Conjunctival melanoma and melanosis: a reappraisal of ... · Perspective Conjunctival melanoma and melanosis: a reappraisal of terminology, classiﬁcation and staging Bertil Damato

Perspective

Conjunctival melanoma and melanosis: a reappraisal ofterminology, classification and stagingBertil Damato MD PhD FRCOphth1 and Sarah E Coupland MBBS PhD FRC Path2

1Ocular Oncology Service, Royal Liverpool University Hospital, and 2Department of Pathology, School of Cancer Studies, University ofLiverpool, Liverpool, UK

ABSTRACT

This paper aims to stimulate debate on the terminology,classification, grading and staging of conjunctival melanosisand melanoma. We audited our results with 76 invasive con-junctival melanomas. Staging according to the sixth edition ofthe Tumour Node Metastasis (TNM) system did not corre-late well with tumour extent and outcome. Approximately50% of invasive melanomas were associated with ‘primaryacquired melanosis with atypia’, a term which in our opinionunderestimates the gravity of this disease. We also founddeficiencies in the grading, terminology and classification ofconjunctival melanocytic abnormalities. In summary, wesuggest that the term ‘primary acquired melanosis’ bereserved for clinical diagnosis. Histologically, this abnormalitycan be categorized more precisely as either ‘hypermelanosis’or ‘conjunctival melanocytic intraepithelial neoplasia(C-MIN)’. ‘Primary acquired melanosis without atypia’ can betermed more accurately as ‘C-MIN without atypia’. In view ofthe high risk of invasive melanoma, we suggest that ‘primaryacquired melanosis with atypia’ be termed ‘C-MIN’ withatypia, with the more severe changes regarded as melanomain situ. To improve objectivity in the reporting of C-MIN, wepropose a scoring system based on horizontal and verticalspread and degree of severity of melanocytic atypia. Wesuggest that the TNM staging system for conjunctivalmelanoma be revised to: (i) include a Tis stage; (ii) takeaccount of tumour size, quadrant and caruncular involve-ment; and (iii) improve staging of any local invasion beyondconjunctiva.

Key words: conjunctival neoplasm, disease-specific mortal-ity, melanoma, ophthalmology, pathology.

INTRODUCTION

The terminology of conjunctival melanotic abnormalities isconfusing. For example, the noun, ‘melanosis’, which refers tomelanotic pigmentation visible to the naked eye, is used toencompass both melanin hyper-secretion and melanocyticproliferation.1 Furthermore, there is disagreement as towhether melanocytic intraepithelial neoplasia with atypia ispre-cancerous or cancerous.2,3 As a result, the same disease ata particular point in the spectrum of malignancy may betermed ‘primary acquired melanosis with atypia’ (PAM withatypia) by one pathologist and ‘melanoma in situ’ by another,possibly risking under-treatment or over-treatment.

Classification of conjunctival melanocytic disorders isvaluable to clinicians and histopathologists considering thedifferential diagnosis of a particular case. Various systemshave been developed (see later text), but in our opinion,these all have limitations, such as including congenital ocularmelanocytosis, which does not involve conjunctiva, beingsub-conjunctival.4

The histological grading of PAM with atypia is variablydescribed using terms such as ‘mild’ and ‘severe’.5 There isscope for a system that grades this conjunctival melanocy-tic intraepithelial neoplasia (C-MIN) more objectively andreproducibly. This would make it easier to detect progressionwhen sequential biopsies are performed over a long periodand would enhance communication between the pathologistand ophthalmologist. Such a scoring system would also facili-tate multicentre collaboration when evaluating treatment.

Conjunctival melanomas are staged according to theTumour Node Metastasis (TNM) system, developed by theAmerican Joint Committee on Cancer and the Union Inter-national Contre Cancer.6 The objectives of this system are tostage disease in a standardized fashion so as to enhanceprognostication, treatment planning and multicentre studies.We feel that the sixth edition categorizes conjunctival mela-nomas into stages that do not correlate adequately with

� Correspondence: Dr Bertil Damato, Ocular Oncology Service, Royal Liverpool University Hospital, Prescot Street, Liverpool L7 8XP, UK. Email:

[email protected]

Received 16 April 2008; accepted 3 October 2008.

Clinical and Experimental Ophthalmology 2008; 36: 786–795doi: 10.1111/j.1442-9071.2008.01888.x

© 2008 The AuthorsJournal compilation © 2008 Royal Australian and New Zealand College of Ophthalmologists

mailto:[email protected]

prognosis. For example, bulbar conjunctival tumours movefrom stage 1 to stage 2 if there is any corneal involvement.

The aims of this article are to highlight shortcomings inthe terminology, classification, grading and staging of con-junctival melanocytic disorders and to stimulate debate onthese topics. It is not the intention to provide an encyclo-paedic review of the published literature, which is extensive.Only references relevant to our suggestions are thereforecited.

NORMAL CONJUNCTIVAL MELANOCYTES

Conjunctival melanocytes are normally dendritic and locatedexclusively in the basal layer of the epithelium, where theyare greatly outnumbered by basal squamous cells. The mel-anocytes secrete melanin into the adjacent epithelium, toprovide protection from ultraviolet light. The amount ofconjunctival melanin is not usually sufficient to be visible tothe naked eye.

MELANOSIS

The term, ‘ocular melanosis’, dates back to the early nine-teenth century.7 Originally, it included uveal and conjunc-tival melanomas. Today, it refers to flat, melanoticpigmentation that is visible on naked eye or slit-lampexamination. The term ‘melanosis’ tends also to be used todescribe histological appearances of hyperpigmentation,both intra- and extracellular.2,4 It is also used, inappropriatelyin our opinion, to describe a proliferation of melanocyteswithin the conjunctival epithelium. We feel it would beuseful to distinguish between these two forms of hyperpig-mentation by using the following terms: (i) ‘hypermelanosis’,to describe over-secretion and increased deposition ofmelanin; and (ii) ‘conjunctival melanocytic intraepithelialneoplasia (C-MIN)’ to refer to an increased population of

melanocytes, which may or may not demonstrate cytologicalatypia.

Primary conjunctival hypermelanosis

This consists histologically of an increased production ofmelanin by melanocytes that are normal in size, location andnumber (Fig. 1). These are the histological features of thecutaneous freckle (ephelis) and could be regarded as suchwhen they arise as small, faint, melanotic macules in lightskinned individuals. Diffuse bilateral conjunctival melanosisis more common in individuals with dark skin, such as thoseof African or Hispanic origin, in whom the widely used term‘racial melanosis’ is accepted. Some have advocated the term‘benign epithelial melanosis’;8 however, according to ourschema, primary conjunctival melanosis is by definitionbenign, because cytological features of malignancy are nec-essarily absent.

Secondary conjunctival melanosis

This term refers to conjunctival melanosis that is secondaryto an increased deposition of melanin in conjunctival epithe-lium as a result of: (i) conjunctival lesions such as inclusioncysts and squamous cell carcinoma (Fig. 2);9 and (ii) systemicconditions, such as Addison’s disease or treatment withcalcium channel blockers.10 The melanin can be intracellularas in squamous cell carcinoma, or extracellular, as in Addi-son’s disease.

NAEVUS

Conjunctival naevi are benign proliferations of melanocyticnaevus cells usually located predominantly in the substantiapropria.11 They vary greatly in their degree of pigmentation.Histologically, the main types are classified as: junctional(Fig. 3), compound, subepithelial, Spitz naevus and blue

a

b

Figure 1. Primary conjunctivalmelanosis. (a) Slit-lamp appearance,showing irregular diffuse conjuncti-val melanosis. (b) Light micrographstained with haematoxylin andeosin, showing a normal density ofdendritic melanocytes, with no fea-tures of atypia and with epithelialhypermelanosis.

Conjunctival melanoma and melanosis 787


naevus. Pure conjunctival intraepithelial naevi can occur butare exceptionally rare.

CONJUNCTIVAL MELANOCYTIC INTRAEPITHELIALNEOPLASIA (C-MIN)

By definition, all acquired conjunctival melanocytic intraepi-thelial proliferations are neoplastic. Some might suggest thatthe term ‘ocular surface melanocytic neoplasia’ should beused, because of the recent vogue for referring to conjuncti-val squamous cell neoplasms as ‘ocular surface squamousneoplasia’.12 We prefer the adjective ‘conjunctival’, because it

includes the palpebral conjunctiva, which is not located onthe ocular surface.

Conjunctival melanocytic intraepithelialneoplasia without atypia (syn. primaryacquired melanosis without atypia, benignconjunctival melanocytosis)

The clinical appearance is that of a unilateral, conjunctival,melanotic macule, which develops in later life (Fig. 4a). Thispigmentation tends to be irregular and to wax and wane,gradually becoming more extensive. It can become multifocal.

a

bFigure 2. Secondary conjunctivalmelanosis, exemplified by pig-mented, keratinized squamous cellcarcinoma on the temporal bulbarconjunctiva adjacent to the limbus.(a) Slit-lamp appearance. (b) Lightmicrograph stained with haema-toxylin and eosin (Black arrowshows melanin within carcinoma;white arrow shows melanin in mac-rophages in stroma).

a

bFigure 3. Conjunctival naevus. (a)Slit-lamp appearance. (b) Lightmicrograph stained with haema-toxylin and eosin. Arrow showsborder between intraepithelial andsub-epithelial naevus. Cysts arepresent in the lower part of thefigure and these are typical of con-junctival naevi.

788 Damato and Coupland


Histologically, there is a diffuse, intraepithelial melano-cytic proliferation, with increased numbers of melanocytesconfined to the basal layer of the conjunctival epithelium(Fig. 4b,c). These melanocytes are either normal or hyper-trophic, possibly with an increased number of dendrites, andshow no cytological features of atypia or of malignancy.

This condition is widely called ‘PAM without atypia’;1,2

however, we consider the term to be imprecise, especially asit also includes hypermelanosis without cellular proliferation(as mentioned earlier).13 One could consider defining thiscondition as ‘conjunctival intraepithelial melanocytic hyper-plasia’; however, by definition, hyperplasia indicates that thecellular proliferation is a reversible response to a physiologi-cal stimulus and it is not known whether this is indeed thecase. We propose the term ‘conjunctival melanocyticintraepithelial neoplasia (C-MIN) without atypia’, because(i) this proliferation of melanocytes within the conjunctivalepithelium does represent ‘neoplasia’; and (ii) there is bothclinical and experimental evidence suggesting that this con-dition can undergo transformation and progress to ‘C-MINwith atypia and ultimately to invasive melanoma’.1

Conjunctival melanocytic intraepithelialneoplasia (C-MIN) with atypia (syn. primaryacquired melanosis with atypia)

Clinically, this is indistinguishable from C-MIN withoutatypia (Fig. 5a). Histologically, it is characterized by a neo-plastic melanocytic proliferation with significant cellularpleomorphism but with no penetration of the basal mem-brane and, hence, no invasion of the substantia propria. The

melanocytic atypia varies in severity, from a proliferation ofsmall polyhedral melanocytes, with only subtle cellular pleo-morphism, to the presence of large melanocytes or epithe-lioid cells, usually with prominent nucleoli and possibly withconspicuous mitotic figures. The atypical melanocytes can beconfined to the basal layer of epithelium or can spread ver-tically to involve more superficial epithelial layers, eventuallyreplacing the entire epithelium. In many cases, the atypicalmelanocytes clump into ‘nests’, which may either be single,multifocal with a ‘skip-like pattern’, or confluent in either theradial or vertical direction (Fig. 5b). In addition, the intraepi-thelial atypical melanocytes can demonstrate a dissociatedsingle-cell (pagetoid) spread.

C-MIN with atypia is widely referred to as ‘PAM withatypia’.1,2,5 We consider the latter term to be imprecisebecause the word ‘melanosis’ is vague, as mentioned earlier,and because it encompasses: (i) melanocytic intraepithelialneoplasia with mild atypia; (ii) conjunctival melanoma in situ;as well as (iii) secondary pagetoid spread from an invasiveconjunctival melanoma. The term ‘conjunctival melanocyticdysplasia’ would be convenient as it incorporates both cyto-logical and architectural disruption as seen in ‘C-MIN’.However, the term ‘dysplasia’ has fallen out of favour inpathology circles, particularly when addressing melanocyticlesions, and is generally applied to pre-malignant changesinvolving epithelial cells. We propose the term ‘conjunctivamelanocytic intraepithelial neoplasia with atypia’ as replace-ment for ‘PAM with atypia’, because this is indeed a neoplas-tic process, as indicated by the presence of significantcytological atypia and the ultimate progression to invasivemelanoma in advanced stages. In addition, this terminology

ab

c

Figure 4. Conjunctival intraepi-thelial melanocytosis withoutatypia. (a) Slit-lamp appearanceshowing irregular pigmentation insuperior bulbar conjunctiva. (b)Light micrograph showing anincreased population of melano-cytes, which are located in the basallayer of the epithelium and whichdo not show cytological atypia (hae-matoxylin and eosin). (c) Lightmicrograph showing melanocytesidentified by immunohistochemistryusing MelanA.



has similar inferences as other intraepithelial neoplasia, inthe conjunctiva and elsewhere.

At present, C-MIN is generally graded as ‘PAM’ with‘mild’ or ‘severe’ atypia.5 Some also use the term ‘moderate’.4

Such terms are subjective, however, and therefore do showpoor inter- and intra-observer reproducibility. We wouldprefer a more objective grading system. The main risk factorspredicting invasive melanoma are degree of cytologicalatypia as well as pattern and extent of intraepithelialinfiltration.1,5,14 We suggest grading C-MIN according to: (i)the pattern of horizontal (radial) spread (i.e. basal, pagetoid,nesting); (ii) degree of vertical spread from basal layertowards surface (i.e. basal, <50% of epithelial thickness,50–90% epithelial thickness and >90% of epithelial thick-ness); and (iii) grade of cytological atypia (i.e. nuclear size,abundance of cytoplasm, mitotic rate and nucleoli). We haveprepared a score sheet to facilitate this process, which we arecurrently assessing for intra- and inter-observer repeatability(Fig. 6). This would be useful both clinically and for researchpurposes.

Assessment of the extent of melanocytic proliferation isenhanced by the use of immunohistochemical stains, such asMelanA/MART-1. Pancytokeratin is useful in demonstratingthe ‘mirror image’ to the MelanA stain, showing the propor-tion of conjunctival epithelial cells (be they surface epitheliaor inclusion cysts) that have been replaced by the atypicalmelanocytes. Imperative in the pathology report are: (i) anindication as to how much of the conjunctival biopsy isoccupied by these changes, in particular, by the most severealterations; and (ii) whether the C-MIN is present in any ofthe surgical resection margins. Further clinicopathologicalstudies are required to correlate these features with the risk ofinvasive disease, with and without treatment, and hence thecriteria for surveillance, biopsy and treatment.

CONJUNCTIVAL MELANOMA IN SITU

Melanoma in situ necessarily precedes all invasive disease,because most conjunctival melanomas arise from intraepithe-lial melanocytes.

Folberg et al. reported that nine out of ten (90%) caseswith vertical (i.e. non-basal) invasion of epithelium by mel-anocytes with atypia progressed to invasive melanoma. Irre-spective of such vertical intraepithelial spread, if themelanocytic atypia amounted to epithelioid cytology, then75% of patients developed invasive melanoma as comparedwith 4 out of 16 cases (25%).1 Sugiura et al. analysed 29 casesof ‘PAM with atypia’ and found invasive disease in 15/16cases with epithelioid melanocytes, even when these wereconfined to the basal layer (i.e. showing ‘lentiginous’growth).14 Shields et al. reported lower rates of invasive mela-noma;5 however, this might be because of treatment. Wefound C-MIN in 50% of 40 patients with invasive conjunc-tival melanoma.15 These histological features are broadlysimilar to those found in skin and mucous membranes (suchas the oral mucosa), which are generally termed ‘melanoma insitu’ (Fig. 7). For these reasons, there is an argument for refer-ring to all C-MIN as ‘conjunctival melanoma in situ’. Confir-mation of such a hypothesis can only be obtained byexamining the genotypic features of the atypical melano-cytes in the various stages of C-MIN and comparing themwith those in the invasive component.

The cut-off between ‘PAM with severe atypia’ and con-junctival melanoma in situ is vague. There seems to be ageneral reluctance to refer to C-MIN as conjunctival mela-noma in situ unless the abnormal melanocytes have com-pletely replaced the surface epithelial cells. This is possiblybecause of concerns that patients and ophthalmologistsmight become alarmed, so that excessive treatment is admin-istered, with the risk of causing unnecessary ocularmorbidity.5 These risks need to be weighed against the risk ofmetastatic death. In Sugiura’s study, four patients with epi-thelioid melanocytes (14%) developed metastases by thecompletion of their study.14 Our concern is that the term‘primary acquired melanosis with atypia’ might result inunder-treatment of early intraepithelial non-invasive neopla-sia, just when any opportunities for eradicating this lethaldisease might be greatest. Our term ‘conjunctival melano-cytic intraepithelial neoplasia’ may be less likely to providefalse reassurance. Interestingly, both Folberg and Sugiura

a bFigure 5. Conjunctival melano-cytic intraepithelial neoplasia. (a)Slit-lamp appearance, with diffuse,flat, conjunctival melanosis, withvariable pigmentation. (b) Lightmicrograph, showing nests of atypi-cal melanocytes in the basal area ofthe epithelium.



Figure 6. Score sheet for histological grading and staging of conjunctival melanocytic intraepithelial proliferation (includes the spectrum ofmelanosis, benign melanocytosis and conjunctival melanoma in situ).



recommend treatment of all patients with any degree ofintraepithelial melanocytic atypia, with the goal of eradicat-ing this disease and minimizing the risk of invasivemelanoma.1,14

Whether intermediate grades of C-MIN are regarded aspre-malignant disease or melanoma in situ, there are manypatients who are managed by delaying treatment until pro-gressive disease is confirmed by sequential biopsy. In suchcases, it is hoped that our scoring system will enhance com-munication between pathologist and surgeon so that pro-gressive disease can be detected and treated more quickly.

Patients and ophthalmologists should not be undulyalarmed by the term ‘conjunctival melanoma in situ’ if it isexplained to them that, with timely and effective treat-ment, the risk of metastatic spread is minimal, because ofthe absence of lymphatic channels in the conjunctivalepithelium.

INVASIVE CONJUNCTIVAL MELANOMA

It is conventional practice to label conjunctival melanomas as‘invasive’ even when this is minimal. Such invasion is a keypoint in disease progression, because it provides the mela-noma with access to lymphatic channels in the conjunctivalsubstantia propria, and hence a route for regional and sys-temic metastases.

Conventionally, invasive conjunctival melanomas areclassified according to whether they originate from intraepi-thelial neoplasia, a naevus or de novo.2 To our knowledge,

however, there is no convincing evidence that derivation ofa melanoma from a naevus influences outcome in any way.We would prefer to classify conjunctival melanomas accord-ing to factors that influence outcome, that is, according to:(i) whether they show a discrete or diffuse infiltrative pattern(Fig. 8a,b); and (ii) whether or not they are associated withC-MIN, which may consist of either pre-existing melanomain situ, secondary spread from the invasive tumour, or both(Fig. 9).

a bFigure 7. Light micrographs of insitu cutaneous melanoma. (a) Hae-matoxylin and eosin. (b) MelanAstain demonstrating the atypicalmelanocytes.

a bFigure 8. Invasive conjunctivalmelanoma. (a) Nodular tumourencircling the limbus. (b) Diffusemelanoma with sub-epithelialinfiltration. Note how the clinicalappearances are similar to those ofmelanocytic intraepithelial neopla-sia (Fig. 5a).

Figure 9. Invasive conjunctival melanoma with adjacent intraepi-thelial disease.



CLINICAL STAGING OFCONJUNCTIVAL MELANOMA

The TNM system stages tumour size and extent for thepurposes of planning treatment, estimating prognosis, evalu-ating outcomes and facilitating multicentre studies. Thesixth edition of the TNM system stages conjunctival mela-nomas as follows: T1 for bulbar tumours; T2 for tumoursinvolving cornea; T3 for tumours involving non-bulbar con-junctiva; and T4 for tumours invading globe, eyelid skin,orbit, nasal sinus or brain.6 This system suffers from severallimitations because it: (i) excludes a Tis stage; (ii) takes noaccount of tumour size, which is known to influence treat-ment, if not survival; (iii) implies that corneal invasionworsens the survival prognosis, although to our knowledgethere is no evidence for this;16 (iv) does not consider tumourquadrant, whereas our data suggest that medial tumour loca-

tion is associated with a worse prognosis (Fig. 10a,b);17 (v)merges all non-bulbar conjunctival areas into one groupwhen our studies suggest that caruncular involvement maybe associated with a worse prognosis than forniceal or palpe-bral location (Fig. 10c,d);18 (vi) suggests that intraocularinvasion is associated with a worse prognosis than caruncu-lar involvement, whereas the opposite is likely to be thecase, the eye not having any lymphatic channels; andbecause it (vii) categorizes eyelid and brain involvementequally, whereas the latter is likely to be associated with alower life-expectancy.

On the basis of our studies, we have made a number ofrecommendations to the current TNM committee revisingthe sixth to the seventh edition, due for publication in 2009.Several of our proposals have been accepted: Stages T1 andT2 invasive melanomas are categorized according tonumber of quadrants of conjunctiva affected. Non-bulbar

Figure 10. (a) Cumulative ‘bubble plot’ showing distribution of invasive conjunctival melanoma according to local tumour recurrence. Eachbubble indicates an area of conjunctiva, with the size of the bubble indicating the percentage of patients showing tumour involvement of thatarea. (b) Kaplan–Meier curves showing time to local tumour recurrence according to coronal tumour location. (c) Cumulative ‘bubble plot’correlating distribution of invasive conjunctiva with metastatic death. (d) Kaplan–Meier curves showing time to metastastic death accordingto involvement of caruncle. In (a) and (c), results from left eye were transposed to right eye. From Damato and Coupland.17



conjunctival melanomas are subdivided according towhether or not the caruncle is involved. T3 tumoursshow local invasion beyond conjunctiva to globe, eyelid,orbit and sinuses. T4 melanomas involve central nervoussystem. Further studies are needed to correlate long-term outcomes with baseline disease at the time of initialtreatment.

CLASSIFICATION OF PIGMENTED LESIONS OFTHE CONJUNCTIVA

Several classifications have been proposed for conjunctivalmelanocytic abnormalities. Two in particular are widelyused and these have been developed by the World HealthOrganisation and the Armed Forces Institute of Pathol-ogy.2,4 Both classifications of conjunctival melanocyticdisease have limitations. First, both include ‘ocular melano-sis’, which is non-conjunctival, and a melanocytosis, nothypermelanosis. Second, both refer to conjunctival mela-noma in situ as ‘primary acquired melanosis’. Third, theWHO classification considers conjunctival melanoma in situas ‘pre-cancerous’ (i.e. pre-neoplastic) instead of cancerous.Furthermore, it does not include naevus in the pre-cancerous group, even though this can give rise to mela-noma, albeit rarely.

We propose a different classification, which overcomesthe problems of the WHO and AFIP systems (Table 1).

SUMMARY AND CONCLUSIONS

In this article, we have highlighted shortcomings in theterminology and classification of melanosis and conjunctivalmelanocytic intraepithelial proliferations and in the stagingof melanoma.

The literature is replete with terms and classifications forintraepithelial melanocytic proliferations, which must haveseemed as valid to their proponents as our suggestions seem tous. We accept that, despite its shortcomings, ‘PAM’ may behere to stay, if only because it comes off the tongue so easily.We would argue, however, that the term ‘C-MIN’ is moreprecise and just as easy to pronounce! Although ocularoncologists recognize the serious implications of intraepithe-lial melanocytic neoplasia, our impression from patient refer-rals to our centre suggests that some general ophthalmologistsand pathologists are being misled by term ‘PAM’ and itspossibly harmless connotations so that treatment may besub-optimal.

In 1985, Folberg et al., in their landmark paper, cited Hen-kind’s astute observation that ‘much of what has been writtenabout pigmented lesions of the conjunctiva is either anec-dotal, speculative, or controversial’.1,18 This situation still per-sists 23 years later. Multicentre studies are required toaccumulate sufficient evidence for progress to occur, butthese in turn depend on adequate and uniform documenta-tion of disease severity and extent, which we hope this paperwill improve.

ACKNOWLEDGEMENT

Research Support: The Ocular Oncology Service in Liver-pool is funded by the National Commissioning Group of theNational Health Service.

REFERENCES

1. Folberg R, McLean IW, Zimmerman LE. Primary acquiredmelanosis of the conjunctiva. Hum Pathol 1985; 16:129–35.

2. Campbell RJ. Histological Typing of Tumours of the Eye and Its Adnexa.Berlin: Springer, 1998; 15–20.

3. Ackerman AB, Sood R, Koenig M. Primary acquired melanosisof the conjunctiva is melanoma in situ. Mod Pathol 1991; 4:253–63.

4. McLean IW, Burnier MN, Zimmerman LE et al. Tumors of the Eyeand Ocular Adnexa. Washington: Armed Forces Institute ofPathology, 1994.

5. Shields JA, Shields CL, Mashayekhi A et al. Primary acquiredmelanosis of the conjunctiva: risks for progression to melanomain 311 eyes. The 2006 Lorenz E. Zimmerman lecture. Ophthal-mology 2008; 115: 511–19.

6. Sobin LH, Witterkind C. TNM Classification of Malignant Tumours,6th edn. New York: Wiley-Liss, 2002.

7. Mackenzie W. Melanosis of the Eyeball. In: A Practical Treatise onthe Diseases of the Eye. Philadelphia, PA: Blanchard and Lea, 1855;689–97.

8. Jakobiec FA. The ultrastructure of conjunctival melanocytictumors. Trans Am Ophthalmol Soc 1984; 82: 599–752.

Table 1. Proposed classification of melanocytic conjunctivallesions

MelanosisPrimary

FreckleRacial

SecondaryOcular diseaseSystemic disease

Conjunctival naevusJunctional naevusCompound naevusSubepithelial naevusSpitz naevus (epithelioid/spindle cell)Blue naevus

Conjunctival melanocytic intraepithelial neoplasia (C-MIN)Without atypiaWith atypia

Invasive conjunctival melanomaWithout intraepithelial neoplasia (i.e. de novo)With primary intraepithelial neoplasiaWith secondary intraepithelial neoplasia (i.e. pagetoid spreadfrom invasive tumour)

Secondary conjunctival melanomaFrom skinFrom uvea

Metastatic conjunctival melanomaFrom cutaneous melanoma

C-MIN, conjunctival melanocytic intraepithelial neoplasia.



9. Shields CL, Manchandia A, Subbiah R et al. Pigmented squa-mous cell carcinoma in situ of the conjunctiva in 5 cases.Ophthalmology 2008; 115: 1673–8.

10. Gloor P, Alexandrakis G. Clinical characterization of primaryacquired melanosis. Invest Ophthalmol Vis Sci 1995; 36: 1721–9.

11. Shields CL, Fasiudden A, Mashayekhi A et al. Conjunctivalnevi: clinical features and natural course in 410 consecutivepatients. Arch Ophthalmol 2004; 122: 167–75.

12. Pe’er J. Ocular surface squamous neoplasia. Ophthalmol Clin NorthAm 2005; 18: 1–13.

13. Guillen FJ, Albert DM, Mihm MCJ. Pigmented melanocyticlesions of the conjunctiva – a new approach to theirclassification. Pathology 1985; 17: 275–80.

14. Sugiura M, Colby KA, Mihm MC Jr et al. Low-risk and high-risk histologic features in conjunctival primary acquired mel-anosis with atypia: clinicopathologic analysis of 29 cases. Am JSurg Pathol 2007; 31: 185–92.

15. Damato BE, Coupland SE. Clinical mapping of conjunctivalmelanomas. Br J Ophthalmol 2008; 92: 1545–9.

16. Tuomaala S, Aine E, Saari KM et al. Corneally displaced malig-nant conjunctival melanomas. Ophthalmology 2002; 109: 914–19.

17. Damato B, Coupland SE. An audit of conjunctival melanomatreatment in Liverpool. Eye 2008; doi: 10.1038/eye.2008.154.

18. Henkind P. Conjunctival Melanocytic lesions: natural history.In: Jakobiec FA, ed. Ocular and Adnexal Tumors. Birmingham:Aescalapus, 1978; 572–82.



Documents

Conjunctival melanoma and melanosis: a reappraisal of ... · Perspective Conjunctival melanoma and melanosis: a reappraisal of terminology, classiﬁcation and staging Bertil Damato