Acceptability of Atropine Eyedrops Administered Sublingually for Sialorrhea Treatment Related to Neurological Conditions

ABSTRACT

Off‐label use of anticholinergic agents (atropine eye drops) administered sublingually are a first‐line treatment in standard clinical practice in France to treat sialorrhea in patients with neurological conditions. The ability and willingness to using and administering such medication have become key factors to ensure safe and effective therapy. Given that the critical aspects for ophthalmic product development differ from those for oral medicine, the objectives of this study were to investigate patient acceptability of atropine eye drops for treating sialorrhea. A multi‐centric, cross‐sectional study using the CAST—ClinSearch Acceptability Score Test methodology was conducted in France between February 2021 and April 2023. In total, 31 evaluations were collected. Most patients were males (74.2%) and the mean age was 51.6 ± 20 years. Poor acceptability was reported in real‐life settings. A lack of a suitable administration device combined with excessive saliva flow and patient disabilities made it difficult to ensure the required dose intake. Due to the bitter taste of the drug, poor palatability of the product appeared to be a key concern for oral administration. Designing a suitable form of atropine in accordance with the specificities of patients with neurological disabilities is needed to ensure the effective treatment of sialorrhea.

1. Introduction

Sialorrhea is a discomfiting and indisposed condition characterized by an unintended excess flow of saliva. Sialorrhea in patients with neurological conditions is most commonly caused by impairment of being able to swallow, due to a lack of coordination in orofacial and palate‐lingual musculature, which is a frequent concern for patients with neurologically disabilities, suffering from Amyotrophic Lateral Sclerosis, Parkinson disease, acquired brain injury, stroke or cerebral palsy. The prevalence of sialorrhea in patients with neurological disabilities varies between 30% and 70% in adults and 37% and 58% in children depending on condition [1, 2, 3, 4]. Although the exact mechanism of hyper‐salivation is still unknown, sialorrhea can also be induced by the use of psychotropic agents as commonly reported in 80% of patients treated with clozapine [5]. Sialorrhea may lead to psychosocial and clinical complications, such as skin irritations, perioral infection and/or life‐threatening aspiration pneumoniae, with a high impact on the daily life of patients and their families [1, 6, 7, 8, 9, 10].

Assessment of sialorrhea is usually carried out with a multidisciplinary approach undertaken by rehabilitation doctors, neurologists, speech therapists, and physiotherapists, and also by patients and their families. Its frequency (constant, frequent, or occasional) and severity can be assessed by using standardized questionnaires such as the Drooling Severity and Frequency Scale, as well as by the number of changes of clothing/towels per day [11, 12]. Reducing sialorrhea can significantly improve patient quality of life, however, treatment remains challenging [13]. Behavioral therapies and motor rehabilitation should be used for treatment. If the results are not satisfactory, pharmacological therapies based on botulinum neurotoxin A (BoNT/A) and/or off‐label use of anticholinergic agents have been described in literature [1, 11]. In France, BoNT/A is indicated for the symptomatic treatment of chronic sialorrhea in adults with neurological conditions related to Parkinson's disease, brain injury, or stroke. However, its use remains limited since it should only be administered by experienced, qualified physicians. Additionally, BoNT/A injections can lead to potential side‐effects, especially dysphagia, eyelid ptosis, dysarthria and hemorrhagic risk in patients receiving antithrombotic drugs such as antiplatelets or anticoagulants [14].

Despite clinical guidelines, the management of antithrombotic treatments may be difficult in such a context since they require either a therapeutic relay or a window to limit the hemorrhagic risk which, conversely, may give rise to potential thrombotic risks that would limit their use. In any case, BoNT/A remains a second line treatment option after anticholinergic agents [15]. The use of anticholinergic agents, available either as a transdermal patch of scopolamine or atropine eye drops administered sublingually constitute first‐line treatment options for sialorrhea [1, 8, 11, 15, 16]. For both pharmacological and practical reasons such as a short elimination half‐life (allowing easy titration), strength and stability, currently marketed atropine 1% eye drops, packaged as 0.4 mL single‐dose container, are mostly used in hospital and ambulatory settings. In clinical practice, atropine is usually titrated drop by drop (one drop contains 250 μg of active substance) to achieve a daily dose of 250 μg–2000 μg administered one to three times per day. Nevertheless, such off‐label use can lead to potential mishaps such as medication administration errors and/or over‐ or under‐dosing as reported in literature [16, 17, 18, 19].

Sialorrhea is observed in adults and children with neurological disabilities and is usually associated with swallowing disorders. The negative impact of swallowing issues concerning oral medicine is well known [20]. The acceptability of medicine (the overall ability and willingness of patients as well as caregivers to administer the medicine as intended [21]) has emerged as a key factor in such populations to ensure patient adherence and consequently, treatment safety and efficacy. Beyond, the patients' ability to swallow, there are many characteristics which determine acceptability of oral medicine, such as palatability (the overall appreciation in relation to smell, taste, aftertaste and feeling in the mouth [21]) and ease of administration. Moreover, there are different critical attributes for ocular medicine development, such as local tolerance, feeling after administration (impact on vision) or usability of the delivery device for eye drops [22]. Therefore, sublingual administration of eye drops may give rise to acceptability issues, because dosage delivery of this type has not been designed for oral administration.

In this original study, we used the CAST—ClinSearch Acceptability Score Test methodology to investigate the acceptability of atropine 1% eye drops used as first‐line treatment of sialorrhea in adult patients with neurological conditions (such as Parkinson's, motor neuron disease, stroke or acquired brain injury). The CAST methodology is made of the following elements: (i) a questionnaire to collect standardized evaluations for medicine intakes; (ii) an acceptability reference framework based on a large set of real‐life observer‐reported outcomes previously collected for many medicine intakes; (iii) a multivariate data analysis process to score acceptability [23]. The objective was to gain insights into this off‐label use to help clinicians in their medical practice.

2. Methods

2.1. Study Design and Setting

An observational, multicentric, cross‐sectional study was conducted in the physical medicine and rehabilitation ward (including the post‐intensive care rehabilitation unit and the long‐term complex care unit) and in the geriatric medicine ward at Raymond Poincaré and Sainte‐Périne Hospitals (AP‐HP, Paris Saclay University) in France between February 2021 and April 2023.

2.2. Inclusion Criteria and Sample Size

Adult patients ≥ 18 years old, residing in France, being treated for sialorrhea, secondary to neurological condition only, with sublingual delivery of marketed atropine 1% eye drops, packaged as 0.4 mL single‐dose container, were eligible. Participants were included on a voluntary basis without any randomization. Evaluations of the drug delivery method in a minimum of 30 patients were targeted to yield an acceptability score using the CAST methodology.

2.3. Data Collection

Acceptability was assessed during the first administration of atropine (single dose) once enrolled in the study. Patients could be on different atropine dosages depending on their clinical response, as a result, the prescribed dose could represent either one or two drops. Acceptability evaluation was carried out once per patient and there was no follow‐up.

The trained pharmacist observing the administration, recorded observable events or behaviors that would describe many aspects of acceptability using a standardized questionnaire: (i) an intake/required dose report (fully, partly, or not taken at all), (ii) a 3‐point facial hedonic scale (positive, neutral, or negative reaction) to describe the patients' reaction during administration, and (iii) needed time by hospital staff to prepare the prescribed dose of medication (from opening the packaging to having the required dose of medication ready to use including any modifications and handling) as well as to administer the prescribed dose of medication (from the required ready‐to‐use dose to the intake completion). In addition, methods used to ease and/or achieve administration was reported. These resulted in five binary variables whether the method was used or not. The variables were: (i) dividing the intake of a prescribed dose which could not be taken as a whole (such as successive sips of a liquid preparation), (ii) altering the intended use of the prescribed dose of medication, in which the dose was modified before administration (such as tablet being crushed into powder) or the dose had to be administered in other ways than those provided in the leaflet (such as the oral administration of an ocular solution), (iii) using food and/or drink, such as mixing the dose with food or drink (with a spoon of applesauce or a glass of apple juice for example) or having a patient drink or eat just before or after administration to mask the drug taste/for easier swallowing, (iv) using a device that was not provided with the medication such as an oral administration syringe provided by another medication, and (v) using restraint in which a patient had to be made to take it.

For each method reported, the pharmacist was requested to specifically write any further information. For example, which drink and/or food was taken, the quantity, and the reasons for the choices made. Since CAST is a generic assessment tool for any medication, some methods could be inapplicable. Other unspecified methods used to help with taking the medication were to be described in a text field.

In addition to observer‐reported outcomes, patient difficulties in sticking out their tongue and opening their mouth were recorded. Furthermore, information on the patient (age, sex, and type of neurological disability) was collected.

2.4. Data Handling

Following data collection, the variables were recoded and transformed when necessary. The sum of times for preparation and administration was made into a categorical variable (preparation and administration time) having three categories classified as short (≤ 20″), medium (from 30″ to 1′), or long (> 1′). This data transformation was based on the distribution of data and the expertise of the research team developing the CAST methodology.

Based on qualitative data, data were recorded during data review when the description in the text field clearly suggested the use of a wrong response option or a non‐selected method.

2.5. Statistical Analysis

For the acceptability score, we used the acceptability reference framework built on 2091 evaluations in elderly people (Figure 1). These evaluations were previously collected to build this reference framework using multivariate analysis [23, 24, 25, 26, 27].

FIGURE 1.

Acceptability reference framework.

First, a multiple correspondence analysis (MCA) summarized the variability between acceptability measures and their mutual key relationships via a three‐dimensional acceptability map (3D‐map) through which resulting dimensions revealed any associations and dissociations of the observed measures that would contribute to explain data variability. Applying dilations to both evaluations (initially positioned at the barycentre of the nine related observed measures) and observed measures (initially at the barycentre of all the evaluations including it) representations, the 3D‐map is a simultaneous optimal representation. The interpretation of the 3D‐map was based on the distance between the elements and thus proximities expressed similarities. Observed measures closed on the map were often selected together in the evaluations. Similarly, evaluations completed in a comparable manner converged on the map.

Subsequently, a hierarchical clustering on principal components and k‐means consolidation grouped the most similar evaluations into two distinct clusters. The clusters were described by the categories significantly overrepresented in their subset of evaluations compared to a random distribution (v‐test value greater than 1.96 and p < 0.05). The “positive” observed measures appeared to be overrepresented in the first cluster which was therefore defined as the “positively accepted” profile (green area), while all “negative” measures were overrepresented in the second cluster defined as the “negatively accepted” profile (red area).

The evaluations collected were positioned on the reference framework. There was no influence on the reference framework, which allowed scoring process implementation. Atropine 1% eye drops were positioned on the map at the barycentre of the evaluations collected; confidence ellipses surrounding the barycentre for all dimension pairs defined an area containing its true position with a 90% probability. Lastly, the medication was classified as positively accepted if the barycentre, with the confidence ellipses surrounding it, belonged to the green area of the reference framework, or not accepted if the barycentre or a part of the confidence ellipses belonged to the red area of the reference framework.

3. Results

3.1. Population

In total, 31 evaluations of atropine 1% eye drop administration were collected. Among the patients included, 23 (74.2%) were men and the mean age was 51.6 ± 20 years (range 18 to 87) (Table 1). Eight patients (25.8%) were 65 years of age or older. Twenty‐seven patients (87.1%) were already being treated with atropine 1% eye drops for sialorrhea prior to inclusion. Thirty patients had swallowing disorders (96.8%), 8 (25.8%) had difficulty opening their mouths and 14 (45.2%) had difficulty sticking their tongues out.

TABLE 1.

Patient characteristics.

Characteristics	N = 31
Age (Years ± Standard Deviation)	51.6 ± 20
Sex ratio (male/female)	2.87
Type of neurological disability
Stroke	18 (58.1%)
Amyotrophic Lateral Sclerosis	4 (12.9%)
Spinal injury	2 (6.5%)
Traumatic brain injury	2 (6.5%)
Acute inflammatory polyradiculoneuropathy	2 (6.5%)
Parkinson and related diseases	2 (6.5%)
Multiple Sclerosis	1 (3.2%)
Impairment
Swallowing disorders	30 (96.8%)
Difficulties for opening the mouth	8 (25.8%)
Difficulties for sticking out the tongue	14 (45.2%)

3.2. Acceptability

According to the acceptability reference framework, atropine 1% eye drops could not be considered as positively accepted (Figure 2). Although there were evaluations positioned in the positive (green) area, the barycentre of total number of evaluations along with 76.0% of the confidence ellipses surrounding it were located in the negative (red) area of the reference framework.

FIGURE 2.

Acceptability of atropine 1% eye drops as treatment for sialorrhea for patients with neurological conditions in France.

The position of the medication on the reference framework reflects the reported measures. Each evaluation corresponded to a combination of an observed measure for the eight variables. A particular combination of observed measures reflected the behavior of users. There were 14 specific combinations of observed outcomes represented by dots (with size proportional to use) on the reference framework (Figure 2).

The following combination (“fully taken”, “neutral reaction”, “short time duration”, “no divided dose”, “no food and/or drink”, “used alteration”, “no extra device” and “no restraint”) reflected an intended medicine use without much problems however with sublingual administration of the ocular solution, was the most used (29.0%). Ten combinations (71.0%) corresponding to 16 evaluations (51.6%) were located in the negative/red area. This was the case for the second most used combination (16.1%), which was the same as the first, with the category “negative reaction” replacing “neutral reaction”. This combination was the closest to the positive/green area. The following combination was the furthest from the positive/green area, and was only used once (3.0%): “fully taken”, “negative reaction”, “long time”, “no divided dose”, “no food and/or drink”, “used alteration”, “no extra device”, and “used restraint”.

Considering each observational variable, we were able to highlight the sublingual administration of the ocular solution for all the patients (Table 2). Such alteration of the intended use of the product negatively impacted the acceptability score. In addition, a poor acceptability score could have been due to the number of negative reactions during the dose administration, which was non‐negligible (38.7%).

TABLE 2.

Observer‐reported outcomes for the acceptability of atropine eyedrops administered sublingually for sialorrhea treatment.

Observer reported outcomes	N = 31
Result intake
Fully taken	25 (80.6)
Partly taken	6 (19.4)
Not taken	0 (0)
Patient reaction
Positive	4 (12.9)
Neutral	15 (48.4)
Negative	12 (38.7)
Preparation and administration time
Short	18 (58.1)
Medium	12 (38.7)
Long	1 (3.2)
Divided dose
No	30 (96.8)
Yes	1 (3.2)
Food and/or drink
No	31 (100.0)
Yes	0 (0.0)
Alteration
No	0 (0.0)
Yes	31 (100.0)
Extra device
No	31 (100.0)
Yes	0 (0.0)
Restraint
No	28 (90.3)
Yes	3 (9.7)

Considering the qualitative information from the text field, six comments highlighted the poor/bitter taste of the drug for the patient. Although most atropine 1% eye drops were fully taken (80.6%), ten comments indicated the difficulty of validating the full administration of the required dose. Indeed, healthcare professionals could have had to “blindly” place a colorless drop under the tongue of patients for whom the sublingual route was difficult to access (such as trismus, difficulty in opening the mouth, or sticking out the tongue). Substantial doubt about the administration of the required dose led us to consider the dose as partly taken. Container usability concerns and the inability to deliver an accurate dose were also indicated in the written feedback.

4. Discussion

In this study, the CAST methodology was used to investigate the acceptability of atropine 1% eye drops, packaged in a 0.4 mL single‐dose container, for the treatment of sialorrhea in patients with neurological conditions. This methodology was developed in pediatrics [28, 29], then transposed to the older population [23], and has previously received a letter of support from the European Medicine Agency [30]. Although most of the study participants were under 65 years of age, the elderly acceptability reference framework was used due to the deteriorated health status of patients with disabling symptoms caused by neurological conditions.

Our findings show that atropine 1% eye drops may not be considered as being positively accepted by patients with neurological disabilities treated for sialorrhea in real‐life conditions. Indeed, the barycentre, along with 76% of confidence ellipses, belonged to the negative area of the acceptability reference framework. This sub‐optimal acceptability was mainly due to the alteration of the intended use of this ocular product, its poor palatability for oral administration, the difficulty of ensuring the full administration of the required dose, and container usability concerns. Off‐label use of this eyedrop in clinical settings due to a lack of appropriate dosages marketed for treating this specific population may explain these findings. Suitable products should be designed to meet patient needs and to ensure the product's intended performance [31].

Although atropine formulated as tablets is marketed in a few countries as an antispasmodic, this format of the medication is not currently available in France. Moreover, conventional tablets may not be optimal for treating patients with swallowing disorders [23, 26]. Sublingual atropine ophthalmic solution is currently the only standardized dosage form available for the treatment of sialorrhea in the absence of an appropriate marketed product. However, the requirements and guidelines for developing an ophthalmic dosage form are very different from those for oral dosage, especially in terms of dose control.

Ophthalmic solutions are available for all dose administrations, whether they are multidose or single‐dose, and come in a range of plastic bottle varieties that deliver drops with variable volumes [32, 33]. The volume and number of drops delivered may depend on the pressure exerted on the bottle and the angle at which the bottle is held. For atropine, one drop is approximately 25 μL containing 250 μg of active substance, however, it can vary. Although, one to four drops (250 μg–1000 μg) are required, the 0.4 mL single‐dose container comprises around 16 drops. In this study, some difficulties using the container for sublingual administration were observed. There may be a need to have an administration device adapted to a specific patient profile (such as in trismus, spasticity, apraxia, and/or uncooperative patients). The lack of suitable administrative devices combined with excessive saliva flow and bucco‐facial apraxia may make it difficult to take the required dose of the colorless liquid solution. These doubts led us to consider a required dose being as only partly taken.

Frequent administrations (one to three times a day) of low doses (1 to 2 drops (250 μg–500 μg)) appear to be quite difficult to achieve in routine clinical practice. The amount of atropine taken may therefore vary from one administration to another. Consequently, such off‐label sublingual use can lead to an increased risk of over‐ or under‐dosage. Two cases of adverse events were recently reported by Michelon et al. (2021) [17] which could highlight the potential risks associated with prescribing atropine 1% eye drops for the treatment of sialorrhea. A post hoc review of the French pharmacovigilance database of all cases reported up to November 2023 revealed that a total 13 cases of adverse events were reported with the use of atropine 1% eye drops administered sublingually. They included six overdoses (incorrect administration) and five prescription errors.

The acceptability of medication remains a complex concept determined by the characteristics of not only the product but also that of the patient [21, 34]. Beyond ease of administration (pharmaceutical form and device), palatability is considered a key attribute for oral medicine acceptability, however cannot be applied for ocular medicines [22]. Atropine in an aqueous solution is known to have a bitter taste [5]. Our study confirms that bad taste may affect acceptability. One nurse pointed out that a patient's grimace was a sign that the atropine drop was actually present under the tongue. Poor palatability of atropine solution designed without taste masking (addition of sweetener or flavor) negatively impacts acceptability. It may be worth noting that oral prescriptions of ocular products could be confusing for nurses, possibly resulting in potential errors with inadvertent administration via the ophthalmic route [18, 19].

Regarding limitations, although the threshold to get an acceptability score using the CAST methodology (a minimum of 30 evaluations) was reached, there was a level of heterogeneity in the sample that should be considered when interpreting the results. The sample size did not permit any exploration of subgroups (a minimum of 30 evaluations per sub‐group to compare acceptability scores). The impact of variable parameters in our sample such as sex, age, type of neurological disability, impairments (i.e., swallowing disorders, difficulties for opening the mouth or difficulties for sticking out the tongue) or previous exposure to treatment could not be studied, which would be of interest to explore in the future. Effectively, the influence of some patient characteristics on acceptability may have been established, in particular by using the CAST methodology. A study by Vallet et al. [35] focusing on antibiotics in pediatrics emphasized the multidimensional nature of acceptability. Other studies have highlighted that sex could be an important factor, since a gap in acceptability between women and men driven by palatability exists [24, 25]. Potential higher sensitivity of under‐represented females to the unpalatable formulation could have had an impact on the acceptability score. Furthermore, this cross‐sectional study does not allow for exploration of changes in acceptability over time for this treatment, which is mainly used as a long‐term therapy.

Our findings suggest that atropine eye drops administered sublingually could not be considered as acceptable for treatment of sialorrhea in patients with disabling neurological conditions. Such off‐label use could expose patients to challenges in adherence and potential administration errors. Despite these limitations, this formulation remains the first‐line option for sialorrhea in routine clinical practice in France. There is therefore a clear need for a dedicated dosage form that ensures standardized dosing, supports safe administration tailored to patients' disabilities and clinical conditions, and enhances overall acceptability—particularly regarding palatability. To meet this unmet need, a new formulation is currently under development: a viscous oral solution of atropine sulfate at 2.5 mg/mL. Improvement in acceptability will require rigorous formal validation. The CAST methodology is particularly appropriate for this purpose, as it enables relative evaluations within the reference framework. The acceptability of the new formulation, assessed under comparable conditions in an upcoming study, can thus be directly compared with that of the atropine 1% eye drops. This forthcoming work may also offer an opportunity to investigate the determinants of acceptability.

Author Contributions

Conceptualization, H.M. and T.V.; methodology, F.R. and T.V.; software, T.V.; formal analysis, T.V.; investigation, H.M. and R.G.; resources, C.L.‐D., J.P. and V.B.; data cu‐ration, H.M. and T.V.; writing – original draft preparation, H.M. and T.V.; writing – review and editing, H.M., C.L.‐D., J.P., R.G., V.B., F.R. and T.V.; visualization, T.V.; supervision, H.M., C.L.‐D., J.P.; project administration, H.M. All authors have read and agreed to the published version of the manuscript.

Funding

The authors have nothing to report.

Ethics Statement

The data presented in this article were collected and analyzed according to the Accept65 protocol. Approvals were obtained from the French advisory committee for data processing in health research and the data protection authority. An exemption from the ethics committee has been issued. Although there was a protocol deviation on subject age, as some patients were under 65 years old, it was considered as minor, as the deteriorated health status of included patients is quite homogenous and not significantly affected by age.

Consent

The authors have nothing to report.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Data S1: AtropineStudy_Dataset.

Data Availability Statement

The data set generated during the study is available in Supporting Information (Table S1).