Introduction

Diabetic retinopathy is a prevalent sight threatening disorder of retina which is caused by abnormalities in capillaries and vessels of retina in patients with diabetes. It is the worldwide leading cause of loss of vision.1 The development of macular edema and proliferative retinopathy are among the important cause of impairment of vision.1, 2

The choroid is highly vascularised tissue that provide blood supply to outer retina, including photoreceptors and retinal pigmented epithelial cells (RPE) especially in fovea region where there is no retinal vasculature.3 It is the only source of oxygen and nutrients to outer retina and RPE and plays an important role in the pathophysiology of the diabetic retinopathy(DR).

Clinical and experimental findings have suggested that choroidal vasculopathy in diabetes might play a role in progression and severity of DR.4, 5, 6 Although diabetic retinopathy is a disease that primarily affect microvasculature of retina, the concept of diabetic choroidopathy came into light when it was first established that there is significantly higher loss of choroidocapillaris in diabetic subjects as compared to aged control healthy controls.7 Several abnormalities of choroid including vascular degeneration, obstruction of choriocapillaries, choroidal aneurysm and choroidal neovascularisation have been reported in histological studies of diabetic eye.4, 7, 8

Choroidal angiopathy has been received less research attention. Until the recent time choroid could have been evaluated only by laser Doppler flowmetry (Doppler flowmeter has also shown decreased choroidal perfusion),9 ultrasound and indocyanine green angiography.10 The techniques though useful for finding vascular abnormalities or any changes in choroidal blood flow but they did not give any anatomical information about RPE or choroidal layers.

Spectral domain optical coherence tomography (SD-OCT) now provides a high quality high resolution cross sectional images of macula. Additionally, enhanced depth imaging (EDI) software provides a very reliable and reproducible estimation of the choroidal thickness. EDI-OCT is unique in having capability of tracking fovea, was recently described as a new method to measure thickness of choroidal in normal and pathologic states.11, 12, 13, 14, 15, 16

Studies on association of thickness of choroid with diabetes or with progression of diabetic retinopathy have shown contradictory results. Some studies have shown increase in choroidal thickness17, 18 while other have shown decrease in thickness of choroid19, 20, 21, 22, 23 in patient with diabetes or with increase in severity of DR.

In view of scarcity of definitive evidence of change in choroidal thickness in association with diabetes/diabetic retinopathy we have done cross-sectional study to compare subfoveal choroidal thickness in nondiabetic and diabetic also in various grades of diabetic retinopathy.

Materials and Methods

All patients having diabetes coming to ophthalmology OPD in a tertiary care center in central India from February 2019 to August 2020 who fulfil the given inclusion criteria as well as met no exclusion criteria were included in the study.

Exclusion criteria

1. History of vitreoretinal surgery

2. Vitreoretinal disorders other than diabetic retinopathy currently or in past.

3. Cataract surgery in past 6 months.

4. Spherical equivalent of refractive error more than or equal to +/- 6D.

5. Any media opacity likely to cause signal strength attenuation in OCT.

6. Signal strength <6/10 in OCT.

7. Panretinal photocoagulation (PRP) treated within 3 months.

8. History of intravitreal anti-VEGF.

After taking a written informed consent a comprehensive history has been taken that included a detailed ocular and systemic (duration of diabetes and antidiabetic medication) history, demography (age, sex) laterality, systemic co-morbidities (hypertension, kidney disease). General examination and systemic examination of associated systemic diseases was done.

The ophthalmic examination included assessment of BCVA on Snellens which was then converted into log MAR value, intraocular pressure, anterior segment examination, fundus examination, digital fundus photography, fundus fluorescein angiography (if required), OCT using Cirrus HD OCT Model 500 were done and relevant investigations were advised.

As per ETDRS classification, patients were categorized into mild, moderate and severe non-proliferative diabetic retinopathy (NPDR) and proliferative diabetic retinopathy (PDR). Each patient with diabetic retinopathy were also grouped on the basis of presence or absence of diabetic macular edema (DME). Patients were also divided into PRP untreated DR and PRP treated DR with history of last session of PRP of more than 3 months.

The SFCT was measured using EDI-HDOCT technique and the scan passing through foveola, was selected. Measurement was taken as the vertical distance between hyper-reflective line of Bruch’s membrane to the innermost hyper-reflective line of chorio-scleral interface. (Figure 1)

Figure 1

Showing SFCT in enhanced depth imaging high definition raster scan of OCT in moderate diabetic retinopathy patient.

Result

During the course of study, a total of 400 eyes of 200 patients were included in this study which were categorized as 320 eyes of patients having diabetes and 80 eyes of age matched healthy controls.

Three twenty eyes were divided on the basis of severity of DR into diabetics with no diabetic retinopathy (50 eyes), mild NPDR (60 eyes), moderate NPDR (62 eyes), severe NPDR (70 eyes), PDR (48 eyes) and PRP treated DR (30 eyes). On the basis of presence of DME, patients were categorized as 75 eyes with DME and 195 eyes without DME.

Patients with DR were also divided on the basis of history of last session PRP more than 3 months into PRP treated DR (30 eyes) and PRP untreated DR (240 eyes).

The base line features of cases and controls including mean age, sex, duration of diabetes and mean BCVA were measured.

The mean SFCT in diabetics and age matched healthy controls were 259.93 ± 53.7 μm and 368.68 ± 10.18 μm respectively. It was significantly lesser in diabetics than age matched healthy controls having p value<0.001.

The mean SFCT values in diabetics without DR was 345.6 ± 9.11μm and diabetics with DR was 244.07 ± 42.29μm. It was significantly greater in diabetic without DR as compared to diabetic with DR (p value<0.001).

While measuring SFCT in different grades of DR, the mean SFCT values were no DR, mild, moderate and severe NPDR and PDR were 345.60 ± 9.11μm, 302.28 ± 19.19μm, 266.56 ± 15.46 μm, 227.27 ± 8.47 μm and 201.4 ± 7.77μm respectively. It implied that SFCT decreased significantly as the severity of DR increased with p value<0.001. The SFCT further decreased significantly in PRP treated eyes (mean SFCT=188.63 ± 10.22μm) with p value<0.001.

The mean SFCT values in diabetic without DME (267.09 ± 56.82μm) and diabetic with DME (236.56 ± 32.62 μm) showed significant differences, it was significantly greater in diabetic without DME than diabetic with DME (p value<0.001).

The mean SFCT in PRP untreated DR eyes and PRP treated DR were 251 ± 39.58μm and 188.63± 10.22μm respectively showed that it significantly greater in PRP untreated DR than PRP treated DR with p value<0.001.

Figure 2

Showing distribution of patients

Figure 3

Showing distribution of patients according to presence of DME

Figure 4

Showing SFCT in non-diabetics and diabetics

Figure 5

Showing SFCT in diabetic with and without diabetic retinopathy

Figure 6

Showing trend of SFCT with severity of DR

Figure 7

Showing SFCT in diabetic with and without DME

Figure 8

Showing SFCT in PRP treated and PRP untreated DR

Discussion

Diabetic retinopathy is the common sight threatening retinopathy. It is the worldwide leading cause of loss of vision.1 The major cause of visual impairment are development of macular edema and proliferative retinopathy.1, 2 The choroid is highly vascularised tissue that provide blood supply to outer retina, including photoreceptors and RPE especially in fovea region where there is no retinal vasculature.3 It is the only source of oxygen and nutrients to outer retina and RPE and plays an important role in the pathophysiology of the DR.

While comparing SFCT in diabetics with age matched healthy controls, the mean SFCT was significantly greater in eyes of healthy controls as compare to eyes of patients having history of diabetes mellitus. The reason for thinning of choroid might be as a result of thinning and dropout of choroidal vessels due to luminal narrowing of capillaries in diabetic retinopathy. Our findings were in consistent with previous studies done by Sudhalkar A et al,24 Verma A et al25 and Wang X et al.26 While Wang H et al27 found no significant difference between two groups in their study.

Comparison of SFCT in diabetics with and without DR revealed that the mean SFCT was significantly greater in diabetic with no DR group as compare to diabetic with DR group. The finding in our study were consistent with study done by Sudhalkar A et al24 and Ambiya V et al.28

When SFCT in various grades of DR decreased significantly as the severity of diabetic retinopathy increased. It further decreased significantly in PRP treated eyes. Wang S et al29 and Sudhalkar A et al24 also found similar observation. Ambiya V et al28 noted a significant moderately negative correlation of SFCT with severity of DR i.e.as compare to healthy control SFCT decreased in mild NPDR and moderate NPDR but not statistically significant however decrease in severe NPDR and PDR was statistically significant. Wang H et al27 and Rewbury R et al30 found that SFCT significantly increase with increasing severity of DR as compare to diabetic eyes and healthy controls.

The mean SFCT was found to be significantly greater in eyes without DME as compare with eyes with DME (p value<0.001). Our finding was consistent with finding noticed by Querques G et al.19 The reduced SFCT may cause hypoxia of retinal tissues which leads to increased expression of VEGF in RPE, pericytes and microvascular endothelial cells and may induce breakdown of blood retinal barrier, that forms the basis of diabetic macular edema. Sudhalkar A et al24 found no significant difference between two groups, while Kim JT et al18 found increase in SFCT in DME group.

The SFCT was compared in PRP treated DR eyes with PRP untreated DR eyes which was found to be significantly greater in PRP untreated DR as compare with PRP treated DR. Similar observation was noted by Kang HM et al21 and Kim JT et al.18 Wang H et al27 also found greater mean SFCT in PRP untreated DR as compare with PRP treated DR but that difference was not statistically significant. Sudhalkar A et al24 observed no significant difference in between PRP untreated DR and PRP treated DR.

PRP causes photoreceptor cells and RPE coagulation, also affects choriocapillaris. It also causes obliteration of choriocapillaris, and the choroidal blood flow is also significantly slower in PRP-treated eyes. In addition to impairment of choroidal circulation caused by PRP itself, improved inner retinal layers oxygenation and decreased VEGF secretion may also contribute to decreased permeability of choroid and alteration of choroidal autoregulation which results in generalized thinning of choroid. Thus, both directly and indirectly, various choroidal changes produced by PRP, may affect choroidal circulation.

Limitations

Conclusion

We conclude that the SFCT could be a marker of severity of diabetic retinopathy. It was decreased in diabetics as compared to nondiabetics and was significantly decreased with increasing severity of DR and in presence of DME. PRP also significantly decreases SFCT.

Conflict of Interest

None declared.

Source of Funding

None.