Clinical Characteristics of Methanol-Induced Optic Neuropathy: Correlation between Aetiology and Clinical Findings

Purpose To show the clinical characteristics, identify the magnetic resonance imaging (MRI) and optical coherence tomography (OCT) features, and observe the visual outcome of methanol-induced optic neuropathy. Methods Clinical data were retrospectively collected from in-patients diagnosed with methanol-induced optic neuropathy in the Neuro-Ophthalmology Department of the Chinese People's Liberation Army General Hospital from January 2016 to January 2021. Results Eight patients were included in this study. The exposure time was 6–34 h for ingestion, 3-4 months for inhalation, and more than ten years for skin absorption. All patients demonstrated bilateral acute visual impairment. Seven of eight patients had other accompanying systemic symptoms. Seven of eight patients demonstrated optic nerve lesions in MRI, and five presented with a hyperintense T2 signal in a “central” type. OCT showed the macular ganglion cell layer and inner plexiform layer (mGCL-IPL) thinning before the peripapillary retinal nerve fiber layer (pRNFL) thinning. The visual improvement was achieved transiently for seven of eight patients after treatment. One patient with a mitochondrial DNA mutation maintained a bilateral no-light perception (NLP) from the onset to the last visit. All patients had poor visual prognoses, with either light perception or NLP. Conclusions Methanol-induced optic neuropathy is a rare bilateral optic neuropathy with a poor visual outcome. A centrally hyperintense T2 signal of the optic nerve is common in methanol-induced optic neuropathy. The thinning of the mGCL-IPL is more sensitive than that of the pRNFL for early diagnosis. A mitochondrial genetic defect may be a predisposing factor for methanol-induced optic neuropathy.


Introduction
Methanol poisoning is a serious life-threatening condition with a mortality rate ranging from 18 to 44% [1]. Persistent visual sequelae are seen in 30-40% of survivors [2]. During the COVID-19 pandemic, the methanol poisoning rate increased unexpectedly in Iran [3,4], Turkey [5], and the United States [6,7]. Most patients were misled by the misinformation that ingesting alcohol or hand sanitizers could protect against COVID-19, while some were caused by alcohol abuse associated with fear and anxiety. Tis resulted in 800 deaths, and many survived with blindness or visual impairment [3][4][5][6][7]. Tus, methanol poisoning continues to be an important issue in public health, emergency, and ophthalmology worldwide.
Methanol is a volatile watery liquid with a pungent alcohol odor. It is naturally contained in fermented beverages and widely used in industry, including cleaning products, antifreeze, pesticides, and fuel sources [8,9]. Accidental oral ingestion is the primary exposure route [1,3,[9][10][11]. Inhalation and dermal absorption associated with occupational exposure or intentional abuse have been increasingly reported [8,[12][13][14][15][16][17][18][19]. Methanol intoxication is caused by its metabolite formate, an inhibitory enzyme of the mitochondrial respiratory chain. Tis resulted in a decreased aerobic production of adenosine triphosphate (ATP) and the impairment of ATP-requiring intracellular reactions, which leads to systemic histotoxic hypoxia and metabolic acidosis [20].
Te optic nerve, which is highly energy dependent, is vulnerable to mitochondrial dysfunction, including methanol poisoning. ATP shortage leads to the interruption of axoplasmic fow and intra-axonal swelling [21]. Te papillomacular bundle fbers, which are small caliber axons rich in mitochondria, are particularly involved [22]. Terefore, methanol-induced optic neuropathy belongs to acquired mitochondrial optic neuropathy, although other mechanisms, including oxidative stress and proinfammatory cytokines, are also reported. Optic nerve histopathology showed mild edematous changes [23] and degeneration of axons and glial cells [2]. Demyelination or axonal necrosis changes are inconsistent in diferent studies [23,24]. Accompanying damage to the retina may involve all its layers [2].
Methanol-induced optic neuropathy is rare in ophthalmic clinical practice. Patients usually come with mild or recovered systemic manifestations and an insidious exposure history, making it challenging to obtain an accurate diagnosis. Some studies reported that magnetic resonance imaging (MRI) abnormalities like T1 enhancement and/or long T2 in the retrobulbar segment of optic nerves were found in methanol-induced optic neuropathy [12,25,26]. Optical coherence tomography (OCT) fnding of ganglion cell layer (GCL) loss selectively involving the papillomacular bundle was also reported [22]. In a four-year follow-up study, methanol-induced optic neuropathy showed progressive chronic axonal loss demonstrated by a depression of the N1P1 amplitude [27].
So far, several questions are worthwhile to be illuminated to advance our understanding. First, do morbidity and the prognosis of visual impairment difer according to the exposure route? Second, as magnetic resonance imaging (MRI) and optical coherence tomography (OCT) are objective and sensitive imaging modes for the optic nerve, can we obtain clues from them to assist in decision-making? Tird, do mitochondrial gene mutations and nutritional status play a role in methanol-induced optic neuropathy?
We conducted this study to interpret the clinical characteristics of methanol-induced optic neuropathy in a Chinese cohort. Te possible diferences in latency, symptom, and visual prognosis caused by exposure variation will be discussed. MRI and OCT features will be given special attention.

Design and Study Population.
Tis is a retrospective cross-sectional single-center study approved by the institutional review board of the Chinese People's Liberation Army General Hospital (PLAGH) and performed in accordance with the Helsinki Declaration. Te clinical records of in-patients diagnosed with methanol-induced optic neuropathy from January 2016 to January 2021 with a follow-up period of over 6 months were collected from the PLAGH database. An experienced ophthalmologist-incharge kept an additional telephone follow-up. Informed consent was obtained from all patients.
Te inclusion criteria were as follows: (1) methanol is the primary exposure; (2) methanol is detected in the blood or urine; and (3) optic neuropathy is confrmed by the acute loss of visual acuity or visual feld, a pupillary defect with light, and visual electrophysiology examination [8,12].

Ophthalmological and Associated Examinations.
Ophthalmological examinations included pupil examination, relative aferent pupillary defect testing, slit-lamp examination, and ophthalmic fundus examination. Bestcorrected visual acuity (BCVA) was tested using the Snellen chart. BCVA below 0.01 was classifed using the semiquantitative scales "fnger count," "hand motion," "light perception (LP)," and "no light perception (NLP)." Te standard protocols for the optic disc cube 200 × 200 circle scan and the macular cube 512 × 128 scan were performed using the spectral-domain OCT device (Carl Zeiss Meditec, USA). To be included, all scans had to have a signal strength ≥6 without motion artifacts. Te optic nerve and brain MRI were imaged using a 3.0T scanner. Te MRI protocols included T1-weighted image (T1WI) and T2-weighted image (T2WI) sequences with fat suppression before and after gadolinium administration.

Laboratory Tests.
Blood tests were performed, including a complete blood count, serum chemistry panel, infectious disease tests (syphilis included), autoimmune tests (antinuclear antibody titer and anti-double-stranded DNA), and serum folic acid and vitamin B12 level tests.
Toxicological tests of blood and urine were performed in the acute phase. Mt-DNA mutation tests were arranged if consent was given.

Statistical
Analysis. SPSS software package version 25 (SPSS Inc., Chicago, IL, USA) was used for statistical analysis. To test the normality of the distribution, the Kolmogorov-Smirnov test and the Shapiro-Wilk test were used. Te level of statistical signifcance was p < 0.05. Quantitative variables with normal distribution were presented as means with standard deviation and ranges. Categorical variables were described as numbers and percentages.

General Information and Exposure.
Te clinical characteristics are shown in Table 1. Eight sporadic patients (fve male and three female) were included in the study, with a mean age of 48.4 ± 10.1 years (31-62 years). From onset to admission to the Neuro-Ophthalmology Department, the average presentation time was 56.1 ± 33.1 days (18 to 109 days). All patients had a mean follow-up time of 25.15 ± 14.10 months (8.6-42.6 months).
As shown in Table 1, fve of the patients exposed through ingestion were male, with patients 1 and 2 drinking industry alcohol by mistake and patients 3, 4, and 5 accidentally drinking draught or counterfeit spirits. Two patients exposed through inhalation were female and were exposed in the workplace. Patient 6 worked in a solid alcohol factory. Patient 7, a barrel cleaner, used industry alcohol as a detergent. She wore gloves but not a mask while at work. Patient 8 was exposed through dermal absorption. She worked in a pipeline and constantly used industry alcohol as a detergent to clean her hands. Te latency from exposure to onset was 6-34 h for ingestion, 3-4 months for inhalation, and more than 10 years for skin absorption (Table 1).
Serum folic acid and vitamin B12 levels were normal for all patients. As shown in Table 1, four patients underwent mitochondrial DNA mutation tests. Mutation m.14502 T > C was detected in one patient (patient 8).

Acute Symptoms.
Only one patient in the ingestion group (patient 1) presented no obvious systemic symptoms. Among the patients, 87.5% (7/8) showed various acute systemic symptoms, including gastrointestinal, cardiopulmonary, renal, and encephalic disorders, and 25% (2/8) with comas received dialysis in the acute phase. One patient (patient 4) had an acute auditory impairment but he recovered six months later. No patient had a central nervous system complaint during their last visit, although two (patients 4 and 7) had a putaminal lesion in the brain MRI (Table 2). Table 3 shows the features of the optic nerve and brain MRI. Among the patients, 87.5% (7/8) had an abnormal optic nerve MRI, including 75.0% (6/8) T1 enhancement (Figures 1(a) and 1(b)) and 75.0% (6/8) hyperintense T2 (Figure 1(c)). Among the seven optic nerve MRI-positive patients, the orbital part was 100% (7/7) involved, and the intracranial part was 42.8% (3/7) involved. Furthermore, 71.4% (5/7) of the patients demonstrated "central" hyperintense T2 signals in the orbital segment, which presented as a "target sign" (Figure 2(a)) in the coronal scan and as a "neutral axis sign" (Figure 2(b)) in the axial scan. In the coronal scan, hyperintense T2 in the intracranial part presented as a "mask sign" (Figure 2(c) (i, ii, iii)). A putaminal lesion was identifed in two patients in the brain MRIs (Figure 3).

MRI and OCT Features.
OCT identifed an apparent decrease in the combined measurement of the thickness of the macular ganglion cell and inner plexiform layer (mGCL-IPL) for all patients. Te peripapillary retinal nerve fber layer (pRNFL) thickness increased in the early stage in one patient ( Figure 4) and decreased in all patients afterward (Figures 5 and 6). Te optic disc seemed normal early and became pale later. Table 4. In our study, all patients experienced bilateral, painless, acute, and dramatic visual loss regardless of the exposure route and the severity of systemic symptoms. All patients had a severe visual loss, including 87.5% (7/8) in NLP and 12.5% (1/8) in LP. Te worst visual acuity was reached within two days in 87.5% (7/ 8) of the patients and on day 10 in 12.5% (1/8).

Visual Function and Treatment. Visual acuity and treatments are listed in
Te patients received multiple treatments. In addition to dialysis for two patients in the emergency department, all patients received high-dose intravenous corticosteroid and nutritional therapy, including vitamin B12, idebenone, folate, and coenzyme Q 10 . Tree patients received plasmapheresis (also called plasma exchange), one received intravenous erythropoietin, and fve received hyperbaric oxygen therapy.
Among the sample, 87.5% (7/8) experienced transient and subtle visual improvement after treatment, but they all ended up having LP or NLP during their last visit. Patient 8, who experienced no visual improvement, carried the mitochondrial DNA mutation m.14502 T > C. She maintained NLP from the onset to the last visit.
Average, standard deviation, lowest, highest, frequency, and ratio values were used in the descriptive statistics of the data.

Discussion
Tis study evaluated the clinical characteristics of methanolinduced optic neuropathy in view of diferent exposure routes. Te incidence of visual and systemic symptoms under diferent exposure routes varied in previous reports. Bebarta et al. found that methanol inhalation was a low-risk factor for visual dysfunction [28]. Givens et al. observed that many inhalational patients developed visual loss [16]. Zhonghua et al. reported that eight inhalational patients developed severe visual loss but no systemic symptoms [12]. In our study, four of fve ingestion patients and all inhalation/dermal exposure patients demonstrated severe visual impairment and systemic symptoms. Tus, inhalation and dermal exposure were as efective as the oral route in producing toxic efects. Te incidence and severity of impairments could be correlated more with internal exposure, such as blood levels, than with the exposure route. Te onset of visual symptoms is hours to two days after ingestion [1], four days to fve years after inhalation [12], and 2-10 h after dermal exposure [10,13]. In our study, low-dose dermal exposure for 10 years also led to acute onset. Terefore, ingestion exposure always leads to an acute onset, while inhalation or dermal exposure may have a long latency (chronic course) before the acute onset. We consider that the diference lies in internal exposure, which is afected by variables such as form (liquid or vapor), concentration, dose, exposure time, and size of the exposure area.
One of the notable fndings in our study was the MRI imaging feature. Te MRI identifed a T1 enhancement or a long T2 in the optic nerve in 87.5% of the patients. Te retrobulbar and the following orbital segment of the optic nerve are most often involved, which is in agreement with previous clinical reports [12,26] and histopathological studies [29], suggesting that the retrolaminar optic nerve is selectively vulnerable to methanol poisoning. Te pathogenesis is presumed to be histotoxic anoxia in a vascular watershed area (the laminar region of the optic nerve), which is the result of direct inhibition of cytochrome oxidase by formic acid, leading to ischemic changes. Additionally, the increasing pressure following oedema in the visual pathway might further aggravate the deterioration due to ischemic     changes. Diferent from other ischemic optic neuropathy such as hyaluronic acid obstruction-induced ischemic optic neuropathy [30], which shows no contrast enhancement in the T1-weighted image, the methanol-induced optic neuropathy shows a T1 enhancement (Figures 1(a) and 1(b)).
Since it is a "metabolite ischemia" due to the action of formic acid inhibiting the cytochrome oxidase, the blood supply is not obstructed. Furthermore, we reported for the frst time the "central" hyperintense T2 signal in methanol-induced optic neuropathy. "Central" hyperintensity ( Figure 2) is diferent from difused or segmental hyperintensity, which has been reported in many publications (also reported in this paper in Figures 1(a)-1(c)). "Central" hyperintense means that the hyperintensity signal is specifcally located in the central axis of the optic nerve. As demonstrated in Figure 2, it presented as a "target sign" in the coronal orbital optic nerve, a "neutral axis sign" in the axial orbital optic nerve, and a "mask sign" in the coronal cranial optic nerve. It was identifed in 62.5% of the patients in our study. A similar feature has been reported in Leber's hereditary optic neuropathy (LHON) [31,32]. According to our clinical observations, it usually presents in mitochondrial optic neuropathies, such as LHON and nutritional optic neuropathy, and accidentally presents in other optic neuropathies, such as radiationinduced optic neuropathy and optic neuritis. We found that the "central" hyperintense T2 lesion was highly suggestive of, although not specifc to, mitochondrial optic neuropathy. We presume that it refects the selective loss of small-caliber fbers, the P-cell population, which are disproportionately present in the papillomacular bundle and pass through the center of the optic nerve. Tese fbers subserve the visual functions of visual acuity, color vision, and high spatial frequency contrast sensitivity [33], which are consistent with the clinical features of dyschromatopsia and cecocentral scotoma in mitochondrial optic neuropathy. Further clinical investigation with a larger sample size may provide more detailed information, especially with a 7.0 T MRI. Te brain MRI identifed a long T2 in the basal ganglia in 25.0% of the patients. Interestingly, these damages are also found in Leber's "plus," mitochondrial encephalomyopathy, lactic acidosis and stroke-like episodes (MELAS) syndrome, and Leigh syndrome, which share the same pathogenic feature as methanol-induced optic neuropathy as a mitochondrial dysfunctional disease [34].
OCT is a valuable tool for detecting ganglion cell alterations in the optic nerve disease [35,36]. In our previous studies, we reported that the total macular thickness loss was a potential predicting factor in ethambutol-induced optic neuropathy [37] and started thinning before pRNFL in LHON [38]. In the present study, we found that the pRNFL could slightly increase in the early stage of methanolinduced optic neuropathy and decrease afterward, but the mGCL-IPL decreased dramatically in the early stage (Figures 4-6). Terefore, the mGCL-IPL is more sensitive than the pRNFL in early methanol-induced optic neuropathy. Mild optic disc edema has been previously reported in methanol-induced optic neuropathy [12]. We considered axon swelling to be induced by an axoplasmic fow disorder because of mitochondrial dysfunction.
Te visual prognosis of methanol-induced optic neuropathy is controversial. According to one study, if a patient survived, most ocular symptoms would improve or resolve [39]. Another study reported high visual morbidity, which was underestimated on discharge [40]. In our study, all patients sufered from severe visual impairment, which may transiently and slightly improve after various treatments but deteriorated afterward. Te co-efect of mitochondrial gene mutation may explain one patient's sufering from NLP without remission in our study.
Tis cohort study has some limitations. First, we used the Snellen instead of the Early Treatment Diabetic Retinopathy  Study (ETDRS) visual chart in the visual acuity test, which could lead to a certain bias in low vision evaluation. Second, the small sample size led to a deviation in the assessment of the visual prognosis. Prospective studies with a larger sample size and a more extended follow-up period are warranted to confrm our fndings and investigate the role of malnutrition in methanol-induced optic neuropathy.

Conclusions
Methanol-induced optic neuropathy is an acquired mitochondrial optic neuropathy with a poor visual prognosis. Te high positive rates of optic nerve MRI and central hyperintense T2 signals are valuable in the recognition of methanol-induced optic neuropathy. Te thinning of the mGCL-IPL is more sensitive than that of the pRNFL for early diagnosis. A mitochondrial genetic defect may be a predisposing factor for it.

Data Availability
Te data involved in this study are owned by the Department of Ophthalmology of Chinese PLA General Hospital. Te data are available from Dr. Huanfen Zhou (zhouzhoueye@ 163.com).

Conflicts of Interest
Te authors declare that there are no conficts of interest.

Authors' Contributions
QS, MS, and YZ are the co-frst authors and contributed equally. QS, MS, QX, and HZ designed and conducted the study. QS, SW, YZ, and ShW collected the data, performed analyses, and did management and interpretation of the data. QS, MS, and HZ prepared the manuscript. QX and HZ performed critical revision of the manuscript, who shared co-corresponding authors. All the authors reviewed and approved the fnal manuscript.