Concordance Between Preoperative Imaging and Surgical Specimen Size in the Evaluation of Intraductal Papillary Mucinous Neoplasms of the Pancreas
*Juan P. Cobar MD1, *Amir Ebadinejad 1, *El-Sayed Ibrahim MD2, *Julia Silverman BS3, *Tony El-Jabbour MD2, Lindsay Bliss MD1, David Curtis MD1, Bret Schipper MD1, *Saverio Ligato MD2, Oscar K. Serrano 1
1Department of Surgery, Hartford Hospital, Hartford, CT; 2Department of Pathology, Hartford Hospital, Hartford, CT; 3School of Medicine, University of Connecticut, Farmington, CT
Background Pancreatic intraductal papillary mucinous neoplasms (IPMN) are considered pre-malignant cystic tumors with a wide array of malignant potential. Consensus guidelines on operative intervention are profoundly influenced by tumor size and imaging features; however, the optimal imagining modality is still undefined. While previous research has focused on evaluating isolated agreement between imaging studies in the workup of IPMN, none have reviewed the concordance in tumor size between preoperative radiological assessments and gross surgical pathology examination. Study Design Patients who had surgical resection of an IPMN between 2014 and 2023 were identified. Pre-operative IPMN measurements on magnetic resonance imaging (MRI) and endoscopic ultrasound (EUS) were analyzed and compared to surgical pathology measurements. Concordance analysis was performed using measures of central tendency for the population. Results We identified 57 patients who had an IPMN resection of whom 25 (44%) had a preoperative MRI and 49 (86%) had a preoperative EUS. The average IPMN size in our cohort was 2.2 + 1.4 cm for MRI and 2.4 + 1.7 cm for EUS. The average IPMN size on pathologic examination was 3.5 + 2.5 cm. While both modalities underestimated the size of the IPMN, EUS showed significant differences between imaging and gross pathology measurements (-1.0 ± 2.7 cm, p=0.010). When comparing over- and underestimate variances within each modality as a function of pathologic size, all were statistically significant (p?0.040), with the variance for EUS having the most profound difference (p?0.002). Conclusion In the surgical evaluation of IPMN patients, MRI and EUS have poor concordance with actual tumor size. These results support the recommendation of utilizing MRI as a follow-up modality in the longitudinal evaluation of IPMN patients. Given that imaging findings are some of the primary characteristics for surgical decision-making, these findings suggest caution when using size with these modalities solely as a determinant of surgical resection.
Back to 2024 Abstracts