Clinical
Vol 3 No 5 • October/November 2012 • International Journal of Ophthalmic Practice
themselves on the FAF image. This is in accordance 
with the study by McBain et al (2007). McBain et al (2007) also concluded that CNV 
has a characteristic pattern only if it is classified as 
classic. In such a case an area of well defined, low 
FAF signal can be seen over the CNV. This is not 
necessarily because the RPE is damaged but because 
the fluid blocks the FAF signal. In some cases a 
ring of increased FAF is seen around the area with 
decreased FAF. Since CNV, classified as occult, grow 
beneath the RPE, and do not block the FAF signal, 
these cases do not show a typical pattern; sometimes 
there is an increased FAF and sometimes no pattern 
at all. In eyes with AMD that are about to develop a 
geographic atrophy increased FAF can be seen. That 
is not the case of a fundus that is about to develop 
CNV (Smith et al, 2006). Lipofuscin levels are not 
an important part of CNV unless it has been there 
long enough to damage the RPE or if it has brought 
on a retinal detachment. It would be good to have 
a technique that could help us distinguish wet 
AMD (which needs referral) from dry AMD (which 
currently cannot be treated). At present it does not 
seem as though FAF can fufil this function. In the case of other wet changes an abnormal FAF 
can be both absent and present depending on in 
what retinal layer they are located and how long they 
have been there. Ruckmann et al (2002) stated that 
acute central serous chorioretinopathy (CSR) shows 
an increase of FAF over the involved area but there 
is so far no explanation of why. Pece et al (2010) saw 
a FAF pattern in fundi with cystoid macular oedema 
due to diabetes. Where the signal derived from is 
unclear. A hypothesis is that the intraretinal fluid in 
the cysts distorted the FAF signal and concluded that 
other techniques were still preferable for diagnosis of 
cystoid macular oedema. Vujosevic et al (2011) stated 
that increased FAF could be seen in patients with 
diabetic retinopathy but that lipofuscin accumulation 
does not seem to be a part of the pathological process 
of diabetes. They suggested that more studies needed 
to be carried out to determine the origin of the FAF 
signal before conclusions can be drawn about what 
the FAF image tells us. Thus FAF is perhaps not the 
best tool for diagnosing abnormalities in patients with 
diabetes, although cotton wool spots and exudates are 
visible in some cases (see, e.g. Figure 15 and 16). In concordance with Bindewald et al (2005) and 
Landa et al (2011) our experience with the Canon CR-2 
PLUS is that AMD shows a lot of variation in FAF 
appearances and that areas of abnormal FAF may or 
may not correspond to hyperpigmentation or drusen. A hypothesis is that these areas precede visible lesions, 
which will later show on the colour fundus image. If 
this were to be the case it would be of great interest to 
investigate whether the FAF findings, not yet visible 
on the colour fundus image, would change when 
Figure 19. Male, 61 years, decimal visual 
acuity 1.0. On the FAF image there is an area of 
hypofluorescence, slightly above and at 2 o’clock 
from the macula. On the colour fundus image 
an epiretinal membrane is found in the corre-
sponding location. The membrane does not cover 
the macula. Figure 20. Female, 66 years, decimal visual 
acuity 0.8. AMD changes and an epiretinal 
membrane over macula. Where the folds of the 
membrane are seen on the coloured image the 
FAF also shows folds which gives an uneven 
pattern over the macula that usually shows even 
hypofluorescence. However, the FAF pattern is 
most likely also influenced by the AMD changes.