Abnormal cerebral blood flow in patients with Leber’s hereditary optic neuropathy

Purpose: The study aimed to unravel abnormal cerebral blood flow (CBF) in patients with Leber’s hereditary optic neuropathy (LHON) using arterial spin labeling (ASL) and to investigate the associations among disrupted CBF, disease duration, and neuro-ophthalmological impairment. Methods: ASL perfusion imaging data was collected from 20 patients with acute LHON, 29 patients with chronic LHON, and 37 healthy controls. We used a one-way analysis of covariance to test the intergroup differences in CBF. Linear and nonlinear curve fit models were applied to explore the associations among CBF, disease duration, and neuro-ophthalmological metrics. Results: Brain regions differed in LHON patients, including the left sensorimotor and bilateral visual areas (p < 0.05, cluster-wise family-wise error correction). Acute and chronic LHON patients demonstrated lower CBF in bilateral calcarine than the healthy controls. Chronic LHON had lower CBF in the left middle frontal gyrus and sensorimotor cortex, and temporal-partial junction than the healthy controls and acute LHON. A significant logarithmic negative correlation was shown between CBF of left middle frontal gyrus and disease duration. A significant linear positive correlation was found between retinal nerve fiber layer thickness and CBF in left middle frontal gyrus, and negative correlations between loss of variance and CBF in left middle frontal gyrus and sensorimotor cortex (p < 0.05, Bonferroni correction). Conclusion: LHON patients exhibited reduced CBF in the visual pathway, sensorimotor and higher-tier cognitive areas. Disease duration and neuro-ophthalmological impairments can influence the metabolism of non-visual areas.


Introduction
LHON is a hereditary neurological disease caused by mitochondrial DNA (mtDNA) mutations characterized by painless progressive decline or loss of vision with male predominance males (Chinnery et al., 2000;Poincenot et al., 2020).The main features of LHON anterior visual pathway lesions are loss of retinal ganglion cells (RGC), progressive thinning of retinal nerve fiber layer (RNFL), dilation of the radial papillary capillary (RPC) around the optic disk, and loss of optic nerve axons, and so on (Asanad et al., 2019;Balducci et al., 2016;Cui et al., 2020;Matsuzaki et al., 2018;Rizzo et al., 2012;Wang et al., 2017).
Most LHON mutations are located in genes encoding mitochondrial complex I, also called nicotinamide adenine dinucleotide dehydrogenase subunit (ND1).Mutations in ND1 change the amino acids of ND1 (Yen et al., 2006) and disrupt ATP function, leading to a reduction in ATP production and cell respiration rate.A rapid reduction in ATP concentration was reported in all three types of LHON mutation-carrying cell cultures (ND1, ND4, and ND6) (Zanna et al., 2003).Thus, ND1 dysfunction in patients with LHON can limit ATP production, triggering brain cell metabolism changes and leading to abnormal brain activity.
Thus, it is expected that LHON mutation would blunt the energy metabolism of the patient's brain tissues and cause central never system (CNS) symptoms.Magnetic resonance spectroscopy (MRS) and arterial spin labeling (ASL) are two frequently used non-invasive MRI techniques to evaluate the brain's energy metabolism.For example, phosphorus MRS ( 31 P-MRS) showed defective occipital lobe energy metabolism in LHON carriers of m.11778G > A (Barbiroli et al., 1995).A proton MRS ( 1 H-MRS) study reported decreased absolute creatine followed by decreased absolute N-acetylaspartate concentration in the normal-appearing white matter of LHON carriers (Ostojic et al., 2009).These studies implied that LHON could cause abnormal brain metabolism, even during the pre-symptomatic period.ASL measures the rate of arterial blood transport to the capillary bed (termed cerebral blood flow (CBF)) by saturating (or labeling) the endogenous arterial blood, which is closely related to brain neural activity and metabolism (Alsop et al., 2015).Changed CBF has been detected in blind and high myopia people (De Volder et al., 1997;Ptito et al., 2005;Wang et al., 2020) and neuropsychiatric diseases (Hays et al., 2016;Jann et al., 2015;Xu et al., 2017).To our knowledge, there is no study reporting brain CBF changes in LHON patients or pre-symptomatic carriers.Thus, detecting the potential CBF changes in LHON patients can help us further understand the patho-metabolic mechanism of LHON.
Based on the metabolic (Barbiroli et al., 1995;Ostojic et al., 2009), functional (Rocca et al., 2011;Vacchiano et al., 2019), and structural abnormalities (Long et al., 2019;Wang et al., 2021;Zhang et al., 2021) reported in LHON patients and carriers, we hypothesized that abnormal CBF would also be identified in both the visual and non-visual pathways; furthermore, we hypothesized that the disease process and visual damage would influence the CBF changes in the LHON.To address these hypotheses, we collected ASL data from 55 LHON patients and 37 healthy controls.We first applied voxel-based data-driven statistics to explore which cerebral regions have abnormal CBF in LHON patients.Then several curve fitting models were used to investigate the potential relationships among CBF abnormalities, disease duration, and neuro-ophthalmological metrics in these brain regions.

Participants
From May 2012 to December 2016, we recruited 55 LHON patients diagnosed in Henan Provincial People's Hospital.Inclusion criteria were: (1) with one of three common mtDNA point mutations (m.3460G > A, m.11778G > A, m.14484T > C); (2) progressive non-painfully loss of vision; (3) no history of neurological, psychiatric, or substance abuse; 3) no other ophthalmic diseases; (4) no visible intracerebral and spinal cord lesions.It should be noted that the participants of the current studies overlapped with 4 previous published works; however, the ASL images and contents of this study were independent of early works (Tian et al., 2022;Wang et al., 2017Wang et al., , 2021;;Zhang et al., 2021).
Six LHON patients were excluded from the study, including 1 for missing neuro-ophthalmological examination, 1 for missing T1 structural MRI images, and 4 for artifacts in ASL images.Thus, the study finally enrolled 20 acute LHON (A-LHON) (disease duration < 1 year, ages ranging from 10 years to 57 years old, 10 minors with lower than 18 years, 18 males, 14 cases with m.11778G > A mutation, 1 case with m.3460G > A, and 5 with m.14484T > C) and 29 chronic LHON (C-LHON) (disease duration ≥ 1 year, ages ranging from 13 years to 53 years old, 5 lower than 18 years old, 20 males, 23 cases with m.11778G > A mutation, 2 cases with m.3460G > A, and 4 cases with m.14484T > C).The 49 LHON patients covered a wide range of disease duration spanning the acute and chronic phases (from 3 weeks to 422 months).No patients displayed extra-ocular symptoms and visible intracerebral lesions.We classified the acute and chronic patients with a threshold at 12 months is based on several early studies on the dynamic trajectory of LHON's clinical progressions (Carelli et al., 2017;Kirkman et al., 2009;Pemp et al., 2021), in which the "acute" stage refers to period with progressive or dynamic vision impairment, while "chronic" stage means period with stable vision loss.We also recruited 37 HC with comparable age and gender distribution as the LHON (ages ranging from 11 years to 44 years old, 27 males, 7 minors with age < 18 years) from the surrounding communities by advertising.The inclusion criteria were identical for LHON patients except for no visual impairment and mtDNA mutations.

Neuro-ophthalmological examination
The corrected visual acuity was evaluated by the logarithm of the minimum angle of resolution (logMAR).Octopus perimeter 101G2 program TOP Strategy (Interzeag AG, Haig-Streit Schlieren, Switzerland) was carried out to obtain the visual field and assess mean defect (MD), mean sensitivity (MS), and loss of variance (LV) (Mateus et al., 2016).The average peripapillary RNFL thickness was measured using a high-resolution spectral-domain optical coherence tomography (Carl Zeiss Meditec, Dublin, CA, USA) with a preset diameter of 3.45 mm.

CBF preprocessing
First, the ASL differential image (perfusion-weighted images, PWI) was obtained by subtracting the label image from the control image.Second, the CBF map was calculated based on the PWI images according to the pcASL model (Wu et al., 2014).Third, SPM12 (http://www.fil.ion.ucl.ac.uk/spm) was used to normalize the CBF map to the Montreal Neurological Institute (Ostojic et al., 2009) space.Specifically, the ASL control images of each subject were coregistered to the T1WI after skull stripping; then, the T1WI of each subject was segmented and normalized into MNI space by an Exponentiated Lie algebra (DARTEL); then, the derived deformation flow field was used to warp the CBF images into the MNI space.Fourth, to control for the heterogeneities in labeling efficiency among participants, the normalized CBF map was then divided by the global mean CBF value to obtain a relative CBF map (rCBF).Finally, all rCBF maps were spatially smoothed with a Gaussian kernel of 6 × 6 × 6 mm 3 full-width at half maximum (FWHM).

Association between abnormal CBF and clinical variables
Among the curve-fitting model, disease duration demonstrated a significantly logarithmic negative correlation with the rCBF of the left MFG (R adj 2 = 0.195, ƒ 2 = 0.269, p = 7.80e -4 , Bonferroni corrected) (Fig. 3, Supplementary Table 5).Moreover, there demonstrated a significant linear Using SPM12, A voxel-based one-way ANOVA was used to compare the rCBF differences among groups with age and gender as co-variables (voxel-wise p < 0.001, corrected at a cluster-level threshold of P < 0.05 using family-wise error [FWE] method).Because the identified big clusters had functional heterogeneous regions, we further separated these clusters into five independent regions of interest (ROIs) with approximately 100 voxels around each peak (Supplementary Method, Supplementary Tables 1-2).and the average rCBF of each ROI for each subject was extracted using the DPABI V4.3 (http://rfmri.org/dpabi).Finally, a post-hoc analysis was conducted to compare the rCBF differences in each ROI between each pair of groups after correcting for age and gender effects using SPSS19.0(p < 0.05/5 ROIs/3 group pairs = 3.33e-3, Bonferroni corrected).
Linear and logarithmic curve fitting were carried out to explore the association between the rCBF values of each ROI and clinical variables (including disease duration and neuro-ophthalmological metrics) after regressing out the gender (with or without) and age effects using SPSS 19.0 (p < 0.05/5 ROIs/5 measures = 0.002, Bonferroni correction).In addition, Bayesian information criteria (BIC) was used to select the model with the best goodness of fit (GOF) (Vrieze, 2012), and Cohen's ƒ 2 was used to evaluate the effect size (Selya et al., 2012).

Demographic and clinical variables
No statistical differences were found in age (One-way ANOVA, F = 2.09, p = 0.13) and gender (Chi-square test, χ 2 = 4.40, p = 0.11) among A-LHON, C-LHON and HC.One-way ANOVA revealed that both A-LHON and

Discussion
To the best of our knowledge, this is the first study to explore CBF metabolism alteration using the ASL technique in LHON.Consistent with our hypothesis, we found that the rCBF of bilateral calcarine gyri was reduced in both A-LHON and C-LHON.In the cortex outside the visual pathway (such as the sensorimotor cortex, middle frontal gyrus, and temporal-parietal junction), the rCBF was reduced in only C-LHON.We identified several pairs of associations among the abnormal CBF, disease duration, and neuro-ophthalmological measures.These results may help us understand the metabolic mechanisms of LHON from the perspective of CBF perfusion.
Besides the involvement of the visual pathway, abnormalities outside the visual pathway have also been reported in LHON patients, including lesions in the basal ganglia (Cui et al., 2020;Mercuri et al., 2017), vestibular nuclei (Miyaue et al., 2019), dentate gyrus (Wang et al., 2021), and inferior olivary nucleus (Nakaso et al., 2012), abnormal white matter integrity (Wang et al., 2021) and connectivities (Rocca et al., 2011).In the present study, we also found decreased rCBF in the sensorimotor cortex, posterior MFG, and temporalparietal junction in the chronic LHON.The sensorimotor cortex plays a central role in somatosensory perception and motor control (Ebbesen et al., 2018).The temporal-parietal junctions were also closely associated with visuospatial perspective-taking (Brugger, 2002) and somatosensory processing (Blanke et al., 2004;Orru et al., 2021).The Several possible mechanisms could explain the decreased CBF in the visual cortex.LHON mutations can lead to mitochondrial dysfunction and energy metabolic defects (Zanna et al., 2003).Thus, the reduced CBF in both the acute and chronic LHON may indicate a direct impairment of the visual cortex.Moreover, the RNFL structure of LHON undergoes an initial thickening at the preclinical and acute stages and then progressively thinning (Asanad et al., 2019;Balducci et al., 2016;Borrelli et al., 2016;Mizoguchi et al., 2015;Tian et al., 2022;Zhang et al., 2014).Previous studies have shown that RNFL thickness is significantly associated with disrupted optic tract integrity (Manners et al., 2022;Wang et al., 2017), atrophied LGN (Jonak et al., 2020), and decreased V1 GMV (Barcella et al., 2010).Thus, the reduced CBF may also reflect the secondary degeneration driven by RGC loss.However, we did not find significant associations between the occipital CBF and RNFL thickness or duration in LHON.Two main factors may drive the insignificant association of visual CBF with clinical measures.First, visual loss could directly cause useless-induced CBF reduction in the visual cortex.Second, early studies

Conclusion
This study provided evidence that LHON patients exhibited reduced CBF not only in the visual pathway but also in sensorimotor and higher-tier cognitive areas.Furthermore, the decreased metabolism of these non-visual areas can be influenced by disease duration and severity of neuro-ophthalmological impairments.

Data Availability
The individual original MRI data will not be shared with the public because the subjects' signed permission of personal data spread had not been approved.We promised that the imaging protocols and statistical analysis results would be accessed to public once the draft has been accepted for publication.Any reader can get the shareable data by email to the corresponding author.
Code Availability Not applicable.

Declarations
Ethical approval The research was approved by the Ethics Committees of Henan Provincial People's Hospital and was carried out in compliance with the Code of Ethics of the World Medical Association (Declaration of Helsinki).
Consent to participate Informed consent was obtained from all participants or their legal guardians if they were children or adolescents.

Consent for publication
The work described has not been published before.It is not under consideration for publication elsewhere.Its publication has been approved by all co-authors, if any.Its publication has been approved by the responsible authorities at the institution where the work is carried out.
Competing Interests All authors declare no conflict of interest.
posterior MFG near the premotor area plays an essential role in executing complex motor responses (Bhoyroo et al., 2022;Kombos et al., 1999).One possible explanation for lower CBF in these regions might be related to non-visual sensory and motor dysfunction in LHON, which could explain the abnormal audiovisual perception and motor control in LHON reported in early studies, such as peripheral neuropathy (Finsterer & Zarrouk-Mahjoub, 2018), dystonia (Saracchi et al., 2013), cerebellar ataxia (Funakawa et al., 1995) and hearing dysfunction (Leng et al., 2015;Rance et al., 2012).
It should be noted that there were no evident extra-ocular symptoms in our recruited patients.Moreover, we only found chronic rather than acute LHON had decreased CBF in the non-visual cortices, and lower CBF of these regions was correlated with severer vision impairment and RGC loss.Because vision loss is unlikely cause secondary degeneration of these non-visual areas, other mechanisms might also contribute to decreased CBF in chronic LHON.One possible explanation is the multi-modal sensory integration deficit after visual deprivation.Previous studies have reported decreased functional connectivity between the primary visual cortex, sensory-motor cortex, and auditory cortex in either congenital or late-onset blind people (Liu et al., 2007;Yu et al., 2008).A recent study also described disrupted structural connectivity between the visual cortex and superior temporal gyrus in the LHON (Zhang et al., 2021).Another possible explanation is the increased efficiency of energy utilization of these areas in more efficiently processing non-visual signals in chronic LHON patients, which was supported by increased non-visual perception (Collignon et al., 2009;Fieger et al., 2006) and task-evoked SMC activity (Burton et al., 2003).This experience-dependent plasticity hypothesis should be validated in the future by combining neuro-activity and CBF measures.
There are still several limitations in this study.First, the number of participants was not large enough to elucidate the effects of potential confounders (such as sex and mtDNA variants) on CBF changes, which need to be verified by larger cohorts in the future.Second, due to the old machine model and full-brain coverage purpose, a thicker (4 mm) slice was used in ASL to improve signal noise ratio, which may discount the precision of CBF quantification.Third, since our study only focused on CBF in LHON, the interpretation of the association between neuroimaging and clinical indicators is not sufficient.In the future, multi-dimensional indexes such as multi-modal MRI, ophthalmic examination, and clinical biochemistry can be established to help us understand the neuroplasticity mechanism of functional alterations in LHON.

Fig. 4
Fig. 4 Correlation between abnormal rCBF and neuro-ophthalmological metrics in patients with LHON and HC.Solid fitting line indicates significant associations.Dashed lines indicate missing associations.Abbreviations: LHON = Leber's hereditary optic neu-

Table 1
Demographic and clinical characteristics