Fallstricke in der Netzhautbildgebung mittels optischer Kohärenztomographie

[Avoiding mistakes in anti-VEGF intravitreal injection therapy]

Intravitreal injection (IVI) of drugs for treatment of various macular diseases is now one of the most frequently performed surgical procedures worldwide. As mostly chronic diseases are treated, the indications for treatment often mean a continuous treatment over years with a corresponding effort regarding spatial, personnel and financial resources. The diagnosis and indications for treatment are nowadays mainly made by spectral domain optical coherence tomography (SD-OCT). The ability to clinically assess and evaluate a fluorescence angiography is less practiced, although these are still a component of the indications for intravitreal injections. Therefore, it can happen that despite all diligence patients may receive anti-vascular endothelial growth factor (VEGF) treatment, sometimes permanently, based on a misinterpretation of the macular diagnosis or disease activity and these indications, once made, are rarely questioned or retracted. Therefore, the aim of this manuscript is to point out possible and typical misinterpretations in the indications or continuation of IVI treatment with anti-VEGF by means of case studies and to sensitize for differential diagnoses.

MINIMAL OPTICAL COHERENCE TOMOGRAPHY B-SCAN DENSITY FOR RELIABLE DETECTION OF INTRARETINAL AND SUBRETINAL FLUID IN MACULAR DISEASES

PURPOSE

To determine the minimal optical coherence tomography B-scan density for reliable detection of intraretinal and subretinal fluid.

METHODS

Spectral domain optical coherence tomography raster scanning (Spectralis; Heidelberg Engineering, Heidelberg, Germany) using a scan field of 20° × 20° of 97 B-scans with an interscan distance (ISD) of 60 μm was performed in 150 eyes of 150 consecutive patients at monitoring visits for intravitreal anti-vascular endothelial growth factor therapy. Using custom software, every other B-scan was repeatedly deleted to generate additional data sets with an ISD of 120 μm (49 B-scans), 240 μm (25 B-scans), and 480 μm (13 B-scans). Two independent reviewers evaluated the data sets for the presence of cystoid spaces of intraretinal fluid and subretinal fluid.

RESULTS

Treatment diagnoses were neovascular age-related macular degeneration (68.0%), macular edema secondary to retinal vein occlusion (20.7%), diabetic macular edema (10.7%), and other retinal diseases (4.0%). Using the source data sets with an ISD of 60 μm, intraretinal fluid was detected in 56.0%, subretinal fluid in 19.3%, and either/both in 68.7%. Compared with these results, the sensitivity of detection of intraretinal fluid and/or subretinal fluid using an ISD of 120 μm, 240 μm, and 480 μm was 99.0% (95% confidence interval, 94.7-100.0; P = 0.5), 97.1% (91.7-99.4; P = 0.1), and 87.4% (79.4-93.1; P = 0.0001), respectively.

CONCLUSION

An increase of ISD up to 240 μm does not significantly impair the detection of treatment-relevant exudative retinal changes in monitoring during intravitreal therapy of macular diseases. These findings are relevant for the choice of optical coherence tomography B-scan density in both routine clinical care and interventional clinical studies.

[Second opinion in retinal imaging]

With the use of digital imaging systems and the possibilities of data exchange, the second opinion is becoming increasingly more important in retinal imaging. For a meaningful application, technical imaging requirements and medical assessment quality requirements have to be fulfilled. Responsibilities should be clearly defined. The aim must be to achieve a significant contribution to ensure high-quality patient care.

[Baseline diagnostics and initial treatment decision for anti-vascular endothelial growth factor treatment in retinal diseases : Comparison between results by study physician and reading centers (ORCA/OCEAN study)]

BACKGROUND

The ORCA module of the non-interventional OCEAN study investigated the use of retinal imaging diagnostics in the clinical treatment of patients undergoing vascular endothelial growth factor (VEGF) inhibitor treatment as part of routine clinical care. This article analyzes the agreement between the diagnosis documented by the treating ophthalmologist and the evaluation of reading centers at baseline as well as the effect on the response to treatment during the course.

METHODS

A total of 396 patients (age 75.4 years) were enrolled in which ranibizumab treatment was indicated by the treating ophthalmologist due to either diabetic macular edema (DME), neovascular age-related macular degeneration (nvAMD) or retinal venous occlusion (RVO). Over a period of 24 months, patient and examination data, treatments and interpretation of retinal imaging data by the treating ophthalmologist were systematically recorded. Furthermore, retinal imaging data were also evaluated by three reading centers.

RESULTS

In 338 out of 396 (85.4%) study eyes, the baseline diagnosis of the treating ophthalmologist was confirmed by the reading centers (DME 87.5%, nvAMD 82.3%, RVO 94.9%). In 17 of the remaining 58 eyes with a discrepant diagnosis, there was at least a consensus with respect to the indications for VEGF inhibitor therapy. The differential diagnoses included a variety of different retinal diseases. During follow-up of up to 3 months, eyes with a consistent diagnosis showed a clear increase in visual acuity (6.4 versus 2.7 letters, p = 0.05) and greater decrease in central retinal thickness (-112.3 versus -24.4 μm, p < 0.0001).

DISCUSSION

The initial treatment decision for anti-VEGF therapy with consideration of the differential diagnoses can be challenging. Accurate evaluation of the clinical and imaging findings along with appropriate expertise appear to be important. The observation of superior initial response in eyes with a consensus of the diagnosis at baseline underlines the relevance of an adequate initial assessment for a successful treatment outcome.

Optical coherence tomography angiography (OCT-A) in an animal model of laser-induced choroidal neovascularization

Aim of the study was to compare optical coherence tomography angiography (OCT-A) and conventional fluorescein angiography (FA) for quantitative analysis of the retinal and choroidal vasculature in the animal model of laser-induced choroidal neovascularization (CNV). Therefore, Dark Agouti rats underwent argon laser photocoagulation to induce CNV at D0. In vivo imaging using combined confocal scanner laser ophthalmoscopy (cSLO)-based FA and OCT-A (Heidelberg Engineering GmbH, Heidelberg, Germany) was performed before and immediately after laser treatment as well as at day 2, 7, 14 and 21. OCT-A en-face images were compared to cSLO images obtained by conventional FA topographic uptake recorded using a series of different pre-defined focus settings. For a quantitative comparison of CNV imaging by OCT-A and FA, CNV area, vessel density, number of vessel junctions, total vessel length and number of vessel end points were analyzed. Subsequent ex vivo analyses of the CNV included immunofluorescence staining of vessels in retinal and RPE/choroidal/scleral flatmount preparations. We found, that OCT-A allowed for high-resolution non-invasive imaging of the superficial, intermediate and deep retinal capillary plexus as well as the choroidal blood vessels in rats. Compared with OCT-A, visualization of CNV progression by invasive FA was less accurate, in particular the deep vascular plexus was visualized in more detail by OCT-A. The area of neovascularization was mainly detected in the deep retinal vascular plexus, outer nuclear layer (ONL), ellipsoid zone (EZ) and the choroid. Within the laser lesions, signs of CNV formation occurred at day 7 with progression in size and number of small vessels until day 21. Due to leakage and staining effects, CNV areas appeared significantly larger in FA compared to OCT-A images (p ≤ 0.0001 for all tested layers). Vessel density, number of vessel junctions, total vessel length and number of vessel end points were significantly higher in intermediate vascular plexus (IVP) and deep vascular plexus (DVP) in OCT-A compared to FA images. Overall, CNV area in flatmounts was similar to OCT-A results and much smaller compared to the area of dye leakage by FA. This study demonstrates that in vivo OCT-A imaging in small animals is feasible and allows for precise analysis of the formation of new blood vessel formation in the animal model of laser-induced CNV. Given its superior axial resolution, sensitivity and non-invasiveness compared to conventional FA imaging, OCT-A opens the door for a more detailed evaluation of CNV development in such a model and, thus, enables the analysis of the response to novel therapeutic interventions in longitudinal in vivo studies.