Purpose

To describe the five-year, single-center, clinical experience of abbreviated MRI (AMRI) protocol for HCC surveillance.

Material and methods

The radiology reports for all AMRIs performed through May 2019, at a single referral center, were retrospectively reviewed. Each AMRI examination was scored as negative (no or only definitely benign observations), subthreshold (observations <10 mm) or positive (observations ≥ 10 mm, not definitely benign). Each patient was categorized as surveillance-negative, surveillance-subthreshold or surveillance-positive according to their highest AMRI score they had during the inclusion period. Blinded to those categories, each patient was classified as positive or negative for HCC using a composite reference standard that integrated all available clinical, imaging and pathology data.

Results

1160 AMRIs were performed in 561 patients. 185 (33%) patients were lost to follow-up. Of the remaining 376 patients, 295 (79%) were surveillance-negative, 23 (6%) surveillance-subthreshold, and 58 (15%) surveillance-positive. Based on the composite reference standard, 292/295 (99%) surveillance-negative and 20/23 (87%) surveillance-subthreshold patients were HCC-negative (true negatives); 29/58 (50%) of surveillance-positive patients were HCC-positive (true positives). Sensitivity, specificity, and accuracy of surveillance AMRI were 86%, 91% and 90%, respectively.

Conclusion

Hepatobiliary-phase AMRI surveillance is feasible clinically in patients in whom US surveillance is compromised. Our preliminary single-center experience suggests it provides higher sensitivity (86% vs 78%) and specificity (98% vs 89%) [4] with a lower loss to follow-up (33% vs 38%) [5] than historically reported US surveillance.

Learning objectives

Purpose

To determine whether elevated liver iron in two populations of American Indians exceeds what is expected from genetics.

Material and methods

Liver MRIs were evaluated for 300 consecutively-enrolled American Indians from two study centers in different parts of the United States as part of the ongoing SHS, stratified for obesity, diabetes, and alcohol consumption (87M, 208F, 5 unknown; mean age 50.7 yrs). For each participant, C282Y and H63D mutations and alcohol consumption history were collected, and PDFF and R2* values were derived from MRIs. Percentages of participants with 3T-equivalent R2* values in the 80-100 s^-1 range and >100 s^-1 were tabulated and compared to expected percentages based on (C282Y/H63D) genetics.

Results

3.3% (10/300) of participants had R2* > 100 s^-1 (range 101 to 818 s^-1; 3 from Site 1, 7 from Site 2), and 10.0% (30/300) of participants had R2* in the 80-100 s^-1 range (29 from Site 1, 1 from Site 2). None of the first 160 participants were homozygous for either mutation. 64.3% (193/300) of participants had PDFF > 6%, a commonly accepted cutoff for fatty liver. 35.9% (99/276) of responding participants reported heavy episodic drinking.

Conclusion

The observed elevated prevalence of inferred liver iron in American Indians at two study centers appears not to be explainable by genetics alone and, the higher prevalence at Site 1 is unexplained. Given the high observed prevalences of fatty liver and heavy episodic drinking, these findings may be related to coincident nonalcoholic and/or alcoholic fatty liver disease, and to population differences between the two sites.

Purpose

To evaluate the agreement of size measurements of focal liver observations on MRI and its impact on Liver Imaging Reporting And Data System (LI-RADS) v2018 diagnostic categories.

Material and methods

Between September 2017 and May 2018, 52 patients with 95 focal liver observations underwent multiarterial-multiphase MRI at 3T. Observations size was assessed by three radiologists, independently, at each imaging phase. In consensus, observations were also scored for the presence of LI-RADS major features. Consensus-feature scores and individual phase measurements were used to assign observations’ LI-RADS category. Inter-reader and intra-reader agreement on size at each imaging phase were evaluated using intraclass correlation coefficient (ICC). Bootstrap-based confidence intervals were estimated. Sizes for each reader in each imaging phase were compared using Spearman’s correlation. To assess of the effect of individual variables on size a multivariable mixed-effect linear regression was used.

Results

Intra-reader agreement was excellent, with the higher variation for reader 3 during the arterial phases (ICC: 0.936 - 95% CI 0.64 – 0.98). Interreader agreement was excellent with ICC varying from 0.938 (95% CI 0.870 – 0.976) to 0.970 (95% CI 0.935 – 0.989) depending on the imaging phase. Changes in LI-RADS categories in relation to size or reader occurred in 43% of the observations. Size variability were related to: i) imaging delay (acquisition time post-contrast injection) (p<0.01), ii) presence of APHE (p=0.01), iii) presence of capsule appearance (p=0.02), and iv) reader (p<0.01).

Conclusion

While intra- and inter-observer agreement rates for size measurement are “excellent” for radiologists, measurements variability across different phases can impact LI-RADS categorization.

Purpose

To assess the interreader agreement for Liver Imaging Reporting and Data System v2018 features using multi-arterial phase and multi-delayed MR images in patients with cirrhosis.

Material and methods

We retrospectively reviewed 52 consecutive cirrhotic patients with a total of 95 LI-RADS 3, 4, 5, or M observations who underwent multi-arterial and multi-delayed phase MRI at 3T. These included 3D fat-suppressed T1-weighted spoiled gradient echo images acquired at 9 time points (6 arterial phases, 3 delayed phases) after injecting 0.1 mmol/kg gadobutrol. The arterial time points were obtained every 4-5 seconds with a high-temporal-resolution view-sharing MRI technique. Three radiologists independently scored the following LI-RADS v2018 imaging features: arterial phase hyperenhancement (APHE), APHE subtype, and corona enhancement on the multi-arterial phase; "washout", “washout” type, and enhancing "capsule" on multi-delayed. Interreader agreement was assessed with intraclass coefficient.

Results

Interobserver agreement differed by each time delay. On multi-arterial phase images, interreader agreement was substantial to excellent for APHE (range, 0.717-0.849), substantial for APHE subtype (range, 0.675- 0.801), and fair to moderate for corona enhancement (range, 0.209-0.490). On delayed phase images, interreader agreement ranged from moderate to substantial (range, 0.552-0.685) for “washout” and fair to moderate (range, 0.383-0.501) for enhancing “capsule”.

Conclusion

Overall iInterreader agreement was excellent for APHE and substantial for type of APHEgood for APHE and type of APHE using mHAP MRI. Multiarterial phase may improve consistency at image features for LI-RADS. Lower agreements were observed for corona enhancement and delayed imaging features.However, further refinement of corona enhancement and delayed imaging features maybe needed.

Learning objectives

Purpose

To describe the five-year, single-center, clinical experience of abbreviated MRI (AMRI) protocol for HCC surveillance.

Material and methods

The radiology reports for all AMRIs performed through May 2019, at a single referral center, were retrospectively reviewed. Each AMRI examination was scored as negative (no or only definitely benign observations), subthreshold (observations <10 mm) or positive (observations ≥ 10 mm, not definitely benign). Each patient was categorized as surveillance-negative, surveillance-subthreshold or surveillance-positive according to their highest AMRI score they had during the inclusion period. Blinded to those categories, each patient was classified as positive or negative for HCC using a composite reference standard that integrated all available clinical, imaging and pathology data.

Results

1160 AMRIs were performed in 561 patients. 185 (33%) patients were lost to follow-up. Of the remaining 376 patients, 295 (79%) were surveillance-negative, 23 (6%) surveillance-subthreshold, and 58 (15%) surveillance-positive. Based on the composite reference standard, 292/295 (99%) surveillance-negative and 20/23 (87%) surveillance-subthreshold patients were HCC-negative (true negatives); 29/58 (50%) of surveillance-positive patients were HCC-positive (true positives). Sensitivity, specificity, and accuracy of surveillance AMRI were 86%, 91% and 90%, respectively.

Conclusion

Hepatobiliary-phase AMRI surveillance is feasible clinically in patients in whom US surveillance is compromised. Our preliminary single-center experience suggests it provides higher sensitivity (86% vs 78%) and specificity (98% vs 89%) [4] with a lower loss to follow-up (33% vs 38%) [5] than historically reported US surveillance.