Choroidal Thickness in Healthy Eyes Measured by Widefield Optical Coherence Tomography

The choroid is the vascular tissue localized between the retina and sclera and has an essential role in the maintenance of the ocular functions. The most important known functions of the choroid are retinal thermoregulation, light absorption, modulation of the intraocular pressure (IOP) and most relevant, the nourishment of the external retinal layers [1].

Until recent years, the study of the choroid has been challenging since it could not be properly visualized. Since Spaide et al. [2] introduced enhanced depth imaging (EDI) optical coherence tomography (OCT) technology, deeper structures were able to be observed. However, one of the limitations of EDI-OCT was its inability to locate the choroidal-scleral interface in certain cases, especially those with thicker choroids [3]. The advancement of swept-source (SS) OCT enabled deeper penetration and allowed for a more detailed visualization of the choroid and choroidoscleral junction [4‒7]. The development of faster and wider field OCT imaging system technologies allows us to scan even the distant periphery for a better understanding of the choroid, in both healthy and pathological eyes using widefield OCT (WF-OCT).

Over the last years, the choroid and its role in posterior segment disease have become an increasing subject of study due to its involvement in retino-choroidal diseases such as pachychoroid spectrum disorders, age-related macular degeneration or myopic maculopathy and its association with age [3, 7, 8], gender [9, 10], axial length (AL) [11], spherical equivalent [12], and IOP [13].

Nevertheless, since most research had focused on choroidal thickness (CT) around the macular area, the pattern of normal peripheral CT is still unknown. It is important to describe the periphery in detail because the majority of choroidal diseases are not only limited to the macula.

The aim of this study was to describe the CT profile up to the periphery, defined as the area beyond the posterior pole to vortex vein ampulla [14] in healthy eyes using WF-OCT. Additionally, secondary purposes of this study were to compare the reliability of manual versus (vs.) automatic measurement and to study age-related thickness variations.

This was a cross-sectional and noninterventional study conducted in a tertiary referral hospital for vitreoretinal pathology. The protocol of the study was approved by the Local Ethics Committee and all procedures adhered to the tenets of the Declaration of Helsinki. Informed consent was obtained from all participants in the study and from the legally designated representative in minor patients.

Between February 2022 and May 2022, 191 eyes of 101 healthy children and adults were recruited. Patients underwent a complete ophthalmological examination that included a slit-lamp evaluation of the anterior segment, IOP measurement, and fundoscopy. Best-corrected visual acuity (BCVA) and AL (IOL Master; Carl Zeiss Meditec, Jena, Germany) were assessed. WF fundus images were obtained by using WF-OCT (Xephilio WF-OCT S1; Canon Corp, Tokyo, Japan). Inclusion criteria were BCVA 20/20, patients willing to participate, AL between 23 and 26 mm, no systemic or ocular diseases and only good-quality images with clear visualization of the choroid were included for analysis. The exclusion criteria were as follows: (1) eyes with AL of less than 23 mm and greater than 26 mm; (2) eyes with a history of ocular surgery other than cataract and refractive intervention; (3) systemic or ocular diseases. Both eyes of each patient were included in the study whether the inclusion criteria were met, in accordance with previous published data [5]. Each patient or patient’s tutor in case of underage patients was asked to fill out and sign an informed consent form.

Image WF-OCT protocol included 12 radial and cross-sectional b-scans centered on the fovea. The field of view of the B-scans was 23 mm, the scan depth 5.3 mm, and a A-scan repetition rate of 100,000 Hz [15]. B-scan example is shown in Figure 1a. To analyze CT similar to ocular morphology, real-shape correction software provided by Canon Inc. was used [16].

CT was defined as the distance between the Bruch membrane and choroid-scleral junction (Fig. 1b). Using the cross-sectional B-scans, CT was manually measured by two masked and independent retina specialists (Fig. 1b, c) and automatically measured using the built-in software of the OCT device (Fig. 1e, f). CT was assessed in the subfoveal (SF) region and in the horizontal and vertical sectors. Five determinations of the CT were performed both manually and automatically every 2,000 μm from the fovea: temporal (T) (T1–T5), superior (S) (S1–S5), inferior (I) (I1–I5), and nasal (N) (N1–N5). Also, the four oblique sectors were automatically measured: supratemporal (ST), infratemporal (IT), supranasal (SN), and infranasal (IN). The CT of the optic disc area was avoided [17]. Mean CT value was the average value of all determinations automatically measured in all sectors for each eye. To study the evolution of CT with age, the study population was divided into 2 subgroups according to age distribution published in previous papers [5, 18]: under 40 years and over 41 years old.

All analyses were performed using the IBM-PSSS statistical software program (IBM-SPSS, v. 28.0.0.0, Chicago, IL, USA). A two-tailed p < 0.05 was considered statistically significant.

Descriptive statistics were provided for normally distributed variables as means with standard deviation for quantitative variable and n (percentage) for categorical variables. The Kolmogorov-Smirnov test was performed to assess the normality or non-normality of the variables.

Categorical variables were compared using χ² test for normally distributed and Fisher’s exact test for the nonparametric variables, and continuous distributed variables were compared using independent. Student t test for normally distributed variables or the nonparametric Mann-Whitney U test for non-normally distributed variables. Pearson correlation was used to determine the correlations for normally distributed variables. The results were expressed in terms of r and p value. As Pearson’s “r value” is interpreted, values between r = 0.30 to r = −0.30 are not considered clinically significant, being a weak or even null correlation (if r = 0). Demographic data, AL, and CT were studied. As CT is a quantitative variable, the interobserver agreement was calculated by using intraclass correlation (ICC) being moderate (≥0.4), good (≥0.6), excellent (≥0.8), and low (<0.4) [19]. Interobserver agreement for CT manually measures and the concordance between manually and automatically CT measures were analyzed.

The inclusion criteria were met by 191 eyes of 101 individuals. Eleven fellow eyes were excluded because in seven of then the AL >26 mm, two had a choroidal nevus and two a corneal leukoma. Out of the total, 67 (66.3%) were female patients. The mean age was 39.05 ± 19.06 years old (range 6–80), the median age was 39 and the interquartile range of age 27.5 years old. Mean AL was 24.09 ± 1.11 mm (range 23.03–25.82). Patients’ data are shown in Table 1.

There was no statistically significant correlation between CT, AL, ethnicity, or gender. WF-OCT allowed both independent retina specialists clear visualization of the retinal pigment epithelium and choroid-scleral junction and therefore measurement of CT in all individuals (100%). The mean SFCT measured manually was 336.55 ± 75.57 μm (range 150.5–467.5) and automatically 343.67 ± 84.18 μm (range 136–508).

When measured automatically, the mean CT in μm ranged as follows: 328.03 ± 81.35 at T1 to 196.71 ± 48.81 at T5; nasal 246.18 ± 90.20 at N1 to 204.00 ± 79.07 at N5; superior 346.51 ± 78.86 at S1 to 150.07 ± 46.89 at S5; inferior 326.98 ± 89.86 at I1 to 126.61 ± 38.62 at I5; ST 322.86 ± 85.30 at ST1 to 215.37 ± 78.69 at ST5; IT 335.51 ± 95.12 at IT1 to 188.16 ± 60.42 at IT5; SN 307.95 ± 88.51 at SN1 to 222.92 ± 76.15 at SN5; IN 311.74 ± 91.33 IN1 to 174.02 ± 72.34 at IN5. Results are shown in Table 2.

The first superior measure was the thickest among all measures in 62.83% of the eyes (n = 120/191 eyes), in the remaining (n = 71/191) it was the SF CT. All other measures tended to decrease in all other sectors as the points distanced from the fovea. The CT measured automatically by quadrant segmentation and manually is shown in Figure 2a and b, respectively.

Interobserver agreement of all manual measures was excellent (p < 0.01) being the ICC: SF 0.93; nasal 0.90; temporal 0.86; superior 0.86, and inferior 0.84. When comparing automatic and manual CT, the ICC in all measurements was excellent (p < 0.01) being the ICC: SF 0.86 (range 0.81–0.89); nasal 0.94 (range 0.91–0.95); temporal 0.90 (range 0.87–0.92); superior 0.90 (range 0.84–0.93); and inferior 0.92 (range 0.90–0.94). Manual versus automatic CT measurements are shown in Figure 2c and d.

A positive linear correlation was found between SFCT and CT in all the quadrants. In the first determination of CT at 2,000 μm from the fovea, a good positive correlation was found I (r = 0.68, p < 0.01), S (r = 0.68, p < 0.01), N (r = 0.60, p < 0.01), and T (r = 0.64, p < 0.01); and in the four sectors, IT (r = 0.75, p < 0.01), IN (r = 0.67, p < 0.01), SN (r = 0.69, p < 0.01), ST (r = 0.72, p < 0.01). The correlation became weaker as the CT distanced from the fovea.

Linear regression analysis showed no difference between continuous age and CT. Regarding different age groups, the mean CT of subjects <40 years old was statistically thicker (257.02 μm) when compared to subjects >40 years old (234.73 μm) (p < 0.01). AL between groups was not statistically significant (p > 0.05). Results are shown in Figures 3 and 4.

Since its introduction, OCT revolutionized the current practice of ophthalmology and OCT imaging is now a standard method for studying macular choroidal pathology. The choroid and its role in posterior segment disease has become a topic of increasing study due to its implication in retino-choroidal diseases. However, since most choroidal diseases are not limited to the macular area, as seen in retinal vasoproliferative tumors, choroidal nevus and peripheral polypoidal choroidal vasculopathy, the study of the choroid up to the periphery is crucial. Previous studies analyzed the choroid up to 6 mm focused on the SF area since SFCT has been a primary parameter to measure CT in normal and diseased eyes [17]. Nowadays, WF-OCT technology is expected to provide new information about the choroidal mid-periphery and its relevance in the appearance of chorioretinal conditions.

This study provides the most comprehensive description and widest area of CT in healthy eyes to date. Xephilio WF-OCT S1; Canon Corp, Tokyo, Japan provided the present study a reliable information about the choroid of healthy patients. Imaging depths of up to 5.3 mm and 23 mm width allowed examiners to visualize the retina and the choroid with high resolution and no peripheral distortion. In addition, this technology allowed faster image acquisition without being an obstacle for children or the elderly during the examination.

The mean SFCT automatic measured was 343.67 and manually measured was 336.55, being the ICC excellent 0.96, p < 0.01. Also, by quadrants, the ICC showed excellent reliability: nasal quadrant 0.90; temporal 0.86; superior 0.86; and inferior 0.84 p < 0.01. Consequently, manual measuring is independent of the explorer.

In the present study, when measured automatically, the thickest point was in the superior quadrant in 63.83% of the sample (>2,000 microns from the fovea) followed by the SF measure in the remaining subjects. This finding is consistent with a recently published study by Touhami et al. [20] Their study included 64 eyes of 33 healthy patients and studied the distribution of the CT in nine subfields based on the ETDRS grid of 9 mm × 9 mm. They found that the thickest choroidal point was most often located superior and temporal to the macula (72.2%). According to our findings, the thickest choroidal point is located in the superior quadrant beyond the macula.

Funatsu et al. analyzed normal peripheral CT in 120 Asian eyes and reported the thickest choroid sector as the ST in 69.17% of the subjects [17]. As in the present study, indocyanine green angiography was not used. Therefore, there is insufficient evidence to attribute the thickest quadrant to the presence of the dominant vortical vein whose presence in 67% of the cases was located in the superior and temporal quadrant by Mori et al. [21] As a consequence, the distribution of the vorticose vein in our sample may be inferred from the findings of the CT mentioned.

Age-dependent thinning has been widely reported in previous studies as blood flow decreases with age in normal subjects [20]. This study considered the conclusions reached by Ruiz-Medrano et al. [5] when analyzing CT in groups of age [5]. They reported that the CT seems to decrease progressively until 40 years, at which point the most significant variation took place: from 313.9 μm in younger than 40 years old subjects to 264.6 μm in older than 40 years old subjects (49 μm difference). Comparably, this study supports the findings of Ruiz-Medrano et al. [5] showing decrease of mean CT in subjects until 40 years of age and above the age of 40 from 257.02 μm to 234.73 μm in our sample (22.2 μm difference) (p < 0.01). No differences were found in mean CT from 40 years of age (p > 0.05). The difference between their results, regarding the mean CT, could be because they only considered the horizontal meridian scan, unlike the current study in which both horizontal and vertical scans were analyzed, the present study being more accurate and representative.

Similarly, Touhami et al. [20] described that >30% of healthy subjects 55 years of age or less having thicker choroids in comparison with older healthy adults. Consequently, and because the majority of the pachychoroid spectrum diseases occur in this range of ages, CT value may not be the only parameter when describing pachychoroids.

There are two limitations to this study. First, the number of participants is quite limited and mostly included Caucasian. Although diseases within the pachychoroid spectrum are more prevalent in Asia, it is necessary to analyze the choroidal characteristics of Caucasian patients. Second, controversy could arise from using both eyes of the same patient; however, the possibility of using both eyes without altering the results has already been demonstrated in previous work by our group.

Despite these limitations, to the best of our knowledge this is the first study in English language literature that analyzes CT using WF-OCT in healthy Caucasian eyes up to periphery. In conclusion in more than 50% of the healthy subjects, the superior sector is thicker than the subfoveolar sector, confirming previous published data. When comparing automatic and manual CT, the ICC in all measurements was excellent, therefore the automatic system is valid and serves as the basis of choroid measurement. These findings imply streamlining daily clinical practice in the study of CT. Regarding age-related thickness variations, the mean CT became significantly thinner from 40 years of age. Further studies are necessary to replicate the current findings with larger cohorts, wider range of age groups, including pediatric population.

The study was conducted in accordance with the Declaration of Helsinki and approved by the Local Ethics Committee of Puerta de Hierro-Majadahonda University Hospital (Madrid, Spain). Approval code PI 128/18, approved on March 9, 2020. Written informed consent was obtained from all subjects to participate in the study and in case of participants aged under 18 written informed consent was obtained from parents/legal guardians.

The authors have no conflicts of interest to declare.

This study was not supported by any sponsor or funder.

Conceptualization, project administration, and methodology: S.B., B.E., J.M.R.-M., M.P., and J.R.-M.; validation and resources: J.M.R.-M.; formal analysis: M.P.; investigation: S.B., B.E., J.M.R.-M., J.R.-M., and M.P.; data curation: S.B., B.E., and M.P.; writing: S.B. and B.E.; writing – review and editing: J.M.R.-M., J.R.-M., and M.P.; visualization and supervision: J.M.R.-M., J.R.-M., and M.P. All authors have read and agreed to the published version of the manuscript.

Blanca Eslava and Sofia Bryan contributed equally to this work.

All the data generated or analyzed during this study are included in this article. Further inquiries can be directed to the corresponding author.