|Year : 2018 | Volume
| Issue : 12 | Page : 1696-1699
Current role of optical coherence tomography angiography: Expert panel discussion
Aniruddha Agarwal1, Dilraj S Grewal2, Glenn J Jaffe2, Michael W Stewart3, Sunil Srivastava4, Vishali Gupta1
1 Department of Ophthalmology, Advanced Eye Center, Post Graduate Institute of Medical Education and Research, Chandigarh, India
2 Division of Retinal Ophthalmology Duke Eye Center, Durham, NC, USA
3 Department of Ophthalmology, Mayo Clinic in Jacksonville, Florida, USA
4 Department of Ophthalmology, Cole Eye Institute Cleveland Clinic, Cleveland, Ohio, USA
|Date of Submission||22-Jun-2018|
|Date of Acceptance||06-Aug-2018|
|Date of Web Publication||19-Nov-2018|
Dr. Vishali Gupta
Department of Ophthalmology, Advanced Eye Center, Postgraduate Institute of Medical Education and Research, Sector 12, Chandigarh - 160 012
Source of Support: None, Conflict of Interest: None
Optical coherence tomography (OCT) angiography is a promising technique that provides depth resolved images of blood flow in the retina and choroid with levels of detail far exceeding that obtained with older forms of imaging and offers several advantages over conventional angiography to visualize the chorioretinal vasculature in a rapid and noninvasive manner. However, as with any evolving imaging technique, there are ongoing challenges in terms of need for new equipment, limitations of imaging capability and software processing techniques, as well as in understanding the implications of the imaging and its correlation with pathophysiology of the retina and choroid. This expert panel discussion is structured to provide an insight into the current state of OCT angiography.
Keywords: Angiography, choroid, imaging, OCT angiography, optical coherence tomography, retina
|How to cite this article:|
Agarwal A, Grewal DS, Jaffe GJ, Stewart MW, Srivastava S, Gupta V. Current role of optical coherence tomography angiography: Expert panel discussion. Indian J Ophthalmol 2018;66:1696-9
|How to cite this URL:|
Agarwal A, Grewal DS, Jaffe GJ, Stewart MW, Srivastava S, Gupta V. Current role of optical coherence tomography angiography: Expert panel discussion. Indian J Ophthalmol [serial online] 2018 [cited 2019 May 20];66:1696-9. Available from: http://www.ijo.in/text.asp?2018/66/12/1696/245600
OCT angiography is a promising diagnostic tool, it is still very much in evolution. There are rapid changes both in hardware with increased scan speeds and a shift toward swept source (SS) engines as well as software changes with improved segmentation and analytical capabilities.
While these hardware and software changes are exciting, there is still debate on the utility and capability of OCT angiography in terms of standardized grading and ability to detect longitudinal changes and its use as an endpoint in clinical trials.
We asked several pioneering OCT angiography researchers for their insight into the current state of OCT angiography:
- Glenn J. Jaffe, MD
Professor of Ophthalmology
Robert Machemer M.D.
Professor of Ophthalmology
Chief, Division of Retinal Ophthalmology
Duke Eye Center, Durham, NC, USA
- Michael W. Stewart, MD
Professor and Chairman of the
Department of Ophthalmology
Mayo Clinic in Jacksonville, Jacksonville, FL, USA
- Sunil Srivastava, MD
Cole Eye Institute
Cleveland Clinic, Cleveland, OH, USA
- Vishali Gupta, MD
Professor of Ophthalmology
Advanced Eye Centre
Postgraduate Institute of Medical
Education and Research, Chandigarh, India
Here, they share their experiences with clinical and research applications with the current generation of OCT angiography machines, current limitations, and major unmet challenges that need to be addressed to drive the applicability of OCT angiography to the next level – both in clinic and for research.
| Question 1: There are Several Different Optical Coherence Tomography Angiography Platforms Currently Available and Others Being Developed. Are the Measurements Across Platforms Comparable?|| |
Currently available spectral domain OCT (SD-OCT) machines use different anatomic landmarks on which to base morphologic measurements such as central subfield thickness and macular volume. Physicians who manage patients with macular diseases and scan patients with different machines from visit to visit need to be aware of these differences and compensate for them when making treatment decisions. Conversion factors that can be used to standardize measurements made by the various SD-OCT machines have been published.
Since visualization of both the retinal and chorioretinal vasculature relies heavily on segmentation analysis but not on the type of thickness measurements acquired with OCT, errors and differences between the two technologies differ. In an analysis of choroidal neovascularization, automated SS-OCT angiography measurements were larger than those acquired with SD-OCT angiography and correlated better with hand-drawn measurements. The abilities to image the retinal vasculature differ however, and I am unaware of studies that directly compare the two technologies in eyes with vein occlusions. Wide-field montage images from SD-OCT angiography will be useful to assess the mid-peripheral vasculature but SS-OCT angiography is better able to image much broader areas of retina and will be more useful for visualizing peripheral retinal perfusion.
A recent survey of retinal specialists found that the majority of them preferred the image quality of SS-OCT angiography over SD-OCT angiography regardless of the retinal pathology, but they believed that the clinical utility of each technology was the same.
| Question 2: What Unmet Challenges Remain in Achieving Standardized Grading of Optical Coherence Tomography Angiography?|| |
Similar to SD and SS-OCT, there are now multiple hardware platforms for OCT angiography, each with their own proprietary image processing software algorithm. OCT angiography image grading is typically performed on enface image slabs which are obtained by segmentation of the different retinal layers into various horizontal depth-resolved slabs such as superficial and deep retinal plexus, as well as the choriocapillaris and choroid. The different OCT angiography hardware platforms each have associated software with different algorithms to resolve the images and to reduce noise and artifact. In addition, boundaries for these retinal and choroidal slabs differ across platforms. SD-OCT had similar challenges since each manufacturer used different boundary identification algorithms and conversion factors had to be developed to allow interchangeable measurements for retinal thickness. However, with OCT angiography, the volume of data that needs to be standardized is much greater and there is an unmet need for further research to standardize these segmentation boundaries and image processing across platforms. Quantification of flow for example within a choroidal neovascular membrane also needs to be standardized. While structural OCT B scans are used in conjunction with OCT angiography slabs, it is as yet not clear whether change in CNV blood flow can be accurately detected with OCT angiography. The second issue is standardization of the metrics used for quantification. Unlike in structural OCT, where retinal thickness and volume are well-recognized outcome measures, such parameters are still being developed and validated in OCT angiography. Metrics like vessel density, perfusion density, blood flow index, skeletonized density, vessel length density, and fractal dimensions are being studied but there is yet no consensus on standardized nomenclature.
| Question 3: How Do You Use Optical Coherence Tomography Angiography to Guide Your Treatment in Diabetic Retinopathy?|| |
- Detection or confirmation of neovascularization of the disk and neovascularization elsewhere (within the field of view)
- Differentiating neovascularization of disk from collaterals
- Identifying small areas of ischemia in macular edema
- Monitoring response to an intervention.
| Question 4: How Do You Use Optical Coherence Tomography Angiography to Guide Your Treatment in Retinal Vein Occlusions?|| |
Let me start by discussing my use of fluorescein angiography in vein occlusions prior to the introduction of OCT angiography, against which I will contrast my current use of OCT angiography. Diagnosis of retinal vein occlusions has always been based on clinical findings, so I have never used fluorescein angiography routinely. However, I found fluorescein angiography helpful for the following reasons:
- Identifying areas of vascular incompetence in eyes where the pattern of hemorrhage did not completely fit the “delta”-shaped distribution characteristic of branch vein occlusions. This was sometimes helpful in establishing an retinal vein occlusion diagnosis in eyes with retinal arterial macroaneurysms or diabetic retinopathy with an asymmetric distribution of hemorrhages.
- Diagnosing enlargement of the foveal avascular zone. This may be important for two reasons:
- Identifying eyes with macular edema that may have limited visual potential even after successful treatment with vascular endothelial growth factor (VEGF) inhibitors. Regardless of the amount of foveal nonperfusion, I still treat retinal vein occlusion-associated macular edema with anti-VEGF therapy, but I may set expectations with the patient regarding improvement in visual acuity.
- If I need to perform a grid laser photocoagulation because pharmacotherapy has failed to satisfactorily resolve the edema, I avoid treating nonperfused retina that is contiguous with the foveal avascular zone.
- Quantifying the total area of retinal nonperfusion. With the advent of ultrawide-field angiography, considerable interest has emerged for calculating the retinal ischemic index. Though I still treat eyes with large areas of peripheral retinal ischemia according to the responsiveness of their macular edema, if I am planning to discontinue or suspend pharmacotherapy, I become concerned about the development of rubeosis irides (CRVOs) or retinal neovascularization (BRVOs). I frequently apply prophylactic scatter photocoagulation to these eyes. Presently, we have no data to support the use of targeted laser photocoagulation of nonperfused peripheral retina to decrease the burden of pharmacotherapy.
Although my use of fluorescein angiography with vein occlusions was case by case, I now use OCT angiography to routinely assess macular perfusion, particularly in the deep vascular plexus since this has been correlated with visual outcomes. I find information interesting and, at times valuable, I do not yet believe that current OCT angiography technology is a “game changer” for the management of retinal vein occlusion. Of course, with OCT angiography, I am limited to assessing macular perfusion, so I still perform a small number of ultrawide-field fluorescein angiograms to assess peripheral perfusion.
| Question 5: How Has the Imaging the Choriocapillaris and Choroid Using Optical Coherence Tomography Angiography Been Useful in Your Uveitis Practice?|| |
As we all know, many of the uveitis entities have their origin from the choroid and/or choriocapillaris and retina gets involved secondarily. Looking at the fluorescein angiography gives an idea about retinal changes mainly. Indocyanine green is the gold standard but cannot be repeated frequently during the follow-up because of its invasive nature. OCT and OCT angiography help me in evaluating choroid/choriocapillaris the following ways:
- Detection of Type 1 choroidal neovascularization when it is in its incipient stage, especially if seen next to a scar. These new vessels are extremely difficult to detect in a scarred retina (as in Serpiginous Choroiditis) on fluorescein angiography but can be delineated very well in the early stages on OCT angiography.
- The choroidal stroma shows multiple dark spots on OCT angiography, indicative of small granulomas, in Vogt-Koyanagi-Harada disease. These granulomas can be seen better once the overlying serous retinal detachment starts resolving. In the pre-OCT angiography era, we used to monitor serous detachment and then choroidal thickness on OCT. However, in our current practice, we persist with the induction therapy as long as we see these dark spots on OCT angiography (even after the serous detachment has resolved and choroidal thickness returned to normal). This helps in treating the subclinical disease and minimizes recurrences.
- OCT helps in both the acute and chronic phase to study the structural damage caused by the disease and determine the visual outcome.
- Choroidal vascularity index is quite useful in determining the vascularity of inflamed versus ischemic choroid.
| Question 6: How Do You Use Optical Coherence Tomography Angiography to Diagnose and Manage Patients With Infectious Uveitis?|| |
Combined OCT and OCT angiography give several important clues in the diagnosis of infectious uveitis:
- The thickening of neurosensory retina with hyperreflective oval deposits on inner retinal surface is quite characteristic of toxoplasmic retinochoroiditis. In addition, one may also see hyperreflective round deposits on posterior hyaloid, often with mirror image on retina in patients with toxoplasmosis. In contrast, patients with viral retinitis show lack of oval deposits on retinal surface.
- OCT is very useful in diagnosing subacute sclerosing panencephalitis that shows central destruction of neurosensory retina with only internal limiting membrane drape over the RPE. OCT angiography shows lack of perfusion in these patients, thus confirming the presence of ischemia
- OCT angiography is also very useful in identifying the ischemia of outer retinal plexus in patients with dengue maculopathy, etc.
- Development of retinal angiomatous proliferation or CNV can complicate TB granulomas and addition of anti-VEGF therapy aids in resolution of these granulomas. OCT and OCT angiography are extremely helpful in identifying these lesions at the earliest.
| Question 7: What are the Top Three Unmet Needs for Optical Coherence Tomography Angiography in Uveitis?|| |
The unmet needs are software, software, and software. I credit the imaging companies as the development of novel tools is progressing and there are multiple groups working on improving our ability to analyze structures like the choroid by reducing artifact. Ultimately, any quantitative analysis that we want to perform will need to be robust and reproducible. In diseased eyes, this will require software that can accurately segment the layers of the retina and choroid and also reduce the impact of artifacts. In normal eyes, this is easier, but in eyes with outer retinal changes from uveitis or cystoid macular edema or thinning from vasculitis, this is a bit more challenging. I would also say we still do not understand the impact of findings on OCT angiography imaging on outcomes as most of our data are retrospective. We need prospective data that allow us to track changes over time and identify patterns which could be predictive of inflammation or visual loss.
| Question 8: Do You Think There is Potential for Development of New Endpoints for Uveitis Clinical Trials Using Optical Coherence Tomography Angiography?|| |
Yes, I think there is potential, but we have a long way to go before we could use it an endpoint. Any measurement tool has to meet certain criteria of reliability, validity, and responsiveness before it can be considered as an endpoint. Additionally, there needs to be a standard accepted on acquisition patterns and software that would automate the measurement of the endpoint. As amazing as the images from OCT angiography are, currently the software analysis is fragmented as each company develops their own custom analysis. Finally, any endpoint would have to show a relation to visual functional outcome or reduction in complication from inflammation. We currently do not have enough prospective data to make that assessment.
| Question 9: How Far Along are We in Being Able to Evaluate the Peripheral Retina Using Optical Coherence Tomography Angiography?|| |
With current SD-OCT, we can get some view of the peripheral retina. With swept-source engines, the peripheral retina is easier to capture and view. It does require a good patient who can fixate for several seconds. The images are often montaged together. Remember though that imaging of the peripheral retina will come with some artifacts and the resolution will drop some depending on the scan pattern. The ability to visualize large vessels is possible. The accuracy of the small capillary vessels imaging will be limited by resolution.
| Question 10: How Do You Evaluate for and Address Artifacts in Optical Coherence Tomography Angiography?|| |
Some artifacts, such as those produced by media opacities and retinal blood vessels, are unavoidable and difficult to compensate for.
Minimizing movement artifacts is critical to obtaining an accurate scan. Simply introducing a competent OCT technician to an OCT angiography machine does not guarantee optimal scanning. OCT technicians need to be adequately trained in not only correct use of the device but also in the science behind angiographic image acquisition. Patient cooperation during scanning is of more importance with OCT angiography since individual points need to be repeatedly scanned over the course of 1 + seconds. Though the current machines have eye-tracking software to minimize motion artifacts, patient cooperation remains an important factor. Patients who are elderly or deaf, or those with orthopedic abnormalities or phobias regarding foreign objects approaching the eyes, may be particularly challenging. Technicians need to exhibit much more patience and skill when performing OCT angiography than they need when performing OCT.
Automated segmentation of the retinal layers is both a blessing and a curse. This function enables the device to create composite files, composed of each layer from the vitreoretinal interface down to the choroid, but segmentation errors are common particularly in diseased retinas. For most eyes undergoing OCT, I am satisfied with the automated analysis provided by the computer and this can be exported as a PDF to remote work stations. However, for OCT angiography, I routinely walk to the machine to perform my own manual segmentation to be sure that each layer has been correctly and optimally identified. As a result, OCT angiography is much more time consuming for both the technician and physician.
| Question 11: What are Some of the Challenges With the Use of the Current Generation of Optical Coherence Tomography Angiography for Clinical Trials in Diabetic Retinopathy and Macular Degeneration?|| |
OCT angiography is currently limited to evaluation as an exploratory endpoint in clinical trials. OCT angiography has been evaluated in exudative age-related macular degeneration to determine the location of neovascularization and longitudinally assess the dimensions of choroidal neovascular membranes. We have also used it in eyes with diabetic retinopathy and diabetic macular edema to quantify the foveal avascular zone area, vessel density, and perfusion density in the superficial capillary plexus.
The two main challenges to achieve standardized grading are to obtain good quality images and accurate software segmentation of the different anatomical boundaries. The automatic segmentation algorithms are highly sensitive to scan quality. In addition, pathology such as retinal pigment epithelial detachments, intraretinal and subretinal fluid, and subretinal hyperreflective material can cause failure of the automated segmentation algorithm. When grading such images, it is critical to recognize image artifacts as a variety of causes include scan quality, motion artifact, and projection artifact among others. While some of the segmentation errors can be manually corrected with the current software iterations, others cannot and often a number of images often cannot be graded accurately. We need improved software algorithms both for artifact reduction and greater interconversion among the platforms.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.