Example of choroidal nevus and photographic field from the Singapore Malay Eye Study.
Example of choroidal nevus and photographic field from the Blue Mountains Eye Study.
Ng CH, Wang JJ, Mitchell P, Amirul Islam FM, Wong TY. Prevalence and Characteristics of Choroidal Nevi in an Asian vs White Population. Arch Ophthalmol. 2009;127(3):314-319. doi:10.1001/archophthalmol.2008.625
To describe the prevalence and characteristics of choroidal nevi in an Asian population and compare this with findings from a white population.
The Singapore Malay Eye Study (SiMES) examined a population-based, cross-sectional, age-stratified, random sample of 3280 Malay persons (78.7% participation rate) aged 40 to 80 years living in Singapore. Comprehensive examination of participants included bilateral retinal photography. Choroidal nevi were graded from photographs using the Blue Mountains Eye Study (BMES) protocol.
The person-specific prevalence of choroidal nevi was 1.4%, with 50 nevi found in 45 participants. This is lower than the 6.5% prevalence seen in white persons in the BMES. However, characteristics of nevi in Malay persons in the SiMES were similar to those of white persons in the BMES by size (SiMES, 1.27 mm; BMES, 1.25 mm; P = .35), shape (P = .58), color (P = .39), location within posterior pole or periphery (P = .30), and nevus margin proximity to the optic disc (P = .29). Features previously identified as indicating growth or malignant potential (including diameter >6 mm, posterior margin touching optic disc, orange pigment, pigment clumping, and retinal edema) were not found in this sample.
Choroidal nevi were detected in 1.4% of Malay persons. There were no significant racial or ethnic differences in nevi characteristics between Malay and white persons.
Choroidal nevi are commonly observed as benign ocular tumors presenting as round or ovoid lesions of increased choroidal pigment, brown-gray or green-gray in color, with detectable but not sharp borders and possible elevation of up to 2 mm.1- 3In clinic-based studies1,4- 10and autopsy series,11,12the reported prevalence of nevi has ranged from 0.2% to 30%. However, there are few population-based data on the prevalence of choroidal nevi. In the only 2 population studies to date on white persons, the prevalence of choroidal nevus was reported to be 1.9% in subjects older than13 years when detected via ophthalmoscopic examination,13but as high as 6.5% in subjects aged 49 years or older when detected from the grading of retinal photographs.3In the only population study of nevi in an Asian population, the Beijing Eye Study reported a prevalence of 2.9% in adult Chinese persons, detected using retinal photographic grading.14
Although largely benign, the clinical significance of choroidal nevi lies in their potential for growth and, rarely, malignant transformation.1,8,15- 22Nevus characteristics associated with growth and malignant potential include larger size (greater diameter and thickness), location (close proximity to optic disc), presence of pigment (orange pigment and pigment clumping), subretinal fluid, and accompanying visual symptoms.15- 18,21,23- 26
It has been suggested that choroidal nevi are uncommon in persons who are not white.27However, there is a paucity of population-based data on the prevalence or characteristics of choroidal nevi in populations that are not white, including Asians. In this article, we describe the prevalence and characteristics of choroidal nevi in an Asian Malay population in Singapore and compare these data with the white population in the Blue Mountains Eye Study (BMES).3
The Singapore Malay Eye Study (SiMES) is a population-based cross-sectional study of 3280 Malay adults aged 40 to 79 years. The Malay race is the third largest ethnic group in Singapore, accounting for more than 4% of the world population with nearly 400 million Malay persons living in Southeast Asia alone.28,29Details of the SiMES design and methodology are reported elsewhere.29- 34Age-stratified random sampling of all Malay adults residing in the southwestern part of Singapore was conducted, selecting 1400 names from each decade (ages 40-49, 50-59, 60-69, and 70-79 years), with an initial 5600 names selected. Of these, 4168 individuals (74%) were eligible to participate. Persons were ineligible if they had moved from their residential address, had not lived there in the last 6 months, or were deceased or terminally ill. Of 4168 eligible individuals, 3280 participated in the study (78.7%); 831 nonparticipants (20%) declined and 57 (1%) were not contactable. Nonparticipants tended to be in the older age group (70-79 years) compared with participants.
All participants underwent a standardized interview and detailed clinical slitlamp examination. Sociodemographic and medical information was obtained, including smoking history and comorbidities.29,31,34,35The hospital institutional review board approved the study, conducted in accordance with the Declaration of Helsinki, with written informed consent obtained from participants
A standardized protocol was used for retinal photography.29,36After pharmacological pupil dilation, two 45° photographs were taken of each eye, one centered at the optic disc and another centered on macula, using a digital retinal camera (Canon CR-DGi with 10D SLR back; Canon, Tokyo, Japan).
A choroidal nevus was defined as an unequivocal pigmented slate blue or green-gray choroidal lesion measuring at least 500 μm in diameter. Choroidal lesions resembling nevi that were partially depigmented were graded as patchy hypomelanotic nevi. Congenital hypertrophy of the retinal pigment epithelium, pigment clumps, and pigmented scars were excluded. There were no melanomas detected in this Malay population.
The nevus features recorded included the number, shape (oval, round, or irregular), color (slate blue, green-gray, or hypomelanotic), location (posterior pole, periphery or overlap of posterior pole, or periphery), position relative to fovea, proximity of posterior margin to the optic disc, quadrantal distribution relative to the optic disc, and presence of orange pigment, pigment clumping, retinal edema, and drusen. Nevus diameter and surface area were measured using Adobe Photoshop CS2, with micrometer per pixel conversion calculated using 4500 μm as the standard distance from the center of the optic disc to the center of the fovea. Nevi were assessed as being macular, subfoveal, or subfoveolar when positioned within a 3000-μm, 750-μm, or 175-μm radius from the fovea, respectively. Drusen characteristics assessed included maximum size (63 μm, C0[distinct and indistinct]; 125 μm, C1; and 250 μm, C2, as used for grading of drusen in age-related maculopathy37), area of nevus involved by drusen, and placement of drusen over the nevus (central, peripheral, or both).
Choroidal nevi were analyzed as a binary outcome variable. Associations with age and sex were analyzed using χ2tests and reported as numbers and proportions. Age-standardized prevalence rates for men and women were computed using direct standardization of our study sample to the Singapore 2000 Census data of Malay citizens and permanent residents.38Nevus characteristics such as size, shape, color, and location were reported in numbers and percentages. Findings were compared with raw data obtained from the BMES in participants of the same age by using the χ2test of proportion for each category. The Fisher exact test was used for any cell frequencies less than 5. Nevus associations with other variables (including cataract, age-related macular degeneration, diabetes, and smoking) were assessed using logistic regression. The association between nevus size and age was assessed using the χ2test. P ≤ .05 was regarded as statistically significant. All analyses were performed in SPSS version 16.0 (SPSS Inc, Chicago, Illinois).
The final study sample consists of 3280 persons aged 40 to 80 years (some participants were aged 80 years at the time of examination). Of these, 20 participants without retinal photographs or with photographs of insufficient quality for grading were excluded, leaving 3260 participants with gradable photographs for analysis. A total of 45 (1.4% of 3260) had 1 or more choroidal nevi. Two subjects (4.4%) had bilateral nevi, 24 (53.3%) had nevi in the right eye only, and 19 (42.2%) had nevi in the left eye only. The maximum number of nevi seen per eye was 2, found in 3 subjects (6.7%). The age and sex distribution of participants with choroidal nevi in this population are outlined in Table 1. In men, a slightly higher prevalence of persons with choroidal nevi was seen in younger age groups, but this was not observed in women. The person-specific prevalence of choroidal nevi was significantly higher in men than women only in the 50- to 59-year age group.
Of the 50 nevi found in 45 participants, 49 (98%) were slate blue or green-gray in color, with only 1 nevus (2%) having a patchy hypomelanotic appearance. A total of 20 nevi (40%) were round, 17 (34%) oval, and 13 (26%) irregular in shape. The mean (SD) nevus diameter was 1.27 (0.47) mm (range, 0.45-2.55 mm), and the mean (SD) nevus surface area was 1.54 (1.13) mm2(range, 0.21-4.22 mm2). No associations were found between nevus size and age.
Twenty-nine nevi (58%) lay within the posterior pole, 17 (34%) were in the midperiphery, and 4 (8%) overlapped the posterior pole and midperiphery. Twenty nevi (40%) were located within or on the border of the macula; 2 (4%) were subfoveal, and 3 (6%) were subfoveolar. The posterior margin of 17 nevi (34%) were within 2 disc diameters of the optic disc, and 33 (66%) were beyond 2 disc diameters. Nevi were located in the upper temporal quadrant relative to the optic disc in most cases (32 nevi, 64%), were lower temporal in 4 cases (8%), upper nasal in 12 (24%), and in the lower nasal quadrant in 2 cases (4%).
Drusen were present overlying 10 nevi (20%). Of the nevi with overlying drusen, the mean (SD) area of nevus involved by drusen was 10.69% (9.59%) of nevus surface area (range, 0.85%-30.25%). Drusen were distributed over the peripheral surface in 6 nevi (60%) and were overlying both the central and peripheral nevus surface in 4 nevi (40%). Large drusen with a size of C2were seen in 30% of nevi, drusen size C1were seen in 30%, and there were C0-sized indistinct drusen in 30% and C0-sized distinct drusen in 10%. The presence of drusen overlying nevi was unrelated to participant age or nevus size. There were no nevi with overlying orange pigment, pigment clumping, or retinal edema.
Table 2compares characteristics of choroidal nevi in the Malay population with those of white persons in the BMES3of the persons aged 49 to 80 years. The nevi in both studies were similar in terms of size, shape, color, proximity of nevus margin to the optic disc, and location within the posterior pole or periphery, although there were differences in the quadrantal distribution of nevi and the proportion of nevi with overlying drusen.
No associations were found between nevi and cataract, age-related macular degeneration, diabetes, smoking history, education level, or occupation. No participants with choroidal nevi had glaucoma. Although 5 nevi (10%) lay within 750 μm of the foveal center, there were no cases where visual impairment was attributable to the nevus.
Our study found a 1.4% prevalence of choroidal nevi in a Malay population in Singapore. A key feature of this study was the ability to compare findings with the white population from the BMES, as similar definitions were used for grading choroidal nevi.3We found a substantially lower prevalence in Malay persons compared with the 6.5% prevalence reported in the BMES.3Similarly, the only other report of choroidal nevi in an Asian population, the Beijing Eye Study, also found a substantially lower prevalence in adult Chinese persons (2.9%)14compared with white persons in the BMES.3
Importantly, despite the differences in prevalence, we show that the key characteristics of nevi were similar in Malay persons in the SiMES and white persons in the BMES. The mean nevus diameter in our population was 1.27 mm, similar to the mean diameter of 1.25 mm found in the BMES.3Nevi in both studies were also similar in terms of shape (P = .58), color (P = .39), location within posterior pole or periphery (P = .30), and proximity of nevus margin to the optic disc (P = .29). The only significant difference was in the frequency of drusen in the nevi (22% in the SiMES and 100% in the BMES; P < .001).
There is a wide range in the reported prevalence of choroidal nevi from previous studies, mostly conducted in white persons.1,3- 13These differences are likely owing to different examination methods used and sample variations across studies. There are few data in the literature on choroidal nevi from populations that are not white.27After correcting for limitations of the photographic field, with the estimation that 25% of nevi would likely be missed on photography, a prevalence of 1.8% in the SiMES can be estimated. This adjustment was derived from Naumann et al,39who suggested from combined clinical and histopathologic investigations that 75% of choroidal nevi are located in the posterior pole and its immediate surroundings, an area comparable with our photographic field.
A limitation of our study is that, as grading was performed by one author (C.H.N.) with adjudication provided by a retinal specialist (A.H.), intragrader or intergrader reproducibility of the grading of nevi was not assessed. There were also differences in the photographic fields used in the SiMES and the BMES. In the SiMES as well as in the Beijing Eye Study,14two 45° photographic fields centered on the disc and macula were used, whereas six 30° photographic fields were used in the BMES (disc, macula, lateral macula, upper and lower temporal arcade, and nasal retina).3The decision to obtain only 2 standard photos in the SiMES was based on the primary aim of SiMES to investigate age-related eye diseases, principally macular diseases29; however, we acknowledge that a wider photographic field would have been ideal in terms of detecting choroidal nevi.
Figure 1and Figure 2show composite views of the photographic fields obtained in the SiMES and the BMES. The illustrations enable subjective comparison of field differences, although we are unable to comment on the exact proportion of nevi that would have been undercounted in the SiMES owing to field differences. However, despite the differences in photographic fields and quadrantal distributions outlined in Table 2, the proportion of nevi found involving the posterior pole (SiMES, 51%; BMES, 45%), midperiphery (SiMES, 38%; BMES, 49%), and overlapping the posterior pole and midperiphery (SiMES, 11%; BMES, 6%) was not significantly different. These findings suggest that differences in the fields photographed alone may partially contribute to, but are insufficient to account for, such a large disparity in prevalence, so that racial differences are likely contributors to the observed differences in nevi prevalence.
As in the BMES,3there was a trend toward decreasing prevalence with increasing age in our study, although this did not reach statistical significance. Ocular media opacity obscuring the view of subtle nevi could partly explain this finding, although it is unlikely to be a major reason owing to the high visibility of nevi. In our study, a higher prevalence of choroidal nevi was seen in younger age groups in men, but not in women. The prevalence of choroidal nevi was significantly higher in men than women only within the 50- to 59-year age group. We speculate that this is a random variation in the data due to small numbers after stratifying the sample into age groups. Interestingly, the Beijing Eye Study also reported that the overall prevalence of choroidal nevi was significantly higher in men than in women.14
Like their cutaneous counterparts, the main clinical concern with choroidal nevi is their rare potential for malignant transformation.1,8,15- 22Previous studies have estimated that one choroidal melanoma may result from 4300 to 8845 nevi per year in white persons.3,8,19This risk gives rise to continuing efforts to identify reliable indicators of growth and malignant potential, as well as continuing surveillance of patients with choroidal nevi.27,40Features associated with risk of transformation include orange pigment (lipofuscin), pigment clumping, and subretinal fluid, which are hypothesized to arise from retinal pigment epithelial damage resulting from enlargement of the underlying nevus.1,17,23,41,42None of these features were observed in any participants in our study, consistent with the very low probability of detecting choroidal melanoma in a population-based sample at a single time point without longitudinal data. Importantly, the BMES 5-year follow-up data also showed no malignant changes over that period.22
Large elevated nevi have been found to have an increased risk of growth and malignant change.1,17,21,23,39,41Past studies proposed that suspicious nevi should include those larger than 6 mm in diameter or thicker than 1 to 3 mm in elevation1,16,21,27or those with posterior margins that touched the optic disc.16Nevi in our study were relatively small, with a mean (SD) diameter of 1.27 (0.47) mm. In our sample, the largest lesion measured 2.55 mm in diameter, and no nevi had posterior margins touching the optic disc, suggesting that most lesions were benign without signs of potential malignancy. We were unable to measure the height of nevi from digital images, although we can subjectively comment that the lesions photographed appeared not to be significantly elevated.
The presence of drusen overlying nevi has been postulated to imply low growth potential, although its significance is equivocal.1,18,23,41,43Various percentages (26% to 98%) of choroidal nevi were reported to have overlying drusen in previous studies, using different grading methods.2,3,12,39,40The presence of drusen overlying nevi has been associated with increasing subject age40as well as increasing nevus size,3suggesting that this sign relates to the chronicity of the nevus. We found no statistically significant associations between drusen and age or nevus size in this population. Our study found that only 20% of nevi had overlying drusen, substantially less than the 98% reported in the BMES.3We cannot offer explanations for such a large discrepancy but speculate that both sample variations and differences in grading (as performed by different graders) could have contributed to this.
Choroidal nevi have been reported to be associated with visual impairment, especially subfoveal nevi.39,40,44- 46Gonder et al46reported decreased visual acuity in 11% of patients referred to an oncology clinic for assessment of choroidal tumors. Shields et al40noted decreased visual acuity in 6.3% of 3422 patients referred to an ocular oncology service at a tertiary referral center. In our population-based study in which 40% of nevi were within the macular area and 10% were subfoveal, there were no cases where impaired visual acuity was attributable to choroidal nevi, consistent with the BMES findings. Given that lesions in Gonder and colleagues' and Shields and colleagues' studies were much larger than those in ours and in the BMES, with an average diameter of 4.3 mm and 4.7 to 5.6 mm, respectively, it is possible that cases referred to oncology assessment centers are selected samples and are likely to have included relatively large nevi, which may not be representative of choroidal nevi found in a usual population.
In conclusion, our study in an Asian Malay sample is among the first reports of population-based data on choroidal nevi in a population that is not white. We documented that the prevalence of choroidal nevi in this Malay population was lower than that in an age-comparable white population. Most nevi were relatively small and did not affect visual acuity. Importantly, we show that characteristics of nevi were largely similar between Asian and white persons.
Correspondence: Tien Y. Wong, MD, PhD, Centre for Eye Research Australia, University of Melbourne, Gisborne St, Victoria 3002, Australia (email@example.com).
Submitted for Publication: May 18, 2008; final revision received November 9, 2008; accepted November 23, 2008.
Financial Disclosure: None reported.
Funding/Support: This study was supported by grants 0796/2003, 0863/2004, and CSI/0002/2005 from the National Medical Research Council and grant 501/1/25-5 from the Biomedical Research Council. Additional support was provided by the Singapore Tissue Network and the Ministry of Health, Singapore.
Additional Contributions: The authors thank the staff and participants in the Singapore Malay Eye Study for their important contributions. We also thank Alex Harper, MBBS, FRANZCO, for providing adjudication in the grading of choroidal nevi in this study and Ava Grace Tan, BS, for production of the illustrations.