Clinical studyT2-SPACE imaging of the cauda equina for the assessment of leptomeningeal metastatic disease
Introduction
Leptomeningeal metastatic disease (LMD) is an uncommon manifestation of a broad range of neoplasms, including solid organ malignancies, haematological malignancies and primary brain tumours [1]. Most commonly, it occurs late in the course of the disease and is associated with a poor prognosis [1]. The diagnosis of LMD can be challenging, due to the broad range of possible clinical presentations. Symptoms and signs may relate to cranial or spinal nerve dysfunction, be secondary to raised intracranial pressure, or be more general, such as headaches, nausea, vomiting and cognitive changes [1]. Management options include systemic chemotherapy (in particular if there is also uncontrolled systemic disease), intrathecal chemotherapy and radiotherapy (for larger discrete nodular and/or symptomatic lesions) [2].
Traditionally, the diagnosis of LMD has required the identification of malignant cells on cytological assessment of the cerebrospinal fluid (CSF) [3]. Cytology has a high specificity, of over 95%, but the sensitivity is relatively low, generally less than 50% [4]. This makes MRI of the brain and spine an important component of the diagnostic paradigm for suspected LMD [5], [1]. Positive MRI findings are adequate for a diagnosis of LMD in the appropriate clinical context, and do not require cytological confirmation [1]. Older studies have shown that the sensitivity of contrast-enhanced MRI for the diagnosis of LMD is approximately 70% [1], with a specificity of 77–100% [4], [3], [6], [1], though one would expect that the accuracy may have since increased due to technological improvements. The key finding of LMD in the spine is abnormal enhancement along the surface of the spinal cord or the nerve roots of the cauda equina, and this may be linear or nodular.
MRI for the investigation of LMD is typically performed with IV contrast, but unenhanced MRI also frequently demonstrates positive findings [7]. An older study by Moulopoulos et al found that precontrast T1-weighted imaging (T1WI) demonstrated abnormal findings – such as loss of CSF clarity, poor definition of the conus medullaris, thickened and clumped nerve roots, and nodules – in 41 of the 48 patients with leptomeningeal enhancement on contrast-enhanced T1WI (ceT1WI) [7]. Of note, however, all patients in this cohort were known to have LMD [7], and the specificity of some of the more subtle findings – such as loss of CSF clarity or poor definition of the conus medullaris – is debatable in a typical clinical setting. Nevertheless, the presence of multiple nodules would be expected to be a more specific feature of LMD, and the addition of post-contrast imaging does not necessarily provide incremental value, as non-malignant differentials such as nerve sheath tumours would also be expected to enhance.
A recent report has suggested that a Magnetic Resonance myelography (MRM) technique utilising a 3-dimensional, heavily T2-weighted, Sampling Perfection with Application optimised Contrasts using different flip angle Evolution (T2-SPACE) sequence of the lumbar spine, performed in the coronal plane, may provide a useful adjunct to the MRI assessment of suspected LMD [8]. MRM is predominantly utilised for the investigation of intracranial hypotension, based on the detection of cerebrospinal fluid (CSF) leaks. A CSF leak is suspected if streaky T2-hyperintensity extends lateral to the nerve root sleeves, with the confidence increasing with greater transverse extension [9]. The heavy T2-weighting allows the distinction between fluid and normal epidural fat [9], though also provides excellent differentiation between CSF and other tissues, in this case the nerve roots [8], [9].
The potential utility of MRM for the investigation of LMD is based on the detection of nodular lesions of the cauda equina which are not readily visualised or confidently characterised on standard sagittal and axial sequences, which are typically performed with 3–4 mm slices and an inter-slice gap. Detection of nodules is facilitated by the high-resolution volumetric nature of the sequence and the high contrast between CSF and the nerve roots (and their associated pathology) [8]. The lack of significant effects of gadolinium administration means that an inability to detect abnormal linear enhancement of the cauda equina is an expected limitation of this technique. As such, post-contrast imaging remains an important component of the MRI protocol, especially if no abnormal nodularity is demonstrated.
Having since added T2-SPACE with an MRM technique to the MRI protocol for the investigation of LMD at our institution, we sought to review our experience.
Section snippets
Materials & methods
Institutional ethics committee approval was obtained. A list of MRI studies of the spine performed at our institution (a dedicated tertiary oncologic hospital) over a 20-month period (from February 2018 – the time at which this sequence was added – to September 2019) was obtained and examinations including T2-SPACE of the cauda equina were identified. T2-SPACE was routinely included in the MRI protocol for the investigation of known or suspected LMD over this period. The imaging protocol for
Results
T2-SPACE imaging of the cauda equina was performed in 59 patients over this period. This included 30 males (51%) and 29 females, with a median age of 58 years (range 18–85 years). The patients’ underlying cancer diagnoses included a variety of haematological malignancies (n = 14), breast cancer (n = 13), melanoma (n = 11, including one patient with primary meningeal melanoma), lung cancer (n = 9), primary central nervous system (CNS) tumours (two medulloblastomas, two pineal primaries and one
Discussion
T2-SPACE added little incremental sensitivity for the detection of LMD beyond standard ceT1WI, but provided other valuable benefits. Importantly, T2-SPACE was able to confidently characterise equivocal findings on T2WI in several patients. While the reporting radiologist will generally suspect, for example, that an apparent nodule at the lateral aspect of the central vertebral canal reflects volume averaging from degenerative changes, true nodular lesions can also occur in this location.
Conclusion
T2-SPACE imaging of the cauda equina is able to characterise equivocal findings on standard T2WI and improves the detection of nodularity when there is a contraindication to IV contrast administration. There is also value in following-up known nodular lesions of the cauda equina, improving temporal comparison and, in the case of suspected benign lesions, potentially negating the need for IV gadolinium administration at follow-up.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
References (11)
- et al.
Therapy of leptomeningeal metastasis in solid tumors
Cancer Treat Rev
(2016) - et al.
A second look at unenhanced spinal magnetic resonance imaging of malignant leptomeningeal disease
Clin Imaging
(1997) - et al.
Unexpected diagnosis of spinal leptomeningeal metastatic disease on MRI myelography
J Clin Neurosci Official J Neurosurg Soc Australasia
(2019) - et al.
Leptomeningeal metastasis from systemic cancer: review and update on management
Cancer
(2018) - et al.
Neuroimaging and cerebrospinal fluid cytology in the diagnosis of leptomeningeal metastasis
Ann Neurol
(1995)