Sensitivity of routine abdominal multidetector computed tomography in patients with pancreatic solid tumour proven with endoscopic ultrasound-guided fine-needle aspiration – Preliminary results from a tertiary referral center

Introduction: The diagnosis of pancreatic tumour, especially in early stages is sometimes challenging. Multi-detector computed tomography (MDCT) is generally accepted as one of the most accessible and sensitive method to demonstrate pancreatic masses, but there is an increasing number of evidence, that endoscopic ultrasound (EUS) is superior to MDCT both in sensitivity and specificity especially when combined with fine-needle aspiration (FNA). The aim of our preliminary retrospective analysis was to determine the value of routine abdominal MDCT in patients with solid pancreatic malignacy proven with EUS-FNA.

Patients and methods: In 2012, 53 pts with the suspicion (clinical signs, positive abdominal ultrasound or ERCP) of pancreatic malignacy were referred to our endoscopic lab for pancreatic EUS-FNA. Based on the radial EUS findings FNA was performed with linear echoendoscope according to our protocol (22 or 25 G FNA-needle with at least 4 needle passes). All patients had contrast-enhanced, dual-phase MDCT with general abdominal protocol before or after the EUS within 2 weeks. Radiologists were blinded according to the EUS-FNA results but were informed about the suspition of pancreatic tumour. Both EUS and MDCT results were analized retrospectively.

Results: Cytology revealed 36 malignancy of 39 FNAs (3 neuroendocrine tumour and 33 adenocarcinoma; 31 pancreatic head, 4 body, 1 tail). The mean tumour size on the radial EUS was 31.8 (19 – 45)mm. MDCTs of the 36 patients with cytologically proven pancreatic malignacy were positive in 22 cases (61.1%). The radiologist saw enlarged pancreatic head in 3 (8.3%), and dilated main pancreatic duct in 2 pts (5.5%) of the remaining 14 cases as indirect signs. More importantly in 9 patients (25%) in whom EUS-FNA proved malignancy the CT result was judged as negative. The overall sensitivity of routine MDCT was 61.1%.

Conclusion: According to our single-center preliminary results, the sensitivity of routine MDCT is not acceptable in cases of highly suspected pancreatic malignacy. Further analysis of our data is needed to evaluate the factors influencing the MDCT results, but it is supposed that in cases with high suspceptibility of pancreatic cancer MDCT with special pancreatic protocol should be performed to increase the sensitivity of the „gold standard” method.