Secondary Logo

Journal Logo

Original Contribution

Gabapentin and Memantine for Treatment of Acquired Pendular Nystagmus: Effects on Visual Outcomes

Nerrant, Elodie MD; Abouaf, Lucie MD; Pollet-Villard, Frédéric MD; Vie, Anne-Laure MD; Vukusic, Sandra MD, PhD; Berthiller, Julien MSc; Colombet, Bettina PharmD; Vighetto, Alain MD; Tilikete, Caroline MD, PhD

Author Information
Journal of Neuro-Ophthalmology: June 2020 - Volume 40 - Issue 2 - p 198-206
doi: 10.1097/WNO.0000000000000807


Acquired nystagmus is a disabling condition resulting in oscillopsia, decreased visual acuity, poor vision-specific quality of life, and reduced social functioning scores (1,2). Pendular nystagmus is characterized by a to-and-fro eye oscillation without the resetting quick phases that differentiate it from jerk nystagmus. The most common causes of acquired pendular nystagmus (APN) are multiple sclerosis (MS) and oculopalatal tremor (OPT) (3). APN in symptomatic OPT is most frequently observed after brainstem or cerebellar lesions. We have previously investigated the functional consequences of APN in the setting of both MS and OPT and have found deterioration of vision-specific health-related quality of life scores in both patient groups (4).

Based on pathophysiological hypotheses, pharmacological treatments for APN have been examined in treatment trials, leading to the proposal of gabapentin or memantine as effective agents (5–9). In these studies, the efficacy of memantine and gabapentin was based on objective ocular motor measures, including nystagmus amplitude and/or velocity. Nevertheless, except for visual acuity that was tested in the study by Thurtell et al (9), functional consequences of nystagmus, including oscillopsia and visual quality of life outcomes, were not evaluated. This represents an important limitation in treatment trials to date for APN.

This cross-over trial evaluated the effects of gabapentin and memantine on APN in patients with MS and OPT. The primary objective was to determine the effect of each treatment on ocular motor and visual functional measures, including visual acuity, subjective assessment of oscillopsia amplitude and direction, and vision-specific quality of life. The secondary aims were to evaluate the association of objective baseline ocular motor parameters and functional measures, as well as the safety and the tolerability of each treatment.


Study Design

We conducted a single-center, open-labeled, controlled cross-over trial to test the effects of gabapentin and memantine in patients with APN. Patients were examined and evaluated before visits 1 and 3, and during visits 2 and 4 while on treatment with each drug as shown in the study flow chart (Fig. 1). All included patients were contacted by phone after 30–37 days following the last posttreatment visit (visit 5).

FIG. 1.:
Flowchart summarizing study design. MS, multiple sclerosis; OPT, oculopalatal tremor.

Study Cohort: Inclusion/Exclusion Criteria

We prospectively included patients with MS and OPT who had pendular nystagmus. All patients presented with chronic APN for at least 6 months (confirmed by physician examination) due to MS or symptomatic OPT. MS was diagnosed according to the McDonald revised 2010 criteria (10). All patients with MS were tested at least 3 months after an acute relapse and/or corticosteroid course. The diagnosis of symptomatic OPT was based on the presence of both pendular nystagmus and synchronous palatal tremor following a focal brainstem or cerebellar lesion; these findings had to be associated with a degenerative olivary hypersignal on FLAIR or T2 MRI. Patients with other ophthalmological disorders that could impair vision, or those with ongoing seizure, severe neurological disability, psychiatric disorder, or other contraindication to gabapentin or memantine therapy, were excluded.

All patients were informed about the design and the purpose of the study. Patients provided informed, written consent to the protocol and study procedures. Ethical approval was received from the National French ethical committee on human experimentation (eudract: 2012-003204-12 and 2014-005548-17), in agreement with French law (March 4, 2002) and the Declaration of Helsinki. The study was registered in a public trial registry (ClinicalTrials: NCT01744444).


After baseline evaluation, patients were randomly assigned with a 1:1 ratio to start either memantine or gabapentin (Fig. 1). In accordance with the Thurtell et al (9) method, we used gabapentin 300 mg 4 times daily and memantine 10 mg 4 times daily with a titration phase of 9 days to reach the full dosages. Visits 2 and 4 were performed when patients had been on full dosage treatment for at least 8 days. A washout period of 30–37 days was observed between the 2 treatment periods. At visit 4, patients chose which treatment to be continued. Both memantine and gabapentin were open-labeled for their use in the study.

Data Collection/Follow-up

The following tests were performed on visits 1 to 4. There was no standardization of the timing of data collection relative to the timing of treatment dosing. The patients were tested in the afternoon between 2 and 4 pm, at least 2 hours after drug dosing.

Evaluation of Functional Consequences of Nystagmus

Clinical Evaluation

Patients underwent a complete ophthalmic examination, including best-corrected distance and near visual acuities, ophthalmoscopy, and complete neurologic, neuro-ophthalmologic, and neuro-otologic examinations. Distance visual acuity measures were expressed in logarithm of the minimum angle of resolution (LogMAR), evaluated on the Early Treatment Diabetic Retinopathy Study scale. Near visual acuity measures were evaluated on the Parinaud scale and then converted into LogMAR. Side effects were ascertained by direct questioning of the participants.

Oscillopsia Measurements

Patients estimated both direction and amplitude of their oscillopsia while viewing a stationary target at distance (5 m) and near (57 cm) locations, with best visual correction in photopic conditions. The oscillopsia measure was obtained binocularly in case of conjugate nystagmus and monocularly of the worst eye in case of disconjugate or monocular nystagmus. Results were expressed in degrees (°). A more precise description of oscillopsia measurements is given in Supplemental Digital Content (see Supplement 1,

Vision-Specific Quality of Life Questionnaire

To examine vision-specific health-related quality of life, we used the 25-Item National Eye Institute Visual Functioning Questionnaire (NEI-VFQ-25) (11).

Objective Nystagmus Parameters

Eye Movement Recordings

Eye movements were recorded in darkness using an infrared video camera mounted on a light tight mask placed in front of the right or the left recorded eye (Synapsys, Marseille, France). A more precise description of eye movement recordings is given in Supplemental Digital Content (see Supplement 1,

Eye Movement Analysis

Based on the movement of each eye position projected in 3 planes, the following values were extracted from the 3D recording (Fig. 2):

  • The dominant plane (horizontal [H], vertical [V], or torsional [T]) as the plane in which the more regular and/or the largest amplitude of eye oscillation was measured for each eye.
  • The mean amplitude, mean frequency, mean intensity, and the mean peak velocity were calculated on the best 10-second period of nystagmus in the defined dominant plane. Intensity of nystagmus was defined as a multiple of the frequency by the amplitude.

FIG. 2.:
3D infrared video-oculography recording (25-Hz frequency) example of the right eye of an OPT patient. Based on this recording, the following values were extracted: the dominant plane (here torsional), the mean amplitude, mean frequency, mean intensity, and the mean peak velocity on the best 10-second period of nystagmus. OPT, oculopalatal tremor.


All statistics were performed by the STATISTICA software package (Statistica 9, Statsoft Inc, 1984–2010).

Wilcoxon signed-rank tests for paired data were performed to compare pretreatment and posttreatment parameters in each treatment group. Spearman correlations were calculated between baseline ocular motor measurements (amplitude, velocity, and frequency) and baseline visual functional parameters (distance visual acuity, near and distance oscillopsia). In patients with asymmetric binocular nystagmus, measurements from the eye with the worse nystagmus were correlated with oscillopsia and with NEI-VFQ-25 scores.


Demographic Data

The study group included 9 men and 7 women with an average age of 43 (36–55) years. Details of patients studied are summarized in Table 1. Examination disclosed pendular nystagmus, mainly in a horizontal or vertical direction, which was clinically symmetric in 4 patients, asymmetric in 9 patients, and monocular in 3 patients. A total of 29 eyes with nystagmus were observed. Ten (62.5%) patients had MS, and 6 (37.5%) presented with OPT after brainstem hemorrhage or surgery for brainstem cavernoma. The average nystagmus duration was 7 (1.5–13) years. Signs of chronic optic neuropathy (based on reduced visual acuity, optic disc pallor, and corresponding visual field defects) were observed in 70% of patients with MS.

Demographic and clinical data
Demographic and clinical data

Baseline Clinical Examination and Eye Movement Recordings

Figure 3 shows examples of eye movement recordings before and during both treatments for a patient with MS (Subject number 1) and a patient with OPT (Subject number 14). We will refer to preG for baseline evaluations during the gabapentin trials and preM for evaluations during the memantine trials. Baseline values are presented in Table 2. Worse baseline near monocular visual acuity was significantly associated with greater nystagmus fast-phase velocity (ρ = 0.37, P < 0.05) and greater nystagmus intensity (ρ = 0.55, P = 0.002), but not with nystagmus amplitude. Baseline distance monocular visual acuity was not correlated with ocular motor measures.

FIG. 3.:
Examples of eye movement recordings for one MS patient (A) and one OPT patient (B) before and on both treatments. Memantine and gabapentin were had effects on nystagmus objective parameters. MS, multiple sclerosis; OPT, oculopalatal tremor.
Mean (and median) of the different measures before (preG and preM) and under (PostG and PostM) treatment by gabapentin (G) or memantine (M), and relative P (the Wilcoxon signed-rank test for paired data)

Greater near oscillopsia measures were associated with greater mean nystagmus amplitude and greater velocity (respectively, ρ = 0.69, P < 0.01 and ρ = 0.64, P < 0.01), but not with visual acuity of the studied eye. Distance oscillopsia measures were not correlated with visual acuity, nystagmus ocular motor parameters, or vision-specific quality of life. Composite scores and peripheral vision scores on the NEI-VFQ-25 were not correlated with nystagmus ocular motor parameters.

Treatment Effects

Twelve of 16 (75%) patients completed the study. The effect of memantine was not analyzed in 2 patients, and the effect of gabapentin was not analyzed in 1 patient. In another patient, no treatment evaluation could be analyzed (see safety and tolerability paragraph below). The baseline values, measurements during treatment and statistical comparisons are presented in Table 2.

We did not find a statistically significant improvement in median distance monocular visual acuity with memantine (0.1 LogMAR, from 20/40 to 20/32, P = 0.09) or gabapentin (0 LogMAR, no change from 20/40, P = 0.055) (n = 25 eyes) treatments. Near monocular visual acuity improved by 0.18 LogMAR (from 20/50 to 20/30) on memantine (P = 0.03) and 0.12 LogMAR (from 20/63 to 20/50) on gabapentin (P = 0.05).

Distance oscillopsia improved on memantine (P = 0.05) and on gabapentin (P < 0.01). Near oscillopsia did not significantly change on memantine (P = 0.3) or on gabapentin (P = 0.3).

The composite score of the NEI-VFQ-25 did not significantly improve (P = 0.6 for both treatments). Nevertheless, gabapentin treatment significantly improved the general vision subscore on the NEI-VFQ-25 by 10 points (P = 0.03) (n = 7). See Figure 4 for functional visual measures and Figure 5 for objective ocular motor measures on eye movement recordings.

FIG. 4.:
Functional visual parameter changes on both treatments. A. Near visual acuity is expressed in LogMAR. B. Far oscillopsia is expressed in degrees. *Significant result. LogMAR, logarithm of the minimum angle of resolution.
FIG. 5.:
Ocular motor parameter changes on both treatments. Amplitudes (A), velocities (B), and frequencies (D) were recorded by video-oculography. Intensity (C) is expressed as multiple of amplitude by frequency. Data points lying below the diagonal indicate an improvement in parameters. *Significant result.

Nystagmus amplitude was significantly decreased on gabapentin (P = 0.02) and on memantine (P < 0.01). Nystagmus velocity improved by 7.0°/second (35%) on memantine (P < 0.01) and by 2.0°/second (11%) on gabapentin (P < 0.01). Nystagmus intensity was significantly decreased on memantine (1.8, P < 0.01) and on gabapentin (1, P = 0.03) (n = 25 eyes).

Nystagmus amplitude and velocity decreased by more than 50% in 7 of 24 eyes (29.2%) of patients treated with memantine and in 1 of 24 eyes (4.2%) of patients on gabapentin treatment. It is noteworthy to mention that in 4 of 24 eyes (16.7%) of 2 patients with MS, nystagmus was abolished by memantine; this was not observed with gabapentin treatment. There was no statistically significant change in nystagmus frequency with memantine or with gabapentin.

Safety and Tolerability of Each Treatment

During the trial period, 4 MS patients and all OPT patients had at least one side effect (10/16 patients, [62.5%]) while taking memantine. During gabapentin dosing, 3 patients experienced at least one side effect (18.8%). Because of side effects, 18.8% of patients discontinued memantine treatment (1 MS and 2 OPT patients), and only 6.7% discontinued gabapentin (1 MS patient). Of patients discontinuing memantine, one patient with OPT required an urgent hospital admission for a manic episode and confusional state. Drowsiness was the most common side effect reported during memantine dosing for MS patients. Other side effects of memantine were increased emotionality (1 OPT patient), irritability and anxiety (3 OPT patients), asthenia (1 MS patient), ataxia (1 OPT patient), neuropathic pain (2 OPT patients), and headaches (1 MS patient). Gabapentin side effects were less common, including slight drowsiness (2 MS patients), ataxia (2 patients), anxiety (1 MS patient), and weight gain (1 MS patient). One MS patient discontinued gabapentin because of drowsiness, vertigo, and ataxia. Gabapentin was well tolerated by OPT patients, improving APN but also lessening neuropathic pain and spasticity.

Overall, at visit 5, 4 patients (25%) chose to continue memantine at a reduced dose. Ten patients (62.5%) chose treatment at a dose of 1,200 mg/day or higher. Two patients (12.5%) decided to discontinue pharmacological treatment.


Our study confirmed that both memantine and gabapentin decrease pendular nystagmus amplitude, velocity, and intensity with no change in nystagmus frequency. This study was designed to examine treatment effects on the visual consequences of nystagmus. Under treatment, near visual acuity improved, and distance oscillopsia decreased despite no change in vision-specific quality of life. Both medications had side effects; these were more frequent on memantine. One serious adverse event was reported in a patient during memantine treatment. Finally, two-thirds of patients preferred to continue with gabapentin.

At baseline, distance and near monocular best-corrected visual acuities were around 20/50 (0.4 LogMAR) in our patients. This low visual acuity could be explained by chronic optic neuropathy as observed in 70% of MS patients in our study. Nevertheless, the correlation of near visual acuity with nystagmus velocity suggests that decreased visual acuity is also dependent on eye oscillations. In addition to decreased visual acuity, patients with acquired nystagmus complain of oscillopsia. In previous treatment trials of acquired nystagmus, oscillopsia was subjectively analyzed either as “decreased” or “absent” (5,6,12), or was not evaluated (7,8,13). In this study, as proposed by Thurtell et al (9), and for the first time in an APN trial, we attempted to objectively evaluate oscillopsia feature by estimating its direction and amplitude while viewing a stationary targets at distance and near locations. Patients reported that distance oscillopsia evaluation was more difficult than that for near because of the need to handle the laser light; however, this problem was resolved by the examiner (See Supplemental Digital Content, Supplement 1, The baseline median estimation of near oscillopsia was around 0.8° and of distance oscillopsia around 0.55°. Moreover, baseline near oscillopsia was highly correlated with nystagmus velocity and amplitude, but not distance oscillopsia. Baseline near oscillopsia magnitude (0.8°) was also closest to median nystagmus amplitude (around 1.1°).

In addition to visual outcomes, ocular instability leads to functional consequences on quality of life. Previous studies have attempted to evaluate the functional consequence of visual disturbances using a 25-level scale (7). The NEI-VFQ-25 questionnaire has been previously shown to be reliable for determining functional visual consequences of neuro-ophthalmologic findings in MS patients (13) and in patients with pendular nystagmus and either MS or OPT (4). In our study, the observed baseline NEI-VFQ-25 composite score was close to previously reported findings for patients with pendular nystagmus (4). However, we did not find any association between the baseline NEI-VFQ-25 composite score or subscores and nystagmus measurements, visual acuities, or degrees of oscillopsia. This may suggest that findings of reduced vision-specific quality of life could be related to other neurological conditions impacting vision, such as ophthalmoplegia or optic neuropathy. This may also suggest that the NEI-VFQ-25 composite score is not specifically designed to assess the functional consequences of nystagmus. The use of a 10-Item Neuro-Ophthalmic Supplement to the NEI-VFQ-25 could be used in future studies to assess quality of life.

Gabapentin and memantine were suggested as effective treatments for APN in previous studies (5–9). In our study, those treatments were associated with a significant change in the velocity of pendular nystagmus (around 11% changes during gabapentin and 35% during memantine therapy). The memantine results are very similar to those found in the study by Thurtell et al (9) in which eye speed decreased by 32.8%; however, our results for gabapentin were quite different, indicating a 27.8% decrease in eye speed. One explanation could be that the population of patients was not similar in both studies. In the study by Thurtell, patients had different types of nystagmus, including jerk nystagmus. Furthermore, we had 2 MS patients in the memantine treatment group for whom nystagmus was abolished. This interesting aspect has already been reported (7,8). Near monocular visual acuity improved in both treatment groups, by 0.18 LogMAR on memantine and 0.12 LogMAR on gabapentin. Distance visual acuity improved on both treatments but was not statistically significant. In the study by Thurtell, only distance visual acuity was measured and showed a significant 0.084 LogMAR improvement with both treatments. The discrepancy between results in both studies and in both distance and near visual acuities should prompt systematic evaluation of both measures. The evaluation of oscillopsia has been characterized by a new approach in assessment of its functional impact. Indeed, we demonstrated that oscillopsia could be greatly improved, especially with gabapentin therapy. Because there were no significant associations of visual acuity and oscillopsia, both should be considered when evaluating the specific and functional consequences of nystagmus such as oscillopsia.

Our study was not designed to compare the effects of treatments on pendular nystagmus. This was due to the uncommon nature of the patient syndromes studied and the complexity of executing treatment trials in patients with visual and neurologic dysfunction. Our goal was to demonstrate the effect of treatments on both ocular motor and functional outcomes in patients with APN. This goal was achieved, yet future studies will be needed to further document and understand the effects of treatments for APN in patients with MS and OPT.

Despite an equivalence-design trial, gabapentin seems to be safer than memantine. Indeed, memantine induced serious side effects and was badly tolerated in our disabled patients. Memantine has been already implicated in inducing neurologic impairment in MS patients (14). Caution should be advised in its use in MS patients, and, based on our data, in patients with OPT as well. Therefore, gabapentin should be used in first-line therapy of APN. Future trials should also consider levels of daily dosing that help symptoms while minimizing side effects. Although efficacy was demonstrated at 40 mg/day of memantine and 1,200 mg/day of gabapentin, our patients chose lower doses when continuing with memantine and higher doses for gabapentin.


Category 1: a. Conception and design: C. Tilikete, A. Vighetto, B. Colombet, and S. Vukusic; b. Acquisition of data: C. Tilikete, E. Nerrant, L. Abouaf, and A.-L. Vie; c. Analysis and interpretation of data: C. Tilikete, E. Nerrant, F. Pollet-Villard, and J. Berthiller. Category 2: a. Drafting the manuscript: E. Nerrant and C. Tilikete; b. Revising it for intellectual content: C. Tilikete, A. Vighetto, and S. Vukusic. Category 3: a. Final approval of the completed manuscript: C. Tilikete, A. Vighetto, S. Vukusic, and J. Berthiller.


The authors thank Léonor Favre, Felix Laborier, Myriam Prost, and Lydia Merabtene for their contribution in data recording.


1. Tilikete C, Vighetto A. Oscillopsia: causes and management. Curr Opin Neurol. 2011;24:38–43.
2. Pilling RF, Thompson JR, Gottlob I. Social and visual function in nystagmus. Br J Ophthalmol. 2005;89:1278–1281.
3. Kang S, Shaikh AG. Acquired pendular nystagmus. J Neurol Sci. 2017;375:8–17.
4. Tilikete C, Jasse L, Pelisson D, et al. Acquired pendular nystagmus in multiple sclerosis and oculopalatal tremor. Neurology. 2011;76:1650–1657.
5. Averbuch-Heller L, Tusa RJ, Fuhry L, et al. A double-blind controlled study of gabapentin and baclofen as treatment for acquired nystagmus. Ann Neurol. 1997;41:818–825.
6. Bandini F, Castello E, Mazzella L, Mancardi GL, Solaro C. Gabapentin but not vigabatrin is effective in the treatment of acquired nystagmus in multiple sclerosis: how valid is the GABAergic hypothesis? J Neurol Neurosurg Psychiatry. 2001;71:107–110.
7. Starck M, Albrecht H, Pöllmann W, Dieterich M, Straube A. Acquired pendular nystagmus in multiple sclerosis: an examiner-blind cross-over treatment study of memantine and gabapentin. J Neurol. 2010;257:322–327.
8. Starck M, Albrecht H, Pöllmann W, Straube A, Dieterich M. Drug therapy for acquired pendular nystagmus in multiple sclerosis. J Neurol. 1997;244:9–16.
9. Thurtell MJ, Joshi AC, Leone AC, Tomsak RL, Kosmorsky GS, Stahl JS, Leigh RJ. Crossover trial of gabapentin and memantine as treatment for acquired nystagmus. Ann Neurol. 2010;67:676–680.
10. Polman CH, Reingold SC, Banwell B, Clanet M, Cohen JA, Filippi M, Fujihara K, Havrdova E, Hutchinson M, Kappos L, Lublin FD, Montalban X, O'Connor P, Sandberg-Wollheim M, Thompson AJ, Waubant E, Weinshenker B, Wolinsky JS. Diagnostic criteria for multiple sclerosis: 2010 revisions to the McDonald criteria. Ann Neurol. 2011;69:292–302.
11. Ma SL, Shea JA, Galetta SL, Jacobs DA, Markowitz CE, Maguire MG, Balcer LJ. Self-reported visual dysfunction in multiple sclerosis: new data from the VFQ-25 and development of an MS-specific vision questionnaire. Am J Ophthalmol. 2002;133:686–692.
12. Strupp M, Schüler O, Krafczyk S, Jahn K, Schautzer F, Büttner U, Brandt T. Treatment of downbeat nystagmus with 3, 4-diaminopyridine: a placebo-controlled study. Neurology. 2003;61:165–170.
13. Claassen J, Spiegel R, Kalla R, Faldon M, Kennard C, Danchaivijitr C, Bardins S, Rettinger N, Schneider E, Brandt T, Jahn K, Teufel J, Strupp M, Bronstein A. A randomised double-blind, cross-over trial of 4-aminopyridine for downbeat nystagmus—effects on slow phase eye velocity, postural stability, locomotion and symptoms. J Neurol Neurosurg Psychiatry. 2013;84:1392–1399.
14. Villoslada P, Arrondo G, Sepulcre J, Alegre M, Artieda J. Memantine induces reversible neurologic impairment in patients with MS. Neurology. 2009;72:1630–1633.

Supplemental Digital Content

Copyright © 2019 by North American Neuro-Ophthalmology Society