Gastroenterology · Surgery · 1 h ago
Quantitative CT features improve IPMN risk model discrimination
A single-center retrospective study of 333 patients undergoing IPMN resection evaluated whether quantitative CT features improve prediction of high-grade dysplasia or malignancy. Adding these features to guideline-based criteria increased the model’s area under the curve from 0.760 to 0.800; external validation is still needed.
- Study included 333 resected IPMN patients at one center.
- Quantitative CT features improved model discrimination over guideline criteria.
- External validation is needed before routine clinical use.
Preoperative imaging can struggle to distinguish intraductal papillary mucinous neoplasms (IPMNs) with high-grade dysplasia (HGD) or malignancy from those with low-grade dysplasia. In this single-center retrospective study, researchers assessed three-dimensional quantitative features on preoperative pancreas-protocol CT scans from 333 patients who underwent resection between 2001 and 2024. Pathology showed low-grade dysplasia in 204 patients, HGD in 56, and malignancy in 73.
The pancreatic index decreased across the groups, from a mean of 75.6% in low-grade dysplasia to 70.5% in malignancy (p=0.003). Pancreatic Hounsfield unit standard deviation was higher in HGD and malignancy than in low-grade dysplasia. A model combining guideline-based criteria with quantitative features—including pancreatic index, pancreatic and cystic attenuation variability, and pancreatic duct volume—had greater discrimination than guideline criteria alone (area under the curve, 0.800 vs. 0.760; p=0.007).
The findings suggest quantitative CT measures may add information to preoperative IPMN risk stratification. However, the cohort included only surgically resected patients, and the study was retrospective and single-center. The authors say external validation is needed before the model’s clinical utility in routine practice can be established.
In practice
For clinicians assessing patients with IPMNs, quantitative features from pancreas-protocol CT may provide additional discrimination when combined with conventional guideline-based criteria. In this retrospective, single-center cohort of 333 patients who underwent resection, the combined model had an AUC of 0.800 versus 0.760 for guideline criteria alone. The findings do not establish that using the model improves patient outcomes or decisions, and applicability to patients not selected for surgery is uncertain. External validation is required before routine use.
AI-generated clinical interpretation · Not editor-reviewed
Is this summary clinically accurate?
Help fellow clinicians: your rating sends inaccurate summaries straight to our editors.
Sign in to rate this summary →Source
Annals of Surgical Oncology: Quantitative CT-Derived Imaging Features Improve Preoperative Risk Stratification of Intraductal Papillary Mucinous Neoplasms ↗This is an automated AI-condensed summary that has not yet been reviewed by an editor. Always consult the full item at the original source.