Optical Coherence Tomography in Age-related Macular Degeneration
Sabater Gozalvo, Alfonso1
De Nova Fernandez-Yañez, Elisa1
Optical coherence tomography (OCT) acquires cross-sectional images with semihistologic resolution. The technology precisely defines the location and nature of the changes in the retina and adjacent structures and objectively evaluates the thickness of the retina and surrounding structures. These capabilities allow detection of newly emerging fluid and/or intraretinal or subretinal tissue and tissue below the retinal pigment epithelium (RPE). OCT provides important information on serous retinal detachments, hemorrhages, and subretinal neovascular membranes that are components of exudative macular degeneration and more precise analysis of anatomic structures and subtypes of neovascular membrane lesions. These capabilities facilitate an understanding of the differences among the classic membranes, hidden membranes, retinal angiomatous proliferation (RAP), and disciform scars in the natural course of the disease and assess the treatment response. The images are presented in a scale of colors (or gray), such that tissues that reflect more light or disperse more light are shown in red and white, respectively, while the ones that reflect or disperse the least light are shown in blue and black. Tissues that moderately reflect light are shown as green or yellow. The color represents the optical properties of the tissues and not the tissues themselves. Therefore, the image is not real but represents the true dimensions of the measured structures. 1-10
The vitreous transmits light without reflecting it and is depicted in black in OCT images (Figure 1). The posterior hyaloid usually is indistinguishable from the retinal surface except when the posterior vitreous detaches and appears as a weakly reflective band. The choriocapillaris and choroid are highly reflective layers, because they are vascular and limit light penetration into the deeper layers. Blood vessels are identified by their high reflectivity and the screening effect generated on adjacent tissues (Table 1).7,8,11
Age-related maculopathy, considered the precursor of age-related macular degeneration (AMD), is defined as the presence of areas of hyperpigmentation or hypopigmentation of the RPE and/or confluent or soft drusen. When soft drusen in the macular region are associated with focal areas of mobilization of pigment (hypopigmentation and hyperpigmentation), there is an increased risk of progression to AMD. Drusen, which are degenerative nodular formations located mainly in Bruch’s membrane, are comprised of proteins, lipids, mucopolysaccharides, and other components that appear in adulthood and tend to increase in size and number over time. Drusen is the AMD characteristic that is detected early in the clinical examinations. On OCT, drusen appear as RPE deformation or thickening that may form irregularities and undulations (Figure 2).
Drusen are classified histologically as hard formations and defined as small hyaline deposits with delimited margins that are considered age-related low-risk changes, and soft drusen, which are deposits of granular or amorphous material considered to be precursors of AMD. Drusen can progress to an atrophic form (dry) or exudative form (wet) of AMD. The risk is based on the lesion number, size, and confluence. 4,6,8,12-15
Atrophic Macular Degeneration
In the macular region, progressive atrophy of the RPE, the outer retinal layers, the choriocapillaris, and dense clusters of drusen can be seen.
OCT can detect decreases in retinal thickness and increases in the reflectivity of the RPE detachment, which results from the decreased ability of the atrophic retinal tissue to attenuate light. The reduced retinal thickness and volume can be determined by the retinal maps that display the areas of greatest atrophy, identify the extent of the atrophy, and monitor progression. Geographic atrophy (GA) represents the final stage of dry AMD. 6,8,10,16,17
Exudative Macular Degeneration
OCT, a fundamental tool in the diagnosis and management of patients with choroidal neovascularization (CNV), allows identification of active neovascular membranes and determination of the extent of the membranes in many cases. The technology is useful for assessing subfoveal involvement and also helps diagnose hidden choroidal neovascular membranes, for which fluorescein angiography (FA) shows confusing patterns. OCT also can monitor treatment responses by providing information about the need for retreatment. 3,8,15,18-23
Well-defined classic CNV appears on OCT as hyperreflective areas in contact with or in front of the RPE; the pathology may be dome-shaped or appear as a thin formation (fusiform or nodular) (Figure 4). Retinal edema is always present to some extent in front of the active membrane; if the retina is thinner than normal, new vessels may be latent. CNV is less evident a few weeks after onset, and only interruption, breakdown, and pronounced thickening of the RPE can be seen.
Intraretinal fluid either can occur diffusely and cause increased retinal thickness and reduced retinal reflectivity or be localized or non-reflective cysts (cystic edema) (Figure 8).
RPE tears are serious complications of AMD and appear during the evolution of a serous or serohemorrhagic detachment of the epithelium. RPE tears also can be secondary to photodynamic therapy (PDT) and laser photocoagulation. Patients with occult CNV membranes and pigmented epithelium detachments are at high risk of developing pigmented epithelium tears. 6,8
Polypoidal choroidopathy is considered a variant of neovascular membranes with serous or hemorrhagic detachment of the retina and RPE in the posterior pole. The pathology is seen on FA and especially by videoangiography with green indocyanine green (ICG). Both tests show dilated vessels in the choroid layer with the characteristic presence of polypoidal structures which usually occur at the termini of vessels at the edge of the vascular network. (Figure 10).8,24-26
Disciform scars form the stage of exudative AMD in which the retinal tissue is replaced by scar tissue (usually vascular). Disciform scars develop with regression of subretinal hemorrhages and retinal edema and hyperplasic elements of the RPE. Disciform scars are usually dry and show marked destruction of all retinal layers. It is less common to find diffuse exudation with elevation of the entire sensory retina at the posterior pole (Figure 12).
Time-Domain vs. Spectral-Domain OCT in AMD
Correlation with Other Techniques in AMD
OCT retinal studies in patients with AMD can complement information obtained with other conventional examinations, such as angiography, or other more modern techniques such as autofluorescence. We discuss combining these techniques in recent studies.
OCT and Angiography
OCT and Autofluorescence
Clinical and Therapeutic Implications of OCT in AMD
OCT is now a routine examination for patients with AMD. Although some studies have advocated a role for OCT in the dry forms of the disease, its use is greater in patients with the neovascular forms, and it is now a main method of study in these patients. From the first studies of PDT, the different patterns of response to treatment were assessed with OCT, and the development of antiangiogenic therapies made OCT a fundamental tool in routine patient management. In recent years, OCT has become increasingly important, to the extent that recent studies have based the need for retreatment on the findings in OCT images.
The better definition achieved with the OCT devices that include SD technology allows more detailed study of the outer retinal layers and RPE, which are important structures in the development of dry AMD. Thus, new SD-OCT instruments can accurately distinguish the presence and size of drusen and RPE changes, making it possible to differentiate the different retinal layers to identify changes at that level.48 Thus, in a group of patients with dry AMD, Schuman and colleagues49 found localized thinning of the photoreceptor layer immediately above the drusen compared to healthy controls, suggesting a degenerative process with cell loss to explain the decreased visual function in this group (Figure 16).
Bearelly and colleagues17 studied the thinning at the edges of the plaques of GA to establish a gradient in the thickness of the photoreceptors layer from the healthy retina to the atrophic plaque. Although they had a small sample size in their study (n = 17), they concluded that SD-OCT allows quantitative measurement of disease progression and postponed for future study the application of the technique. Moreover, to study progression of the dry forms, other studies have been designed to correlate the findings with the SD-OCT scans with other techniques such as the autofluorescence mentioned previously. To this end, Stopa and colleagues47 analyzed a series of patients with dry AMD and correlated SD-OCT images of areas of GA, and isolated hard drusen and soft coalescing drusen with retinographies and images with autofluorescence. Thus, in addition to finding different patterns of reflectivity for each type of drusen, they observed that certain patterns of hyperreflectivity in some drusen and the overlying retina corresponded to increased or decreased autofluorescence at these points. This established a certain morphology-function relationship with both scans (Figure 17).
Finally, in some cases Stopa and colleagues identified small hyporeflective spots in retinas with soft coalescent drusen and drusenoid PEDs corresponding to subretinal fluid, which were not found in other examinations. In conclusion, it appears that the study of dry AMD need a combined approach with different and complementary techniques, with autofluorescence highlighted. OCT plays an important role both from a descriptive and clinical standpoint and provides anatomic information on retinal structures as a primary exploration and a complementary test.50
PrONTO Study (Prospective Optical Coherence Tomography Imaging of Patients with Neovascular Age-Related Macular Degeneration Treated with intraOcular Ranibizumab): The injection protocols established in multicenter phase III studies of ranibizumab for neovascular AMD, the ANCHOR and MARINA trials, consisted of monthly intravitreal injections for 2 years for all study eyes that resulted in substantial gains in VA in various intermediate controls and at the final examination.51-53 However, the dose of 24 injections carried a human and economic burden that made it difficult and impractical to administer. In phase I and II studies of ranibizumab before these studies, an extension study was conducted in which administration of new additional doses were left to the discretion of the investigator based on the presence of diffusion on FA or intraretinal or subretinal fluid on OCT. This subanalysis showed that often the presence of retinal fluid could be detected much earlier by OCT than FA, leaving the door open for a greater role of OCT in treatment.
Considering this, Fung and colleagues at the Bascom Palmer Eye Institute in Miami, Florida, proposed the PrONTO Study. In that study, after three monthly loading doses, monthly retreatments of ranibizumab were administered for 2 years according to the criteria in Table 3. Three criteria are based on OCT analysis (decreases exceeding five letters of VA [ETDRS] with fluid in the macula detected by OCT, increases exceeding 100 microns in central retinal thickness compared to previous lowest value measured by OCT or the recurrence of fluid on OCT in a previously dry eye) (Figure 18).54
The good results obtained with the varying numbers of injections based on OCT led to the inclusion of OCT as a primary criteria in new multicenter studies such as the SAILOR or SUSTAIN studies. Those studies included a larger number of patients than in the original early study (n = 40) and confirmed the role of OCT in patient monitoring.
OCT and Reviews: “Treat & Extend.”
In patients with wet AMD treated with antiangiogenic drugs, the frequency of the injection was not the only parameter assessed. The high number of regular examinations also is a substantial burden for patients, families, and the health system. Several studies (PIER, EXCITE) proposed revisions to the fixed protocols after 1 month, but the results were worse than those obtained with monthly revisions, with both fixed reinjection protocols and those based on OCT. 6,57 Therefore, it is necessary to establish specific criteria to control AMD and allow more time between retreatments without losing the treatment effectiveness. Once again, OCT had a fundamental role in this approach. In this sense, Spaide proposed a revision scheme called Treat and Extend; after three monthly loading injections, patients were reevaluated in 6 weeks, at which time the physicians determine whether there are signs of activity by funduscopy and edema or intraretinal or subretinal fluid by OCT (Figure 20).58
The results depend not only on treatment but also the frequency of subsequent examinations by the protocol presented in this proposal, which takes a leading role in OCT. The use of this treatment regimen is being evaluated in a study conducted at Wills Eye Institute, Philadelphia, Pennsylvania. The preliminary results are promising, with a lower frequency of examinations, i.e., an average of 7.4 injections annually and similar VA outcomes to those obtained in the PRONTO study in a greater number of patients (n=92).59
Prognostic Factors in OCT
Several OCT parameters are associated with decreased VA and different responses to other antiangiogenic treatments regarding VA, central thickness findings, and prognosis after treatment. Singh and colleagues retrospectively analyzed a group of patients with wet AMD treated with bevacizumab (Avastin, Genentech, south San Francisco, CA) and concluded that in addition to VA pretreatment the total retinal thickness measured by OCT affected the final VA and the total decrease in the post-treatment retinal thickness, which was lower in patients who had received previous treatments.60 Keane and colleagues also found that the main factor associated with decreased VA was the volume of subretinal tissue, and in fewer cases, thickening of the neurosensory retina, without a significant association with the total volume of subretinal fluid and RPE detachment and VA.61 However, these parameters did not justify the variability found in the VA for similar values of subretinal tissue volume or thickening of the neurosensory retina. The authors pointed to the complex pathophysiology of the neovascular membranes and the limitations of the TD-OCT used to explain the results. Sayanagi and colleagues reported that SD-OCT is a superior generation of TD-OCT than its predecessors for assessing the activity of the neovascular membranes and changes in AMD after ranibizumab treatment, and Kiss and colleagues pointed to the RPE status of neovascular membranes as the main predictor of VA in patients treated with ranibizumab in addition to the conventional parameters such as central retinal thickness.37-62 Several studies also reported a significant correlation between a hyperreflective band indicating the integrity of the junction between the inner and outer segments of the photoreceptors and higher VA in patients treated with ranibizumab.63-64 It appears that parallel to the development of new devices, structures assume special importance, such as the junction of the inner and outer segments of the photoreceptors, which until now was only identified, as shown by studies with prototypes with high-definition axial resolution of 3.5 microns as we will discuss later in this chapter.65-66 Lee and colleagues reported that the presence of posterior vitreomacular adhesion on OCT was associated with CNV in a large series of patients with AMD (n = 251). Those authors suggested that this finding is a possible risk factor for subretinal membrane development because of chronic vitreomacular traction on the retina, opening the door to a possible surgical approach in patients not responding or resistant to drug treatment (Figure 21).67-68
Ahlers and colleagues at the Medical University of Vienna led by Ursula Schmidt-Erfurth, MD, pointed to a new parameter as a prognosis factor in patients with AMD, i.e., the optical density ratio of the subretinal fluid detected by OCT.69 The authors suggested that this ratio may be an indirect way to measure the integrity of the barrier and therefore be useful for differential diagnosis between different exudative macular diseases such as central serous chorioretinopathy and to assess the response to antiangiogenic drugs (Table 4).
OCT and Wet AMD in Clinical Practice
To reach a consensus on the criteria of treatment and frequency of revisions, daily management of these patients is individualized, and the decisions are based on clinical examination and qualitative analysis of OCT images.2 The main signs of activity of the neovascular membranes in the OCT are the presence of intraretinal or subretinal fluid and RPE detachments and tears. These OCT findings should be evaluated biomicroscopically for the presence of fibrosis in disciform scars, which are final and irreversible stages of the disease that sometimes can be shown on OCT. The presence of any of these tomographic signs, the patient’s VA, and the ophthalmoscopic and angiographic appearance of the lesions should be evaluated by ophthalmologists to reach treatment decisions and revisions in each case until the current prospective multicenter studies shed light on results based on conclusive evidence.
New Techniques and Prototypes in AMD
OCT and Confocal Scanning Laser Ophthalmoscopy
Some new SD-OCT instruments (Spectralis HRA-OCT, Heidelberg Engineering, Inc., Heidelberg, Germany; Spectral OCT/SLO, Opko/OTI, Inc., Miami, FL) are equipped with this method, which simultaneously captures digital funduscopic images (infrared, red free autofluorescence, FA and ICGA) and OCT images from the same light source. This enables clinicians to correlate the two images pixel by pixel and allows a detailed topographic survey and tomographic retinal structures.70 In AMD, these devices are used especially in studies attempting to correlate OCT and autofluorescence in GA.71-72
Displaying the Choroid with Depth-Enhanced Optical Imaging OCT
The principle of OCT requires penetration of the emitted beam through the structures observed before being received for analysis. Thus, with the standard technique, images of the full-thickness retina to the RPE, Bruch’s membrane and choroid immediately adjacent space can be obtained, but the remainder of the choroid is beyond the capabilities of the instrument. To study this layer, Spaide developed a system consisting of an approximating scanner to the eye for multiple inverted images that then are processed to obtain a higher quality image in which the choroid can be seen in greater depth. Several diseases have been studied using this technique. This technique may be relevant for studying wet AMD associated with RPE detachment and especially in occult CNV, which often cannot be visualized with conventional OCT. 40,41,73-76
New Prototypes: High-Speed Resolution Ultrahigh OCT
SD instruments reach axial resolutions close to 5 microns using a source that emits light at a certain wavelength and detection system based on the Fourier principle. To obtain even higher resolutions, several groups have designed devices that vary primarily based on the light source. A group at Tufts Medical Center, Boston, developed a prototype that uses a broadband laser superluminescent diode combined with a Fourier detection system that can achieve axial resolutions up to 3.5 microns and allows precise analysis of the outer retinal layers, such as the photoreceptors layer, the RPE, and Bruch’s membrane.77 The investigators have published several studies highlighting the importance of these strata in the pathogenesis of dry AMD, the ability for early detection of intraretinal and subretinal fluid, and the status of the photoreceptors before and after ranibizumab treatment in a series of patients with wet AMD.4,66,78,79
RPE Study of Polarization-Sensitive OCT
SD instruments identify external retinal structures but provide limited information on the region of the RPE, making it possible to assess only the status based on hyperreflectivity. To explore differences in the reflectivity of the RPE in the various AMD subtypes, an OCT prototype has been designed that incorporates a receiver capable of obtaining measurements of intensity and polarization at each point of study. In a study conducted in 44 patients with drusen, GA, disciform scarring, and neovascular forms, the authors described the different patterns of involvement of the RPE and suggested future uses of the device to assess the status, progression, and even disease prognosis.80
Last revision: October 2011 by Alfredo Garcia-Layana