Ординатура / Офтальмология / Учебные материалы / Retinal Vascular Disease Joussen Springer

II 11

11 II

Two basic patterns of abnormalities can be found during angiography: hyperfluorescence and hypofluorescence.

Hyperfluorescence is defined as an increase in fluorescein detection in a particular structure compared to standard angiography. It can be caused by an increased permeation of dye (e.g., damage to the integrity of the blood-retinal barriers), newly developed structures containing fluorescein (e.g., neovascularizations of the retina or choroid) or an increased visibility of normal fluorescein concentrations (e.g., window defects allowing better visualization of choroidal fluorescence through a window in the retinal pigment epithelium).

The term “leakage” refers to a subgroup of hyperfluorescence and is defined as passage of fluorescein visible on angiography across barriers which it normally does not cross in a significant amount (e.g., inner and outer retinal blood barrier). It can further be subdivided into “staining” (leakage into a solid structure, e.g., scar tissue) or “pooling” (leakage into preformed hollow spaces, e.g., serous pigment epithelial detachment) or a combination of both. The terms “leakage” and “staining” have recently been used to distinguish active from fibrotic choroidal neovascularizations following photodynamic therapy. Here, “leakage” is defined as increasing leakage through the late phases of angiography in contrast to “staining” as a sign of inactive CNV (usually attenuation of hyperfluorescence towards the late phase).

Hypofluorescence is defined as a decrease of fluorescein detection compared to standard angiography. It can be caused by blockage of exciting or emitting light (e.g., by haemorrhages) or a decrease of fluorescein containing structures or their fluorescein content (e.g., occlusion of retinal or choroidal vessels).

11.5Quantitative Evaluation of Fluorescein Angiography

By means of densitometric measurements the filling times of arterioles, venules, or defined areas of the retina or choroid can be evaluated from fluorescein angiography [8]. Early, middle, and late phases are differentiated; the examiner precisely specifies the time after injection of the fluorescein or the time after the dye first appears on the optic disk.

Arm-retina time: The normal arm-retina time is about 10 – 20 s. A prolonged arm-retina time indicates a generalized circulatory disorder (e.g., cardiovascular insufficiencies, carotid occlusion).

Arterial filling time should be faster than 1 s. Slow arterial filling indicates reduced perfusion pressure that may be related to a stenosis of the central retinal

or ophthalmic arteries, increased intraocular pressure, or severe impairment of systemic circulation.

Arteriovenous passage time is the time period between the inflow of the dye into retinal arteries and

the appearance in the corresponding venules. Vari- II 11 ous studies have indicated that arteriovenous pas-

sage time is prolonged in retinal vascular disorders like diabetic retinopathy, retinal vascular occlusion, inflammatory retinal and choroidal disease, and glaucoma.

11.6 Avoiding Unnecessary Angiographies

As with all invasive diagnostic tests, fluorescein angiography should only be employed if the diagnosis cannot be established without the test, the angiography serves as a treatment guideline or is needed for interpretation of the course of a disease during fol- low-up examinations. FA should not be used as a substitute for a thorough clinical examination or for documentation purposes only.

Due to the increasing demand for FA following the development of new treatment strategies for the treatment of exudative AMD, the capacities for FA are nowadays often stretched to the limit within tertiary centers. In addition to the potential side effects of angiography and the health-economic aspects of performing unnecessary examinations, this is another reason to avoid FAs that are not needed for the above reasons.

Appropriate communication with the photographer is essential. Before angiography, the photographer should be sure about the structures to be examined and particular points of importance during the angiography (e.g., which phase of angiography, which region of the retina, which magnification is to be used and additional exposures of the periphery or the fellow eye). The setting up of standardized forms for angiography and standard protocols for particular diseases (e.g., for exudative AMD or diabetic macular edema) within the department is advisable.

In our experience, superfluous angiographies are frequently performed in the following situations:

Non-exudative AMD. For documentation purposes, color photographs or autofluorescence examinations are sufficient.

End-stage exudative AMD. There is no therapeutic option available at present and the diagnosis can usually be established without angiography. Clinically significant macular edema in diabetic retinopathy. The diagnosis is based on ophthalmoscopy and not on angiography. FA should only be performed if laser treatment is planned (to identify sources of leakage and areas of ischemia). Optical coherence tomography (OCT)

seems to be particularly helpful for monitoring the course of clinically significant macular edema (CSME) following treatment with laser photocoagulation or intravitreal triamcinolone.

11 II Retinal venous occlusions. The diagnosis can usually be established without angiography. The use of FA for the differentiation of ischemic versus non-ischemic venous occlusions is highly controversial. An evidenced-based necessity of FA is at present established only in branch vein occlusions according to the guidelines of the Branch Vein Occlusion Study (see Chapter 21.3).

Intraocular tumors. Often, angiography is inadequately included in the standard workup of intraocular tumors. In experienced institutions, the diagnosis can be established based on clinical examination and ultrasonography in most cases.

If the diagnosis cannot be established by other methods, angiography is seldom able to achieve that essential bit of information needed for differential diagnosis.

References

1.Blacharski PA (1985) Twenty-five years of fluorescein angiography. Arch Ophthalmol 103(9):1301 – 2

2.Bressler NM (2002) Verteporfin therapy of subfoveal choroidal neovascularization in age-related macular degeneration: two-year results of a randomized clinical trial including lesions with occult with no classic choroidal neovascularization – verteporfin in photodynamic therapy report 2. Am J Ophthalmol 133(1):168 – 9

3.Early Treatment Diabetic Retinopathy Study Research Group (1995) Focal photocoagulation treatment of diabetic macular edema. Relationship of treatment effect to fluorescein angiographic and other retinal characteristics at baseline: ETDRS report no. 19. Arch Ophthalmol 113(9):1144 – 55

4.Gass JD (1997) Stereoscopic atlas of macular diseases: Diagnosis and treatment, 4th edn. Mosby, St. Louis

5.Macular Photocoagulation Study Group (1991) Subfoveal neovascular lesions in age-related macular degeneration. Guidelines for evaluation and treatment in the macular photocoagulation study. Arch Ophthalmol 109(9):1242 – 57

6.The Branch Vein Occlusion Study Group (1984) Argon laser photocoagulation for macular edema in branch vein occlusion. Am J Ophthalmol 98(3):271 – 82

7.Wessing A (1968) Fluoreszenzangiografie der Retina. Lehrbuch und Atlas. Georg-Thieme Verlag, Stuttgart

8.Wolf S, Arend O, Reim M (1994) Measurement of retinal hemodynamics with scanning laser ophthalmoscopy: reference values and variation. Surv Ophthalmol 38 Suppl: S95 – 100

205

II 12

diagnostic imaging data. Time-consuming evaluations of color images for retinal diseases such as diabetic retinopathy were still based on subjective analyses by human graders. Angiographic assess-

12 II ments still required training and expertise, and were plagued by inconsistencies and intergrader variability [32, 38]. Few investigators took advantage of the newly available digital information from these images to perform advanced quantitative analyses.

During that time, Carl Zeiss Meditec, Inc. (Dublin, CA) was quietly developing a fledgling technology that would soon revolutionize the field – OCT. Introduction of this disruptive technology required the development of novel hardware and software as well as a new market in which to sell the instrument. Most importantly, Zeiss chose to make a radical departure from industry standards. Instead of simply providing a fundus image to be viewed and interpreted by the clinician, Zeiss chose to automate the extraction of quantitative information from OCT images. This was a dramatic departure from the industry norm, which only required that fundus images be provided for subjective evaluations, without measurements or automated assessments.

With slit-lamp biomicroscopy firmly established as the gold-standard clinical evaluation and angiography as the principal tool for further disease clarification, many clinicians in the mid 1990s were skeptical of this new technology. Soon, however, unique benefits of noninvasive cross-sectional imaging became apparent. For example, vitreomacular traction syndrome (VMT), which was previously clinically unapparent in many patients, was able to be visualized and diagnosed with confidence using OCT. Macular holes and cysts could also be visualized and monitored. Furthermore as intravitreal therapies for retinal diseases were introduced, OCT’s value in the quantification of retinal thickening and subretinal exudation became certain. The optimal integration of OCT data with other clinical findings to synthesize a clinical care plan is still an evolving subject.

12.2.1 Optical Coherence Tomography

Essentials

Provides both cross-sectional visualization and clinically relevant quantitative measurements

Based on low-coherence interferometry

Measures reflectivity of tissue interfaces with axial resolution less than 10 μm

Good reproducibility in patients with diabetes

A complete discussion of OCT is beyond the scope and purpose of this chapter. The reader is referred to several excellent texts on this topic [27, 30, 33, 34]. Briefly, OCT is based on the principle of lowcoherence interferometry. Akin to B-scan ultrasonography, OCT uses differences in the reflection of light, instead of sound echoes, to render twodimensional images (tomograms) of the retina. Various light sources, typically superluminescent diodes or lasers with very short pulses (i.e., femtosecond lasers), are used to generate broad bandwidth light. The depth or axial resolution of OCT is based on the bandwidth of these sources, and the lateral resolution is determined by the diameter of the focused probe beam.

The light is split into two different paths: a reference beam projected inside the instrument and a sample beam focused on the tissue of interest. In time domain OCT, differences in the time of flight for these two light paths are measured using a Michel- son-type interferometer. In Fourier domain OCT, these differences are characterized with a spectrometer and Fourier-based mathematical calculations. Differences in the optical characteristics of ocular tissues result in the different reflectivity intensities that are measured by OCT.

A single point of light reflected off the retina forms an A-scan, which contains information about the axial location of these tissue interfaces. These single points of light can be laterally aligned to form B-scan images, which often use a false-color display to depict interface intensities: highly reflective interfaces are rendered in white and red, medium level reflections are shown as yellow and green, and features with low reflectivity are depicted in blue. OCT data can be viewed en face as a C-scan or in dense three-dimensional cubes (3D OCT) by capturing B- scans or C-scans in rapid succession. Whereas the axial resolution of clinical ophthalmic echography is limited to greater than 100 μm, differences less than 10 μm can be discerned with conventional OCT instruments. Newer instruments, potentially coupled with adaptive optics devices, can resolve structures that are separated by less than 2 μm in the axial direction [109].

Scan acquisition is painless for the subject, but requires cooperation and steady fixation. Due to inherent speed limitations in conventional timedomain OCT technology, a radial pattern of scan line capture is often used as a compromise between acquisition time and imaging density. Even with this compromise, time-domain OCT instruments require the subject to maintain steady fixation for many seconds at a time, which may be difficult for patients with macular diseases [15, 38]. Therefore, use of an external fixation light for the fellow eye has been

advocated when the acuity of the eye being examined is less than 20/300 [38].

Software included with conventional OCT instruments attempts to measure the thickness of the retina by first identifying the anteriormost highly reflective layer (inner border of neurosensory retina) and then identifying the posterior extent of the retina just anterior to the highly reflective retinal pigment epithelium (RPE) layer [15]. Many studies have verified the reproducibility of these OCT measurements by acquiring several measurements of the same eye at the same visit [3, 6, 9, 25, 28, 29, 45, 49, 50, 56, 58, 112]. OCT measurements also appear to be reproducible in patients with diabetes. Baseline differences in retinal thickness measurements between normal subjects and diabetic patients without clinically significant macular edema (CSME) may exist and be subject to diurnal variation [9, 25, 43, 82]. Male gender, high BMI, and longer axial lengths may also be associated with increased retinal-thickness measurements and may need to be considered in clinical studies utilizing OCT [112]. Investigators have also found that high-resolution scans acquired after mydriasis provide more consistent results [75], and total macular volume measurements may be more consistent than foveal thickness values [6].

These studies provide evidence that OCT is both accurate and reproducible. Quantitative measurements of retinal thickness provide clinicians with actual numbers which can be used to treat patients with complex diseases. In addition, the images themselves are an excellent way to assess and document clinical status and help patients to understand and visualize their diseases.

12.2.2 Biomicroscopic Examination

Essentials

OCT may be more sensitive for the detection of CSME than non-contact or contactlens biomicroscopy

Diffuse or subclinical (between 200 and 300 μm) retinal thickening and absence of hard exudates in the central macula both predict failure of slit-lamp biomicroscopy for the detection of CSME

Diagnosis by OCT cannot yet replace the clinical diagnosis of CSME

Macular edema is a common final pathway for many retinal vascular diseases, and its clinical detection is critical for the prevention of vision loss. One of the best studied examples of this is CSME. The clinical definition of CSME is based on the assessment of the

diabetic fundus with stereoscopic slit-lamp biomicroscopy through a pharmacologically dilated pupil [18]. Since non-contact lens biomicroscopy may fail to detect CSME in as many as 10 % of cases, many cli-

nicians advocate the use of contact lens biomicrosco- II 12 py to prevent blinking and eye movements, increase

axial magnification, and facilitate the detection of subtle areas of edema [5, 10]. Regardless of the technique chosen, the results of this time-consuming examination are subjective, ephemeral, and at best semiquantitative.

OCT, on the other hand, is a high-resolution, cross-sectional, objective, and reproducible method of documenting the retinal thickness and morphology of macular diseases. Several studies suggest that OCT may be more sensitive than slit-lamp biomicroscopy for detecting small changes in retinal thickness [28, 54, 91, 116]. Lattanzio and colleagues documented a progressive increase in mean foveal thickness in patient groups advancing from no edema (228 μm) to non-clinically significant macular edema (322 μm) to clinically significant macular edema (476 μm) [44]. This suggests that OCT’s graded measurements may be more sensitive to small changes than categorical assessments by clinicians. Yasukawa et al. found that all eyes with CSME by non-contact biomicroscopy had detectable retinal thickening by OCT, whereas almost one-third of the eyes graded clinically as non-edematous had retinal thickening by OCT [117]. Brown et al. discovered that even with contact lens examination, more than half of cases with retinal thickness greater than 200 μm by OCT had clinically undetectable retinal thickening and almost one-fifth of cases with retinal thickness greater than 300 μm were also undetectable [5]. In their final assessment, they found that almost one-fourth of diabetic macular edema (DME) cases discovered with OCT had undetectable thickening on clinical examination. Other investigators have also found that OCT has a higher sensitivity than clinical examination for the detection of other features of diabetic retinopathy such as intraretinal cysts and subretinal fluid [67].

Investigators using OCT as the gold-standard definition of DME have found that diffuse thickening, thickening between 200 and 300 μm, and absence of hard exudates in the central macula are all predictors for the failed detection of CSME by slit-lamp biomicroscopy [5, 10, 28, 116]. Since more than half of the errors in the diagnosis of DME may be on the conservative side [8], some clinicians advocate the use of OCT if macular edema is not detected using standard clinical methods [5]. However, Browning also found a substantial percentage of patients undergoing focal laser therapy for CSME who did not have a single OCT zone thickened beyond normal ranges [7].

Since OCT was developed after the completion of the Early Treatment Diabetic Retinopathy Study, we do not have robust evidence for its use as an adjunct or replacement for slit-lamp biomicroscopy, nor do we

12 II have guidance on treatment recommendations for subclinical or OCT-evident macular edema. Nevertheless, even in the absence of proof from robust clinical trials, OCT is still becoming a standard of clinical practice for the evaluation of macular diseases, and will likely become a permanent feature in treatment and management decisions for diabetic macular edema.

12.2.3 Stereoscopic Fundus Photography

Essentials

Reading center protocols for fundus image assessment are not typically accessible to the clinician in practice

OCT may be a more objective and quantitative method of standardizing retinal thickness assessments both in clinical trials and in clinical practice

Diagnosis by OCT cannot yet replace the photographic diagnosis of CSME

Fundus photograph grading systems developed for the purposes of clinical trials provide a reproducible, semiquantitative assessment of disease stage, but are relatively complicated, time-consuming, and difficult to use in daily practice [25]. The excellent spatial resolution of conventional fundus imaging makes it unlikely to be supplanted in the foreseeable future. However, variability in the axial resolution of stereoscopic photographs, either from photographic technique or image quality, calls into question the advisability of basing future macular edema trials on this imperfect analysis. Furthermore, the continued use of reading center evaluations in clinical trials propagates a fundamental disconnect between clinical trials and clinical practice: conclusions from many trials are based on photographic examinations at centralized reading centers, while clinical treatment decisions are made by physicians in practice who have neither the time nor the specific training to implement reading center protocols. Therefore, conclusions and recommendations from those studies may be improperly applied to patient care, directing some patients to unnecessary treatments and others to improper diagnosis or monitoring [8].

OCT is a widely available method of standardizing clinical data collection, and it is a more objective and quantitative modality than fundus photography. Current OCT instrumentation does not possess the

spatial resolution of conventional fundus imaging devices, but it can render cross-sectional and topographic findings in greater detail. In addition, the standardized numerical results provided by OCT examinations have the potential to facilitate both the conduct of research as well as the ultimate application of its recommendations to clinical practice. Strom and colleagues recently attempted to bridge the gap between traditional reading center assessments and modern OCT examinations by demonstrating good agreement between stereoscopic fundus photograph grading and OCT measurements for the detection of diabetic macular edema [98]. Although they suggest that stereo fundus photographs may be more sensitive than OCT for the detection of retinal thickening, it is likely that reproducible, quantitative results will ultimately establish OCT as a better objective evaluation of retinal thickness than expert subjective assessments of fundus images.

12.2.4 Fluorescein Angiography

Essentials

Fluorescein angiography and OCT each provide unique pieces of information that may be necessary for the appropriate management of patients with retinal vascular diseases

OCT retinal thickness measurements may correlate better with visual acuity in DME than fluorescein angiography

OCT retinal thickness measurements may not correlate well with visual acuity in venous occlusive disease

Fluorescein angiography (FA) is a highly effective method of evaluating retinal vascular disorders because it provides important information about macular perfusion and patterns of leakage. Many morphologic changes, such as intraretinal edema and neurosensory retinal detachments, are not specifically imaged with angiography, although the dynamic process of fluorescein leakage adequately demonstrates the underlying compromise of the blood-retinal barrier and disturbances in the equilibrium between fluid extravasation and reabsorption [67]. Furthermore, current methods of FA interpretation require expert knowledge and are subjective, qualitative, and time-consuming [32, 38].

OCT provides unparalleled cross-sectional rendering of in situ retinal morphologic changes such as cystoid edema and subretinal fluid. Although existing OCT instruments do not provide dynamic infor-

mation about fluid movement within or underneath the retina, they automatically extract reproducible measurements of retinal thickness and volume. In fact, many investigators have found that these parameters may have a stronger correlation with visual acuity in DME than FA findings [11, 25, 28, 29, 35, 44, 59, 66, 70]. Other groups have found close associations between FA and OCT findings [2]. One study by Kang and colleagues found good correlations between focal leakage on FA and intraretinal edema on OCT, as well as between diffuse cystoid leakage on FA and outer retinal or subretinal fluid accumulation on OCT [39]. In contrast to these observations, studies in venous occlusive disease have failed to demonstrate strong correlations between OCT measurements and FA findings [11, 45].

Therefore, it appears that FA and OCT each have a unique role in the assessment of retinal vascular diseases. Treatment decisions that are based on the presence of fluid within or under the retina may be better assessed by OCT, whereas disease entities characterized by dynamic leakage may be better assessed with angiography. Since neither modality currently provides a complete answer, concurrent FA and OCT may be required to sufficiently assess the extent of damage from retinal vascular diseases. However, further developments in OCT technology, such as Doppler OCT and OCT angiography, may provide a comprehensive solution in a single imaging modality.

12.2.5 Ultrasound

Essentials

The axial resolution of noninvasive B-scan ultrasound is less than OCT

Echography may be an option for the detection of macular edema in cases with poor posterior pole visibility or poor patient cooperation

B-scan ultrasonography is a useful method for imaging the vitreous, the peripheral retina, and the posterior pole in cases in which media opacities or patient cooperation prohibit adequate visualization through the pupil. The resolution of clinical echography devices, however, is limited by their operating frequency, and cannot approach the resolution of OCT when used to noninvasively image the posterior pole. In one study by Lai et al., B-scan ultrasonography was shown to have a high degree of sensitivity (91 %) and specificity (96 %) for detecting retinal thickening when compared to OCT measurements [41]. However, this same study demonstrated only a mod-

erate correlation (r = 0.65) between qualitative thickness measurements made with the ultrasound device and quantitative measurements made with OCT [41]. Therefore, echography has not been shown to

be as effective at quantifying macular thickening as II 12 OCT, and should most likely be reserved for use in

the select cases outlined above.

12.3 Diagnosis

OCT is a useful tool for the diagnosis and management of retinal vascular disorders because of its ability to measure retinal thickness and render intraretinal details. When coupled with other imaging modalities, such as FA, OCT may also be helpful in understanding the pathogenesis of the disorder and in optimizing treatments. OCT may also be useful for the diagnosis of macular edema in the presence of media opacities, such as asteroid hyalosis, that prevent accurate biomicroscopic assessment [7]. Furthermore, diagnoses such as the vitreomacular traction syndrome that may have been missed in past decades are readily apparent with cross-sectional imaging. And, even in cases with easily observable findings, OCT can be used to establish a baseline macular volume and monitor subsequent responses to treatment. Serial measurements can then be compared to values obtained from a normal population to enable earlier and more accurate diagnosis and treatment [28].

Retinal vascular disorders typically produce three categories of findings on OCT: retinal thickening, intraretinal cystoid changes, and/or serous detachments of the neurosensory retina. The pattern of findings on OCT can be helpful in making a diagnosis or in determining the patient’s prognosis, and will be discussed in more detail below [11, 70].

12.3.1 Intraretinal Edema

Essentials

Intraretinal edema is evident as retinal thickening associated with decreased optical backscattering

Branch retinal vein occlusions produce a characteristic asymmetric pattern of swelling

Retinal edema by OCT should be scrutinized for evidence of vitreomacular traction

Edema within the retina can be detected on OCT as an area of retinal thickening associated with decreased optical backscattering [11, 15, 28, 66]. When displayed in false color, these areas appear

darker than adjacent, nonedematous tissue (Fig. 12.1). If the edema occurs adjacent to or within the fovea, the typical foveal depression may be reduced, eliminated, or even inverted. Branch retinal vein occlusions (BRVOs) produce a characteristic, abrupt disparity in retinal thickness along the horizontal raphe that is characteristic of the asymmetric edema seen in this disorder (Fig. 12.2).

Otani’s definition of outer, sponge-like swelling further specifies that intraretinal edema is typically ill-defined and widespread [66]. This is in contrast to cystoid macular edema (discussed below), which may have more focal and well-defined areas of involvement. Thickening from primary intraretinal edema should be differentiated from the thickening that may occur secondary to retinal traction [15]. A brief examination of the vitreoretinal interface for the presence of an epiretinal membrane or evidence of traction from the posterior vitreous should elucidate this cause.

12.3.2 Cystoid Macular Edema

Essentials

Cystoid spaces are round, hyporeflective regions on OCT, typically in the outer retinal layers

Pseudophakic cystoid changes may occur in the inner retina

Findings on OCT correlate relatively well with fluorescein angiography

Cystoid changes are spread evenly across the macula in CRVO, while they respect the horizontal raphe in BRVO

Foveal cystoid changes may be a feature of idiopathic juxtafoveal telangiectasis

Cystoid macular edema (CME) may be a common final pathway for many retinal vascular diseases including diabetic macular edema, retinal vein occlusions, hypertensive retinopathy, and idiopathic juxtafoveal telangiectasis. Histopathologic reports indicate that intraretinal cystoid spaces in CME may vary in their location and size according to the