History of Indocyanine Green Angiography

The ICG dye has been used in medicine since the mid-1950s as a method for continuously recording time-varying dilation in whole blood for applications initially in the field of cardiology, such as the measurement of cardiac output in valvular and septal defects. In fact, it is still used today in a device designed to determine simultaneous measurement of cardiac output, blood volume, and hepatic clearance. For ophthalmic angiography, Kogure and Choromokos first tried to use the ICG molecule for infrared absorption fundus angiography. Their efforts were improved by Hochheimer and Flower, who performed elegant studies to determine the potential use for infrared absorption angiography of the choroid in the aging and glaucomatous eye. This work culminated in a classic description of infrared absorption choroidal angiography performed simultaneously with FA in 1976. Patz joined Flower to investigate possible clinical applications for ICG angiography; however, their methodology was inhibited by the minimal fluorescence elicited by the molecule and the primitive technology available at that time to record it. The advent of digital angiography with a highly weighted camera in the near infrared and high-speed imaging capability made it possible for ICG angiography to acquire clinically meaningful images. The excitatory and fluorescing wavelengths used in ICG angiography enhanced penetration posteriorly into the choroid with reduced interference by the RPE and serosanguineous manifestations in the fundus. These advances permitted our group to first describe the use of digital ICG angiography in the ophthalmic setting in 1992. Although imaging of the choroid was vastly improved by this technique, practical clinical value was initially restricted by the limits of resolution of the photographic systems and the lack of convincing evidence that such information was worthwhile in the clinical setting compared to other available methods of diagnosis. Also influencing its use was the lack of enthusiastic support by some retinal specialists throughout the United States, many at leading institutions where their database related to chorioretinal diseases was exclusively film-based, utilizing FA. Poor digital resolution and the lack of stereopsis were reasons for slowing down progress and acceptance of digital imaging and ICG angiography at these centers, which were led by many experts in the field. These clinical scientists, referred to by the ophthalmic photography population as “stereo-film based fluorescein advocates,” resisted the conversion to digital imaging in general—for good reason. The sharper images on film and the stereoscopic capability were the standards of that era. Eventually, there was improvement in digital resolution, initially a near doubling effect. As well, the cameras required to record infrared images also improved dramatically. These developments set the stage for further use of ICG angiography, which required a digital system. However, the most prominent force in converting to digital imaging for clinical scientists throughout the United States was the decision by regulatory agencies to require a digital database for their clinical trials. This policy obliged teaching centers throughout the United States to convert to the digital system. Curiously, Kodachrome film is actually no longer available. So, the conversion to digital was inevitable.

Although I was initially an FA film-based stereo advocate, our retinal group in New York continued to study the choroidal circulation with ICG angiography, particularly in neovascular age-related macular degeneration (AMD) and a few other selected abnormalities of the choroid. The concept was to image the retinal circulation with FA and the choroidal circulation with ICG angiography. In the beginning, I was particularly interested in using ICG angiography to study 2 aspects of neovascular AMD, the so-called choroidal “feeder vessel,” which presumably perfused a larger neovascular complex, and the precise nature of a vascularized pigment epithelial detachment (vascularized PED) since that lesion was known to contain subpigment epithelial vessels in the form of a choroidal neovascular membrane as well as serous exudation. The intent was to treat the “feeder vessel” and the neovascular component of a vascularized PED with laser photocoagulation, the only available method of therapy at that time. This concept, which was designed to obliterate or infarct the membrane, did not progress beyond multiple encouraging but uncontrolled studies. There was some anecdotal evidence that both approaches could benefit patients, but no clinical trial emerged to provide the convincing evidence needed for determining an acceptable treatment method. Specifically, a clinical trial with randomization, prospectivity, matched controls, and reasonable follow-up to establish safety and efficacy was never realized. Alas, the feeder vessel technique and photocoagulation of the neovascular component of the PED, or the so-called “hot spot,” were never accepted as worthwhile forms of therapy by the general retinal public. Essentially, this therapeutic laser concept was destined to yield to proven alternative methods using photodynamic therapy (PDT) with verteporfin (Visudyne) and anti-vascular endothelial growth factor (VEGF) pharmacologic treatment. However, continued improvement in ICG angiography provided high-speed, highly resolved images, which set the occasion for its use in other selective chorioretinal diseases.

To assess the relevance and the clinical value of ICG angiography in the management of chorioretinal diseases, 10 years after its introduction, in 2003, an international group of qualified retinal specialists reviewed a total of 376 articles on the technique. Based on published, peer-reviewed studies, which provided evidence-based clinical information, they concluded that there were relatively few specific indications for ICG angiography. They strongly recommended ICG angiography to 1) identify polypoidal choroidal neovascularization (polypoidal CNV); 2) document “occult” or type-1 choroidal neovascularization (CNV) beneath the RPE in AMD, particularly with a significant vascularized PED; and 3) evaluate recurrent CNV. They also recommended ICG angiography with some enthusiasm to 1) identify feeder vessels in neovascular AMD; 2) study chronic central serous chorioretinopathy (CSC); and 3) evaluate posterior uveitis for conditions such as multiple evanescent white dot syndrome (MEWDS), acute multifocal placoid pigment epitheliopathy (AMPPE), nonspecific vasculitis, Vogt-Koyanagi-Harada disease (VKH), angioid streaks, and birdshot chorioretinopathy. They also concluded that ICG angiography was a well-established technology with a distinct advantage for imaging relatively few chorioretinal diseases, and that it had contributed to our understanding of the pathophysiologic mechanisms in many ocular diseases.

So, what has emerged with ICG angiography and the practice patterns of physicians since this revealing article? Actually, not much. One reason is the introduction and development of high-resolution SD-OCT. This noninvasive imaging system provided a more precise, high-resolution histologic and physiological representation of the posterior fundus from the vitreoretinal interface, through the inner retina, the RPE, and recently, the choroid. OCT also provided a 3-dimensional reconstruction of the macula with the capability for automated comparisons for point-to-point correlations in time during the course of an abnormality. Recently another feature to the system, enhanced depth imaging (EDI), now provides clearer imaging of the choroid, revealing atrophic as well as infiltrative changes in a variety of disease processes. This sophisticated, noninvasive technology has become the main method for retinal specialists to document exudative and degenerative manifestations at all levels of the posterior fundus.

International Use of Indocyanine Green Angiography

Given the advantages of OCT in its ability to image the choroid, what is left today for ICG angiography in the clinical setting? To answer this question, I invited experts in the retina and uveitis specialties to comment on their experience and personal practice preferences for diagnosing and managing patients with ICG angiography. Each has used ICG angiography extensively with the most advanced technology, avoiding the pitfalls of retrofitted photographic cameras, which were never on a par with standard factory-assembled ICG angiography systems. These physicians do not refer patients to a remote ICG angiography location to receive hard copy for image analysis; rather, they review the study digitally on remote, personal monitors or viewing stations. The image interpretation is distinctly improved with this approach with selective analysis, magnification, enhanced contrast, and comparison to previous acquisitions. These retinal and uveitis specialists have also published extensively on the use of ICG angiography. Each was asked to comment on the question, “How do you use ICG angiography in the clinical setting today?” Their responses were as follows:

Tomohiro Iida, from Fukashima, Japan, reported that he and other Japanese retinal specialists usually perform ICG angiography on all of their patients with neovascular AMD. This is not surprising since there is a well-known predilection for polypoidal CNV in Asians. The vast majority of Japanese patients develop this form of neovascularization in AMD with more than 50% of them having active polypoidal leakage and/or bleeding. An additional percentage has an inactive or consolidated-regressed plaque of polypoidal CNV that is more visible on ICG angiography since the sub-RPE vessels are not obscured by an intact pigment epithelium. Furthermore, Japanese retinal physicians use this knowledge to incorporate the use of verteporfin (Visudyne) photodynamic therapy or PDT as a primary preference or in conjunction with anti-vasogenic pharmacotherapy. The first PDT clinical trial for AMD in Japan showed significant benefit for treating neovascular AMD compared to other studies from western countries, where polypoidal CNV is present in only 8% or 9% of newly diagnosed neovascular AMD patients.

Giselle Soubrane, from Creteil, France, uses ICG angiography in practically all of her neovascular AMD patients, a mixed racial population. She believes that ICG angiography offers a better understanding of the neovascularization process. Felice Cardillo-Piccolino, from Genoa, Italy; Ursula Schmidt-Erfurth, from Vienna, Austria; Giovanni Staurenghi, from Milan, Italy; and Sebastian Wolf, from Bern, Switzerland, have a similar approach for the use of ICG angiography in their neovascular AMD patients. When polypoidal CNV is suspected in any patient, they regard ICG angiography as the principal imaging modality. Cardillo-Piccolino, Schmidt-Erfurth, Soubrane, and Staurenghi also believe that the use of ICG for neovascular AMD in their patients exceeds the use by retinal specialists in the United States. Wolf says that he is more selective based on the clinical manifestations, and he actually uses ICG angiography less today than in the past. Staurenghi is the only one in this group of retinal specialists who still occasionally uses high-speed ICG angiography for the feeder vessel technique in neovascular AMD.

These retinal specialists also use ICG angiography in the clinical setting to evaluate other neovascular maculopathies. For example, Soubrane uses ICG angiography to identify lacquer cracks, the usual originating site for CNV in pathologic myopia. Schmidt-Erfurth uses ICG angiography in patients with an acquired vitelliform detachment. They each recommend ICG angiography in such patients to rule out the possibility of CNV beneath a turbid or hemorrhagic detachment before instituting interventional forms of therapy. Both also feel that ICG angiography is essential for imaging retinal angiomatous proliferation (RAP), so-called Type 3 neovascularization in AMD, originating from the retina to form a retinal-choroidal anastomosis.

In Japan, according to Iida, half fluence PDT under ICG angiography guidance is also used for chronic CSC. Although OCT with EDI can detect a thick choroid as a marker in CSC, each of the retinal specialists uses ICG angiography to image the multifocal areas of choroidal permeability, a marker for the disease, and a guide for half fluence PDT when indicated. This therapeutic approach has become their preferred standard of care. Cardillo-Piccolino also believes that ICG angiography is of importance when these patients are misdiagnosed because of the presence of polypoidal CNV masquerading as chronic CSC.

All of these retinal specialists feel that ICG angiography is the best way to image a choroidal hemangioma. The tumor’s vasculature is clearly delineated with ICG angiography, delineating it from the surrounding exudative detachment, which stains brilliantly with FA. Late leakage from pooling of the ICG molecule in the subretinal space is seen in contrast to the disappearance of the dye from the vascular lesion (“wash-out”). Soubrane feels that specific, intrinsic vascular patterns of choroidal melanomas vs suspicious nevi can also be determined with ICG angiography, but this was not recommended by the other retinal specialists.

Carl Herbort, from Lausanne, Switzerland, a prominent uveitis-retinal specialist, recommended the use of ICG angiography to evaluate selective cases of posterior uveitis. He feels that at least half of the posterior uveitis conditions originate in or predominantly involve the choroid and are best investigated with ICG angiography. He prefers ICG angiographic imaging for choroidal inflammatory syndromes, and he is puzzled why many uveitis specialists, particularly in the United States, seldom require or rely on this form of imaging in the management of their patients. Herbort says that uveitis is by far the principal clinical area in which ICG angiography is most useful, if not essential. He believes that ICG angiography makes it possible to classify and treat a choroiditis according to its pathophysiology and not purely by its clinically evident manifestations. His classification of choroidal inflammatory diseases is based on the perfusion of the choriocapillaris documented with ICG angiography. For example, in MEWDS, multifocal choroiditis, and AMPPE, there is nonperfusion of the choroid, in contrast to birdshot chorioretinopathy, VKH disease, sympathetic ophthalmia, tuberculosis, and sarcoidosis, which present with granulomatous infiltrations.

International Use of Indocyanine Green Angiography

Given the advantages of OCT in its ability to image the choroid, what is left today for ICG angiography in the clinical setting? To answer this question, I invited experts in the retina and uveitis specialties to comment on their experience and personal practice preferences for diagnosing and managing patients with ICG angiography. Each has used ICG angiography extensively with the most advanced technology, avoiding the pitfalls of retrofitted photographic cameras, which were never on a par with standard factory-assembled ICG angiography systems. These physicians do not refer patients to a remote ICG angiography location to receive hard copy for image analysis; rather, they review the study digitally on remote, personal monitors or viewing stations. The image interpretation is distinctly improved with this approach with selective analysis, magnification, enhanced contrast, and comparison to previous acquisitions. These retinal and uveitis specialists have also published extensively on the use of ICG angiography. Each was asked to comment on the question, “How do you use ICG angiography in the clinical setting today?” Their responses were as follows:

Tomohiro Iida, from Fukashima, Japan, reported that he and other Japanese retinal specialists usually perform ICG angiography on all of their patients with neovascular AMD. This is not surprising since there is a well-known predilection for polypoidal CNV in Asians. The vast majority of Japanese patients develop this form of neovascularization in AMD with more than 50% of them having active polypoidal leakage and/or bleeding. An additional percentage has an inactive or consolidated-regressed plaque of polypoidal CNV that is more visible on ICG angiography since the sub-RPE vessels are not obscured by an intact pigment epithelium. Furthermore, Japanese retinal physicians use this knowledge to incorporate the use of verteporfin (Visudyne) photodynamic therapy or PDT as a primary preference or in conjunction with anti-vasogenic pharmacotherapy. The first PDT clinical trial for AMD in Japan showed significant benefit for treating neovascular AMD compared to other studies from western countries, where polypoidal CNV is present in only 8% or 9% of newly diagnosed neovascular AMD patients.

Giselle Soubrane, from Creteil, France, uses ICG angiography in practically all of her neovascular AMD patients, a mixed racial population. She believes that ICG angiography offers a better understanding of the neovascularization process. Felice Cardillo-Piccolino, from Genoa, Italy; Ursula Schmidt-Erfurth, from Vienna, Austria; Giovanni Staurenghi, from Milan, Italy; and Sebastian Wolf, from Bern, Switzerland, have a similar approach for the use of ICG angiography in their neovascular AMD patients. When polypoidal CNV is suspected in any patient, they regard ICG angiography as the principal imaging modality. Cardillo-Piccolino, Schmidt-Erfurth, Soubrane, and Staurenghi also believe that the use of ICG for neovascular AMD in their patients exceeds the use by retinal specialists in the United States. Wolf says that he is more selective based on the clinical manifestations, and he actually uses ICG angiography less today than in the past. Staurenghi is the only one in this group of retinal specialists who still occasionally uses high-speed ICG angiography for the feeder vessel technique in neovascular AMD.

These retinal specialists also use ICG angiography in the clinical setting to evaluate other neovascular maculopathies. For example, Soubrane uses ICG angiography to identify lacquer cracks, the usual originating site for CNV in pathologic myopia. Schmidt-Erfurth uses ICG angiography in patients with an acquired vitelliform detachment. They each recommend ICG angiography in such patients to rule out the possibility of CNV beneath a turbid or hemorrhagic detachment before instituting interventional forms of therapy. Both also feel that ICG angiography is essential for imaging retinal angiomatous proliferation (RAP), so-called Type 3 neovascularization in AMD, originating from the retina to form a retinal-choroidal anastomosis.

In Japan, according to Iida, half fluence PDT under ICG angiography guidance is also used for chronic CSC. Although OCT with EDI can detect a thick choroid as a marker in CSC, each of the retinal specialists uses ICG angiography to image the multifocal areas of choroidal permeability, a marker for the disease, and a guide for half fluence PDT when indicated. This therapeutic approach has become their preferred standard of care. Cardillo-Piccolino also believes that ICG angiography is of importance when these patients are misdiagnosed because of the presence of polypoidal CNV masquerading as chronic CSC.

All of these retinal specialists feel that ICG angiography is the best way to image a choroidal hemangioma. The tumor’s vasculature is clearly delineated with ICG angiography, delineating it from the surrounding exudative detachment, which stains brilliantly with FA. Late leakage from pooling of the ICG molecule in the subretinal space is seen in contrast to the disappearance of the dye from the vascular lesion (“wash-out”). Soubrane feels that specific, intrinsic vascular patterns of choroidal melanomas vs suspicious nevi can also be determined with ICG angiography, but this was not recommended by the other retinal specialists.

Carl Herbort, from Lausanne, Switzerland, a prominent uveitis-retinal specialist, recommended the use of ICG angiography to evaluate selective cases of posterior uveitis. He feels that at least half of the posterior uveitis conditions originate in or predominantly involve the choroid and are best investigated with ICG angiography. He prefers ICG angiographic imaging for choroidal inflammatory syndromes, and he is puzzled why many uveitis specialists, particularly in the United States, seldom require or rely on this form of imaging in the management of their patients. Herbort says that uveitis is by far the principal clinical area in which ICG angiography is most useful, if not essential. He believes that ICG angiography makes it possible to classify and treat a choroiditis according to its pathophysiology and not purely by its clinically evident manifestations. His classification of choroidal inflammatory diseases is based on the perfusion of the choriocapillaris documented with ICG angiography. For example, in MEWDS, multifocal choroiditis, and AMPPE, there is nonperfusion of the choroid, in contrast to birdshot chorioretinopathy, VKH disease, sympathetic ophthalmia, tuberculosis, and sarcoidosis, which present with granulomatous infiltrations.