New strategy for Liver Imaging Reporting and Data System category M to improve diagnostic performance of MRI for hepatocellular carcinomas ≤ 3.0 cm

Purpose We aimed to determine a new strategy for Liver Imaging Reporting and Data System category M (LR-M) criteria to improve the diagnosis of HCC <= 3.0 cm on magnetic resonance imaging (MRI). Methods A total of 463 pathologically confirmed hepatic observations <= 3.0 cm (375 HCCs, 32 other malignancies, 56 benignities) in 384 patients at risk of HCC who underwent gadoxetate-enhanced MRI were retrospectively analyzed. Two radiologists evaluated the presence of major, ancillary, and LR-M features according to LI-RADS v2018. Of the ten LR-M features, those significantly associated with non-HCC malignancy were identified using multivariable logistic regression analysis, and new LR-M criteria for improving the diagnosis of HCC were investigated. Generalized estimating equations were used to compare sensitivity and specificity of LR-5 for diagnosing HCC using the new LR-M criteria with values calculated using the original LR-M criteria. p < 0.05 was considered to indicate a significant difference. Results Of ten LR-M features, rim arterial-phase hyperenhancement, delayed central enhancement, targetoid restriction, and targetoid transitional-phase/hepatobiliary-phase appearance were independently significantly associated with non-HCC malignancy (adjusted odds ratio >= 6.2; p <= 0.02). Using the new LR-M criteria (two or more of these significant features), the sensitivity of LR-5 for diagnosing HCC was higher than that with the original LR-M criteria (69% [95% confidence interval 64-73%] vs. 65% [61-70%], p = 0.002), whereas the specificity was similar (90% [82-95%] vs. 92% [83-96%], p = 0.28). Conclusion The new LR-M criteria (two or more significant features) can improve the sensitivity of LR-5 for diagnosing HCC <= 3.0 cm, without compromising specificity. [GRAPHICS] .

Automated summary

This study aimed to improve the diagnosis of HCC <= 3.0 cm by investigating new Liver Imaging Reporting and Data System category M (LR-M) criteria. The study analyzed a total of 463 hepatic observations and identified several features significantly associated with non-HCC malignancy. The new LR-M criteria showed improved sensitivity for diagnosing HCC without compromising specificity.