**3. Results**

This study included 35 patients with DE accompanying NOP features. The mean age was 55.6 ± 11.7 years, and 27 patients (77.1%) were women. There were no differences in baseline characteristics according to sex (data not shown). Only 1 of the 24 patients treated with gabapentin experienced a side-effect (mild tremor).

Table 1 shows the demographics and medical history of the patients enrolled in this study. There was no history of systemic comorbidity, ocular surgery, and trauma in group 1. Systemic comorbidities including rheumatologic, neurologic, and phycological disorders were more frequent in group 2 than in group 3 (*p* = 0.034). No differences were observed between the previous ocular histories of groups 2 and 3.


**Table 1.** Demographics and personal history of patients with dry eye accompanied by neuropathic ocular pain features.

M, male; F, female; LASIK, laser-assisted in situ keratomileusis; LASEK, laser-assisted sub-epithelial keratectomy. Compared using Fisher's Exact test.

Table 2 presents the comparisons of the clinical DE and MG parameters of the three groups. The CSS of groups 2 and 3 were lower than those of group 1 (*p* = 0.048 and *p* = 0.033). There were no differences between the CSS and other clinical parameters of groups 2 and 3.


**Table 2.** Clinical parameters in patients with dry eye accompanied by neuropathic ocular pain features according to the treatment response.

All values are presented as mean ± SD. OSDI, ocular surface disease index; TBUT, tear break-up time; CSS, corneal staining score; MG, meibomian gland. Comparison using one-way analysis of variance, and post hoc analysis using Tukey's honestly significant difference test with Bonferroni adjustment.

Table 3 shows the changes in the OPAS scores of groups 1 and 2. Improved OPAS scores of ocular pain severity and associating factors were noted after treatment in group 1 (both *p* = 0.026). Group 2 exhibited improved OPAS scores of ocular pain severity, pain other than eyes, and QoL (*p* = 0.011, *p* = 0.026, and *p* = 0.011, respectively).


**Table 3.** Changes in the Ocular Pain Assessment Survey score in the topical treatment and gabapentin response groups.

All values are presented as median (interquartile range). Comparing using Wilcoxon signed-rank test.

The comparisons of the OPAS scores between groups 2 and 3 are presented in Table 4. The scores of "enjoying life/relations with other people" associated with the QOL and pain aggravated by mechanical and chemical stimuli were lower in group 2 than in group 3 (*p* = 0.019, *p* = 0.003, and *p* = 0.004, respectively).


**Table 4.** Comparison of the Ocular Pain Assessment Survey score between the gabapentin response and non-response groups.

> \* Compared using independent t-test. † Compared using Mann–Whitney U test.

## **4. Discussion**

Pain is an unpleasant sensory and emotional experience associated with actual or potential tissue damage, or described with respect to such damage, and can be classified into nociceptive and neuropathic pain [26]. Nociceptive pain is caused by actual or threatened damage to tissue due to the activation of nociceptors. In contrast, neuropathic pain is caused by a lesion or disease of the somatosensory nervous system. Repeated peripheral nerve injury can lead to peripheral sensitization, and prolonged peripheral ectopic pain initiates central sensitization [13,26].

DE is a multifactorial disease of the ocular surface, which is accompanied by ocular symptoms [1]. At times, patients experience ocular pain that affects their QoL. The discordance between the ocular symptoms and signs sugges<sup>t</sup> an underlying neuropathic pain etiology; in such cases, ocular pain could be refractory to conventional topical DE treatment [5,8,13]. Gabapentin is the first-line treatment for systemic neuropathic pain in conditions such as fibromyalgia, postherpetic pain, and diabetic neuropathy [14]. It is an anti-convulsant drug that reduces the release of multiple excitatory neurotransmitters by acting on the α2δ subunit of the voltage-gated calcium channels, thus decreasing central sensitization [7,15].

However, limited data are available to support the use of systemic neuropathic pain medication for NOP associated with DE. A prospective, placebo-controlled study demonstrated that gabapentin reduced postoperative pain after photorefractive keratectomy [27]. However, a recent randomized pilot study showed that pregabalin, which has a similar mechanism with gabapentin, failed to prevent DE symptoms after laser-assisted in situ keratomileuses [28]. Ongun et al. [19] showed that gabapentin was more effective for the treatment of severe DE with NOP compared to topical treatment. In the present study, we aimed to analyze the differences in the clinical manifestations between groups according to treatment response in patients with DE accompanied by NOP features.

Our results showed that the frequency of other comorbidities such as rheumatologic, neurologic, and psychological disorders was higher in the gabapentin response group. Gabapentin has pharmacologic characteristics, binding to voltage-sensitive calcium channels at the α2δ subunit, affecting their function as well as influencing receptor trafficking [15]. It can secondarily influence

gamma-aminobutyric acid and glutamate tone and activity via this mechanism [15]. Therefore, gabapentin can relieve not only neuropathic pain but also general systemic symptoms, such as mood, sleep, vasomotor symptoms, etc. [15,29]. This explanation corresponds to the results seen in Table 4, i.e., the significant improvement in non-ocular pain and QoL.

In contrast, the topical treatment response group had no ocular history including surgery and trauma, with more severe CSS compared to the other groups. The sensory neurons of the ocular surface and nociceptors could actually be injured in patients who had ocular surgery and trauma leading to neuroinflammation associated with peripheral and central sensitization [5,8,13]. Topical anti-inflammatory agents such as topical steroids and cyclosporin could decrease the release of proinflammatory neuropeptides and cytokines from injured nerves, a ffecting nociceptive pain and peripheral sensitization [9]. However, improvement of corneal nerve morphologic status and central sensitization has not been demonstrated. Therefore, patients with previous ocular history, including surgery and trauma, may not respond to topical treatment and require systemic NOP treatment.

The evoked pain in response to chemical and mechanical stimulation tended to be greater in the gabapentin non-response group than in the gabapentin response group. Evoked pain, including allodynia and hyperalgesia, is provoked or increased pain in response to stimulation [26]. These manifestations are common clinical characteristics of neuropathic pain; however, their underlying mechanisms are complex and diverse depending on the provoking stimulus [30]. The e fficacy of gabapentin for the alleviation of systemic neuropathic pain was proven by several randomized, double-blind placebo-controlled studies [14]. Nevertheless, few studies have specifically focused on the treatment of evoked pain. Furthermore, one study showed that mechanical allodynia was a negative predictor of the overall e ffect of pregabalin in patients with postherpetic neuralgia [30,31].

Dieckmann et al. [9] suggested the proparacaine challenge test for di fferentiating between peripheral and central neuropathic pain and proposed a corresponding treatment strategy for DE patients with neuropathic pain etiology. The results of our study showed that the treatment response was related not only to the degree of central sensitization before treatment, but also to the patients' systemic comorbidity, ocular history, ocular surface status, and presence of evoked pain. Gabapentin can cause side e ffects including CNS depression (drowsiness, dizziness, headache, etc.) and mood problems; thus, consultation with a neuropsychiatrist may be needed [15]. In this study, only one patient experienced a mild tremor as a side-e ffect for gabapentin treatment. Therefore, we believe gabapentin may be tolerable in DE patients associated with NOP. In addition, our results will help clinicians predict in which DE patients associated with NOP gabapentin treatment will be more e ffective.

Our study had some limitations. First, it was designed retrospectively and the sample size was small. Patients were recruited from a single tertiary center; thus, the findings may not be representative of the general DE population. Further prospective and longitudinal studies with a large sample size are required in the future. Second, the threshold value for determining the group classification may not have been a representative value. We classified patients with an improvement of 2 or more points on the WBFPS as the treatment response group; however, this might be not a standardized cut-o ff value. Moreover, we did not analyze the results of the proparacaine challenge test. It was di fficult to classify the enrolled patients based on the central or peripheral phenotype of pain, since the majority of participants showed a mixed phenotype. Third, the extent of actual nerve damage was not measured using in vivo confocal microscopy. However, to the best of our knowledge, this was the first study to analyze the di fferences in the clinical manifestations between groups according to treatment response in DE patients with NOP features.

In conclusion, gabapentin could be successful for the treatment of DE patients with NOP features who have systemic comorbidities including rheumatological, neurological, and psychological disorders, and less evoked pain in response to mechanical and chemical stimuli. Topical treatment for DE with NOP features could be successful for patients who have a corneal staining and no ocular history including surgery and trauma.

**Author Contributions:** Conceptualization, K.C.Y.; methodology, H.-J.Y., J.K.; software, H.-J.Y.; validation, H.-J.Y., K.C.Y.; formal analysis, H.-J.Y.; investigation, H.-J.Y.; resources, H.-J.Y., J.K.; data curation, H.-J.Y., J.K.; writing—original draft preparation, H.-J.Y.; writing—review and editing, H.-J.Y., K.C.Y.; visualization, H.-J.Y.; supervision, K.C.Y.; project administration, K.C.Y.; funding acquisition, K.C.Y. All authors have read and agreed to the published version of the manuscript.

**Funding:** This study was supported by a gran<sup>t</sup> of the Korea Health Technology R&D Project through the Korea Health Industry Development Institute (KHIDI), funded by the Ministry of Health and Welfare, Republic of Korea (HR20C0021050020) and the Chonnam National University Hospital Biomedical Research Institute (BCRI 20072).

**Conflicts of Interest:** The authors declare no conflict of interest.
