differential diagnosis

Is It a Nevus or Melanoma?

You can differentiate these extremely similar choroidal lesions. Here's how.

Kimberly K. Reed, O.D., F.A.A.O., Davie, Fla.

A Caucasian male in his 60s recently presented for his routine eye exam. The patient was relaxed and happy, and I was rewarded with a satisfying “Oh wow! That's so much clearer!” at the refraction's end. It was the kind of exam that makes you glad you practice optometry. But, during the posterior segment dilated fundus exam, my heart sank. There, as plain as day, was a large, raised lesion in the posterior pole. It was mostly non-pigmented, but it had a small area on the nasal border that was dusky-gray. I didn't detect any vascular abnormality on or near it, and there weren't drusen overlying it. Overall, it looked ominous. My mind began picking its way through its filed-away facts about choroidal lesions, so I could arrive at a diagnosis and determine the best management for the patient.

What follows are these filed-away facts and how I ultimately proceeded with the patient mentioned above. Keep in mind that a nevus can rarely convert to a melanoma, and a choroidal melanoma can result in vision loss and/or enucleation and even death. As a result, it's crucial you're prepared to diagnose these very similar-appearing, yet different conditions.

Review of anatomy

The choroid is the posterior aspect of the uveal tract, anteriorly comprised of the iris and ciliary body. Located in between the retina and the sclera, the choroid is very richly vascularized and contains pigment cells called melanocytes. These cells, along with the retinal pigment epithelium (RPE) cells, comprise the visible “color” of the fundus. In addition, melanocytes are atypical in both nevi and melanomas of the uveal tract, causing most of these lesions to look gray, blue-gray or greenish-gray, although some can be non-pigmented (amelanotic), like the lesion in my patient's eye.

Fundus photograph of a benign gray choroidal nevus in the inferior arcade. Note the small size and lack of orange pigment and subretinal fluid.

Because one of the most critical aspects in describing any posterior segment lesion is its size, it's useful to remember that the diameter of the optic disk is just about 1.6mm, on average. We are accustomed to estimating the size of retinal findings in disk diameters (DD), but it's more clinically accurate to estimate, or, better yet, measure and record the size of suspicious areas in millimeters, in at least two dimensions. The disk is a handy estimation tool, but it's not a precisely accurate “ruler” for measuring lesion size. For example, you might note a pigmented lesion that's roughly 1DD long by roughly 2DD wide. You could also record this as “approximately 1.5mm vertical x 3mm horizontal.” Even better, digital fundus photography allows a much more accurate measurement.

Choroidal nevi

Choroidal nevi are most commonly seen in Caucasians, with an incidence of about 6% to 7% of that population.^1,2 They are less frequently seen in Hispanics and Asians and least frequently seen in blacks. The characteristics of choroidal nevi:

► Spot-like appearance. In more than 90% of cases, nevi appear as “dark spots” that can be gray, blue-gray, greenish-gray or brownish-gray.² The remaining ones are considered “amelanotic,” because they don't display the typical dark coloration.

► Drusen. This suggests chronicity and supports the nevus diagnosis. Still, drusenoid lesions on top of a choroidal nevus may actually be small RPE detachments, which are common to melanomas.

► Small in diameter. On average, nevi tend to be roughly 1DD in size, or about 1.5 mm.¹

► Tend to be symptomless. It is rare for a choroidal nevus to cause symptoms. If, however, the lesion leaks retinal fluid, which is associated with abnormal blood growth, the patient can experience starbursts and vision loss due to a retinal detachment or degeneration. Again, this is far more common with melanomas than with nevi.³

Fundus photograph of a suspicious choroidal nevus in a subfoveal location. This patient's nevus was suspicious due to the presence of orange pigment, subretinal fluid (note the track of old fluid inferotemporal to the nevus), location in the posterior pole and symptoms of metamorphopsia (due to the subretinal fluid).

Choroidal nevi can present with atypical characteristics as well. For instance, although we've been taught that lipofuscin (orange pigment) at or on a presumed choroidal nevus is actually a choroidal melanoma, this is not an absolute diagnostic sign of malignancy, as some choroidal nevi have lipofuscin as well.¹ Keep in mind, however, that its presence is a risk factor for malignant transformation to a melanoma. The presence of lipofuscin is a sign of cellular death, commonly seen in association with macular degeneration. It is best seen with fundus autofluorescent (FAF) imaging.⁴ Also, nevus growth is not a firm diagnostic indicator of a choroidal melanoma. Some nevi do “grow” through time, but those that do, do it very slowly. In addition, although choroidal nevi are generally “flat,” they can be slightly raised. In fact, the thickness cutoff point for choroidal nevi is generally accepted to be about 2.0mm.^1,2,5,6 When you think about the size of the optic disk, that's actually fairly thick. Finally, subretinal fluid (SRF) is likely the best indicator that a choroidal nevus may actually be a melanoma. However, SRF can also appear in choroidal nevi, particularly when they have associated subretinal neovascularization.

Through about five years, 2.6% of all nevi, and 10.3% of “suspicious” nevi, converted into melanomas.^6-8

Fundus autofluorescent imaging (FAF) of a small, amelanotic choroidal melanoma in the superior arcade. Note the bright clusters, which represent lipofuscin on the evolving collar button of the tumor.

Choroidal melanomas

Choroidal melanomas are the most common primary malignant intraocular tumors, with about 1,500 new cases reported every year in the U.S.⁹ (Of interest, beneath the ozone hole in Australia, there is almost twice the per capita incidence. This suggests a role for ultraviolet irradiation and sunglass protection.) Choroidal melanomas are the second most common primary malignant melanoma in the body. Estimated five-year survival rates are between 70% to 90%, depending primarily, but not exclusively, on the choroidal melanoma's size when first diagnosed.^9,10 In general, though, the larger the choroidal melanoma, the worse the prognosis for both vision and metastasis.¹¹

Although melanomas can form “de novo,” meaning they can start as malignant lesions, most arise from pre-existing choroidal nevi. The following are the characteristics of choroidal melanomas:

► Spot-like appearance. Just like nevi, melanomas are usually dark in color, but occasionally can be amelanotic.⁹

► Large in diameter. Choroidal melanomas are, on average, larger in size than nevi. Most are greater than 7mm in diameter. Therefore, a large nevi should place it in the suspicious or possible melanoma category.^1,2

► 2.0mm thickness or greater. Melanomas tend to be thicker in size than nevi, specifically, 2.0mm and greater.

► Asymmetry. One half of the lesion tends to look different than the other half of the lesion. Still, this should not be used as an independent diagnostic marker.

► Border irregularity. The edges of the lesion tend to be notched, uneven or blurred, though this characteristic should not be used as an independent diagnostic marker either.

► Color irregularity within the lesion. Some melanomas are uniform in color, but more often they appear with areas that are light, medium and/or darkly pigmented, randomly distributed throughout the lesion.

► Lipofuscin is likely present. Most melanomas have lipofuscin on their surface. That said, because lipofuscin can also be present in choroidal nevi, you should not use this as an independent diagnostic marker.

Fundus photograph of a large mushroom-shaped choroidal melanoma in the superior quadrant with a collar button and secondary serous retinal detachment.

► SRF is likely present. This is another frequent sign, though as is the case with lipofuscin, SRF alone does not define malignancy.¹²

► Dome-shaped. Most melanomas are dome-shaped.^7,13 However, when melanomas grow in apical height, or toward the vitreous cavity, they can break through Bruch's membrane. When this occurs, the shape of the tumor may look like a mushroom or “collar-button” with the “stalk” located in the choroid and passing through Bruch's membrane, and the mushroom's “cap” acting as the extension of the tumor as it grows outward. The result: Nothing can restrict its growth pathway in the retina. Mushroom-shaped tumors are almost always melanomas, and a collar-button is diagnostic of a melanoma.

► Tend to be symptomless. Choroidal melanomas, as is the case with choroidal nevi, are typically symptomless. That said, if a melanoma is located near the lens, it could push the natural lens and cause irregular astigmatism. In addition, it can leak fluid below the retina, causing a retinal detachment and subsequent starbursts and floaters. Also, if the melanoma is in the macula, it can grow below the fovea, creating hyperopia. Further, it can grow into and destroy the fovea, inflicting metamorphopsia, vision loss and/or color perception changes.¹¹

► Rapid growth. Choroidal melanomas are characterized by rapid enlargement.

As is the case with choroidal nevus, a choroidal melanoma can present with atypical characteristics as well. These include sentinel blood vessels, vitreous hemorrhage or a dark brown spot on the sclera (indicative of extrascleral extension).

I've had success in diagnosing lesions as “suspicious” for choroidal melanoma by using the acronym “SPOTS,” which is based on several researchers' data.^7,14,15 The first “S” stands for symptoms. If the patient has any of the outlined symptoms, this is not a good sign. “P” is for position in the fundus. Statistically, lesions in the posterior pole, and especially those adjacent to or even touching the optic disk, carry a higher risk factor for a choroidal melanoma. “O” stands for orange pigment, a suggestive sign. “T” is for thickness, using 2.0mm as the defining line between likely benign and likely malignant. The last “S” stands for SRF. One of these signs prompts suspicion. The more of these you rack up, however, the more likely the diagnosis is melanoma rather than nevus.

Required diagnostic tools/management

Because most patients who have a choroidal nevus or choroidal melanoma are symptomless, the diagnosis of either is usually made at the annual eye exam. The following diagnostic devices matched with the knowledge of the clinical features of each condition will help you to achieve a definitive diagnosis.

► Fundus photography. Due in part to the advent of low-cost digital photography, it is currently the standard of care to photograph choroidal nevi. Because choroidal nevi can convert into melanomas, and lesion size plays major role in differential diagnosis, a baseline fundus photograph is essential to monitoring for growth. Further, it is essential you fully evaluate the fundus of each and every patient, and carefully note the size of any potentially suspicious lesions. Several retinal imaging devices include image comparison software, enabling you to carefully monitor lesions for any increase in size. Or, you can use detailed fundus drawings, carefully marking the position of overlying retinal vasculature for monitoring lesion size increase at follow-up visits. Often, it isn't easy to obtain reimbursement for serial photographs. Your experience will vary, but if your clinical evaluation suggests an ophthalmoscopic change, prompting confirmation with photographs, it might make any appeals for reimbursement denial easier. (See “How do I Code for That?” below.)

If the lesion is small, flat and doesn't have any of the other “SPOTS” characteristics, it's safe to monitor the patient carefully for progression. Follow-up between three months for new onset or suspicious lesions and 12 months for older established lesions can be justified based upon the individual circumstances. Large, more suspicious lesions, particularly those that are elevated, are typically evaluated with B-scan ultrasound, OCT and fluorescein and indocyanine green angiography as well as FAF imaging. Patients with suspicious lesions should be referred to a retinologist or ocular oncologist for further evaluation and possible treatment, especially when the lesion carries risk factors for transformation. Unless you routinely perform ultrasonography, a raised lesion may prompt a referral as well. In addition, these patients are often referred for an abdominal imaging study (MRI, CT or ultrasound), as the liver is the most likely site for metastic melanoma development. Depending on your comanagement relationships, you can either arrange this testing yourself, or work through the patient's primary care provider.

► B-scan ultrasound. B-scan ultrasound, with a superimposed A-scan, provides information about the “acoustics” of the inside of the lesion. Choroidal melanomas are classically described as having low-to-medium internal reflectivity. This can differentiate them from highly reflective hemangiomas or variably reflective metastic choroidal tumors. When a mass is echolucent, or acoustically hollow, it is likely filled with fluid, like a choroidal or retinal detachment. Other less common B-scan ultrasound findings associated with choroidal melanoma are choroidal excavation (into the sclera), extrascleral tumor extension, exudative retinal detachment and vascularity (seen on dynamic examination.)⁷

► OCT. This is the best way to see small amounts of subretinal or intraretinal fluid. OCT may also be helpful in confirming that a suspicious lesion isn't located in the retina. Enhanced depth OCT (EDI-OCT) has also shown to be useful for measuring choroidal nevi, though it is not yet widely available.¹⁶ Specifically, EDI-OCT was shown to depict the exact location of a lesion, its dimensions, depth and demarcation from the surrounding healthy tissue.

► FAF and angiograhy. These tests can evaluate internal vascular patterns that may help differentiate melanomas from other choroidal tumors and uncover obscure tumor margins. For instance, in addition to demonstrating the presence of orange pigment, FAF can help to show that a tumor is exhibiting leakage, which is a risk factor for malignant transformation. Angiography is also helpful in demonstrating that a suspicious tumor has intrinsic vascularity or a “double circulation pattern.” However, many choroidal nevi and small choroidal melanomas have similar angiographic pattens. Angiography is also helpful in monitoring post-treatment tumor progression and for radiation retinopathy.

Many practitioners wonder whether CT Scan and/or MRI should be employed for “making the diagnosis.” While large, highly vascularized tumors will appear with characteristics that are strongly suggestive of melanoma, other malignancies will not, making the use of CT and MRI not particularly helpful in those ophthalmoscopically borderline cases.

How do I Code for That?
The following are ICD-9 and CPT codes for various diagnoses and procedures. (Check cms.gov for updates to the ICD-10 coding, and use any modifiers as required.) Average reimbursement is an informal average based on a Google search.

Choroidal melanoma treatment

For many years, the only treatment option was enucleation.¹⁷ Biopsy is not performed to differentiate between choroidal nevi and melanomas. This is because of the small size of choroidal nevi, concerns about potential “seeding” of tumor cells to adjacent tissues or to the blood supply as well as intraocular complications. Further, due to advances in methods of diagnosis and testing, few eyes are enucleated that ultimately don't contain melanomas. For example, the Collaborative Ocular Melanoma Study (COMS) reported a mere 0.4% misdiagnosis rate.

Though “seeding” is a worry with biopsies of all types, no published evidence exists that biopsy increases the chance that during the procedure one could disrupt and dislodge tumor cells from the lesion itself into the surrounding tissue, or worse yet, into the blood supply.

The COMS has shown that treatment by surgical placement of a radioactive plaque directly beneath the tumor yields nearly the same prognosis for mortality, as does enucleation for medium-sized tumors. As a result, plaque placement is now a preferred treatment for lesions of medium size. Large tumors are typically treated by enucleation, though many can now be treated with globe-sparing larger plaques.

Other less commonly employed treatment options: proton beam irradiation, block excision of the tumor, laser photocoagulation, transpupillary thermotherapy laser and orbital exenteration, the latter of which is a last-resort for tumors showing widespread orbital extension.¹⁸

My patient

The aforementioned patient's lesion is definitely in the “suspicious” category. After evaluation with OCT, digital photography, and ultrasonography, it was determined that the patient's lesion was about 1.0mm in apical height, which is less than the 2.0mm “cutoff” for high suspicion of malignancy. However, I did obtain a second opinion, given the overall lesion size, which was very close to 5.0mm. As a result, we have decided to take a watchful, waiting approach, and I've scheduled the patient for a follow-up appointment in three months. OM

Special thanks to Eyecancer.com, Paul Finger, M.D. and Kimberly Chin, O.D. for their input on this article.

References

1. Mashayekhi A, Siu S, Shields CL, et al. Slow enlargement of choroidal nevi: a long-term follow-up study. Ophthalmology 2011 Feb; 118(2):382-8.
2. Sumich P, Mitchell P, Wang JJ. Choroidal nevi in a white population: the Blue Mountains Eye Study. Arch Ophthalmol 1998 May;116(5):645-50.
3. The Eye Cancer Network. Choroidal Nevus. www.eyecancer.com/patient/Condition.aspx?nID=60&Category=Choroidal+Tumors&Condition=Choroidal+Nevus. Accessed July 25, 201.
4. Chin K. Finger PT. Autofluorescence characteristics of suspicious choroidal nevi. Optometry 2009; Mar;80(3):126-30.
5. You QS. Xu L, Jonas JB, et al. Change in choroidal nevi during a 5-year follow-up study: the Beijing Eye Study. Br J Ophthalmol 2010 May;94(5):575-8.
6. Kaiserman I, Kaiserman N, Pe'er J. Long term ultrasonic follow up of choroidal naevi and their transformation into melanomas. Br J Ophthalmol 2006 Aug;90(8):994-8.
7. Gass JD. Problems in the differential diagnosis of choroidal nevi and malignant melanoma. XXXIII Edward Jackson Memorial lecture. Trans Sect Ophthalmol Am Acad Ophthalmol Otolaryngol. 1977 Jan-Feb;83(1):19-48.
8. Augsburger JJ. Is observation really appropriate for small choroidal melanomas? Trans Am Ophthalmol Soc 1993;91:147-51.
9. The Collaborative Ocular Melanoma Study Group. Factors predictive of growth and treatment of small choroidal melanoma. COMS Report #5, Arch Ophthalmol 1997 Dec;(115):1537-1544.
10. Damato B, Eleuteri A, Taktak AF, et al. Estimating prognosis for survival after treatment of choroidal melanoma. Prog Retin Eye Res. 2011 Sep 30(5):285-95.
11. Eye Cancer Network. Choroidal Melanoma. www.eyecancer.com/Patient/Condition.aspx?nID=62&Category=Choroidal+Tumors&Condition=Choroidal+Melanoma. Accessed July 25, 2011.
12. The Collaborative Ocular Melanoma Study Group. Arch Ophthalmol 2001 Jul; 119(7):951-965.
13. Newman H, Chin KH, Finger PT. Subfoveal choroidal melanoma: pretreatment characteristics and response to plaque radiation therapy. Arch Ophthalmol 2011 Jul;129(7):892-8.
14. Shields CL, Shields JA, Kiratli H, et al. Risk factors for growth and mestasasis of small choroidal melanocytic lesions. Ophthalmology 1995 Sep;102(9):1351-61.
15. Shields CL, Cater J, Shields JA, et al. Combination of clinical factors predictive of growth of small choroidal melanocytic tumors. Arch Ophthalmol 2000 Mar; 118(3):360-4.
16. Basdekidou C, Wolff B, Vasseur V, et al. Flat Choroidal Nevus inaccessible to ultrasound sonography evaluated by enhanced depth imaging optical coherence tomography. Case Report Ophthalmol. 2011 May;2(2):185-8.
17. Finger PT. Radiation therapy for choroidal melanoma. Surv Ophthalmol. 1997 Nov-Dec;42(3):215-32.
18. The Collaborative Ocular Melanoma Study Group. http://www.jhu.edu/wctb/coms/. Accessed July 24. 2011.

Dr. Reed is an associate professor and director of Educational Effectiveness & Outcomes Assessment at the Nova Southeastern University College of Optometry in Fort Lauderdale, Fla. E-mail her at kimreed@nova.edu, or send comments to optometricmanagement@gmail.com.