Fluorescein angiography (FA) allows study of the circulation of the retina and choroid in normal and diseased states. Photographs of the retina are taken after intravenous injection of sodium fluorescein, an orange-red crystalline hydrocarbon with a molecular weight of 376 daltons that diffuses through most of the body fluids. It is available as 2–3 mL of 25% concentration or 5 mL of 10% concentration in a sterile aqueous solution. It is eliminated primarily through the liver and kidneys within 24–36 hours via the urine. Eighty percent of the fluorescein is protein-bound, primarily to albumin, and not available for fluorescence; the remaining 20% is unbound and circulates in the vasculature and tissues of the retina and choroid, where it can be visualized.
Fluorescence occurs when a molecule is excited by light of a certain wavelength that raises the molecule to a higher energy state and then allows it to release a photon of light to bring it back to its original state. To image this fluorescence, special excitation and barrier filters are required. Sodium fluorescein fluoresces at a wavelength of 520–530 nm (green) after excitation by a light of 465–490 nm (blue). To obtain a fluorescein angiogram, white light from the camera flash unit is passed through a blue (excitatory) filter, and blue light enters the eye. The blue light, with its wavelength of 465–490 nm, excites the unbound fluorescein molecules circulating in the retinal and choroidal layers or that have leaked out of the vasculature and stimulates them to emit a longer-wavelength yellow-green light (520–530 nm). Both the emitted yellow-green fluorescence and some degree of reflected blue light from structures that do not contain fluorescein exit the eye and return to the camera. A yellow-green (barrier) filter on the camera lens blocks the reflected blue light, permitting only the yellow-green light, which has originated from the fluorescein molecules, into the camera.
The image formed by the emitted fluorescence is recorded on either black-and-white, high-contrast 35-mm film, videotape, or a digital camera. The 35-mm film permits higher-resolution images of the retinal vessels and choroid and is generally easier to use than videotape for capturing stereoscopic frames and stereoscopic viewing. Newer digital systems offer high-resolution images rivaling those of 35-mm film and can adjust contrast and brightness to highlight certain details; they can also zoom in on areas of concern, which is not possible with film-based images. Digital images are seen immediately by the photographer, who can adjust for focus and problems during the procedure. This is not possible with film. Finally, digital imaging systems allow easy image archiving and retrieval, thus offering the capability of quickly comparing images over time for diagnosis and treatment.
To properly interpret a fluorescein angiogram, it is vital to understand retinal anatomy. The retina has a dual blood supply. The central retinal artery and retinal circulation serve the inner half of the retina, and the endothelial cell tight junctions provide the inner blood–retina barrier. Normally, neither bound nor unbound fluorescein can pass through this barrier. The choroidal circulation serves the outer half of the retina, and the RPE provides the outer blood–retina barrier. Fluorescein particles that are not bound to protein can pass through the fenestrated walls of the choriocapillaris but do not normally pass through the RPE or zonulae occludentes between adjacent RPE cells to gain access into the subretinal space. Therefore, fluorescein from the choroid cannot enter the neurosensory retina unless the RPE has a defect. Although the fluorescence in the choroid is partially blocked by the pigment in the RPE, it is visible as deep, diffuse background fluorescence.
Fluorescein is injected into a peripheral vein and enters the ocular circulation via the ophthalmic artery 8–12 seconds later, depending on the rate of injection and the patient’s age and cardiovascular status. The retinal and choroidal vessels fill during the transit phase, which ranges from 10 to 15 seconds. Choroidal filling is characterized by a patchy choroidal flush, with the lobules often visible. Because the retinal circulation has a longer anatomical course, these vessels fill after the choroidal circulation. The arterial phase of the angiogram occurs after the choroidal phase, with filling of the retinal arteries. The arteriovenous phase begins with complete filling of the retinal arteries and capillaries and completes with laminar filling of the retinal veins. This phase, which usually occurs approximately 1 minute after dye injection, is considered the peak phase of fluorescence, where the most detail is evident in the fovea. Over the next few minutes, the dye recirculates, with a gradual decline in fluorescence. In the late phases of the angiogram, the choroid, Bruch’s membrane, and the sclera stain. The larger choroidal vessels are often seen as hypofluorescent areas against this hyperfluorescent background.
Fluorescein can leak out of retinal capillaries into the retina only when the capillary endothelium is damaged, as in diabetic retinopathy. Similarly, fluorescein can leak from the choriocapillaris through pigment epithelial cells into the subretinal space and the retinal interstitium only when the latter are abnormal, as in central serous chorioretinopathy. Thus, patterns of hyperfluorescence and stereoscopic images yield valuable information about leakage from retinal vessels or through abnormal pigment epithelium. Abnormalities seen with FA can be grouped into 3 categories, associated with one of the following:
Autofluorescence is fluorescence that can be seen before the fluorescein dye is injected; this is caused by naturally highly reflective substances, such as optic disc drusen. Hypofluorescence occurs when normal fluorescence is reduced or absent; it is present in 2 major patterns:
vascular filling defect
Vascular filling defects occur where the retinal or choroidal vessels do not fill properly, as in nonperfusion of an artery, vein, or capillary in the retina or choroid. These defects produce either a delay or a complete absence in filling of the involved vessels.
Blocked fluorescence occurs when the stimulation or visualization of the fluorescein is blocked by fibrous tissue or another barrier, such as pigment or blood, producing an absence of normal retinal or choroidal fluorescence in the area.
Blocked fluorescence is most easily differentiated from hypofluorescence due to hypoperfusion by evaluating the ophthalmoscopic view, where a lesion is usually visible that corresponds to the area of blocked fluorescence. If no corresponding area is visible clinically, then it is likely an area of vascular filling defect and not blocked fluorescence. By evaluating the level of the blocked fluorescence in relation to the retinal circulation, one can determine how deep the lesion resides. For example, when lesions block the choroidal circulation but retinal vessels are present over this blocking defect, then the lesions are above the choroid and below the retinal vessels.
Hyperfluorescence occurs when there is an excess of normal fluorescence; it is seen in several major patterns:
transmission, or window, defect
Leakage refers to the gradual, marked increase in fluorescence throughout the angiogram when fluorescein molecules seep through the pigment epithelium into the subretinal space or neurosensory retina, out of retinal blood vessels into the retinal interstitium, or from retinal neovascularization into the vitreous. The borders of hyperfluorescence become increasingly blurred, and the greatest intensity of hyperfluorescence is found in the late phases of the study, when the only significant fluorescein dye remaining in the eye is extravascular. Leakage occurs, for example, in CNV (Fig 2-1), in microaneurysms in telangiectatic capillaries in diabetic macular edema, and in neovascularization of the disc.
Staining refers to a pattern of hyperfluorescence where the fluorescence gradually increases in intensity through transit views and persists in late views, but its borders remain fixed throughout the angiogram. Staining results from fluorescein entry into a solid tissue or similar material that retains the fluorescein, such as a scar, drusen, optic nerve tissue, or sclera (see Fig 2-1B).
Pooling refers to the accumulation of fluorescein in a fluid-filled space in the retina or choroid. At the beginning of the angiogram, the fluid in the space contains no fluorescein and is not visible. As fluorescein leaks into the space, the margins of the space trap the fluorescein and appear distinct, as seen, for example, in an RPE detachment in central serous chorioretinopathy (Fig 2-2D). As more fluorescein enters the space, the entire area fluoresces.
A transmission defect, or window defect, refers to a view of the normal choroidal fluorescence through a defect in the pigment or loss of pigment in the RPE, such as shown in Figures 2-1A and 2-1B. In a transmission defect, the hyperfluorescence occurs early, corresponding to filling of the choroidal circulation, and reaches its greatest intensity with the peak of choroidal filling. The fluorescence does not increase in size or shape and usually fades in the late phases of the angiogram, as the choroidal fluorescence becomes diluted by blood that does not contain fluorescein. The fluorescein remains in the choroid and does not enter the retina.
Autofluorescence describes the appearance of fluorescence from the fundus captured prior to intravenous fluorescein injection. It is seen with structures that naturally fluoresce, such as optic nerve drusen and lipofuscin.