Association of the choroidal vascularity index with myopic traction maculopathy: A preliminary case-series report

INTRODUCTION

The choroid is a vascular structure that supplies most oxygen and nutrients to the retina and retinal pigment epithelium (RPE). Due to its role, choroidal thickness has been used as a robust tool in clinical research, as it affects several ophthalmic diseases.^[1] However, choroidal thickness does not differentiate the vascular flow between the two stromal and luminal vascular components, and the measurements are also not reproducible. By determining the area of dark and light pixels corresponding to lumina (luminal area; LA) and stromal areas (stromal choroidal area; SCA) using noninvasive images, major components of the choroid can be represented by the vascular lumen of the vessels.^[2] This parameter has been described in recent years and is calculated from optical coherence tomography (OCT) images through image binarization and quantification; it is called the choroidal vascularity index (CVI). It is a novel tool to assess choroidal structure using non-invasive imaging procedures and techniques.^[3-5] The CVI is calculated as the ratio of LA to total choroidal area (TCA), presented as a percentage.^[6] Since the CVI reflects vascular changes in the choroid, it provides a unique opportunity to obtain perfusion information about the choroid.

Multiple studies have linked the association of the CVI with disease pathogenesis and progression. These include genetic, retinal, and macular diseases.^[5,7,8] These studies suggest that the CVI can potentially be developed as a unique marker for retinal diseases. Since retinal diseases involve multiple factors, such as inflammation, edema, and leakage, the CVI has been suggested to be a potential marker because it represents the potential alterations that affect the integrity of the vascular network within the heavily blood-supplied retinal structure, that is, the choroid.^[7,9] Therefore, it is believed that quantification of vascular factors can be a true representation of the ongoing disease pathogenesis.

Myopic traction maculopathy (MTM), a sight-threatening condition in individuals with high myopia, has received considerable attention for timely diagnosis and management, as it leads to multiple retinal pathologies if untreated on time.^[10] Highly myopic eyes with macular foveoschisis and myopic foveoretinal detachment (FRD) naturally progress to the formation of macular holes (MHs).^[11,12] Highly myopic eyes may develop macular foveoschisis in almost 34% of patients.^[13-15] It is believed that macular foveoschisis and FRD are the precursors of MTM. Identifying these conditions early in pathogenesis would potentially help prevent vision loss. Morphological changes in high myopia are usually assessed using Doppler ultrasound and fundus fluorescein angiography (FFA). However, these methods have limitations, as Doppler ultrasound provides low-resolution images, and FFA is an invasive procedure. Monitoring MTM and treatment outcomes longitudinally through non-invasive reproducible methods would be a significant achievement for managing this disease.^[16] Previously, vascular density in the macular region has been studied as a clinical measure in high myopia patients; however, the findings are variable.^[16] CVI may be a sensitive tool because it accounts for microcirculatory changes in the retina,^[17-19] and there have been reports of a functional correlation of visual acuity with vessel density in the macula and choroid.^[20] Therefore, it is hypothesized that the CVI will predict the development of MTM in high myopia cases. For this study, the most recent MTM classification was adopted.^[15,21] According to this classification, stage 1 disease is myopic foveoschisis (MF), stage 2 disease is foveoretinal detachment (FRD), stage 3 disease is myopic MH, and stage 4 disease is MH retinal detachment (MHRD).

Previously, CVI has been observed to be negatively affected by age.^[22] Therefore, there have been several attempts to determine age-related normative data for CVI. It has also been reported that the CVI is not affected by confounding factors such as blood pressure, axial length, and intraocular pressure.^[23] Therefore, it has been argued that the CVI may be a sensitive marker of disease progression. CVI distribution has also been reported to be highly consistent in the subfoveal, central submacular, and total submacular choroid. Recently, Tang et al.^[24] assessed the choroidal thickness and vascularity index in high myopia cases after macular buckling surgeries. They found that the CVI increased at 1 and 3 months postoperatively and that the choroid thickened in the early post-operative period. Although many studies on high myopia have evaluated the CVI in their patient cohorts, there has been a substantial lack of comparative data between disease and age-matched control groups. Therefore, we aimed to evaluate the association of the CVI with different stages of MTM for its potential use as a marker of disease progression and pathogenesis. To the best of our knowledge, the present study is the first to compare CVI values across different stages of MTM with the CVI of normal control subjects. Therefore, these preliminary findings are expected to provide significant clinical insight into managing MTM.

MATERIAL AND METHODS

Study design

In this retrospective, interventional, one-surgeon, case‒ series report, one normal emmetropic eye, one healthy myopic eye [Figure 1], and four operated and structurally fully resolved myopic eyes at different stages of MTM (surgery group) were evaluated using OCT. The surgery group included consecutively enrolled patients from April 2017 to June 2022 who underwent a vitrectomy procedure using different internal limiting membrane (ILM) removal surgical techniques for symptomatic MTM. Only selected eyes at each stage of the condition where surgery was successful were included in the study. The disease was staged as shown in [Table 1]. The long-term post-operative enhanced high-definition (HD) 9-mm horizontal OCT images were binarized, and the TCA and LA were quantified using ImageJ software. The CVI (%) was calculated as the ratio of LA to TCA.

Close

Figure 1:

Normal control groups. (a) Enhanced high-definition (HD) 9-mm horizontal B scan designed to depict more detail of the intraretinal structure and choroidal layers in a normal eye. (a1) Corresponding horizontal B scan with binarized processing of the subfoveal choroidal stroma and luminal vascular visualization of the subfoveal choroidal vessels to obtain the choroidal vascularity index (CVI) of a normal emmetropic eye. The selected subfoveal area is clearly delineated with the yellow dotted line. (b) HD 9-mm horizontal B scan of a healthy highly myopic eye with enhanced choroidal vessel visualization to quantify the CVI. (b1) Corresponding binarized image with a red dotted line clearly delineating the selected subfoveal choroidal vascular area.

Export to PPT

Table 1: Summary of study groups and inclusion criteria.

Study group	Inclusion criteria
Control emmetropic (n=1) Control high myopia (n=1)	Spherical equivalent refractive error of +0.25 diopters and axial length of 20.72 mm Spherical equivalent refractive error of −13.0 diopters and axial length of 29.5 mm
Surgical treatment group (n=4)	One eye exhibiting myopic foveoschisis One eye with evidence of foveoretinal detachment One eye with evidence of a macular hole One eye with macular hole retinal detachment

This study was conducted in the Retina Department at Oftalmologia Integral ABC, Mexico City, Mexico. The institutional review board approved the study, and written informed consent was obtained from all patients per institutional guidelines; no reference number is provided for retrospective studies by this institution. In this report, the stage of the disease, the study details and the inclusion criteria are listed in [Table 1].

The participants were matched for age, sex, study period, and follow-up duration. Patients were evaluated and followed longitudinally. None of the selected eyes included in the study received intravitreal injections or laser photocoagulation during the study period. In addition to the selection criteria described in [Table 1], none of the eyes had evidence of patchy foveal-affected chorioretinal atrophy, diffuse macular chorioretinal atrophy, or quiescent or active myopic choroidal neovascularization (CNV) according to the ATN classification.^[25] The diagnosis of different stages of MTM was confirmed using magnified spectral-domain OCT (SD-OCT) findings. All four eyes were followed up for more than 6 months with different time frames of the post-operative follow-up period. The four examined eyes included in this report underwent a post-operative long-term perfusion evaluation with OCTA and binarization of the enhanced HD 9-mm horizontal B scan images to calculate the CVI.

RESULTS

The patient demographic data and preoperative clinical characteristics are summarized in [Table 2]. The pre-operative and post-operative functional changes and the type of surgical macular approach according to the MTM stage are described in [Table 3]. The postoperative quantifications of the CVI, LA, and TCA are summarized in [Table 4].

The quantification of the TCA, LA, and CVI of the control eyes compared with the post-operative TCA, LA, and CVI in the different post-operative MTM stages is shown in [Figure 2a-c], respectively.

Close

Figure 2:

(a) TCA. (b) LA and (c) CVI among the six cases. TCA: Total choroidal area, LA: Luminal area, CVI: Choroidal vascularity index.

Export to PPT

The correlated statistical analysis of the post-operative TCA and the best-corrected visual acuity (BCVA) are plotted on the X- and Y-axes, respectively. Each dot represents data from one case. The blue line depicts the linear regression of the data [Figure 3a]. Likewise, the correlation analyses of the postoperative LA and the BCVA are plotted on the X- and Y-axes, respectively. Each dot represents data from one case. The blue line depicts the linear regression of the data [Figure 3b]. Finally, the correlated analysis of the post-operative CVI and the BCVA are plotted on the X- and Y-axes, respectively. Each dot represents data from one case. The blue line depicts the linear regression of the data [Figure 3c].

Close

Figure 3:

(a) Correlation between TCA and BCVA. (b) Correlation between LA and BCVA. (c) Correlation between CVI and BCVA. TCA: Total choroidal area, LA: Luminal area, CVI: Choroidal vascularity index, BCVA: Best-corrected visual acuity.

Export to PPT

Surgical case 1 (MFs)

A 56-year-old symptomatic male complained of metamorphopsia and visual loss in his right eye over 7 months. The pre-operative visual acuity was 0.78 in logMAR units (20/120 Snellen equivalent), and the applanation ocular tension was 11 mmHg. This eye had an axial length of 27.32 mm with PS, and the preoperative OCT findings were consistent with macular thickening due to schisis of the internal and external retina layers [Figure 4a]. After a 31-month follow-up, the center of the macular region appeared thinner with a normal foveal profile, no evidence of a MH and final vision of 0.48 in logMAR units (20/60 Snellen equivalent). Furthermore, some recognizable SD-OCT biomarkers, such as a normal foveal contour and internal and external neuroretina lines, and total restoration of the central subfoveal ellipsoid zone (EZ), including at the IS/OS line and the external limiting membrane, were observed [Figure 4a1]. The long-term post-operative perfusion evaluation was abnormal, with a CVI index of 53.3%, lower than the one in the control healthy myopic eye [Figure 4a2].

Close

Figure 4:

Surgical cases. (a) Pre-operative OCT findings were consistent with myopic foveoschisis due to macular thickening and schisis-like thickening of the internal and external retina layers. (a1) After a 31-month follow-up, the center of the macula region looked thinner with a normal foveal profile and without evidence of a macular hole. SD-OCT biomarkers, such as a normal foveal contour and internal and external neuroretina lines with total restoration of the central subfoveal EZ, were observed. (a2) Postoperative long-term binarized image with a CVI index of 53.3%. (b) Myopic FRD with subretinal fluid and a very thin foveal superficial layer. (b1) Long-term postoperative horizontal B scan with diffuse thinning of the retinal layers and evidence of FRD flattening. (b2) The CVI calculated from the binarized image was 49.6%. (c) Highly myopic eye with moderate posterior staphyloma. Evidence of a full-thickness myopic macular hole with tractional elongation of Henle’s layer and thickening of the macula without evidence of macular detachment. (c1) The postoperative structural evaluation showed a flat macula with a recovered foveal profile, an ELM line lucency defect and a well-preserved RPE layer. (c2) The CVI was 47.4%. (d) Preoperative SS-OCT of an extensive retinal detachment-related myopic macular hole. (d1) This image depicts the long-term postoperative appearance with extrafoveal chorioretinal atrophic areas, an irregular foveal profile, thin foveal roof, nicely closed macular holes, macula reattachment without evidence of residual subretinal fluid, attenuated internal and external retina layers, subfoveal IS/OS (EZ), and an ELM line. (d2) The quantified choroidal perfusion indices were lower than the one obtained in the healthy myopic eye, with a very low CVI of 40.9%. The selected subfoveal area is clearly delineated by the red and yellow dotted lines in the binarized panels. OCT: Optical coherence tomography, CVI: Choroidal vascularity index, EZ: Ellipsoid zone, FRD: Foveoretinal detachment, ELM: External limiting membrane, SD-OCT: Spectral-domain optical coherence tomography.

Export to PPT

DISCUSSION

The present study investigated the CVI as a potential biomarker for MTM. Data were compared between two eyes in the control group (one healthy emmetropic eye and one healthy highly myopic eye) and an intervention group (patients with MTM related to high myopia). The results showed that the CVI was significantly lower in the disease group than in the control healthy eyes. These findings imply that the CVI may indicate disease pathology in MTM. A lower CVI in diseased eyes suggests that choroidal vascular defects caused by pathologic myopia or surgery procedures can influence the manifestation, progression and functional outcomes in the different stages of surgically resolved MTM. This study was not designed to quantify the preoperative CVI in the different stages of this condition; consequently, it was not possible to comparatively evaluate the pre-operative CVI with the postoperative values to infer vascular damage hypothetically caused by transoperative alterations of the choroidal perfusion.

Pre-operative patient characteristics were similar and were directly comparable except for visual acuity, which was expected to be poor in the disease group. We found that the TCA and the CVI differed between the normal subjects and MTM patients. However, the CVI was similar among the surgical groups in the four patients. The results support the use of the CVI as a noninvasive biomarker; it distinguished patients in the MTM group from patients in the healthy control group. It has been reported that age and CVI are inversely related in healthy eyes.^[22] The results also showed that the ILM surgical techniques were inversely correlated with the CVI, but we could not find a statistically significant relationship due to the low number of eyes. This would be a novel finding and signal that this procedure likely does not favor better long-term visual recovery in MTM patients. Regardless of this observation, a surgical intervention significantly improved visual acuity in all participants.

In the present study, we compared the CVI across different stages of MTM with that of the medical control eyes. The CVI for disease stages 1, 2, and 3 differed from that of the medical control group. Because of the small sample size in this preliminary case-series report, no normative data were obtained for each disease stage. Consequently, the data may not be directly comparable for making conclusions, but the results indicate that quantification of the CVI can predict disease severity across MTM stages. Agrawal et al.^[4,5,7] reported the normative database for CVI from a population-based study on 345 healthy eyes through choroidal enhanced depth imaging OCT scans. They reported that the mean subfoveal CVI was 65.61 ± 2.33%. The CVI in the medical group in the present study was 65.35 ± 4.03%, which is comparable to published data.^[4] The CVI in the surgical group was significantly less than that in the medical group and was markedly variable across patients, with a median value of 47.8 ± 5.20%. Yazdani et al.^[2] evaluated the CVI in low myopic and emmetropic eyes. They reported that myopes showed a slightly higher CVI than emmetropes.^[2] Although we presented data from high myopes and the findings are not comparable to those of Yazdani et al.,^[2] the contrasting results indicate that further studies would be required to quantify the CVI in myopic subjects.

The CVI has been proposed as a novel OCT-based parameter to quantify structural changes in underwent scleral buckling procedures revealing low macular perfusion indices,^[27] choriocapillaris flow reduction in placoid chorioretinitis^[28] and type 1 age-related submacular neovascularization,^[29] and eyes with multiple retinal dystrophies, such as retinitis pigmentosa (RP), Stargardt disease, cone-rod dystrophy, Best disease and Bietti crystalline dystrophy.^[30] The literature suggests loss of the choriocapillaris with the overlying RPE in RP, as seen in ex vivo human eyes, suggesting the presence of choroidal vascular changes in inherited retinal diseases.^[31] Wei et al.^[30] reported that the CVI was decreased significantly in patients with retinal diseases, with a mean CVI of 52 ± 9%, whereas the CVI in normal eyes was 70 ± 3%. These findings are comparable to those from our study cohort. Earlier, Tan et al.^[32] also reported that the mean CVI in eyes with RP was significantly lower than that in normal eyes. The authors considered that a lower CVI in choroidal vascular defects in RP could influence the manifestation and progression of the disease, which potentially makes this parameter a surrogate marker in monitoring and analyzing the progression of RP.^[32] The literature suggests that retinal venular and arteriolar oxygen saturation is significantly higher in patients with RP, which also correlates with functional damage.^[33] Iovino et al.^[34] reported that the CVI is decreased significantly in patients with CME associated with RP, indicating a potential role of altered choroidal hemodynamics in the development of CME in RP. Through a retrospective comparison of the CVI and subfoveal choroidal thickness (SFCT) in patients with Stargardt disease and age- and sex-matched healthy controls, Ratra et al.^[35] concluded that the CVI is a more robust tool than the SFCT for assessing choroidal changes in Stargardt disease and can be a potential surrogate marker for disease monitoring. They reported that the CVI was significantly decreased in eyes with the disease compared to those in the normal control group, and there was a negative association between visual acuity and CVI.^[35] The increased choroidal vascularity observed in posterior uveitis is interpreted as an indirect indicator of intraocular inflammation, as reported by Agrawal et al.^[7] and Kim et al.^[9]

Decreased CVI has been reported in tubercular multifocal serpiginous choroiditis.^[3] Agrawal et al.^[4,5,7] also studied the application of the CVI in monitoring Vogt-Koyanagi-Harada disease. They observed that the CVI decreased over time, similar to the choroidal thickness. Choroidal thickness is useful for monitoring initial response to treatment but is not a robust tool to evaluate long-term outcome. Therefore, the CVI is proposed as a better outcome measure in chorioretinal diseases.^[5] Koh et al.^[36] assessed the CVI in patients with AMD and controls. They determined that the SFCT was similar across the study groups, but the CVI was significantly lower in patients with AMD.^[36] Ng et al.^[37] evaluated choroidal structural changes in eyes with myopic CNV treated with anti-VEGF over 12 months. They reported thinning of the subfoveal choroid without alteration in the CVI in treated eyes. More recently, Hwang et al.^[38] reported significantly decreased CVI in cases of branch retinal vein occlusion and macular edema. The authors suggested that the fellow eye CVI could be a useful supplementary prognostic biomarker in these diseases. Another possible mechanism for decreased CVI was considered to be the extracellular fluid shift from the retina to the choroid. Another cause of decreased CVI in BRVO is the pressure effect of choroidal congestion.^[8] The CVI has also been reported to be significantly lower in eyes with central serous retinopathy with CNV.^[39] It is suggested that choroidal vascular alterations may be associated with CNV in these conditions. The reason for the decreased CVI in the MTM cases could be vascular, trophic-tractional degeneration due to high myopia and histopathological changes from PS.

The major limitation of the present study is the small sample size, which prevented us from obtaining normative data for each stage of the disease to facilitate a comparison of outcomes with other diseases. In addition, the correlation of the CVI with retinal functional damage assessment techniques such as multifocal ERG, automated microperimetry and autofluorescence imaging may further increase the understanding of the disease process and the potential to develop noninvasive methods as disease biomarkers. However, those methods were not included in the present study, as we focused only on quantifying the CVI across normal subjects and different stages of the disease.

There has been significant development in imaging tools and assessment techniques for retinal diseases using objective and noninvasive techniques. The CVI is one of the innovative tools recently proposed as a novel marker for choroidal pathogenesis. The present study also showed that the CVI might indicate MTM development in highly myopic eyes. Despite the small sample size, the results are conclusive and strongly support the study’s hypothesis. However, the CVI is not a unique marker of MTM, as it has been previously described for a wide range of retinal diseases. Notwithstanding this specificity limitation, we propose that comparison of the CVI between the two eyes and with normal subjects would be helpful in monitoring pathology in MTM. Further studies are warranted, especially to determine the range of the CVI values for each stage of the disease so that direct comparisons can be made for objective analysis of disease progression.

CONCLUSION

The CVI was strongly associated with MTM and correlated with disease severity in high myopia cases. Further prospective studies analyzed with uni- and multivariate regression formulas are required to determine statistically reliable quantitative data at each stage of the disease for better use of the CVI as a potential biomarker in disease pathogenesis.