P1715 - High Rates of Unnecessary Surgical Resections For Benign Pancreatic Cystic Lesions (PCLS): A Systematic Review and Meta-Analysis

Priyata Dutta, MD¹, Elit Quingalahua, MD², Fadi Hawa, MD³, Jean M. Chalhoub, MD⁴, Un Jung Lee, PhD⁵, Gabriele Capurso, ⁶, Jorge D. Machicado, MD^7
1Trinity Health Ann Arbor Hospital, Ypsilanti, MI; ²Central Michigan University College of Medicine, Ann Arbor, MI; ³University of Michigan, Ann Arbor, MI; ⁴Staten Island University Hospital, Northwell Health, Staten Island, NY; ⁵Biostatistics Unit, Office of Academic Affairs, Northwell Health, Staten Island, NY; ⁶San Raffaele Hospital, Roma, Lazio, Italy; ⁷University of Michigan Health, Ann Arbor, MI

Introduction: Surgical resection is often indicated for patients with IPMNs with high-grade dysplasia/invasive adenocarcinoma (HGD/IC) or other cysts with malignant behavior. However, pre-operative histologic diagnosis is often not feasible and surgical decisions rely on clinical factors, imaging findings, and cyst fluid analysis. Thus, PCLs are sometimes misclassified, resulting in unnecessary surgeries for benign indications (e.g. IPMN with low-grade dysplasia [LGD], serous cystadenoma [SCA]) and potential harm related to pancreatic surgical complications. In this systematic review, we aimed to estimate the global prevalence of A) incorrect pre-operative diagnoses of resected PCLs and B) surgical resections for benign PCLs (IPMNs with LGD and SCAs.

Methods: A health science librarian searched MEDLINE, EMBASE, and Cochrane Central. Observational studies or case series of ≥ 50 patients reporting surgical histopathology of resected PCLs were included. We calculated the prevalence of surgically resected PCLs with A) incorrect preoperative diagnosis; B) final histopathology of IPMN with LGD; and C) final histopathology of SCA. A proportional meta-analysis was carried out using random-effect models and using a logit transformation. Prevalence estimates are reported as proportions with corresponding 95% confidence intervals (CIs). Sensitivity analysis was conducted using the "leave-one-out" technique.

Results: Out of 3,239 screened studies, 17 (n=5,945) met the eligibility criteria. An incorrect pre-operative diagnosis of PCL type was made in 38% (95%CI 29-47%; I2 98%) of surgically resected PCLs (13 studies, n=4,132). On sensitivity analysis, the prevalence of incorrect pre-operative diagnosis ranged between 33 to 39%. Of all resected PCLs, final histopathology revealed IPMN with LGD in 27% (95% 19-37%; I2 94%) and SCA in 15% (95% CI 12-19%; I2 88%). On sensitivity analysis, this proportion ranged from 25 to 30% for IPMNs with LGD and 14 to 16% for SCAs. Among PCLs with a final diagnosis of IPMN, the proportion of LGD was 64% (95% 54-73%; I2 92%), ranging from 61 to 66% on sensitivity analysis.

Discussion: We found that an incorrect preoperative diagnosis is made in ~1/3 of patients who have surgical resection for PCLs. Pancreatic surgeries for benign IPMNs and SCAs account for ~40% of all PCL surgeries, which may represent an over-treatment with unjustified comorbidity. Further refinements in the pre-operative diagnosis of PCLs are necessary.


Disclosures:


Priyata Dutta, MD¹, Elit Quingalahua, MD², Fadi Hawa, MD³, Jean M. Chalhoub, MD⁴, Un Jung Lee, PhD⁵, Gabriele Capurso, ⁶, Jorge D. Machicado, MD⁷. P1715 - High Rates of Unnecessary Surgical Resections For Benign Pancreatic Cystic Lesions (PCLS): A Systematic Review and Meta-Analysis, ACG 2024 Annual Scientific Meeting Abstracts. Philadelphia, PA: American College of Gastroenterology.