Primary oesophageal melanoma—recognition and evolution of management

Melanoma is one of the most common malignancies diagnosed in New Zealand with a reported incidence of approximately 55 per 100,000; one of the highest in the world.¹ The vast majority of cases worldwide are cutaneous melanoma, whereas primary oesophageal melanoma (POM) only accounts for <0.05% of malignant melanomas and 0.1–0.2% of all oesophageal tumours.² There are only a few hundred case reports of this entity worldwide.³ In this report, we examine the case of a woman in her 50s of mixed Māori and European heritage who presented with progressive oesophageal dysphagia to solids over 3 months and an associated 5kg weight loss.

Case report

Our patient was referred for direct-access endoscopy by the ear, nose and throat (ENT) service who had been managing a benign right thyroid nodule and came across her dysphagia incidentally. Past medical history includes a known atrial septal defect, uterine leiomyoma and previous caesarean section. Her body mass index was 38. She never smoked and had minimal alcohol consumption. She underwent upper gastrointestinal endoscopy, which identified a large partially obstructing oesophageal mass with stigmata of recent bleeding in the lower third of the oesophagus, 35cm from the incisors (Figure 1).

Staging scans were performed and an enlarged epigastric lymph node with mild fluorodeoxyglucose (FDG) uptake was suspicious for nodal metastasis (Figure 2 and 3). No other FDG avid nodes were seen, nor distant metastatic disease. Histology from the endoscopic biopsy revealed squamous mucosa with high grade poorly differentiated epithelioid cells that were positive for SRY-related HMG-box gene 10 (SOX10) and negative for cytokeratins, neuroendocrine and squamous markers. The features were consistent with melanoma. BRAF mutational analysis was negative. Complete skin, pelvic and anorectal examination has not revealed another primary melanoma lesion.

After a melanoma multidisciplinary meeting (MDM) discussion she was referred for surgical resection with curative intent. She underwent a two-stage minimally invasive oesophagectomy and pyloroplasty without serious complications. The resected specimen showed 6 foci of invasive melanoma arising from in situ melanoma, with the largest measuring 42x25x20mm (Figure 4 and 5). Deepest invasion was to submucosa without lymphovascular or perineural invasion. Resection margins were clear. All 36 resected nodes were negative for malignancy. Final pathological staging is pT1bN0M0. Additional targeted mutational analysis of the resection specimen did not show mutations in KIT, KRAS, NRAS, GNA11 and GNAQ genes. Adjuvant therapy was not recommended by MDM, but she will have surveillance CT at 6-month intervals for 3 years, then annually for 5 years under the upper gastrointestinal surgery service.

View Figure 1–5.

Discussion

The first histologically confirmed case of POM was described in literature in 1952.⁴ A recent systematic review found that the mean age of diagnosis is 61.⁵ About half of cases have metastasised to other organs by the time of diagnosis, and the overall prognosis is poor with an estimated median survival of 10 months and 2–5% at 5 years.³ The pathogenesis and risk factors for POM remain poorly understood but it is known that 4–8% of the general population have melanocytes present in the oesophageal mucosa.^5–7

The clinical manifestations of POM include dysphagia (75%), retrosternal pain (69%), epigastric discomfort (8.6%) and overt gastrointestinal bleeding (8.6%).^3,5 While the vast majority (90%) of cases are found in the middle and lower third of the oesophagus, only 73–85% of POM cases exhibit melanocytic pigmentation.^3,5 Thus, POM is still a differential for amelanotic polypoid lesions in the oesophagus necessitating histopathological examination for a confirmed diagnosis.⁷ Presence of melanin granules with tumour cells is a cornerstone criterion for the histopathological diagnosis of POM.⁸ Twelve percent of cases involved multifocal lesions. Immunohistochemistry for S-100 protein, Melan-A, HMB-45, SOX10 and KBA.62 also aid in diagnosis.³ Of note, the commonly detected BRAF mutation for cutaneous melanoma is typically not detected in POM, which may instead harbour mutations in other genes such as KIT, influencing the selection of targeted immunotherapy agents.⁸

Like other solid organ malignancies, staging of POM is crucial for both treatment decision and prognosis, as up to 25% of cases worldwide used PET/CT to rule out distant metastasis.⁵ Both lymph node metastases and the overall TNM staging with the American Joint Commission on Cancer staging classification for the oesophagus have been demonstrated to be independent predictors for POM prognosis.⁹ Surgical resection is the gold standard for treatment with curative intent in POM without distant metastases, and while there is no established standard systemic therapy in the adjuvant or metastatic settings, there is increasing evidence that immunotherapy offers significantly extended overall survival compared to limited efficacy from chemotherapy or locoregional radiotherapy.^3,5 There is high disease recurrence after curative resection for POM occurring in 40% of cases. The median time to recurrence is 6 months.⁵ The risk of recurrence extends to at least 5 years, which highlights the importance of surveillance post-resection of POM.¹⁰

Our case report fits with the classic descriptions and features of POM but has an amelanotic endoscopic appearance as well as favourable pathological staging after receiving standard of care surgical resection. Her favourable prognosis is partly attributable to the prompt recognition by the ENT team that progressive dysphagia requires urgent endoscopic evaluation. Due to its rarity, post-resection management remains poorly defined and underinvestigated. Further studies in this area will help address the high disease recurrence and improve overall survival.