Zosteriform metastasis from melanoma

Approximately 10% of metastases from all primary neoplasms involve the skin, but for malignant melanoma the figure is 44%.¹ In some cases of melanoma this is the presenting feature, either because the primary lesion has regressed completely or because it has been unnoticed or ignored by the patient. Occasionally the melanoma has originated at an extracutaneous site such as the retina or the anal canal.

Metastases from cutaneous melanoma normally present as flesh coloured papules or nodules in the skin. Only about a third are pigmented or ulcerated. We report a case in which cutaneous metastases from a melanoma imitated herpes zoster. This presentation is known as zosteriform metastasis; it also occurs with other neoplasms.

Case report

A 73 year old white man presented with a three week history of painful, pruritic vesicles on a background of erythema on the right frontal area of the scalp (figure). The lesion had not responded to self prescribed topical antibiotics and antiseptics. The patient had grown up in South Africa and had a history of excessive exposure to the sun. He had previously developed three basal cell carcinomas and had numerous actinic keratoses. Five years earlier a malignant melanoma, Breslow thickness 1.25 mm, had been excised from his right shoulder. At that time there had been no evidence of metastatis. He had been followed up regularly and there had been no evidence of recurrence. He also had a five year history of stage Ib mycosis fungoides.

Clinical examination showed that the lesion lay within the area supplied by the ophthalmic branch of the right trigeminal nerve. A provisional diagnosis of herpes zoster (shingles) was made, and he was treated with 800 mg of oral aciclovir five times a day. There was no history of herpes zoster at the same site or elsewhere.

He was reviewed seven days after starting aciclovir. The lesions had extended slightly but otherwise remained unchanged. At that point a diagnosis of plaque stage mycosis fungoides was considered, and the scalp lesion was biopsied. Histological examination showed that the dermis was heavily infiltrated by non-pigmented malignant cells, which were epithelioid in appearance and forming nests; immunostaining showed that these cells were metastatic melanoma.

The area was too extensive to excise. After discussion with the patient the lesion was treated with electron beam radiotherapy (40 Gy in 15 fractions). The response was dramatic and within a few weeks the lesion regressed almost completely except for some residual macular erythema.

Discussion

Many different malignant tumours can metastasise to the skin but the commonest primary sources are tumours of breast, stomach, lung, and uterus. These lesions usually present as firm papules or nodules, both of which may ulcerate; occasionally they are inflammatory, sclerotic, bullous, or vesicular. Zosteriform metastasis is less well known; it may arise from adenocarcinoma of the lung,²,³ carcinoma of the prostate,⁴,⁵ Kaposi's sarcoma,⁶,⁷ transitional cell carcinoma of the bladder,⁸ or malignant melanoma.⁹,¹⁰ Zosteriform metastases are usually painful, tender, or pruritic and consist of vesicles on a background of erythema, imitating the appearance of shingles. They are commonly confined to a single unilateral dermatome, adding to the potential for misdiagnosis. Our patient presented with zosteriform metastases which we initially misdiagnosed as herpes zoster.

It is obviously important to recognise metastatic disease. Firstly, it may indicate an occult malignancy and call for a vigorous search for the primary lesion, which may be internal; if the primary lesion is cutaneous it may have been unnoticed or ignored by the patient. Secondly, metastatic disease may represent progression of a known malignancy, requiring a change in management.

The mechanism for the zosteriform appearance of metastatic disease is not known. It has been postulated that recent herpes zoster might induce infiltration of malignant cells in a Koebner-like phenomenon.⁶,¹¹ Our patient had no history of herpes zoster or any other skin lesion in this area. Perineural lymphatic spread of malignant cells has been suggested as a mechanism²,⁴,⁵,¹²; invasion of the dorsal root ganglia with subsequent peripheral spread is another hypothesis.³,⁸,¹² Our case also highlights the value of histological examination in the management of skin lesions that do not respond to treatment.

Figure.

Zosteriform metastasis from malignant melanoma