Article Text


Indocyanine green guided laser photocoagulation in patients with occult choroidal neovascularisation
  1. Andreas W A Weinberger,
  2. Hildegard Knabben,
  3. Uta Solbach,
  4. Sebastian Wolf
  1. Augenklinik der Med Fak der RWTH Aachen, Aachen, Germany
  1. Dr S Wolf, Augenklinik der Med Fak der RWTH Aachen, Pauwelsstrasse 30, 52057 Aachen, Germany.


AIMS To determine whether indocyanine green (ICG) guided laser photocoagulation of occult choroidal neovascularisations (OCNV) is beneficial for patients with occult choroidal neovascularisation secondary to age related macular degeneration (AMD).

METHODS A prospective pilot study was performed in 21 eyes with OCNV secondary to AMD that could be identified extrafoveolarly or juxtafoveolarly in an early ICG angiographic study. Laser photocoagulation was applied to the neovascular membrane identified in the early ICG angiographic study.

RESULTS Visual acuity ranged from 20/400 to 20/20 (logMAR 0.54 (SD 0.29)) before and hand movements and 20/30 (logMAR 0.81 (0.69)) at the last follow up after laser photocoagulation. During the follow up (30 (13) months) vision improved in four eyes (two lines), in seven eyes the initial visual acuity could be stabilised (two lines), in five eyes vision dropped moderately (three to five lines), and in five eyes vision decreased severely (six or more lines). Recurrences (seven patients) or persistent CNV (six patients) was observed in 13 patients.

CONCLUSION This preliminary study of ICG guided laser photocoagulation of occult extrafoveal and juxtafoveal choroidal neovascularisations suggests that this technique may improve the visual prognosis of these patients. Further prospective controlled studies are necessary to confirm these data.

  • indocyanine green angiography
  • occult choroidal neovascularisation
  • age related macular degeneration
  • laser photocoagulation
  • experimental treatment

Statistics from

Age related macular degeneration (AMD) is the most common cause of vision loss in the Western world.1 In the majority of cases, the loss of central vision is secondary to exudative changes and fibrovascular scarring following choroidal neovascularisation (CNV) or vascularised pigment epithelial detachment (PED). These lesions are commonly associated with subretinal exudation, blood, and lipid deposition. The Macular Photocoagulation Study demonstrated that laser photocoagulation of well defined CNV improved the visual outcome compared with the natural course.2 3 However, most patients with newly diagnosed exudative maculopathy secondary to AMD have poorly defined or occult choroidal neovascularisation (OCNV) that is not amenable to photocoagulation therapy by the standards recommended by the Macular Photocoagulation Study Group.4

Recently, indocyanine green angiography (ICGA) has become a clinical tool in the diagnosis and management of choroidal neovascular membranes secondary to AMD.5-11 The infrared absorption and emission spectrum of indocyanine green dye helps to delineate OCNV in many cases since it is capable of penetrating the retinal pigment epithelium, exudates, and, to some degree, even overlying blood.12

In general, OCNV can be classified into two groups depending on the clinical manifestation11: group 1 includes patients presenting with OCNV in association with PED, group 2 includes patients with CNV beneath the retinal pigment epithelium without a distinct PED. The latter corresponds to the type 2 CNV defined by the MPS.13

We and others have reported previously that OCNV defined by the MPS standards could be converted into well defined CNV in up to 50% of cases by ICGA.14-16 However, depending on the imaging technique various approaches were used to delineate OCNV.6 11 14 17 18 In this study we report our results of laser photocoagulation of OCNV that could be visualised in the early ICGA studies using a scanning laser ophthalmoscope.

Patients and methods

A total of 175 patients with AMD who were followed at the outpatient department of the University Hospital Aachen were recruited to the study. Inclusion criteria were age older than 51 years and the presence of OCNV in the fluorescein angiogram (FA). Patients were excluded if they had undergone previous treatment for AMD such as drusen photocoagulation, laser photocoagulation, or radiation therapy for exudative AMD or had any other eye disease that affected visual acuity.

Patients were included in the pilot study of ICG guided argon laser photocoagulation if the early ICGA studies revealed a juxtafoveal or extrafoveal choroidal neovascular network. Additional inclusion criteria were visual acuity equal to or better than 20/400, size of CNV smaller than four disc areas, and duration of symptoms less than 6 weeks. After the experimental nature of the treatment was explained to the patients informed consent was obtained.

Initial examination included a postrefractive ophthalmological examination to determine the best corrected Snellen visual acuity, followed by slit lamp biomicroscopy, Goldmann applanation tonometry, indirect and direct ophthalmoscopy, colour fundus photography, fluorescein angiography (FA), and ICGA. Our technique of ICGA using a scanning laser ophthalmoscope (SLO 101, Rodenstock, Germany; HRA, Heidelberg Engineering, Germany) has been described elsewhere.8 After diagnostic ICGA was performed the neovascular membrane was traced and then transposed to a FA picture using a customised digital image manipulation software (NIH image 1.6). We performed ICG guided laser photocoagulation with argon green to the neovascular membrane. Laser photocoagulation was performed according to a standard protocol similar to the MPS guidelines. The area of the visible extrafoveal choroidal neovascular network was covered with laser confluent burns. In addition, treatment was to extend 100 μm beyond the boundaries of the neovascular membrane. In general, we used a 200 μm spot with a 0.5–1.0 second duration.

Follow up visits were scheduled at 2 weeks, 6 weeks, 3 months, 6 months, and in 6 month intervals thereafter. At each visit a postrefractive ophthalmological examination to determine the best corrected Snellen visual acuity, followed by slit lamp biomicroscopy, Goldmann applanation tonometry, indirect and direct ophthalmoscopy, colour fundus photography, FA, and ICG were performed. In cases of membrane persistency (neovascular network identifiable with ICGA or FA within 6 weeks after initial photocoagulation) or recurrence (identifiable membrane later than 6 weeks after initial photocoagulation) patients received additional laser treatment.

At each follow up examination visual acuity was compared with the initial visual acuity and was considered “improved” if visual acuity improved by two or more lines, as “stable” if visual acuity did not change beyond two lines to the better or worse, as “moderate loss” if vision dropped three to five lines, and as “severe loss” if vision dropped more than six lines. For statistical analysis (Student’s t test, significance if p <0.05) visual acuity data were transformed to log MAR values.


A total of 21 eyes of 21 patients (12 female, 11 male) with a minimal follow up of 12 months were included in the pilot study of ICG guided argon laser photocoagulation. The patients’ ages ranged from 65 to 85 years (mean 73 (SD 5) years). Two eyes had a pigment epithelium detachment (group 1 OCNV, Table 1). Five additional patients showed an extrafoveal plaque-like lesion on late phase ICGA without an identifiable CNV network on early frames but were not included in the study.

Table 1

Visual acuity of study eyes and outcome during follow up graded as “improved” (I), “stabilised” (S), “decreased” (D), or “severely decreased” (SD)

Initial mean CNV size as seen in ICGA was 0.72 (0.49) disc diameters, mean initial exudation extension was moderate (1.48 (0.512)) on a scale from 0 to 4 (0 = no exudation, 4 = massive exudation). Duration of symptoms before presenting to the clinic was 2 (1.1) weeks. Seven patients had recurrent CNV, six patients persistent CNV. Four recurrences were observed within 6 months after initial treatment, two recurrences between 6 and 12 months, and one recurrence 18 months after initial treatment. Patients received up to three re-treatments during follow up. Persistent CNV was located within the laser photocoagulation scar while recurrence was observed in five cases at the foveal edge of the laser photocoagulation scar. No correlation between the initial appearance, location, or size of the OCNV and the rate of recurrence or persistence could be detected. Additionally, the differences between the tracing of ICGA defined CNV and OCNV in FA could not be related to the appearance of recurrences or persistence. Table 2 shows the number of eyes with persistent or recurrent neovascular membranes and the number of eyes that were re-treated at each follow up visit. Persistent CNV were best identified in ICGA, whereas recurrences were seen easier in FA. Treatment was performed using either ICGA or FA (Table 2).

Table 2

Obliteration after laser photocoagulation observed in fluorescein angiograms (FA) and ICG angiograms (ICGA) during follow up

The mean follow up was 30 (13) months, ranging from 12 to 48 months. Initial visual acuity ranged from 20/20 to 20/400 (mean log MAR 0.54 (0.29)). Final visual acuity was counting fingers (CF) to 20/25 (mean log MAR 0.81 (0.69)). Best visual acuity during follow up ranged from 20/20 to 20/800 (mean log MAR 0.34 (0.33)).

Stabilised or improved visual acuity was noted in 76% of eyes, whereas 10% of the eyes experienced a severe visual loss after 6 months of follow up. At 12 months of follow up 66% of eyes had stabilised or improved vision, 14% had a severe visual loss. After 24 months 50% of eyes had stabilised or improved vision while 29% had severe visual loss. Finally, after 36 months the percentage of eyes with stabilised or improved vision decreased to 30% and 40% had experienced a severe visual loss (Tables 3 and 4). The location of recurrent CNV (subfoveal or not) was of greater importance to visual outcome than the number of re-treatments applied (Table 1). However, with an increasing number of recurrences the final visual outcome deteriorated. Mean logMAR was initially 0.54 (0.29), after 12 months 0.56 (0.45), after 24 months 0.79 (0.51) and after 36 months 1.01 (0.80) (see Table 4, Figs 1, 2, 3, 4, 5, 6, 7, 8). The increase of the logMAR value after 36 months is statistical significant compared with baseline values.

Figure 1

Cumulative percentage of patients with severe visual loss. Data from the Macular Photocoagulation Study (MPS)13and from Bressler et al31 for the natural course of occult CNV are included in the diagram.

Figure 2

Scattergram depicting initial and final visual acuity. Points within the dotted lines indicate stabilised visual acuity, points above improved, and points under these lines decreased visual acuity.

Figure 3

Fundus photograph of a patient with occult CNV (FA) and visible neovascular network in the early ICG angiographic study, visual acuity was 20/70.

Figure 4

Fluorescein angiogram of a study patient showing diffuse leakage in the macular area and cystoid macular oedema.

Figure 5

Early phase indocyanine green angiogram reveals a discrete neovascular network above the fovea with a feeding vessel at its nasal superior rim. The white arrows delineate the boarders of the CNV, the dark arrow points to the feeding vessel. Same patient as in Figure 3.

Figure 6

Follow up ICG angiogram 18 months after initial treatment. Visual acuity was 20/200. One re-treatment was performed. Same patient as in Figure 3.    

Figure 7

Follow up fluorescein angiogram 18 months after initial treatment. Visual acuity was 20/200. One re-treatment was performed. Same patient as in Figure 3.

Figure 8

Fundus photograph of the patient 18 months after initial laser photocoagulation, visual acuity was 20/200.

Table 3

Percentage of eyes with improved, stabilised, decreased, and severely decreased visual acuity during follow up

Table 4

Visual acuity in logMAR units (mean (SD)) from study entry to 36 months of follow up


The treatment of patients with newly diagnosed, exudative AMD with an OCNV remains a challenge. At present, guidelines provided by the Macular Photocoagulation Study3 19 require the presence of a well defined CNV diagnosed by fluorescein angiography. However, more than 80% of patients with newly diagnosed exudative AMD present with an untreatable or occult lesion.4 For this reason, new imaging techniques along with alternative treatment such as submacular surgery,20 radiation therapy,21-24 or antiangiogenic treatment25 are under investigation to offer better diagnosis and treatment to patients with exudative AMD and OCNV. ICGA has been shown to be a valuable tool in identifying occult CNV and to convert them into visible CNV either as neovascular capillary network in the early phase angiogram or as plaque-like hyperfluorescence in the late phase angiogram.6 7 11 14-16 26 27 Recently, it has been demonstrated that plaque-like hyperfluorescence can be found in the late phase ICGA images not only in eyes with exudative AMD.28 Therefore, plaque-like hyperfluorescence may not only represent active OCNV but other pathological changes of the RPE complex. In our study we included only eyes with OCNV that demonstrated a visible neovascular network in the early ICGA. These eyes may represent a special subgroup of eyes with OCNV that may be very similar to eyes with classic CNV. In our series of 175 patients we were able to identify the OCNV by ICGA in 94 cases. Most of them were subfoveal; therefore, only 12% of eyes with OCNV could be included in this study.

Previous pilot series of ICG guided laser photocoagulation have been encouraging and showed resolution of the exudative findings in 56% of cases. However, the mean follow up of patients treated has been 6 months only.6 Our data show that patients benefit from laser photocoagulation of OCNV which are identifiable in an early ICGA study. With follow up time the percentage of eyes with visual loss increases overall, but compared with the natural history and radiation therapy the prognosis is still better.13 29 30 Compared with the data published by the Macular Photocoagulation Study Group19 the visual outcome of our study patients compares with those who received laser photocoagulation for well defined extrafoveal and juxtafoveal membranes. This is not surprising since the ICGA basically converts OCNV into a visible neovascular network similar to a well defined CNV; thus, making them suitable for laser photocoagulation treatment. Anatomical differences between these membrane types seem to have little influence on the overall outcome.

Recurrence or persistence was observed in about 60% of cases. This number is similar to findings in classic CNV. Neither the initial appearance, location, size of the OCNV, nor differences between the tracing of ICGA defined CNV and OCNV in FA could be correlated with the appearance of recurrence or persistence. FA proved to be more sensitive than ICGA in detecting recurrence of CNV at the rim of the laser photocoagulation scar (except within the first 2 weeks after treatment) while persistence within the scar area was better defined by ICGA.

In order to validate our findings, further prospective controlled studies are required. Our data suggest that laser photocoagulation of choroidal neovascular membranes which are occult by the means of fluorescein angiography but well defined in the early ICGA study is beneficial for visual prognosis.


Proprietary interest: None


View Abstract

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.