Table 2 Summary of cross-sectional and case-control studies on the role of OCTA in assessing diabetic choroidopathy
Study (Quality) | Newcastle-Ottawa Quality Assessment scale | Device | Population Sample size | Outcome measures | Findings |
|---|---|---|---|---|---|
Thaker et al. 2023 [29] | N/A | SS-OCTA | NPDR: 110 patients (176 eyes) | -Choroid circulation flow void | - Choroid circulation flow void was significantly associated with more severe NPDR. |
Bandello et al. 2023 [50] | 9 | -Total: 60 cases -Treatment-naïve mild DR (ETDRS 20–35): 30 -HC: 30 | The CC FD% was associated with foveal mesopic sensitivity (β: -0.234, p = 0.046), parafoveal mesopic sensitivity (β: -0.312, p = 0.032), and parafoveal dark-adapted sensitivity (β: -0.282, p = 0.048). | ||
Deng et al. 2023 [31] | 9 | -T2DM (not severe DR, no DME): 102 cases (194 eyes) -HC: 28 cases (56 eyes) | -CCF -ERG | - CCF was decreased from the control to the nondiabetic retinopathy (NDR) to DR group. In patients with diabetes, CCF was correlated with ERG parameters (coefficient index=-0.601, P < 0.001 with 16 Td-s implicit time; coefficient index=-0.687, P < 0.001 with 32 Td-s implicit time; coefficient index = 0.933, P = 0.035 with 32 Td-s amplitude). CCF was correlated with the RPE thickness and the level of HbA1c (both P = 0.001). | |
Wang et al. 2023 [30] | N/A | SS-OCTA | -T2DM: 1692 cases | -CC FD density -CC FD size -CC FD number | - After controlling for confounding factors, individuals with DR had a significantly higher FD density compared to those without DR. Specifically, the differences were 1.61% (95% CI 1.04 to 2.18; p < 0.001) for individuals with mild NPDR, 2.23% (95% CI 1.76 to 2.70; p < 0.001) for those with moderate NPDR, and 3.31% (95% CI 2.27 to 4.36; p < 0.001) for those with severe NPDR, compared to individuals without DR. - The higher FD number and size were correlated with more severer degrees of DR (all p < 0.05). - Adding FD density to conventional risk factors significantly improved the ability to distinguish DR from NDR patients, with an AUC of 0.829 (95% CI 0.804 to 0.855; p < 0.001). |
Viggiano et al. 2023 [52] | 9 | SS-OCTA | -T1DM (n = 40) including: NDR: 12, and mild NPDR: 28 -HC: 10 | -OCTA: CC FD% -AO: CD, LDi, HPi | - A close relationship was showed between cone metrics and CC FD in NPDR group. - The percentage increase in CC FD was linked to an increase in LDi (p = 0.035). Moreover, the rise in CC FD% was found to be correlated with a decrease in CD (p = 0.042) and HPi (p = 0.017). |
Tan et al. 2023[20] | 8 | -HC: 27 cases (35 eyes) -DM: 75 cases (132 eyes), including: NDR: 62 eyes NPDR: 51 eyes, and PDR: 19 eyes | -CC flow void density, % (3*3 mm; fovea-centered) | - CC parameters were strongly altered with DR stages (p < 0.01). - The CC parameters were more effective in distinguishing between the group with no DM & NDR, compared to the retinal parameters (with AUC of 0.954 vs. 0.821, p = 0.006). - Combining both retinal and choroidal microvasculature parameters in a classification model resulted in a significant enhancement in distinguishing between DR and the absence of DR, compared to using each parameter independently (p = 0.029). | |
Xiong et al. 2022 [17] | 9 | SS-OCTA | -DM: 71 cases -High myopia: 71 -DM + High myopia: 71 -HC: 71 | -CC perfusion index | The mean CC perfusion index for the control, diabetes, high myopia, and diabetes with high myopia groups were 91.11 ± 0.84%, 90.16 ± 1.46%, 89.80 ± 1.42%, and 89.36 ± 1.19%, respectively (P < 0.001). |
Kaderil et al. 2022 [49] | N/A | SD-OCTA | -Total: 58 patients (95 eyes), including: 1- NPDR: 59 2- PDR: 36 (2 A: VA ≤ 0.2 logMAR; 2B: > 0.2) | -CC plexus flow area | - CCP flow area was lower in group 2B (All, p < 0.05). |
Liu et al. 2022 [41] | 7 | SS-OCTA | -NPDR: 42 cases (71 eyes) -PDR: 31 cases (53 eyes) -HC: 30 cases (51 eyes) | -choroid perfusion | - The choroid perfusion exhibited significant differences across different areas and amongst the three groups. |
Zhang et al.,2021 [10] | 7 | -36 diabetic patients (54 eyes), including: NDR: 31 eyes, Mild NPDR: 19 eyes, and Moderate NPDR: 4 eyes -32 control subjects (54 healthy eyes) | -CC blood flow signal density | Within the diabetic group, the CC blood flow signal density in the macular area (diameter = 2000 μm) exhibited a significant reduction compared to the healthy control group (P < 0.05). | |
Ryu et al. 2021 [39] | 7 | -DM NDR: 49 -NPDR: 51 -PDR: 38 -HC: 52 | -CC vessel density | In NDR group, CC vessel density was found to be decreased only in the perifoveal area (with p-values of 0.823 for the foveal area, 0.631 for the parafoveal area [1–3 mm], and 0.039 for the perifoveal area [3–6 mm]). Multivariate linear regression analyses showed significant associations between DR severity and all retinal and choroidal microvascular indices. | |
Zlatanović et al. 2021 [16] | 8 | SD-OCTA | -T2DM without DR: 83 cases (166 eyes); age: 59 ± 14 y -HC: 33 cases (66 eyes) | -CC flow area and vessel density | - There was a statistically significant reduction in the CC flow area and vessel density in the NDR group compared to healthy subjects. |
Parravano et al. 2021 [53] | N/A | -T1DM, including: NDR: 18 NPDR: 25 | -EZ “normalized” reflectivity, -mfERG rRADs -FDn, FDa, & FD% | -The mean values for EZ “normalized” reflectivity, mfERG RAD, CC FDn, and FD% were not significantly different between the NPDR and NDR groups (p > 0.05), but FDa was significantly higher in NPDR eyes (p < 0.05). -In NPDR eyes, there was a significant linear relationship between decreased mfERG RADs and increase in either CC FDa and FD%. -In NDR eyes, EZ “normalized” reflectivity was negatively correlated with CC FD%. | |
Stulova et al. 2021 [19] | 7 | -T1DM NDR: 41 patients (76 eyes) -HC: 31 (55 eyes) | -CC FDs | -Patients with T1DM showed an increase in both the density and mean size of FDs. | |
Ra et al. 2021 [42] | N/A | -T2DM: 152 cases (282 eyes), including: NDR: 114 NPDR: 79 S-NPDR: 48 PDR: 41 | -CC vascular density -Choroidal vascular density | -CC vascular density was lower in patients with PDR compared to those with NPDR (P < 0.05). -Significant positive correlations were found between superficial and deep retinal vascular density and CC vascular density (all P < 0.001), while choroidal vascular density showed negative correlations with these parameters (also P < 0.001). -Both retinal and CC vascular density were negatively correlated with diabetic retinopathy grade (all P < 0.001), while choroidal vascular density had a weak positive correlation (P = 0.030). | |
Ucgul et al. 2021 [55] | N/A | -T2DM, including: NDR without MA: 20, NDR + MA: 20, and Mild NPDR + MA:30 | - CC vascular density | There were significant inverse correlations between diabetes duration, creatinine, urea, serum Na, and certain CC vessel density values (with p-values below 0.05 for all correlations). | |
Ghassemi et al. 2021 [43] | N/A | -DR 97 cases (188 eyes) | -CC vascular density (macula 3*3 mm) | - There was a consistent pattern of decreasing CC vascular density from normal cases to cases of NDR and NPDR, followed by a slight increase in the PDR stage, but the vascular density never reached normal levels. - BCVA was positively correlated with CC vascular density of foveal area. | |
Agra et al. 2021 [27] | 8 | -T2DM NDR: 30 cases (30 eyes) -HC: 30 cases (30 eyes) | -CC flow area | -Patients with diabetes showed a slight increase in CC flow area compared to controls (mean area of 22.6 ± 3.9 mm2 vs. 22.3 ± 4.6 mm2, p = 0.017). -There was a positive correlation between fasting blood glucose levels and CC flow area (p = 0.034). | |
Loria et al. 2021 [15] | 7 | SS-OCTA | -HC: 17 -NDR: 30 -Minimal NPDR:22 -Moderate NPDR:30 -Severe NPDR:16 -PDR: 5 | -CC Flow void area | -There was a significant positive correlation between CC flow void area and DR stage. Furthermore, FVA-CC was significantly higher in diabetic patients without DR compared to healthy individuals (P = 0.008). |
Ro-Mase et al. 2020[51] | 8 | SD-OCT | -T2DM: 26 cases including: NDR: 4, NPDR:12, and PDR:10 -HC: 13 | -CC FD -Microperimetry: Retinal sensitivity -AO: HPi | -In patients with NPDR and PDR, there was a significant correlation between foveal and CC FD and retinal sensitivity (fovea, r = -0.58; P = 0.046 and r = -0.82; P = 0.003; parafovea, r = -0.59; P = 0.044 and r = -0.72; P = 0.019, respectively), but not in control and NDR groups. -No significant differences were found in HPi across all groups. |
Dai et al. 2020 [1] | 8 | SS-OCTA | -DM: 45 with DR -HC: 27 -Extreme axial lengths of > 25.2 mm or < 23.2 mm were excluded | -CC FD (5-mm from 6*6 mm scans) | - Compared to control eyes, diabetic eyes had a significantly higher CC FD% (12.34 ± 4.14% vs. 8.82 ± 2.61%, P < 0.001) and larger mean CC FD size (2151.3 ± 650.8µm2 vs. 1574.4 ± 255.0 µm2, P < 0.001), with both values being 1.4-fold greater on average. - There was no significant difference in CC FD% or mean CC FD size between eyes with NPDR and those with PDR (P = 1.000 and P = 1.000, respectively). |
Saif et al. 2020 [9] | 7 | SS-OCTA | -HC: 16 eyes -NDR: 16 eyes -NPDR: 16 eyes -PDR: 16 eyes | -Choroidal vascular density | -There was a decrease in choroidal vascular density in diabetic patients compared to healthy control group. |
Ashraf et al. 2020[35] | N/A | 3*3 mm macular scan | -225 cases Type 1 or Type 2 DM (352 eyes), including: Mild NPDR:183, Moderate NPDR:71, Sever NPDR or PDR:98 | -CC FD | Among eyes without predominantly peripheral lesions, the mean CC flow density decreased with increasing DR severity (mild NPDR, 69.7% [6.2%]; moderate NPDR, 67.6% [5.6%]; severe NPDR or PDR, 67.1% [5.6%]; P = 0.01). -However, in eyes with predominantly peripheral lesions, the mean CC flow density did not appear to change with increasing DR severity (mild NPDR, 67.1% [5.6%]; moderate NPDR, 69.3% [4.6%]; severe NPDR or PDR, 68.3% [5.6%]; P = 0.49). |
Dai et al. 2020 [6] | 8 | SS-OCTA | -NDR: 16 cases (16 eyes) -HC: 16 | -CC FD% -CC FD size | -In diabetic eyes, compared to controls, there were significant increases in mean FD% and mean FD sizes in the central 1.0-mm disk (P = 0.011 and P = 0.017), the central 1.5-mm rim (P = 0.003 and P = 0.009), the central 2.5-mm rim (P = 0.018 and P = 0.020), and the entire 5.0-mm disk (P = 0.009 and P = 0.008). -There were no significant differences in any retinal vessel quantitative parameters between the two groups (all P > 0.05). |
Gendelman et al. 2020 [7] | N/A | SS-OCTA | -90 DM cases (160 eyes), including: NDR: 33, Mild NPDR:17, Moderate NPDR: 8, Severe NPDR: 10, and PDR:22 | -CC FD | -Age and severity of DR were significantly and positively associated with FD% in all studied regions, with a greater effect observed in the two centermost regions. -The increase in FD percentage per year of age varied by region: 0.12 for the inner (p < 0.001), 0.09 for the middle (p < 0.001), 0.05 for the outer (p < 0.001), and 0.06 for the full-field (p < 0.001). -The increase in FD percentage per increase in DR severity stage also varied by region: 0.65 for the inner (p < 0.0087), 0.56 for the middle (p < 0.0012), 0.33 for the outer (p < 0.045), and 0.36 for the full-field (p < 0.018). |
Lupidi et al. 2020 [37] | 7 | -29 diabetic patients with Level 20 DR severity score -HC: 20 | -CC vascular perfusion density -SCP and DCP vascular perfusion density | In diabetic patients, vascular perfusion density values were significantly lower in the DCP (25.1% vs. 26.5%; p = 0.04) and CC (71.2% vs. 86.6%; p = 0.0001) compared to controls. -A negative linear correlation was found between CC vascular perfusion density and DCP vascular perfusion density in diabetic patients, while a positive linear correlation was observed between the same parameters in controls. | |
Borrelli et al. 2020 [54] | 7 | -NPDR: 30 -HC: 30 | -EZ normalized reflectivity -CC perfusion density | -The NPDR group had impaired perfusion in both retinal and choroidal vasculature, and a lower “normalized” reflectivity compared to controls (0.73 ± 0.19 vs. 0.96 ± 0.25, P < 0.0001). -Multiple regression analysis showed a significant direct association between EZ “normalized” reflectivity and CC perfusion density in NPDR patients (P = 0.025) but not in controls (P = 0.476). | |
Forte et al. 2020 [14] | 7 | SS-OCTA | -T1DM NDR: 17 -T2DM NDR: 17 -23 HC | -CC voids | -There was a higher frequency of CC voids in both DM Type 1 and DM Type 2 compared to controls (P = 0.003 and P < 0.001, respectively). |
Sacconi et al. 2019 [23] | 9 | SD-OCTA | -T1DM NDR: 34 cases (34 eyes) -HC: 27 | -CC perfusion density | -OCTA showed lower deep capillary plexus perfusion density in diabetics than the control group, but no significant differences were found in other retinal/choriocapillaris plexuses or foveal avascular zone area. |
Mastropasqua et al. 2019 [46] | 8 | SS-OCTA | -94 cases (95 eyes), including: NDR: 25, Mild NPDR: 23, Mod/sever DR: 26, and PDR: 20 -HC: 25 | -CC, SCP and DCP perfusion density (central vs. peripheral) | -The prediction of disease worsening had excellent specificity and good sensitivity, particularly in the central and temporal sectors across all plexuses. |
Yang et al. 2019 [4] | 8 | -HC: 43 -NDR: 56 -Mild DR: 43 -Moderate DR: 54 -Severe DR: 38 -PDR: 48 | -CC flow density | -As DR progressed, there was a downward trend in CC flow density in the CC layer. -Significant differences were found between mild NPDR and moderate NPDR and between severe NPDR and PDR in the comparison of CFD in the CC layer using both 3-mm and 6-mm scan patterns (P = 0.003, P = 0.001). | |
Yang et al. 2019 [38] | 8 | -HC: 40 -Mild DR: 40 -Moderate DR: 40 -Severe DR: 40 -PDR: 40 | -CC vascular density | -As DR severity increased, there was a reduction in macular perfusion in the SCP, DCP, and CC. -Vessel density in the DCP was found to be a better indicator of DR severity, with an area under the curve, sensitivity, and specificity of 0.967, 92.5%, and 93.1%, respectively, compared to vessel density in the SCP and CC. | |
Conti et al., 2019 [26] | 7 | SD-OCTA | -HC: 37 eyes -NDR: 31 eyes -NPDR: 41 eyes -PDR: 27 eyes | -CC capillary perfusion density | -Eyes affected by both NPDR and PDR exhibited a significant reduction in choriocapillaris capillary perfusion density when compared to controls. In contrast, diabetic eyes without retinopathy did not display a noteworthy change. -The whole-image capillary perfusion density of the CC declined by 8.3% in NPDR eyes (p < 0.01) and by 7.1% in PDR eyes (p < 0.01). Additionally, parafoveal capillary perfusion density in the choriocapillaris showed an 8.9% decrease in NPDR eyes (p < 0.01) and an 8.2% decrease in PDR eyes (p < 0.01). |
Rodrigues et al. 2019 [36] | N/A | SD-OCT | -NPDR: 56 cases (101 eyes) | -CC vascular density | -Univariate analysis revealed that higher ETDRS level was associated with several OCTA parameters, including parafoveal SCP density (OR = 0.87, 95% CI 0.76–0.99, p = 0.039), parafoveal DCP density (OR = 0.79, 95% CI 0.72–0.87, p < 0.001), and CC density (OR = 0.89, 95% CI 0.80–0.99, p = 0.036). -However, after adjusting for relevant clinical features, only parafoveal vessel density in the DCP remained a significant predictor of NPDR ETDRS level (OR = 0.54, 95% CI 0.32–0.92, p = 0.024). |
Li et al. 2019 [48] | 8 | SD-OCTA | -T-2DM: 97 patients (NDR to PDR) -HC: 48 | -Foveal and parafoveal CC vascular density -Foveal flow area in CC plexus | -Compared to controls, diabetic patients had lower flow area in CC plexus and vascular density in all three layers. -In the NDR group, foveal flow area in CC plexus decreased significantly compared to controls. In mild NPDR, parafoveal VD decreased significantly in all three layers compared to NDR, particularly in the temporal and nasal areas. In moderate NPDR, vascular density reduction extended to the inferior area in SCP and DCP compared to mild NPDR. In severe NPDR, there were progressive losses of vascular density in all layers compared to moderate NPDR. In PDR, the superior vascular density in SCP increased significantly compared to severe NPDR. |
Li et al. 2018 [18] | 7 | SD-OCTA | -T2DM NDR: 42 cases (42 eyes) -HC: 40 | -CC flow area | -Reduced CC flow area was significantly more obvious in the NDR subjects than in the control subjects (1.94 ± 0.28 versus 2.05 ± 0.11 mm2, p = 0.02). |
Cao et al. 2018 [56] | 7 | SD-OCTA | -T2DM NDR: 71 -HC: 67 | -CC vessel density | -NDR group showed a significant reduction in average vessel density of SCP, DCP, and CC (p < 0.001, p < 0.001, and p = 0.006, respectively). |
Carnevali et al. 2017 [24] | 7 | SD-OCTA | -T1DM NDR; 25 cases -HC: 25 | -CC vessel density | -Comparison of the DCP between diabetic and control eyes showed a significant decrease in vessel density in the diabetic group [0.464 ± 0.016 vs. 0.477 ± 0.014, respectively (p = 0.005)]. -However, no significant differences were found in vessel density in the all-retina plexus, SCP, or choriocapillaris. |
Dimitrova et al. 2017 [25] | 9 | SD-OCTA | -DM NDR: 29 -HC: 32 | -CC vessel density | -Compared to control subjects, NDR patients had decreased vessel densities in both superficial and deep retina [44.35% ± 13.31% and 31.03% ± 16.33% vs. 51.39% ± 13.05%, P = 0.04; and 41.53% ± 14.08% vs. control, P < 0.01, respectively]. -They also observed a tendency towards decrease in CC vessel density in diabetic eyes without DR compared to healthy subjects; however, the difference did not reach statistical significance. -There was a significant negative correlation between CC density and diastolic blood pressure in the NDR group (r = − 0.42, P = 0.02). |
Choi et al. 2017 [5] | 8 | SS-OCTA | -HC: 32 cases (63 eyes) -PDR: 7 (9 eyes) -NPDR: 16 (29 eyes) -DM NDR: 28 (51 eyes) | -CC flow impairment | -CC flow impairment was observed in patients with either PDR or NPDR. Furthermore, CC flow impairment was found in 24 out of 51 diabetic eyes with no DR. |