Pupillary Light Reflex and Multimodal Imaging in Patients With Central Serous Chorioretinopathy

Abstract

Purpose: The purpose of this study was to investigate and compare the corresponding alterations of the pupillary response between acute and chronic central serous chorioretinopathy (CSC) and between different disease categories.

Methods: We recruited patients with unilateral acute and chronic CSC. An eye tracker was applied to determine the pupillary light reflex (PLR) and evaluate the following PLR metrics in healthy eyes: pupil diameter, diameter changes, including relative constriction amplitude (AMP%), and re-dilation ratio (D1%). Baseline optical coherence tomography (OCT), and fluorescein and indocyanine green angiography (FA/ICGA) were performed to analyze the relationship between pupillary response and retinal/choroidal architecture.

Results: In total, 52 patients were enrolled, including 25 with acute CSC and 27 with chronic CSC. Compared to the chronic CSC group, the acute CSC group displayed a significantly larger baseline pupil diameter (BPD; of 5.51 mm, P = 0.015), lower AMP% (34.40%, P = 0.004), and higher D1% (93.01%, P = 0.002), indicating sympathetic overactivity. On OCT, the total macular volume was positively correlated with the D1% (r = 0.48, P = 0.005) and negatively with AMP (r = −0.47, P = 0.007). On ICGA, the intense choroidal vascular hyperpermeability (CVH) group displayed a greater BPD than the nonintense CVH group. Additionally, 9 cases with later recurrent episodes following therapy showed a lower AMP% and higher D1% than the nonrecurrent group.

Conclusions: The PLR revealed sympathetic excitation in patients with acute CSC. The stronger D1% was significantly associated with greater total macular volume, and it may be a potential biomarker for predicting the later recurrence of CSC.

The pupillary light reflex (PLR) is the process by which the pupil diameter constantly adjusts to light level fluctuations to control how much light enters our eyes.¹ It has been recognized over recent years that a quantitative neurophysiologic tool of the PLR may provide alternative biomarkers for investigating and understanding mechanisms involved in ocular or brain dysfunction.^2–6 Parasympathetic-innervated pupillary constriction is controlled by the Edinger-Westphal nucleus, whereas the sympathetic-innervated pupillary dilator muscle is stimulated via the hypothalamic nuclei.^7,8 Therefore, dynamic pupillometry, which can provide different pupil metrics to effectively quantify the pupillary responses, is a promising and convenient assessment method for autonomic nervous system (ANS) activity.^3,9 

Increasing evidence indicates that sympathetic nervous system (SNS) hyperexcitability in central serous chorioretinopathy (CSC)^10–12 is associated with some psychologic risk factors, including chronic stress, type-A behavior pattern, and aggressive personality.^13,14 Choroidal abnormalities, including thickened choroid, diffuse hyperpermeable capillaries, and choroidal vessel dilation, have been observed in CSC through multimodal imaging techniques, such as optical coherence tomography (OCT), fluorescein angiography (FA), and indocyanine green angiography (ICGA).¹⁵ Choroidal blood flow is thought to be innervated mainly by the ANS and the sympathetic arousal may lead to the disturbance of choroidal vascular circulation and homeostasis, and participate in the pathogenesis of CSC.¹³ With recent advances in pupillometry, PLR metrics currently include pupil diameter, constriction amplitude (AMP), and re-dilation ratio (D1%), which can provide quantitative data, reflecting the activity of the sympathetic or parasympathetic tone. Our previous study discovered that patients with CSC had sympathetic activation, parasympathetic inhibition, a higher total mood disturbance, and more negative mood scores than healthy volunteers, as examined by a combination of pupillary responses, heart rate variability (HRV), and the Profile of Mood States questionnaire.⁵ 

Acute CSC is typically a self-limiting disease with a good recovery; however, recurrent or chronic CSC can lead to hypoxic damage to the neurosensory retina and photoreceptor cell death, and thereby result in irreversible vision loss.^16,17 Therefore, a convenient and efficient method for predicting recurrent or chronic CSC is highly recommended. Consequently, PLR metrics were applied to investigate the interdependencies with the multimodal imaging parameters and to determine whether these metrics as physiologic biomarkers can differentiate between the disease categories and predict later CSC recurrence. 

Methods

Subjects

Our previous study compared healthy subjects and patients with CSC regarding pupillary responses and HRV, but not anatomic or imaging characteristics.⁵ Combining PLR data from this previous study with more recent cases and cases in which HRV measurements failed, the current cases were collected from April 2020 to September 2022. Participants were recruited from the Department of Ophthalmology at the Hyogo Medical University (Japan) after providing written informed consent. The procedure in this study was approved by the Hyogo Medical University ethics committee in accordance with the Tenets of the Declaration of Helsinki. 

The inclusion criteria were patients with unilateral CSC that was confirmed by OCT showing serous subretinal fluid and by FA detecting one or more leakage points. Acute CSC was defined as experiencing an acute episode of visual symptoms or subretinal fluid within 6 months. Chronic CSC was defined as having a symptomatic episode or subretinal fluid that lasted more than 6 months. Exclusion criteria included the following: (1) age-related macular degeneration, diabetic retinopathy, uveitis, optic neuritis, or other retinal disorders; (2) systemic diseases of diabetes mellitus, head trauma, or peripheral neuropathy; and (3) currently taking anticholinergics, sympathomimetics, systemic steroids, or other medications that may affect the pupil measurements. 

We obtained the following demographic and medical information from each patient: age, sex, medical history, medication usage, blood pressure, best corrected visual acuity (BCVA), spherical equivalent, treatment regime, and the recurrence status during the follow-up period. A recurrence was defined as a serous retinal detachment that occurred following the previous episode of subretinal fluid resolution. 

Pupillary Light Reflex

In this study, the pupil diameter was captured by the screen-based eye tracker Tobii Pro Spectrum (Tobii Pro AB, Danderyd, Sweden) at a sampling rate of 300 hertz (Hz). A detailed description of this device and the experimental process were given in our previous study.⁵ The experiment was conducted between 12:00 and 15:00. Patients were provided with detailed instructions before the commencement of the PLR task. In a dimly lit room, patients sat comfortably on a chair placed 60 cm away from a black screen with a red cross in the center on which to fix their gaze. Two minutes of dark adaptation were followed by white light stimulation (red, green, blue [RGB] = 255, 255, and 255) with an intensity of 13 cd/m² lasting 1 second and covering the entire screen for the PLR. The intensity of the white light was measured at a distance of 60 cm from the screen which is a 23.8-inch monitor (EIZO FlexScan EV2451; EIZO Corporation, Japan) with a resolution of 1920 × 1080 pixels (16:9) and a refresh rate of 60 Hz. The PLR test was performed on one eye (the healthy eye) after the calibration and the diseased eye was covered with an eye mask. 

The synchronization of stimuli and pupil recordings by the eye tracker facilitated the following PLR metrics⁵ evaluation (Fig. 1A): (1) pupil diameters: 1-second baseline pupil diameter (BPD) before the flash onset; (2) constriction phase (parasympathetic nervous system innervation): minimum pupil diameter (MIN), AMP (BPD − MIN), the ratio of AMP to BPD (AMP%) as the relative amplitude, maximum constriction velocity and acceleration (MCV and MCA, respectively); time from flash onset to the minimum pupil diameter (T1) and maximum velocity (T2); and (3) re-dilation phase (SNS innervation): pupil diameter (D1) at 5.5 seconds (3.5 seconds after light offset) and the ratio of D1 to the BPD as the re-dilation ratio (D1%). For the PLR trial analysis, blinks or signal loss in excess of 50% were eliminated, and the valid PLR trial was linearly interpolated to fix the noise and blinking artifacts, and a 10-point moving average filter was utilized subsequently to smooth the data. 

Figure 1.

 

(A) Pupillary light reflex responses to white light stimulation in the acute and chronic central serous chorioretinopathy (CSC) groups. Baseline pupil diameter and minimum pupil diameter were larger in the acute CSC group than in the chronic CSC group (P = 0.015 and P = 0.003, respectively). (B) Normalized pupil diameter changed over time between the two groups, there was a significantly lower AMP% and higher D1% in the acute CSC group (P = 0.004 and P = 0.002, respectively). The lighter shading above and below represents the standard error of the mean. Abbreviations: AMP%, relative constriction amplitude; D1%, ratio of the re-dilation pupil diameter at 5.5 seconds to the baseline pupil diameter.

Multimodal Imaging Classifications

We further elaborated the retinal imaging features of each patient by the combination of FA and ICGA (HRA2; Heidelberg Engineering, Heidelberg, Germany) and Heidelberg Spectralis OCT (Heidelberg Engineering). Two independent readers (authors X.Z. and T.S.) conducted the analyses and evaluations of pertinent imaging data, reconciled differences through discussion, and finally established the following parameters and classifications: (1) the number of leakage points: single leakage spot or multiple leakage spots (≥ 2 leakage spots); (2) the intensity of fluorescein leakage on FA: faint or mottled fluorescence as nonintense or increasing fluorescence as intense grade; (3) choroidal vascular hyperpermeability (CVH): substantially brighter hyperfluorescent areas on ICGA approximately 10 minutes after injection was considered to be intense CVH, whereas areas that were slightly brighter or not brighter than the background fluorescence was termed nonintense CVH; (4) total macular volume (TMV) obtained automatically by Spectralis OCT; (5) subfoveal choroidal thickness (SCT) perpendicularly from the Bruch membrane to the chorioscleral boundary; and (6) the thickness of the choriocapillaris layer was calculated by subtracting the thickness of the Haller's layer (from the inner margin of the large choroidal vessel layer to the outer margin of the sclera layer) from the SCT.¹⁸ 

Study Outcomes and Statistical Analysis

The baseline multimodal imaging characteristics and PLR performance were compared between the acute and chronic CSC groups as primary outcomes. The secondary outcome was the difference in the PLR between the later recurrent CSC group and the nonrecurrent CSC group. We also compared the PLR between different study categories (i.e. single leakage spot group vs multiple leakage spots group, and intense CVH group versus nonintense CVH group). Subsequently, correlation analyses between PLR metrics and imaging characteristics (i.e. TMV value and SCT value) in patients with CSC were performed. 

For continuous variables, the t-test was used to compare normally distributed data, whereas the Mann–Whitney U test was used to compare skewed data distributions. Correlations between the PLR metrics and anatomic/clinical variables were calculated by Pearson's or Spearman's correlation coefficients. A P value < 0.05 was considered to be statistically significant. The BCVA was measured with a Landolt C acuity chart. Decimal BCVA values were converted to logarithm of the minimum angle of resolution units (logMAR). We used JMP Pro (version 15; SAS Institute Inc., Cary, NC, USA) and GraphPad Prism (version 7.0; GraphPad Software, San Diego, CA, USA) for analyses. 

Results

Baseline Characteristics

Among the patients with valid PLR measurements, a total of 25 patients with acute CSC (22 men and 3 women; mean age = 50.7 ± 10.8 years) and 27 patients with chronic CSC (24 men and 3 women; mean age = 48.5 ± 6.4 years) were enrolled. The two groups did not differ with respect to age, sex, blood pressure, or other basic clinical characteristics (Table 1). Without recognized visual abnormalities in the healthy eye, the mean BCVA of this eye in acute CSC and chronic CSC was 20/17, 20/16, respectively, in the Snellen equivalent. Additionally, there was no significant difference in the baseline mean BCVA between the two groups regarding the CSC eye (P = 0.832). For the multimodal imaging findings, there were no significant differences in the OCT findings, FA pattern, or ICGA intensities between these two groups. 

Table 1.

 

Baseline Demographics, Clinical Information, and Multimodal Imaging Characteristics Between the Acute and Chronic CSC Groups

 

PLR in the Acute and Chronic CSC Groups

Differences in the PLR measurements for the acute and chronic CSC groups were analyzed and summarized in Table 2, and the corresponding PLR trace changes over time for each group were presented in Figure 1. We discovered that the BPD and MIN were significantly larger in the acute CSC group compared with the chronic group (P = 0.015 and P = 0.003, respectively; see Fig. 1A). In addition, the acute CSC group had a significantly diminished constriction response (AMP% = 34.40% vs. 39.12%, P = 0.004; Fig. 1B) and a significantly greater dilation response (D1% = 93.01% vs. 89.75%, P = 0.002; see Fig. 1B) than the chronic CSC group. 

Table 2.

 

Pupillary Light Reflex of the Acute and Chronic CSC Groups

 

Association Between PLR and Imaging/Clinical Variables

We investigated the mean PLR differences among 48 cases by ICGA examination, which were divided into the intense CVH (n = 19) and nonintense CVH groups (n = 29). Normalized pupil diameter changed over time between these two groups are shown in Figure 2A. As displayed in Table 3 and Figure 2B, a larger BPD (5.61 mm vs. 5.01 mm, P = 0.002) and MIN (3.69 mm vs. 3.11 mm, P = 0.004) were found to be indicative of a significantly intense CVH on ICGA. However, no significant differences were found in other study categories, such as the leakage spot group (see Table 3). 

Figure 2.

 

(A, B) Differences in the PLR between the intense CVH (n = 19) and nonintense CVH groups (n = 29). The intense CVH group showed a significantly larger BPD (5.61 mm vs. 5.01 mm, P = 0.002) and MIN (3.69 mm vs. 3.11 mm, P = 0.004). (C, D) Normalized pupil diameter changed over time between the later recurrent and nonrecurrent CSC groups. There was a significantly lower AMP% in the later recurrent CSC group (P = 0.034). Besides, the later recurrent CSC group demonstrated a significantly higher re-dilation response (D1% = 93.77, n = 9) than the nonrecurrent group (D1% = 90.66, n = 40; P = 0.032). The lighter shading or dotted lines above and below represent the standard error of the mean. Abbreviations: CSC, central serous chorioretinopathy; CVH, choroidal vascular hyperpermeability; PLR, pupillary light reflex; BPD, baseline pupil diameter; MIN, minimum pupil diameter; AMP%, relative constriction amplitude; D1%, ratio of the re-dilation pupil diameter at 5.5 seconds to the baseline pupil diameter.

Table 3.

 

Pupillary Light Reflex of the Multimodal Imaging Groups

 

We conducted correlation analyses between PLR metrics and anatomic characteristics in patients with CSC who completed PLR and OCT examinations on the same day (n = 33; Fig. 3). Among them, TMV data were unavailable in one patient. The TMV correlated negatively with AMP (Spearman r = −0.47, P = 0.007, n = 32; Fig. 3A), but positively with the D1% (Spearman r = 0.48, P = 0.005, n = 32; Fig. 3B). Nevertheless, there was no significant correlation between the subfoveal choroidal thickness and the AMP or D1% (n = 33; Figs. 3C, 3D). A substantial positive correlation between the choriocapillaris layer thickness and AMP% (Spearman r = 0.52, P = 0.002, n = 33) was found in our study. 

Figure 3.

 

Correlation analyses between PLR metrics and imaging characteristics in patients with CSC who underwent PLR and OCT examination on the same day. (A, B) Total macular volume correlated negatively with AMP (Spearman r = −0.47, P = 0.007, n = 32), but positively with the D1% (Spearman r = 0.48, P = 0.005, n = 32). (C, D) There was no significant correlation between subfoveal choroidal thickness and AMP or the D1% (n = 33). Dotted lines represent 95% confidence bands. Abbreviations: PLR, pupillary light reflex; CSC, central serous chorioretinopathy; OCT, optical coherence tomography; AMP, constriction amplitude; D1%, ratio of the re-dilation pupil diameter at 5.5 seconds to the baseline pupil diameter; N.S., not significant.

PLR in the Recurrent Group

Among the 52 CSC cases, photodynamic therapy was administrated to 42 eyes (80.8%) and focal laser therapy to 5 eyes (9.6%), whereas 5 eyes (9.6%) underwent observation. Furthermore, 9 cases experienced recurrent episodes during the follow-up period with a mean recurrence interval of 3.8 ± 2.0 months (range = 1–8 months). As shown in Table 4 and Figures 2C, 2D, there was a significantly larger MIN and lower AMP% in the later recurrent CSC group (P = 0.025 and P = 0.034, respectively). Besides, the later recurrent CSC group demonstrated a substantially higher re-dilation response (D1% = 93.77) than the nonrecurrent group (D1% = 90.66, P = 0.032, n = 40; see Fig. 2D), whereas three patients were lost to follow-up and were not included in this statistical analysis. In addition, the recurrence interval among the recurrence group was negatively correlated with D1% (Pearson r = −0.74, P = 0.022, n = 9), with a greater D1% suggesting a shorter interval. 

Table 4.

 

Pupillary Light Reflex of the Later Recurrent and NonRecurrent CSC Groups

 

Discussion

Our primary objective was to evaluate the performances of the pupillary response in different stages of CSC. The reduced AMP% and larger BPD and D1% indicated that enhanced sympathetic activity was associated with acute CSC compared with chronic CSC. Likewise, the patients with later recurrent CSC displayed higher SNS activity as evidenced by the greater D1% than in those without recurrence. In particular, a higher D1% might signal a shorter recurrence interval. Moreover, the interdependencies between the PLR metrics and fundus imaging characteristics demonstrated that intense CVH and a higher TMV may indicate sympathetic excitation. 

As a network of causal relations, stress, elevated plasma cortisol levels, and enhanced sympathetic activity are intimately inter-related and involved in CSC pathogenesis.^13,19 Psychologic stress cannot only result in hyperactivity of the adrenomedullary-sympathetic system, which triggers noradrenaline release, causing pupil dilation, but also stimulates cortisol secretion through the hypothalamic-pituitary-adrenal axis.^13,20 Assessing stress or sympathetic activity is challenging and subject to error when measured in humans.²¹ In our prior study, we found that patients with CSC exhibited sympathetic excitation, parasympathetic inhibition, and total mood disturbance compared with healthy subjects.⁵ Therefore, in the current study, we used pupil dynamics, a sensitive and quantitative physiologic marker, to gauge disease activity and the corresponding alterations in the ANS. 

Pupillometry as a noninvasive technique is widely utilized to measure changes in pupil diameter and evaluate the dynamic balance between sympathetic and parasympathetic activity.^1,3 PLR was carried out with a healthy eye to prevent the low vision in the afflicted eye from affecting the assessment of the ANS. In response to light, the amplitude of pupil contraction depends on the activity of the iris sphincter muscle, which is regulated by the parasympathetic system. By contrast, the iris dilator muscle is primarily controlled by the α1-adrenergic sympathetic pathway, which results in increased BPD and D1% values when the sympathetic nerve is activated.^5,22 Furthermore, Hopstaken et al. emphasized that the BPD is frequently correlated with physiologic arousal levels, reflecting the innervation of sympathetic fibers.²³ Therefore, compared with chronic CSC, acute CSC demonstrated a decreased AMP% and larger BPD and D1%, which, respectively, indicated parasympathetic nervous system attenuation and sympathetic enhancement. 

A previous study found that patients with active CSC exhibited significantly higher salivary alpha-amylase activity levels than those with inactive CSC, suggesting pathologic sympathetic activation in the acute phase.²¹ Likewise, Bernasconi et al. found that patients with acute CSC have a higher ratio of low-frequency to high-frequency components of the HRV analysis, an indicator of sympathetic excitation.²⁴ According to their findings, sympathetic activity was higher in acute and acute recurrent CSC than in chronic CSC. Our present intriguing findings that pupillary responses are different in acute and chronic CSC may provide some new evidence for aberrant sympathetic activity and stress responses contributing to disease onset or progression. 

Detailed investigation of the relationship between the multimodal fundus imaging and pupillary response has not been conducted in patients with CSC to date. The present study showed that intense CVH, as identified by ICGA, was associated with sympathetic excitation, as indicated by a larger BPD and MIN. Moreover, the TMV correlated positively with the D1% but negatively with the AMP, which may be consistent with a case study that discovered that fluid leakage increased appreciably in a patient suffering from anxiety, a similar mechanism to stress-induced CSC.²⁵ Another report found that serum mitochondrial DNA (mtDNA) levels were substantially higher in acute CSC compared with the healthy group and serum mtDNA levels were positively correlated with the height of subretinal fluid.²⁶ They proposed that physical and/or psychologic stress might result in an imbalance between the SNS and parasympathetic nervous system and elevate the serum mtDNA levels, subsequently triggering Toll-like receptors to increase choroidal vasculature permeability.²⁶ This may be in agreement with our results that the positive correlation between the D1% and TMV and intense CVH indicated higher SNS activity. In a rat model, we observed choroidal thickness was substantially increased after restraint stress but decreased within one day.²⁷ However, although we did not find any significant relationships between SCT and pupil parameters in the current study, we did find a positive correlation between the choriocapillaris layer thickness and AMP%. One of the morphologic characteristics in CSC was the focal attenuated inner choriocapillaris layer that was located above the dilated choroidal vessels.²⁸ Therefore, the choriocapillaris layer thickness may provide an intuitive representation of the state of the autonomic nerves. In view of these findings, a combination of fundus imaging and pupillometry can be used to gain insight into a patient's sympathetic state. 

Furthermore, Saito et al. revealed that the macular choroidal blood flow velocity determined by laser speckle flowgraphy was high in the acute stage of CSC and gradually decreased with disease regression.²⁹ Enhanced SNS activity may contribute to choroidal artery vasoconstriction mediated by α-adrenergic receptor, and secondary passive overflow into the peripheral large choroidal veins, resulting in choroidal blood flow elevation and perfusion abnormalities in eyes with acute CSC.^29–31 Additionally, sympathetic β-adrenoceptor activation may contribute to cardiac output elevation and thus consequently to the choroidal blood overflow, resulting in subretinal fluid accumulation.³⁰ In the current study, the quantitative PLR parameters by providing insight into the fundus imaging manifestations and ANS abnormalities in CSC may contribute to further comprehension of the mechanism of its development. 

In addition, pupillary responses demonstrated higher sympathetic activity in the later recurrent group than in the nonrecurrent group. Our prior study found lower blood 5-hydroxytryptamine (5-HT) concentrations in recurrent CSC compared with chronic CSC.¹¹ The sympathetic premotor neurons associated with cardiovascular function can be directly inhibited by the 5-HT within the rostral ventrolateral medulla,³² and reduced 5-HT can enhance vulnerability to psychosocial stress.^33,34 Mukherji et al. revealed that patients with recurrent CSC have increased circulating cortisol compared with non-chorioretinal patients.³⁵ In another study, neither increased cortisol levels nor serum testosterone levels were associated with chronic CSC.³⁶ Additionally, the present study revealed a negative relationship between the recurrence interval and D1%, with a higher D1% indicating a shorter interval. Consequently, we hypothesized that patients with CSC are more likely to relapse if their SNS is overexcited. Prompt diagnosis and management of recurrent CSC are of great significance for the protection of vision. However, due to limited recurrent cases and that recurrent CSC may tend to be associated with chronic CSC, we were unable to draw firm conclusions and need to conduct further research with a longer follow-up period. 

The main limitation of the current study was the time lag between pupil examination and fundus imaging results. However, the interval between OCT and pupil examination was more than 10 days in only 7 patients, and between angiography and pupil experiment was more than 1 month in only 6 patients. As a result of the short follow-up period and limited recurrent cases, relevant results should be interpreted carefully. It would be meaningful to further verify with more cases whether a higher D1% can predict the later recurrence and determine its relationship to the recurrence interval. Nevertheless, pupillary response data may assist medical workers in customizing stress-reduction strategies and identifying the necessity for long-term follow-up and prompt treatment for patients with CSC or stress-related disorders in the future. 

In conclusion, the differences between acute and chronic CSC in terms of their abnormal pupillary responses, as well as the differences of PLR in imaging features of the retina and choroidal vasculature, revealed important insights into the relationship between sympathetic activation and anatomic abnormalities in CSC. Quantitative PLR metrics may serve as efficient and simple biomarkers for disease activity. 

Acknowledgments

Supported by a Hyogo Medical University Grant: “Hyogo Innovative Challenge”; Hyogo Medical University Diversity Grant for Research Promotion under MEXT Funds for the Development of Human Resources in Science and Technology, Initiative for Realizing Diversity in the Research Environment (Characteristic Compatible Type); the Mishima Saiichi Memorial Ophthalmology Research International Grant; and Grants-in-Aid for Scientific Research (22K09803). 

Previous Meeting Presentation: The findings in this study were presented at the biennial meeting of the International Society of Eye Research (ISER), Gold Coast, Queensland, Australia, February 19–23, 2023. 

Disclosure: X. Zhou, None; H. Fukuyama, None; T. Sugisawa, None; Y. Okita, None; H. Kanda, None; Y. Yamamoto, None; T. Araki, None; F. Gomi, None