Diagnostic approach to subcutaneous nodules in patients with neuroendocrine tumours treated with depot somatostatin analogs: a cross-sectional study.

BACKGROUND

The presence of cutaneous nodules in patients with gastroenteropancreatic neuroendocrine tumours (GEP-NETs) receiving depot somatostatin analogs (SSAs) is a diagnostic challenge as differential diagnosis between injection site reactions and metastases is essential.

OBJECTIVE

To characterize the clinical, radiological, cytological and histopathological features of subcutaneous nodules in patients with GEP-NETs treated with SSAs.

MATERIALS AND

METHODS

Retrospective, cross-sectional study of patients with GEP-NETs treated with SSAs in whom subcutaneous nodules were detected on routine abdominal computed tomography (CT) scans. High resolution and colour Doppler ultrasonography was performed. Those patients with inconclusive radiological studies went through fine-needle aspiration cytology (FNAC) and/or biopsy.

RESULTS

Twelve patients (five males, seven females) were included (six midgut carcinoid NETs, six pancreatic NETs). Three patients received intramuscular depot octreotide, seven subcutaneous lanreotide, and two both treatments. CT scan findings were nonspecific. Sonography revealed a hyperechoic pattern in recent injections, and a hypoechoic pattern with a characteristic hyperechoic peripheral rim in long-term injections (more than 3 months after injection). On colour Doppler sonography, nodules showed no signs of intralesional vascularity. Fine-needle aspiration cytology (FNAC) was performed in five patients, revealing a characteristic acellular proteinaceous material. Biopsy in four patients showed different reactional infiltrates around the acellular material.

CONCLUSIONS

High resolution and colour Doppler ultrasonography may be very useful for the differential diagnosis of subcutaneous nodules in patients with GEP-NETs treated with SSAs. FNAC and a biopsy are useful tests for confirmation of the diagnosis in patients with inconclusive findings. We propose a management algorithm.