State-of-the-Art Imaging Techniques in Metastatic Spinal Cord Compression

Abstract

Metastatic Spinal Cord Compression (MSCC) is a debilitating complication in oncology patients. This narrative review discusses the strengths and limitations of various imaging modalities in diagnosing MSCC, the role of imaging in stereotactic body radiotherapy (SBRT) for MSCC treatment, and recent advances in deep learning (DL) tools for MSCC diagnosis. PubMed and Google Scholar databases were searched using targeted keywords. Studies were reviewed in consensus among the co-authors for their suitability before inclusion. MRI is the gold standard of imaging to diagnose MSCC with reported sensitivity and specificity of 93% and 97% respectively. CT Myelogram appears to have comparable sensitivity and specificity to contrast-enhanced MRI. Conventional CT has a lower diagnostic accuracy than MRI in MSCC diagnosis, but is helpful in emergent situations with limited access to MRI. Metal artifact reduction techniques for MRI and CT are continually being researched for patients with spinal implants. Imaging is crucial for SBRT treatment planning and three-dimensional positional verification of the treatment isocentre prior to SBRT delivery. Structural and functional MRI may be helpful in post-treatment surveillance. DL tools may improve detection of vertebral metastasis and reduce time to MSCC diagnosis. This enables earlier institution of definitive therapy for better outcomes.

Keywords: Bilsky scale; CT; MRI; deep learning; image-guided radiotherapy; metal artifact reduction; metastatic epidural spinal cord compression; metastatic spinal cord compression; stereotactic body radiotherapy; stereotactic radiosurgery.

Figure 1

Metastatic spinal cord compression is classified with a six-point scale, also referred to as the Bilsky grading scale. Legend: red = tumour; purple line = dura; light blue = cerebrospinal fluid (CSF); yellow = spinal cord. The gradings are as follows—Bilsky 0: tumour that is confined to the bone (i.e., without epidural involvement); Bilsky 1a: tumour with epidural involvement but without indentation of the thecal sac; Bilsky 1b: tumour with epidural involvement and indentation of the thecal sac but without spinal cord contact; Bilsky 1c: tumour with epidural involvement and spinal cord contact without cord compression; Bilsky 2: tumour with epidural involvement and compression of the spinal cord but without obliteration of the surrounding CSF spaces; Bilsky 3: tumour with epidural involvement and severe compression of the spinal cord with complete obliteration of the surrounding CSF spaces [2,21].

Figure 2

Axial T2-weighted (a), pre-contrast T1-weighted (b) and post-contrast fat-suppressed T1-weighted (c) MR images of a lung carcinoma patient with metastatic spinal cord compression (MSCC) at the level of T6. Epidural tumour extension is noted with thecal sac indentation but no spinal cord abutment (solid arrows), best demonstrated on the T2-weighted sequence (Bilsky 1b). Marrow infiltration by the tumour is indicated by hypointensity on the T1-weighted sequence, with corresponding enhancement seen on the post-contrast sequence. Axial T2-weighted (d), pre-contrast T1-weighted (e) and post-contrast fat-suppressed T1-weighted (f) MRI images of the same patient at a higher level (T4) show high-grade MSCC with complete obliteration of the CSF spaces, again best demonstrated on the T2-weighted sequence (Bilsky 3). Prominent enhancement of the epidural and vertebral component of the tumour is seen in post-contrast images.

Figure 3

Lateral radiograph (a), sagittal (c) and axial (e) T2-weighted metal artifact reduction sequence (MARS) MR image of a patient with CFR-PEEK screw implants in the spine. Lateral radiograph (b), sagittal (d) and axial (f) T2-weighted MARS (“WARP”) MR image of a patient with titanium screw implants in the spine. The titanium screws are partially visualized in (d) due to the slice orientation and are best seen in the T9, L1 and L2 vertebral bodies. In contradistinction to the titanium implants, CFR-PEEK implants are radiolucent on radiographs and result in less metal-related artifacts on MRI. Axial T2-weighted image of the patient with titanium implants without MARS demonstrate significantly increased geometric distortion and signal losses, rendering assessment of the vertebral body and contents of the spinal canal difficult (g).

Figure 4

Sagittal pre-contrast T1-weighted (a), post-contrast fat-suppressed T1-weighted (b), T2-weighted (c) and STIR (d) sequences of a 70-year-old male patient with lung cancer that had metastasized to the spine. Axial post-contrast fat-suppressed T1-weighted sequences at the level of L1 (e) and T12 (f), and axial T2-weighted sequences at the level of L1 (g) and T12 (h) of the same patient. There is a pathological L1 compression fracture (solid arrow) with diffuse fatty marrow replacement extending to the posterior elements by the tumour. A convex posterior border is demonstrated as well as an enhancing epidural component. This causes low-grade (Bilsky 1c) MSCC. At the level of T12 (open arrow), no significant marrow replacement is seen and the posterior elements return normal signal intensities. There is subtle retropulsion of fracture fragments and no enhancing epidural component. These findings suggest a T12 osteoporotic compression fracture.

Figure 5

CT Myelogram of a lung cancer patient with suspected metastatic spinal cord compression (MSCC) at the level of C4, without annotations (a) and with annotations (b). The epidural disease is indicated by the yellow colourwash and the spinal cord by the blue colourwash in (b). At this level, the epidural tumour abuts the spinal cord with partial obliteration of the contrast-opacified subarachnoid space (Bilsky 2 MSCC). More superiorly at the level of C3 (c), no evidence of MSCC or epidural disease is seen.

Figure 6

A 51-year-old rectal cancer patient with spine metastasis and concurrent inferior vena cava obstruction with resultant dilatation of the epidural venous plexus. Axial CT images in the soft tissue window (a), and axial post-contrast fat-suppressed T1-weighted (b) MR sequences of the spine. The dilated epidural veins (solid arrows) mimic an enhancing soft-tissue lesion in the epidural space and can be mistaken for metastatic spinal cord compression (MSCC). Another 68-year-old lung cancer patient with spine metastasis. Axial CT images in the soft tissue window (c), and axial post-contrast fat-suppressed T1-weighted (d) MR sequences of the spine. Enhancing epidural disease (dotted arrows) causes high-grade (Bilsky 2) MSCC.

Figure 7

Titanium screw implants (a) and CFR-PEEK implants (b) on CT imaging. CFR-PEEK implants are radiolucent and result in less metal-related artifacts including beam hardening artifacts.

Figure 8

Axial CT-images in a 53-year-old patient with spinal implants with (a) and without (b) a post-processing metal artifact reduction (MAR) algorithm. While there is a noticeable reduction in metal-related artifacts with the MAR algorithm, there is an apparent linear lucency (solid arrow) across the right screw at the pedicle region. This was a new streak artifact inadvertently produced by the MAR algorithm and comparison with the non-MAR images is important to avoid mistaking these new artifacts as implant fractures.

Figure 9

Axial post-contrast fat-suppressed T1-weighted MRI. (a) Metastatic lesion in the L1 vertebral body (yellow arrow). (b) CT and MR fusion for stereotactic body radiotherapy (SBRT) planning; 27 Gy over 3 fractions delivered using volumetric modulated arc therapy. Clinical target volume (CTV—blue outline), planning organ at risk volume (PRV, cord—red outline), 95% isodose (orange colour wash).

Figure 10

Sagittal T2-weighted image (a) of a 69-year-old female patient with metastatic rectal cancer to the spine shows a pathological T5 vertebral fracture. Axial post-contrast fat-suppressed T1-weighted images of the same patient, post separation surgery and planning for stereotactic body radiotherapy (SBRT) (b), and 6 months post-treatment (c). There is reduced tumour enhancement and bulk between (b) and (c) due to favourable post-SBRT response.

Figure 11

Spinal instrumentation with titanium screws in a patient with spinal metastasis. Standard sagittal STIR (a) and post-contrast fat-suppressed T1-weighted (b), and axial T2-weighted images at the level of T12 (e) and L4 (f). Six-month follow-up MRI was performed using AIR^TM Recon DL with sagittal STIR (c) and post-contrast fat-suppressed T1-weighted (d), and axial T2-weighted images at the level of T12 (g) and L4 (h). AIR Recon DL improves resolution and soft tissue contrast, significantly improving the diagnostic quality of images while reducing scan times.

Figure 11

Spinal instrumentation with titanium screws in a patient with spinal metastasis. Standard sagittal STIR (a) and post-contrast fat-suppressed T1-weighted (b), and axial T2-weighted images at the level of T12 (e) and L4 (f). Six-month follow-up MRI was performed using AIR^TM Recon DL with sagittal STIR (c) and post-contrast fat-suppressed T1-weighted (d), and axial T2-weighted images at the level of T12 (g) and L4 (h). AIR Recon DL improves resolution and soft tissue contrast, significantly improving the diagnostic quality of images while reducing scan times.