The Specificity Rates of Clarus 500™ Ultra-Wide-Field Retinal Imaging for Detecting Peripheral Retinal Lesions in Medium-to-High Myopia Eyes Open Access

Myopia is a prevalent ocular condition that has become a significant global public health concern [1‒3]. Holden et al. [4] estimated that the number of myopic individuals (−0.30 diopters [D] or greater) would be projected to reach 4,758 million (49.8% of the population), and the number of individuals with high myopia (−5.00 D or greater) would be expected to reach 938 million (9.8%) by 2050 [3]. Due to progressive axial elongation of the globe, myopia, especially high myopia, may lead to substantial visual loss [3, 5]. Degenerative changes in the macula and chorioretinal atrophy elevate the risk of ocular complications such as retinal detachment, myopic macular degeneration, and myopic choroidal neovascularization [3, 5, 6]. Vongphanit et al. [7] investigated the evidence of myopic retinopathy and verified that higher myopia increased the risk of visual impairment [3, 8]. It is crucial to effectively treat myopia-related retinal pathologies.

Multimodal imaging techniques are being utilized for prevention and early diagnosis of various retinal diseases in the field of ophthalmology [9]. In addition to traditional tools such as slit-lamp examination using a non-contact lens and Goldmann three-mirror lens, new technologies including optical coherence tomography (OCT), OCT angiography, and multiple ultra-wide-field (UWF) color fundus imaging have been adopted, particularly for early screening.

Clarus 500™ is an advanced UWF fundus camera system that can capture images with a wide field of view, facilitating a more straightforward assessment of peripheral retinal abnormalities [9‒11]. Therefore, how does the consistency of Clarus 500™ compare with other methods, such as slit-lamp examination? Does its performance characteristics facilitate the detection of peripheral retinal lesions in medium-to-high myopic eyes? This study aimed to evaluate the consistency of Clarus 500™ in screening retinal issues in patients with medium-to-high myopia. A comparative analysis will be conducted to assess the sensitivity and specificity of Clarus 500™ imaging system in myopic eyes.

This was single-center, diagnostic test. The study collected data from 221 patients, encompassing 436 eyes. All enrolled patients were aged between 18 and 50 years old. The refractive parameters of the patients were determined through subjective refraction under cycloplegia (Nidek RT-5100). The axial length of the patients was examined by IOLMaster (Carl Zeiss). Intraocular pressure (IOP) was initially measured using the non-contact air-puff tonometer (Tonometer-10, Canon, Tokyo, Japan). When non-contact air-puff tonometer-measured IOP exceeded 21 mm Hg, Goldmann applanation tonometer was used for remeasurement, and the remeasured values were below 21 mm Hg.

Data were collected from 221 myopic patients (185 females, 83.71%; 36 males, 16.29%), including 436 eyes (220 right, 216 left). The mean age was 28.9 ± 6.0 years (ranging from 18 to 49 years). The average spherical equivalence refraction was −8.33 ± 2.81 D (ranging from −20.63 D to −0.50 D). The average axial length was 26.60 ± 1.24 mm (ranging from 23.79 to 32.31 mm). The average IOP of study eyes was 15.05 ± 2.78 mm Hg (ranging from 8.30 to 24.70 mm Hg). The detailed demographics and ocular characteristics of the patients are summarized in Table 1.

Another skilled masked ophthalmologist used a slit-lamp non-contact +120 D lens for the retinal assessment. Prior to examination, the patient’s pupils were completely dilated after the administration of tropicamide phenylephrine eye drops (0.5% tropicamide and 0.5% epinephrine; Santen Pharmaceutical, Japan). Patients were instructed to gaze in nine cardinal directions to facilitate a comprehensive examination. Each retinal degenerative lesion and its corresponding location were meticulously documented for further analysis.

After pupil dilation, an experienced imager carried out the image-collection procedure. Mydriatic UWF fundus images were obtained at different gaze positions (front view, superior temporal, superior nasal, inferior temporal, and inferior nasal gazes) using an automated UWF imaging system.

Figure 1 shows an automated collage. The images were exported as JPEG files with a resolution of 6,604 × 4,274 pixels for subsequent analysis. Images deemed of inadequate quality were retaken. A senior retinal specialist independently conducted an analysis of all the images. We categorized the retinal lesions into three main groups: retinal degeneration, retinal holes/tears, and retinal hemorrhage (online suppl. Fig. 1; for all online suppl. material, see https://doi.org/10.1159/000546494).

First, we conducted a descriptive analysis of the baseline characteristics of all patients. Continuous variables were expressed as mean ± standard deviation, while categorical variables were represented as n (%). Next, to evaluate the diagnostic effectiveness of Clarus 500™, we calculated the sensitivity and specificity as follows: sensitivity = true positive/(true positive + false negative), specificity = true negative/(true negative + false positive). To assess the agreement between the slit-lamp examination with non-contact +120 D lens and Clarus 500™, the observed agreement proportion (i.e., the number of eyes for which both assessments agreed) and Cohen’s kappa coefficient along with its 95% confidence interval (95% CI) were utilized. According to the guidelines provided by Landis and Koch [12], the interpretation of the kappa coefficient is as follows: poor if kappa < 0, slight if 0–0.20, fair if 0.21–0.40, moderate if 0.41–0.60, substantial if 0.61–0.80, and almost perfect if 0.81–1.00. All statistical analyses were performed using R, version 4.3.0 (R Program for Statistical Computing), and the two-sided cutoff value of statistical significance was set at p < 0.05.

Retinal lesions were detected in 13/436 (2.98%) participants via mydriatic slit-lamp examination with non-contact +120 D lens, and in 5/436 (1.15%) participants through mydriatic examination with the Clarus 500™ examination. Here, 10/13 (76.92%) and 4/5 (80%) had retinal degeneration. Following slit-lamp examination with non-contact +120 D lens, lattice degeneration was found in 10 eyes (2.29%), retinal holes/tears in 2 eyes (0.46%), and retinal hemorrhage in 1 eye (0.23%). After examination with Clarus 500™, lattice degeneration was observed in 4 eyes (0.92%), and retinal holes/tears were observed in 1 eye (0.23%).

More than one treatable peripheral lesion was noted in 1 eye (0.23%) (online suppl. Fig. 2). Refer to Figure 2 for detailed information.

In comparison with slit-lamp examination with non-contact +120 D lens, Clarus 500™ exhibited a specificity of 99.77% (95% CI: 98.70%, 99.99%) for peripheral lattice degeneration, 100% (95% CI: 99.15%, 100%) for peripheral retinal holes/tears, and 100% (95% CI: 99.16%, 100%) for retinal hemorrhage. The concordance for lattice degeneration (kappa = 0.42, 95% CI: 0.10, 0.75) and retinal holes/tears (kappa = 0.67, 95% CI: 0.05, 1) diagnosis showed higher consistency (both p < 0.01). The detailed results for the sensitivity, specificity, and kappa values per category are shown in Table 2.

To efficiently detect and evaluate peripheral retinal holes and degeneration in myopic patients, utilization of advanced technological products is crucial. This study employed Clarus 500™ imaging system and slit-lamp with non-contact lens as a preoperative assessment tool for myopia surgery, demonstrating significant practical utility.

In modern outpatient settings, aimed at aligning more closely with clinical practice, this study employed a conventional non-contact lens as the benchmark to assess the efficacy of the UWF retinal imaging system in detecting retinal lesions. Our findings revealed that UWF retinal imaging with Clarus 500™ exhibited moderate and substantial consistency in identifying lattice degeneration and retinal holes/tears among myopic patients.

We analyzed the differences between the two detection methods. Upon analysis, it was observed that using slit-lamp examination with non-contact +120 D lens revealed more peripheral retinal degenerations. Among them, 7 eyes (77.78%) showed more peripheral lattice degeneration, 1 eye (11.11%) showed more peripheral retinal holes/tears, and 1 eye (11.11%) showed more peripheral hemorrhages (refer to online suppl. Fig. 3). The detailed images are shown in online supplementary Figure 4.

The auto-montage of the four images from Clarus 500™ may have certain limitations. First, the retina has a distinctively curved structure. Most myopia cases are linked to excessive axial elongation, which can potentially lead to the development of posterior staphyloma or ocular deformities [13]. During the peripheral image acquisition process with Clarus 500™, images could be distorted, leading to blurred areas and unclear edge lesions. Second, the successful capture of peripheral images requires precise eye positioning over a specific duration. Any eye movement that occurs during this process can result in blurred peripheral images. Moreover, the study’s predetermined shooting angles may have led to overlooking certain areas or underrepresenting peripheral retinal lesions, as illustrated in Figure 3.

Slit-lamp examination with a non-contact +120 D lens offers advantages, such as non-contact operation, ease of use, strong stereoscopic perception, clear imaging, wider visual range, and strong penetration ability [14, 15]. However, it requires skill from clinicians. Clarus 500™ enhances the quality and comfort of patients’ medical services. It provides a comparable field of view, saves time, and offers true-color images that are crucial for screening for retinal diseases, including peripheral lesions.

Our research also found that Clarus 500™ demonstrated high specificity, with a specific rate of 99.77% (95% CI: 98.70%, 99.99%) for peripheral lattice degeneration. A previous study on routine patients found that Clarus 500™ UWF imaging yielded results consistent with those of standard slit-lamp examination and detected more retinal findings [16]. Highly myopic patients may benefit from using non-mydriatic optomap UWF imaging to identify potential peripheral retinal lesions, which is another UWF retinal imaging system [17]. If time saving and image quality are prioritized, especially for screening retinal diseases, Clarus 500™ may be more suitable.

Previous studies showed that UWF imaging missed nearly half of horseshoe retinal tears, which means it alone lacks the sensitivity to rule out the presence of horseshoe retinal tears [18]. We analyzed the quadrants where the Clarus 500™ failed to detect lesions (as shown in Table 3), which disclosed its possible limitations in specific parts of the retina. This finding is consistent with the understanding that no single imaging method can detect all retinal lesions. To enhance the detection of lesions, we recommend multimodal imaging. Combining the Clarus 500™ with OCT or fundus fluorescein angiography offers complementary data, compensating for individual tools’ drawbacks and enabling more accurate diagnoses and informed clinical decisions [19].

This study has several limitations. First, the sample size was relatively small, and the number of patients with retinal lesions was limited. Second, the analysis was confined to peripheral lattice degeneration and peripheral retinal holes/tears, without inclusion of other retinal abnormalities. Additionally, we were unable to draw conclusions on the impacts of factors like axial length and quadrant on the Clarus 500™ detection rate. Future research conducted in retina clinics, where a higher proportion of patients present with retinal pathologies, may enable a more comprehensive evaluation of Clarus 500™ [19].

Clarus 500™ demonstrates high specificity in detecting peripheral retinal holes/tears and exhibits consistent reliability when compared with slit-lamp examination using non-contact +120 D lens in evaluating lattice degeneration.

The authors thank the patients who joined the study and the staff in the Eye and ENT Hospital of Fudan University for their support.

This study was approved by the Ethical Committee of Eye and ENT Hospital of Fudan University Review Board (2024097), and the study was conducted in accordance with the tenets of the Declaration of Helsinki. Written informed consent was obtained from all patients in the study.

The authors have no conflicts of interest to declare.

This work was supported by the National Natural Science Foundation of China (Grant No. 82371091, 82201207), Shanghai Rising-Star Program (Grant No. 21QA1401500), and Natural Science Foundation of Shanghai Municipality (Grant No. 23ZR1409200).

Fangcheng Xu and Xiaojun Hu: conceptualization, data curation, formal analysis, investigation, methodology, validation, and writing – original draft, review, and editing; Yunzhe Wang, MD, and Ruoyan Wei, MD: data curation, and writing – review and editing. Yongfu Yu, MD, PhD, and Yuwei Peng: data curation, methodology, validation, and writing – review and editing. Meiyan Li, MD, PhD, and Haixiang Wu, MD, PhD: conceptualization, writing – review and editing, providing funding, and supervision of the project.

Fangcheng Xu and Xiaojun Hu contributed equally as first authors.

All data generated or analyzed during this study are included in this article. Further inquiries can be directed to the corresponding author, E-mail: [email protected]; [email protected].