Zweifel et al. characterized outer retinal tubulation (ORT) based on OCT. ORTs are branching tubular structures found in the outer nuclear layer (Fig. 15.5); however, ORT are not pathognomonic for nAMD (Zweifel et al. 2009). ORTs are present in other eye diseases as well, such as retinitis pigmentosa or in association with angioid streaks (Ellabban et al. 2012; Tulvatana et al. 1999). Histopathological analysis shows that ORTs largely consist of cones in various phases of degeneration lacking outer segments and inner segments, morphologically altered mitochondria, and external limiting membrane delineating the luminal wall (Schaal et al. 2015). On SD-OCT, ORT has a tubelike appearance with hyporeflective centers and hyperreflective borders. En face images reveal branching networks emanating from neovascular lesions (Wolff et al. 2012). When reviewing OCT scans, care must be taken to distinguish ORT from intraretinal cysts or subretinal fluid without a hyperreflective border. ORTs are commonly present outside the foveal central 1-mm area, but progress centrally in the course of the disease (Lee et al. 2014). A recent study based on the CATT (comparison of anti-VEGF treatment trails) population suggests that ORTs are an important finding on OCT as eyes with ORT have significantly worse visual acuity (VA) at baseline and after 2 years than eyes without ORT change (Lee et al. 2014). VA outcomes were consistent with previous findings in which ORT had been associated with significantly worse VA; however, under anti-VEGF treatment, an increase in VA was still recorded (Faria-Correia et al. 2013).

Type 1 neovascularization corresponds mainly to the occult form of CNV and is defined as hyperreflective areas posterior of the RPE in SD-OCT, which is frequently associated with PED. The 3D en face map of the RPE layer reveals a dome-shaped area with an area of fibrovascular material (Malamos et al. 2009). Enhanced depth imaging (EDI) helps to visualize and localize the content of the whole fibrovascular PED. The en face projection of the hyperreflective material perfectly matches the hyper-fluorescent neovascularization in ICG angiography (Coscas et al. 2012).

Often type 1 CNV lesions are clinically silent for a long time because the overlaying neurosensory retina remains dry and the retinal architecture is maintained. Clinically manifest type 1 CNV lesions are characterized by exudative features on OCT, where mostly subretinal fluid can be seen. Subretinal fluid on OCT is a clear sign to initiate anti-VEGF therapy. With disease progression, disruption of the blood-retinal barrier and alterations of the RPE layer occur, and consequently intraretinal cystoid spaces appear on OCT.

Today, there is widespread consensus that polypoidal choroidal vasculopathy is a subtype of type 1 nAMD. But whether or not it could be a distinct entity of the choroidal vasculature is still discussed. It has a high prevalence in Asian patients and is more frequent in Asian male than female patients as opposed to Caucasian people where it occurs more often in female than in male patients (Imamura et al. 2010). Clinical signs of polypoidal lesions are very diverse, which is why initial diagnosis is confirmed by ICG angiography, which identifies polypoidal dilatations and branching networks of choroidal vessels.

On SD-OCT, PCV neovascularizations appear as sub-RPE lesions, with multiple large and small PEDs attaching to the outer surface of the RPE, as well as branching communications. With SS-OCT, polypoidal structures were identified between the RPE and Bruch’s membrane. Branching vascular networks above and below Bruch’s membrane as well as choroidal vascular abnormalities were found. En face OCT maps located under the RPE visualize the entire topography of polypoidal lesions best. This has been shown in SD and SS-OCT systems (Alasil et al. 2015; Imamura et al. 2010; Saito et al. 2008; Sayanagi et al. 2015), the latter systems being superior. En face view may also reveal the “hematocrit sign,” describing a lesion where blood is presumed to be separated into corpuscular and serous components (Imamura et al. 2010).

In type 2 CNV, the RPE layer is penetrated by the neovascular tissue, which grows within the subretinal space. As the origin of the subretinal neovascular complex lies within the choroid, type 2 CNV is connected to a type 1 choroidal lesion. Type 2 neovascular membranes are identified as hyperreflective and thickened areas anterior to the RPE in SD-OCT. Because of the vessel leakage, intraretinal fluid in the form of cysts is one of the first signs visible on OCT. SD-OCT has been shown to be the method of choice in type 2 CNV: FA and ICG angiography underestimate the size of the neovascular complex as the area of the associated edema cannot be delineated in FA and ICG angiography (Sulzbacher et al. 2011). Type 2 lesions usually improve in tissue morphology with anti-VEGF therapy by regress in retinal thickness, RPE regrowth, and ellipsoid zone/interdigitation zone rearrangement, indicating improved tissue function. CNV diameters, however, remained stable, which is particularly true for the occult components of the CNV lesion (Framme et al. 2010; You et al. 2012).

The type 3 AMD subtype is also known as retinal angiomatous proliferation (RAP) and is characterized by neovascularization, presumed to originate from the retina (Sayanagi et al. 2014). Type 3 was first described in 1996 as a deep retinal anomalous complex with advanced Bruch’s membrane changes and severe visual loss (Hartnett et al. 1996) and is considered the second most frequent form of AMD (Jung et al. 2014) with varying numbers reported.

Today we know that depending on the stage of the disease, intraretinal neovascularizations are not limited anterior of the RPE, as a retinochoroidal anastomosis is present on SD-OCT in the late stage of the disease (Yannuzzi et al. 2001). Also IRC in combination with PED are a common characteristic (Rouvas et al. 2010; Schmidt-Erfurth et al. 2014). Outer retinal layer disorganization becomes evident on OCT with progression of the disease (Querques et al. 2013).

CRT was first used in the ANCHOR and MARINA trials and served as a secondary outcome measure (Brown et al. 2006; Rosenfeld et al. 2006). Both studies employed a fixed treatment regimen with monthly intravitreal injections. CRT change became the most frequent criteria for retreatment decisions when nAMD treatment was individualized applying PRN (pro re nata, treatment as needed) treatment based on predefined retreatment criteria (Fung et al. 2007; Holz et al. 2011; Schmidt-Erfurth et al. 2011). The Comparison of Age-Related Macular Degeneration Trials (CATT) later suggested that retreatment should be based on a zero-tolerance strategy of fluid seen on OCT by eradication as early as possible (Martin et al. 2012). Switching from low-resolution TD-OCT to high-resolution SD-OCT by eradication led to more retreatment numbers being reported (Busbee et al. 2013) because the high resolution and large number of scans possible with SD-OCT revealed small occurrences of fluid. Major et al. affirmed the superiority of SD-OCT over TD-OCT as it revealed more exudative disease activity (Major et al. 2014).

Today CRT-based PRN treatment is still the most commonly used regimen in Europe (Schmidt-Erfurth et al. 2014). Therefore, it is not surprising that all recent OCT devices provide false-color-coded retinal thickness maps that give a first impression of macular topography (Fig. 15.7). The maps are usually divided into nine subfields (Early-Treatment-Diabetic-Retinopathy-Study-Group 1985); the center field equates to the CRT, which is defined as the average retinal thickness within a 1-mm circular area centered on the fovea. Great attention should be paid to OCT B scans as increases in retinal thickness comprise a multitude of pathological retinal changes.

Despite its popularity, treatment based solely on CRT is already outdated. CRT only correlates with visual function in treatment-naïve patients and during the loading phase (first 3 months with monthly intravitreal anti-VEGF injections) (Bolz et al. 2010). Simader et al. were able to show that the change in CRT from baseline does not correlate with best-corrected visual acuity (BCVA) after a 3-month loading interval (Simader et al. 2014).

Various pharmaceutical trials have considered the reduction in CRT a therapeutic success, although it is known that CRT is not necessarily associated with improved retinal function (Chakravarthy et al. 2012; Cohen et al. 2013; Martin et al. 2011; Roberts et al. 2014). The mismatch between CRT and BCVA emphasizes the need for OCT biomarkers that better indicate a functional benefit from nAMD therapy.

The integrity of the external limiting membrane (ELM) and the ellipsoid zone as assessed by OCT correlates well with visual acuity in patients with nAMD. Müller cells and photoreceptors are assumed to connect at the ELM layer. Thus, ELM together with ellipsoid zone is considered a criterion that directly reflects photoreceptor function. However, ELM is no predictor for individual loss or recovery in BCVA, but rather mirrors the current functional state of the retina (Roberts et al. 2014). It has to be kept in mind that ELM integrity is not pathognomonic for nAMD. A correlation with visual acuity has also been found in retinal detachment, retinal vein occlusion, and macular hole (Chhablani et al. 2013; Landa et al. 2012; Wakabayashi et al. 2009; Wolf-Schnurrbusch et al. 2011). A small body of evidence suggests that the ellipsoid zone status in nAMD is a good indicator for BCVA after anti-VEGF injections (Sayanagi et al. 2009; Ueda-Arakawa et al. 2012). However, adjacent lesions make it very difficult to delineate the state of the photoreceptor layers in OCT, for example, shadowing or signal transmission occurs due to the changed retinal architecture in disease.

The optical density ratio (ODR) might be a valuable biomarker in nAMD as it correlates well with BCVA under anti-VEGF therapy and may be useful for differentiation as well as prognosis (Ahlers et al. 2009). As described previously, the ODR compares the optical density of fluid accumulation in or under the retina to the optical density of the vitreous body. Optical density ratios change in the course of the disease because the blood-retinal barrier regains function under anti-VEGF therapy and prevents the CNV from leaking. A high optical density signal indicates increased reflectivity of the fluid accumulation, which is assumed to be caused by the protein concentration in the SRF (Baek and Park 2015) and is therefore thought to be a direct indicator for the blood-retinal barrier function (Ahlers et al. 2009; Baek and Park 2015; Neudorfer et al. 2012). Further, Ahlers et al. showed that ODR changes correlate well with BCVA changes under anti-VEGF therapy (Ahlers et al. 2009).

We have mentioned outer retinal tabulation previously as findings of the intraretinal space, containing degenerated photoreceptors, almost exclusively cones, and Müller cells (Schaal et al. 2015). ORT may be misdiagnosed as new fluid in the retina which lacks the hyperreflective border. The prevalence of ORT in the first year of anti-VEGF treatment is around 17 % and increases to over 40 % after 4 years (Dirani et al. 2015). ORT may be considered a valuable OCT biomarker for future clinical trials as BCVA and retinal sensitivity (microperimetry) were shown to be significantly worse in nAMD eyes with ORT than ORT-free eyes, both at baseline and after 1–2 years’ follow-up (Faria-Correia et al. 2013; Lee et al. 2014; Dirani et al. 2015; Iaculli et al. 2015).

So far, most clinical trials investigating flexible PRN treatment have shown that PRN is inferior to strict monthly injections. The first clinical anti-VEGF trials using various treatment regimen have given insight into detailed morphologic changes in sub-/intraretinal fluid compartments and their correlation with functional outcomes such as BCVA.

Three pathologic changes affecting central retinal morphology have been described in nAMD patients; intraretinal cystoid fluid (IRC, Fig. 15.4), subretinal fluid (SRF, Fig. 15.2a), and pigment epithelial detachment (PED, Fig. 15.1) (Jaffe et al. 2013; Schmidt-Erfurth et al. 2015; Simader et al. 2014). A comprehensive subgroup analysis of the VIEW II data (Schmidt-Erfurth et al. 2014) showed a clear association between the PRN regime and visual acuity loss in patients presenting with primary PED at baseline (Schmidt-Erfurth et al. 2015). A significant drop in BCVA was found in this subgroup contrasting with the loss of one or two letters after gross averaging of the whole study population. After stratification for retinal morphologies, it became evident that all patients other than those with initial PED have constant BCVA readings. This finding exemplifies the need for solid OCT biomarkers to provide the best individualized disease management.

Under anti-VEGF therapy, IRC and SRF were resorbed in 70–80 % of patients, whereas PED vanished in only 50 %. Patients with SRF only had good BCVA at baseline followed by improvement under therapy. Patients with IRC with or without PED had poor BCVA at baseline and did not gain as many letters as the first group under anti-VEGF therapy. However, PED did regress under continuous therapy, but switching from fixed to flexible PRN treatment led to a reactivation of this subretinal lesion. We suggest that the elevation of the retinal pigment epithelium (RPE) is the morphologic correlate for nAMD activity, and the associated loss in visual acuity caused by secondary degenerative intraretinal cystoid fluid formation is the functional manifestation. This is in good agreement with the finding that the occurrence of IRC is correlated with a loss in BCVA in the eyes with PED (Schmidt-Erfurth et al. 2015). The eyes with subretinal fluid (SRF) had the best visual prognosis. IRC and fibrovascular PED at baseline influenced BCVA prognosis negatively (Ritter et al. 2014; Schmidt-Erfurth et al. 2015).

These results clearly contrast with the current convention that PRN is inferior to a fix OCT-based therapy regimen, highlighting the need to consider morphologic OCT variables in the future. However, final conclusions regarding retinal morphologic retreatment criteria are still lacking, and further studies are needed to identify or confirm OCT biomarkers. As for now, expert guidelines recommend that IRC, SRF, and PED should be incorporated into treatment protocols (Schmidt-Erfurth et al. 2014). Monthly monitoring is advised using SD-OCT because of its ability to visualize even small morphologic changes, which is essential when monitoring a disease with the potential to progress rapidly leading to irreversible loss of vision if treatment is deferred.

Doppler OCT obtains high-resolution OCT images and information about functional blood flow and velocity (Wang et al. 1995; Werkmeister et al. 2008). Wang et al. were the first to report total retinal blood flow (Wang et al. 2007), and other groups followed (Baumann et al. 2011; Blatter et al. 2013; Dai et al. 2013; Doblhoff-Dier et al. 2014). Various eye diseases such as age-related macular degeneration (Ehrlich et al. 2009; Pemp and Schmetterer 2008), diabetic retinopathy (Clermont and Bursell 2007; Pemp and Schmetterer 2008; Wang et al. 2011b), and glaucoma (Costa et al. 2014; Schmidl et al. 2011) were shown to be linked to early changes in blood flow. Early changes in blood flow regulation can be detected by using flicker light stimulation. The response to flicker light was altered in nAMD and normalized after anti-VEGF treatment (Lanzl et al. 2011). Today, flickering light stimulation is incorporated into Doppler OCT setups (Clermont and Bursell 2007; Leitgeb 2007; Pemp and Schmetterer 2008; Wang et al. 2011a, b), with the aim of early diagnosing ocular diseases such as nAMD.

From a clinical point of view, visualization of the retinal and choroidal vasculature is the most advanced application of Doppler OCT. The optical Doppler effect creates motion-related image contrast in vessels, which can be extracted by software-based post-processing (Schwartz et al. 2014). This is why this technique is also called optical coherence angiography or OCTA (Makita et al. 2006). OCTA has already been commercialized based on an SD-OCT system (e.g. Angiovue, Optovue, Fremont, CA, USA), allowing fast and detailed visualization of the retinal and choroidal microcirculation (Fig. 15.8). With improvements in the technology, it has become possible to distinguish CNV from the surrounding retinal or choroidal vasculature in en face (coronal) projections as well as to quantify the CNV area (Flores-Moreno et al. 2015; Jia et al. 2012; Moult et al. 2014) and visualize CNV vessel architecture in the course of anti-VEGF treatment. This treatment results in abnormal vessel configurations, which can possibly be explained by treatment-related pruning of vessel sprouts (Spaide 2015). So far studies using OCTA have mostly had a descriptive character, but lately obtaining quantitative information such as vessel density, flow index, and areas of nonperfusion (Jia et al. 2015) has also become possible, adding valuable information, for example, in nAMD follow-up. Also, a recent study investigating CNV of various origins concluded that OCTA is ideal for identifying neovascularization and adjacent fluid noninvasively and without the use of contrast agents (de Carlo et al. 2015). The same conclusions were drawn in a study investigating the value of OCTA for early identification and treatment response of type 3 neovascularization (RAP) (Dansingani et al. 2015). The main advantage of OCTA is that it is label-free, given the risks associated with the dye fluorescein (López-Sáez et al. 1998), the current gold standard for nAMD diagnosis (Schmidt-Erfurth et al. 2014a, b; AAO 2015; Chakravarthy et al. 2013). Future nAMD treatment will probably be guided by the presence or absence of flow-related contrast or areas of non- or hyperperfusion on OCTA, which could indicate CNV activity.