Conjunctival Nevi: Clinical Features and Natural Course in 410 Consecutive Patients

Objectives  To describe the clinical features of a conjunctival nevus and to evaluatethe lesion for changes in color and size over time.

Design  Retrospective, observational, noncomparative case series.

Participants  Four hundred ten consecutive patients with conjunctival nevi.

Main Outcome Measures  The 2 main outcome measures were changes in tumor color and size.

Results  Of the 410 patients, 365 (about 89%) were white, 23 (about 6%) wereAfrican American, 8 (2%) were Asian, 8 (2%) were Indian, and 6 (1%) were Hispanic.The iris color was brown in 55% (229/418), blue in 20% (85/418), green in20% (83/418), and not indicated in 5% (21/418). The nevus was brown in 65%,tan in 19%, and completely nonpigmented in 16%. The anatomical location ofthe nevus was the bulbar conjunctiva (302 eyes, 72%), caruncle (61 eyes, 15%),plica semilunaris (44 eyes, 11%), fornix (6 eyes, 1%), tarsus (3 eyes, 1%),and cornea (2 eyes, <1%). The bulbar conjunctival lesions most commonlyabutted the corneoscleral limbus. The nevus quadrant was temporal (190 eyes,46%), nasal (184 eyes, 44%), superior (23 eyes, 6%), and inferior (21 eyes,5%). Additional features included intralesional cysts (65%), feeder vessels(33%), and visible intrinsic vessels (38%). Cysts were clinically detectedin 70% of histopathologically confirmed compound nevi, 58% of the subepithelialnevi, 40% of the junctional nevi, and 0% of the blue nevi. Of the 149 patientswho returned for periodic observation for a mean of 11 years, the lesion colorgradually became darker in 5% (7 patients), lighter in 8% (12 patients), andwas stable in 87% (130 patients). The lesion size was larger in 7% (10 patients),appeared smaller in 1% (1 patient), and was stable in 92% (137 patients).There were 3 patients who developed malignant melanoma from a preexistingcompound nevus (2 cases) or blue nevus (1 case) over a mean interval of 7years.

Conclusions  Conjunctival nevus is a benign tumor most often located at the nasalor temporal limbus and rarely in the fornix, tarsus, or cornea. Over time,a change in tumor color was detected in 13% (20/149) and a change in tumorsize was detected in 8% (12/149).

The conjunctival nevus is a common benign ocular tumor. It can manifestwith a spectrum of clinical features.¹ Recentarticles about the conjunctival nevus have primarily focused on the histopathologicfeatures of this entity.^2-8 Overthe past 2 decades, there have been only 2 relevant studies on the long-termnatural history of conjunctival nevi.^4,6 Bothstudies focused on the pathologic correlation of excised lesions with littleinformation on the varied clinical features. In this article, to our knowledge,we provide for the first time in the recent literature a detailed accountof the clinical variations of conjunctival nevi in 410 consecutive patients.We provide information regarding the frequency of change in pigmentation andsize of these benign tumors. A comparison of the clinical appearance of conjunctivalnevi with conjunctival melanoma is also made.

The clinical records of all patients with a conjunctival nevus, examinedand treated on the Ocular Oncology Service, Wills Eye Hospital, Thomas JeffersonUniversity, Philadelphia, Pa, between July 1, 1974, and May 30, 2002, werereviewed. Clinical data were gathered retrospectively regarding features ofthe patient and the conjunctival nevus. The clinical data were then analyzedwith regard to 2 main outcome measures—color change within the nevusand size change of the nevus.

These data included patient features at the initial examination suchas age, race (African American, Hispanic, Asian, or white), and sex (femaleor male). Data were recorded regarding cutaneous lesions (nevus, dysplasticnevus syndrome, or malignant melanoma), other conjunctival lesions (primaryacquired melanosis, pinguecula, pterygium, or squamous cell neoplasia), andchoroidal lesions (nevus or melanoma). Data were recorded on iris color (blue,green, brown, or hazel). A history regarding the conjunctival nevus includedprevious documented growth of the lesion (present or absent), previous treatmentof the lesion (none or excisional biopsy), symptoms (spot, inflammation, pain,or none), and the duration of these symptoms.

The ocular data included best-corrected Snellen visual acuity, designationof the eye involved (right or left), and the intraocular pressure. The tumordata included the anatomical location (bulbar conjunctiva, fornix, tarsalconjunctiva, plica semilunaris, caruncle, or cornea), quadrant location ofthe tumor epicenter (superior, temporal, inferior, or nasal), proximity tothe limbus (expressed in millimeters), anterior tumor margin (cornea, limbus,bulbar conjunctiva, or fornix), posterior tumor margin (cornea, limbus, bulbarconjunctiva, or fornix), largest basal dimension (expressed in millimeters),largest thickness (expressed in millimeters), elevation status (flat or elevated),color (amelanotic, tan, or brown), intralesional cysts status (present orabsent), and the number of cysts per nevus, feeder vessels (present or absent),and intrinsic vessels (present or absent).

Follow-up examinations were generally made at 6- to 12-month intervals.In this study we included only patients with photographic follow-up for comparisonof clinical data. The follow-up data included the type of color change withinthe nevus (darker, lighter, or no change) and the type of nevus size change(larger, smaller, or no change).

The general information about patient demographics is listed in Table 1. The visual acuity on the initialexamination was 20/20 to 20/50 in 403 patients (97%), 20/60 to 20/100 in 5patients (1%), and 20/200 or worse in 8 patients (2%). Data regarding generalocular findings are listed in Table 2.Data regarding tumor findings are listed in Table 3 and in Figure 1, Figure 2, Figure 3, and Figure 4.There were 180 patients (43%) who reported that the lesion appeared to haveenlarged over time before the date of the initial examination. The tumor wasmost commonly located on the bulbar conjunctiva (72%). Of all 418 lesions,the quadrant location of the tumor was temporal in 46% and nasal in 44%; theanterior margin was immediately at the limbus in 48% and behind the limbusin 25%. The median size was 3.5 mm in basal diameter and 0.5 mm in thickness.The lesion was most commonly pigmented (84%) and had intralesional cysts (65%).Feeder vessels (33%) and intrinsic vessels (38%) were found less often. Tumortreatment is listed in Table 4 andincluded observation (62%) or excisional biopsy alone or with cryotherapy(38%). The most common reasons for excisional biopsy included our concernfor malignant change based on clinical features (16%), recent growth (8%),color change (<1%), or recurrence of previously excised lesion (2%). Theclinical features suggestive of possible melanoma include older patient ageat recognition of nevus; corneal, forniceal, or palpebral involvement; prominentfeeder vessels; lack of intrinsic cysts; and personal or family history ofcutaneous melanoma or dysplastic nevus syndrome. Other reasons for excisionincluded patient's concern for melanoma (7%) or cosmetic appearance (4%).

The histopathologic diagnoses in the 148 excised lesions are listedin Table 5 and included compoundnevus (70%), subepithelial nevus (4%), junctional nevus (3%), and blue nevus(3%) (Figure 5). Of these 148 nevi,tumor pigmentation was found in 82% of those classified as compound, 86% ofthose classified as subepithelial nevi, 100% of those classified as junctionalnevi, and 100% of those classified as blue nevi. Cysts were noted in 70% ofthe compound nevi, 58% of the subepithelial nevi, 40% of the junctional nevi,and 0% of the blue nevi.

Of the 149 conjunctival nevi followed up for a mean of 11 years withoutbeing excised, the tumor showed a color change in 13% and a size change in8% (Table 6 and Figure 6). The clinical features of the 3 patients with nevi thatevolved to malignant melanoma (over a mean of 7 years) are listed in Table 7. No patients developed melanomametastasis.

Epibulbar melanocytic lesions include conditions of the conjunctivalepithelium, stroma, and sclera. Some of these conditions include conjunctivalracial melanosis, primary acquired melanosis, secondary melanosis, nevus,and melanoma, as well as ocular melanocytosis and extraocular extension ofuveal melanoma.¹ The clinical features of thesepigmented conditions occasionally overlap and cause diagnostic confusion.Moreover, an amelanotic conjunctival nevus can resemble other nonpigmentedconditions including inflamed pingueculum, episcleritis, conjunctival cyst,allergic conjunctivitis, foreign body granuloma, lymphangioma, and squamousepithelial neoplasia. The differentiation of these various conditions is importantas it implies diverse ocular and systemic prognoses. The diagnosis is usuallymade based on the clinical features and occasionally confirmed with histopathologicfindings. With regard to the conjunctival nevus, the diagnosis is typicallymade by recognition of the classic clinical features using slitlamp biomicroscopy.In this article, we describe a comprehensive overview of the various presentationsof the conjunctival nevi in 410 consecutive patients.

Few articles have focused on the subject of the clinical features ofconjunctival nevi. Most previous articles have described the pathologicalfindings with generalizations, but little description of the clinical findings.^2-11 In1965, Jay⁹ reported the pathological featuresof benign nevi of the conjunctiva. He gathered basic clinical informationonly on patient age at diagnosis and noted that the patients most commonlyseen by an ophthalmologist were between the ages 10 to 29 years and typicallyclaimed that the lesion was first detected when the patient was younger than9 years. Other clinical features of the excised tumors were unavailable. Henkind,¹⁰ in 1978, provided a comprehensive summary of conjunctivalmelanocytic lesions with minor generalizations on the clinical findings ofconjunctival nevi. Since then, other articles on this condition have concentratedon related histogenesis, light microscopic findings, and ultrastructural featuresof common and rare subtypes of nevi.^2-5,7,8,11

In 1996, a Danish study by Gerner et al⁶ provideda clinicopathologic study on 343 conjunctival nevi. They described the followingtumor locations: the bulbar conjunctiva in 33%, caruncle in 29%, limbal conjunctivain 27%, and at the eyelid margin in 1%. The patients were seen most commonlybetween the ages of 10 and 19 years. All but 3 patients were white and only1 nevus evolved into a malignant melanoma. Further specific clinical informationwas not reported.

To our knowledge, our report is the first to delineate the specificclinical features of conjunctival nevi. We found the mean patient age at theinitial manifestation was 32 years and most commonly, this tumor was foundin white subjects (89%) (Table 1).Many patients had cutaneous nevi (19%) and 1% of patients each reported ahistory of basal cell carcinoma, dysplastic cutaneous nevi,¹² andprevious cutaneous malignant melanoma. Unlike choroidal melanoma or iris melanoma,which is usually found in patients with blue irides,^13,14 conjunctivalnevi were most commonly found in patients with brown irides (55%) (Table 2). Additional choroidal nevi werenoted in 7% of the white patients, consistent with the expected number inthe white population.¹⁵

With regard to the symptoms and clinical appearance of the tumor, mostpatients reported noticing a spot on the eye (88%), 3% noted inflammation,less than 1% experienced related pain, and 10% of patients were unaware ofthe lesion. The symptoms were present for a mean of 10 years. Enlargementor color change in the lesion over the years prior to our examination wasreported by 43% of patients, but such enlargement was rarely supported byphotographs. The tumor was most commonly found in the bulbar conjunctiva (72%),caruncle¹⁶ (15%), or plica semilunaris (11%)(Table 3). Rarely was the tumorfound in the fornix (1%), tarsal conjunctiva (1%), or within the cornea (<1%).We, therefore, suggest that any pigmented lesion in the fornix, tarsus, orcornea might be considered to be a condition other than a nevus.¹⁷ Thisis especially important if the pigmented lesion is circumscribed and extendsinto the stroma, in which case malignant melanoma might be considered.^18-20 In comparison, of150 consecutive patients with conjunctival melanoma reported by Shields etal,¹⁹ the tumor location was the bulbar conjunctiva(92%), caruncle (1%), plica semilunaris (1%), fornix (3%), and tarsal conjunctiva(4%).

Those tumors on the bulbar conjunctiva were a mean of 1 mm (median,0 mm) from the limbus. In only 2 patients was there corneal involvement andboth were unusual cases with small corneal stromal nevi.²¹ Cornealinvolvement from a pigmented conjunctival lesion should suggest primary acquiredmelanosis or racial melanosis if the conjunctival pigmentation is within theepithelium, but, importantly, should raise consideration for conjunctivalmelanoma if the conjunctival pigmentation has thickness and extends into thestroma.¹ Conjunctival nevi generally stop abruptlyat the limbus and typically do not involve the corneal epithelium or stroma.Overhang of the cornea from a large conjunctival nevus is possible, but invasionof the cornea by a nevus would be distinctly unusual. On the other hand, conjunctivalmelanoma often grows beyond the limbus into the cornea.

The tumor involved the temporal (46%) or nasal (44%) quadrants of theconjunctiva more than the superior (6%) or inferior (5%) quadrants. This distributionalong the temporal and nasal quadrants is more common with conjunctival nevuscompared with conjunctival melanoma, as conjunctival melanoma has been foundin the superior (16%), nasal (17%), and inferior (22%) quadrants less commonlythan in the temporal (63%) quadrant.¹⁹

The conjunctival nevi ranged in basal dimension from 0.2 to 30 mm, witha mean of 4.1 mm. The mean thickness was estimated to be less than 1 mm. Incomparison, conjunctival melanoma has been detected at a mean size of 8 mmin basal dimension and 2 mm in thickness.¹⁹ Allof the patients with large nevi over 10 mm in basal dimension had prominentintralesional cysts to suggest the diagnosis and most had excisional biopsyfor histopathologic confirmation. The largest lesions were diffusely multicysticand poorly circumscribed, often resembling a cystic conjunctival lymphangiomaor lymphangiectasia.

The biomicroscopic appearance of the nevus was critical to its diagnosis.The tumor was most commonly brown (65%) and less often tan (19%) or completelynonpigmented (16%) appearing as a gelatinous translucent mass. Cysts wererecognized in 65% of the nevi, feeder vessels in 33%, and intrinsic vesselsin 38%. Interestingly, cysts were most common in compound nevus (70%), decreasinglycommon in subepithelial nevus (58%) and junctional nevus (40%), and absentin blue nevus (0%). Conjunctival melanoma has been recorded as brown in 68%,tan in 19%, and completely nonpigmented but with prominent intrinsic vesselscasting a pink rather than gelatinous color in 11% of these lesions.¹⁹ In contrast, however, conjunctival melanoma rarely,if ever, displays intralesional cysts. Feeder vessels are prominent with conjunctivalmelanoma. Thus, the importance of recognition of tumor cysts is a key pointin differentiating conjunctival nevus from malignant melanoma as many otherfeatures overlap.

Excision of conjunctival nevi was performed in 38% of the cases, mostlyfor reasons to rule out melanoma or other tumor, recent growth in the lesion,or cosmetic concern (Table 4).Of the 148 lesions for which histopathologic results were available, the findingsrevealed the following in descending order: compound nevus (70%), subepithelialnevus (24%), combined nevi (4%) (ie, compound and blue nevi or subepithelialand blue nevi), junctional nevus (3%), and blue nevus (3%) (Table 4). Gerner et al,⁶ in a clinicopathologicreview of 343 conjunctival nevi, found similar distribution of nevi with compoundnevi (78%), supepithelial (intrastromal) nevus (15%), junctional nevus (6%),and blue nevus (<1%). Of 57 nevi excised from children and teenagers, McDonnellet al⁴ identified compound nevi (72%), subepithelialnevus (4%), junctional nevus (21%), blue nevus (2%), and Spitz nevus (2%).⁴ These results suggest greater junctional activityin children with conjunctival nevi.

In our series, a correlation of the clinical features with the specifichistopathologic diagnosis was provided (Table 5). Compound and junctional nevi were found in younger agedgroups, whereas subepithelial and blue nevi were in slightly older aged groups.Bulbar conjunctival location was found with all 4 groups, but caruncular orplical tumor locations were seen only with compound or subepithelial nevi.Both compound and subepithelial nevi showed a variation in color types, butjunctional and blue nevi were always brown. Cysts and feeder vessels wereabsent in all 4 patients with a blue nevus, but present in the other types.

The natural history of the 149 observed conjunctival nevi in our seriesrevealed photographically documented change in tumor color in 13% and in sizein 8%. Both were gradual and visible only on careful comparison of photographsover years. The change in apparent tumor size could be related to neoplasticgrowth, but we suspect more commonly it is related to enlargement of the intrinsiccysts, increased pigmentation in previously amelanotic regions of the nevus,or, importantly, inflammation within the nevus. Zamir et al²² foundthat 75% of the excised conjunctival nevi in children showed some degree ofinflammatory infiltrate, some of which showed alarming clinical growth withoutmalignant melanoma formation histopathologically. In our group, 3 patientsdeveloped malignant melanoma; all were white; 2 had cutaneous abnormalitiesincluding dysplastic nevus syndrome in one patient and a family history ofcutaneous malignant melanoma in the other patient. The tumors were dark brownin all cases and lacked cysts in 2 patients. Following excision, evidencefor melanoma was confirmed, along with underlying compound nevi in 2 patientsand blue nevus in 1 patient. From a reverse perspective, conjunctival melanomahas been found to originate from preexistent nevus in 4% of the patients,primary acquired melanosis in 57%, de novo in 39%, that and not specifiedin 6%.¹⁹ Nooregaard et al²⁰ foundmelanomas originate from a nevus in 16% of the patients, primary acquiredmelanosis in 36%, and de novo in 47%.

There are limitations in this study that should be realized. First,the cohort of patients was derived from a tertiary care ocular oncology centerand, thus, may represent a biased group with more suggestive features. Thus,the 3 patients who developed conjunctival malignant melanoma likely representa higher rate than expected in the general population of patients with conjunctivalnevi. Second, there was incomplete photographic follow-up on some of the patientsin this cohort. This could inflate the worst-case scenario as the stable patientmight choose no further follow-up, whereas those with prominent lesions ortumor growth might choose close follow-up. Third, data collection were retrospectiveand although the collection was extensive, some data were incomplete. Fourth,not all patients had histopathologic confirmation of the nevus because manywere followed up conservatively without surgery.

Conjunctival nevi display a spectrum of clinical features from heavypigmentation to a complete lack of pigmentation, from diffuse confluence ofcysts to a complete lack of cysts, and from a tiny dotlike size lesion toextensive tumors occupying 1 or 2 quadrants of the ocular surface. The naturalhistory of conjunctival nevi is benign with minor gradual changes of pigmentationin 13% of the patients and gradual change in size in 8% of patients.

Corresponding author and reprints: Carol L. Shields, MD, Ocular OncologyService, Wills Eye Hospital, 840 Walnut St, Philadelphia, PA 19107.

Submitted for publication May 20, 2003; final revision received September7, 2003; accepted October 1, 2003.

This study was supported by the Eye Tumor Research Foundation, Philadelphia,Pa (Dr C. L. Shields); the Macula Foundation, New York, NY (Dr C. L. Shields);the Center for Eye Research, Tehran, Iran (Dr Mashaykhi); the Rosenthal Awardof the Macula Society, Cleveland, Ohio (Dr C. L. Shields); and the Paul KayserInternational Award of Merit in Retina Research, Houston Tex (Dr J. A. Shields).