A Systematic Comparative Study on the Diagnostic Value of Transabdominal Ultrasound in Patients with Pancreatic Cystic Lesions

Abstract

Pancreatic cystic lesions are a frequent incidental finding in abdominal imaging. Despite its usually benign background, a small fraction exhibiting features suspicious for cancerous development demands continuous follow-up or surgical removal. Current guidelines advocate magnetic resonance imaging and endoscopic ultrasound to evaluate the risk of malignancy, whereas transabdominal ultrasound is perceived as subordinate imaging. The objective of this study was to analyze cyst detection rates of latest-generation ultrasound machines compared to magnetic resonance imaging, computed tomography, and endosonographic ultrasound and to determine inter-rater reliability. The results showed that large cysts facilitate their visualization by transabdominal ultrasound while detection rates are independent of the anatomical part of the pancreas in which they were sited. Changes in the pancreatic duct width, a connection to the pancreatic duct system, and the architectural characteristics of cysts are poorly recognized by transabdominal ultrasound compared to magnetic resonance imaging and endoscopic ultrasound. Computed tomography imaging is preferred over transabdominal ultrasound to detect calcifications and regional lymphadenopathy. Even if conducted by experienced investigators, transabdominal ultrasound examinations fail to agree with magnetic resonance imaging scans regarding cyst detection rates (κ = 0.093).

Keywords: pancreatic cyst; pancreatic intraductal neoplasms; ultrasound imaging.

Figure 1

Diagnostic accordance between TAUS and final discharge summaries. (A) Frequency of different PCL entities as reported by TAUS and discharge summaries (in absolute numbers). (B) Frequency of different PCL entities as reported by TAUS (in percentage numbers). (C) Degree of overall accordance between TAUS and final discharge summaries. (D) TAUS-related causes of discrepant findings. Transabdominal ultrasound, TAUS; intraductal papillary mucinous neoplasm, IPMN; main duct type-IPMN, md-IPMN; branch duct type-IPMN, bd-IPMN; mucinous cystic neoplasm, MCN; post-pancreatitis fluid collection, PFC; pancreatic cystic lesion, PCL.

Figure 2

Diagnostic accordance between TAUS and MRI scans. (A) Patients with dual imaging combining TAUS with MRI or other type of imaging. (B) Patients (in absolute numbers) with MRI and TAUS assigned to the categories “consistent findings”, “divergent findings”, “PCL not further classified”, and “Pancreas not examinable by TAUS”. (C) Distribution of consistent findings. (D) Inconsistent findings between TAUS and MRI. (E) Venn diagram: non-classifiable PCL on TAUS and MRI. (F) Diagnostic categorization of PCLs evaluated by MRI (in absolute numbers). (G) Juxtaposition of MRI-confirmed bd-IPMN with TAUS-based assessment (in absolute numbers). Transabdominal ultrasound, TAUS; magnetic resonance imaging, MRI; intraductal papillary mucinous neoplasm, IPMN; main duct type-IPMN, md-IPMN; branch duct type-IPMN, bd-IPMN; mucinous cystic neoplasm, MCN; post-pancreatitis fluid collection, PFC; pancreatic cystic lesion, PCL.

Figure 3

Diagnostic accordance between TAUS and EUS. (A) Patients with dual imaging combining TAUS with EUS or other type of imaging; (B) Patients (in absolute numbers) with EUS and TAUS assigned to the categories “consistent findings”, “divergent findings”, “PCL not further classified”, and “Pancreas not examinable by TAUS”. (C) Distribution of consistent findings. (D) Inconsistent findings between TAUS and EUS. (E) Venn diagram: non-classifiable PCLs on TAUS and EUS. (F) Diagnostic categorization of PCLs evaluated by EUS (in absolute numbers). Transabdominal ultrasound, TAUS; endoscopic ultrasound, EUS; intraductal papillary mucinous neoplasm, IPMN; main duct type-IPMN, md-IPMN; branch duct type-IPMN, bd-IPMN; mucinous cystic neoplasm, MCN; serous cystic neoplasm, SCN; post-pancreatitis fluid collection, PFC; pancreatic cystic lesion, PCL.

Figure 4

Diagnostic accordance between TAUS and CT scans. (A) Patients with dual imaging combining TAUS with CT or other type of imaging. (B) Patients (in absolute numbers) with CT and TAUS assigned to the categories “consistent findings”, “divergent findings”, “PCL not further classified”, and “Pancreas not examinable by TAUS”. (C) Distribution of consistent findings. (D) Inconsistent findings between TAUS and CT. (E) Venn diagram: non-classifiable PCLs on TAUS and CT. (F) Diagnostic categorization of PCL evaluated by CT (in absolute numbers). Transabdominal ultrasound, TAUS; computed tomography, CT; intraductal papillary mucinous neoplasm, IPMN; main duct type-IPMN, md-IPMN; branch duct type-IPMN, bd-IPMN; mucinous cystic neoplasm, MCN; serous cystic neoplasm, SCN; post-pancreatitis fluid collection, PFC; pancreatic cystic lesion, PCL.

Figure 5

(A) Quantification of cysts per patient and imaging modality. (B) Comparison of PCL size determined by MRI with TAUS measurements; (C) Intrapancreatic localization of PCLs arranged by anatomical segment. Pancreatic cystic lesion, PCL; transabdominal ultrasound, TAUS; endoscopic ultrasound, EUS; magnetic resonance imaging, MRI; computed tomography, CT.

Figure 6

Discriminatory features to be used for detection of high-risk PCL. (A) TAUS examinations with at least one proven PCL per patient. (B) Architecture of PCL. (C) Connection of PCL to the pancreatic duct system. (D) Dilation of the pancreatic main duct. (E) Change in caliber width of the pancreatic main duct. (F) Dilation of the common bile duct. (G) Parenchymal and intracystic calcifications. (H) Irregular internal structuring of PCL. (I) Presence of regional lymphadenopathy. (J) Compression of neighboring organs by PCL. Pancreatic cystic lesion, PCL; transabdominal ultrasound, TAUS; endoscopic ultrasound, EUS; magnetic resonance imaging, MRI; computed tomography, CT.