Comparison of MRI sequences for evaluation of multiple sclerosis of the cervical spinal cord at 3 T
Introduction
MRI plays a pivotal role in the early diagnosis of multiple sclerosis (MS) in patients who present with a clinically isolated syndrome. Modified Barkhof criteria, based on MRI findings, form part of the Revised McDonald's International Panel (IP) criteria when proving lesion dissemination in space and time, and were selected for their high specificity (78%) and sensitivity (73%) [1], [2]. Spinal cord imaging often forms part of the initial assessment, and is specifically recommended when presenting symptoms which occur at the level of the spinal cord do not resolve or brain MRI results are equivocal [3] as cord lesions are felt to be more specific for MS. A cervical cord lesion can substitute for one of the brain lesions when proving dissemination of lesions in space and time [4] (Table 1). Given the high prevalence (up to 90% [5], [6]) of cervical cord lesions, MRI is also used to monitor progress of disease, particularly in patients undergoing treatment with newer immunomodulatory drugs, particularly those enrolled in Phase II and III drug trials.
The current Consortium of MS Consensus Guidelines 2006 MRI cervical spine imaging protocol (Table 2) does not include the short tau inversion recovery sequence (STIR), despite earlier reports of increased sensitivity of this sequence for lesion detection [6], [7], [8], [9] and its frequent use in many centres.
The greatest criticisms of this sequence have largely arisen from 1.5 T imaging, and include increased noise, flow and movement artefact [5], [7], [8]. Recently, a so-called “TI weighted STIR” (T1STIR) with a shorter time to echo (TE) and longer time to invert (TI) than the traditional STIR, was proposed as a more sensitive inversion recovery (IR) sequence to evaluate MS plaques of the spinal cord at 3 T [10]. We therefore include this sequence in our protocol when time permits.
The aim of our study was therefore to evaluate the sensitivities of our routine protocol, which includes a traditional STIR sequence, and the “T1 weighted STIR” (T1STIR) sequence with the currently recommended T2 sequence in detecting demyelinating lesions on our 3 T system.
Section snippets
Patients
A retrospective analysis of all patients with multiple sclerosis who presented for assessment to our institution between May and October 2009 and who had visible cervical cord lesions was undertaken.
MR image acquisition
All imaging was performed with a 3.0 T Siemens imager (Trio™; Siemens Medical Solutions, Germany) equipped with a Quasar gradient system (45 mT/m with slew rate of 200 T/m/s) and a 13 element phased-array coil. All patients had been examined with the routine departmental protocol of sagittal and axial
Results
25 patients were included in this study. All 25 had been imaged with the T2 and STIR sequence, with a total of 104 ROIs from 87 lesions. Just under half of these patients (11) were imaged with all three sequences (T2, STIR and the additional TISTIR), with a total of 43 ROIS from 32 lesions.
All lesions on the STIR and T1 STIR sequence were statistically significantly different from background cord and were also all readily visible. We found approximately 10% of lesions (9/87) were not
Discussion
We found the traditional STIR sequence was 8 times more likely to show superior lesion:cord contrast ratios when compared with the FSE T2 sequence, which is consistent with the recent study which also examined these sequences at 3 T in 12 patients [10]. Our findings are also consistent with Campi et al. [7] and Hittmair et al. [8], both studies in which contrast noise ratios were calculated for 17 lesions at 1.5 T and STIR was found to be superior.
The greater statistical lesion detection rate of
Conclusion
The traditional STIR sequence shows statistically significant superior contrast ratios and superior lesion detection rate when compared with the T2 sequence at 3 T. The TISTIR sequence also shows potential as a “new” sequence to evaluate demyelinating plaques of MS at 3 T, given its statistically significant superior contrast ratios when compared with both traditional STIR and T2 sequences. These findings are consistent with those of earlier smaller studies. We propose that the current guidelines
Acknowledgements
We would like to thank A/Prof Julie Pascoe, Epidemiology and Biostatistics Unit, Department of Clinical and Biomedical Sciences, Barwon Health Campus, University of Melbourne, for her advice regarding statistical analysis.
References (17)
- et al.
Comparison of two MR sequences for detection of multiple sclerosis lesions of the spinal cord
Am J Neuroradiol
(1996) - et al.
Comparison of different MR imaging criteria to predict conversion to clinically definite multiple sclerosis
Am J Neuroradiol
(2000) - et al.
Standardized MR imaging protocol for multiple sclerosis: Consortium of MS Centers consensus guidelines
Am J Neuroradiol
(2006) - et al.
Diagnostic criteria for multiple sclerosis: 2005 revisions to the “McDonald Criteria”
Ann Neurol
(2005) - et al.
Differentiation of multiple sclerosis from other inflammatory disorders and cerebrovascular disease
Radiology
(2002) - et al.
Comparison of three MR sequences for the detection of cervical cord lesions in patients with mutliple sclerosis
Am J Neuroradiol
(1999) - et al.
Comparison of MRI pulse sequences for investigation of lesions of the cervical spinal cord
Neuroradiology
(2000) - et al.
Spinal cord lesions in patients with multiple sclerosis: comparison of MR pulse sequences
Am J Neuroradiol
(1996)
Cited by (46)
2021 MAGNIMS–CMSC–NAIMS consensus recommendations on the use of MRI in patients with multiple sclerosis
2021, The Lancet NeurologyImaging of the spine and spinal cord: An overview of magnetic resonance imaging (MRI) techniques
2021, Revue NeurologiqueShould spinal cord MRI be systematically performed for diagnosis and follow-up of multiple sclerosis? Synthesis
2020, Revue NeurologiqueCitation Excerpt :MAGNIMS consensus guidelines, for spinal cord lesion identification includes a combination of two sets of sagittal images with different contrasts: T2-weighted FSE, short tau inversion recovery (STIR), proton density-weighted sequence (PD) or phase-sensitive inversion recovery (PSIR) [15,21]. STIR seems to be more sensitive than T2-weighted sequences to the detection of lesions but provides more artefacts [30,31]. PSIR appears to be very sensitive especially in cervical cord, but is currently not available or used in all centers [32].
Pearls and Pitfalls of Spine Imaging
2019, Radiologic Clinics of North AmericaImaging Approach to Myelopathy: Acute, Subacute, and Chronic
2019, Radiologic Clinics of North America
- 1
Tel.: +61 03 5226 7032; fax: +61 03 5226 7072.