Fundus autofluorescence becomes mainstream in optometry

Found in many OCT devices, this technology's clinical utility is expanding.

Issue: February 2016

ByDoug Rett, OD, FAAO

In the last 5 years or so, fundus autofluorescence has been receiving increasing attention in the posterior segment literature.

Fundus autofluorescence (FAF) has been studied for several decades, but lately there are many commercially available imaging devices on the market, and their penetration into the mainstream has allowed for many interesting discoveries. Today most companies that make a fundus camera, an optical coherence tomographer or a widefield device have an option for FAF. Here I will bring to light some of the conditions that are being studied by FAF, how it is changing our management and what is next for this exciting new field.

Like fluorescein angiography, a FAF image represents everything in the fundus that emits light only at a certain wavelength. In the case of FAF, most imaging systems filter out any wavelength less than 500 nm. The peak emission of FAF near the fovea is 630 nm.

Imagine that looking at a FAF image is like looking at a picture painted only with that wavelength of paint. If the image is black in one area, it does not mean nothing is there; it means that nothing that emits light in that certain wavelength is there. The main thing that emits light at that wavelength in the fundus is lipofuscin (LF). FAF is essentially a picture of where LF is in the fundus.

The imaging system sends in (toward the fundus) visible blue light around 488 nm. This excites LF, which emits a yellowish light around 630 nm; this is allowed to pass through the filter into the camera’s sensor. Most FAF imaging systems interpret this data in a black-and-white representation, to enhance the contrast between LF-containing tissue and tissue that does not contain LF.

LF is important because you can use it to determine if the retinal pigment epithelium (RPE) is healthy, damaged or absent. Sometimes this is obvious on fundus exam, but many times it can be very subtle and other times it can be completely hidden. The RPE is responsible for phagocytosis and lysosomal digestion of the outer segments of photoreceptors, which allows photoreceptors to reset their excitability. The miracle of vision is hard work on the RPE: over the course of a lifetime, it is estimated each RPE cell will phagocytose 3 billion outer segments. As we age, the RPE cells start to incompletely breakdown these outer segments, which causes LF to build up. Certain components of LF, such as A2E and all-trans retinal, are toxic to RPE and retinal cells and can cause apoptosis of these cells past a certain level. An FAF is, at its simplest, a measure of how much lipofuscin is in the patient’s RPE.

So an older person should have a brighter FAF than a younger person, simply because there is more LF built up in the RPE by that age. However, there are conditions that cause a change in the FAF signal.

Figure 1. Color photo of the right eye in 2010.

Figure 2. Color photo of the left eye in 2010.

Figure 3. FAF of the right eye in 2010.

Figure 4. FAF of the left eye in 2010.

Images: Rett D

Age-related macular degeneration

The most obvious application for FAF is to document how a patient with nonexudative AMD progresses. Sometimes RPE atrophy is difficult to detect on funduscopy. Keep in mind that we label RPE atrophy as “geographic” if it is bigger than 175 microns in diameter (about one-eight disc diameter), and if any geographic atrophy (GA) at all is present, the Age-Related Eye Disease Study recommends that the patient be started on antioxidant supplementation. It is important to look for the presence of GA, and FAF is a nice way to do it.

For instance, take a look at Figure 1, the color photo of a patient’s right eye. Most of us can detect the GA in the fovea extending temporally. But take a look at Figure 3, the FAF of the right eye in 2010. Notice that smaller spot of GA just nasal to the fovea. Note how this patient progresses over the years. FAF does a much better job distinguishing exactly how much RPE is actually nonfunctioning than traditional fundus photography or even funduscopy.

The term “junctional zone” is prevalent in FAF literature and could be important in identifying which patients are at risk of their GA increasing. The junctional zone is a ring of even brighter autofluorescence surrounding an area of atrophy; it glows because there is more LF concentrated in the RPE here. Some studies have found that the presence of a junctional zone predicts that the GA will expand at a more rapid rate, which makes sense, but some studies have found no correlation to the progression rate. Remember, the typical rate of GA expansion is 139 microns/year (or about the width of a retinal vein as it leaves the disc). The patient in the accompanying photos did not have a junctional zone, but did progress much quicker than average. Microperimetry has been performed in these junctional zones and found that the sensitivity is significantly less than unaffected retina, implying that the excess LF is already limiting photoreceptor function, even though the RPE cells still exist.

PAGE BREAK

FAF in AMD is about more than just GA. The Fundus Autofluorescence in Age-related Macular Degeneration Study Group (FAM) is a large, ongoing clinical trial in Germany focusing on FAF (Holz, et al.). FAM started collecting data in 2006 and has published a number of interesting articles in the last decade that have increased our knowledge on many aspects of FAF. Some of the group’s seminal work has come from studying the patterns of FAF in AMD, specifically looking at whether a certain pattern of FAF is predictive for a rate of GA progression. This could help clinicians predict how fast a patient may lose vision. FAM found that progression rates in eyes with banded and diffuse patterns at the junctional zone were significantly higher than in those without abnormalities or with focal FAF patterns. Figure 5 is an example of a banded pattern.

Central serous chorioretinopathy

FAF is helpful in identifying central serous chorioretinopathy (CSCR). In its acute phase, the condition will cause blisters of subretinal fluid, which can be somewhat difficult to find clinically, but relatively easy to find with an OCT. However, oftentimes CSCR will cause blisters in the nonfoveal macula, and the patient will be asymptomatic, and eventually the fluid will regress. It will leave an area of disrupted RPE where the blister was, and several of these patches of disruption can help lead to a diagnosis of CSCR. The patches are subtle to see clinically, but using FAF they pop out. They appear as mottled, a combination of glowing and dimmed LF patches. In a patient in whom you suspect might have CSCR, if you see several patches of mottled RPE on FAF, that diagnosis would be strengthened.

Some FAF imaging systems also use near-infrared excitation (as opposed to the short-wavelength of the traditional FAF). The traditional FAF discussed here used 488 nm to excite and detects at wavelengths higher than 500 nm; near-infrared (NIR) excites at 787 nm and detects only wavelengths higher than 800 nm. This makes for a weak return signal that is blocked by most cataracts, but it is helpful for young patients, such as those in whom you might expect to see CSCR. The autofluorescence that corresponds to NIR results from the melanin content in the RPE, but mostly the choroid. Thus, for conditions caused by a dysfunction in the choroid, such as CSCR, it is a useful diagnostic tool and will often reveal large areas of abnormal choroid or choriocapillaris that were previously invisible.

Figure 5. FAF of the right eye in 2012.

Figure 6. FAF of the left eye in 2012.

Figure 7. FAF of the right eye in 2015.

Figure 8. FAF of the left eye in 2015.

Other ocular disease

In relation to choroidal problems, FAF is uniquely helpful in differentiating choroidal melanoma from nevus, because it can help identify the presence of LF. On FAF, even benign choroidal nevi will exhibit an intrinsic autofluorescence, as will any small drusen overlying them, but choroidal melanoma will demonstrate a marked glow due to the LF present in the lesion.

Much work is being done on FAF and Plaquenil (hydroxychloroquine, Sanofi-Aventis) toxicity, as there is some thought that FAF abnormalities will present before OCT structural change or 10-2 central field loss. FAF is one of the three diagnostic tests that is now required for patients to have while being followed for Plaquenil toxicity. They must have spectral domain OCT, FAF or multifocal electroretinogram annually when high risk (Brown).

Interestingly, FAF is being used to evaluate for normotensive glaucoma progression. The thought is that as perfusion to the optic nerve head decreases, peripapillary atrophy increases; this could be easily followed with FAF, and the clinician could even check for the presence of a junctional zone. Of course, FAF is handy is detecting disc drusen, especially in adults when the drusen have migrated more anterior in the disc. There is even work being done to help differentiate the confusing white dot syndromes by their FAF patterns.

Finally, an interesting article out of Japan and Switzerland looked at the change in FAF after cataract surgery (Nagai, et al.). Researchers gave half their patients clear IOLs and half yellow-tinted IOLs and measured the progression of FAF in both groups. After 2 years the group with clear IOLs had higher FAF and a higher incidence of AMD findings.

It is important to do FAF and especially worth the small extra effort if you already have it built into your OCT device. Like most things, the more you do it, the more you will discover.

References:
Browning DJ. Am J Ophthalmol. 2013; doi:10.1016/j.ajo.2012.09.025.
Holz FG, et al. Am J Ophthalmol. 2007;doi:10.1016/j.ajo.2006.11.041.
Holz FG, et al. Atlas of Fundus Autofluorescence Imaging. New York, NY: Springer-Verlag Berlin Heidelberg; 2007.
Nagai H, et al. J Cataract Refract Surg. 2015;doi:10.1016/j.jcrs.2015.01.017.
Schmitz-Valckenberg S, et al. Invest Ophthalmol Vis Sci. 2006;doi:10.1167/iovs.05-0892.
Schmitz-Valckenberg S, et al. Surv Ophthalmol. 2009;doi:10.1016/j.survophthal.2008.10.004.
Sepah YJ, et al. Saudi J Ophthalmol. 2014;doi:10.1016/j.sjopt.2014.03.008.
Shields C, et al. Br J Ophthalmol. 2008;doi:10.1136/bjo.2007.130286.
Wester ST, et al. Ophthalmic Surg Lasers Imaging Retina. 2010;doi:10.3928/15428877-20091230-15.
Yehoshua Z, et al. Ophthalmic Surg Lasers Imaging Retina. 2013;doi:10.3928/23258160-20130313-05.

For more information:
Doug Rett, OD, FAAO, sees patients and works with students and residents at the optometry clinic for the VA Boston Healthcare System. He can be reached at doctorrett@gmail.com.

Disclosure: Rett has no relevant financial disclosures.