Safety and effect of topical neostigmine ophthalmic solution in animal models

Abstract

The study aimed to evaluate the safety and effect of various dosages of topical neostigmine ophthalmic solution (TNOS) in animal models and establish a recommended dosage for further studies on myasthenia gravis (MG) diagnosis. A placebo-controlled, sequential ascending dose study was conducted in healthy rabbits. Eighteen eyes were randomized to receive one of three concentrations (1.0, 1.5, and 2.5 mg/mL) of TNOS. Pupillary sizes over time were compared with control group that received normal saline solution (NSS). Mean pupillary size of the 1.5 mg/mL group was significantly smaller than control at 30 to 180 min after instillation (maximal mean difference (TNOS - NSS); mMD − 2.58, 95% confidence interval (95% CI) -4.24 to -0.93, p = 0.006). The mean pupillary size of the 2.5 mg/mL group was significantly smaller than the control at 60 to 90 min (mMD − 3.13, 95% CI -5.07 to -1.18, p = 0.005). No significant difference in pupillary size between 1.5 and 2.5 mg/mL groups was observed at any time points. No systemic or ocular complications were observed in any concentration. The 1.5 mg/mL TNOS was the lowest efficacious dose in this study without adverse effects and was considered a promising starting dose for further MG studies.

Introduction

Myasthenia gravis (MG) is an autoimmune disease caused by antibodies against nicotinic acetylcholine receptors (AChR), resulting in muscle weakness. Most patients initially present with ocular symptoms, including ptosis and diplopia, and some of them have isolated ocular MG without generalized conversion, especially after two years from the onset of symptoms¹. Currently, diagnosis of MG can be made by several methods, including clinical, serologic, electrophysiologic, and pharmacologic testing. The clinical diagnosis remains challenging due to clinical fluctuation, degree of symptoms, and similarity of manifestations to other conditions². Serum acetylcholine receptor antibody (AChR-Ab), the most specific diagnostic test, is useful in adults with generalized MG; however, sensitivity is only 50% in cases with ocular MG. On the other hand, single-fiber electromyography (SF-EMG) is the most sensitive test, though it requires the patient’s cooperation and tolerability, and the provider’s expertise in performing the test. Moreover, both serologic and electrophysiologic facilities are not available in many hospitals^3–5.

Acetylcholinesterase inhibitor (AChEI) plays a role in both the diagnosis and treatment of MG by inhibiting the breakdown of acetylcholine, thus increasing the amount of acetylcholine to interact with the nicotinic AChR and, consequently, alleviating muscle weakness. Intravenous edrophonium chloride, a short-acting AChEI, and longer-acting intramuscular neostigmine have been used as the pharmacologic test of MG. An alleviation of ptosis or diplopia after an injection of both AChEIs supports the diagnosis. Despite their high sensitivities and specificities, they are not commonly utilized due to serious cholinergic side effects, including cardiac arrest, bradycardia, and bronchospasm. The procedure may cause patient discomfort (e.g., gastrointestinal distress and injection-related side effects) and requires medical providers for drug administration and complication monitoring. Aside from nicotinic effects, neostigmine can stimulate muscarinic AChR existing in the iris sphincter muscle and lead to pupillary constriction^6–9.

Accordingly, we aimed to develop topical neostigmine ophthalmic solution (TNOS) for potential use in MG diagnosis. The primary objective is to evaluate the safety of various dosages of TNOS in preclinical animal models. The secondary objectives are to evaluate the cholinergic effect of TNOS on pupillary size and to identify an optimal dosage as a candidate starting dose for further studies regarding MG diagnosis in patients with ocular muscle weakness. This preclinical safety and pharmacodynamic study was performed in healthy rabbits by measuring pupillary response as a surrogate for cholinesterase inhibition. Thus, any extrapolation to MG diagnosis remains speculative until the TNOS is investigated in disease models or humans.

Methods

Topical neostigmine ophthalmic solution preparation

Standard commercially available parenteral form of neostigmine methyl sulfate solution (2.5 mg/mL; 1 mL/ampule) (Neostigmine GPO^®, Government Pharmaceutical Organization, Bangkok, Thailand) was used. The TNOS was prepared under aseptic conditions by a well-trained ophthalmologist (CO) on the day of the experiment. Neostigmine solution was diluted with normal saline solution (NSS) to the prespecified concentrations of 1.0, 1.5, or 2.5 mg/mL. The remaining substance in the ampule was discarded.

Sterility testing

The 10-mL diluted 1.5 mg/mL TNOS in sterile standard eye dropper bottles were kept in four different conditions for 28 days after TNOS preparation. The conditions included refrigerated control, refrigerated control with simulated drug use (i.e., squeezing one drop to simulate patient use once a day and immediately putting back into the storage), room temperature, and room temperature with simulated drug use. The sample size for each group consisted of three replicates of eye dropper bottles. To enumerate bacterial and fungal colonies, the diluted drug from each eye dropper bottle was cultured on 0, 7, 14, 21, and 28 days after drug preparation. Blood agar (Biomedia^®, Nontaburi, Thailand) and chocolate agar (Biomedia^®, Nontaburi, Thailand) were used to detect bacteria, and Sabouraud dextrose agar (SDA; HiMedia^®, Maharashtra, India) was used for fungal culture. A 50-µL sample from each bottle was dropped and then spread on the surface of the agar medium using a sterile spreader. To determine the bacterial colonies, plates were monitored on day 1, 3, and 5 after incubation at 35˚C with a 5% carbon dioxide (CO₂) condition. To observe the fungal colonies, SDA plates were observed on day 1, 3, 5, 7, and 30 after incubation at 25˚C. When microorganisms were not detected during the incubation periods based on the types of culture medium, it was reported as “no growth”.

Preclinical animal models

Study design and ethical considerations

We conducted a placebo-controlled, sequential ascending dose study in animal models. The study protocol was approved by the Institutional Animal Care and Use Committee (IACUC) of Chulalongkorn University in accordance with university regulations and policies governing the care and use of laboratory animals (COA No. 2273032). The study was carried out following the Ethical Principles and Guidelines for the Use of Animals for Scientific Purposes, edited by the National Research Council of Thailand, and the authors complied with the Animal Research: Reporting of In Vivo Experiments (ARRIVE) guideline.

Animal preparation and housing system

Nine New Zealand white, 12-week-old, healthy female rabbits (National Laboratory Animal Center (NLAC), Mahidol University, Thailand) weighing 2.0 to 2.5 kg were used in the experiment. The rabbits were kept in a strictly hygienic conventional housing system of the Chulalongkorn University Laboratory Animal Center. The husbandry conditions were a controlled temperature of 20 ± 1 °C, a humidity of 50 ± 20%, standard fluorescent light, and a fixed standard light cycle (12-hour dark:12-hour light). A conventional laboratory diet was used for feeding, with an unrestricted supply of drinking water.

Administration of neostigmine and control

Nine rabbits were randomized into three groups (n = 6 eyes/group) to receive the different concentrations of TNOS, including 1.0, 1.5, and 2.5 mg/mL. One rabbit from each group was randomized to receive NSS instillation into both eyes as a concurrent control (n = 6 eyes) only on the day before TNOS instillation. Then, for each TNOS concentration group, a 50-µL TNOS was instilled into both eyes of each rabbit daily for 7 days to investigate the safety and effect of prolonged repeated administration. The TNOS instillation was conducted in a stepwise manner from low to high concentrations of TNOS groups (Fig. 1a). The solution was placed in the conjunctival sac after gently pulling the lower eyelid away from the eyeball, and then the lids were held close for a few seconds to prevent loss of solution. The experiment in the next escalating TNOS concentration would be terminated if a rabbit developed any severe adverse events related to TNOS instillation (criteria described in the next section). The rabbit with severe adverse events would be euthanized, and further testing with more animals would not be conducted. The euthanasia process would comply with the American Veterinary Medical Association (AVMA) Guidelines for Euthanasia of Animals: 2020 Edition, using intravenous sodium pentobarbital followed by CO₂ inhalation. If the rabbits were healthy at the end of the experiment, they would be transferred to the Faculty of Veterinary Science, Chulalongkorn University for educational purposes.

Fig. 1.

Demonstrates experimental procedure and outcome measurements (a) Demonstrates the administration of NSS and TNOS in each group of rabbits. (b) Demonstrates the timing of outcome measurements. NSS = normal saline solution; TNOS = topical neostigmine ophthalmic solution.

Regarding dose rationale, the maximal concentration of available parenteral neostigmine is 2.5 mg/mL, which also produced a significant miotic effect in a previous study¹⁰. This study extended the dose to lower concentrations, based on the modified Fibonacci sequence recommended in the dose-escalation design, to determine the lowest efficacious dose.

Assessment of safety

To investigate safety, general physical and ocular examinations were performed. The vital signs (i.e., pulse rate (PR), blood pressure (BP), respiratory rate (RR), and oxygen saturation (SpO₂)) and ocular examinations were daily evaluated prior to NSS or TNOS administration, 1 h, and 24 h after each instillation. The conjunctivae, cornea, and iris were observed under a handheld slit lamp biomicroscope for signs of ocular irritation or serious eye damage. Grading of ocular lesions using the Organization for Economic Cooperation and Development (OECD) Guidelines for the Testing of Chemicals regarding the in vivo eye irritation/serious eye damage was performed¹¹. A dilated fundus examination with an indirect ophthalmoscope was performed at 7 days after the first instillation. Intraocular pressure (IOP) was measured before instillation and at 15, 30, 60, 90, 120, 180 min, and 24 h after the first instillation. The measurement was performed three times at each time point with a rebound tonometer (iCare^®, Helsinki, Finland). Rabbits were weighed daily from the time before the experiment until 7 days after the first instillation (Fig. 1b). The humane endpoints included unmanageable severe pain or distress (e.g., repeated pawing/rubbing of the eye, excessive blinking/tearing, loss of appetite, inactivity, or weight loss of more than 20% compared to the baseline before the experiment) or presence of the following ocular conditions, e.g., corneal perforation or significant corneal ulceration including staphyloma, hyphema, grade 4 corneal opacity, absence of pupillary response to light lasting for 72 h, ulceration/necrosis/sloughing of the conjunctival/nictitating membrane¹¹.

Assessment of effect

To evaluate the effect of TNOS, the pupillary size was daily measured by photography with a calibration marker before NSS or TNOS administration and at 15, 30, 60, 90, 120, 180 min, and 24 h after instillation under constant controlled daylight illumination of the housing system (350 lx) (Fig. 1b). The camera model was Sony^® α6000 E-mount camera with an APS-C sensor, with the Sigma 70 mm f/2.8 DG Macro lens (Sony Corp., Bangkok, Thailand) and a Metz^® Mecablitz 15 MS-1 digital ring flash (Metz Consumer Electronics GmbH, Zirndorf, Germany). The distance between the camera and the rabbit’s eye was 30 cm, and the camera settings included an aperture of f/22, a shutter speed of 1/125, and an ISO of 100. The pupillary size was measured in millimeters by ImageJ (US National Institutes of Health, Bethesda, Maryland, USA).

All measurements were performed by a single neuro-ophthalmologist who was blinded to treatment applications to reduce bias. To control inter-day variability, all assessments were performed at a similar time of the day to avoid diurnal variation under the same controlled environment, and all rabbits were restrained using the same methods at all time points.

Statistical analysis

The sample size followed the OECD guidelines for the Testing of Chemicals regarding the in vivo eye irritation/serious eye damage. The guidelines recommended three albino rabbits per group for testing ocular irritation in an animal study¹¹.

To evaluate the sterility, the species of microorganisms would be reported, and samples with positive microbial contamination would be presented as frequencies with percentages.

To evaluate safety, ocular abnormalities and adverse events would be presented as frequencies and percentages. The IOP was calculated as mean difference (MD) from baseline with 95% confidence interval (95% CI) and analyzed with repeated measures two-way analysis of variance (ANOVA) with Greenhouse-Geisser correction. To evaluate the effect of TNOS, pupillary size was described as mean ± standard deviation (SD) and MD between each TNOS group and control with 95% CI over an observed period. Repeated measures two-way ANOVA with Greenhouse-Geisser correction was used to evaluate differences in pupillary size among various dose levels and control over the observed period. Tukey’s method was used for multiple comparisons among time points.

A p-value less than 0.05 was considered statistically significant. All analyses were conducted using GraphPad Prism version 10.0.0 (GraphPad Software, Boston, Massachusetts, USA).

Results

Sterility testing

The mean temperatures of the refrigerated control and room temperature groups were 4.9 °C (3.3 to 5.7) and 21.9 °C (21.0 to 23.0), respectively. No microorganism was detected in any sample across all conditions during the observed time points.

Animal characteristics

The mean ± SD of baseline body weight (BW) of all rabbits was 2.3 ± 0.1 kg. The BW of all rabbits increased throughout the experimental period without a significant difference in mean BW among groups at any stage. All outcome measurements were completed without any missing data.

Safety

No systemic or ophthalmic complication was observed in all rabbits. Vital signs were stable at every observed time point. The mean ± SD of overall baseline vital signs included PR 258.0 ± 25.6 beats/min, BP 113.7 ± 14.3/77.0 ± 11.2 mmHg, RR 41.9 ± 2.0 breaths/min, and SpO₂ 95.2 ± 3.2%, compared to vital signs at 60 min after TNOS instillation which included PR 242.4 ± 19.1 beats/min, BP 116.7 ± 12.9/74.9 ± 13.3 mmHg, RR 42.2 ± 2.2 breaths/min, and SpO₂ 96.4 ± 2.8%. Signs of ocular irritation, serious eye damage, or even mild or transient side effects were not detected throughout the experimental period. The ocular anterior segment was unremarkable, and no retinal break or retinal detachment was detected on the dilated fundus examination. In the 1.0 mg/mL group, IOP significantly decreased compared to baseline at 60 min (MD −1.78 mmHg, 95% CI −3.20 to −0.35, p = 0.020) and 120 min (MD −1.83 mmHg, 95% CI −3.47 to −0.19, p = 0.032) after instillation. In the 2.5 mg/mL group, IOP significantly decreased at 120 min (MD −1.78 mmHg, 95% CI −2.81 to −0.74, p = 0.005) and 180 min (MD −2.22 mmHg, 95% CI −4.13 to −0.32, p = 0.027). In the 1.5 mg/mL group, there was no significant change in IOP at any time point compared to baseline IOP. There was no significant difference in mean IOP among groups at the same time points. No rabbit showed any signs of severe pain, distress, or significant weight loss. Thus, after completing the experiment, all rabbits were transferred to the Faculty of Veterinary Science, Chulalongkorn University.

Effect of topical neostigmine ophthalmic solution on pupillary size

The baseline mean pupillary sizes of rabbits were comparable among all groups. The 1.5 and 2.5 mg/mL TNOS produced significant miosis with no significant difference between them, while the 1.0 mg/mL TNOS did not produce a significant effect. Pupillary sizes measured on the first day were analyzed and shown in Table 1. The mean pupillary sizes were significantly smaller compared to the control from 30 to 180 min in the 1.5 mg/mL group and from 60 to 90 min in the 2.5 mg/mL group. The maximal miotic effects compared to the control were observed at 90 min in the 1.5 mg/mL group (MD −2.58 mm, 95% CI −4.24 to −0.93, p = 0.006, degree of freedom (DF) 6.695) and at 60 min in the 2.5 mg/mL group (MD −3.13 mm, 95% CI −5.07 to −1.18, p = 0.005, DF 6.294). No statistically significant difference in pupillary size between 1.5 and 2.5 mg/mL groups was found at any time point. The maximal miotic effect was observed at 60 min in the 1.0 mg/mL group (MD −1.50 mm, 95% CI −3.34 to 0.35, p = 0.110, DF 6.438) without significant effect at any observed time point. Based on the findings, TNOS had an estimated onset of miosis at 15 to 30 min, peak effect at 60 to 90 min, duration of action until 180 min, and the miotic effect resolved at 24 h (Table 1; Figs. 2 and 3).

Table 1.

Mean pupillary size of the control, 1.0 mg/ml, 1.5 mg/ml, and 2.5 mg/ml neostigmine groups at each time point after instillation on the first day.

Time	Dose	Mean ± SD (mm)	TNOS vs. NSS		1.5 mg/mL vs. 2.5 mg/mL
			MD (95% CI) (mm)	Adjusted p-value	MD (95% CI) (mm)	Adjusted p-value
0 min	NSS (n = 6)	6.68 ± 0.47 6.72 ± 0.54 6.50 ± 0.24 6.62 ± 0.71
	1.0 mg/mL (n = 6)		0.04 (−0.85 to 0.95)	0.999
	1.5 mg/mL (n = 6)		−0.18 (−0.87 to 0.53)	0.851	−0.12 (−1.16 to 0.93)	0.982
	2.5 mg/mL (n = 6)		−0.06 (−1.15 to 1.03)	0.998
15 min	NSS (n = 6)	6.75 ± 0.14 5.89 ± 1.21 5.76 ± 0.84 5.49 ± 1.67
	1.0 mg/mL (n = 6)		−0.86 (−2.69 to 0.96)	0.397
	1.5 mg/mL (n = 6)		−0.99 (−2.25 to 0.27)	0.116	0.27 (−2.23 to 2.77)	0.983
	2.5 mg/mL (n = 6)		−1.26 (−3.78 to 1.26)	0.356
30 min	NSS (n = 6)	6.56 ± 0.23 5.23 ± 1.17 4.46 ± 1.34 4.26 ± 1.61
	1.0 mg/mL (n = 6)		−1.33 (−3.09 to 0.42)	0.128
	1.5 mg/mL (n = 6)		−2.10 (−4.11 to −0.09)	0.042*	0.20 (−2.43 to 2.83)	0.995
	2.5 mg/mL (n = 6)		−2.30 (−4.71 to 0.10)	0.059
60 min	NSS (n = 6)	6.34 ± 0.48 4.84 ± 1.24 3.81 ± 1.09 3.21 ± 1.31
	1.5 mg/mL (n = 6)
			−1.50 (−3.34 to 0.35)	0.110
	1.5 mg/mL (n = 6)		−2.53 (−4.15 to −0.91)	0.006*	0.60 (−1.55 to 2.74)	0.828
	2.5 mg/mL (n = 6)		−3.13 (−5.07 to −1.18)	0.005*
90 min	NSS (n = 6)	6.52 ± 0.47 5.07 ± 1.29 3.94 ± 1.12 4.22 ± 1.20
	1.0 mg/mL (n = 6)		−1.45 (−3.35 to 0.46)	0.137
	1.5 mg/mL (n = 6)		−2.58 (−4.24 to −0.93)	0.006*	−0.28 (−2.33 to 1.77)	0.974
	2.5 mg/mL (n = 6)		−2.30 (−4.08 to −0.53)	0.015*
120 min	NSS (n = 6)	6.28 ± 0.30 5.44 ± 0.94 4.68 ± 0.68 4.99 ± 0.89
	1.0 mg/mL (n = 6)		−0.84 (−2.22 to 0.56)	0.261
	1.5 mg/mL (n = 6)		−1.60 (−2.60 to −0.58)	0.005*	−0.31 (−1.72 to 1.12)	0.909
	2.5 mg/mL (n = 6)		−1.29 (−2.61 to 0.03)	0.055
180 min	NSS (n = 6)	6.39 ± 0.40 5.61 ± 0.89 5.18 ± 0.70 5.46 ± 0.69
	1.0 mg/mL (n = 6)		−0.78 (−2.11 to 0.54)	0.287
	1.5 mg/mL (n = 6)		−1.21 (−2.27 to −0.16)	0.026*	−0.28 (−1.51 to 0.94)	0.892
	2.5 mg/mL (n = 6)		−0.93 (−1.97 to 0.11)	0.081
24 h	NSS (n = 6)	6.59 ± 0.17 6.65 ± 0.52 6.57 ± 0.34 6.80 ± 0.49
	1.0 mg/mL (n = 6)		0.06 (−0.71 to 0.82)	0.994
	1.5 mg/mL (n = 6)		−0.02 (−0.53 to 0.48)	0.998	−0.23 (−0.99 to 0.53)	0.780
	2.5 mg/mL (n = 6)		0.21 (−0.52 to 0.93)	0.767

SD = standard deviation; TNOS = topical neostigmine ophthalmic solution; NSS = normal saline solution; MD = mean difference; 95% CI = 95% confidence interval. *Adjusted p-value < 0.05 (Tukey’s method for multiple comparisons).

Fig. 2.

Demonstrates mean pupillary size among the control, 1.0 mg/mL, 1.5 mg/mL, and 2.5 mg/mL topical neostigmine ophthalmic solution groups at 0, 15, 30, 60, 90, 120, 180 min, and 24 h after instillation on the first day. NSS = normal saline solution.

Fig. 3.

Demonstrates pupils of the rabbits before (the first row) and 60 min after (the second row) instillation of (a) normal saline solution, (b) 1.0 mg/mL, (c) 1.5 mg/mL, and (d) 2.5 mg/mL topical neostigmine ophthalmic solution.

The pupillary sizes measured on the second to seventh day were analyzed and demonstrated a similar trend (Fig. 4). However, the magnitude of pupillary constriction was maximum on the first day, with attenuation on subsequent days (Fig. 5).

Fig. 4.

Demonstrates mean pupillary size over the observed time points among the 1.0 mg/mL, 1.5 mg/mL, and 2.5 mg/mL topical neostigmine ophthalmic solution groups on the second to seventh day of instillation compared to NSS. NSS = normal saline solution.

Fig. 5.

Demonstrates mean pupillary size over the observed time points in the 7-day instillation of 1.0 mg/mL, 1.5 mg/mL, and 2.5 mg/mL topical neostigmine ophthalmic solution.

Discussion

There was no adverse event related to the TNOS application found in the study. This finding was consistent with the previous study in rabbits, which reported that no ocular inflammation was observed after instillation of 2.5 mg/mL neostigmine¹⁰. In our study, ocular examinations were performed at 1 h after instillation to evaluate immediate side effects and were daily examined for 7 days before escalating to the higher dose level to evaluate prolonged safety and effects of repeated administrations. The dilated fundus was examined for signs of retinal break or retinal detachment, and the IOP monitoring was done to detect possible side effects of cholinergic agents. Cholinergic agonists could reduce IOP by promoting ciliary muscle contraction and the subsequent miotic effect to increase aqueous outflow^12,13. The authors considered a change of mean IOP more than 2 mmHg as a clinically significant threshold. Although a significant IOP decrease was statistically found in some observed time points, mostly, the degree of change seemed clinically insignificant. The use of cholinergic agonists in some clinical settings, including exfoliation syndrome, phacomorphic glaucoma, and malignant glaucoma, may aggravate paradoxical angle closure and elevate IOP; however, IOP spikes were not observed in our study¹⁴. Nevertheless, potential angle closure risk should be carefully considered, especially in patients with the aforementioned conditions. The effect of TNOS on IOP in human studies should be further investigated due to species differences. In addition to ocular adverse effects, vital signs were also monitored due to the possible cardiopulmonary side effects of systemic neostigmine^3,6. The vital signs were stable throughout the experimental periods; thus, the authors speculated that TNOS was minimally absorbed into systemic blood flow and could minimize the risk of systemic side effects compared to intravenous and intranasal neostigmine, which have a bioavailability of nearly 100%^15,16. However, to confirm this speculation, pharmacokinetic measurements, e.g., plasma neostigmine levels or cholinesterase activity, and an initial human safety trial with careful cardiopulmonary monitoring, should be performed.

Significant pupillary constriction was observed in both 1.5 and 2.5 mg/mL groups, without a significant difference between these two groups. Pupillary constriction demonstrated the potential effect of local cholinesterase inhibition of TNOS. Considering clinical significance, the miotic effect was more than 2 mm compared to the control in both concentrations from 30 to 90 min, consistent with the significant miotic effect observed after eye instillation of 2.5 mg/mL neostigmine in a previous study¹⁰. However, a trial of other doses of neostigmine has not been performed, and details of the outcome were limited. The mean pupillary size was not demonstrated after 90 min and was reported to have no significant difference only at 24 h¹⁰. Our study extended the novelty and originality by testing multiple doses of TNOS (1.0, 1.5, and 2.5 mg/mL) and proposed the lower efficacious dose of 1.5 mg/mL. Moreover, this study also provided detailed safety and efficacy profiles. The mean pupillary size was measured in more frequent intervals until 180 min after instillation based on the pharmacologic properties of neostigmine, and at 24 h to evaluate the residual action. Significant pupillary constriction was observed at 30 to 180 min in the 1.5 mg/mL group. In comparison with intramuscular neostigmine, the onset of ptosis alleviation occurs within 15 min, with the peak effect at 30 min, and the duration of action lasts for several hours. Regarding clinical application, the recommended timing of diagnostic evaluation from our findings appeared similar to that of the intramuscular neostigmine test, where ptosis is reevaluated at 30 to 45 min¹⁷. The duration of action of TNOS may provide enough time for clinical evaluation in uncooperative patients, children, or patients with diplopia who require measurement of ocular alignment. This long duration may also be enough for use as a therapeutic purpose of MG. Nevertheless, several factors, including the pharmacokinetic barrier at the neuromuscular junction and the doses required, should be considered. The authors suggested further studies measuring drug levels in ocular tissues and functional tests, e.g., eyelid elevation following TNOS administration in experimental autoimmune myasthenia gravis (EAMG) models with ptosis, should be performed to prove this hypothesis. The pupillary size was repeatedly measured for 7 days to ensure the effect of TNOS on pupillary size and evaluate the effect of multiple applications. The findings showed a similar trend of mean pupillary size; however, the maximal degree of pupillary constriction occurred on the first day of instillation in all dose levels. A possible mechanism was desensitization of the cholinergic receptors on the iris sphincter muscle after TNOS administration. This hypothesis was supported by a previous study on pupillary size and light response after topical application of cholinomimetics. The miotic response to the second cholinomimetic application, 24 h following the first drug application, was reduced. An overshoot of the resting pupillary diameter and reduction of the light response remained after the recovery of miosis¹⁸. The effects on pupillary size or the levator palpebrae superioris muscle following repeated neostigmine administration have not been reported. Similar desensitization might occur with nicotinic AChR on the levator palpebrae superioris muscle; however, the diagnosis of MG usually requires only a single administration. The pupillary size in our study was measured from photography instead of direct measurement by a plastic ruler to increase accuracy and to prevent the sympathetic surge of rabbits from fright, which can cause pupillary dilatation.

According to the safety and preliminary potential efficacy of TNOS in this study, the findings suggested that the 1.5 mg/mL TNOS may be the lowest clinically efficacious dose for local cholinesterase inhibition without adverse effects. The authors suggested that 1.5 mg/mL TNOS could be a promising starting dose for further investigation in EAMG models or initial human safety trials. Given the cholinesterase-inhibiting effect of neostigmine, which could enhance both muscarinic and nicotinic AChRs⁶, the authors proposed that 1.5 mg/mL TNOS could potentially stimulate nicotinic AChR on the levator palpebrae superioris muscle. Sufficient drug penetration to produce the local effect of TNOS on the levator palpebrae superioris muscle was supported by the previous report of eyelid elevation after 2.5 mg/mL neostigmine instillation in MG patients¹⁰. Although pupillary constriction confirmed the local AChEI effect of 1.5 mg/mL TNOS on muscarinic AChR in the iris sphincter muscle, these AChRs are different subtypes and exist on different ocular muscles. Thus, it did not guarantee improvement of skeletal muscle strength in the levator palpebrae superioris muscle. Further studies in EAMG models or pilot dose-escalation studies in ocular MG patients with careful cardiopulmonary and ocular monitoring should be performed to investigate the effect of TNOS on ptosis alleviation.

The sterility testing suggested that TNOS could be stored at either room temperature or in the refrigerator for at least 28 days without preservatives, even in the simulated patient use group. However, the authors primarily aimed to develop TNOS as a diagnostic tool; thus, single instillation of TNOS was used immediately after dilution. Considering using TNOS as a therapeutic agent or diagnostic test for multiple patients in certain periods, the diluted drug should be kept for multi-day use. In such situations, several factors, e.g., prolonged environmental exposure, temperature changes, and user contamination, could affect sterility, and preservatives may be required. Moreover, stability should be investigated because the stability of parenteral neostigmine cannot be guaranteed after dilution.

Despite a high proportion of MG patients having isolated ocular symptoms, to the authors’ knowledge, no topical ophthalmic drop directly alleviates myasthenic ptosis in current practice, which may be potentially applied as a diagnostic tool^1,12,19–21. Current diagnostic methods still have some limitations regarding sensitivity, specificity, safety, tolerability, and availability. Systemic AChEI tests, despite good diagnostic performance, carry the risk of systemic side effects and the complexity of the procedure. Serum AChR-Ab has low sensitivity in ocular MG patients. Although a novel method of AChR-Ab test using live cell-based assay increases sensitivity, it is time-consuming and not widely available. SF-EMG requires patient’s tolerability and provider’s expertise. The potential strengths of using TNOS as a diagnostic tool over using other current diagnostic methods are a safer profile and less invasiveness^3–5,22. While previous studies reported ptosis alleviation after the use of topical alpha-adrenergic agonists stimulating sympathetically innervated Müller muscle, which could elevate the upper eyelid only 1 to 2 mm^23–25. The authors proposed a further study to evaluate the effect of TNOS on ptosis alleviation due to several reasons. Firstly, given the cholinergic effect of neostigmine on the interrupted levator palpebrae superioris, which has a more prominent effect on eyelid elevation than the Müller muscle^19,25, we hypothesized that it could elevate the eyelid in MG patients greater than the alpha-adrenergic agonists. Secondly, the beneficial properties of neostigmine over other AChEIs are its solution form, which is easy to prepare as an eyedrop formulation. Additionally, a longer duration of action than edrophonium is appropriate for clinical evaluation, and it does not cross the blood-brain barrier, unlike physostigmine. Lastly, commercial parenteral neostigmine is a widely available drug currently used in general anesthesia⁶.

The potential strength of this study is that, to the authors’ knowledge, this was the first dose-finding study of topical neostigmine in animal models. The present study provided a safety profile and compared the effects of TNOS among various dosages. This suggested the optimal dosage for use as a starting dose in further human studies to avoid too many patient exposures to subtherapeutic doses while preserving safety. The criteria of adverse events for early termination of the experiment were clearly defined. The effect of TNOS was observed in frequent intervals and longer duration after each drug administration. Thus, the onset, peak effect, and duration of action could be estimated from this study. Moreover, the outcomes were measured in prolonged multiple applications.

The first limitation of this study is that pupillary constriction demonstrated local cholinesterase inhibition but did not directly measure the neuromuscular effect on eyelid muscles, since the primary objective of this study is to evaluate the safety and pharmacodynamics in preclinical animal models prior to human clinical studies. Second, although the sample size was sufficient for assessing safety and irritancy, the study might have had insufficient power to detect subtle pharmacodynamic differences and the potential for pseudoreplication. Lastly, drug formulation parameters associated with patient tolerability, e.g., pH and osmolarity, and pharmacokinetic data to confirm systemic absorption or tissue concentration should be further investigated.

No systemic or ocular adverse effects were observed in all dose levels of TNOS. The 1.5 and 2.5 mg/mL TNOS significantly caused pupillary constriction in rabbits without a significant difference between groups. Based on the findings, 1.5 mg/mL was the lowest concentration that provided significant pupillary constriction without observable adverse effects in healthy rabbits and was considered a candidate starting dose for future studies. This incremental advancement laid the groundwork for further investigations to validate the effect of TNOS on ptosis alleviation in disease models and human studies.

Author contributions

All authors conceptualized and designed the experiment. CO, KR, and KM conducted the experiment. CO and YC performed the data analysis. CO drafted the original manuscript and prepared all figures. CO, SJ, SA, and YC revised the manuscript. All authors approved the submitted manuscript.

Funding

The 90th Anniversary of Chulalongkorn University Scholarship, Ratchadapiseksompotch Fund

Ratchadapiseksompotch Fund, Faculty of Medicine, Chulalongkorn University, grant RA67/030.

Data availability

Data is provided within the manuscript.

Declarations

Competing interests

The authors declare no competing interests.