Introduction
Liver Imaging Reporting and Data System (LI-RADS) 3 is an indeterminate category with an approximately 30-40% malignancy risk on contrast-enhanced MRI.1 LR-3 lesions require monitoring at 3–6-month intervals, which may be time-consuming and costly.2 We aim to determine the outcomes of LR-3 lesions in cirrhotic patients and identify predictors of malignant and benign outcomes.
Methods
A retrospective chart review was conducted on patients with LR-3 lesions identified between 2019-2021. Variables collected included imaging features, radiologist experience, lesion outcome, and coexistence of other lesions. Patient variables included age, sex, cause of cirrhosis, and liver treatment status (treatment-naïve versus previously treated).
Results
Among 795 LR-3 observations in 427 patients (41% with prior treatment), 86% of lesions demonstrated arterial phase hyperenhancement, and 17% showed ancillary features. The highest grade coexisting lesion was LR-5 in 26%, LR-4 in 19%, and LR-3 in 35%. Over four years, 26% of LR-3 lesions upgraded, 58% downgraded, and 16% remained stable. No significant difference in outcome was observed based on radiologist experience or major LI-RADS criteria. LR-3 lesions in treatment-naïve livers had lower upgrade rates (17% vs 38%, p<0.001; HR 0.5, p<0.001) and higher downgrade rates (69% vs 44%, p<0.001; HR 1.78, p<0.001) compared to post-treatment livers. Among LR3 lesions coexisting with other LR3 lesions, 57% downgraded (p < 0.001; HR 0.76, p=0.026) and 24% upgraded (p < 0.001; HR 1.62, p=0.032). In contrast, when coexisting with LR4/5 lesions, 45% downgraded (p < 0.001; HR 0.66, p=0.004) and 37% upgraded (p < 0.001; HR 1.74, p=0.014). Isolated LR-3 lesions downgraded in a substantial 73% of cases (p < 0.001) and only 16% upgraded (p < 0.001). Alcohol use was the only cause of cirrhosis significantly associated with upgrade (HR 1.61, 95% CI, p = 0.01). Lesions with ancillary features were approximately twice as likely to upgrade and 34% less likely to downgrade (HR 2.07, 95% CI, p = 0.004; HR 0.64, 95% CI, p = 0.03).
Conclusions
Liver treatment status, coexistence of other LIRADS lesions, alcoholic cirrhosis, and ancillary features are associated with malignant outcomes. Delineating LR-3 outcome predictors may guide surveillance and reduce resource burden.
References
Lee, Sunyoung et al. “Percentages of Hepatocellular Carcinoma in LI-RADS Categories with CT and MRI: A Systematic Review and Meta-Analysis.” Radiology vol. 307,1 (2023): e220646. doi:10.1148/radiol.220646
Chernyak, Victoria et al. “Liver Imaging Reporting and Data System (LI-RADS) Version 2018: Imaging of Hepatocellular Carcinoma in At-Risk Patients.” Radiology vol. 289,3 (2018): 816-830. doi:10.1148/radiol.2018181494