]. After needle positioning, a pasty cement is injected under real-time imaging control in order to stop the injection when a satisfactory filling is obtained or a leakage is detected [10‐13, 16]. Extraosseous cement leakage in the vicinity of neural structures (typically, sacral canal and foramina, and posterior aspect of the acetabulum) and into the hip joint should be avoided [9, 10, 12, 13]. Neural pain due to cement leakage next to a nerve can be treated with cortisone infiltrations around the affected nerve [10, 13]. Symptomatic leakage to the hip joint may uncommonly require surgical removal of the cement [9].

Pelvic fixation with percutaneous screws can be used as a treatment for non-displaced fractures or to prevent fractures in patients with lytic metastases (Fig. 24) [14, 15]. The approach and the screw length can be safely planned with CT. In cases of osteolytic metastases, this technique is ideally combined with cementoplasty to allow a better fixation and support [15].

All these described techniques can be used alone or in combination to obtain better pain control and mechanical support [5, 15, 16]. When combining techniques, it is usually feasible to use the same coaxial access to the target area, in order to reduce the risk of complications along the approach and to reduce the procedural time (Figs. 11, 12, 19 and 24).