PET/CT in the management of metastatic cervical lymphadenopathy from unknown primary site: a seven years retrospective study.

Abstract

Objectives: To evaluate the accuracy of PET/CT for the diagnostic evaluation of patients presenting cervical node metastasis of suspected unknown primary; furthermore to understand its relative clinical utility and relevance when compared to classic endoscopic investigation approach.

Materials & methods: A retrospective study was pursued, collecting information from clinical files of all patients who presented to the Portuguese Institute of Oncology - Oporto, from January 2005 to December 2011, with cervical node metastases whose primary hadn't been found, despite clinical examination and standard imaging (CT scan or MRI) and therefore were submitted to a PET/CT. Among those presenting with non-supraclavicular metastasis patients were subsequently analyzed according to: histopathology; those who performed examination under anaesthesia (EUA) for biopsies either before of after PET/CT.

Results: Eighty nine patients were included in the study. Detection rate was 32.6% with no statistically difference between those with supraclavicular metastases and those with metastases in higher cervical levels (p = 0.24). In this last group (n = 76), 43% patients had had PET/CT and an endoscopy associated with biopsies of the upper aerodigestive tract in different orders, to complete diagnostic workup in cases where the first performed was inconclusive. No statistically difference was found between these two methods (p = 0.25). Most of noticed false negatives were microscopic lesions located deep in the palatine tonsils.

Conclusions: PET/CT showed to be an useful tool when searching for primary tumours whether metastasis were supraclavicular or located in higher levels of the neck. Despite its good accuracy and detection of tumours previously undetected by EUA with biopsies (missed mainly due to sampling error), up-front negative scan shouldn't preclude performing endoscopies. Being evident that both tools are helpful. It was not possible in this study to find any evidence that could show which one of these two exams should be performed first.