Introduction

Multiple sclerosis (MS) and neuromyelitis optica spectrum disorder (NMOSD) are severe demyelinating diseases in central nervous system (CNS) with similar clinical and imaging findings. Accurate differentiation of these diseases is essential for treatment decisions and thus affects prognosis¹. Currently, differentiating MS and NMOSD relies on antibody testing in clinical settings. It is warranted to explore imaging biomarkers for differentiation of MS and NMOSD. This study aimed to determine the differences of MS and NMOSD using physiological and quantitative MRI.

Methods

Study sample: Patients with aquaporin-4 positive NMOSD or MS and age- and sex-matched healthy controls (HC) were recruited from January 2022 to September 2023. All participants underwent brain physiological and quantitative MR imaging. The study protocol was approved by institutional review board and written consent form was obtained from all patients and healthy volunteers. MR imaging protocol: The MR imaging was performed on a 3.0T MR scanner (Ingenia, Philips Healthcare, Best, The Netherlands) with 32-channel head coil. The imaging protocol included phase-contrast (PC), T1-weighted magnetization prepared rapid acquisition of gradient echo (MPRAGE), water extraction with phase contrast arterial spin tagging (WEPCAST)² and T2-Relaxation-Under-Spin-Tagging (TRUST)³ imaging sequences. MR image analysis: The brain structures in T1-weighted MPRAGE images were automatically segmented using MRICloud platform⁴ and the brain volume (BV) was measured. Global cerebral blood flow (CBF) was measured on PC images at 4 main feeding arteries (left/right internal carotid arteries and left/right vertebral arteries)⁵ and T1-weighted MPRAGE images. As shown in Figure 1, WEPCAST MR images were used to selectively measure the amount of labeled water in the superior sagittal sinus, which could give an estimation of the global extraction fraction (E) of water. Together with global CBF, blood brain barrier (BBB) permeability to water was assessed as permeability-surface area product (PS)⁶: $$PS=-\ln(1-E)\cdot CBF$$According to the relationship between blood T2 and oxygenation, the TRUST MR images were processed to obtain whole-brain venous oxygenation (Yv) according to established procedures³. Then, the whole-brain oxygen extraction fraction (OEF) was calculated as: $$OEF=\frac{Ya-Yv}{Ya}\cdot 100\%$$where Ya was the arterial oxygenation level, measured by a pulse oximetry during the TRUST MRI scans. An example of TRUST MRI for the assessment of OEF is shown in Figure 2. All image analyses were conducted using MATLAB R2022a software. Statistical analysis: The one-way analysis of variance (ANOVA) test was conducted to compare MRI quantitative measurements among different groups. Then the univariate and multivariate logistic regression were explored to analyze the association between MRI quantitative measurements and different groups. P value <0.05 was considered as statistically significant. All statistical analyses were performed on SPSS Statistics 26.0.

Results

A total of 46 participants were enrolled, including 16 NMOSD patients (40.4±10.1 years old, 16 females), 15 MS patients (38.1±12.5 years old, 14 females), and 15 healthy controls (HC) (45.7±10.0 years old, 15 females). Clinical characteristics of study population were listed in Table 1. The results of one-way ANOVA test (Table 2) showed that the water PS of NMOSD patients was significantly higher than that of HC (138.8±29.0 ml/100g/min vs. 118.7±17.4 ml/100g/min, P=0.019), while no significant difference was found between HC and MS patients and between NMOSD patients and MS patients (all P>0.05). The OEF of MS patients was significantly lower than that of HC (34.2±4.5% vs. 38.8±4.7%, P=0.031), while no significant difference was found between HC and NMOSD patients and between NMOSD patients and MS patients (all P>0.05). Multivariate logistic regression analysis showed that, after adjusting for clinical confounding factors of age and BMI, OEF remained significantly different between NMOSD patients and MS patients (OR=0.216, 95%CI: 0.057-0.817, P=0.024), but such differences in OEF and PS were no longer statistically significant between HC and NMOSD patients (P=0.073) and between HC and MS patients (P=0.099), respectively (Table 3).

Discussion

Our study provides evidence that both NMOSD and MS patients have cerebral physiological impairments. The potential mechanism of NMOSD is the breakdown of the tight connections between epithelial cells caused by Inflammatory infiltration, and therefore leading to the disruption of the blood-brain barrier. However, due to impaired mitochondrial function, MS uniquely leads to the decline of brain tissue oxygen metabolism, which is also the key to distinguish it from NMOSD. Future studies with larger sample size are warranted.

Conclusion

Our study demonstrated that water PS and whole-brain OEF might be potential indicators for NMOSD and MS, respectively. In addition, OEF is the key to distinguish MS from NMOSD.

Acknowledgements

No acknowledgement found.