Pancreatic cancer with portal vein invasion diagnosed by endoscopic ultrasound.

An 82-year-old woman with abdominal pain for the preceding 6 months was referred for endoscopic ultrasound (EUS) to evaluate the presence of pancreatic cancer. The patient had a history of diabetes and hypertension. A computed tomography (CT) scan showed a nodular image on the pancreatic body, hypovascularized, measuring 2.5 cm × 2.0 cm, without signs of vascular involvement [Figure 1]. It was also observed, nodular images on the liver, the visceral peritoneum and peri-hepatic lymphonodes.

Figure 1

Computed tomography images of the pancreatic lesion without vascular invasion

One month later, the patient underwent an EUS that showed a hypoechoic heterogeneous lesion in the body of the pancreas, measuring 4.0 cm × 3.5 cm, with vascular invasion of the portal vein [Figure 2]. Nodular masses in the visceral peritoneum and moderate ascites were also observed [Figure 3]. An EUS fine-needle aspiration of the pancreatic mass was performed with a 22-guage Echo-Tip® needle (Cook Medical Inc., Limerick, Ireland) [Figure 4] and a sufficient cytology material was obtained after three punctures, suggesting malignant cells [Figure 5]. After discussion with the surgery department, it was opted to perform a celiac plexus neurolysis, by injecting absolute alcohol (20 mL) and lidocaine (5 mL) through a fenestrated needle on the celiac plexus.

Figure 2

Endoscopic ultrasound view of a hypoechoic lesion in the body of the pancreas with portal vein invasion

Figure 3

Endoscopic ultrasound view of nodular images in visceral peritoneum and ascitis

Figure 4

Endoscopic ultrasound (EUS) view of EUS fine-needle aspiration

Figure 5

Cytology smear suggesting malignant cells

The optimal approach to preoperative imaging assessment of pancreatic cancer remains unclear in the literature. In a prospective study, Soriano et al.,[1] demonstrated that helical CT had the highest accuracy in assessing extent of primary tumor (73%), locoregional extension (74%), vascular invasion (83%), distant metastases (88%), tumor TNM stage (46%), and tumor resectability (83%), whereas EUS had the highest accuracy in assessing tumor size and lymph node involvement (65%). Furthermore, in a retrospective analysis conducted by Zhang et al.,[2] CT angiography, can provide reliable information for vascular involvement and general resectability of pancreatic malignant tumors.

On the other hand, in a retrospective study conducted by Buchs et al.,[3] EUS was the most valuable imaging modality in assessing vascular invasion (especially for venous invasion) for pancreatic cancer, with an accuracy of more than 80%. In a 2013 meta-analysis,[4] with a total of 1330 patients, CT scan showed lower sensitivity than EUS for vascular invasion (58% vs. 86%); however, the specificities for vascular invasion (95% vs. 93%) were comparable in studies where both imaging techniques were performed. In other systematic review published in 2007, Puli et al.,[5] showed that the pooled sensitivity of EUS in diagnosing vascular invasion was 73% and the pooled specificity was 90.2%.

The patient underwent a CT scan as CT angiography was not available, but no signs of vascular involvement were observed. Although the EUS didn’t change the treatment on this case, it was possible to demonstrate a portal vein invasion, proceed histological diagnosis and perform celiac plexus neurolysis.