Preoperative diagnostic failure in gallbladder cancer: Influence of tumor location and size on imaging precision

Preoperative imaging diagnosis of early-stage gallbladder cancer (GBC) remains challenging. The effectiveness of different imaging modalities and clinical factors to diagnose GBC have not been fully investigated. We identified risk factors for preoperative diagnostic failure of GBC, including tumor location (hepatic vs. peritoneal) and size relative to different imaging approaches.

Patients undergoing curative-intent GBC resection (2000-2023) were identified from an international, multi-institutional database. The primary outcome was successful preoperative GBC diagnosis based solely on imaging without biopsy. Multivariable logistic regression identified risk factors associated with diagnostic failure, and the impact of different imaging modalities was assessed.

Among 293 patients, preoperative GBC diagnosis was successfully made in 164 (56.0 %) patients. Hepatic-sided tumors were less common among undiagnosed versus diagnosed patients (18.6 % vs. 44.5 %; p = 0.033). On multivariable analysis, hepatic-sided location (OR:0.13 [0.02-0.76]; p = 0.025, ref:peritoneal-sided) and tumor size ≥2.0 cm (OR:0.11 [0.01-0.88]; p = 0.035) were associated with lower odds of preoperative imaging diagnostic failure. Among tumors <2.0 cm, peritoneal-sided lesions had a higher risk of diagnostic failure than hepatic-sided, with the risk gap widening as size decreased. MRI/MRCP (39.5 % vs. 65.2 %) and EUS (5.4 % vs. 16.5 %) were used less often among undiagnosed patients compared to diagnosed ones (both p < 0.001), while CT use was similar (84.5 % vs. 85.4 %; p = 0.993). The failure of preoperative imaging diagnosis decreased as more imaging modalities were combined compared with CT alone: 65.1 % for CT only versus 17.4 % for CT plus MRI/MRCP or EUS.

Peritoneal-sided tumors and lesions <2.0 cm were associated with higher preoperative diagnostic failure risk among patients with GBC, especially when a single imaging modality was utilized. Combining different imaging modalities may improve preoperative diagnosis among this subset of patients.