A. C. E. Mccartney

Prognostic factors in primary malignant melanoma of the conjunctiva: a clinicopathological study of 256 cases.

A series of 256 consecutive cases of invasive primary conjunctival malignant melanomas was examined to identify clinical and histopathological prognostic factors. The follow up period varied between 0.3 and 45.9 years (mean 9 years, median 6.3 years). The 5 year survival rate was estimated at 82.9%, the 10 year survival rate at 69.3%. Multiple regression analysis with the Cox proportional hazards model was used to assess sex, age, and a number of baseline features of conjunctival malignant melanoma as possible prognostic factors influencing melanoma related mortality. In assessing each potential prognostic factor, the effects of all other factors were taken into account in the modelling process. Tumours in unfavourable locations--that is, those involving the palpebral conjunctiva, fornices, plica, caruncle, and lid margins, were associated with 2.2 times higher mortality compared with (epi)bulbar melanomas. Patients with mixed cell type tumours had about three times higher mortality compared with those with pure spindle cell melanomas, and histological evidence of lymphatic invasion by tumour cells was also a prognostic feature, carrying a fourfold increase in the death rate. Multifocal tumours were associated with a fivefold increase in mortality among those with tumours in favourable (epi)bulbar locations, but were not prognostic in patients with melanomas in unfavourable sites. The death rate was significantly higher in those with initial tumour thickness of more than 4 mm, but only among patients with unfavourably located melanomas. Sex, age, and clinical origin of the tumour (primary acquired melanosis, pre-existing naevus, or de novo) were not useful prognostic indicators in this study.

Orbital exenteration in 95 cases of primary conjunctival malignant melanoma.

The role of orbital exenteration in the management of malignant melanoma of the conjunctiva has been underexplored. The outcome in 95 patients with this condition, who underwent exenteration as a primary treatment (n = 36) or after failure of other treatment (n = 59) for early to advanced stages of the disease, was evaluated. The majority of treated cases had multicentric melanomas sited at prognostically unfavourable locations. In the group of tumours with a maximum thickness of 1.0 mm no melanoma related mortality was noted. Melanomas thicker than 1.0 mm were associated with a mortality varying between 33% and 50%, independent of whether exenteration was performed as primary or secondary treatment. An especially poor outcome was noted for the group of caruncular melanomas despite exenteration. These findings indicate that total eradication of tumour should be performed at an early stage. For this purpose, a combination of debulking surgery and adjunctive cryotherapy or beta radiotherapy is more appropriate than orbital exenteration which causes disfigurement and blindness. Exenteration of the orbit should be reserved as a palliative procedure for advanced stages of neoplastic disease.

Impression cytology of conjunctival melanosis and melanoma.

Impression cytology using cellulose acetate paper has been used in various ocular surface disorders as a simple non-invasive diagnostic test. To assess its value in differentiating melanocytic tumours, 24 patients with a range of pigmented lesions of the conjunctiva were examined using this technique. Cytological and histological diagnoses were compared in 23 cases. In 73% of cases impression cytology predicted the histological diagnosis by detection of superficial atypical melanocytes and their proportion relative to benign epithelial cells. This pilot study shows impression cytology to be a useful diagnostic aid in the differentiation of pigmented tumours of the bulbar conjunctiva.

Malignant melanomas of the conjunctiva, nasal cavity, and paranasal sinuses.

Four patients had malignant melanoma involving the conjunctiva, nasal cavity, and paranasal sinuses. In each patient, the conjunctival melanoma preceded recognition of the lesion in the nasal cavity by months or years. Each patient had one or more conjunctival biopsy specimens that documented melanoma, and in each patient the melanomas evolved in primary acquired melanosis of the conjunctiva. Histologic examination of the epithelium adjacent to the tumor in the nasal cavity failed to show either atypia or diffuse melanosis. We believe that these melanomas in the nasal cavity and paranasal sinuses are most likely to evolve as regional metastases, although it is possible that they arise de novo or as related foci in accordance with the multicentric characteristics of conjunctival melanoma. If a patient with a conjunctival melanoma has symptoms referable to the nasal cavity or paranasal sinuses, ophthalmologists should refer the patient to an otolaryngologist.

Premalignant melanosis of the conjunctiva and the cornea in xeroderma pigmentosum.

Xeroderma pigmentosum is a rare autosomal recessive dermatosis. The neoplastic changes in sunlight-exposed areas of the skin and eyes may be related to the impaired replication of ultraviolet radiation-damaged DNA. A 38-year-old Greek woman is reported with a mild form of xeroderma pigmentosum and primary acquired melanosis with atypia of her right limbal conjunctiva and cornea. The development of this precursor of conjunctival malignant melanoma in a xeroderma pigmentosum patient may support the putative role of sunlight exposure in malignant transformation of conjunctival melanocytes.

Pigmentation of the Eyelid Margin Accompanying Conjunctival Melanoma

In a review of patients with histologically verified malignant melanoma of the conjunctiva, we found 17 patients who had heavy pigmentation of the ipsilateral eyelid margin that was visible on external examination. Among nine patients in whom the histologic characteristics of the skin pigmentation could be studied, the features ranged from an increase in pigmentation of melanocytes within the basal layer to invasive malignant melanoma. The presence of heavy skin pigmentation in association with conjunctival melanoma appeared to be an ominous sign, even when the eyelid did not show histologic features of malignancy. Twelve of the 17 patients (70%) died of metastases after histologic verification of the conjunctival melanoma.

Nasal and orbital recurrence of conjunctival melanoma 21 years after exenteration.

In our experience 5% of invasive malignant melanomas of the conjunctiva arising from areas of primary acquired melanosis with atypia spread to the ipsilateral nasal cavity and paranasal sinuses. Twenty one years after orbital exenteration for multicentric conjunctival melanoma an 82-year-old man was seen with an orbital recurrence, which had extended to the nasal cavity and paranasal sinuses through the nasolacrimal duct without invading the mucosa. This previously undescribed way of spread after the longest symptom-free interval following exenteration ever reported is illustrated.

AgNOR counts in conjunctival malignant melanoma lack prognostic value.

Using a silver staining technique, argyrophilic nucleolar organiser region-associated proteins (AgNORs) have been studied in routinely processed paraffin sections of 46 invasive malignant melanomas (MM) of the conjunctiva. The aim of this retrospective study was to assess the value of the AgNOR method as a prognostic indicator for this neoplasm. The 46 cases were divided into two groups: (A) 14 cases of MM that metastasised and caused death of the patient within 5 years of (histological) diagnosis, and (B) 32 cases of MM that did not metastasise and in which patients survived beyond 5 years. The mean of the AgNOR counts per nucleus was 7.03 (95% CI: 5.81-8.24) in group A, and 7.15 (95% CI: 6.53-7.77) in group B. A comparison using a multifactor analysis of variance (ANOVA) model, which corrected for possible confounding effect of tumour thickness, site, and cell type showed no significant difference in AgNOR counts between groups A and B (p = 0.8). Analysis by the Cox proportional hazards regression model showed that survival was not influenced significantly by the mean AgNOR number (hazard ratio: 0.92). Whereas the AgNOR technique may be used to distinguish benign from malignant melanocytic lesions of the conjunctiva, we conclude it has no value in predicting the outcome for patients with conjunctival MM.