A Scoping Review of Pharmacologic and Non-Pharmacologic Interventions to Treat Chronic Ocular Pain

Abstract

Background:

Chronic Ocular Pain (COP) is disabling and a common reason for referral to neuro-ophthalmologists. Many patients with COP suffer from amplified or dysfunctional pain processing in the peripheral or central nervous system, rather than a problem with the eye itself. The extent and evidence for therapeutic interventions targeting pain in the peripheral and central nervous system among patients suffering from COP is unknown.

Evidence Acquisition:

We performed a search of Ovid MEDLINE, Embase.com, Scopus.com, and Clinicaltrials.gov. We limited our search to English language articles published since 2015. We included observational and experimental studies of pharmacologic and non-pharmacologic interventions (e.g., vagal nerve stimulation) used in adults 18 years or older with COP for at least 3 months. Studies were included if impact on pain intensity, functioning, or quality of life were measured as primary, secondary, or exploratory outcomes. All articles were independently screened by two team members for inclusion. Data extraction was performed independently by two team members. Any disagreements during the screening and data extraction process were adjudicated by a third team member and discussed as a group.

Results:

Our search identified 9656 studies from 2015 to 2026. Of those, 21 met our inclusion/ exclusion criteria. The study designs were mostly retrospective: five case reports, nine observational, one case-control study; three prospective studies were non-randomized, observational studies. three studies were randomized clinical trials, but one did not have published results. Most (67%, n=14) were studies of pharmacologic interventions including all three completed prospective studies. Pharmacologic interventions included gabapentin, cryosim-3 applied to the eyelids, autologous serum tears, botulinum toxin type A injections, and nerve blocks. The six studies including non-pharmacologic interventions used transcutaneous electrical nerve stimulation, trigeminal ganglion stimulation, and pulsed radiofrequency procedures. The three randomized controlled clinical trials were of transcutaneous electrical nerve stimulation, topical ocular cooling device, and topical OK-101 ophthalmic solutions at different concentrations. The most common pain measures used across studies were pain intensity scales (range 0–10), such as the numeric rating scale. Improvements in COP were reported for most studies. Reporting of adverse events occurred in eight (43%) studies.

Conclusions:

Although COP is a common symptom, we found few studies of pharmacologic and non-pharmacologic treatment interventions with a few randomized controlled trials. Rigorous studies of treatment interventions aimed at reducing COP are needed. We provide recommendations for both the conduct of future studies and a current approach to COP treatment.

Introduction

Chronic ocular pain (COP) describes pain in one or both eyes present for at least three months. COP represents a leading cause of eye-related healthcare visits in the United States (US) and a common source of referrals for neuro-ophthalmologists.¹ COP can occur in isolation, but may also be present in a number of ophthalmic and neurologic conditions, most commonly dry eye disease and migraine. Treatments for COP are often aimed at addressing sources of pain in the eye itself, such as restoring tear film homeostasis, but for many patients, objective signs of ocular abnormalities do not exist or explain the degree of pain. As a result, COP often goes unaddressed leading to decreased quality of life^2,3 and visual functioning,⁴ and higher odds of suicidal ideation,⁵ in comparison to people without COP.

Mounting evidence indicates that many of these individuals with COP have neuropathic or nociplastic mechanisms underlying their pain. Neuropathic pain arises from injury to the nerves in the peripheral nervous system (PNS) or central nervous system (CNS). Nociplastic pain is distinct from neuropathic pain as there is no injury to the peripheral or CNS.⁶ Likewise, in people with nociplastic pain, there is no clear evidence of tissue injury to account for the pain. Instead, pain arises from amplified and/or dysregulated neural signaling and sensory processing within the CNS.

Multimodal therapies emphasizing non-pharmacologic treatments alongside pharmacologic interventions acting on the PNS and CNS are fundamental to managing neuropathic and nociplastic pain.⁷ Non-pharmacologic treatments include lifestyle changes (e.g., exercise, sleep hygiene), psychotherapeutic interventions (e.g., cognitive behavioral therapy), and procedures (e.g., nerve blocks, transcutaneous electrical nerve stimulation). Pharmacologic interventions include medications aimed at modulating the PNS and CNS. A scoping review of interventions, both pharmacologic and non-pharmacologic, that act on pain arising from the PNS or CNS in individuals with COP has not been previously performed. Furthermore, a preliminary search for any systematic reviews or scoping reviews on the topic were identified in MEDLINE or the Cochrane Database of Systematic Reviews.

Our aim is to provide a scoping review summarizing the extent and evidence for pharmacologic and non-pharmacologic interventions targeting underlying neuropathic or nociplastic pain mechanisms in patients with COP. Additionally, we summarize the impact of these interventions on relevant pain outcomes and compare the studied interventions in COP to those used to treat other common regional or widespread pain syndromes so as to identify promising interventions for further research. We chose to perform a scoping review to provide clinicians with a single resource that summarizes the current state of COP therapeutics aimed at addressing neuropathic and/or nociplastic pain. These data can support neuro-ophthalmologists and other clinicians caring for patients with COP and guide future therapeutic research efforts.

Methods

A scoping review was undertaken to identify all studies from 2015 to 2026 of interventions that have been used to treat COP and understand the impact of these interventions on pain intensity and functioning when compared to placebo or standard care. Our review was conducted in accordance with the Joanna Briggs Institute (JBI) methodology for scoping reviews.⁸ A study protocol was deposited in the Deep Blue Repository.⁹

We included studies with adults 18 years or older in which treatment of COP was the primary objective. We defined COP as any symptom of discomfort arising in the area in or around the eyes for three months or more. This includes symptoms such as irritation, burning, stabbing, itching, and sensation of dryness. We excluded photophobia as a standalone symptom. We chose three months as this is a standard definition of “chronic” across pain conditions. The concepts of interest were number of studies, including unpublished, and published studies, participant characteristics, pain measures, intervention types, and strengths and weaknesses of the studies. We included any ambulatory or emergency health care setting. We included both experimental and quasi-experimental study designs including randomized controlled trials, non-randomized controlled trials, before and after studies (pre-post designs) and interrupted time-series studies. Analytical observational studies including prospective and retrospective cohort studies, case-control studies and analytical cross-sectional studies were considered for inclusion. We considered descriptive observational study designs including case series, individual case reports and descriptive cross-sectional studies for inclusion. Qualitative studies were also considered. Reviews were excluded but were used to scan for additional primary literature references.

Search strategy

The search strategy aimed to locate both published and unpublished studies. An initial limited search of MEDLINE was undertaken to identify articles on the topic. The text words contained in the titles and abstracts of relevant articles, and the index terms used to describe the articles were used to develop a full search strategy for Ovid MEDLINE, Scopus.com, Embase.com, and Clinicaltrials.gov (eMethods). The search strategy, including all identified keywords and index terms, were adapted for each included database and/or information source. The reference list of all included sources of evidence was screened for additional studies. Studies published in English since 2015 were included to reflect the advances in the understanding of COP mechanisms and subsequent mechanism-targeted treatment over the past 12 years.

Study selection

Following the search, all identified citations were collated and uploaded into Covidence and duplicates removed. Titles and abstracts were screened by NK with adjudication by LBD for assessment against the inclusion criteria for the review. Potentially relevant sources were retrieved in full. The full text of selected citations were independently assessed in detail against the inclusion criteria by NK and AS. Reasons for exclusion of sources of evidence at full text that do not meet the inclusion criteria were recorded. Any disagreements that arose between the reviewers at each stage of the selection process were resolved through discussion and with an additional review by LBD. The results of the search and the study inclusion process are presented in a flow diagram.

Data extraction, analyses, and reporting

Data were independently extracted from papers included in the scoping review by NK and AS. The data extracted included specific details about the study participants, concepts, and key findings relevant to the review question. Any disagreements that arose between the reviewers were resolved through discussion with LBD. The accuracy of the data extraction was independently verified by MY and HA. The study authors were contacted to request missing or additional data when needed. The extracted data are presented in tables to reflect a summary of the results that aligns with the objective and research question of this scoping review. Our scoping review was reported according to the Preferred Reporting Items for Systematic Reviews and Meta-analyses extension for scoping reviews (PRISMA-ScR; eTable 1).¹⁰

Results

The database search identified 9656 records for inclusion. No additional studies were identified through other sources (e.g., content experts). The search results were entered into Covidence and 2335 duplicates were removed resulting in 7321 studies. The titles and abstracts of these records were screened and an additional 7271 were excluded. Full text review occurred for 50 articles. Of these, 29 were excluded due to wrong outcomes, study design (e.g., narrative review), patient population, or inability to find full text article. Notably, two retrospective cohort studies investigating low dose naltrexone¹¹ and nortiptyline¹² for COP were not included due to lack of data confirming duration of pain. In total, 21 articles met all eligibility criteria for inclusion in this review.^13–32 Figure 1 shows a flow diagram for identifying eligible studies.

Figure 1.

Flow Diagram of the Study Inclusion Process

Data extracted for each study is included in Table 1. Most studies (n=11) were published 2019–2021;^{13,14,17,20–22,24,27,28,32,33} three from 2015–2018,^15,16,19 and seven 2022–2026.^{18,23,25,29–31,34} Most studies were retrospective cohort studies (n=9)^18–25,32 or case reports (n=5).^13–17 Additionally, we identified one study protocol for a randomized controlled trial (completed, not published),³¹ one retrospective case-controlled study,³³ and three non-randomized prospective studies (1 pre-post and 2 observational cohort designs).^27–29 One study was a randomized controlled pilot study³⁰ and another was a randomized controlled trial.³⁴ The interventions prospectively evaluated include high-frequency TENS (randomized controlled pilot study),³⁰ a topical ocular cooling device,³⁴ gabapentin in combination with artificial tears and cyclosporine A ophthalmic,²⁸ cryosim-3 (Transient Receptor Potential Melastatin 8 agonist) applied topically to the eyelid,²⁷ onabotulinum toxin type A used in a modified migraine pattern.²⁹ The retrospective case control study investigated autologous serum tears.³³ The remaining retrospective and case report studies investigated gabapentin,^14,22 a nerve or ganglion stimulation strategy, such as TENS, or pulsed radiofrequency therapy delivered either in isolation or combined with a pharmacologic intervention (onabotulinum toxin A, intrathecal fentanyl),^{15,16,19–21,24,25,32} or nerve blocks^13,17,22,23 or onabotulinum toxin A injections¹⁸ either in isolation or combined with other interventions. Concurrent use of other medications (e.g., gabapentin) was not an exclusion criteria for any studies. One study was a retrospective chart review of all the various therapies tried including oral, topical, and adjuvant therapies, which included TENS.³²

Table 1.

Characteristics of included studies

Author, Year	Study Design	Study period/ follow-up duration	Sample Size (by arm/group)	Key Eligibility Criteria	Intervention/Exposure	Baseline characteristics (age, sex, etc)	Pain outcome measured and effect estimates	Adverse events
Michael et al., 2020	Case report	Unknown	1	N/A	Oral gabapentin 300 mg twice daily	37 years, M. Left eye pain for 8 months following complicated LASIK	Pre-treatment NRS: Not reported Post-treatment NRS: 0, 3 weeks after starting gabapentin	Not reported
Yalamananchili et al., 2019	Case report	3 years	1	N/A	Total of 8 bilateral orbital injection of triamcinolone acetonide 40 mg/cc and bupivacaine 0.25% mixed 50:50 with injection of 3–4mL. Recent change to dexamethasone implant (Ozurdex; Allergan, Irvine, CA) and 0.25% bupivacaine	21 years, M. Bilateral eye pain post-refractive surgery and failed artificial tears, corticosteroid drops, nonsteroidal drops, and oral medications (NSAIDs, opioids, and pregabalin)	Pre-treatment NRS: 7–9 Post-treatment: 1–3 Improvement lasting 4–9 months before repeat injection required	Not reported
Li et al., 2015	Case report	Unknown	1	N/A	Combined left supraorbital nerve stimulation and 100 units onabotulinum toxin type A	73 years, M. 1 year of postherpetic pain in left V1 dermatome	Pre-treatment VAS: 8–9 Post-treatment VAS: 2–3	Not reported
Sayegh et al., 2016	Case report	>1 year	1	N/A	Intervention 1: Trigeminal ganglion stimulation via implantation of a deep brain stimulator electrode Intervention 2: Intrathecal infusion system delivering fentanyl and bupivacaine	32 years, W. 9 years of photoallodynia and “bilateral dry eye” following LASIK. Normal ocular surface exam and partial improvement with proparacaine	“well controlled”	Intervention 1: Migration of electrode lead and explanation Intervention 2: Revision needed 2 weeks after implantation for catheter migration and cerebrospinal fluid collection around the pump
Duerr et al., 2019	Case report	7 months	1	N/A	Supraorbital, supratrochlear, infratrochlear, infraorbital nerve blocks: 4 mL of 0.5% bupivacaine mixed with 1 mL of 80 mg/mL methylprednisolone acetate and 0.5–1mL injected at each site. On week later, had repeat infraorbital and infratrochlear nerve blocks.	66 years, M. Left pain after retinal detachment and scleral buckle failed autologous serum tears, corticosteroid drops, diclofenac drops, cyclosporine drops, gabapentin, diclofenac	Pre-treatment NRS: 8 Post-treatment NRS: 1–2	None
Locatelli et al., 2025	Retrospective cohort	4–6 weeks	27	Veterans with neuropathic ocular pain symptoms (light sensitivity, wind/air sensitivity, persistent pain despite dry eye treatment) undergoing treatment with botulinum toxin	Inca- or Onabotulinum toxin type A 35–100 U into corrugators, procerus, frontalis	52.6 ± 10.7 years, 55.6% F	Degree of pain improvement (range 0–3; 0 = none, 1 = mild, 2 = moderate, 3 = marked change) 74.1 % (n = 20) with any improvement. 25 % (n = 5) mild, 45 % (n = 9) moderate, 30 % (n = 6) marked improvement.	Not reported
Sivanesan et al., 2018	Retrospective cohort	24 hours	14	Veterans with Neuropathic ocular pain determined by specific descriptors such as burning, sensitivity to light (e.g. photophobia) and wind and/or high discordance between symptoms and signs	TENS, bilateral electrode placement along the ocular midline above the brow and another at the temple	47.1 ± 12.3 years, 21.4% F	Mean±SD Defense and Veterans Pain Rating Scale (0–10) Baseline/ Endpoint Right eye: 4.54±3.18 to 1.92±2.50 (p=0.01) Left eye: 4.46±3.36 to 2.00±2.38 (p=0.01)	Mild, transient epiphora during treatment (n = 1); worsening pain requiring early termination (n = 1); no new adverse events at 24-hour follow-up
Zayan et al., 2020	Retrospective cohort	>3months	10	I: Veterans with neuropathic pain and treated nociceptive sources, response in clinic E: contraindications to TENS	TENS using the RS Medical RS-4i Plus device	48.6 ± 12.5 years, 20% F	Post- vs pre-treatment NRS: decreased by 27.4% (mean rank=5.6, Z=−2.1, p=0.02)	None
Liu et al., 2021	Retrospective cohort	3 months	32	I: Post herpetic neuralgia of the ophthalmic branch E: skin/tissue infection at puncture site, coagulopathy, severe cardio/ cerebrovascular disease	Pulsed radiofrequency of the ophthalmic branch of the trigeminal nerve (COSMAN Radiofrequency Therapy Apparatus, USA) at 42°C for 8 min	69 ± 15 years, 37.5% F	VAS Pre-treatment: all VAS scores ≥4 Post-treatment: 2 people with pain ≥4 at each time point	Recurrence of burning pain requiring additional medical therapy (n = 2; 6.25%), no loss of corneal reflex in any patient
Small et al., 2019	Retrospective cohort	variable	19 (8 gabapentinoid; 11 with nerve blocks 10 of which were refractory to gabapentinoid)	Treated in the oculofacial pain clinic with ICD10 code H57.XX and neuropathic corneal pain	Gabapentin 400mg daily-1200mg three times daily, pregabalin 150mg twice daily Supraorbital, supratrochlear, infratrochlear, infraorbital nerve blocks: 4 mL of 0.5% bupivacaine mixed with 1 mL of 80 mg/mL methylprednisolone acetate and 0.5–2mL injected at each site. Greater occipital added if pain over that nerve and pterygopalatine added if all ineffective	Gabapentinoid: 46 ± 14.4 years, 50% F Nerve block: 54 ± 19.8 years, 36% F	NRS Gabapentinoid: 2 complete resolution, 3 ≥ 50% improvement, 1 <50% improvement, 2 no improvement Nerve blocks: 7 improved, 4 no change	Not reported
Patel et al., 2021	Retrospective cohort	Variable means reported (range 266 to 838 days)	101	I: Treated in the oculofacial pain clinic with ICD10 code for H57.XX and presence of one or more pain features: sensitivity to wind and light, symptoms>signs, abnormal corneal sensitivity, persistent pain after topical anesthetic, cutaneous periocular allodynia E: pain resolved with treatment of nociceptive sources	Oral, topical, and adjuvant (eg. TENS, nerve blocks)	55 ± 17 years, 64.4% F	Clinical improvement in pain (no change, mild, moderate, marked, resolved) Patients with migraine-like pain frequently experienced pain improvement, while postsurgical patients had the lowest response rates. Patients with a longer follow-up and who tried more therapies experienced more significant relief	Not reported
Lee et al., 2023	Retrospective cohort	Up to 20 weeks (visits every 2 weeks, up to 10 visits)	204	I: Ocular neuropathic pain despite at least 3 months of ocular surface management E: contraindications to nerve blocks	Greater occipital nerve block with 2mL of 0.4% lidocaine solution injected on both sides	62.5 years, 86.3% F Median (IQR) NRS pre-treatment= 5 (4,7)	Estimated pain reduction per treatment = −0.55, p < 0.001 when considering pain =10 on NRS at baseline	Recurrence of ocular neuropathic pain (n = 2); neuritis (n = 4); transient facial hypoesthesia (n = 2); skin irritation (n = 1)
Mehra et al., 2021	Retrospective cohort	3–6 months	18	I: neuropathic pain (persistent pain after topical anesthesia symptoms>signs, improvement with in office use E: contraindications to TENS, use<3 months	Cutaneous trigeminal nerve stimulation of supratrochlear and supraorbital branches (Cefaly®, Cefaly Technology, Herstal, Belgium)	57.5 ± 14.5 years, 67% M	Mean±SD NRS over the past 24 hours Pre-treatment: 6.2± 2.1 Post-treatment (3 m) 4.4±3.2 (p=0.02) Post-treatment (6 m) 4.3±3.0 (p < 0.01)	Sedation with TNS use, persisting throughout follow-up (n = 15, 83%)
Olcucu et al., 2024	Retrospective cohort	60 seconds post-treatment	22	I: peripheral or mixed peripheral/ central neuropathic corneal pain defined as symptoms>signs, refractory to conventional therapy, history of a corneal nerve lesion E: No change in pain with proparacaine (presumed completely central), contraindication to nerve stimulation	Extranasal neurostimulation	48.9 ± 3.3 years, 77.3% F	Mean±SD VAS Baseline: 5.59±0.41 Endpoint: 2.59 ±0.43 (p<.001)	Not reported
Aggarwal et al., 2019	Retrospective case-control	3.8 ±0.5 months (range 1–8 months)	28 (16 cases; 12 age- and sex-matched controls)	I: Severe corneal pain (≥7/10), refractory to all previous treatment E: Confounding ocular diseases (e.g., uveitis)	Autologous serum tears 20%	Cases: 61.8 ± 4.4 years, 63% F Controls: 56.4 ± 3.1 years, 50% F	Mean±SD VAS Baseline: 9.1±0.2 (range 8–10) Endpoint: 3.1±0.3 (range 0–4), (p<.001)	Not reported
Yoon et al., 2021	Prospective, pre-post design	1 month	15	I: Dry eye disease with neuropathic ocular pain features (defined as pain≥4, unresponsive to topical treatment, symptoms>signs)	Topical TRPM8 agonist, cryosim-3, a TRMP8 agonist, application to the eyelid 4 times daily in addition to treatment for dry eye disease	59.5 ± 13.0 years, 60% F	Mean±SD OPAS Baseline: 30.60±12.84 1 month: 21.53±10.84 (p= 0.015)	Not reported
Ongun et al., 2021	Nonrandomized prospective cohort	6 weeks	72; 36 each in 2 groups	I: Severe DED according to OSDI score, Schirmers <5mm, TBUT <5seconds. Inclusion into group 2 if “neuropathic component” E: Confounding ocular diseases (e.g., uveitis)	Group 1: artificial tears + topical cyclosporine Group 2: artificial tears + topical cyclosporine + oral gabapentin titrated up to 1800–2400 mg/day	Group 1: 43.2 ± 6.3 years, 55.6% F Group 2: 46.1 ± 6.9 years, 58.3% F	Mean±DS OSDI: Baseline/Endpoint Group 1 70.11±18.04 49.41±16.70 (p<.001) Group 2: 66.82±16.10 31.13±11.52 (p<.001) Between group differences at 6 weeks, p<.001	Not reported
Reyes et al., 2023	Prospective, pre-post design	4–6 weeks	12	I: Pain and light sensitivity suggestive of neuropathic ocular pain. E: Confounding ocular diseases (e.g., uveitis)	Onabotulinum toxin A (35 units total: 5units in procerus, 10 units in the corrugators, and 20 units in the frontalis)	Mean age (SD): 53.5 (9.6) years; 33% (n=4) male	Mean±SD NRS 1week recall: Pre=6±2.5, Post=4.1±1.7, p=0.06 Mean±SD NPSI-Eye total: Pre=42.5±17.9, Post= 33.5±14.6, p=0.04 Mean±SD DEQ5: Pre=15.8±4.3, Post=14.6±4.1, p=0.22 Mean±SD OSDI total: Pre=89.6±16.7, Post: 62.54±25.1, p=0.53	Not reported
Hamrah et al., 2024	Randomized controlled trial protocol	12 weeks	48	I: neuropathic corneal pain symptoms; reduced corneal nerve density with microneuromas on IVCM E: Confounding ocular diseases, <50% proparacaine response	Topical OK-101 ophthalmic solution (0.05% or 0.1%)	Not reported	VAS, primary outcome but results not reported	Not reported
Galor et al., 2026	Randomized controlled trial	8 weeks	31	I: Age 22–90 years, chronic ocular surface symptoms ≥3 months with VAS ≥4 E: Allergy to device components, corneal disease	Novel topical cooling device (ETX-4143) applied to the anesthetized eye for 4 minutes	Mean age (SD) 62.3 (17.5) years, 80% female in invervention group	Mean change-from-baseline ±SD COP-Q: Intervention=- 5.24 ± 2.46 vs. sham − 2.86 ± 2.27, nominal p = 0.04	70% of the intervention and 36.4% of the sham cohort reported an AE. Transient conjunctival hyperemia was most common with the intervention. No serious AEs
Felix et al., 2026	Randomized controlled pilot trial	6 months	37	I: Average eye pain intensity ≥4/10; stable medication regimen ≥3 months; naïve to TENS; neuropathic-like ocular pain characteristics E: Confounding ocular diseases, contraindications to TENS	High frequency TENS using the Cefaly® device 3 times/weekly for 6 months compared to low frequency TENS	Mean age (SD) 58 (12) years, 51% female	Mean NRS within 24 hours: significant reduction in both treatment groups Mean NRS baseline compared to 3 months and 6 months in both treatment groups: no change Mean (SD) NPSI-Eye subscores of pressing pain (2.50 [5.50] to 1.50 [2.50], p = .03) and paroxysmal pain (1.50 [3.50] to 0.00 [1.50], P = .02) reduced in the high frequency TENS group only at 3 months, but not 6 months	19 with reported adverse event. Headaches or migraines (n = 10), other noted adverse effects included forehead rash, forehead pain and numbness, ear sensitivity, and head “fullness” or an unbalanced sensation

Abbreviations: IQR = interquartile range; SD = standard deviation; N = number of participants; VAS = visual analog scale (0–10 range); TENS = transcutaneous electrical nerve stimulation; OSDI = Ocular Surface Disease Index (0–100 range); TBUT = tear break-up time; DEQ5 = Dry Eye Questionnaire 5 (0–22 range); COP-Q, Chronic Ocular Pain Questionnaire (NRS = Numerical Rating Scale (0–10 range); FCS = Fluorescein corneal staining; F=Female, M=male; IVCM = in vivo confocal microscopy; BCDVA = best-corrected distance visual acuity; I= inclusion criteria; E= exclusion criteria

Pain outcome measures varied across the studies (Table 1). Validated measures included the visual analogue scale (VAS), numeric rating scale (NRS), ocular surface disease index (OSDI), ocular pain assessment survey (OPAS), neuropathic pain symptom inventory adapted to the eye (NPSI-Eye), Dry Eye Questionnaire 5 (DEQ5), and pain relief scale. Some studies used subjective ratings (e.g., mildly improved). All studies reported an improvement in pain or ocular symptom burden with either one or a combination of interventions tried. Reporting of adverse events occurred in 8 (40%) studies.^{16,17,19–21,23,24,30} The remaining studies did not comment on whether any adverse events occurred.

Discussion

Our review revealed few studies of pharmacologic and non-pharmacologic interventions to treat COP since 2015. We observed an average of 2 studies annually although no studies were published in 2017, 2022, or 2025. Nearly all published studies were retrospective cohort studies or case reports and one was a randomized controlled pilot trial of high-frequency TENS.³⁰ A second randomized controlled clinical trial has been completed for OK-101, but the results have not been published.³¹ The other prospective studies to date have assessed gabapentin,²⁸ cryosim-3 applied to the eyelids,²⁷ and onabotulinum toxin type A²⁹ all reporting an improvement post-intervention. Given the burden of COP in the population, investment in rigorous research studies across the therapeutic research pipeline are greatly needed.

Few interventions have been assessed in COP compared to other chronic pain conditions. Current guideline recommendations on best practices for chronic pain management emphasize the use of both pharmacologic therapies and non-pharmacologic therapies as part of a multimodal therapeutic strategy.³⁵ Non-pharmacological interventions can include lifestyle interventions to promote exercise and sleep, as well as psychotherapy given the high rates of comorbid mood disorders among people with chronic pain. In studies of other chronic pain conditions, non-pharmacologic therapies can be as effective as pharmacologic interventions although notably, long term data are often lacking.^36–38 None of the studies we examined used psychotherapy or lifestyle interventions. The only non-pharmacologic interventions investigated in COP are neuromodulating devices, such as TENS and other nerve stimulators or radiofrequency procedures and a topical cooling device, which is not commercially available. Except for two randomized trials, all were case reports or retrospective cohort studies. In the randomized pilot study, a significant reduction in pain intensity was found in the first 24 hours in both the high and low-frequency TENS groups, but these effects were not sustained at 3 and 6 months although the high-frequency group did experience a significant reduction in paroxysmal pain and pressing pain as measured by single questions in the NPSI-Eye.³⁰ In general, these types of interventions are hypothesized to work by blocking the pain signals from peripheral nociceptors from reaching the CNS or by stimulating the release of endorphins and other neurotransmitters that activate opioid receptors.^39,40 Across other chronic pain conditions, improvement with these devices and procedures are mixed due to the lack of standardized protocols across modalities, challenges with blinding in clinical trials, and small sample sizes. Additionally, real world barriers to implementation include a lack of insurance coverage and high costs limiting widespread use.⁴⁰

Among pharmacologic interventions used in people with >3 months of COP, most studies have focused on gabapentin. Gabapentin and pregabalin are often considered a first-line therapy for chronic neuropathic pain and nociplastic pain (e.g., fibromyalgia), and chronic head and facial pain conditions (e.g., migraine) as it targets voltage gated calcium channels and reduces the release of excitatory neurotransmitters, thereby reducing abnormal signaling in the PNS and CNS. However, the effects are modest across most chronic pain trials with a number needed to treat of approximately 6–10 to obtain a 50% or more reduction in pain.⁴¹ In the only prospective study of gabapentin use in individuals with COP, 72 patients with dry eye disease (based on ocular surface signs) and neuropathic corneal pain (defined using painDETECT, a general neuropathic pain measure) were treated with artificial tears and cyclosporine ophthalmic.²⁸ Thirty-six participants were additionally treated with gabapentin titrated to 1800–2400mg daily. After 6 weeks, OSDI scores, which are a measure of ocular surface symptom burden inclusive of pain, were still considered moderate-to-severe but were overall reduced in both groups and significantly improved (both statistically and clinically) among those receiving gabapentin compared to the group that did not receive gabapentin (no gabapentin: 49.41±16.70 vs gabapentin: 31.13±11.52; p<.001). Notably, Schirmers test scores and tear break-up time were also significantly better in the gabapentin group compared to the non-gabapentin group, raising the possibility that the improvement in ODSI scores could have been driven by improvement in the ocular surface parameters rather than improvement from gabapentin.

Additional pharmacologic interventions have been assessed both retrospectively and prospectively largely targeting peripheral neuropathic mechanisms. One retrospective observational study investigated the use of greater occipital nerve blocks (2 mL of 0.4% lidocaine) every 2 weeks to treat 204 patients with “ocular neuropathic pain” and found a significant decrease in the pain relief scale per unit of treatment.²³ It is not clear how the diagnosis of ocular neuropathic pain was made making it difficult to draw conclusions. However, numerous, small retrospective studies used nerve blocks in other locations (supraorbital, supratrochlear, infratrochlear, infraorbital, pterygopalatine ganglion, retrobulbar) with at least mild relief lasting for variable durations. Similarly, botulinum toxin injections (all with inca- or onabotulinum toxin A) have been assessed in one prospective²⁹ and two retrospective studies.^15,18 The exact mechanism by which botulinum toxin works to relieve pain is unclear, but is believed to work in part via inhibition pain signaling peptides, such as calcitonin gene related peptide. In a small prospective study of 12 participants using a pre- post- design, onabotulinum toxin A was used in a modified migraine pattern without the use of occipital, paraspinal, or trapezius injections and demonstrated a significant reduction in NPSI-Eye scores (a peripheral neuropathic pain measure adapted to the eye) 4–6 weeks following injection (NPSI-Eye total: 42.5 ± 17.9 (n=12) vs 33.5 ± 14.6 (n=10); p=0.04).²⁹ The reduction in pain intensity was not significant and there was no significant reduction in light sensitivity, which was the primary outcome for the study. In a larger retrospective study of 27 individuals, 74% (n=20) reported an improvement in pain as measured by a subjective pain report of degree of change.¹⁸ Most improvements were perceived as moderate or marked.

Other medications that are among the most prescribed medications for chronic pain, such as serotonin norephinephrine reuptake inhibitors (SNRIs) and tricyclic antidepressants (TCA), have not been studied in well-defined populations of COP with pain for at least 3 months. A recent Cochrane review and network meta-analysis ultimately concluded that two SNRIs, duloxetine and milnacipran had the best evidence for the treatment of chronic pain, although many drugs lacked enough high-quality studies for any definite conclusions.⁴² Norepinephrine and serotonin may work by decreasing pain signaling in descending pain pathways and reducing inflammatory cytokines, not by acting primarily as antidepressants.⁴³ These medications appear to be effective in patients with neuropathic or nociplastic pain and may have a role in the treatment of COP but studies are needed.

Unique to COP is the use of autologous serum tears for patients with a presumed neuropathic component. Autologous serum tears are made from an individual’s serum per institutional and U.S. Food and Drug Administration guidance. Serum contains nerve growth factor and neurotrophic factors, which are hypothesized to promote neuron survival and axonal sprouting in the cornea. In a retrospective case control study of 16 cases with neuropathic corneal pain (defined as pain>7/10, refractory to standard topical ophthalmic treatment, and no confounding ocular diseases) and 12 sex and age matched controls, VAS scores significantly improved in the cases (baseline 9.1 ± 0.2 vs endpoint (mean 4 months) 3.1±0.3; p<.001) as did the features on corneal confocal microscopy.³³

Aside from the lack of prospective studies, other substantial limitations exist throughout the current literature. First, many terms are used to describe neuropathic/nociplastic COP, such as neuropathic corneal pain, but these terms have not been codified and clear diagnostic criteria have not been delineated. Furthermore, mechanistic terms, such as “neuropathic” pain in particular, are not consistently used in the ophthalmic literature in accordance with the IASP definitions. Instead, it is often used to describe “centralized” pain whether due to nerve injury or dysfunctional sensory processing. Second, pain outcomes vary widely across studies and often do not follow recommendations from the pain consortia on pain outcomes. The Initiative on Methods, Measurement, and Pain Assessment in Clinical Trials and guidelines from the Veterans Health Administration Pain Measures Work Group recommend that pain outcomes should include the assessment of both pain intensity and interference. VAS and NRS scales were most often deployed but often provide only a snapshot of pain intensity at the moment, which may not reflect pain across an entire day or week. Third, the adverse events and side effects of treatments are frequently not assessed in COP treatment studies. Side effects of treatments are important to understand to support shared decision making and ensure that treatments do not worsen quality of life. Side effects can also to be leveraged to treat a concurrent issue. For example, for a patient that is struggling with sleep disturbance, a TCA at bedtime can be helpful in improving sleep in addition to pain. Last, it is not surprising that most of the published literature shows an improvement in COP with the interventions studied given the nature of the study designs. However, it is important to note that even in randomized controlled trials, placebo responses in chronic pain studies can be substantial accounting for approximately 50% of the treatment response in some trials.⁴⁴ Careful planning of trials with run in periods to assess for regression to the mean, functional outcomes (not just self-report), and scripted interactions with patients are just a few of the strategies to consider when planning COP studies. Based on the results of the current review and limitations, we have provided recommendations for future studies (Table 2) and our current approach to the treatment of patients with COP arising entirely or in part from underlying neuropathic or nociplastic mechanisms (Figure 2).

Table 2.

Recommendations for Future Studies

1Randomized clinical trials of multimodal therapeutics are needed in chronic ocular pain. Selecting therapies that have proven effective in the management of other primary headache (e.g., migraine) and neuropathic pain conditions (e.g., small fiber neuropathy), may be beneficial.

2Definitions and criteria for ocular pain subtypes should be standardized and published to allow for comparison across treatments and improve translation to clinical care.

3Outcome measures should include pain intensity and pain interference as recommended by Initiative on Methods, Measurement, and Pain Assessment in Clinical Trials, as well as quality of life outcomes. This will not only ensure that our treatments lead to improvements in pain and functioning but also allow for comparison across treatments and pain conditions.

4Adverse events and side effects should be assessed and reported in all future studies regardless of study design.

Figure 2.

Principles of treating chronic ocular pain from neuropathic or nociplastic mechanisms

While this scoping review provides a comprehensive overview of the current literature on COP several limitations should be acknowledged. First, we only included English language studies and thus, might be missing studies in the non-English language medical literature. Second, our review does not offer quantitative synthesis, limiting the ability to assess effect sizes or statistical associations. The paucity of studies prevented us from conducting a systematic review and the inclusion of a wide range of study designs introduces heterogeneity, making synthesis challenging. Third, given the design of our study as a scoping review, we did a qualitative assessment of study quality across studies but did not conduct a formal quantitative appraisal of study quality. Finally, scoping reviews are designed to map evidence rather than to draw definitive, hypothesis-driven conclusions, and thus findings should be interpreted in that context.

Despite the fact that COP is a leading reason people seek eye care and are commonly referred to neuro-ophthalmology, research on pharmacologic and non-pharmacologic interventions for COP remains limited, with a notable absence of randomized controlled trials. Neuromodulatory devices, while promising in theory, have been evaluated primarily in small, uncontrolled studies, and lifestyle and psychotherapeutic interventions—key components of multimodal pain management for other chronic pain syndromes—have not been explored in COP. Among pharmacologic therapies, gabapentin remains the oral agent studied most in clearly defined COP populations, and evidence for other commonly used pain medications is notably lacking. Furthermore, heterogeneity in pain outcome measures, variably applied diagnostic criteria for COP, and poorly reported adverse events and side effects limit the generalizability of existing data. These gaps underscore the need for rigorous, prospective studies, standardized definitions and outcome measures, and a broader evaluation of both pharmacologic and non-pharmacologic strategies in COP.

Abbreviations:

IQR

interquartile range

SD

standard deviation

N

number of participants

VAS

visual analog scale

TENS

transcutaneous electrical nerve stimulation

OSDI

Ocular Surface Disease Index

TBUT

tear break-up time

DEQ5

Dry Eye Questionnaire 5

NRS

Numerical Rating Scale

FCS

Fluorescein corneal staining

IVCM

in vivo confocal microscopy

BCDVA

best-corrected distance visual acuity

P

pharmacologic

N

neuromodulatory / nerve procedures