Conjunctival Nevi

Clinical Features and Therapeutic Outcomes

Laurent Levecq, MD,1 Patrick De Potter, MD, PhD,1 Jacques Jamart, MD, MSc2

Objective: To determine the epidemiology and the clinical and therapeutic outcomes of conjunctival nevi
and to identify the clinical variables statistically associated with operative excision.
Design: Prospective, observational, noncomparative case series.
Participants: Two hundred fifty-five patients with the clinical diagnosis of conjunctival nevus.
Methods: Consecutive cases of conjunctival nevi managed at a single institution were studied to identify the
clinical risk factors for operative excision.
Main Outcome Measures: Reasons for operative excision.
Results: Of the 255 patients who were periodically observed for a mean of 5.3 years (range, 1–11), nevi were
clinically diagnosed in 140 females and 115 males and modified operative excision was performed in 75 patients
(29%). The decision of operative excision was made by the surgeon in 13 cases (17%) and by the patient in 62
cases (83%). In those 13 patients, the operative decision was prompted by our concern for possible malignant
transformation based on suspicious biomicroscopic features in 10 patients (13%) and photographically docu-
mented tumor growth in 3 patients (4%). For the other 62 patients who elected to undergo surgery, their reasons
for excision included patient’s concern for cancer in 34 cases (45%), cosmetic arguments in 9 cases (12%), and
patient’s request owing to lesion-induced ocular surface irritation in 19 cases (25%). Comparison between
groups showed that the clinical factors at initial visit that were statistically predictive of surgical excision were the
older age of the patient (P ⫽ 0.001), the largest basal tumor diameter (P⬍0.001), tumor location (P ⫽ 0.023), and
presence of clear cysts (P ⫽ 0.013), of intrinsic vasculature (P⬍0.001), of prominent feeder vessels (P⬍0.001),
and of corneal involvement (P ⫽ 0.008). None of the excised lesions showed histopathologically malignant
features.
Conclusions: In our series, documented tumor growth of conjunctival nevus remained relatively a uncom-
mon event with a incidence of 4%. Conjunctival nevi in older patients, associated with dilated feeder vessels,
prominent intrinsic vasculature, and corneal involvement were more likely to be treated with operative excision.
Financial Disclosure(s): The authors have no proprietary or commercial interest in any materials discussed
in this article. Ophthalmology 2010;117:35– 40 © 2010 by the American Academy of Ophthalmology.

Benign conjunctival lesions of melanocytic origin may be
separated into 2 broad categories: nevi and other benign
disorders such as complexion-associated pigmentation, pig-
mentation associated with systemic disorders, ephelis, and
lentigo.1 Nevi may be further subdivided according to the
age of the first clinical presentation: congenital (present at
birth or appearing within the first 6 months of life) or
acquired.1 An acquired conjunctival nevus generally be-
comes clinically apparent in the first or second decade of
life as a discrete, variably pigmented, slightly elevated,
sessile lesion that may contain clear cysts and usually lo-
cated in the interpalpebral conjunctiva.2– 6 Dilated feeder
conjunctival or episcleral vessels are commonly lacking.
Transformation into malignant melanoma of the conjunctiva
is a rare event, particularly in young patients; however, its
true incidence is unknown. Best management is usually
periodic observation with photographs.
Despite numerous studies on long-term natural history of
conjunctival nevi and their pathologic features,1,4 – 6 few
relevant articles regarding clinical course and variations in
change in pigmentation and size and course of these benign
tumors are available.7 Herein, we have analyzed the epide-
miology and the clinical and therapeutic outcomes of con-
junctival nevi referred to an Ocular Oncology Center and
tried to identify the clinical variables statistically associated
with surgical excision.
Materials and Methods
In this prospective, nonrandomized study, the clinical records of
all patients clinically diagnosed with conjunctival nevus and fol-
lowed at the Ocular Oncology Unit, Cliniques Universitaires St-
Luc, Brussels, between October 1997 and October 2007 and with
a minimum of 12 month follow-up were reviewed. This study
followed the principle of the Declaration of Helsinki and ethics
approval was obtained from the Cliniques Universitaires St-Luc
Institutional Review Board. The clinical diagnosis of conjunctival
nevus was based on patient’s history and biomicroscopic features
of the lesion evaluated by one of the authors (PDP).1– 6 All patients
with an acquired conjunctival lesion becoming clinically apparent
in the first or second decade of life and presenting as a unifocal,
variably pigmented, and slightly elevated lesion that may contain
clear cysts, intrinsic vasculature, feeder vessels, and well-defined
edges and movable over the globe were included in our study. The
clinical data were recorded by one of the authors (PDP) and then

© 2010 by the American Academy of Ophthalmology ISSN 0161-6420/10/$–see front matter 35

Published by Elsevier Inc. doi:10.1016/j.ophtha.2009.06.018
 Ophthalmology Volume 117, Number 1, January 2010
analyzed with regard to the main outcome of the excision of the
lesion.
These patient data included patient age at initial consultation at
the Ocular Oncology Unit, time interval between patient’s aware-
ness of the lesion and his referral, patient gender (female or male),
and race (Caucasian, African, Asian, or Hispanic). The tumor data
included anatomic conjunctival location (bulbar, palpebral, plica,
or caruncular), quadrant location on bulbar conjunctiva if involved
(superior, nasal, inferior, or temporal), and corneal involvement
(absent or present). Additional tumor features included largest
basal diameter (mm), pigmentation (amelanotic, partially pig-
mented, or pigmented), intralesional clear cysts (absent or present),
prominent intrinsic vasculature (absent or present), and dilated
feeder vessels (absent or present). We did not record lesion thick-
ness because clinical thickness estimation would be most inaccu-
rate and because we did not perform ultrasound biomicroscopic
measurement in our patients.
The decisions for management were based on the surgeon’s
concern (PDP) for possible malignant transformation (tumor
growth and suspicious biomicroscopic changes such as increase in
intrinsic vascularity, increase in feeder vessels, or change in pig-
mentation) and on patient’s motivation (patient’s concern for can-
cer, cosmetic appearance, or ocular surface irritation). The therapeutic
options included periodic observation and modified operative excision
with a no-touch technique and adjuvant cryotherapy.8 Additional
alcohol corneal superficial epitheliectomy was performed when
corneal involvement was documented.8 The histopathologic data
and reasons for operative excision were recorded.
Numerical variables are exposed as mean values ⫾ standard
deviations, and by medians into brackets. Qualitative, ordinal, and
numerical variables were compared between groups with the chi-
square, Cochran, or Wilcoxon rank-sum tests, respectively. The
influence of various parameters on the choice of treatment was also
studied by multivariate logistic regression, with backward selec-
tion of variables by likelihood ratio test. All tests are 2 tailed and
were performed using SPSS 15.0 statistical software (SPSS Inc.,
Chicago, IL).
Results
Throughout a 10-year period, 255 conjunctival nevi were clinically
diagnosed in 140 females and 115 males. The mean age at the
initial visit was 31 years (range, 1–90) and the mean follow-up was
5.3 years (range, 1–11). Clinical characteristics of evaluated le-
sions are shown in Table 1.
Treatment approaches were observation in 180 patients (71%)
and modified operative excision in 75 patients (29%). Among
those 75 patients, the decision of surgical excision was made by
the surgeon (PDP) in 13 cases (17%) and by the patient in 62 cases
(83%). In those 13 patients, the surgical decision was prompted by
our concern for possible malignant transformation based on sus-
picious biomicroscopic features in 10 patients (13%) and photo-
graphically documented tumor growth in 3 patients (4%). For the
other 62 patients who elected to have surgery, their reasons for
surgical excision included patient’s concern for cancer in 34 cases
(45%), cosmetic arguments in 9 cases (12%), and patient’s request
owing to lesion-induced ocular surface irritation in 19 cases (25%).
In all 75 excised lesions, histopathologic diagnoses included
compound nevus in 55 cases (74%), subepithelial nevus in 16
cases (21%), and junctional nevus in 4 cases (5%).
The clinical parameters were subsequently statistically ana-
lyzed for possible association with surgical excision (Table 1). The
clinical factors at the initial visit that were statistically predictive
of surgical excision were the greater patient age, largest basal
tumor diameter, tumor location, and presence of clear cysts, of
36
intrinsic vasculature, of prominent feeder vessels, and of corneal
involvement.
Surgical excision was performed for 11% of lesions in the
inferior quadrant, 14% of palpebral conjunctival tumors, 22% of
lesions in the nasal quadrant, 22% of lesions in the temporal
quadrant, 38% of lesions in the superior quadrant, 41% of tumors
of the caruncle, and 50% of tumors of the plica. A multivariate
logistic regression analysis confirmed the statistically significant
predictive value of all previously cited factors, except tumor lo-
cation. When we analyzed clinical data within the group of patients
who underwent surgical excision (Table 2) the predictive factors of
surgical excision contemplated by the surgeon (PDP) were the
greater patient age, interval between first awareness and first visit,
and presence of intrinsic vessels and feeder vessels.
In the subgroup of lesions surgically excised per the patient’s
request (Table 3), the larger basal diameter and presence of clear
cysts and the temporal location of the conjunctival nevus were
statistically associated with patient’s motivation for surgical
excision.
Discussion
A variety of tumors and simulating lesions can occur in the
conjunctiva. These tumors are generally classified into
melanocytic and nonmelanocytic tumors based on biomi-
croscopic and histopathologic features. In the series by
Shields et al9 of 1643 conjunctival tumors, 53% were of
melanocytic origin and 47% of nonmelanocytic origin.
The most common lesion in this large clinical series was
conjunctival nevus, accounting for 28% of all conjuncti-
val tumors and 52% of those classified as melanocytic
tumors.9
The diagnosis of conjunctival nevus is typically made by
on the classic clinical features using slit-lamp biomicros-
copy and occasionally confirmed with histopathologic ex-
amination.1–7 In our series, the tumor was most commonly
pigmented (51%), partially pigmented (28%), or com-
pletely amelanotic (21%). Intralesional clear cysts were
documented in 57% of nevi, prominent intrinsic vascu-
larity in 21%, and feeder vessels in 27%. Intralesional
cysts were more often recognized among the 410 nevi evalu-
ated by Shields et al; they emphasized the importance of
recognition of cysts in differentiating conjunctival nevus
from melanoma.9
Shields et al7 provided a relevant report by delineating
the specific clinical features of conjunctival nevi and their
natural course. They found the mean patient age at initial
manifestation was 32 years with symptoms present for a
mean of 10 years and most nevi (89%) were found in the
white population. Our findings were similar, with a mean
patient age at initial consultation of 31 years; 85% of nevi
were found in Caucasians. Similar to the data of Shields et
al data regarding tumor location, nevi were most commonly
found in the bulbar conjunctiva (67%) with involvement of
the temporal (52%) and nasal (30%) quadrants more than
the superior (13%) and inferior quadrants (5%), caruncle
(22%), and plica semilunaris (9%) in our series. Rarely did
the tumor involve the palpebral conjunctiva (3%).
Operative excision with a no-touch technique and cryo-
therapy was performed in 29% of our cases. Patient’s mo-
tivation (cancerophobia, cosmetic appearance, and ocular
 Levecq et al 䡠 Conjunctival Nevi

Table 1. Clinical Characteristics of 255 Conjunctival Nevi

Total Observed Lesions Excised Lesions

(N ⴝ 255) (N ⴝ 180) (N ⴝ 75) P
Patient gender NS
Female 140 100 40
Male 115 80 35
Patient race NS
Caucasian 218 155 63
African 32 20 12
Asian 3 3 0
Hispanic 2 2 0
Mean patient’s age (yrs) 31⫾19 [29] 29⫾18 [26] 37⫾19 [35]
Interval between first awareness NS
and first visit (yrs)
⬍1 65 43 22
⬍5 41 37 4
⬍10 19 13 6
⬎10 82 53 29
Unknown 48 34 14
Eye NS
Right 138 97 41
Left 117 83 34
Largest basal diameter (mm) 4.3⫾2.4 [4] 3.9⫾2.3 [3.6] 5.3⫾2.3 [5.1] ⬍0.001
Pigmentation NS
Amelanotic 53 37 16
Partially pigmented 71 53 18
Pigmented 131 90 41
Clear cysts 0.013
No 109 68 41
Yes 146 112 34
Intrinsic vessels ⬍0.001
No 110 89 21
Minimal 91 68 23
Prominent 54 23 31
Feeder vessels ⬍0.001
No 186 146 40
Yes 69 34 35
Anatomic location 0.023
Bulbar conjunctiva
Temporal quadrant 89 69 20
Nasal quadrant 51 40 11
Superior quadrant 21 13 8
Inferior quadrant 9 8 1
Caruncle 56 33 23
Plica 22 11 11
Palpebral conjunctiva 7 6 1
Anatomic location NS
Bulbar conjunctiva
Touching limbus 104 77 27
Distance from limbus 151 103 48
Corneal involvement 0.008
No 207 160 57
Yes 38 20 18

NS ⫽ not significant.

irritation) was the main reason (87%) for excision. Our
concern for malignant transformation triggered by suspicious
clinical features and documented tumor growth prompted us to
perform surgery in 13% of the patients. Despite the lack of
statistical analyses in prior reports, the most common indi-
cations for surgical removal, were documented tumor
growth,1,6,7 limbal location with corneal involvement,1,7,10
and prominent feeder vessels.7 In our series, the most sig-
nificant clinical factors at initial visit that were statistically
predictive of surgical excision were greater patient age, larger
basal tumor diameter, presence of clear cysts, presence of
intrinsic vasculature, presence of prominent feeder vessels,
and corneal involvement. In the series by Shields et al, exci-
sional biopsy was performed in 38% of the cases, mostly to
rule out melanoma or other tumor, recent tumor growth, cos-
metic concern, or recurrence of an excised lesion.7

37
 Ophthalmology Volume 117, Number 1, January 2010

Table 2. Clinical Characteristics of Surgically Removed Conjunctival Nevi

Total Surgeon’s Decision Patient’s Request

Recorded Excision Motivation (N ⴝ 75) (N ⴝ 13) (N ⴝ 62) P
Patient gender NS
Female 40 7 33
Male 35 6 29
Patient race NS
Caucasian 63 10 53
African 12 3 9
Mean patient’s age (yrs) 37⫾19 [35] 53⫾15 [52] 34⫾19 [32] 0.001
Interval between first awareness 0.011
and first visit (yrs)
⬍1 22 8 14
⬍5 4 0 4
⬍10 6 1 5
⬎10 29 2 27
Unknown 14 2 12
Eye NS
Right 41 4 37
Left 34 9 25
Larger basal diameter (mm) 5.3⫾2.3 [5.1] 5.0⫾1.6 [5.0] 5.4⫾2.4 [5.1] NS
Pigmentation NS
Amelanotic 16 3 13
Partially pigmented 18 0 18
Pigmented 41 10 31
Clear cysts NS
No 41 10 31
Yes 34 3 31
Intrinsic vessels 0.007
No 21 1 20
Minimal 23 2 21
Prominent 31 10 21
Feeder vessels 0.003
No 40 2 38
Yes 35 11 24
Anatomic location NS
Bulbar conjunctiva
Temporal quadrant 20 2 18
Nasal quadrant 11 1 10
Superior quadrant 8 0 8
Inferior quadrant 1 0 1
Caruncle 23 5 18
Plica 11 5 6
Palpebral conjunctiva 1 0 1
Anatomic location NS
Bulbar conjunctiva
Touching limbus 27 2 25
Distance from limbus 48 11 37
Corneal involvement NS
No 57 11 46
Yes 18 2 16

NS ⫽ not significant.

In the subgroup of lesions surgically excised per patient
request (Table 3), the larger basal diameter and presence of
clear cysts were identified as the most significant variables
associated with ocular irritation. The temporal location of
the conjunctival nevus was statistically associated with pa-
tient’s cosmetic appearance.
Any pigmented lesion at the limbus that straddles onto
the peripheral cornea or growing conjunctival lesions should
be considered clinically to be a malignant melanoma.1,7
Although corneal involvement was found to be associated
with surgical decision, none of our excised nevi showed his-
topathologic features of malignancy.
Whereas our prospective data collection was exten-
sive, the sample of patients may be considered too small
for multivariate statistical analysis and the results of the
logistic regression analysis should be interpreted cau-
tiously. The mean follow-up of 5.3 years may be consid-
ered too short for those conjunctival nevi whose changes

38
 Levecq et al 䡠 Conjunctival Nevi

Table 3. Clinical Characteristics of Surgically Removed Conjunctival Nevi per Patient Request

Patient Anxiety Patient Irritation Cosmetic Reasons

(N ⴝ 34) (N ⴝ 19 ) (N ⴝ 9) P
Patient gender NS
Female 20 7 6
Male 14 12 3
Patient’s race NS
Caucasian 30 16 7
African 4 3 2
Mean patient’s age (yrs) 34⫾20 [33] 37⫾19 [34] 26⫾12 [26] NS
Interval between first awareness NS
and first visit (yrs)
⬍1 8 4 2
⬍5 3 1 0
⬍10 2 2 1
⬎10 15 8 4
Unknown 16 3 2
Eye NS
Right 21 12 4
Left 13 7 5
Larger basal diameter (mm) 4.7⫾2.1 [4.8] 7.0⫾2.5 [6.5] 4.2⫾1.5 [4.5] 0.001
Pigmentation NS
Amelanotic 8 5 0
Partially pigmented 7 7 4
Pigmented 19 7 5
Clear cysts ⬍0.001
No 23 1 7
Yes 11 18 2
Intrinsic vessels NS
No 12 5 3
Minimal 8 9 4
Prominent 14 5 2
Feeder vessels NS
No 22 10 6
Yes 12 9 3
Anatomic location 0.013
Bulbar conjunctiva
Temporal quadrant 12 2 4
Nasal quadrant 4 6 0
Superior quadrant 2 6 0
Inferior quadrant 1 0 0
Caruncle 12 4 2
Plica 3 1 2
Palpebral conjunctiva 0 0 1
Anatomic location NS
Bulbar conjunctiva
Touching limbus 13 11 1
Distance from limbus 21 8 8
Corneal involvement NS
No 26 13 7
Yes 8 6 2

NS ⫽ not significant.

in tumor size may are gradual and identified over years.
Among other limitations of this study, one may argue the
biased profile of our group of patients referred to a tertiary care
ocular oncology facility, the lack of histopathologic con-
firmation of all conjunctival lesions clinically diagnosed
as nevi, and the lack of anterior segment optical coher-
ence tomography or ultrasound biomicroscopic evalua-
tion for tumor thickness measurements.
References
1. Folberg R, Jakobiec FA, Bernardino VB, Iwamoto T. Benign
conjunctival melanocytic lesions: clinicopathologic features.
Ophthalmology 1989;96:436 – 61.
2. Shields JA, Shields CL, eds. Atlas of Eyelid and Conjunctival
Tumors. Philadelphia: Lippincott Williams & Wilkins; 1999:
243–51.

39
 Ophthalmology Volume 117, Number 1, January 2010

3. Shields CL, Shields JA. Conjunctival tumors in children. Curr
Opin Ophthalmol 2007;18:351– 60.
4. Grossniklaus HE, Green WR, Luckenbach M, Chan CC. Con-
junctival lesions in adults: a clinical and histopathologic re-
view. Cornea 1987;6:78 –116.
5. McDonnell JM, Carpenter JD, Jacobs P, et al. Conjunctival
melanocytic lesions in children. Ophthalmology 1989;96:
986 –93.
6. Gerner N, Norregaard JC, Jensen OA, Prause JU. Conjunctival
naevi in Denmark, 1960 –1980: a 21-year follow-up study.
Acta Ophthalmol Scand 1996;74:334 –7.
7. Shields CL, Fasiuddin A, Mashayekhi A, Shields JA. Con-
junctival nevi: clinical features and natural course in 410
consecutive patients. Arch Ophthalmol 2004;122:167–75.
8. Shields JA, Shields CL, De Potter P. Surgical management of
conjunctival tumors: the 1994 Lynn B. McMahan lecture.
Arch Ophthalmol 1997;115:808 –15.
9. Shields CL, Demerci H, Karatza E, Shields JA. Clinical sur-
vey of 1643 melanocytic and nonmelanocytic conjunctival
tumors. Ophthalmology 2004;111:1747–54.
10. Henkind P, Benjamin JV. Conjunctival melanocytic lesions:
natural history. Trans Ophthalmol Soc U K 1977;97:373–7.

Footnotes and Financial Disclosures

Originally received: March 5, 2009. Financial Disclosure(s):
Final revision: June 6, 2009. The authors have no proprietary or commercial interest in any materials
Accepted: June 10, 2009. discussed in this article.
Available online: November 5, 2009. Manuscript no. 2009-321.
1
The Ocular Oncology Unit, Cliniques Universitaires St-Luc, Université Correspondence:
Catholique de Louvain, Brussels, Belgium. Patrick De Potter, MD, PhD, Ocular Oncology Unit, Ophthalmology
2
Center of Biostatistics and Medical Documentation, Cliniques Universitaires Department, Cliniques Universitaires UCL St-Luc, 10 Avenue Hippocrate,
de Mont-Godinne, Université Catholique de Louvain, Yvoir, Belgium. 1200 Brussels, Belgium. E-mail: Patrick.Depotter@uclouvain.be.

40