Liver Imaging Reporting and Data System in Patients at High Risk for Hepatocellular Carcinoma in the Memphis Veterans Affairs Population

Limitations

Limitations to our study occurred because radiologist 3 did not review all of the images nor score them. Radiologist 3 was limited to 12 images where there was disagreement and was limited to 2 scores to choose from for each image. Further limitations include that this study was performed at a single center. Our study focused on one imaging modality and did not include ultrasounds or MRIs. We did not compare the demographics of our patients with those of other VA hospitals. The radiologists interpreted the images individually, and their subjectivity was another limitation.

Conclusion

In the MVAMC population, LI-RADS showed a good sensitivity, specificity, and diagnostic accuracy for CT surveillance scans in patient at high risk for HCC at an earlier time point than did standard evaluation by very experienced CT radiologists. Higher LI-RADS scores on surveillance CTs had good diagnostic accuracy for the probable future diagnosis of HCC, whereas lower LI-RADS scores had a good diagnostic accuracy for probable benign nodules. Utilizing the LI-RADS algorithm on all surveillance CTs in patients at high risk for HCC may lead to obtaining MRIs or follow-up CTs sooner for suspicious nodules, leading to an earlier diagnosis of HCC and possible earlier and more effective intervention.