Postoperative Evaluation of Retinal and Choroidal Perfusion in Diabetic Tractional Retinal Detachment

Abstract

Optical coherence tomography angiography (OCT-A) is useful for quantitative analyses of different choroidal and retinal vascular plexuses. Highlighting postoperative choroidal and retinal perfusion outcomes in patients who have undergone successful tractional retinal detachment (TRD) repair is crucial for understanding the impact of this condition on postoperative visual acuity. This chapter describes postoperative perfusion outcomes, such as vessel density (VD) quantified in the superficial and deep capillary plexuses of the retina and choroidal perfusion markers, such as the choroidal vascularity index (CVI) and choriocapillaris flow area (CFA). In this analysis, superficial and deep capillary plexuses were quantified, and the CVI and CFA were significantly lower in the surgical group (P = 0.0011), with median CVIs of 57.95% and 2.28 mm2 in the control group and 44.41% and 1.38 mm2 in the surgical group, respectively. Definitive correlations were shown between alterations in the structure of the retina and choroid after surgery and visual dysfunction in diabetic individuals. The CVI and CFA can be used as quantitative measures to evaluate choroidal damage in postoperative patients with traction retinal detachment. The CVI serves as a dependable quantitative biomarker for evaluating the progression of diabetic retinopathy (DR) or for tracking postoperative eyes.

1. Introduction

The prevalence of diabetes mellitus (DM) is increasing worldwide, affecting an estimated 285 million adults (6.4%) in 2010 [1]. It is considered to be the main cause of preventable visual morbidity and blindness worldwide. Until 2008, the prevalence of vision-threatening proliferative diabetic retinopathy (PDR) was 4.4%, with a general prevalence of diabetic retinopathy (DR) of 28.5% among adults with diabetes in the United States. DR is caused by chronic microangiopathy, which induces progressive uncontrollable ischemic changes in the inner and outer retina, damaging the permeability of the blood-retinal barrier (BRB) [2].

The main causes of vision loss in the diabetic population include various clinical types of diabetic macular edema (DME) and severe complications of high-risk PDR, including active recurrent vitreous hemorrhage (VH) and tractional retinal detachment (TRD) of different configurations and extensions as the final fibroproliferative stage of PDR; which require surgical repair because of the high risk of severe vision loss [3, 4, 5]. Chronic nonresponsive macular edema associated or not associated with vitreomacular traction and the occurrence of epiretinal membrane (ERM) proliferation are also common causes of vision loss and visual disability and constitute the most prevalent challenges in ophthalmology and diabetic eye care [3, 4, 5].

An approach used to manage DR-related bleeding complications involves repetitive injectable intervention protocols consistent with intravitreal injection of different specific target agents against signaling proteins involved in the angiogenesis cascade called vascular endothelial growth factors (anti-VEGFs) to improve BRB stability [6]. They are widely available for use worldwide, especially for the treatment of aggressive proliferative DR [6]. Similarly, some patients are boosted with steroids, periocular injections, or intravitreal extended-release devices to improve long-term visual outcomes. Panretinal photocoagulation (PRP) is the gold standard of care, but currently approved practices in developed countries prefer this approach as second-line treatment [7].

Sivaprasad et al. [8] defined a protocol for diabetic retinopathy titled Clinical efficacy and mechanistic evaluation of aflibercept for proliferative diabetic retinopathy (CLARITY) and the Research Network Protocol, which has guided medical approaches for treating DME and PDR bleeding complications. TRD involving the macula is considered a late and severe sight-threatening complication, and surgical approaches such as advanced updated control-perfusion pars plana vitrectomy (PPV) have been routinely performed [7]. Extensive research has failed to predict the factors associated with better visual outcomes after PPV [9, 10, 11, 12].

Optical coherence tomography angiography (OCT-A), a novel noninvasive digital advanced imaging technology, has emerged to facilitate the visualization of blood flow in the most critical eye tissues. OCT-A technology exploits moving red blood cells based on the diffractive particle movement principle to critically resolve vessel location and characterize the different vascularized tissues of the eye. OCT-A imaging produces three-dimensional composites of the layers of the vascular choroid and the different retinal capillary plexuses without the use of contrast agents, which can be visualized with high-definition depth-encoded and en-face slabs; the different vascularized tissue slabs merged along with the structural OCT B scans to be analyzed, and detailed blood flow imaging of the superficial and deep retinal vascular plexuses and choriocapillaris flow can be obtained with quantified options, which were not well visualized with previous imaging modalities [13]. Because this imaging technique overcomes several significant limitations of previous dye-based techniques for evaluating the choroid, OCT-A has become a key tool for evaluating the choroid, offering new and important insights into the pathogenesis of many retinal and choroidal disorders. This technology has increased the sensitivity of detailed imaging of different manually or digitally selected choriocapillaris submacular regions and specific areas of interest at the level of the deep and superficial retinal capillary networks [13]. OCT-A perfusion analysis uses motion as a contrast mechanism to visualize the location of moving intravascular cells [13]. OCT-A is a novel, noninvasive, free-dying technology capable of avoiding indocyanine green dye (ICG) dark artifacts of retinal and choroidal vascular features that are present owing to the leakage of dye molecules [14, 15, 16]. This high-resolution technique provides capillary-level details such as histological information [17, 18, 19, 20].

OCT-A has been useful for assessing DR and clear media PDR, as it delineates foveal avascular zone (FAZ) abnormalities, microvascular occlusions, end-vascular-occluded fine vascular segments, the presence of saccular or fusiform microaneurysms and clearly demonstrates the characteristics of the deep capillary plexus (DCP) and superficial capillary plexus (SCP). In addition to OCT-A, the Optovue AngioVue system enables quantitative analysis of the different capillary plexuses of the retina and choroid, serving as a marker for the diagnosis and surveillance of medical or surgical approaches [21]. Fundus fluorescein angiography (FFA) cannot visualize the details of peripapillary radial vascular or deep capillary networks [16, 17, 22]. OCT-A has been demonstrated to be a superior digital technique for the quantitative analysis of different vessel density (VD) metrics and objectively evaluates choroidal and retinal perfusion patterns to detect all vascular-relevant findings in the eyes of patients with DR [22, 23]. OCT-A techniques also reliably detect abnormal vascularization at any level in choroidal or retinal tissue (Figure 1) [24, 25, 26].

Figure 1.

OCT-A image of the superficial capillary plexus (SCP) (a) and deep capillary plexus (DCP) (b) showing fine vessel abnormalities and vessel tortuosity (VT) in a diabetic eye. Multiple small vascular areas of nonperfused areas are observed throughout the macula in both plexuses. (c) In the bottom image, a corresponding structural high-definition (HD) image shows some degree of disorganization of the retina inner layers (DRIL) with well-demarcated outer retina structures pinpointed by automatic segmentation of the green lines. The purple and red dots represent vascular structures in the choroid and retina, respectively. (d) Corresponding binarizing processing of the choriocapillaris flow area (CFA) of 2.104 mm² in the selected protocol for a total area of 3.142 mm². This modified multipanel figure was obtained from a previous publication [27] by Quiroz-Reyes et al. and is used under the Creative Commons Attribution 4.0 International License (https://creativecommons.org/licenses/by/4.0/).

OCT-A is a potentially valuable imaging tool for improving DR treatment outcomes and monitoring injectable intervention protocols. In addition, deleterious changes in retinal and choroidal perfusion markers and patterns that occur after vitrectomy for TRD or due to the progression of diabetic microangiopathy may be detected and consequently treated with injectable interventions such as sustained anti-VEGF therapy [19].

This chapter examines the importance of retinal and choroidal perfusion evaluations in diabetic patients with DR and TRD who have undergone PPV and their role in guiding treatment to improve functional outcomes. We explored the influence of surgical approaches on postoperative superficial and deep retinal perfusion plexuses and the correlation of the choroidal vascularity index (CVI) and choriocapillaris flow area (CFA) with final best-corrected visual acuity (BCVA). After a thorough analysis of PubMed and scholarly literature, complemented by the authors’ experience, we aimed to increase the comprehension of TRD management and ultimately improve patient visual outcomes for this serious sight-threatening condition by detecting postoperative retinal and choroidal ischemic alterations using sophisticated OCT-A imaging technology.

1.1 Preoperative and postoperative functional and structural evaluation protocol

Surgical patients must be evaluated by recording BCVA values, biomicroscopy, fundus examination with different fundoscopic lenses, indirect ophthalmoscopy (different TRD clinical cases with clear media are presented in Figures 2 and 3), and high-resolution vitreoretinal mode B-scan ultrasonography. Horizontal and vertical cross-sectional images of the macular region should be obtained using SD-OCT in eyes with clear media. In this study, OCT images were captured using a Ret-Vue-3.4 OCT (Optovue, Inc., Fremont, CA, USA). The measurement of axial length using coherence laser interferometry (IOL Master 700; Carl Zeiss Meditec AG, Oberkochen, Germany) is highly recommended. In eyes with opaque media, TRD should be confirmed and characterized by high-resolution B-scan ultrasonography. Postoperative microstructural evaluations should be performed using an SD-OCT device (images shown in this chapter were obtained using a Spectralis OCT; Heidelberg Engineering, Heidelberg, Germany) and a swept-source DRI OCT Triton device (Topcon Medical Systems, Inc.). We regularly performed postoperative perfusion evaluations, quantitative VD measurements, and quantitative perfusion evaluations of the choriocapillaris flow using an OCT-A device (images in this chapter were captured using RTVue XR OCT Avanti with AngioVue Software; OptoVue Inc., Fremont, CA, USA). This equipment uses a specialized split-spectrum amplitude-decorrelation angiography software algorithm to acquire 70,000 A-scans/s to construct OCT-A volumes consisting of 304 Å–304 A-scans, achieving a high axial resolution at depths of up to 5 μm and minimizing motion artifacts. Each OCT-A cube scan comprised 304 Å–304 A-scans within a 3 mm × 3 mm square centered around the fovea, which yielded 304 B-scans. Each B-scan output displays an average of at least two scans. Segmentation of the SCP, DCP, outer retinal layer, and CFA slabs is obtained using AngioVue software. The central subfield thickness (CSFT) is automatically calculated using standardized algorithms contained in the device software. Perfusional choroidal measurements were obtained following a previously published protocol using the aforementioned OCT-A device [27]. CFA should be performed by segmenting the choriocapillaris subfoveal plexus (CSP) slabs using an OCT-A instrument that can manually locate an area of interest and digitally calculate the VD in a preselected area. Only high-quality images with a standard signal strength index (SSI) > 46 provided by the software for scan quality are used for perfusion analysis.

Figure 2.

Different types of diabetic tractional retinal detachment (TRD). (a) Magnified wide-field Optos color image of exuberant fibrovascular tissue proliferation over the posterior pole and along the vascular arcades that contract and detach the macula. (b) Another image of diabetic TRD with macular involvement. The TRD extends to the nasal equator of the retina with vertical subretinal demarcation lines. This modified multipanel figure was obtained from a previous publication [27] and was generated under the Creative Commons Attribution 4.0 International License (https://creativecommons.org/licenses/by/4.0/).

Figure 3.

Different types of diabetic tractional retinal detachment TRD. (a) Macula-off combined with tractional and rhegmatogenous retinal detachment (RRD). 57082. (b) Combined tractional and RRD in a patient with proliferative diabetic retinopathy (PDR). 57083. The images in this modified figure were originally published on the Retina Image Bank® website. Manish Nagpal, MD. Photographer Gayathri Mohan. Retina Foundation. Retina Image Bank. Jun 10, 2020. © The American Society of Retina Specialists.

1.1.1 Image binarization for CVI quantification and the CFA measurement protocol

CVI values are calculated by quantifying the luminal area (LA) and total choroidal area (TCA) from SD-OCT images of the macula using ImageJ analysis software (version 1.53; USA, public domain). High-definition (HD) 9-mm OCT-B images are converted into an 8-bit format (Figure 4a) and adjusted using the autothreshold technique (Niblack autothreshold). Using the polygon tool, the area of the subfoveal choroid is then digitally selected on the map of the total choroidal area (TCA) 750 μm nasally and 750 μm temporally from the foveal center in the direction of the horizontal plane and vertically from the RPE-Bruch’s membrane region to the inner scleral border in the direction of the vertical plane (demarcated by the dotted red line in Figure 4b). The stromal vascular tissue area is determined by the number of white pixels, and the luminal area (LA) at the choroid is determined by applying the threshold tool and quantifying the number of dark pixels once the binarized image is obtained (Figure 4c). The dark-to-light pixel ratio is then expressed as a percentage and defined as the CVI, as previously described by Agrawal et al. [29, 30]. The protocol study method for binarization was validated in a previous study [28]. The CFA is obtained by automated binarization and segmentation of the CSP slabs using the RTVue XR OCT Avanti with AngioVue Software (OptoVue Inc., Fremont, CA, USA) and automatically calculated from a 3.142 mm² evaluation area (Figure 4d).

Figure 4.

A figure depicting the method used for quantifying the CVI in healthy eyes. (a) Binarized image designed to depict the intraretinal structure and choroidal layers in greater detail in a healthy emmetropic eye. (b) Binarizing processing in a 9-mm HD image. (c) Magnified image within the yellow square showing binarized processing of the subfoveal choroidal stroma and luminal vascular visualization of the subfoveal choroidal vessels to obtain a choroidal vascularity index (CVI) of 62.8%. The selected subfoveal area of choroidal flow is clearly delineated by the red-white dotted line. (d) Binarizing processing of the normal choriocapillaris flow area (CFA). This modified multipanel figure was obtained from a previous publication [28] and was generated under the Creative Commons Attribution 4.0 International License (https://creativecommons.org/licenses/by/4.0/).

1.2 Surgical approaches

A technique using a standard 25-gauge, self-sealed, three-port, control-perfused PPV is recommended by the authors in this section. Local sedation is recommended when using this approach. In addition to central vitrectomy, triamcinolone acetonide (Kenalog 40 mg/mL; Bristol-Myers, New York, NY, USA)-assisted removal of the cortical vitreous from the surface of the retina should be performed, with a focus on the detection of preretinal fibrovascular tissue bands. Detaching or peeling condensed hyaloids in cases of complex vitreoretinal interfaces requires a combination of delamination with segmentation maneuvers and with visco-delamination techniques for epiretinal fibrovascular tissue to avoid iatrogenic retinal breaks. Any foci of bleeding from the fibrovascular tissue should be endodiathermized throughout the procedure. Following previously described steps, the fibrovascular membranes are dissected using membrane stripping techniques, and the residual epiretinal tissue is peeled to release macular traction. The focus of the surgery should be on tangential traction, and surgical macular evaluation and revision should be performed using 0.15% trypan blue ophthalmic solution (Membrane Blue; Dutch Ophthalmic, USA) to stain the cortical vitreous remnants or epiretinal membrane (ERM) proliferation. A 0.25 mg/mL (0.025%) diluted isomolar solution (pH 7.4) of Brilliant Blue G (BBG) dye is recommended for ILM staining when ILM traction is suspected. If there are any rhegmatogenous lesions, the lesion should be diathermally marked, macular reattachment should be achieved, and argon laser-assisted scatter panretinal photocoagulation (PRP) should be performed. In retinal zones with shallow RD due to the presence of subretinal fluid without rhegmatogenous lesions, PRP should be avoided or cautiously applied because the laser may induce small retinal breaks (Figures 5 and 6). A selected gas tamponade, such as a long-acting nonexpandable perfluoropropane (C₃F₈) gas mixture at 15%, sulfur hexafluoride (SF₆), or silicone oil (SO) lighter than water 1000 or 5000 Cks (centistokes), should be administered at the end of the surgical procedure based on the surgeon’s preference.

Figure 5.

Surgical approach and postoperative structural and perfusion findings. (a) Wide-field Optos color fundus image of an eye with long-term type 2 diabetes mellitus and abundant proliferative fibrovascular preretinal tissue that detached the retina involving the macula. (b) Surgical image showing a severe TRD; the macula is off, and fibrovascular tissue emerges from the optic nerve, detaching the retina and temporal vascular arcades. The detached macula is released when trypan blue dye is used to stain vitreous or fibrovascular remnants and facilitate membrane stripping. This surgical image also depicts a distorted and stained macular surface with evidence of dragged macular tissue due to proliferative fibrous tissue remnants over the optic nerve and along the inferior temporal arcade. (c) Surgical magnified image showing clear reattachment of the macula after air-fluid exchange. (d) Corresponding final long-term fundus photo showing an attached retina and extensive panretinal photocoagulation (PRP) scars. (e) Long-term postoperative superficial capillary plexus (SCP) showing end-vessel abnormalities and an enlarged foveal avascular zone (FAZ) with multiple capillary dropouts. (f) Corresponding deep capillary plexus (DCP) with ischemic changes and vascular deficiencies. (g) Long-term postoperative 9-mm HD horizontal B-scan depicting an abnormal foveal profile, disorganization of the retinal inner layers (DRIL), and abnormal outer retina layer markers. (h) The image within the red-white dotted line clearly delineates a low CVI of 54.4%. (i) Abnormal CFA of 1.880 mm². This modified multipanel figure was obtained from a previous publication [27] and was generated under the Creative Commons Attribution 4.0 International License (https://creativecommons.org/licenses/by/4.0/).

Figure 6.

Surgical approach and postoperative structural and perfusion findings. (a) Surgical view of a very complex diabetic TRD involving the macula in a hammock configuration. (b) The macula was traction-released, and passive blood aspiration and retinal exploration were mandatory. (c) Image of the final surgical step depicting air-fluid exchange with a completely reattached retina and extensive PRP. (d) and (e) Postoperative en-face OCT-A images depicting irregular filling at the level of the SCP and DCP insufficiently filled due to low vessel density (VD) and capillary dropout with an irregular and enlarged perifoveal capillary network (PCN). (f) Long-term postoperative image showing diffuse retinal thinning on the temporal macular side; the macular contour is indistinguishable, indicating severe nasal macular cystic edema, DRIL, and external limiting membrane (ELM) subfoveal discontinuities. (g) Postoperative binarizing processing of this image showed an abnormal CVI of 57.6% within the red-white dotted line. (h) Close to normal CFA of 2.104 mm². This modified multipanel figure was obtained from a previous publication [27] and was generated under the Creative Commons Attribution 4.0 International License (https://creativecommons.org/licenses/by/4.0/).

1.3 Postoperative structural and perfusion outcomes in patients with TRD

Postoperative SD-OCT findings (Figures 5–7) revealed foveal contour profile abnormalities, retinal nerve fiber thickness (RNFT) alterations, disorganization of the retinal inner layers (DRIL), diffuse retinal layer thinning with central subfield thickness (CSFT) abnormalities, ellipsoid zone (EZ) disruption, and external limiting membrane (ELM) line abnormalities in a subset of surgical eyes. Based on the OCT-A perfusion data, both the SCP and DCP exhibited extensive perfusion abnormalities, and subfoveal CVI and CFA analyses revealed subnormal values in all the examined patients, with an enlarged and irregular perifoveal capillary network (PCN) with lower-than-normal VD values in the analyzed capillary plexuses (Figures 5–7). A summary of the postoperative structural and perfusion outcomes is shown in Table 1.

Figure 7.

Surgical approach and postoperative structural and perfusion findings. (a) Surgical magnified image depicting a combined tractional and RRD. (b) Transoperative image after intraocular silicone oil (SO) injection at the end of the surgical procedure. (c) Postoperative cross-sectional horizontal SD-OCT B-scan image showing an irregular macular profile with some mild superficial wrinkling of the temporal region along with extrafoveal cystic chronic residual edema with deep hyperreflexive dots, mainly over the nasal side of the macula. ELM line discontinuities were observed, and the subfoveal ellipsoid zone (EZ) was disrupted. (d) Corresponding long-term postoperative binarized image where the red-white dotted line clearly delineates an abnormal CVI of 56.2%. (e) Binarizing of the choriocapillaris to obtain an abnormal CFA of 1.825 mm². This modified multipanel figure was obtained from a previous publication [27] by Quiroz-Reyes et al. and is used under the Creative Commons Attribution 4.0 International License (https://creativecommons.org/licenses/by/4.0/).

SD-OCT (n = 30) and OCT-A (n = 14) findings	Mean	SD
Preoperative evolution of TRD	11.6 weeks	±2.3 weeks
Mean postoperative TRD resolution	3.6 weeks	±1.7 weeks
Mean postoperative follow-up	11.4 weeks	±5.7 weeks
SD-OCT marker	n	Percentage
Postoperative CSFT abnormalities	9	30%
Presence of DRIL	13	43.3%
EZ abnormalities	23	76.6%
ELM line abnormalities	21	70.1%
Plexus
SCP and DCP abnormalities	14	87.5%
PCN abnormalities	12	75%
Decreased VD	10	62.5%
Decreased CVI	14	100%
Decreased CFA	14	100%

Table 1.

Preoperative and postoperative TRD evolution. Postoperative structural and perfusion outcomes.

CFA, choriocapillaris flow area; CSFT, central subfield thickness; CVI, choroidal vascularity index; Deep capillary plexus; DRIL, disorganization of the retinal inner layers; EZ, ellipsoid zone; ELM, external limiting membrane; OCT-A, optical coherence tomography angiography; PCN, perifoveal capillary network; SD, standard deviation; SD-OCT, spectral-domain optical coherence tomography; SCP, superficial capillary plexus; TRD, tractional retinal detachment; VD, vessel density These modified data were published previously [27] by Quiroz-Reyes et al. and is used under the Creative Commons Attribution 4.0 International License (https://creativecommons.org/licenses/by/4.0/).

1.4 Postoperative functional performance in patients with TRD

The final TRD resolution and final visual acuity (VA) correlated with the OCT-A outcomes are shown in Figures 8 and 9, respectively. In Figure 10, comparisons between the preoperative, 3-month, and final postoperative BCVA values are shown. In Figure 11, the BCVA repeated measures that estimated marginal mean values are explained, and the figure shows the mean points for each stage (pre, 3 months, and final) with a 95% confidence interval. In Figure 12, comparisons of preoperative, 3-month, and final postoperative BCVA between eyes that underwent more than one surgery and eyes that underwent only one surgery are shown. In Figure 13, a total of 44 eyes were plotted, including 14 postoperative TRD patients and 30 age-matched healthy control eyes. The Mann-Whitney nonparametric test did not reveal significant differences in age between the control and diabetes cohorts (P = 0.125). The final BCVA, as represented by the logMAR scale, was significantly greater in the diabetes group (P < 0.0001), with a median logMAR of 0.000 in the control group and 0.480 in this subset of the diabetic group. As shown in Figure 14, the CVI values calculated from OCT measurements were significantly lower in the diabetic group (P = 0.0011), with a median CVI of 57.95% in the control group and 44.41% in the diabetic group. A moderately positive correlation was found between the CVI and BCVA in the control cohort (r = 0.535) (P = 0.015). The correlation between the CVI and BCVA in the diabetes cohort was close to zero (r = −0.005) and was not significant (P = 0.986). The BCVA and CVI exhibited broader distributions in the diabetes cohort than in the control cohort, indicating greater variations in visual function and retinal morphology in diseased eyes. Figures 8–14 summarize the structural, functional, and perfusional outcomes previously published by the authors [27].

Figure 8.

Final TRD resolution correlated with OCT-A findings. This modified figure was published previously [27] by Quiroz-Reyes et al. and is used under the Creative Commons Attribution 4.0 International License (https://creativecommons.org/licenses/by/4.0/).

Figure 9.

Final postoperative BCVA correlated with OCT-A findings. This modified figure was published previously [27] by Quiroz-Reyes et al. and is used under the Creative Commons Attribution 4.0 International License (https://creativecommons.org/licenses/by/4.0/).

Figure 10.

Comparison of the preoperative, 3-month, and final postoperative BCVA values. This figure was published previously [27] by Quiroz-Reyes et al. and is used under the Creative Commons Attribution 4.0 International License (https://creativecommons.org/licenses/by/4.0/).

Figure 11.

BCVA repeated measures estimated marginal means. The figure shows the mean points for each stage (preoperative, 3-month, and final) with the 95% confidence interval. This figure was published previously [27] by Quiroz-Reyes et al. and is used under the Creative Commons Attribution 4.0 International License (https://creativecommons.org/licenses/by/4.0/).

Figure 12.

Comparison of the preoperative, 3-month, and final postoperative BCVA values between eyes that underwent more than one surgery (vitrectomy revision) and eyes that underwent only one surgery (primary vitrectomy). This figure was published previously [27] by Quiroz-Reyes et al. and is used under the Creative Commons Attribution 4.0 International License (https://creativecommons.org/licenses/by/4.0/).

Figure 13.

Final BCVA across study cohorts. The BCVA values collected on the Snellen scale were converted to logMAR units and evaluated across study cohorts. The control eyes had significantly lower BCVA values in logMAR units than did the diabetic eyes. The P value is indicated by * (**** = P ≤ 0.0001). The bars represent the means and standard deviations; the data for each patient are plotted as dots. BCVA, best-corrected visual acuity.

Figure 14.

Distribution of the CVI and BCVA. Individual patient data were plotted with points colored according to the study cohort. Linear regression lines are shown as solid-colored lines, with 95% intervals shown as dotted lines.

1.4.1 CVI and CFA measurements

Compared to those of the control eyes, the surgically treated TRD eyes had significantly smaller LA, TCA, CVI, and CFA values (all P < 0.0001) (Table 2). No significant differences were found in LA or TCA values in the PPV group (both p > 0.05). However, the PPV group had significantly lower CVI (P = 0.0125) and CFA (P = 0.0003) values than those in the control group.

Parameters	Vitrectomy group (mean values)	Control group (mean values)
LA (mm^2)	0.189	0.198
TCA (mm^2)	0.312	0.342
CVI (%)	44.41	57.95
CFA (mm^2)	1.38	2.28

Table 2.

Mean choroidal measurements across treated eyes (n = 14).

CFA, choriocapillaris flow area; CVI, choroidal vascularity index; LA, luminal area; TCA, total choroidal area.

As the state of the choroid and retinal postoperative perfusion may influence visual function, a correlation analysis between the choroidal parameters (CVI and CFA) and BCVA was performed. Positive correlations were observed among BCVA, the CVI, and the CFA (Table 3). Because BCVA was measured in logMAR units, these findings suggested a positive correlation between the two choroidal perfusion parameters and visual function; that is, a lower CVI and CFA correlated with a worse BCVA (P = 0.0003 and P = 0.0037, respectively).

Choroidal parameters	Control eyes (n:30)	TRD eyes (n:14)
CVI	−0.124	−0.117
CFA	−0.211	−0.355

Table 3.

Correlation coefficients between BCVA and choroidal parameters.

CFA, choriocapillaris flow area; CVI, choroidal vascularity index; TRD, traction retinal detachment.

1.5 Representative surgical cases

Based on the surgical experience of the authors, representative and complex surgical cases of diabetic macula-off TRD in which vitreous surgery was performed at different stages of TRD, with the estimation of the corresponding long-term postoperative and enhanced binarized OCT imaging CVI and CFA calculations, are shown in multipanel Figures 15–17.

Figure 15.

Surgical Patient 1. A 42-year-old male patient with 18 years of type 2 diabetes had irregular metabolic control. His last glycosylated hemoglobin level was 7.7. The patient had PDR and panretinal photocoagulation (PRP) in both eyes. The patient presented with progressive visual acuity (VA) loss in his right eye after 12 weeks of evolution. Ophthalmological examination revealed a pseudophakic RE with cloudy media. The TOA was 16 mmHg, with no manifestations of anterior segment neovascularization, signs of moderate vitreous bleeding, or evidence of a fibrovascular tissue stalk that detached the posterior pole, mainly at the level of the macula and inferotemporal vascular arcade (a). B-mode ultrasonography confirmed a diagnosis of TRD with macular involvement. The patient underwent an uneventful vitrectomy and endophotocoagulation-like PRP, during which macular reapplication was achieved. Long-term postoperative quantitative VD evaluation of the SCP was considered to be lower than normal, as shown by white numbers on the different subfoveal regions on the ETDRS-like grid (b). Postoperative final evaluation at 14 months showed an abnormal FAZ area on the DCP and SCP slabs (c) and (d) (yellow arrows). DRIL (orange arrows) and postoperative cystic macular edema were detected via B-scan OCT, and white arrows pinpoint outer retina layer markers (e). The binarized image showed a low CVI of 56.2% (f). (g) The CFA was abnormal with a value of 1.887 mm². The final BCVA was 20/80 (0.60 logMAR units). This modified multipanel figure was obtained from the International Journal of Ophthalmology & Visual Science. 2021;6(4):187–198. DOI: 10.11648/j.ijovs.20210604.12 and is used under the Creative Commons Attribution 4.0 International License (https://creativecommons.org/licenses/by/4.0/).

Figure 16.

Surgical Patient 2. A 59-year-old female patient with chronic complications of PDR presented evidence of vitreous hemorrhage on several occasions (a). She was unable to satisfactorily undergo laser PRP, and because of the presence of fibroglial tissue, the patient was not managed preoperatively with intravitreal bevacizumab. She had experienced a significant decrease in her VA for 3 weeks, and ultrasound examination revealed complex TRD of the posterior pole, complex vitreoretinal interface, and macular involvement. She underwent an uneventful vitrectomy, as depicted in the surgical image (b). Fifteen months after vitrectomy, the patient was evaluated using OCT-A. An angiogram of the papilla and macula showed restored perfusion close to normal but still with extensive yuxtafoveal capillary dropouts, an irregular and enlarged FAZ (yellow circle), and interruptions of the capillary margin at the SCP level (c) (white arrows). Postoperative OCT-A evaluation revealed a quantitative VD pattern that was considered lower than normal, which indicated important capillary dropout in the vicinity of the FAZ area. The pericapillary net (PCN) was abnormal, with a lower-than-normal VD. Long-term postoperative horizontal B-scan SD-OCT showed irregular foveal contours and diffuse irregular macular thickening due to the presence of macular edema with a large intraretinal subfoveal cyst (white asterisk), evidence of DRIL, and abnormal outer retinal biomarkers on the SD-OCT scan (d). The binarizing process showed a very low CVI of 53.7%, which was clearly defined in the red-white dotted selected area (e). (f) An abnormal CFA of 1.783 mm² is shown. The final BCVA was 20/120 (0.78 logMAR units). This modified multipanel figure was obtained from the International Journal of Ophthalmology & Visual Science. 2021;6(4):187–198. DOI: 10.11648/j.ijovs.20210604.12 and is used under the Creative Commons Attribution 4.0 International License (https://creativecommons.org/licenses/by/4.0/).

Figure 17.

Surgical Patient 3. A 57-year-old male patient with diabetes mellitus for 22 years presented with progressively decreasing vision in the left eye with 14 weeks of disease evolution and no history of laser PRP or antiangiogenic therapy. Ophthalmoscopic examination and B-mode ultrasonography revealed severe and extensive macula-off TRD, as shown in panel (a). The patient underwent uneventful PPV and phacoemulsification with surgical reattachment of the macula without complications, as shown in the corresponding surgical image (b). At 19 months postsurgery, perfusion evaluation was performed using OCT-A, which revealed an SCP with multiple vascular deficiencies in the VD evaluation, an abnormally enlarged FAZ area, and loss of shape of the corresponding PCN, as shown in the binarizing image with the ETDRS grid quantified with white numbers in its subregions (c). OCT-A evaluation of the DCP showed extensive and diffuse capillary dropouts (white arrows) (d). Long-term postoperative OCT examination revealed severe diffuse irregular thickening of the macula, intraretinal microcysts in the middle layers of the retina, and the presence of extrafoveal DRIL; biomarkers of the outer retinal layers were not recognizable, and atrophic subfoveal tissue was detected (e). (f) Binarized image processing showed a very low CVI of 46.8%. (g) An abnormal CFA of 1.775 mm² was quantified. The final BCVA of this eye was 20/200 (1.00 logMAR units). This modified multipanel figure was obtained from a previous publication [27] by Quiroz-Reyes et al. and is used under the Creative Commons Attribution 4.0 International License (https://creativecommons.org/licenses/by/4.0/).

2. Role of the postoperative CVI in assessing choroidal vascular changes in TRD patients

The CVI has proven to be informative in determining the condition of choroidal perfusion across various treatment scenarios [31, 32]. Quantitative metrics, such as the CVI and CFA, have been used to diagnose, categorize, and guide treatments for retinal diseases [33]. Hayreh demonstrated that the submacular choroidal supply is affected by chronic ischemic conditions such as age-related macular degeneration (AMD) and PDR because of the multiple zones of altered perfusion at the level of the short posterior ciliary arteries. A decrease in the CVI at baseline is indicative of choroidal ischemia in patients with DR. Moreover, the CVI is a useful postoperative indicator for monitoring disease incidence, surgical influence, and treatment response [34]. Employing noninvasive techniques, such as CVI estimation [35] and OCT-A perfusion indices [36], to monitor disease and treatment outcomes could significantly enhance the management of different conditions, such as postoperative TRD.

3. Postoperative CVI and CFA alterations in TRD patients

Studies have determined the relationship between CVI and CFA in the development and progression of various diseases, including genetic, retinal, and vascular conditions [34, 37, 38]. The CVI can be used as an indicator of retinal and choroidal vascular disease. Since retinal diseases involve multiple factors, such as inflammation, edema, and leakage, the CVI and CFA have been suggested as potential markers because they reflect vascular perfusion changes of paramount and critical importance that affect the postoperative integrity and recovery of the vascular network within the highly blood-supplied structure known as the choroid [39, 40, 41].

Our study showed that postoperative OCT-A perfusion values were lower in surgical patients with a longer preoperative duration of TRD and in those who underwent more than one surgical procedure (see Figure 12). Surgical intervention with TRD reapplication significantly improved visual acuity in all patients (see Figure 13). Nevertheless, postoperative perfusion outcomes revealed that the CVI was significantly lower in the surgical group than in the control group (see Figure 14). The CVI and CFA may indicate disease progression in DR patients, and a lower CVI in postoperative eyes could be linked to the absence of recovered choroidal ischemic vascular defects caused by DR or tractional flow deprivation, the impact of surgical procedures or the influence and effects of antiangiogenic therapies. In conclusion, choroidal and retinal perfusion evaluations highlight the potential of the CVI and CFA as biomarkers for RD and their significance in differentiating healthy eyes from eyes with surgically resolved TRD and poor final BCVA. However, further research is needed to explore and better quantify CVI and CFA values in diabetic patients without DR, those with high-risk PDR, and those with resolved TRD to better understand their implications as prognostic markers for DR.

4. Literature review and discussion

Small-gauge instrumentation, control-perfusion techniques, bimanual PPV, wide-field viewing, improved fluidics, longitudinal pre- and postcutting endodiathermy, and precise tissue control techniques have all improved the anatomical outcomes of TRD surgical approaches. However, functional outcomes continue to be disappointing. Other injectable interventions using medical intravitreal antiangiogenic agents have been shown to alleviate postoperative residual DME, postoperative bleeding, and ischemic tissue complications due to severe BRB damage in TRD-associated PDR patients. The analysis of the results in this chapter was based on the surgical experience published by the authors [27] and its correlation with the literature review up to the publication date of this chapter.

It has been demonstrated that visual improvement is better in patients with DR who are receiving anti-VEGF therapy than in those who have received laser treatment [24]. The different options for clearing dense VH in patients who cannot undergo laser treatment include waiting and monitoring for the hemorrhage to reabsorb, performing additional laser treatment, administering injectable intervention with anti-VEGF in the absence of TRD, or performing vitrectomy. Recurrent bleeding with membrane formation and TRD occurrence were risk factors associated with the former method. The Writing Committee for Diabetic Retinopathy for the Clinical Research Network compared the clinical efficacy of intravitreal ranibizumab with that of PRP in clearing VH [9]. They concluded that patients with PDR and high-risk characteristics of severe vision loss treated with injectable intervention using sustainable intravitreal ranibizumab had better visual outcomes at 1 year than those treated with standard PRP care. Protocol S, based on injectable intervention using an anti-VEGF algorithm for PDR, assessed neovascularization status for 2 years in patients with high-risk PDR without prior PRP [42]. These authors concluded that the Diabetic Retinopathy Clinical Research Network (DRCR.net) protocol provided excellent outcomes for 2 years in patients who received intravitreal antiangiogenic treatment for PDR following the protocol guidelines [42].

Decreased perifoveal capillary blood flow caused by microangiopathy results in chronic ischemia of retinal tissue. According to a comprehensive assessment of the long-term postoperative macular perfusion indices, in functional eyes with a BCVA better than 20/200 (1.0 logMAR unit equivalent), retinal capillary nonperfusion was observed to be associated with photoreceptor damage according to the SD-OCT biomarkers, with evident alterations in EZ continuity and the presence of subfoveal DRIL. Macular ischemia is reflected in profound capillary deficiencies in the DCP, low CVI, and CFA VD deficiencies that cause choriocapillaris and outer retinal changes, and deleterious perfusion defects in central photoreceptors. When the central retina remains detached from the RPE, the risk of outer retinal layer ischemia increases because of capillary dropout at the DCP and choriocapillaris reflected in the low VD in the CFA slabs. Postoperative TRD eyes showed outer retinal layer abnormalities that colocalized with areas of enlargement in the FAZ area with a very irregular and ischemic PCN, and diffuse or localized macular capillary nonperfusion areas were detected with VDs lower than the mean VD in the SCP and DCP [27]. Therefore, photoreceptor damage with EZ abnormalities, DRIL, and macular ischemia are the most common factors contributing to limited improvement in vision after successful diabetic TRD surgical reattachment [27].

The current literature does not provide much information on the potential of quantitative retinal and choroidal OCT-A markers to predict visual outcomes in patients undergoing vitrectomy for TRD. There are many technical difficulties in quantitatively assessing pre- and early postoperative retinal and choroidal perfusion in patients with TRD. Cloudy media due to hemorrhage, the presence of preretinal fibrovascular complexes, and media opacities with complexities in the vitreoretinal interface are not technically allowed for preoperative evaluation via OCT-A perfusion imaging. We routinely evaluated the postoperative correlation of six OCT-A perfusion indices, the FAZ area, the FAZ shape, the VD in the SCP and DCP, the CVI, and the CFA, with the final BCVA and structural biomarkers such as the CSFT, EZ, and ELM line characteristics. The presence of subfoveal DRIL also determines how these SD-OCT and OCT-A perfusion biomarkers change longitudinally over time and how they are correlated with BCVA.

Unlike FFA, OCT-A evaluations facilitate in-depth analyses of vascular alterations in individuals with DR [17, 18, 19, 20], acting as a key technological tool for evaluating perfusion in the different vascular layers in the choroid [43]. Patients with macular ischemia exhibit structural changes in both retinal capillary plexuses. In these patients, postoperative assessment of perfusional indices using OCT-A may be a useful predictor of the final VA [44, 45, 46].

Recent reports [25, 44, 45, 46, 47] have shown that a poorer VD in both the superficial and deep capillary networks is correlated with DRIL appearance, EZ abnormalities, and CSFT with a poorer final BCVA. Freiberg et al. [45] quantified FAZ area dimensions and symmetry in patients with DR compared to normal controls using OCT-A. These authors found that OCT-A imaging could pinpoint the abnormal integrity of the capillary network surrounding the FAZ area, which corresponds to the PCN [44, 45]. Other researchers have reported that outer parafoveal VD in the superficial layer may predict improvement in VA after injectable intervention with ranibizumab injection. VD perfusion indices may be useful for evaluating the perfusion status of the macular area after surgery and predicting final vision. Several authors have previously reported that an increased FAZ area or contour irregularities in the PCN contour, an increased nonperfusion area, and decreased VD perfusion indices at the central macula in OCT-A may predict worsening DR [46, 47, 48]. Here, we report the potential of evaluating both retinal capillary plexus-quantified values and choroidal perfusion indices as indicators of complementary treatment for improving postoperative macular ischemia.

Photoreceptor disruption due to macular ischemia contributes to poor visual outcomes. The impaired blood flow associated with subfoveal DRIL generated by decreased blood perfusion combined with cell apoptosis due to retinal ischemia, which is mainly associated with lower-than-normal VD values in the SCP, CVI, and CFA, contributes to poor postoperative visual results. A large FAZ area associated with poor visual acuity (VA) recovery is a common outcome [27, 28, 49].

The analyzed data indicate that DRIL resolves over time and that DRIL resolution is a good predictor of visual improvement [14]. Spontaneous or pharmacologic (antiangiogenic) postoperative improvement in the appearance of subfoveal DRIL represents an anatomical marker of more normal morphology and potentially better visual recovery [27]. Researchers have not determined whether reversibility potentially decreases with increasing DRIL duration or whether this change can be induced. However, these findings have important implications for the timing of TRD [49].

We presumed that DRIL may be a biomarker of tissue damage. Therefore, further histological assessments are needed to determine the mechanisms by which the ischemic GCL-IPL complex, RNFL thickness, and neurodegenerative damage affect VA. However, the functional outcomes in a case series [27] and literature review [20, 21, 26, 45, 46, 47, 48, 49, 50] suggested that increased subfoveal DRIL length, indicating less viable remaining tissue, is associated with worse visual outcomes.

Reports on PDR have shown that disruption of the outer retina, mainly of the EZ and ELM line markers, is a significant predictor of visual prognosis [46, 48, 51]. We found that subfoveal DRIL, ELM line integrity, and EZ characteristics correlated well with postoperative evolution and predicted final visual outcomes [27].

Moreover, certain perfusion indices, such as VD at different subregions of the macula according to the ETDRS grid, and structural biomarkers, such as DRIL, are important for postoperative monitoring once the retina has been successfully reattached and the eye is tamponade-free, which precludes detailed structural and perfusion evaluation. An important issue is that when the retina becomes detached along the macula, it must be managed in a timely manner to ensure that perfusion, mainly to the outer layers, resumes. Postoperative multimodal evaluation is mandatory for better evaluation and improvement of functional outcomes [6, 27] using injectable interventions or sustained anti-VEGF agents [21, 52].

The choroid is the vascularized layer of tissue that provides the most important blood supply to the RPE and outer retinal layers. A restricted blood supply to the choroid adversely affects the outer retinal layers, potentially disrupting tissue homeostasis and causing diabetic choroidopathy and, consequently, different degrees of ischemic chorioretinal tissue damage [43].

The CVI is a recently introduced quantification parameter that represents the vascular status of the choroid based on OCT images [29]. The CVI is a more precise and robust measurement of choroidal vascular characteristics than previous measurements, such as choroidal vessel diameter and choroidal thickness (CT), both of which have important limitations [53]. CVI abnormalities can be used as indicators of choroidal vasculature disease; however, the CVI cannot capture choriocapillaris flow, which is also an important characteristic of the inner choroidal vasculature and can be quantified using OCT-A binarized imaging [54].

The CVI reveals the status of the choroidal vasculature in a variety of diseases and allows treatment monitoring [31, 52, 55, 56, 57, 58]. In vivo, noninvasive imaging of the choroid is relevant for both disease diagnosis and treatment monitoring but can be challenging because this structure is located beneath multiple layers of tissue. Recent advancements in enhanced depth OCT and rapidly evolving OCT-A technology have proven to be useful tools for evaluating retinal and choroidal circulation without the need for dye injection [43, 59].

Improved visualization of the vascular choroid has resulted in quantitative evaluation of the choroidal vasculature. The CVI was first described by Agrawal et al. [29] in 2016 as a relatively new and reliable biomarker of the choroidal vasculature. This imaging technology was developed for a more accurate assessment of choroidal vascular layers [53]. Since its introduction, the CVI has gained popularity; additionally, the CVI has been applied in diabetic choroidopathy, uveitis, central serous chorioretinopathy, retinal vein occlusion, myopic traction maculopathy (MTM), and many other conditions [29, 34, 56, 60, 61, 62, 63]. Recent reports have shown that CVI is a promising biomarker for the diagnosis and follow-up evaluation of retinal diseases, among which diabetic retinopathy and postoperative perfusion evaluation are important for patients with severe resolved TRD [64]. At the molecular level, recent reports have suggested that retinal cell apoptosis due to DR is related to neurofilament deposition in the RNFL accompanied by a high level of extracellular glutamate, which results in Müller cell damage and is likely generated by alterations in retrograde flux, an increase in neurotoxic debris [65], and uncontrolled hyperglycemia causing GCC-IPL disturbances and RNFL thinning [66].

In recent research, some authors have concluded that DME in patients with a higher CVI, higher EZ reflectivity, the presence of SRF, and the absence of DRIL at baseline are more likely to gain visual acuity after anti-VEGF treatment. The CVI may serve as a novel biomarker for visual response to anti-VEGF treatment in DME patients [67].

Quiroz-Reyes et al. have shown how critical CVI and CFA are used to evaluate perfusion in different pathologies, such as GRT-related RRD postoperative evaluation, and surgically resolved MTM eyes, which generally exhibit larger superficial postoperative FAZs, smaller CFAs, lower VDs, more structural macular defects, and thinner CSFTs [35, 68]. Better functional, structural, and perfusion index outcomes were observed in highly myopic eyes that underwent early surgery. However, the postoperative perfusion changes in patients with TRD are somewhat disappointing and do not correlate well with the functional outcomes, as shown in this chapter.

Currently, there are few and conflicting data on the predictive indicators for diabetic retinopathy available. The choroidal thickness (CT) is a characteristic that is unstable and dependent on numerous circumstances. Furthermore, the results of earlier research on choroidal thickness were inconsistent. Regardless of whether they have DR or not, diabetic patients often have a lower CVI than healthy controls. Additionally, DR patients had a lower CVI than non-DR individuals. The CVI is a sensitive and early biomarker for the onset of DR [69].

Jing et al. [70] have described a noteworthy correlation between declining BCVA and CVI in individuals with DR. A reduction in CVI was the factor most strongly associated with visual impairment, according to multivariate regression analysis. This study revealed a correlation between the CVI and BCVA. Thus, we propose that tracking CVI changes could offer a new target for interventions meant to stop the worsening of visual impairment in DR patients as well as a reliable measure for predicting future loss of visual acuity.

This chapter was subject to some constraints that are typical of retrospective analyses with a small sample size. We observed that this was a result of implementing rigorous exclusion criteria, which were specifically designed to restrict our trial to patients who had effectively undergone surgical treatment for diabetic TRD. Nevertheless, it effectively utilized a meticulously controlled group design and employed SD-OCT and OCT-A for pre- and postoperative evaluations. Additionally, we examined the visual outcomes and their correlation with perfusion markers. We anticipate that the scientific retina community will find our paper valuable and timely, considering the scarcity of published literature on perfuse macular assessment in this situation.

The merits of this study lie in its comprehensive data collection on the occurrence of diseases in a substantial number of colleague eyes, as well as its ability to detect postoperative perfusional alterations to be improved with some kind of anti-VEGF injectable interventions. The study revealed that the contralateral eyes exhibited notable changes in perfusion that were strongly correlated with BCVA values. Crucially, all eyes underwent vitrectomy surgery promptly after the development of TRD, resulting in documented detrimental functional and perfusional outcomes.

5. Future prospects

The study of DR and its connection to choroidal perfusion through the CVI and CFA has provided vital knowledge about the fundamental mechanisms and clinical features of this disorder. However, there are other areas that need to be investigated more in order to improve our understanding and improve patient care.

1. Longitudinal investigations involve conducting research over an extended period to examine how DR progresses and its connection with changes in the CVI and CFA values. These studies can offer vital insights into the natural development of the disease and help uncover the possible factors that contribute to its advancement.

2. Exploring the effects of genetic and environmental factors on the development and progression of DR and surgical forms of PDR and their connection to the CVI and CFA values can offer valuable information about an individual’s vulnerability and potential preventive actions.

3. Utilizing developments in imaging technology, such as OCT-A, could provide new opportunities for researching choroidal perfusion and its impact on enhancing the postoperative functional outcomes in PDR. By integrating several imaging techniques, it is possible to obtain extensive information regarding alterations in the choroidal blood vessels.

4. Examining the impact of various treatment modalities, such as medications, injections into the eye, and surgical procedures, on the advancement of CVI and DR could result in the creation of more successful therapeutic approaches.

5. Collaborative large-scale investigations involve numerous sites and patient cohorts. These efforts can result in larger studies, increase the statistical significance of the findings, and enhance the generalizability of the conclusions.

6. Developing animal models that replicate the pathophysiology of DR and allow for the evaluation of choroidal perfusion can offer further understanding of the causes of the disease and potential targets for treatment.

7. When comparing the changes in choroidal perfusion in DR with other retinal disorders such as AMD or MTM, it might be useful in identifying distinct patterns of choroidal perfusion that are specific to DR.

8. Examining the functional correlations between the modifications in CVI and CFA values and visual acuity as well as other functional outcomes after surgery should assist doctors in predicting results and determining treatment decisions.

6. Conclusions

The choroidal vasculature provides the outer retina with its circulatory supply. Even before clinical symptoms appear, the choroidal vascularity index (CVI) assessment may be a useful early indicator of DR. There is a correlation between the degree of retinopathy and CVI in diabetic patients. As the severity of the retinopathy grows, the CT and CVI decrease. The CVI is a more accurate measure of disease development. In the eyes of patients with DR, DME, and those who are successfully reattached, the CVI and CFA biomarkers dramatically changed. In addition to alterations in the retinal vasculature, changes in the choroid also contribute to changes in DR.

Significant associations were observed between postoperative changes in the structure of the choroid and visual impairment in diabetic patients. The CVI can serve as a quantitative metric for assessing choroidal damage in patients with DR. The CVI can be used as a reliable quantitative biomarker to track the progression of DR or for monitoring postoperative eyes. Healthcare professionals should carefully observe alterations in the blood vessels of the choroid in patients with DR and acknowledge that preserving the CVI may be a crucial treatment objective in order to impede the advancement of DR. Patients with DR experience choroidal vascular changes, including a decrease in the CVI. The CVI decreased as the length of DM increased and showed a correlation with visual impairment. This finding suggests that the CVI could serve as a reliable imaging biomarker for monitoring the progression of DR.

Our results suggest severe and remarkable postoperative detrimental structural and perfusion changes in resolved TRD, which correlate with disappointing long-term visual results and may be aggravated by persistent chronic ischemic damage to the macula. The correlations among DRIL resolution, SD-OCT structural findings, other OCT-A perfusion markers, and functional outcomes yielded significant conclusions. Some eyes showed recovery of structural biomarkers with improved VD indices and capillary remodeling at the level of the deep and superficial vascular plexuses in the macula and lower-than-normal CVI and CFA values. The FAZ remained irregular and enlarged without recovery data, indicating that the FAZ may be a prognostic indicator of the worst final visual acuity. Postoperative diabetic macular ischemia of the photoreceptor and the outer and inner retinal layers is responsible for poor visual outcomes, possibly because of the low vascular perfusion in the visual photoreceptor cells in retina and mainly because of low choroidal indices with consequent cellular ischemic damage. The structural and perfusion data described here expand our knowledge of DRIL and other biomarkers obtained using SD-OCT and OCT-A, suggesting that perfusion indices such as the CVI and CFA could be used as clinical and prognostic biomarkers for TRD. Further research to determine the optimal timing for surgical intervention in patients with TRD involving the macula secondary to PDR is needed to minimize irreversible damage to retinal tissue and to vision. Careful longitudinal ophthalmological structural and perfusion evaluations, especially in patients with early-stage active PDR, are highly recommended to prevent tractional complications. Postoperative functional results are often disappointing, even in cases of successful anatomical reapplication of the macula, as described by the authors and the extensive literature reviewed in this chapter.

Acknowledgments

We express our deep appreciation to the technical staff of the Retina Department at Oftalmologia Integral ABC (Nonprofit Medical and Surgical Organization), Mexico City, Mexico, which is affiliated with The Postgraduate Study Division at the National Autonomous University of Mexico.

Conflict of interest

The authors declare no conflicts of interest.

Funding

No funding or grant support was received for this study.

Patient consent

Written informed consent was obtained from all patients in accordance with the institutional guidelines.

Disclaimer

All the scientific comments issued in this chapter are solely the responsibility of the authors and not of the institutions with which they are affiliated.

List of abbreviations

AVC	average vessel caliber
BCVA	best-corrected visual acuity
BRB	blood-retina barrier
CFA	choriocapillaris flow area
CME	cystoid macular edema
CSFT	central subfield thickness
CT	choroidal thickness
CVI	choroidal vascularity index
C3F8	perfluoropropane
Cks	centistokes
DM	diabetes mellitus
DME	diabetic macular edema
DRIL	disorganization of the retinal inner layers
DR	diabetic retinopathy
DCP	deep capillary plexus
ELM	external limiting membrane
ERM	epiretinal membrane
EZ	ellipsoid zone
FFA	fundus fluorescein angiography
FAZ	foveal avascular zone
GCL-IPL	ganglion cell layer-inner plexiform layer
HD	high-definition
ICG	indocyanine green dye
INL	inner nuclear layer
logMAR	logarithm of the minimum angle of resolution
MTM	myopic traction maculopathy
OCT-A	optical coherence tomography angiography
OPL	outer plexiform layer
PDR	proliferative diabetic retinopathy
PCN	perifoveal capillary network
PPV	pars plana vitrectomy
PFO	perfluoro-n-octane
PRP	panretinal photocoagulation
RNFL	retinal nerve fiber layer
RRD	rhegmatogenous retinal detachment
RPE	retinal pigment epithelium
SO	silicone oil
SSI	signal strength index
SCP	superficial capillary plexus
SF6	sulfur hexafluoride
SRF	subretinal fluid
TRD	traction retinal detachment
VA	visual acuity
VD	vessel density
VEGF	vascular endothelial growth factor
VH	vitreous hemorrhage
VT	vessel tortuosity