Differentiation of solid-pseudopapillary tumors of the pancreas from pancreatic neuroendocrine tumors by using endoscopic ultrasound.

BACKGROUND AND AIM

To differentiate solid-pseudopapillary tumors (SPTs) of the pancreas from pancreatic neuroendocrine tumors (pNETs) by endoscopic ultrasound.

METHODS

We retrospectively reviewed all patients with SPTs and pNETs who underwent endoscopic ultrasound (EUS) from May 2012 to August 2018 at the Fudan University Shanghai Cancer Center. We included patients confirmed pathologically with a surgical biopsy or with endoscopic ultrasound-guided fine needle aspiration (EUS-FNA). The demographic data of the patients, characteristics of the lesions and overall survival data of patients with these two diseases were further compared.

RESULTS

A total of 147 pNET patients and 21 SPT patients were included in our study. The mean ages of the patients in the SPT and pNET groups were 35.95years and 54.30years, respectively. There were more females in the SPT group than in the pNET group (71.43% vs. 40.82%). The patients in the pNET group had significantly more lymphatic metastases and visceral organ metastases than the patients in the SPT group. A larger proportion of pNET lesions than SPT lesions had homogeneous echo patterns and were hypervascular. Cystic components and calcification components were more often observed in the SPT lesions than in the pNET lesions. In the multivariate logistic regression analysis, the hypervascularization (OR: 6.528, 95% CI: 1.562-27.285, P=0.010) and cystic component (OR: 0.106, 95% CI: 0.019-0.597, P=0.011) variables resulted in the best discrimination of patients with SPTs from patients with pNETs. Survival among patients with SPTs was higher than that among patients with pNETs at all points in the follow-up period.

CONCLUSIONS

SPTs tended to occur in younger people and were more common in women. Pancreatic neuroendocrine tumors tended to form metastases more often than SPTs. The blood supply and cystic components of the lesions may have novel potential diagnostic utility for differentiating SPTs from pNETs.