LI-RADS (Liver Imaging Reporting and Data System)

What is measured: the probability that a liver observation in a patient at risk (cirrhosis, chronic hepatitis B) is hepatocellular carcinoma. How: multiphase contrast CT or MRI read for the major features (size, non-rim arterial phase hyperenhancement, non-peripheral washout, an enhancing capsule and threshold growth) and ancillary features, giving LR-1 to LR-5, plus LR-M for a malignancy that may not be HCC (cholangiocarcinoma, for example) and LR-TIV for tumour in a vein; LR-5 is about 95 percent specific but only 60 to 70 percent sensitive, so many small cancers sit in LR-3 or LR-4. Contrast ultrasound has its own CEUS LI-RADS, and treatment-response LI-RADS (LR-TR viable or non-viable) is used after ablation or embolisation. What a result changes: LR-5 establishes the diagnosis without a biopsy, feeds BCLC staging and treatment (resection, ablation, transplant listing under OPTN rules, which use LI-RADS), and starts alpha-fetoprotein and imaging follow-up; LR-3 and LR-4 lead to short-interval repeat imaging or biopsy; LR-M leads to biopsy because the treatment differs. Where it matters: early, intermediate and advanced HCC.