The diagnosis of choroidal melanoma is clinical and supported by ancillary imaging findings. Recent advances in multimodal imaging allow for more accurate tumor characterization and a sophisticated approach to monitoring tumor response to therapy. Fundus photography, ultrasonography, fundus autofluorescence, angiography and optical coherence tomography play a central role in the practical management of choroidal melanoma. In particular, the development of ultra-widefield imaging systems and the expanded uses of OCT have impacted patient care significantly. Here, I’ll review the state of the art when it comes to imaging choroidal melanoma.
​

Fundus Photography
​

 
 
Figure 1. Choroidal melanoma in the macula captured with standard color fundus photography documents typical features: a dome-shaped pigmented choroidal mass with overlying orange pigment, the absence of drusen and the presence of subretinal fluid.

Since the development of the fundus camera in the early 1900s, progress in the field of digital imaging has revolutionized the ability to capture clinical observations. The early fundus camera introduced by Zeiss and Nordensen in 1926 provided a 20-degree field of view.1 Standard modern mydriatic fundus cameras now capture 30 to 50 degrees of the fundus and allow for excellent documentation of features pertinent to choroidal melanoma, such as the degree of pigmentation, the presence or absence of drusen, lipofuscin (or “orange pigment”), overlying changes in the retinal pigment epithelium, and subretinal fluid (Figure 1). The standard field can be increased by fusing images to create a montage. In the Diabetic Retinopathy Study, “7-standard” overlapping 30-degree fields were viewed collectively to provide 75 degrees of the fundus.2 Standard color fundus photography provides advantages, including accurate depiction of color and high image resolution; however, standard fields aren’t ideal when choroidal melanoma is extensive or located in the anterior fundus.
​

More recently developed non-mydriatic, ultra-widefield imaging systems capture up to 200 degrees of the fundus in a single image. For peripherally located tumors, these systems allow for more complete documentation of tumor margins and associated features such as exudative retinal detachment, particularly when multiple clock hours of the fundus are involved (Figure 2). The ability to accurately photograph tumors located anterior to the equator at regular time intervals following primary therapy is important for assessing response to therapy and monitoring for recurrence at tumor margins.3 While advantageous in this regard, ultrawidefield imaging has limitations. In comparison to standard color fundus photography, ultra-widefield systems have less accurate color representation. Additionally, the spherical shape of the globe results in distortion of the peripheral image, producing artifactual changes in the shape and apparent size of tumors located more anteriorly. This can be overcome by using a multimodal approach, such as combining clinical impressions from ophthalmoscopy and tumor dimensions from ultrasonography to more accurately characterize overall tumor size.
​

Fundus Autofluorescence
​

FAF can be performed using both standard and ultra-widefield imaging systems. This modality can be particularly helpful in confirming the presence of overlying lipofuscin, a frequent clinical feature of choroidal melanoma. With FAF, hyper-autofluorescence results when lipofuscin is exposed to intense blue light with a wavelength of 488 nm.4 Lipofuscin accumulates in the retinal pigment epithelium and in macrophages, but its appearance is variable depending upon the pigmentation of the underlying lesion. For highly melanocytic tumors, lipofuscin will characteristically be bright orange. For more lightly pigmented or amelanotic choroidal melanoma, lipofuscin may appear “ruddy” brown. Other benign tumors, such as circumscribed choroidal hemangioma, can also have overlying lipofuscin, but its presence may be nearly undetectable by ophthalmoscopy. FAF is useful for identifying lipofuscin, and although not pathognomonic for choroidal melanoma, its presence is supportive of the diagnosis.5

Angiography
​

 
 
Figure 2. Ultra-widefield color fundus photography completely captures a choroidal melanoma in which the anterior margin is overlying the equator. Following treatment with I-125 plaque brachytherapy, The tumor demonstrates regression with a ring of surrounding chorioretinal atrophy.

Fluorescein angiography is available in both standard format and in ultrawidefield imaging platforms that offer 100to 200-degree fields of view. Ultra-widefield systems may provide superior capture ability for more peripherally located choroidal melanoma (Figure 3). FA is also useful for assessing and monitoring treatment-related side effects in the retina, such as capillary changes and areas of radiationinduced non-perfusion. Indocyanine green angiography, also available in standard and ultra-widefield formats, provides superior characterization of the choroidal vasculature and is particularly helpful for documenting features of choroidal melanoma (Figure 4), such as intrinsic vascularity (the “double-circulation sign”).
​

Optical Coherence Tomography
​

 
 
Figure 3. Ultra-widefield fluorescein angi- ography of choroidal melanoma. Intrinsic vasculature is present within the tumor.

Since its introduction, OCT has become widespread in ophthalmic practice.6 Spectral-domain OCT has essentially replaced conventional timedomain systems due to its superior speed, sensitivity and resolution. In particular, enhanced-depth-imaging OCT has become a valuable tool for characterizing the choroid.7 This technique takes advantage of the increased depth of field in the inverted image that results from placing the SD-OCT device closer to the eye.
​

While ultrasonography remains the gold standard for measuring tumor dimensions, EDI-OCT can be particularly helpful for evaluating small choroidal melanoma. Ultrasonography provides millimeter-level resolution, but for tumors less than 1 mm in thickness, EDI-OCT can provide micronlevel information related to tumor thickness and surface topography (Figure 5). In such cases, EDI-OCT has the potential to visualize a tumor completely and to differentiate it from the normal surrounding choroid.8 In one series of 23 lesions composed of amelanotic choroidal nevus, melanocytic choroidal nevus, choroidal melanoma, circumscribed choroidal hemangioma and choroidal metastasis, tumor thickness could be measured by EDI-OCT, but not by ultrasonography in 10 cases (in each case, lesions were less than 1 mm thick).8 EDI-OCT image quality is affected by tumor pigmentation, however. Amelanotic tumors may be somewhat easier to characterize, as they demonstrate less shadowing artifact and have a more homogenous, medium reflectivity.7 Highly pigmented lesions are more likely to demonstrate posterior shadowing, which compromises visualization beyond the anterior tumor surface.9
​

EDI-OCT has also been used in an attempt to distinguish between small choroidal melanoma and suspicious choroidal nevi. In one series of 37 eyes with small choroidal melanoma, choroidal shadowing and choriocapillaris thinning was found all cases.10 Compared with choroidal nevi of similar size, statistically significant EDI-OCT features of small choroidal melanoma included: intraretinal edema; loss of photoreceptors; loss of the external limiting membrane; loss of the inner segment-outer segment junction; irregularity of the inner plexiform layer; and irregularity of the ganglion cell layer. Elongated photoreceptors were observed overlying small choroidal melanoma in 49 percent of eyes, but this finding wasn’t observed in choroidal nevi.10
​

Swept-source OCT has recently been used to characterize choroidal lesions and may provide some advantages for pigmented tumors.11 SS-OCT uses a wavelength-tunable laser and a dual-balanced photodetector that provides superior imaging speed. Its adaptability to longer wavelengths allows for imaging of the choroid and greater penetration of melanin.12 In one series of 30 eyes with choroidal nevi, SS-OCT enabled visualization of intralesional details such as vessels, cavities and granularity. For melanocytic nevi, SS-OCT was superior for depicting intralesional characteristics (vessels, granularity, abnormal choriocapillaris) compared to EDI-OCT.11 In another series of 85 choroidal lesions evaluated by SS-OCT, the majority of melanocytic choroidal lesions could be successfully characterized. In this series, multivariable analysis revealed several factors significantly associated with optimal image quality, including: lesion location closer to the fovea; lighter pigmentation; and smaller diameter.13

 
 
Figure 4. Ultra-widefield indocyanine green angiography of choroidal melanoma demonstrating intrinsic vasculature.

The newest form of OCT, optical coherence tomography angiography, is a non-invasive (i.e., dye-free) technique for imaging the retinal and choroidal vasculature. OCTA collects information regarding retinal and choroidal blood flow by comparing consecutive B-scans. OCTA may also help to distinguish choroidal melanoma from benign nevi by identifying unique vascular patterns. In a series of 11 patients (six with choroidal melanoma and five with choroidal nevus), OCTA demonstrated a hyporeflective mass with no significant deformity of the choroidal vasculature and an intact retinal pigment epithelium-Bruch’s membrane complex in all cases of choroidal nevus. In contrast, OCTA demonstrated an obscured RPE-Bruch’s membrane complex and outer retinal layer in cases of choroidal melanoma.14 The average choriocapillaris flow rate associated with choroidal melanoma was only 55.7 percent, compared to the normal choriocapillaris’s flow rate of 62.8 percent (p=0.01). Additionally, axial and peripheral feeder blood vessels were noted to be more dilated and tortuous for eyes with choroidal melanoma compared with benign nevi.14

 
 
Figure 5. EDI-OCT of a suspicious pigmented choroidal lesion with overlying orange pigment and subretinal fluid. The lesion demonstrated significant and quantifiable growth over the course of six months and was ultimately treated with I-125 plaque brachytherapy.

OCTA can also be useful for both qualitatively and quantitatively assessing iatrogenic changes within the retina and choroid following treatment for choroidal melanoma. Images can be segmented to determine the degree of capillary dropout and nonperfusion in the superficial and deep layers of the retina, and in the choriocapillaris (Figure 6). A recent cross-sectional study compared OCTA findings in 10 eyes with choroidal melanoma imaged prior to therapy and 15 irradiated eyes with clinically apparent radiation retinopathy and/or optic neuropathy. In eyes with radiation-induced side effects, peripapillary retinal capillary density (PPCD) was lower in the treated eye and correlated with the radiation dose to the optic nerve as well as with the visual acuity outcome. In contrast, no significant difference was observed in PPCD in eyes with melanoma prior to irradiation compared with normal fellow eyes.15

Future Directions
​

 
 
Figure 6. OCTA showing a progressive decrease in macular capillary density 6 (A, D), 18 (B, E), and 24 (C, F) months following treatment with proton beam irradiation in the superficial retinal layers (top row) and deep retinal layers (bottom row).

The widespread use of multimodal imaging for choroidal melanoma allows for superior characterization of tumors in a field that’s continually evolving. No one imaging technique alone serves as a gold standard for management. It’s the combination of imaging modalities that allows for superior tumor assessment and a more thorough understanding of the clinical features of choroidal melanoma. For diagnostically challenging cases, newer imaging strategies may provide a non-invasive means to differentiate choroidal melanoma from benign and simulating lesions. Multimodal imaging also allows greater opportunity to study treatment-related side effects. This is especially relevant, as radiation-dose protocols continue to be refined,16 newer therapies will arise,17 and the role of anti-vascular endothelial growth factor in preventing radiation retinopathy is currently being investigated.18 Newer technologies will address issues such as the more accurate depiction of colors, particularly for ultra-widefield systems; the development of smaller and more portable devices; and cost-effectiveness.
​

Dr. Aronow is an assistant professor of ophthalmology at the Ocular Melanoma Center and Retina Service at Massachusetts Eye and Ear/Harvard Medical School. She can be reached at mary_aronow@meei.harvard.edu. She has no financial interest in any of the products discussed in the article.
​

1. Ciardella A, Brown D. Wide field imaging. In: Agarwal A, ed. Fundus Fluorescein and Indocyanine Green Angiography: A Textbook and Atlas. New York: Slack, 2007:79-83.
2. The Diabetic Retinopathy Study Research Group. Diabetic retinopathy study. Report Number 6: Design, methods, and baseline results. Report Number 7: A modification of the Airlie House classification of diabetic retinopathy. Invest Ophthalmol Vis Sci 1981;21:1-226.3. Mirchia K, Turell ME, Singh AD. Imaging modalities for uveal melanoma. European Ophthalmic Review 2012;6:1:56-63.
4. Schmitz Valckenberg S, Holz FG, Bird AC, et al. Fundus autofluorescence imaging: Review and perspective. Retina 2008;28:385-409.
5. Shields JA, Rodrigues MM, Sarin LK, et al. Lipofuscin pigment over benign and malignant choroidal tumors. Trans Sect Ophthalmol Am Acad Ophthalmol Otolaryngol 1976;81:5:871-81.
6. Huang D, Swanson EA, Lin CP, et al. Optical coherence tomography. Science 1991;254:1178-81.
7. Spaide RF, Koizumi H, Pozzoni MC. Enhanced depth imaging spectral-domain optical coherence tomography. Am J Ophthalmol 2008;146:4:496-500.
8. Torres VL, Brugnoni N, Kaiser PK, et al. Optical coherence tomography enhanced depth imaging of choroidal tumors. Am J Ophthalmol 2011;151:586-93.
9. Shah SU, Kaliki S, Shields CL, et al. Enhanced depth imaging optical coherence tomography of choroidal nevus in 104 cases. Ophthalmology 2012;119:1066-72.
10. Shields CL, Kaliki S, Rojanaporn D, et al. Enhanced depth imaging optical coherence tomography of small choroidal melanoma: Comparison with choroidal nevus. Arch Ophthalmol 2012;130:850-6.
11. Francis JH, Pang CE, Abramson DH, et al. Swept-source optical coherence tomography features of choroidal nevi. Am J Ophthalmol 2015;159:1:169-76.
12. Mrejen S, Spaide RF. Optical coherence tomography: Imaging of the choroid and beyond. Surv Ophthalmol 2013;58:387-429.
13. Ung C, Laíns I, Silverman RF, et al. Evaluation of choroidal lesions with swept-source optical coherence tomography. Br J Ophthalmol 2018 Mar 31 [Epub ahead of print].
14. Ghassemi F, Mirshahi R, Fadakar K, Sabour S. Optical coherence tomography angiography in choroidal melanoma and nevus. Clin Ophthalmol 2018;12:207-14.
15. Skalet AH, Liu L, Binder C, et al. Quantitative OCT angiography evaluation of peripapillary retinal circulation after plaque brachytherapy. Ophthalmol Retina 2018;2:3:244-50.
16. Patel AV, Lane AM, Morrison MA, et al. Visual outcomes after proton beam irradiation for choroidal melanomas involving the fovea. Ophthalmology 2016;123:2:369-77.
17. Kines RC, Varsavsky I, Choudhary S. et al. An infrared dyeconjugated virus-like particle for the treatment of primary uveal melanoma. Mol Cancer Ther 2018;17:2:565-74.
18. Kim IK, Lane AM, Jain P, et al. Ranibizumab for the prevention of radiation complications in patients treated with proton beam irradiation for choroidal melanoma. Trans Am Ophthalmol Soc 2016;114:T2.