EUS-guided fine needle aspiration-based clues to mistaken or uncertain identity: serous pancreatic cysts.
High specificity cyst fluid biomarkers are the most effective method for diagnosing serous cystic neoplasms preoperatively, suggesting a multidisciplinary approach for management.
Where it sits
this study against the rest of the pancragen (kedw) corpusSummary and findings
The study evaluated the effectiveness of imaging and EUS-FNA-based analyses in diagnosing serous cystic neoplasms (SCN) in pancreatic cysts. In a surgical cohort of 62 patients, CT/MRI and EUS predicted SCN in 7% and 31% of cases, respectively. High specificity cyst fluid biomarkers were the most effective, correctly identifying SCN in 93% of cases.
Abstract
<h4>Background/objectives</h4>Pancreatic serous cystic neoplasms (SCN) present a diagnostic challenge given their increasing frequency of detection and benign nature yet relatively high rate of misdiagnosis. Here, imaging and analyses associated with EUS-guided fine-needle aspiration (EUS-FNA) are evaluated for their ability to provide a correct preoperative diagnosis of SCN.<h4>Methods</h4>A surgical cohort with confirmed pathological diagnosis of SCN (n = 62) and a surveillance cohort with likely SCN (n = 31) were assessed for imaging (CT/MRI/EUS) and EUS-FNA-based analyses (cytology/DNA analysis for Von Hippel-Lindau [VHL] gene alterations/biomarkers).<h4>Results</h4>In the surgical cohort, CT/MRI and EUS respectively predicted SCN in 4 of 58(7%) and 19 of 62(31%). Cyst fluid cytology and VHL alterations predicted SCN in 1 of 51(2%) and 5 of 21(24%), respectively. High specificity cyst fluid biomarkers (vascular endothelial growth factor [VEGF]/glucose/carcinoembryonic antigen [CEA]/amylase) correctly identified SCN in 25 of 27(93%). In the surveillance cohort, cyst fluid biomarkers predicted SCN in 12 of 12(100%) while VHL alterations identified SCN 3 of 10(30%).<h4>Conclusion</h4>High specificity cyst fluid biomarkers provided the most sensitive means of diagnosing SCN preoperatively. To obtain a preoperative diagnosis of SCN at the highest level of certainty, a multidisciplinary approach should be taken to inform appropriate SCN management.
Background
Pancreatic serous cystic neoplasms (SCN) are increasingly detected but often misdiagnosed due to their benign nature. Accurate preoperative diagnosis is crucial for appropriate management. This study investigates the utility of imaging and EUS-FNA-based analyses in improving diagnostic accuracy for SCN.
Methods
The study involved a surgical cohort with confirmed SCN (n=62) and a surveillance cohort with likely SCN (n=31). Imaging techniques (CT/MRI/EUS) and EUS-FNA-based analyses, including cytology and DNA analysis for VHL gene alterations, were assessed. High specificity cyst fluid biomarkers were also evaluated.
Results
In the surgical cohort, CT/MRI predicted SCN in 7% of cases, while EUS predicted SCN in 31% of cases. Cyst fluid cytology and VHL alterations predicted SCN in 2% and 24% of cases, respectively. High specificity cyst fluid biomarkers were the most effective, correctly identifying SCN in 93% of cases. In the surveillance cohort, cyst fluid biomarkers predicted SCN in 100% of cases, while VHL alterations identified SCN in 30% of cases.
Interpretation
The study suggests that high specificity cyst fluid biomarkers are the most reliable method for preoperative SCN diagnosis, outperforming imaging and other EUS-FNA-based analyses. However, the small sample sizes for certain diagnostic methods and the focus on a surgical cohort may limit the generalizability of the findings. A multidisciplinary approach is recommended to enhance diagnostic certainty.
Key findings
- CT/MRI predicted SCN in 4 of 58 (7%) cases.
- EUS predicted SCN in 19 of 62 (31%) cases.
- Cyst fluid cytology predicted SCN in 1 of 51 (2%) cases.
- VHL alterations predicted SCN in 5 of 21 (24%) cases.
- High specificity cyst fluid biomarkers identified SCN in 25 of 27 (93%) cases.
Limitations
- Small sample size for VHL alterations.
- Reliance on a surgical cohort.
- Limited generalizability due to cohort characteristics.