Reply:

We would like to express our gratitude to the authors for their appraisal of our recent article.¹ As correctly pointed out, we made the decision to include all lesions irrespective of size, which contradicts the recommended guidelines of the North American Imaging in MS Cooperative (https://www.naimscooperative.org/). The stipulated size criteria are based on subjective opinions rather than rigorous scientific validation. The size limitation is an arbitrary restriction intended to enhance specificity at the expense of reduced sensitivity due to inherent imaging limitations at a lower field strength. It is crucial to emphasize that these size criteria have mostly been investigated at standard field strengths without validation of their impact on diagnostic performance at ultra-high-field MR imaging.

Use of these criteria can pose challenges because it can lead to an overestimation of the predictive value in which enrollment is restricted by lesion size, thereby excluding patients with small lesions that may exhibit the central vein sign (CVS) at histology or 7T MR imaging. Al-Louzi et al² used ex vivo 7T MR imaging and histology to demonstrate that lesions with false-negative MR imaging findings for the CVS had a mean maximum dimension of <3 mm, showing that the absence of the CVS on standard MR imaging is partly technical rather than a purely pathophysiologic feature. It has also been demonstrated that adjusting the lesion size cutoff results in a compromise between specificity and sensitivity.² Hence, it is firmly established that the imposition of such artificial constraints will diminish specificity in accurately diagnosing MS using the CVS and may not be suitable at ultra-high-field MR imaging in which a small CVS can be detected more reliably. Undoubtedly, our predictive value of the CVS in the 3T data is anticipated to be lower than that of studies applying size constraints, but our data also reveal that this reduction in sensitivity and increase in specificity are artificially induced when evaluated on corresponding 7T images (as discussed below). Our findings are additionally supported by Tallantyre et al,³ who also did not apply the size criteria at 3T and 7T and reported 100% sensitivity and specificity for the 7T CVS. Their results further demonstrate that 7T T2* identified the CVS in >80% of lesions with a volume of <50 mm³.³

Next, the author also focused on the prevalence and probability of lesion characteristics, The distribution of lesions will differ among cohorts with MS and those without due to the known pathophysiology of these lesions, but periventricular lesions are not a requirement for MS. Kilsdonk et al⁴ have demonstrated with 7T MR imaging that restricting lesions to the deep white matter improves specificity from 88% to 94% but significantly reduces sensitivity from 100% to 81%. Furthermore, we found that 6.6% (4/61) of patients with MS did not have a periventricular lesion, yet they had the CVS in at least 2 small subcortical lesions on 7T MR imaging, with 92.3% (12/13) of these lesions having the CVS. If we were to apply the aforementioned logic and constraints to just these 4 cases, our sensitivity would decrease from 100% to 93.4%.

In conclusion, because our results, along with those of others, demonstrate the exceptional sensitivity and specificity of 7T MR imaging without the need to impose constraints on lesion size, it would be illogical to apply artificial constraints to 7T imaging in an attempt to increase specificity, which is already at 100%.³ While we acknowledge that size constraints may improve specificity in lower-field-strength MR imaging at the expense of false-negative results, our findings suggest that these constraints are likely unnecessary at ultra-high-field MR imaging.

References

1.
1. Okromelidze L,
2. Patel V,
3. Singh RB, et al
. Central vein sign in multiple sclerosis: a comparison study of the diagnostic performance of 3T versus 7T MRI. AJNR Am J Neuroradiol 2023;45:76–81 doi:10.3174/ajnr.A8083 pmid:38164557
2.
1. Al-Louzi O,
2. Manukyan S,
3. Donadieu M, et al
. Lesion size and shape in central vein sign assessment for multiple sclerosis diagnosis: an in vivo and postmortem MRI study. Mult Scler 2022;28:1891–902 doi:10.1177/13524585221097560 pmid:35674284
3.
1. Tallantyre EC,
2. Morgan PS,
3. Dixon JE, et al
. A comparison of 3T and 7T in the detection of small parenchymal veins within MS lesions. Invest Radiol 2009;44:491–94 doi:10.1097/RLI.0b013e3181b4c144 pmid:19652606
4.
1. Kilsdonk ID,
2. Wattjes MP,
3. Lopez-Soriano A, et al
. Improved differentiation between MS and vascular brain lesions using FLAIR* at 7 Tesla. Eur Radiol 2014;24:841–49 doi:10.1007/s00330-013-3080-y pmid:24317461

Main menu

User menu

Search

American Journal of Neuroradiology

References

News and Updates

Resources

Opportunities

American Society of Neuroradiology