Article Date: 8/1/2009

Assessing SD-OCT Technology

Assessing SD-OCT Technology

Is the new generation of OCT devices a milestone in diagnosis and management of retinal and optic nerve diseases?

JEROME SHERMAN, O.D., F.A.A.O., New York, N.Y.

Since the first papers on optical coherence tomography (OCT) were published in the early 1990s, there has been a growing acceptance of the value that this diagnostic tool can bring to eye care, particularly when dealing with retinal diseases. The capability of OCT to tease out and measure the retinal nerve fiber layer (RNFL) also makes it an ideal technology for the early detection of glaucoma as well as non-glaucomatous optic neuropathies.1

Still, I don't think we quite recognize the value that these devices bring to our practices. Here, I'll explain their value, offer advice on what to look for in an OCT and provide several case examples.

The future standard of care

Today, a dilated ophthalmoscopic examination is the standard of care, but in the next five to 10 years, I expect that the OCT will supplement or perhaps even replace the ophthalmoscope as the standard of care for a retinal examination. The primary reason for my prediction is the increased sensitivity found with OCT imaging, particularly the new generation of spectral domain OCT (SD-OCT) systems (also known as Fourier domain, high definition or 3D-OCT). An excellent example of this ability is illustrated by an informal study that my colleagues and I conducted into the incidence rate of Bergmeister's papilla (BP).

In the informal review of 100 SD-OCT scans of the optic disc in 50 optometry students (age = 20 to 30) here at the State University of New York (SUNY) School of Optometry, 73% of the eyes scanned revealed evidence of BP. In every case in which the OCT found BP, it was not observed during direct ophthalmoscope exam. The BP was often present in only one or two of the 128 B-scan slices taken with the SD-OCT. Because of its high resolution, the SD-OCT scans allowed it to reveal this congenital and benign structure.

Earlier diagnosis

Because BP can be easily identified with SD-OCT, I would argue that other subtle findings, such as early disc neovascularization, will be identified with SD-OCT earlier than with other Specdiagnostic tests. What the results of this informal study tell us is that OCT may enable us to see things that we are currently missing on a daily basis. This may be very significant for many conditions, such as disc neovascularization in diabetics, which may be picked up earlier than with standard ophthalmoscopy. This may explain why some diabetic eyes with classified non-proliferative disease present with vitreal hemorrhages at a later date in front of the disc.

By seeing more, we should be able to diagnose and then intervene at a much earlier stage.

Comparable to ERG

In the 20 months since my colleagues and I began using an SD-OCT system, we have found that earlier intervention is possible. A recent example involved a nine-year-old girl who was referred to us for reduced vision in both eyes. One of the possible diagnoses was retinal degeneration. Normally, if the back of the eye appears okay during a dilated retinal exam, we would turn to an electroretinography (ERG) exam. Today, we perform an SD-OCT exam first and, in a consecutive series of 20 cases, we have found that the OCT predicts the results of the ERG exam. This reduces the need for ERG tests that require specialized equipment and can be traumatic to young patients, such as this one. The same is true for fluorescein angiography (FA), which can cause between 5% and 10% of patients to become sick from the injected dye. There is an overall benefit to being able to use the SD-OCT in place of these more invasive tests while still having the same level of diagnostic accuracy. An added advantage is the ability to keep the patient in your practice without the need to refer elsewhere for these more invasive tests.

Today, the SUNY State College of Optometry uses SD-OCT on three groups of patients:

► Patients that have an observable retinal lesion on ophthalmoscopy.

► Those that have reduced vision and/or symptoms without any obvious fundus abnormalities.

► Normal patients in order to learn what normal and abnormal findings are present in patients whose clinical exams show normal results.

Advances in technology

SD-OCTs are a new generation of instruments. Earlier generation time domain OCTs (TD-OCTs) use a single detector. The new generation of OCTs use a spectrometer, and the resulting data undergoes Fourier analysis. SD-OCT devices acquire tens of thousands of Ascans per second. Of the seven new SD-OCTs that are now on the market (see the SD-OCT table below), axial resolution averages around 6μm rather than about 10μm for the original generation OCT. The core components of the seven SD-OCTS are quite similar, and the differences lie primarily in peripheral technologies that are incorporated into the seven different products. For example, the spectralis from Heidelberg Engineering (Vista, Calif.) combines an angiographic system capable of both FA and ICG (indo-cyanine green). The Cirrus HD-OCT (Carl Zeiss Meditec, Dublin, Calif.) offers anterior chamber imaging that does not require an add-on lens. Topcon Medical Systems (Paramus, N.J.) has paired a fundus camera with its SD-OCT, the 3D OCT. Optovue (Fremont, Calif.) was the first to add the option of a corneal-anterior segment module (CAM) to its SD-OCT, the RTVue-100. At present, we work with the TDOCT (Stratus, CZM), The Cirrus HD-OCT, the Topcon 3D OCT, the Spectralis and the RTVue- 100 — essentially, a first-generation and four second-generation systems.

Where TD-OCT was, in general, used for analysis of crosssections of lesions identified by other procedures, we may find SD-OCT valuable for the screening of retinal and optic nerve disorders. The most commonly used program with the Topcon 3D and the Cirrus HD takes 128 or (512 point) cross-sections, each a 6mm, B-scan slice. The SD-OCT provides actual data at each of the 128 points. Since SD-OCT technology can differentiate the RNFL from the remaining retinal structures, posterior pole scans that include the optic nerve head measure both the RNFL and macula thickness. Therefore, SDOCT can conceivably detect both glaucoma and macular disorders simultaneously.

More timely diagnosis?

The fundamental question is whether the additional resolution and number of scans with SDOCT results in enhanced detection and a more timely diagnosis. If this is correct, does this lead to improved patient care? I believe the answer is "yes," and as we learn more about SD-OCT, I expect it to replace other means of diagnostic testing in many patients who have suspected retinal or optic nerve disease.

Case examples

The following case examples discuss the benefits of SD-OCT.

Peripapillary retinoschisis (PPRS) — Retinoschisis in the macula of highly myopic eyes is well documented with TD-OCT, yet cases of 360° isolated peripapillary retinoschisis (PPRS) without macula schisis have not been reported. In one case of PPRS, identified initially with SD-OCT, a TD-OCT was then performed and demonstrated to be capable of documenting the abnormality. However, vitreal retinal traction in several of the 128 SD-OCT slices also appeared to reveal the etiology of the PPRS.

Genesis of posterior staphyloma — Posterior staphyloma is a well-recognized finding in high myopia and is typically documented using the binocular indirect ophthalmoscope, fundus photos and B-scan ophthalmic ultrasound. The posterior "bulge" is an important contributor in the rating of myopia as the sixth leading cause of blindness worldwide. SD-OCT — especially using the 3D presentation — appears to document early presentation. In one of our cases that involved a 7.0D myope that had 20/20 best-corrected- visual-acuity, the two obvious bulges were not aligned with the macula (See figure 1). Such a bulge in the macula would predictably lead to an increase in myopia since a 1mm bulge corresponds with a 3.0D myopic shift.

Figure 1: Here's a 7.0D myope with 20/20 best-corrected-visualacuity, where the two obvious bulges (both invisible to ophthalmoscopy) were not aligned with the macula.

Outer and inner segment (IS/OS) junction of the photoreceptors — Although the IS/OS junction has been identified with TM-OCT, little or no published literature exists on the use of the IS/OS junction in the detection of outer retinal disorders using commercially available equipment.2 This may be due to the inconsistency in visualizing the IS/OS junction in normal eyes. But, if the junction is present in virtually all normal patients, its absence can confirm the presence of outer retinal disease. As a result, you could argue that the IS/OS junction, or perhaps more appropriately termed the photoreceptor integrity line (PIL), can be used as a marker for structural photoreceptor integrity (see figure 2).3 To confirm our hypothesis, we conducted a retrospective study of 350 consecutive normal eyes and 50 consecutive abnormal eyes that had outer retinal disease. We found the PIL present in every normal eye and abnormal (either absent, discontinuous or disorganized) in every eye that had outer retinal disease. This included retinitis pigmentosa, cone dystrophy, geographic atrophy of the retinal pigment epithelium, cases of retinal drusen, choroidal neovascularization membranes, macular holes, acute zonal occult outer retinopathy (AZOOR), acute multifocal placoid pigment epitheliopathy (AMPPE) and high myopia.

Figure 2: The normal photoreceptor integrity line (PIL) under the fovea and macula predicts normal vision.

The most devastating form of dry age-related macular degeneration is geographic atrophy of the retinal pigment epithelium (GARPE) when the lesion is centered in the fovea. However, some GARPE patients have normal visual acuity that suggests the photoreceptors in the fovea and macula are relatively intact. The SD-OCT of a GARPE patient (with 20/20 visual acuity) fortunately demonstrated a normal PIL in both eyes under the fovea. The presence of the normal PIL under the fovea and macula predicts normal vision, making this case an excellent example of this clinically useful relationship.

You can use the SD-OCT PIL as a bio-marker of photoreceptor integrity. It's the first ocular imaging device with this crucial capability. OM

References available upon request.

Dr. Sherman is a distinguished teaching professor of clinical sciences at SUNY State College of Optometry. He serves as president of the Optometric Retina Society. E-mail him at

Optometric Management, Issue: August 2009