Noninvasive OCT angiography allows for detailed visualization of retinal vasculature
Click Here to Manage Email Alerts
September’s announcement by Carl Zeiss Meditec that its AngioPlex OCT angiography technology had received 510(k) clearance from the FDA set the stage for other systems to likely also soon become commercialized in the U.S. The technology allows extreme close-up imaging of the retinal vasculature for assessing retinal vascular diseases and holds potential for guiding treatment decisions and monitoring patient responses to therapy.
“Over the past 20 years, OCT has developed rapidly as a ‘noninvasive’ method of imaging for medical diagnosis and defining activity criteria, such as intra/subretinal fluid, hyper-reflective dots and dense intraretinal areas,” Gabriel J. Coscas, MD, emeritus professor of ophthalmology at the University of Paris XII in Créteil, France, said. But over the past 2 years, it was found that additional information useful for treatment decisions can now be gained with OCT angiography (OCTA).
“This huge progress consists of having both functional and morphological assessment from a single dye-less examination,” Coscas said. “The rapid and noninvasive nature of OCTA allows for easier follow-up of morphological and functional changes in prolonged, repetitive and even monthly evaluations.”
Capabilities of OCTA
SriniVas R. Sadda, MD, president and chief scientific officer of the Doheny Eye Institute in Los Angeles, is attracted to OCTA in part because of the ability to visualize the retinal microcirculation noninvasively without the use of dye, while at the same time reaping the benefits of OCT imaging.
“[OCTA offers] a tremendous amount of contrast. In many ways it provides more detail and better resolution than traditional fluorescein angiography,” Sadda said, citing the ability to visualize the radial peripapillary capillary network near the optic nerve.
“That network is so fine and so tightly packed that with a traditional fluorescein angiogram we could never see it. But now with OCTA, we are able to observe the network,” Sadda said.
This visualization also extends to the central part of the macula.
“With OCTA, you have a lot more confidence for assessing retinal capillary perfusion status, which can be very helpful in understanding various retinal vascular diseases, with diabetic retinopathy and retinal vein occlusions being the most common,” Sadda said.
OCTA is also being used to evaluate macular telangiectasia. “We are able to see the vascular anomalies in so much more detail,” he said.
Another advantage of OCTA’s superior contrast is the ability to glean more quantitative data. Whereas fluorescein angiography typically relies on subjective categories to define mild, moderate or severe loss of blood vessels, OCTA has the potential to use automatic computer algorithms to quantify the level of capillary loss, Sadda said.
“[Optovue] is starting to provide numerical retinal vascular data such as flow density analysis. We can expect that all companies in short order will add quantitative metrics. Once you have quantitative metrics, you can follow disease much more precisely and understand the relationship between these anatomic parameters and other visual function measures,” Sadda said.
Challenging the gold standard
From Sadda’s perspective, the most important benefit of OCTA is that it provides depth-resolved visualization.
“[With standard fluorescein angiography], in the absence of excellent stereopsis, you are unable to tell which layer those retinal vascular abnormalities are coming from because the image you see is an amalgam,” he said.
Unlike fluorescein angiography, which is a two-dimensional image, OCTA displays a three-dimensional image.
“Therefore, you can peel off layer after layer, and look at the blood vessels in each layer,” Nadia K. Waheed, MD, MPH, an associate professor of ophthalmology at Tufts University School of Medicine in Boston, said. “We are seeing fine vessels as well as larger vessels. Some of these vessels can be indicative of whether disease is active and in the future may be indicative of how well someone is responding to treatment and the recurrence rates.”
A benefit of AngioPlex OCTA is that it color codes the separate layers of the retina “in a beautiful layout, so that you can see the vessels in the different layers,” Waheed said.
Amir H. Kashani, MD, PhD, an assistant professor of ophthalmology at Keck School of Medicine at the University of Southern California in Los Angeles, finds the prospect of performing noninvasive angiography in patients with chronic diseases on a more frequent basis to be enticing.
“With OCTA, you can look at capillary flow and blood flow almost at will,” he said. “This provides us an opportunity to look at aspects of disease that we were unable to look at before.”
Kashani said that fluorescein angiography, which has been the mainstay of evaluating retinal vascular diseases for many decades, is still a good test.
“But it is invasive, so we cannot do it all the time,” he said. The test involves injecting a fluorescein dye into a patient’s veins, usually in the arm, and then watching that dye circulate through the eye over a span of 5 minutes to 15 minutes, depending on the patient, with a special camera that can visualize the dye. “We then evaluate what is happening with the blood flow in the eye, based on where the dye is flowing,” Kashani said.
Despite the fanfare surrounding OCTA, Kashani believes there will continue to be a role for fluorescein angiography in assessing these diseases.
“But in large part, it will be replaced with OCTA as we come to understand the technology better and understand the findings better,” he said.
Published data
Kashani has been using AngioPlex, a part of the Cirrus HD-OCT system, for nearly 2 years, which at his institution represents several hundred eyes and for his network of universities about 1,000 eyes total. He finds the system “very reliable” with “excellent resolution compared to histology,” as reflected in a study of five healthy human subjects that he co-authored in Ophthalmic Surgery, Lasers and Imaging Retina. Kashani and colleagues have since evaluated several additional normal subjects, with results supporting the original findings.
“[The system is] very easy to use and intuitive from our standpoint,” Kashani said.
In another study published in Ophthalmic Surgery, Lasers and Imaging Retina, 33 subjects with varying severity of diabetic retinopathy underwent OCTA with the Zeiss system.
“We showed that OCTA has the capability to image, without injection of any dye, almost all of the clinically relevant features in diabetic patients,” he said.
Kashani was also the principal investigator of a study, which is pending publication in Retina, that evaluated OCTA in 25 subjects with different types of retinal vein occlusions.
“We showed again that OCTA can show us all of the clinically relevant features in patients who have vein occlusions in their macula,” he said.
Minimizing noise, artifact
Because OCTA technology is in early development, there are many artifacts that affect one’s ability to accurately interpret images. “These include motion artifact and segmentation artifact,” Sadda said.
“Motion artifact is one of the big issues with OCTA, that is, people moving their eyes as you are taking successive scans,” Waheed said. The Zeiss system incorporates real-time motion-tracking technology, which helps reduce motion artifact.
Waheed and colleagues have mostly used the AngioVue Imaging System from Optovue for performing OCTA for more than 2 years now. This device uses a unique algorithm, the split-spectrum amplitude decorrelation algorithm (SSADA), to reduce noise in the OCTA images, Waheed said. The system also uses a separate special algorithm, Motion Correction Technology, exclusively licensed from MIT, to reduce motion artifact by cross-referencing two different scans. This results in a fast acquisition time while maintaining a high-quality image.
The DRI OCT-1 Triton from Topcon has a swept-source system employing a laser at 1050 nm wavelength that scans at a much faster speed than a standard spectral domain laser, according to Waheed, who uses her own prototype swept-source OCT system (not from Topcon).
“As a result, image acquisition is quicker, and there is better penetration into the deeper layers of the eye, such as the choroid, for greater resolution,” she said.
Perhaps the most challenging artifact is projection artifact, which refers to a shadow being cast by an overlying retinal blood vessel, according to Sadda.
“As a result, a vessel that you think is in a deeper layer is actually a projection of a vessel in a more superficial layer,” he said.
Despite these challenges, Sadda expects OCTA to become a routine part of a clinician’s practice within the next 2 years.
The Spectralis diagnostic imaging platform from Heidelberg Engineering “uses a full-spectrum amplitude decorrelation algorithm to obtain OCTA images and provide images with a very high motion contrast, thus rendering a clear depth-resolved visualization of the different retinal and choroidal vascular layers,” Coscas said.
“Nowadays, we can easily use the Spectralis for most of the different maculopathies,” he said, including exudative age-related macular degeneration, choroidal neovascularization of any etiology, diabetic maculopathy, retinal vascular occlusion (evaluating not only cystoid macular edema but also ischemic components) and vasculopathies.
An important feature of the Spectralis OCTA system is the TruTrack active eye tracking.
“This represents a very reliable method to acquire OCT volume scans without motion artifacts,” Coscas said. The eye-tracking method is based on the simultaneous acquisition of fundus and OCT images, “making it possible to perform a continuous real-time quality check of the OCT data during the exam. This process ensures that only accurate OCT-images are stored.”
The device also allows for “a very precise segmentation of the volume scan, avoiding any risk of superimposed images and obtaining a specific localization of the CNV related to the retinal pigment epithelium and Bruch’s membrane,” Coscas said.
He was the lead author of a recent article in Retina that compared OCTA using Spectralis with traditional multimodal imaging in patients with exudative AMD for guiding treatment options. The prospective case series of 80 eyes described the use of very thin sections of CNV lesions and the analysis of five criteria of activity or quiescence of CNV in exudative AMD.
“These criteria could be valuable during post-treatment follow-up for evaluating treatment efficacy,” Coscas said.
Another player in the OCTA space is Nidek, in the form of prototype OCTA software for its RS-3000 Advance OCT device, according to Sadda.
“As with many other OCTA software and devices, this system is still a work in evolution, as people try to continue to improve upon the performance of the algorithms,” he said.
AMD and diabetic eye disease
At Retina 2015, Waheed presented on OCTA of wet AMD and diabetes.
“Using OCTA will most likely be paradigm-changing as far as treatment and development of treatment,” she said. “However, right now, a lot of people are skeptical about the clinical utility of OCTA because current treatment paradigms for diabetic retinopathy and wet macular degeneration almost exclusively use structural OCT to guide treatment and do not rely on any kind of angiographic information.”
But as more cases of wet AMD are investigated with OCTA, “we are seeing things that we have not seen before on fluorescein,” Waheed said. “One of the reasons for this is that OCTA provides beautiful depth resolution.”
Early studies using OCTA on wet AMD “have been really promising,” Waheed said. “Not only is OCTA able to reliably detect CNV in wet AMD, but it might also be used as treatment follow-up to guide us in the future about treatment decisions.”
Likewise, for diabetes, OCTA “shows us in very fine detail and in very high resolution the vascular changes that are taking place in diabetes,” Waheed said. “Unlike fluorescein, which can only show us the superficial layer of blood vessels in diabetics, OCTA can also show us changes in the deep retinal vessels from diabetes. We are starting to visualize changes now on imaging that we traditionally were able to see only in pathologic specimens.”
Waheed and colleagues have also demonstrated that OCTA allows the detection of changes in the vasculature of diabetics before observing clinical changes in the clinic or by fundus photography. Hence, new clinical trial endpoints might be created.
“The disadvantage of OCTA in diabetics is that is has a limited field of view compared to the widefield imaging we are able to obtain with fluorescein,” Waheed said. “However, the OCTA device companies are working to overcome this with wider scanning fields and montage images.”
For many years now, with traditional OCT, subretinal or intraretinal fluid accumulation “was a test of treatment efficacy to select the best schedule of intravitreal injections,” Coscas said. “Despite multiple efforts, though, a large number of patients remain as non-responders or partial responders.” OCTA could help clinicians make the best decisions “by analyzing activity criteria and blood flow in the network — not limited to exudative reaction and fluid accumulation — before initial treatment and during follow-up,” he said.
A study by Spaide in the American Journal of Ophthalmology consisted of a retrospective case series of 17 eyes of 14 patients with treated CNV secondary to AMD that were evaluated with OCTA using SSADA employed by the AngioVue platform.
“Dr. Spaide shows the progressive changes in CNV during a long period of follow-up, the development of quiescent vessels (arterialization or maturation) and the pattern of the recurrences with tiny capillary vessels,” Coscas said.
In addition, recent papers and lectures on OCTA from France, Italy and Germany at different European meetings, including Euretina and the European Association for Vision and Eye Research, “show an active interest in the development of this fascinating new method, which will substantially enrich conventional multimodal imaging,” Coscas said. For instance, OCTA “allows for visualization of well-perfused CNV. Coupled with the conventional OCT B-scan, which assesses the fluid accumulation, OCTA may become the choice method to evaluate the activity or quiescence of the lesion and suggest treatment and re-treatment decisions.”
Expanding technology
Waheed was a co-author of a prospective case series, published in Ophthalmic Surgery, Lasers and Imaging Retina, of three healthy eyes and two diseased eyes of three patients using widefield montage OCTA for improved visualization of the retinal vasculature in the posterior pole. Images were created by combining nine 3 mm × 3 mm OCTA scans from the AngioVue Imaging System from Optovue.
“Given time, we have the ability to expand OCTA technology into widefield imaging by stitching together smaller pieces of images to obtain a wide field of view,” Waheed said.
Besides diabetes, widefield imaging is advantageous for nearly every other retinal vascular disease, including central retinal vein occlusion, branch retinal vein occlusion and neovascularization in patients with vasculitis, according to Waheed, co-author of a review of OCTA that appeared earlier this year in the International Journal of Retina and Vitreous.
“Although OCTA is showing us new and unique pathologies in wet AMD, diabetes, vein occlusions and retinal vascular diseases, larger clinical trials are needed to establish new clinical trial endpoints and patient outcomes,” she said.
Still, Waheed unequivocally believes OCTA is a viable noninvasive option for visualizing the retinal vasculature in detail.
“It typically takes 3 to 4 seconds to secure an image, and the overall patient time in the room is probably closer to 10 minutes because you take multiple images,” she said. This is much less time-consuming than fluorescein angiography, which normally takes 15 to 20 minutes.
Moreover, because the patient is not stuck with a needle in his arm, “you can pretty much get an OCTA every single visit,” Waheed said.
“The patient must not move at all, there must be sufficient signal strength, and the camera needs to be aligned and focused,” she said.
For day-to-day clinical practice, Kashani believes OCTA has potential application in all of the aforementioned diseases, but predicts the first two markets will likely be for diabetes and vein occlusion because “these patients have very well-known ischemic findings that can be mapped out by OCTA and that we can follow very closely now.”
“[OCTA] is a very powerful, noninvasive method of assessing the retinal vasculature,” Kashani said. “I think it is truly going to revolutionize the way we evaluate the retina in the next 10 years, just as OCT revolutionized the way we looked at the retina in the past 20 years.” OCTA is also “much more user-friendly and patient-friendly than fluorescein angiography,” he said. – by Bob Kronemyer
- References:
- Coscas GJ, et al. Retina. 2015;doi:10.1097/IAE.0000000000000766.
- de Carlo TE, et al. International Journal of Retina and Vitreous. 2015;doi:10.1186/s40942-015-0005-8.
- de Carlo TE, et al. Ophthalmic Surg Lasers Imaging Retina. 2015;doi:10.3928/23258160-20150610-03.
- de Carlo TE, et al. Ophthalmology. 2015;doi:10.1016/j.ophtha.2015.01.029.
- de Carlo TE, et al. Retina. 2015;doi:10.1097/IAE.0000000000000882.
- Ishibazawa A, et al. Am J Ophthalmol. 2105;doi:10.1016/j.ajo.2015.04.021.
- Jia Y, et al. Ophthalmology. 2014;doi:10.1016/j.ophtha.2014.01.021.
- Jia Y, et al. Proc Natl Acad Sci USA. 2015;doi:10.1073/pnas.1500185112.
- Kuehlewein L, et al. Invest Ophthalmol Vis Sci. 2015;doi:10.1167/iovs.15-16510.
- Matsunaga D, et al. Ophthalmic Surg Lasers Imaging Retina. 2014;doi:10.3928/23258160-20141118-04.
- Matsunaga DR, et al. Ophthalmic Surg Lasers Imaging Retina. 2015;doi:10.3928/23258160-20150909-03.
- Muakkassa NW, et al. Retina. 2015;doi:10.1097/IAE.0000000000000836.
- Nagiel A, et al. JAMA Ophthalmol. 2015;doi:10.1001/jamaophthalmol.2015.0668.
- Spaide RF. Am J Ophthalmol. 2015;doi:10.1016/j.ajo.2015.04.012.
- Spaide RF, et al. JAMA Ophthalmol. 2015;doi:10.1001/jamaophthalmol.2014.3616.
- For more information:
- Gabriel J. Coscas, MD, can be reached at Centre Hospitalier Intercommunal de Créteil, 40, Avenue de Verdun, 94010 Créteil, France; email: gabriel.coscas@gmail.com.
- Amir H. Kashani, MD, PhD, can be reached at 1450 San Pablo St., Suite 4806, Los Angeles, CA 90033; email: ahkashan@usc.edu.
- SriniVas R. Sadda, MD, can be reached at 1355 San Pablo St., Suite 100, Los Angeles, CA 90033; email: ssadda@doheny.org.
- Nadia K. Waheed, MD, MPH, can be reached at Tufts Medical Center, 800 Washington St., Box 450, Boston, MA 02111; email: nwaheed@tuftsmedicalcenter.org.
Disclosures: Coscas reports he is a consultant for Heidelberg Engineering. Kashani reports he has received research support from Carl Zeiss Meditec. Sadda reports he has served as a consultant and received research support from Optos and Carl Zeiss Meditec and has been a consultant for Topcon and Nidek. Waheed reports she has been a consultant for Iconic Therapeutics, a contract researcher for Carl Zeiss Meditec and a member of the speakers bureau for ThromboGenics.
For what retinal conditions would you consider OCTA as a primary modality for monitoring or diagnosing disease, and for what indications would you consider it to be adjunctive?
OCTA can detect, monitor CNV
OCTA could be used as a primary modality for the detection of choroidal neovascularization and to monitor its area. Adjunctive indications include the visualization of retinal capillary dropout (in glaucoma and diabetic retinopathy) and choroidal capillary dropout (in dry AMD), for which better segmentation and measurement tools are needed. OCTA is also very sensitive in detecting retinal neovascularization (in diabetic retinopathy), but cannot be considered a primary tool for this because of the limited imaging area of current technology.
David Huang, MD, PhD, is Peterson Professor of Ophthalmology and professor of biomedical engineering, Casey Eye Institute, Oregon Health and Science University. Disclosure: Huang reports he receives patent royalties from Carl Zeiss Meditec and Optovue and has stock ownership in and receives speaker travel support, research grants and material support from Optovue.
OCTA is important adjunct to SD-OCT
Unlike conventional angiography, OCTA does not require dye administration and is noninvasive. Several studies have documented a potentially important role of OCTA in the evaluation of macular disease. In neovascular age-related macular degeneration, OCTA has for the first time provided detailed microvascular images of the structure and morphology of type 1, type 2 and type 3 neovascular complexes, and has permitted qualitative and quantitative analysis of neovascular lesions before and after anti-VEGF therapy.
OCTA does not, however, provide any information regarding leakage, edema and exudation, and therefore, it is likely to serve an important adjunctive role alongside spectral-domain OCT, but may potentially replace conventional angiography in the evaluation of macular disorders.
OCTA allows direct anatomical assessment of neovascular complexes and may be able to provide additional biomarkers to evaluate the progression of these neovascular lesions and their response to anti-VEGF therapy.
David Sarraf, MD, is clinical professor of ophthalmology, Stein Eye Institute, UCLA. Disclosure: Sarraf reports he receives research grants from Genentech, Optovue and Regeneron.