A retrospective analysis of the medical records of 65 DRS patients who visited the department of pediatric ophthalmology at Seoul National University Children's Hospital between 2010 and 2017 was carried out. Diagnosis of DRS was based on the finding of horizontal gaze deficit(s) together with one or more of the following characteristics: (1) palpebral fissure narrowing, (2) globe retraction, and (3) excessive upshoot and/or downshoot. Patients with history of prior ophthalmic surgery, ocular trauma, or orbital disease and those without a record of angle of strabismus at primary gaze were excluded from this study.

An analysis of epidemiological data (sex, age at diagnosis, laterality of the involved eye, strabismus surgery performed after diagnosis, and other systemic or central nervous system diseases) and clinical findings (angle of deviation at primary gaze, upshoot, downshoot, palpebral fissure-narrowing, globe retraction, abnormal head position, refractive status, and stereoacuity) was conducted. The angle of deviation was measured using the alternate prism cover test, and the refractive status was evaluated with cycloplegic refraction. Also, stereoacuity was assessed with the Worth four-dot test and Randot stereotest.

In this study, we propose a new classification system for DRS that sorts patients into three types according to the following criteria: patients whose angle of exotropia at primary gaze exceeded 3 prism diopters (PDs) were classified as “Exo-Duane,” patients whose angle of strabismus at primary gaze did not exceed 3 PDs were classified as “Ortho-Duane,” and patients whose angle of esotropia at primary gaze exceeded 3 PDs were classified as “Eso-Duane.” The numerical criteria of 3 PDs is an arbitrary value. The standard was set within a narrow range since most patients with DRS present with a small angle of deviation at primary gaze less than 10 PDs, and abnormal head posture is concurrent even with a small angle of deviation; moreover, globe retraction, upshoot, or downshoot are often accompanied by only a small amount of ocular movement. Angle of strabismus was the measurement value at first medical examination.

Regarding stereoacuity, patients with DRS with stereoacuity of 100 arcseconds or less were considered to have good stereoacuity, while those with stereoacuity greater than 100 arcseconds were considered to have poor stereoacuity.

Strabismus surgery was indicated for cosmetic purposes in patients with large face turn, strabismus in the primary position, excessive upshoot or downshoot in adduction, or severe globe retraction.

Statistical analyses were performed using descriptive statistics, Fisher's exact test, the Kruskal-Wallis test, and/or the Mann-Whitney U-test where appropriate to compare results among the three DRS types. Also, analysis of covariance was used for age adjustment, and the Wilcoxon signed rank test was applied to compare preoperative and postoperative stereoacuity. All analyses were performed using the IBM SPSS Statistics ver. 23.0 (IBM Corp., Armonk, NY, USA), and comparisons were regarded as statistically significant for p-values less than 0.05. Bonferroni correction was applied for group-wise comparisons of each subtype. This study was approved by the institutional review board of Seoul National University Hospital (1907-184-1051) and the study protocol followed the tenets of the Declaration of Helsinki. Informed consent was waived due to the retrospective nature of the study.

In our study, Ortho-Duane comprised the largest proportion of patients (53.8%) among the three subgroups, followed in order by Eso-Duane (33.8%) and Exo-Duane (12.3%), which is a different finding from previous reports. O'Malley et al. [6], Shauly et al. [7], and Lee and Chang [8] instead reported that the majority of patients had esotropia (or esodeviation), and decreasing proportions of patients had orthotropia and exotropia (or exodeviation) in order, whereas Isenberg and Urist [9] and Park et al. [10] reported that orthotropia was the most common group followed equally by exotropia and esotropia. The differences in results among the studies may be due to the arbitrary criteria used to define the angle of strabismus in our study as compared to the other studies. Regarding average refractive errors, no statistically significant differences among the subgroups after age adjustment were observed, with a partial eta squared (η²) value of 0.056, which means that age had a small effect on refractive error in our patients.

Further, we also did not detect any sex preponderance, which is similar to the findings of previous studies on Korean patients with DRS by Lee and Chang [8] and Park et al. [10] but inconsistent with those reporting a female preponderance among non-Korean patients with DRS [5111213]. The disparity of results among research groups may be related to demographic differences among the patients involved in each study and also be attributable to the small number of patients included in our study.

We observed in our cohort that unilateral cases were more frequent than bilateral cases, and the left eye was more commonly involved than the right eye, findings similar to those of previous studies [5111213]. Bilateral DRS accounted for 4.6% of all patients, which was less than the range of 10% to 24% of all DRS cases noted in previous reports [111214]. This difference in results between our study and others may be due to the limited number of patients in our study, as mentioned above.

In our study, 46.2% of all patients showed abnormal head position, with 38.5% of patients showing head turn, which is less than the proportion of 50% of patients reported by Lee and Chang [8] and that of 54.6% reported by Kekunnaya et al. [13] but more than that of 16.7% reported by Park et al. [10]. With regard to abnormal head positions, head turn was more frequent than head tilt (38.5% vs. 3.1%). Since horizontal deviation rather than vertical deviation such as hypertropia at primary position is more common in patients with DRS, compensatory head posture which is mainly head turn rather than head tilt may occur more frequently. In each subgroup, the proportion of patients with normal head posture was higher in the Ortho-Duane group than in the Exo-Duane or Eso-Duane group, though the former comparison was without a statistically significant difference (p = 0.028 and p < 0.001, respectively). Conversely, the proportion of patients with head turn was higher in the Exo-Duane or Eso-Duane group than in the Ortho-Duane group, with no significant difference for the former comparison (p = 1.000 and p < 0.001, respectively). These differences in the proportion of patients in regard to head posture among the groups could also be explained by the compensatory mechanism used to achieve binocular vision in the same context.

The proportion of patients accompanying hypertropia at primary gaze was 50.0% in the Exo-Duane group, which was largest, followed by 4.5% of the Eso-Duane and 2.9% of the Ortho-Duane groups. Although patients with DRS are known to have no vertical movement deficits, this result indicates that hypertropia at primary gaze could present as a mild form of upshoot in Exo-Duane and Eso-Duane groups.

With regard to refractive error, hyperopia was the most common, found in 67.2% of patients. This result is in agreement with those of most preceding studies [11]. However, the results of analysis by subtype revealed that myopia was most common in Exo-Duane group, whereas hyperopia was most common in Ortho-Duane and Eso-Duane groups. Since only seven patients were included in the Exo-Duane subgroup, generalization of these results to all DRS patients has some limitations.

In our study, 93.1% of patients had gross stereoacuity and 82.1% had good stereoacuity, which is consistent with the results of former studies that reported DRS patients have good stereoacuity since they try to maintain fusion with compensatory head posture [2]. However, the comparison of stereoacuity between subgroups showed no statistically significant differences. All seven patients with available postoperative stereoacuity data showed gross stereoacuity, and 71.4% of them had good stereoacuity after surgery; nevertheless, it was difficult to evaluate whether surgery was a significant influencing factor for improvement of stereoacuity since records of stereoacuity both before and after the surgery were available in only two patients and evaluation of stereoacuity was not possible due to the young age of some of the patients. In the two patients with available data, mean stereoacuity was improved without statistical significance after surgery (p = 0.317). Further studies with long-term follow-up periods and larger numbers of patients are required to better evaluate the effect of strabismus surgery on improvement of stereoacuity.

DRS is one of congenital cranial dysinnervation disorders, which are congenital, nonprogressive, sporadic, or familial developmental anomalies of the cranial nerves characterized by abnormal eye, eyelid, and/or facial movements [15]. The traditional Huber's classification proposes three subtypes of DRS according to EMG pattern and corresponding ocular movement deficits. However, Huber's classification has several limitations. First, EMG studies are not feasible in all cases in practice. Second, diversity of ocular presentations and atypical EMG patterns, which arise from a wide range of pathophysiology of DRS, prevent ease of classification of the subtypes clinically. Third, Huber's classification does not encompass various characteristics of DRS such as vertical deviations and anomalous head positions [11]. Lastly, in Huber's classification, definite, or normal or mild adduction and/or abduction limitation(s) are somewhat subjective judgments that could differ among examiners; as such, a novel classification that enables better communication between ophthalmologists is required. In this regard, our study has strengths in that it proposed a clinically more convenient and objective new classification method of DRS using the angle of strabismus at primary position rather than Huber's traditional classification and analyzed various features of DRS. Moreover, to the best of our knowledge, our study is one of only a few studies to analyze demographics and characteristics of DRS in an Asian population except one patient, which allows for comparison of differences between Asian and Caucasian populations. Although there were several studies that classified DRS patients based on deviation at primary position [1617], our classification system is differentiated from them in that we used the criterion of 3 PD at primary gaze in classifying subgroups. Our study has limitations due to the relatively small number of patients included and because the data were collected through a retrospective analysis of medical records. Further studies analyzing changes of characteristics over time in larger populations and using magnetic resonance imaging may provide more information in understanding the pathophysiology of DRS.

In conclusion, the newly proposed classification of DRS according to type and angle of strabismus at primary gaze was practically useful and showed potential for validity as an objective guideline in the clinical setting.