Abstract
The choroid is the middle layer of the eye, a very vascular and pigmented tissue whose role in several ophthalmological pathologies has been clearly established already. However, it was not until the last few years that we have been able to reliably measure and quantify its shape and thickness in a precise manner.
Ultrasound technology and indocyanine green angiography were the first techniques used for the study of the choroid, and they still maintain its use and clinical indications for the diagnosis and management of several pathologies. But it was the advent of optical coherence tomography (OCT) that meant the greatest breakthrough in choroidal imaging.
In this chapter, we will discuss past, current, and future image modalities for the study of the choroid, with a special focus on OCT and its latest developments.
Keywords
Choroidal imaging techniques, Optical coherence tomography, OCT, ultrasonography, Angiography, En face, OCT angiography
Introduction
The choroid is a highly vascular, pigmented tissue located between the retina and the sclera. The term is derived from the Greek words “form” and “membrane.” It measures 0.22 mm at the posterior pole and from 0.1 to 0.15 mm in its most anterior aspect in a postmortem pathologic sample study. Blood enters the choroid through the short posterior ciliary arteries and is distributed in three layers: Haller’s, Sattler’s, and choriocapillaris. Arterial blood flows consecutively through each of these layers and is then collected by venules. These venules converge in ampoules, which form the vorticose veins, and leave the eye through its equator. The choroid is responsible for the oxygenation and nourishment of the outer retina.
Few years ago, ultrasound and indocyanine green (ICG) angiography were the only imaging techniques available to assess the choroid. However, the understanding of the choroid has rocketed during the past decade followed by several advances in imaging technology that have granted it faster and easier visualization and measurement.
Ultrasonography
In 1956, ultrasonic techniques were first used for the diagnosis of ocular diseases. Some used the so-called intensity modulation technique (B-scan) that required the immersion of the eyes in water, whereas an acoustic tomogram of the eye could be obtained by a scanning movement of a crystal in front of the eye. Some others used time–amplitude methods (A-scan), in which the axis x of the screen forms the time axis, and axis y forms the amplitude axis of the echo.
Although the normal choroid could not be measured, and some authors stated that lesions under 4.0 mm in height could not be fully evaluated, choroidal melanomas and detachments could be found in cases they penetrated into the vitreous cavity for at least 1.5 to 2.0 mm. Subretinal coagulated blood presented a difficult differential diagnosis at that time.
Contact US scanners were introduced in the 1970s, with continuous evolutions and sensitivity improvements. During the last two decades, the revolution of digital format technology brought changes in examination technique, storage of data, and further improvements in image quality ( Fig. 4.1 ).
Choroidal melanoma is the most frequent intraocular malignant tumor, and despite the appearance of new imaging techniques, ultrasound is still of great use. Before its advent, melanomas were only suspected when a visible mass could be seen through clear media, and even in cases where a mass is visible, diagnosis may not be so clear. But it is in case of opaque media when ultrasound is of greatest use. Ultrasound patterns of choroidal masses are still critical to establish a good differential diagnosis. Low resolution, unavailability of quantitative data, and more importantly, poor reproducibility and operator dependency are major limitations of ultrasound.
Angiography
Fluorescein angiography (FA) and ICG angiography have been performed for decades to obtain useful clinical information about the retina and the choroid.
FA was developed in the 1960s for the study of choroidal tumors, and was mainly used to study retinal vasculature, so some of the first authors to use this dye studied choroidal circulation during the earliest phases of the angiogram or through areas of retinal atrophy, which made choroid vessels more visible and easier to distinguish.
Meanwhile, ICG was the first dye used in photographic industry and was first applied for clinical purposes in 1972, when Flower et al . tried to image and describe choroidal vasculature. Indocyanine is a lipophilic and hydrophilic substance with high protein-binding properties (up to 98%). These show greater molecular weight than albumin, which grants indocyanine a lower vascular permeability and tissue penetrance. This differentiates it from fluorescein and allows us for a better study of choroid vasculature. It is metabolized by the liver and suffers biliary excretion.
It is injected intravenously using concentrations of 5 mg/mL and in order to capture its circulation both an excitation and a barrier filter with peaks of 805 and 835 nm, respectively, are necessary. Pictures are usually taken from 8–10 up to 40 min after the injection of the dye. Later studies suggested that earlier times of the ICG could be useful to locate feeder vessels of choroidal neovascularization (CNV) complexes, and so, help focal treatment guidance.
It was, in fact, age-related macular degeneration (AMD) CNV that centered the vast majority of the studies performed using ICG, but has also been helpful for the investigations on the physiology of certain chorioretinal inflammatory disorders, as well as anterior segment disorders.
The choroidal vasculature is best demonstrated on ICG fluorescence angiograms and is not limited to cases of diseased pigment epithelium and choriocapillaris as in the case of FA ( Fig. 4.2 ). Given the choroid is a tridimensional tissue and the fact that the images that are obtained are displayed in a two dimensions, it is mandatory to have a proper knowledge of choroid anatomy in order to perform an accurate interpretation of the angiogram. Normal anatomic variations of blood drainage show an asymmetric pattern in up to 50% of patients, with preference for one of the vorticose veins, which may lead to misinterpretations.
ICG proves itself especially useful for the detection of recurrent CNV in difficult cases, such as pigment epithelial detachments (PED) or areas adjacent to laser scars. It is more valuable than FA for the location of CNV beneath subretinal or subRPE hemorrhages because of the greater penetration properties that infrared light grants. ICG angiography has become the most useful tool for the detection macular, extramacular, or peripapillary polyps. It can also be useful to identify the feeder vessel to a CNV or choroidal leakage/dilated choroidal vessels in CSC patients.
Optical Coherence Tomography
The appearance of optical coherence tomography (OCT) and its development represents a clear progress in choroidal imaging as it provides deeper, higher resolution imaging of the eye layers with brief acquisition times. This technology started its development in the 1980s, the first images being similar to ecography A-scans. Crosssectional, two-dimensional images were developed in 1992, with first in-vivo human retinal scans taking place in 1993.
Time domain OCT (TD-OCT) was initially available to study the posterior segment, but because of its poor penetration below the retinal pigment epithelium (RPE) and low resolution, it could not be used for choroidal imaging purposes. In 2006, spectral domain OCT (SD-OCT) became commercially available with several improvements such as the possibility to combine multiple images of one same spot to reduce noise and enhance the final result, or “eye tracking” software, which help us make sure that images are obtained in the same exact location. However, despite its obvious advantages over TD-OCT, signal roll-off with depth and signal attenuation by pigmented tissues or media opacities still made it difficult to obtain good quality choroidal images in most eyes. Then, Spaide et al . applied a new technique called enhanced depth imaging OCT (EDI-OCT) technology, which provided consistent choroidal visualization in most eyes and allowed reproducible thickness measurements. SD-OCT showed some problems related to wavelength-dependent light scattering and a clear decay in resolution and sensitivity with increasing displacement from zero-delay. Spaide proposed a technique that consisted of using the SD-OCT closer to the eye such that an inverted image is obtained, bringing deeper structures closer to the zero-delay line. Images obtained this way allowed for a better ability to visualize the choroid. The most recent technology for OCT imaging is longer-wavelength, high-penetration, and swept-source OCT (SS-OCT). This device employs a 1050 nm wavelength tunable laser as a light source operated at 100,000 Hz. It can perform image averaging of up to 96 B-scans at each location. For the study of the choroid, the reference mirror is placed at the deeper position of the retina, so that the sensitivity was higher in the choroid. Its one-line scanning mode can produce OCT images containing 1,024 axial scans with a scan length of up to 12 mm. This sampling space in object space corresponds to 11.7 μm/pixel. According to reports, choroidal thickness (CT) can be reliably measured with this device in up to 100% of patients ( Fig. 4.3 ).
Retinal vessels appear hyperreflective in OCT images, whereas choroidal vessels are hyporeflective. This difference seems to be related to the velocity of blood cells through them. The blood flow velocity inside choroidal vessels is much higher than that of retinal vessels. OCT imaging is based on interferometry, where an interference fringe is detected to construct the intensity of the signal. Fast speed of the blood cells makes the interference fringe vanish; thus, no signal is observed inside the choroidal vessels. On the other hand, blood cell speed is relatively slow in the case of retinal vascular structures, which contributes to the interference signal as a hyperreflective signal.
The role of the choroid in several pathologies has already been proved. They include central serous chorioretinopathy, AMD, and polypoidal choroidal vasculopathy, high myopia, posterior uveitis, and choroidal tumors among others. If choroidal variations play a role in retinal diseases, the normal CT profile and its morphology must be known, so as to be able to point out variations as they appear.
Choroidal Thickness
Age dependant thinning is a key factor to establish the role of the choroid in retinal conditions, as has been shown in several studies. Margolis et al . studied a group of 54 eyes of 30 52.4-year-old subjects with a mean SE of −1.3 D finding a mean subfoveal CT (SFCT) of 287 μm. These results are similar to Flores-Moreno’s and Manjunath et al .’s who published a mean SFCT of 292 in 96 eyes of sixty-two 52.6-year-old patients and 272 μm in 34 eyes of 51.1-year-old patients, respectively. Chhablani et al . analyzed 211 eyes from 115 healthy 42.8-year-old Indian subjects and found them to have a mean central macular thickness of 216.4 μm, which was thinner than in other reports. Xu reported a mean SFCT of 266 μm in the nondiabetic patient group of his study (1795 subjects). Pediatric choroid has also been studied by SS-OCT and found to be thicker than that of adult patients.
Our group has conducted several studies on this subject using SS-OCT technology, manually measuring the CT of 276 eyes from 154 healthy patients ranging from 3 to 95 years of age. Both MCT and SFCT showed high correlation with age, and both decrease as patients grow older, showing statistically significant differences between age groups. This reduction was found to be progressive until 40 years of age, at which point the most significant variation takes place. No statistically significant differences were found between men and women. Following linear regression models, a mean reduction of 10 μm in MCT per decade (one every year) and 8.87 μm of SFCT per decade can be predicted.
This line of investigation came up with the fact that both eyes are not exactly symmetrical. Mean nasal CT was statistically thicker in the right eye than in the left eye (228 vs 212 μm, p =0.0002). Chen et al . also found a trend toward a thicker nasal CT (14 μm) in right eyes. It is difficult to explain the strong and consistent evidence that right eyes have thicker nasal choroid than left eyes in a healthy population. One possibility is a differential in blood flow between the two eyes due to lack of anatomic symmetry at the aortic arch. Such asymmetry has been suggested to explain the differences in the incidence and prevalence of vascular pathologies between right and left eyes with respect to metastatic bacterial endophthalmitis and retinal artery occlusion. Choroidal circulation is generated from the short posterior ciliary arteries that penetrate through the sclera around the optical nerve in a variable number between 10 and 20. Therefore, the nasal choroid (choroid between fovea and optical nerve) is directly supplied from these short posterior ciliary arteries. Furthermore, short posterior ciliary arteries are branches of the ophthalmic artery which is a branch of the internal carotid artery and the common carotid artery. The right common carotid artery is a branch of the brachiocephalic trunk, whereas the left common carotid artery emerges directly from the aorta, and may presumably be responsible of a more proximal and direct blood flow to the right carotid artery. Given that, most of the choroidal structure is vascular tissue (the mean vessel density in outer choroid is 87%), a supposed higher blood flow in the right versus left short posterior ciliary arteries may explain why the nasal choroid is thicker in the right eye, as stated in our study, and previously reported by Chen et al .
Choroidal Morphology
The horizontal macular CT profile has already been described in healthy people. The choroid is normally thickest in the subfoveal region, decreasing slowly and progressively toward the temporal aspect and more steeply toward the optic nerve, with a bowl-shaped contour. Adhi et al . described significant alterations in choroidal morphological features in most of eyes with diabetic retinopathy (DR). They found an irregular temporal choroidoscleral interface inflection point in 8 of 9 eyes with nonproliferative DR (89%), 9 of 10 eyes with proliferative DR (90%), and in 13 of 14 eyes with diabetic macular edema (93%) compared with 0 of 24 in controls. The presence of an irregular temporal choroidoscleral interface inflection contour with focal thinning of the choroid has been detected in DR and AMD. Dolz-Marco et al . reported a bilateral case of focal inferotemporal scleral bulge with a concave choroidal thinning surrounded by normal choroidal tissue. The authors hypothesized that this finding could be related to the inferior oblique muscle leading to inward compression of the choroid, not being able to study the scleral wall, as it was not entirely visible.
Our group performed a study on 276 eyes from 154 patients. A concave or bowl-shaped contour of the horizontal choroidoscleral interface was found in 87.2% of the eyes analyzed and a temporal choroidoscleral interface inflection was identified in 12.8% of the eyes ( Fig. 4.4 ). Mean age of the patients with temporal choroidoscleral interface inflection was 16±19 years versus 36±25 years in the group with bowl-shaped contour ( p =0.001; student’s t test for unpaired data).
