Liver Reporting and Data System

The general schema of LI-RADS is to attempt to classify observations as either definite HCCs (LR 5) or definitely benign (LR 1). If some, but not all, features of an HCC or benign observation are present then the categories LR 4 (probably HCC) and LR 2 (probably benign) and can be used, respectively. The middle category, LR 3, is used for indeterminate observations that do not have specific features to allow them to be characterized as more probably benign or more likely HCC. LR3 also includes observations that cannot be categorized as probably benign or probably malignant due to equivocal imaging features.

An LR 1 observation is definitely benign. The radiologist is 100% certain the observation is benign. If there is any doubt that the observation is benign or if the observation has features suggestive of HCC, it should not be categorized LR1. Due to the large number of possible benign entities, LI-RADS does not specifically delineate the features suggestive of benign entities. The LI-RADS supplement discusses benign entities commonly encountered in the cirrhotic liver. Table 1 lists examples.

An LR 2 observation is probably benign. While the radiologist is not 100% certain the observation is benign, the radiologist has a high degree of confidence that the observation is benign. If there is more than minimal doubt that the observation is benign or if the observation has features suggestive pf HCC, it should not be categorized LR2. Due to the large number of possible benign entities, LI-RADS does not specifically delineate the features suggestive of benign entities. The LI-RADS supplement discusses probable benign entities commonly encountered in the cirrhotic liver. Table 2 lists examples.

An LR 3 observation is of intermediate probability for HCC. This category is intentionally designed to include all observations that lack unequivocal features of LR 4 or 5 and also lack unequivocal benign features of LR 1 or 2. Additionally, if an observation has imaging features for LR 4 or 5 but the features and size are stable for 2 years then it becomes LR 3.

An LR 4 observation is probably but not definitely an HCC. The observation has some, but not all, of the features required for LR 5 categorization. Considerations include size, presence of arterial enhancement, washout and growth >10 mm in one year. In other words, LR 4 observations may lack arterial enhancement OR lack one or both major features for HCC, depending on their size. LR 4 observations include those with probable tumor in the lumen of a vein.

An LR 5 observation is a "definite" HCC. The radiologist is 100% certain the obervation is HCC and that, if surgical resection is performed, the observation can be reliably identified at pathology. The imaging features are specific enough to warrant liver transplantation without the need for a biopsy (assuming no contraindications to transplantation). This includes arterially hyperenhancing masses that washout and/or grow >10 mm in one year, depending on the size, as well as observations with definite tumor within lumen of vein. If there is any doubt that the observation is HCC, it should not be categorized LR5.

Definite: With 100% confidence

Diagnostic: Demonstrating features that are so characteristic of the entity that there is a 100% likelihood that the finding represents the entity.

Diameter: Maximum dimension of the observation on the imaging sequence or contrast phase where the margins are best defined. Do not include perilesional enhancement (perfusional abnormality)

Disappearance: Unequivocal absence of the observation based on resolution of all associated imaging features and the absence of any findings suggesting a space-occupying lesion

Early arterial phase: : Post-contrast injection image where the aorta and hepatic artery are brightly enhanced but the hepatic veins are not yet enhanced by antegrade blood flow. Note that with short injection times, the aorta may no longer be enhanced due to short transit time.

Growth rate: Change in size per unit of time. If imaging follow up is done at greater than 1 year interval, then apply a similar rate of growth as given in the criteria.

Growth: Interval increase in the measurement of the maximum diameter of the observation. For hyperenhancing observations, growth may refer to entire observation or the hyperenhancing component.

Heterogeneous enhancement: Hyperintensity or increased attenuation after contrast agent administration that varies across the structure (use of more specific descriptors is recommended, such as mosaic, peripheral, nodular, serpiginous, etc)

Mosaic architecture: An observation that appears to consist of nodules with differing appearances (enhancement, attenuation, intensity). This term can also be applied to lesions with internal enhancing septations.

Hyperenhancement: Imaging observation which in whole or in part unequivocally enhances more than background liver AND has unequivocal higher attenuation or intensity than background liver on contrast-enhanced images

Arterial phase hyperenhancement: Imaging observation which in whole or in part unequivocally enhances more than background liver AND has unequivocal higher attenuation or intensity than background liver during the arterial phase. Note that vascular imaging dynamics may be delayed in patients with portal hypertension. Therefore, arterial phase phenomena may be more apparent on images acquired during the nominal portal venous phase than during the nominal arterial phase in such patients

Hypoenhancement: Imaging observation which in whole or in part unequivocally enhances relative to precontrast imaging AND has an attenuation or intensity that is unequivocally less than background liver on contrast-enhanced images

Arterial phase hypoenhancement: Imaging observation which in whole or in part unequivocally enhances relative to precontrast imaging AND DOES NOT have unequivocal higher attenuation or intensity than background liver during the arterial phase. Note that vascular imaging dynamics may be delayed in patients with portal hypertension. Therefore, arterial phase phenomena may be more apparent on images acquired during the nominal portal venous phase than during the nominal arterial phase in such patients

Portal venous or later phase hypoenhancement: Imaging observation which in whole or in part unequivocally enhances relative to precontrast imaging AND has an attenuation or intensity that is unequivocally less than background liver during the portal venous or later phase. In an observation that demonstrates arterial phase hypoenhancement, the observation must have an attenuation or intensity during the portal venous or later phase that is unequivocally less than on an earlier phase. (The determination of PV or later phase hypoenhancement should be applied to the same portion of the observation for which the determination of arterial phase enhancement is made. Caution should be applied if an observation is surrounded by dense fibrosis; in such cases, apparent PV or later phase hypoenhancement may represent normal enhancement of a hypertrophic pseudomass rather than hypoenhancement of a malignant mass. ie, the surrounding tissue is hyperenhanced.)

Indeterminate: Probability of HCC cannot be defined due to insufficient information or equivocal imaging features

Intermediate: Probability of HCC is between probably benign and probably malignant

Isoenhancement: Imaging observation which in whole or in part enhances relative to precontrast imaging AND has an attenuation or intensity equal to background liver on contrast-enhanced images

Arterial phase isoenhancement: Imaging observation which in whole or in part enhances relative to precontrast imaging AND has an attenuation or intensity equal to background liver during the arterial phase

Late arterial phase: Post-contrast injection image where the portal vein is enhanced but the hepatic veins are not yet enhanced by antegrade blood flow

Masslike: Observation which in the judgment of the interpreting radiologist represents a space occupying lesion. Features suggesting a mass include: Portal venous or delayed phase ring (capsule or pseudocapsule), displacement of intraparenchymal structures, focally abnormal liver contour, visibility on unenhanced, diffusion, T2-, or T2* weighted images, discrete margins, or rounded shape.

Nonenhancement: Imaging observation that does not enhance relative to the pre-contrast imaging. The lack of any enhancement suggests that there is no viable tissue

Non-hyperenhancement: Imaging observation that enhances relative to the pre-contrast imaging AND has attenuation or intensity equal to or less than background liver

Non-masslike: Observation which in the judgment of the interpreting radiologist does not represent a space occupying lesion

Observation: Area that demonstrates imaging features that are different from the adjacent liver parenchyma. This term is preferred over lesion, since some observations may actually represent areas that are not histologically distinct from the adjacent parenchyma, such as perfusion abnormalities or artifacts.

Portal venous or later phase: Refers to images obtained in which antegrade enhancement of the hepatic veins is clearly seen.

Stable: Absence of definite change in radiological features between scans from different dates. This term may refer to size, pattern of enhancement, or other radiological characteristic.

Stable imaging features: Observation where the radiological appearance on each sequence and phase of imaging is unchanged between two different dates.

Unequivocal: Absolutely no doubt that the feature is present. If there is any concern about the presence of the feature, it should be considered "equivocal".

Criteria

• Imaging features diagnostic of a benign entity (see Table 1 for examples) OR
• Definite disappearance in absence of treatment

LI-RADS Table 1: Examples of Observations That May Be Categorized as Definitely Benign

Cyst
Hemangioma
Focal fat deposition
Focal fat sparing
Hypertrophic pseudomass interpreted as definitely benign
Wedge-shaped perfusional alterations
Confluent fibrosis
Focal scars
Some arterial-phase non-hyperenhancing atypical nodules
   Homogeneous siderotic nodules

Notes:

Table 1 purposely does not include hepatic adenoma and focal nodular hyperplasias. The reason is that adenomas and large focal nodular hyperplasias are rare in cirrhotic livers.

Do not classify as LI-RADS 1 observations that have one or more ancillary features that favor the diagnosis of HCC. Table 5 lists ancillary features that favor the diagnosis of HCC.

Perfusion alterations usually have no signal intensity abnormality on unenhanced images. If an observation is associated with signal intensity abnormalities on unenhanced images, a perfusion abnormality is unlikely.

Criteria

• Imaging features suggestive of a benign entity (see Table 2 for examples) OR
• Stable imaging features ≥ 2 years AND no increase in diameter for ≥ 2 years AND does not meet criteria for LI-RADS 1 or 4 or 5 OR
• Probable disappearance in absence of treatment OR

Note
Do not classify as LI-RADS 1 or 2 observations that have one or more ancillary features that favor the diagnosis of HCC. Table 5 lists ancillary features that favor the diagnosis of HCC

LI-RADS Table 2: Examples of Observations That May Be Categorized as Probably Benign

Atypical cyst (or probable cyst?)
Atypical hemangioma (or probable hemangioma?)
Atypical focal fat deposition (or probable focal fat hemangioma?)
Atypical focal fat sparing (or probable focal fat sparing?)
Hypertrophic pseudomass interpreted as probably benign
Rounded perfusional alterations*** (nodular arterial phase hyperenhancement, NAPH)
Patchy (changed from "florid") perfusional alterations***
Atypical confluent fibrosis (probable confluent fibrosis?)
Atypical focal scars (probable focal scars?)
Some arterial-phase non-hyperenhancing atypical nodules****
    Small (<2 cm) heterogeneous siderotic nodules
    Small (<2 cm) T1-hyperintense nodules
    Small (<2 cm) steatotic nodules
    Large (≥ 2cm) otherwise unremarkable nodules
Progressively enhancing observations

* Table 2 purposely does not include hepatic adenoma and focal nodular hyperplasias. The reason is that adenomas and large focal nodular hyperplasias are rare in cirrhotic livers.

** Do not classify as LI-RADS 2 observations that have one or more ancillary features that favor the diagnosis of HCC. Table 5 lists ancillary features that favor the diagnosis of HCC.

*** Perfusion alterations usually have no signal intensity abnormality on unenhanced images. If an observation is associated with signal intensity abnormalities on unenhanced images, a perfusion abnormality is unlikely.

**** Arterial-phase non-hyperenhancing atypical nodules may be categorized as LI-RADS 2 at the discretion of the radiologist. In general, atypical nodules categorized as LI-RADS 2 should not have major or ancillary features of HCC. Arterial-phase non-hyperenhancing atypical nodules with one or more major or ancillary features of HCC usually should be categorized as LI-RADS 3 or 4.

Criteria

• An observation that does not meet unequivocal criteria for LI-RADS 1, 2, 4, or 5 (see Table 3 for specific criteria) OR
• An observation that meets criteria for LI-RADS 4 or 5 with stable imaging features for ≥ 2 years AND no increase in diameter for ≥ 2 years

LI-RADS Table 3: Criteria for LI-RADS 3 Observations

A. <20mm

• Masslike, arterial-phase hyperenhancement, with no Major Feature AND not definitely benign AND not probably benign
• Masslike, arterial-phase hypoenhancement, with no or 1 Major Feature AND not definitely benign AND not probably benign
• Non-masslike AND not definitely benign AND not probably benign
• An observation that cannot be categorized as LI-RADS 1, 2, 4, 5 because of equivocal imaging features
• An observation that meets criteria for LI-RADS 4 or 5 with stable imaging features for ≥ 2 years AND no increase in diameter for ≥ 2 years

B. 20mm or greater

• Masslike, arterial-phase hypoenhancement, with no Major Feature AND not definitely benign or probably benign
• Non-masslike AND not definitely benign AND not probably benign
• An observation that cannot be categorized as LI-RADS 1, 2, 4, 5 because of equivocal imaging features

An observation that meets criteria for LI-RADS 4 or 5 with stable imaging features for ≥ 2 years AND no increase in diameter for ≥ 2 years

Criteria

A. <20mm

• Masslike, arterial-phase hyperenhancement, with only one Additional Major Feature (see below*) OR
• Masslike, arterial-phase isoenhancement or hypoenhancement, with two Additional Major Features (see below*) OR
• Probable tumor within lumen of vein (see Table 6)

B. 20mm or greater

• Masslike, arterial-phase hyperenhancement, with no Additional Major Feature (see below*) OR
• Masslike, arterial-phase isoenhancement or hypoenhancement, with one or two Additional Major Features (see below*) OR
• Probable tumor within lumen of vein (see Table 6)

LI-RADS Table 6: Features of Tumor Within Lumen of Vein

Definite Tumor Within Lumen of Vein

Both of the following:

• Unequivocal arterial phase hyperenhancement in lumen of vein
• Unequivical venous phase hypoenhancement in lumen of vein

Probable Tumor Within Lumen of Vein

One or more of the following:

• Occluded vein with expanded lumen
• Lumen of vein enhances but the late venous or delayed phase enhancement of lumen does not match that of patent veins
• Markedly restricted/impeded diffusion within lumen of vein
• Ill-defined hypoenhancing tubular structures in expected course and location of veins
• Arteries present within lumen of vein (do not apply this criterion if the arteries may represent cavernous transformation around a bland thrombus or occlusion)
• Probable arterial phase hyperenhancement in lumen of vein*
• Probable venous phase hypoenhancement in lumen of vein

*Do not apply this criterion to portal veins if the hyperenhancement is definitely or likely due to late timing of the arterial phase, arterio-portal shunting, or retrograde (hepatofugal) flow)

* Note: Do not apply this criterion to hepatic veins if the hyperenhancement is definitely or likely due to arterio-venous shunting or retrograde (reflux) flow into the hepatic veins from the right atrium and suprahepatic vena cava)

Criteria

A. ≥10 & <20 mm

• Masslike, arterial-phase hyperenhancement, with 2 Additional Major Features (see below*) OR
• Definite tumor within lumen of vein (see Table 6)

B. ≥ 20 mm

• Masslike, arterial-phase hyperenhancement, with 1 or 2 Additional Major Features (see below*) OR
• Definite tumor within lumen of vein (see Table 6)

Notes
Radiologists may with caution apply ancillary features that favor benignity to classify as LI-RADS 3 or 4 observations that meet major feature criteria for LI-RADS 4 or 5. Table 4 lists ancillary features that favor benignity.

Radiologists may with caution apply ancillary features that favor the diagnosis of HCC to classify as LI-RADS 4 or 5 observations that do not meet major feature criteria for the category. Table 5 lists ancillary features that favor the diagnosis of HCC.

Do not assign LI-RADS category scores to observations previously treated by locoregional ablative or embolic therapies.

Additional major features of masslike HCC

Criteria

• Portal venous phase or later phase hypoenhancement relative to liver OR
• Increase in diameter by ≥ 10 mm within one year