Contents lists available at ScienceDirect

The Ocular Surface

journal homepage: www.elsevier.com/locate/jtos

Original Research

Characterization of tear production in subjects with dry eye disease duringintranasal tear neurostimulation: Results from two pivotal clinical trialsJohn D. Shepparda,∗, Gail L. Torkildsenb, Joel A. Geffinc, Jung Daod, David G. Evanse,George W. Ouslerf, Jasmine Wilsong, Stephanie N. Babag, Michelle Senchynah, Edward J. Hollandi

a Virginia Eye Consultants and Eastern Virginia Medical School, Norfolk, VA, USAb Andover Eye Associates, Andover, MA, USAc The Eye Care Group, Waterbury, CT, USAd Cornea and Cataract Consultants of Arizona, Phoenix, AZ, USAe Total Eye Care, Memphis, TN, USAf Ora, Inc., Andover, MA, USAg Allergan plc, South San Francisco, CA, USAh Allergan plc, Irvine, CA, USAi Cincinnati Eye Institute and the University of Cincinnati, Cincinnati, OH, USA

A R T I C L E I N F O

Keywords:Tear productionIntranasal neurostimulationDry eye disease

A B S T R A C T

Purpose: The intranasal tear neurostimulator (ITN) activates the nasolacrimal pathway, which is involved withbasal and bolus tear secretion. These studies characterized the acute and long-term effectiveness of the ITN instimulating tear production in subjects with dry eye disease (DED).Methods: Study 1: Randomized, double-masked, dual-controlled, 1-day crossover. Study 2: Single-arm, open-label, 180-day prospective cohort. Eligible subjects had basal unstimulated Schirmer test (with anesthesia)≤10 mm and intranasal cotton swab–stimulated Schirmer test at least 7 mm greater in the same eye, and OcularSurface Disease Index® ≥13 and ≥ 23, in Studies 1 and 2, respectively. Study 1: Subjects (n = 48) received threerandomized test applications: active intranasal, extranasal (active control), and sham intranasal (inactive con-trol) stimulation, 3 min/application with 1-hour minimum between applications. Primary outcome measure wasthe difference in Schirmer test scores during active intranasal and control applications. Study 2: Subjects(n = 97) performed intranasal neurostimulation for ≤3 min/application, 2–10 times/day. Primary outcomemeasure was the difference in Schirmer scores (stimulated minus unstimulated) at day 180. Both studies re-corded device-related adverse events (AEs).Results: Study 1: Schirmer scores (mean ± SEM) were significantly greater (p < 0.0001) with active intranasal(25.3 ± 1.5 mm) vs extranasal (9.5 ± 1.2 mm) and sham (9.2 ± 1.1 mm) applications. Study 2: Schirmerscores were significantly greater (p < 0.0001) with ITN stimulation vs unstimulated at day 180(17.3 ± 1.3 mm vs 7.9 ± 0.7 mm). No serious device-related AEs were reported in either study.Conclusion: The ITN was well-tolerated and effective in stimulating tear production with acute and long-term usein DED.Clinicaltrials.gov identifier: NCT02680158 and NCT02526290.

1. Introduction

Dry eye is a multifactorial disease of the ocular surface character-ized by a loss of homeostasis of the tear film, in which tear film in-stability and hyperosmolarity, ocular surface inflammation and da-mage, and neurosensory abnormalities play etiological roles [1]. Signsand symptoms of dry eye disease (DED) negatively impact patients’

quality of life and include ocular discomfort and visual disturbances[2–5]. More severely affected patients can experience a quality-of-lifedeficiency comparable to moderate-to-severe angina pectoris [6]. DEDis a fiscal burden to patients and society [7–9].

Recommended staged management of DED is based on severity.Mild DED is generally managed with artificial tears in addition toeducation and environmental modifications [10]. As severity

https://doi.org/10.1016/j.jtos.2018.11.009Received 6 September 2018; Received in revised form 13 November 2018; Accepted 21 November 2018

∗ Corresponding author. Virginia Eye Consultants, 241 Corporate Blvd., Suite #210, Norfolk, VA, 23502, USA.E-mail address: docshep@hotmail.com (J.D. Sheppard).

The Ocular Surface 17 (2019) 142–150

1542-0124/ © 2018 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/BY-NC-ND/4.0/).

T

progresses, additional strategies include anti-inflammatory therapies(eg. topical cyclosporine, steroids, or lifitegrast), lid hygiene, punctalocclusion, thermal pulsation, amniotic membrane bandage, autologousserum, and in the most severe cases, surgery [10,11].

Recently, a novel intranasal tear neurostimulator (ITN; TrueTear™,Allergan plc, Dublin, Ireland) received marketing authorization by theUnited States Food and Drug Administration for temporarily increasingtear production in adult patients. The ITN delivers randomized micro-current patterns to the nasal mucosa, which stimulates the nasolacrimalpathway, leading to activation of the lacrimal functional unit (LFU) andendogenous tear production [12–14]. Under normal conditions, af-ferent sensory neuronal signals from the eyelids, cornea, and con-junctiva provide feedback to the LFU via the oculolacrimal pathway,and help maintain homeostasis of the ocular surface (Fig. 1A) [15]. AsDED progresses, afferent nerve sensitivity deteriorates, compromisingLFU responsiveness to ocular surface dryness. The alternative approachto promoting tear secretion from the LFU emulates endogenous naso-lacrimal pathways (Fig. 1B) [16,17]. Afferent neurons in the nasalcavity detect and help expel foreign bodies or irritants from the nose byinducing tear secretion, thereby irrigating nasal cavities via the naso-lacrimal duct [18,19]. In addition to bolus tear secretion, the nasola-crimal pathway also accounts for approximately one-third of basal tearproduction [12]. Activation of the LFU via this pathway is a primarycompensatory mechanism for addressing ocular surface dryness[12,20–22]. Studies suggest this pathway contributes to 35% of basaltear secretion via nasal breathing [12], and disruption of the pathwayappears to be a contributing factor in DED [23,24].

An ITN prototype was previously investigated in dry eye subjectsduring two pilot studies: 1-day, double-masked, dual-control, rando-mized, crossover study (n = 16) [Lewis R, Sierra P. Presented atAmerican Society of Cataract and Refractive Surgery (ASCRS) 2016],and 180-day, prospective, open-label, single-arm study (n = 40) [25].In the 1-day study, mean Schirmer scores were significantly higherduring use of a single application of active intranasal stimulation versusactive extranasal and sham intranasal controls [Lewis R, Sierra P.ASCRS 2016]. In order to confirm whether the capacity to generatetears with the ITN was undiminished after daily use, in the open-labelstudy, subjects used the ITN at least four times daily. Stimulation withthe ITN significantly increased Schirmer scores during use versus un-stimulated scores while ocular staining and symptom scores were alsosignificantly reduced from baseline [25]. No significant safety concernswere reported in either study [Lewis R, Sierra P. ASCRS 2016; 25].Based on the totality of these results, two pivotal studies were con-ducted to characterize the acute and long-term effectiveness and safetyof the ITN in stimulating tear production in subjects with DED with a

single application on 1 day (Study 1) and with daily use for 180 days(Study 2).

2. Methods

Both studies were designed and performed in accordance with theDeclaration of Helsinki, and International Council for Harmonisationguidelines for Good Clinical Practice, and complied with the HealthInsurance Portability and Accountability Act of 1996 and local laws.Protocols were approved by a central independent review board (AspireIndependent Review Board, Santee, CA, USA) and subjects providedwritten informed consent before any procedures were performed.

2.1. The intranasal tear neurostimulator

The ITN is a handheld device designed to deliver microcurrents tothe intranasal mucosa (Fig. 2A). The disposable tips, which are insertedinto the nasal cavities, are attached to the rechargeable base unit. Whenthe device is powered on, the microcurrent strength can be adjusted

Fig. 1. The integrated lacrimal functional unit and oculolacrimal neural pathway (A), and the nasolacrimal neural pathway (B).

Fig. 2. The intranasal tear neurostimulator (A), placement of the devices duringactive intranasal application (B), active extranasal application (C), and shamintranasal application (D), sham device (E). Note: for the sham application (D),the intranasal posts had a limited insertion depth, in contrast to the fully in-serted depth during active intranasal application (B).

J.D. Sheppard et al. The Ocular Surface 17 (2019) 142–150

143

from level 0 (no stimulation) to 5 (up to 5 mA), as depicted by illu-minated light-emitting diodes (LEDs). The disposable hydrogel-con-taining tips are designed to correctly and comfortably target the ap-propriate area of nasal mucosa. Subjects were instructed to control thelocation of stimulation by the depth and angle of insertion and adjustthe intensity of stimulation by pressing the plus (+) or minus (−)buttons on the device until they experienced a “tearing” sensation, notunlike the preliminary urge to sneeze. Stimulation could be stopped atany time by pressing the minus button or withdrawing the ITN from thenostrils.

2.2. Study 1: 1-day crossover

2.2.1. Design and subjectsThis was a prospective, two-visit, randomized, double-masked,

dual-control, crossover study (ClinicalTrials.gov identifier:NCT02680158) conducted at two United States sites. The clinical hy-pothesis tested in this study was that the tear production associatedwith the use of the active intranasal ITN is greater than the tear pro-duction associated with the use of the active extranasal ITN or the shamdevice applied intranasally.

Eligibility was assessed at a screening visit conducted 45 to 1 day(s)prior to enrollment. Subjects were required to be ≥ 22 years of age, andin order to confirm aqueous-deficient dry eye with the capacity forreflex tears, were required to have a basal unstimulated Schirmer score(with anesthesia) of ≤10 mm/5 min in at least one eye and an in-tranasal cotton swab–stimulated Schirmer score ≥7 mm higher in thesame eye, and, to confirm at least mild dry eye symptoms, an OcularSurface Disease Index® (OSDI) score ≥13 (no more than three responsesof “not applicable”). To avoid exacerbating dry eye symptoms, subjectsmust have been willing to refrain from wearing contact lenses on theday of the study visit. Key exclusion criteria were chronic or recurrentepistaxis, coagulation disorders, or other conditions that could lead tothe risk of clinically significant increased bleeding; history of nasal/sinus surgery or significant trauma; vascularized nasal polyp, severelydeviated septum, or severe nasal airway obstruction; punctal or in-tracanalicular plug in either eyelid; intraocular surgery within the last 3months or refractive surgery within the past 12 months in either eye;and history of corneal transplant in either eye.

2.2.2. Study procedureDuring the single study visit (day 0), subjects performed three ap-

plications: intranasal stimulation (active application [Fig. 2B]), extra-nasal stimulation by the active device to nasal skin outside the nares(active control [Fig. 2C]), and sham intranasal stimulation (inactivecontrol [Fig. 2D]). Subjects were trained by qualified site personnel onthe proper use of the ITN and each of the two controls. Subjects wererandomly assigned in a 1:1:1:1:1:1 ratio to one of six possible appli-cation sequences using a computer-generated randomization schedulestratified by site. Each application lasted approximately 3 min with a 1-hour minimum rest between applications. Subjects and investigatorassessing the primary effectiveness outcome were masked to the iden-tity of the applications.

The sham device was similar in appearance to the ITN but the in-tranasal posts had a limited insertion depth to minimize mechanicalintranasal stimulation (Fig. 2E). Although no microcurrents were de-livered to the disposable tips, the LED lights on the sham device illu-minated to indicate subjects’ selection of microcurrent strength.

2.2.3. Study assessmentsAt the screening visit, a basal Schirmer test with anesthesia (eg.

0.5% proparacaine hydrochloride or equivalent) and a cotton swab–-stimulated Schirmer test were performed bilaterally. The study eye wasdefined as the eye with the greatest increase in tear production bycotton swab stimulation or, if there was no difference, the eye with thelower unstimulated Schirmer score. If there was no difference for either

measure, the right eye was selected as the study eye. Schirmer testscores were also evaluated in qualifying fellow eyes (eyes that met allinclusion criteria). OSDI questionnaire [26], corneal fluoresceinstaining (National Eye Institute [NEI] grading), corrected distance vi-sual acuity (CDVA) using the Early Treatment Diabetic RetinopathyStudy chart, biomicroscopy, and a nasal speculum examination wereperformed.

At the study visit (day 0), corneal staining, CDVA, and biomicro-scopy were assessed prior to the first device application. Schirmer testwith anesthesia was performed during each 3-min ITN application. Assoon as the anesthetic drops were instilled, the subject was instructed tokeep their eyes gently closed for 1 min. After opening the eyes and al-lowing the eyes to recover for approximately one additional minute,excess moisture in the inferior fornix was gently removed with a spear.Schirmer strips (35 mm × 5 mm) were placed in each eye at the junc-tion of the middle and lateral thirds of the lower eye lid. Test stripsremained in the eye until 5 min had elapsed; if the maximum score(35 mm) was reached beforehand, the time to reach the maximum wasrecorded. The investigator was masked to the application and order oftesting. In addition, the subject was not informed of the order of testingnor which application was the active device.

Vital signs (heart rate [HR], blood oxygen saturation [SpO2], andblood pressure [BP]) were measured before, during, and after eachapplication, and subjects were questioned about the occurrence ofcardiovascular and pulmonary symptoms after each application. Afterthe final application, biomicroscopy and nasal speculum examinationwere performed, and CDVA and corneal fluorescein staining were as-sessed. Adverse events (AEs) were evaluated for all subjects who ap-plied the ITN.

2.2.4. Study endpointsThe primary effectiveness outcome measure was the difference in

Schirmer scores during active intranasal device application versus thetwo control applications. The primary safety outcome measure was theincidence of device-related AEs. Additional safety outcomes includedchanges in CDVA, corneal fluorescein staining, biomicroscopy, andnasal speculum findings from baseline or screening.

2.2.5. Statistical analysisFor the primary effectiveness outcome, Schirmer test scores were

summarized by mean ± standard error of the mean (SEM) and median(range), and analyzed as the response variable in a crossover linearmodel, with sequence, application, period, and the application byperiod interaction as fixed effects, and subject (sequence) as a randomeffect. This model accounted for the correlation among observationswithin a subject. Pairwise comparisons were performed between activeand control (active and sham) applications using least square meandifferences from the crossover model and a one-sided significance levelof α = 0.025. Non-parametric Schirmer test results were rank-trans-formed prior to fitting the crossover linear model. Scores were alsocompared between active and control applications using a one-waypaired t-test or a Wilcoxon signed rank test.

Based on power analyses using data from the pilot study [Lewis R,Sierra P. ASCRS 2016] six subjects per application sequence (36 intotal) would provide > 99% power to demonstrate statistical super-iority of Schirmer test scores for the active versus each control appli-cation assuming a 10 mm difference, a crossover mean-squared-error of5.8, and a one-sided α of 0.025. Statistical analyses were performedusing SAS® version 9.2 or higher (SAS Institute Inc., Cary, NC).

2.3. Study 2: 180-day open-label

2.3.1. Design and subjectsThis prospective, six-visit, single-arm, open-label study was con-

ducted at three United States sites (ClinicalTrials.gov identifier:NCT02526290). Subjects were screened for eligibility 60 to 3 days prior

J.D. Sheppard et al. The Ocular Surface 17 (2019) 142–150

144

to enrollment. Study 2 shared the same key eligibility criteria as Study 1with the exception of an OSDI score ≥23, and additional inclusioncriteria of an NEI Scale corneal fluorescein staining score ≥2 in at leastone corneal region, with a sum of ≥4 for all corneal regions in the sameeye. These additional criteria facilitated enrollment of a stable mod-erate to severe DED population best suited to assess ITN effectiveness.The clinical hypotheses were that 180 days of neurostimulation withthe ITN was well-tolerated and, after 180 days of use, neurostimulationcontinues to result in a significant increase in acute tear productioncompared to endogenous tear production prior to stimulation.

2.3.2. Study procedureSubjects were instructed to use the ITN between 2 and 10 times

daily with no more than 3 min per use for 180 days. The base unitrecorded frequency of use, stimulation level, and duration.

2.3.3. Study assessmentsStudy visits included screening, day 0 (baseline), and four follow-up

visits (7, 30, 90, and 180 days). At screening, Schirmer tests with an-esthesia were performed bilaterally, before and during intranasal neu-rostimulation with a cotton swab. The study eye was defined as pre-viously in Study 1. At each subsequent visit, Schirmer tests (withanesthesia) were performed bilaterally, according to the same criteriaas Study 1.

Corneal staining (NEI grid, 5 regions; 0 [no staining] to 3 [severestaining]) and symptom assessments (OSDI questionnaire [26]; dry eyesymptom [DES] questionnaire, using a visual analog scale, 0 [no dis-comfort] to 100 [maximum discomfort] mm) were performed atscreening and on days 0, 7, and 30. Subject satisfaction with the ITNwas assessed at day 180 according to whether dry eye symptoms werebetter/somewhat better/the same/worse than before ITN use.

Safety assessments included AEs, CDVA, biomicroscopy, nasal en-doscopy, and sensitivity of olfaction using the University ofPennsylvania Smell Identification Test (UPSIT) [27].

2.3.4. Study endpointsThe primary effectiveness endpoint was acute stimulated tear pro-

duction in the study eye at day 180, as measured by the differencebetween Schirmer test score during stimulation and before stimulation(unstimulated). Secondary effectiveness endpoints included acute sti-mulated tear production in the study eye at days 0, 7, 30, and 90.Exploratory analyses included change from baseline in corneal staining,

symptoms (OSDI, DES questionnaire), and subject satisfaction at day180. The primary safety measure was assessment of device-related AEs.

2.3.5. Statistical analysisDemographic, baseline, effectiveness, and safety data were char-

acterized using descriptive statistics. Comparisons between baselineand follow-up values were analyzed using paired t-test or Wilcoxonsigned rank test with a two-sided significance level of α = 0.05; for theprimary and secondary effectiveness endpoints, paired t-test orWilcoxon signed rank test was used with a one-sided significance levelof α = 0.025. Missing values were not imputed. A subgroup analysiswas conducted to examine acute tear production at day 180 with re-spect to age, sex, race, baseline Schirmer score, and study site.

Acute tear production in the 180-day pilot study ranged from 4.8 to9.5 mm with a standard deviation (SD) of 6.9–8.8 mm [25]. Using aconservative estimate of 4.5 mm with a SD of 9.0 mm, a minimumsample size of 53 subjects was required for 80% power to demonstrate astatistically significant increase between the paired stimulated andbasal unstimulated Schirmer test results, using a one-sided test at an αof 0.025. Up to 100 subjects were enrolled to provide adequate safetyexperience and power.

All statistical procedures were performed using SAS® version 9.2 orhigher (SAS Institute Inc., Cary, NC).

3. Results

3.1. Study 1: Acute tear production

3.1.1. Demographics and baseline characteristicsA total of 48 subjects were randomized and completed the study

(Fig. 3A). Mean age was 56.9 years, and the majority of subjects werefemale (81.3%) (Table 1).

3.1.2. Primary effectiveness outcomeThe mean (SEM) Schirmer score in the study eye was significantly

greater during active intranasal stimulation (25.3 ± 1.5 mm) versusactive extranasal (9.5 ± 1.2 mm) and sham intranasal (9.2 ± 1.1 mm)applications (p < 0.0001 for all comparisons; Table 2 and Fig. 4A);similar results were observed in the qualified fellow eye (Fig. 4B), andin the analysis by study site. Analysis using the crossover model showeda lack of sequence or carryover effect from one application to another.

Fig. 3. Subject disposition: (A) Study 1: 1-day crossover; and (B) Study 2: 180-day open label.

J.D. Sheppard et al. The Ocular Surface 17 (2019) 142–150

145

3.1.3. Safety outcomesA total of four AEs, all mild in severity, were reported by four

(8.3%) subjects in the study. These included three non-ocular AEs―oneeach of transient lightheadedness (possibly device-related), exacerba-tion of hypertension (not device-related; hypertension diagnosed priorto day 0, and resolved after the study was repeated following moreintensive antihypertensive therapy), and intermittent nose itching(device-related), all of which resolved without sequelae. The onlyocular AE reported was corneal abrasion of a single eye (not device-related), possibly due to multiple Schirmer insertions. No serious AEswere reported and no subject withdrew from the study due to AEs.

No clinically significant effects on vital signs were observed duringor after intranasal stimulation. During intranasal application, small,transient increases in mean HR (∼6 bpm) and mean BP (∼9 mmHgsystolic, ∼8 mmHg diastolic) were observed. The only cardiovascularor pulmonary symptoms reported were one of lightheadedness (activeintranasal), and one of “rapid” heartbeat (sham application). Nasalspeculum examinations showed no changes in any subject throughoutthe study.

A ≥ 2-grade increase in corneal fluorescein staining from baselinewas observed in 8.3%–41.7% of the 96 (48 right and 48 left) eyes of allsubjects, in any region, after the final application (Table 3). The ma-jority of eyes (92/96 [95.8%]) showed no clinically relevant (< 2-line)change in CDVA from baseline after the final application. There were noclinically significant biomicroscopy findings other than the cornealabrasion AE noted above.

3.2. Study 2: 180-day open-label

3.2.1. Demographics and baseline characteristicsA total of 97 subjects were enrolled, of whom 89 (92%) completed

the 180-day period. (Fig. 3B). The mean age was 61.1 years, and themajority were female (79.4%) (Table 1).

3.2.2. Effectiveness outcomesIn the study eye, the mean ± SEM Schirmer score was significantly

higher during stimulation (17.3 ± 1.3 mm) versus unstimulated(7.9 ± 0.7 mm) yielding an increase of 9.4 ± 1.2 mm at day 180

Table 1Demographics and clinical characteristics in Study 1 and Study 2 at screening.

Study 1 (1-daycrossover)(n = 48)

Study 2 (180-dayopen-label)(n = 97)

DemographicAge, years

Mean (SD) 56.9 (13.2) 61.1 (10.0)Median (range) 58.0 (25–86) 61.0 (39–84)

Sex, n (%)Female 39 (81.3) 77 (79.4)Race

White 45 (93.8) 78 (80.4)Black or African

American2 (4.2) 16 (16.5)

Asian 1 (2.1) 3 (3.1)Clinical characteristicUnstimulated Schirmer score in study eye, mma

Mean (SD) 5.5 (3.0) N/AMedian (range) 5 (0–10)

Intranasal cotton swab–stimulated Schirmer score in study eye, mma

Mean (SD) 32.0 (5.9) N/AMedian (range) 35 (8–35)

OSDI scoreMean (SD) 41.5 (16.4) 52.2 (19.3)Median (range) 40.9 (15–83) 50.0 (22.9–100.0)

Corneal fluorescein staining (NEI)Total score, study

eyeN/A

Mean (SD) 7.5 (2.5) [n = 97]Median (range) 7.0 (4.0–14.0)

Total score,qualified felloweye

N/A

Mean (SD) 7.5 (2.5) [n = 45]Median (range) 7.0 (4.0─14.0)

N/A = not applicable; NEI = National Eye Institute; OSDI = Ocular SurfaceDisease Index; SD = standard deviation.

a Schirmer tests were conducted with anesthesia. The maximum possibleSchirmer score was 35.0 mm. Test was conducted for 5 min or until maximumwas reached.

Table 2Schirmer test scores during a single use of active intranasal, extranasal and sham intranasal control applications (Study 1), and stimulated and unstimulated Schirmertest scores at day 180 (Study 2).

Schirmer scores,a mm/5 min Study 1: 1-day crossover

Intranasal Extranasal Sham

Study eye(n = 48)

Fellow eyeb

(n = 35)Study eye (n = 48) Fellow eyeb (n = 35) Study eye (n = 48) Fellow eyeb (n = 35)

Mean (SEM) 25.3 (1.5) 26.6 (1.7) 9.5 (1.2) 9.2 (1.4) 9.2 (1.1) 10.9 (1.6)Median (range) 29.5 (0–35) 30.0 (7–35) 7.0 (0–35) 7.0 (0–35) 7.5 (0–35) 8.0 (0–35)LS mean difference between intranasal and

controls (95% CI)– – 15.8 (13.2–18.4) 17.3 (14.3–20.2) 16.1 (13.5–18.7) 15.7 (12.8–18.7)

p value for difference – – < 0.0001 < 0.0001 < 0.0001 < 0.0001

Schirmer scoresa, mm/5 min at day 180 Study 2: 180-day open-label

Study eye (n = 89) Fellow eyeb(n = 39)

Stimulated Unstimulated Stimulated Unstimulated

Mean (SEM) 17.3 (1.3) 7.9 (0.7) 17.0 (1.8) 6.7 (0.9)Median (range) 13.0 (0.0–35.0) 6.0 (0.0–28.0) 14.0 (0.0–35.0) 5.0 (0.0–23.0)Mean difference (SEM) [95%CI] 9.4 (1.2) 10.4 (1.6)

[7.1–11.7] [7.0–13.3]p value (paired t-test) < 0.0001 < 0.0001

CI = confidence interval; LS = least squares; SEM = standard error of the mean.a Maximum Schirmer score was 35.0 mm.b Qualifying fellow eyes met all of the eligibility criteria but with less severe dry eye than the study eye.

J.D. Sheppard et al. The Ocular Surface 17 (2019) 142–150

146

(p < 0.0001 paired t-test; Table 2 and Fig. 4C). Similar results wereobserved at days 0, 7, 30, and 90 (secondary effectiveness endpoints;p < 0.0001 paired t-test; Fig. 4C), with the day 0 result representingthe point of greatest tear production. Similar increased tear productionwas seen in the qualified fellow eyes (Table 2 and Fig. 4D), and allprespecified subgroups (Table 4).

Significant improvements from baseline (paired t-test) were ob-served with the ITN in corneal fluorescein staining for the total score(−0.47) and inferior region (−0.21) at day 7 in the study eye (bothp = 0.009).

Significant improvements from baseline for all DES questionnaireitems (p ≤ 0.028, paired t-test) at days 7 and 30, and the OSDI totalscore at days 7 (−5.93; p < 0.0001) and 30 (−4.61; p = 0.0013) wereobserved. For the satisfaction survey results at day 180 the majority ofsubjects reported their dry eye symptoms were better or somewhatbetter (61.9%) than before using the ITN; approximately one-third ofsubjects reported no change (29.9%) and two subjects (2.1%) reportedworsening of symptoms (survey not completed by six subjects [6.2%]).

3.2.3. Device application metricsSubjects participated in the study for a combined total of 16,068

days (44.0 years) and applied the ITN for a combined total of 11,800days. The ITN was used 27,338 times with an overall application timeof 34,726 min, and was applied for a mean ± SD of 1.7 ± 1.5 times

per day, with a mean daily application time of 2.16 ± 2.66 min perday.

3.2.4. Safety outcomesThirty-six device-related AEs were reported by 30 (30.9%) subjects

(Table 5); all were mild in severity and temporary, and no device-re-lated serious AEs were reported. Most device-related AEs occurredduring the first month of the study (22/36, 61.1%). The most frequentdevice-related AEs were nasal discomfort (n = 10; 10.3%) and nose-bleeds (n = 5; 5.2%) (Table 5); the most frequent non–device-relatedAEs were cold/flu symptoms (18.6%), sinus infection, pneumonia andgastrointestinal AEs (each 3.1%).

Three subjects discontinued because of AEs: mild, intermittent nasalpain, subsequently diagnosed as a sinus infection that was not device-related (n = 1); mild left nostril pain lasting 4 days (n = 1); and lungadenocarcinoma that was not device-related (n = 1). The majority ofsubjects had no clinically significant change in CDVA, biomicroscopy,and nasal endoscopy findings at day 180. There was no notable effecton smell identification based on the UPSIT.

4. Discussion

Findings from these pivotal studies confirm the acute and long-termeffectiveness of the ITN in stimulating tear production in dry eye

Fig. 4. Acute tear production with the intranasal tear neurostimulator. Study 1: mean Schirmer test scores in study eyes (A) and qualified fellow eyes (B) duringactive intranasal, and extranasal and sham control stimulations. Study eyes, n = 48; fellow eyes, n = 35. *p < 0.0001, stimulated versus unstimulated; error barsrepresent standard error of the mean. Study 2: Stimulated and unstimulated mean Schirmer score in study eyes (C) and qualified fellow eyes (D) at each study visit.Error bars represent standard error of the mean; number of eyes assessed at each study visit is indicated in parentheses. *p < 0.0001 for stimulated compared withunstimulated (paired t-test).

J.D. Sheppard et al. The Ocular Surface 17 (2019) 142–150

147

subjects.In the 1-day crossover study, a single application of the ITN induced

significantly greater acute tear production than active extranasal or

sham intranasal applications. The magnitude of this increase is con-sistent with the prototype ITN pilot study, which employed the samecrossover design and test applications [Lewis R, Sierra P. ASCRS 2016].

Following daily use of the ITN for 180 days, acute tear productionwas significantly increased with stimulation using the ITN comparedwith unstimulated production. The ITN was effective at all earlier timepoints and plateaued at day 7 out to day 180, relative to a slightlygreater value at day 0. The hypothesis underlying this observation isthought to be adaptation to mechanical stimulation as generatedthrough insertion of the ITN. The ITN was effective across a broadpopulation including subjects with mild DED (ie. OSDI ≥13) in Study 1,and prespecified subgroups stratified by baseline Schirmer scores inStudy 2, which included 42 subjects with baseline study eye Schirmerscores ≤5 mm. Improvements in corneal staining were observed forsome regions and time points, but results were inconsistent.Improvements in dry eye symptoms from baseline were statisticallysignificant at days 7 and 30, and over 60% of subjects reported animprovement in dry eye symptoms at day 180.

There were no serious or clinically significant safety issues reportedfollowing single application or after daily ITN use for 180 days.

The only ocular AE reported in the 1-day study, corneal abrasion,was unrelated to the ITN, and together with the observed increase incorneal staining, was expected given the repeat Schirmer testing over aperiod of several hours. Otherwise, there were no clinically significantchanges in vital signs, nasal examinations, biomicroscopy findings, orCDVA. Clinically insignificant and transient changes in HR and BP wereconsidered to be similar to those occurring during daily activities, suchas postural adjustments or consuming a cold drink [28–30].

Following frequent daily use of the ITN, equating to over 27,338cumulative stimulations, there were no device-related serious AEs re-ported in the 180-day study. The most common device-related andnon–device-related AEs (nosebleeds and cold/flu symptoms, respec-tively) may have been at least partially weather-related, as the studywas performed during winter when nasal desiccation, nosebleeds, andupper respiratory infections are common [31,32].

Limitations of the 1-day crossover study include the intensity oftesting and abbreviated time frame of exposure to the ITN. Limitationsof the 180-day study include the single-arm, open-label nature, rela-tively mild corneal staining severity at baseline, and the limited racialheterogeneity of the study population. Both studies included subjectswith impaired basal tear production but normal functioning reflextears, and as such the results from these studies may not be applicable

Table 3Mean (SD) change in corneal staining from baseline (prestimulation) to post-stimulation in all eyes, and the proportion of all eyes with 2-grade or greaterchange in corneal staining from baseline after the final application (Study 1).

Corneal region Study 1: 1-day crossoverAll subjects (N = 96 eyes)

Right eyes(n = 48)

Left eyes (n = 48)

Inferior, mean (SD)Baseline 1.46 (0.82) 1.52 (0.87)Post-stimulation 2.58 (0.71) 2.31 (0.83)Change from pre- to poststimulation 1.13 (0.84) 0.79 (0.97)

Superior, mean (SD)Baseline 0.71 (0.59) 0.77 (0.59)Poststimulation 1.21 (0.65) 1.25 (0.70)Change from pre- to poststimulation 0.50 (0.88) 0.48 (0.85)

Central, mean (SD)Baseline 0.85 (0.90) 0.81 (0.91)Poststimulation 1.83 (1.08) 1.67 (1.12)Change from pre- to poststimulation 0.98 (1.00) 0.85 (1.15)

Temporal, mean (SD)Baseline 1.08 (0.71) 1.02 (0.56)Poststimulation 1.88 (1.04) 1.79 (1.07)Change from pre- to poststimulation 0.79 (1.09) 0.77 (1.08)

Nasal, mean (SD)Baseline 1.29 (0.85) 1.31 (0.78)Post-stimulation 1.94 (0.93) 2.04 (0.99)Change from pre- to poststimulation 0.65 (0.96) 0.73 (0.96)

≥ 2 Grade Increase in NEI Scale from Prestimulation to Poststimulation, n (%)Inferior 20 (41.7) 14 (29.2)Superior 4 (8.3) 6 (12.5)Central 13 (27.1) 13 (27.1)Temporal 15 (31.3) 17 (35.4)Nasal 9 (18.8) 10 (20.8)

Table 4Subgroup analysis of acute tear production assessed by Schirmer Scores at day180 (Study 2).

Parameter Study 2: 180-day open-labelSchirmer score, mean (SEM), mm

p value (pairedt-test)

Stimulated Unstimulated Meandifference

Age< 50 years

(n = 14)16.7 (3.4) 6.3 (1.5) 10.4 (3.2) 0.003

50 to < 60 years(n = 24)

19.2 (2.8) 9.8 (1.6) 9.4 (2.2) 0.0001

60 to < 70 years(n = 35)

17.4 (2.0) 7.3 (0.8) 10.1 (1.8) < 0.0001

≥70 years(n = 16)

14.6 (2.4) 7.9 (1.8) 6.8 (2.9) 0.018

SexFemale (n = 71) 17.7 (1.4) 8.3 (0.8) 9.4 (1.4) < 0.0001Male (n = 18) 15.6 (2.7) 6.3 (1.1) 9.3 (2.1) 0.0002

RaceWhite (n = 71) 16.4 (1.4) 7.6 (0.8) 8.8 (1.3) < 0.0001Non-white

(n = 18)20.8 (2.8) 9.4 (1.3) 11.4 (2.5) 0.0002

Baseline Schirmer score0–5 mm (n = 42) 15.3 (1.9) 6.5 (1.0) 8.9 (1.8) < 0.00016–10 mm

(n = 47)19.0 (1.7) 9.2 (0.8) 9.8 (1.5) < 0.0001

Study site1 (n = 29) 14.8 (2.1) 6.3 (1.2) 8.4 (2.0) 0.00012 (n = 34) 18.4 (2.0) 8.2 (1.0) 10.1 (1.7) < 0.00013 (n = 26) 18.7 (2.6) 9.3 (1.3) 9.4 (2.5) 0.0004

SEM = standard error of the mean.

Table 5Incidence of adverse events and device-related adverse events (Study 2).

Study 2: 180-day open-labelAll subjects (n = 97)a

Adverse events (ocular and non-ocular)Number of AEs 126Number (%) with at least one AE 58 (59.8)Device-Related AEsNumber of AEs 36Number (%) with at least one AE 30 (30.9)

Nasal pain/discomfort/burning 10 (10.3)Nosebleed 5 (5.2)Transient electrical discomfort 5 (5.2)Nasal congestion 3 (3.1)Facial pain 2 (2.1)Headaches 2 (2.1)Trace blood, dot heme in nostril 2 (2.1)Lightheadedness 1 (1.0)Nasal ulcers 1 (1.0)Periorbital pain 1 (1.0)Runny nose 1 (1.0)Sore eye 1 (1.0)Sinus pain 1 (1.0)

a Some subjects reported more than one AE.

J.D. Sheppard et al. The Ocular Surface 17 (2019) 142–150

148

to those with evaporative dry eye but normal aqueous tear production,or impaired reflex tears. Nevertheless, there is evidence that use of theITN results not only in the secretion of the aqueous component, but alsoof mucin [33,34], protein [35] and lipid [36,37], suggesting that theITN promotes secretion of a complete endogenous tear. Additional long-term studies focusing on the impact of daily use of the ITN on diversepopulations such as those with evaporative DED, the improvement onDED signs and symptoms, and quality of life may provide valuable dataguiding the potential utility of the ITN in DED management.

In summary, these findings demonstrate that the ITN is capable ofstimulating significant near to intermediate tear production that ismaintained throughout a 180-day course of chronic use. The ITN waswell tolerated following acute and chronic treatment. These resultssuggest that intranasal neurostimulation could be a potential option tostimulate endogenous tear production in patients with DED. The uniqueand innovative ITN provides a first-in-class, pathophysiologically re-levant, novel mechanism of action that may prove complementary,adjunctive, additive or synergistic to a wide variety of other manage-ment approaches available for dry eye patients.

Funding

Both studies were sponsored by Allergan plc, Dublin, Ireland.Allergan participated in the development of the study concept anddesign, interpretation of the data, preparation and review of themanuscript, and the decision to submit the article for publication.

Financial disclosures

The authors provide the following disclosures: J.D.S. is a consultantto AbbVie, Aerie Pharma, Alcon, Aldeyra, Allergan plc, Bausch & Lomb,Bio-Tissue, Clearside, EyeGate Research, Kala Pharmaceuticals,NovaBay, Noveome, Omeros, Parion, Portage, Quidel, Santen, ScienceBased Health, Shire, Sun Pharmaceuticals, TearLab, TearScience, andTopivert; has received research funding from Alcon, Aldeyra, Allergan,Bausch & Lomb, EyeGate Pharma, Kala Pharmaceuticals, Parion,Santen, and Shire and honoraria from AbbVie, Alcon, Allergan plc,Bausch & Lomb, Bio-Tissue, EyeGate Research, Omeros, Santen, ScienceBased Health, Shire, Sun Pharmaceuticals, TearLab, and TearScience;has received travel reimbursement from AbbVie, Alcon, Aldeyra,Allergan, Bausch & Lomb, Bio-Tissue, Clearside, EyeGate Research, KalaPharmaceuticals, NovaBay, Omeros, Portage, Santen, Science BasedHealth, Shire, Sun Pharmaceuticals, TearLab, TearScience, andTopivert; and is an investor in Lacrisciences, EyeGate Research,NovaBay, TearLab, RPS, and Noveome. G.L.T. is a consultant toAllergan, plc and Ora, Inc. J.A.G. and J.D. have no conflict of interest todisclose. D.G.E. received research funding from Alcon, Allergan plc,Auven Therapeutics, Encore Vision Inc., Johnson & Johnson Vision, andKala Pharmaceuticals. G.W.O. is an employee of Ora, Inc. J.W. andS.N.B. were employees of Allergan plc (formerly Oculeve, Inc.) whenthe study was conducted. M.S. is an employee of Allergan plc. E.J.H. is aconsultant for Alcon, Allergan plc, Bausch & Lomb, Kala Corporation,Mati Therapeutics, Omeros, PRN, Senju Pharmaceuticals, Shire,TearLab, and TearScience and has received research funding fromAlcon, Allergan plc, Mati Therapeutics, Omeros, PRN, and SenjuPharmaceuticals; and has received honoraria from Alcon, Allergan,Bausch & Lomb, Omeros, Senju Pharmaceuticals, and TearScience.

Acknowledgements

The authors would like to thank the subjects who participated inthis study. Writing and editorial assistance was provided to the authorsby Kakuri M. Omari, PhD and Stuart Murray, MSc, of EvidenceScientific Solutions, Philadelphia, PA, USA, and was funded by Allerganplc. All authors met the ICMJE authorship criteria. Neither honorarianor payments were made for authorship.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.jtos.2018.11.009.

References

[1] Craig JP, Nichols KK, Akpek EK, et al. TFOS DEWS II definition and classificationreport. Ocul Surf 2017;15(3):276–83.

[2] Buchholz P, Steeds CS, Stern LS, et al. Utility assessment to measure the impact ofdry eye disease. Ocul Surf 2006;4(3):155–61.

[3] Friedman NJ. Impact of dry eye disease and treatment on quality of life. Curr OpinOphthalmol 2010;21(4):310–6.

[4] McDonald M, Patel DA, Keith MS, Snedecor SJ. Economic and humanistic burden ofdry eye disease in Europe, North America, and Asia: a systematic literature review.Ocul Surf 2016;14(2):144–67.

[5] Miljanović B, Dana R, Sullivan DA, Schaumberg DA. Impact of dry eye syndrome onvision-related quality of life. Am J Ophthalmol 2007;143(3). 409–5.e2.

[6] Schiffman RM, Walt JG, Jacobsen G, Doyle JJ, Lebovics G, Sumner W. Utility as-sessment among patients with dry eye disease. Ophthalmology2003;110(7):1412–9.

[7] O'Brien PD, Collum LM. Dry eye: diagnosis and current treatment strategies. CurrAllergy Asthma Rep 2004;4:314–9.

[8] Sheppard JD. Guidelines for the treatment of chronic dry eye disease. Manag Care2003;12(12 Suppl):20–5.

[9] Stapleton F, Alves M, Bunya VY, et al. TFOS DEWS II epidemiology report. Ocul Surf2017;15(3):334–65.

[10] Jones L, Downie LE, Korb D, et al. TFOS DEWS II management and therapy report.Ocul Surf 2017;15(3):575–628.

[11] American Academy of Ophthalmology Cornea/External Disease PPP Panel HoskinsCenter for Quality Eye Care. Preferred Practice Pattern® Guidelines. Dry eye syn-drome. https://www.aao.org/preferred-practice-pattern/dry-eye-syndrome-ppp–2013. Accessed July 2018.

[12] Gupta A, Heigle T, Pflugfelder SC. Nasolacrimal stimulation of aqueous tear pro-duction. Cornea 1997;16(6):645–8.

[13] Baroody FM, Shenaq D, DeTineo M, Wang J, Naclerio RM. Fluticasone furoate nasalspray reduces the nasal-ocular reflex: a mechanism for the efficacy of topical ster-oids in controlling allergic eye symptoms. J Allergy Clin Immunol2009;123:1342–8.

[14] Brinton M, Chung JL, Kossler A, Kook KH, Loudin J, Franke M, et al. Electronicenhancement of tear secretion. J Neural Eng 2016;13:016006.

[15] Stern ME, Gao J, Siemasko KF, Beuerman RW, Pflugfelder SC. The role of the la-crimal functional unit in the pathophysiology of dry eye. Exp Eye Res2004;78:409–16.

[16] Baraniuk JN, Kim D. Nasonasal reflexes, the nasal cycle, and sneeze. Curr AllergyAsthma Rep 2007;7(2):105–11.

[17] Zilstorff-Pedersen K. Quantitative measurements of the nasolacrimal reflex. ArchOtolaryngol 1965;81:457–62.

[18] Holzer P. Capsaicin: cellular targets, mechanisms of action, and selectivity for thinsensory neurons. Pharmacol Rev 1991;43(2):143–201.

[19] Philip G, Baroody FM, Proud D, Naclerio RM, Togias AG. The human nasal responseto capsaicin. J Allergy Clin Immunol 1994;94(6 Pt 1):1035–45.

[20] Krupin T, Cross DA, Becker B. Decreased basal tear production associated withgeneral anesthesia. Arch Ophthalmol 1977;95:107–8.

[21] Heigle TJ, Pflugfelder SC. Aqueous tear production in patients with neurotrophickeratitis. Cornea 1996;15:135–8.

[22] Tsubota K. The importance of the Schirmer test with nasal stimulation. Am JOphthalmol 1991;111:106–8.

[23] Dartt DA. Neural regulation of lacrimal gland secretory processes: relevance in dryeye diseases. Prog Retin Eye Res 2009;28(3):155–77.

[24] Rocha EM, Alves M, Rios JD, Dartt DA. The aging lacrimal gland: changes instructure and function. Ocul Surf 2008;6(4):162–74.

[25] Friedman NJ, Butron K, Robledo N, Loudin J, Baba SN, Chayet A. A nonrandomized,open-label study to evaluate the effect of nasal stimulation on tear production insubjects with dry eye disease. Clin Ophthalmol 2016;10:795–804.

[26] Schiffman RM, Christianson MD, Jacobsen G, Hirsch JD, Reis BL. Reliability andvalidity of the ocular surface disease Index. Arch Ophthalmol 2000;118:615–21.

[27] Doty RL, Shaman P, Kimmelman CP, Dann MS. University of Pennsylvania SmellIdentification Test: a rapid quantitative olfactory function test for the clinic.Laryngoscope 1984;94(2 Pt 1):176–8.

[28] Ma J, Hu W, Wu D, Zhang Q, Peng C, Cao K, et al. Drinking cold water increasesblood pressure in healthy young students. Blood Press Monit 2014;19:118–9.

[29] MacDonald JR, Hogben CD, Tarnopolsky MA, MacDougall JD. Post exercise hypo-tension is sustained during subsequent bouts of mild exercise and simulated ac-tivities of daily living. J Hum Hypertens 2001;15:567–71.

[30] Pickering TG, Harshfield GA, Kleinert HD, Blank S, Laragh JH. Blood pressureduring normal daily activities, sleep, and exercise. Comparison of values in normaland hypertensive subjects. J Am Med Assoc 1982;247:992–6.

[31] Eccles R. An explanation for the seasonality of acute upper respiratory tract viralinfections. Acta Otolaryngol 2002;122:183–91.

[32] Monto AS. Epidemiology of viral respiratory infections. Am J Med 2002;112(Suppl6A):4S–12S.

[33] Gumus K, Schuetzle K, Pflugfelder SC. Randomized, controlled, crossover trialcomparing the impact of sham or intranasal tear neurostimulation on conjunctival

J.D. Sheppard et al. The Ocular Surface 17 (2019) 142–150

149

goblet cell degranulation. Am J Ophthalmol 2017;177:159–68.[34] Dieckmann G, Jamali A, Pondelis N, Kataguiri P, Salem Z, Franke M, Senchyna M,

Hamrah P. In vivo confocal microscopy demonstrates intranasal neurostimulation-induced goblet cell alterations in patients with dry eye disease. Invest OphthalmolVis Sci 2017;58(8):2694.

[35] Woodward AM, Senchyna M, Franke M, Holdbrook M, Argüeso P. Effect of in-tranasal neurostimulationon tear protein content in patients with dry eye. InvestOphthalmol Vis Sci 2017;58(8):2673.

[36] Pondelis NJ, Dieckmann G, Kataguiri P, Abbouda A, Salem Z, Franke M, SenchynaM, Hamrah P. Intranasal neurostimulator induces morphological changes in mei-bomian glands in patients with dry eye disease. Invest Ophthalmol Vis Sci2017;58(8):2235.

[37] Green KB, Kamat M, Franke M, Holdbrook M, Senchyna M. Tear total lipid con-centration in patients with dry eye following intranasal neurostimulation. InvestOphthalmol Vis Sci 2017;58(8):2693.

J.D. Sheppard et al. The Ocular Surface 17 (2019) 142–150

150