LI-RADS shows low NPV in per-patient liver tumor assessment

Per-patient grouping for liver tumor status based on LI-RADS and treatment response shows low negative predictive value (NPV) in detecting residual or untreated tumors, according to research published December 19 in Radiology.  A team led by Omar Hassan, MD, from the University of California, San Francisco studied hepatocellular carcinoma (HCC) patients who underwent local-regional therapy followed by liver transplant. The team found that per-patient stratification of tumor status at either pretransplant CT or MRI based on the LI-RADS version 2018 treatment response algorithm showed low NPV in this area. “The relatively low negative predictive value may contribute to clinical under-staging, delaying care for patients who need to undergo transplants and potentially allocating transplants to patients out of the Milan criteria,” the Hassan team wrote. The LI-RADS treatment response algorithm grades the diagnosis of HCC and recurrence after local-regional therapy. The system has high specificity, but low sensitivity. However, the researchers noted that the emphasis on specificity can lead to disease under-staging. This could result in poorer post-transplant outcomes. Hassan and colleagues studied the NPVs of pre-transplant CT and MRI assessment for viable HCC on a per-patient basis. They used the LI-RADS treatment response algorithm and used explant pathology as the reference standard. The team included patient records from 206 HCC patients who underwent a liver transpla...
Source: AuntMinnie.com Headlines - Category: Radiology Authors: Tags: MRI Abdominal Imaging Source Type: news