Hepatocellular carcinoma (HCC) remains a leading cause of cancer-related mortality worldwide, and accurate imaging surveillance is critical for the early diagnosis and detection of recurrence after curative treatment. A recent study by Kim et al. in Clinical and Molecular Hepatology compared non-contrast magnetic resonance imaging (NC-MRI) with contrast-enhanced computed tomography (CECT) for diagnosing recurrent HCC after two years, according to the KLCA-NCC practice guidelines [
1]. Given the disadvantages of CECT, including ionizing radiation exposure and contrast-related adverse events, these findings carry important clinical implications. However, the prior study was limited by a small number of recurrence-positive cases and a mean tumor size of only 1.3 cm (1.0–1.7 cm). Considering the similar epidemiologic background of HCC in Korea and China, we sought to validate these results in a larger patient cohort who underwent both NC-MRI and CECT, and to explore whether the conclusions could be extended to broader clinical settings.
We therefore conducted a paired comparative analysis of NC-MRI and CECT in patients with histopathologically confirmed HCC to directly assess their preoperative sensitivity. Patients who underwent surgical resection at the Affiliated Drum Tower Hospital of Nanjing University Medical School between January 2020 and June 2022 were retrospectively reviewed. After applying inclusion and exclusion criteria, 72 patients with both NC-MRI and CECT were eligible (
Fig. 1A). Two experienced abdominal radiologists (12 and 15 years of practice) independently reviewed all images while blinded to clinical information and other imaging results, with discrepancies resolved by consensus. Lesions were classified as positive or negative for HCC according to KLCANCC imaging criteria [
2,
3].. During the arterial phase, the gross classification of the tumor was determined. Round or quasi- round tumors were considered Type I, lobed tumors were considered Type II, and irregular tumors were considered Type III (
Supplementary Fig. 1) [
4].
Among the 72 patients with pathologically confirmed HCC, NC-MRI (T2W+DWI) missed only 2 lesions, whereas CECT (arterial-phase hyperenhancement with washout) missed 14, yielding sensitivities of 97.2% (70/72) and 80.6% (58/72), respectively (
Fig. 1B). The median tumor diameter was 3.10 cm (IQR, 2.25–4.80 cm,
Supplementary Table 1). Given that not all HCCs display typical washout, we further stratified lesions by washout status (
Supplementary Table 1) to assess its impact on diagnostic sensitivity. On CECT, tumors with washout were more frequently categorized as type I morphology compared with those without washout (
Fig. 1C,
P<0.05). None of the included patients had gross vascular invasion. Among 14 patients with pathologically confirmed microvascular invasion (MVI), CECT detected 11 cases, which was not significantly different from the detection rate in MVI-negative patients (44/58,
P>0.05). For NC-MRI, both false-negative cases occurred in MVI-negative patients, and the sensitivity did not differ significantly between MVI-positive and MVI-negative groups (
Supplementary Table 2). Representative cases of each lesion type are illustrated in
Supplementary Figure 1. Overall, NC-MRI achieved higher sensitivity than CECT for preoperative HCC detection, underscoring its potential diagnostic advantage.
In this head-to-head paired comparison, NC-MRI demonstrated a significantly higher detection rate for HCC compared with CECT. These findings align with the work of Kim et al., but also extend it by providing direct paired data before surgery.1 In Kim’s study, the sensitivity of CECT for HCC was reported as 8/23 per lesion and 8/22 per patient. Upon our re-evaluation, the correct values should be 8/22 per lesion and 8/21 per patient. It should be noted that one patient had only para-aortic lymph node metastasis without any intrahepatic lesion. According to the KLCA-NCC practice guidelines, the “wash-in and wash-out” criteria apply specifically to intrahepatic lesions, reflecting the dual blood supply of the liver and the predominantly arterial supply of HCC. Therefore, including this patient in the analysis of intrahepatic lesions was inappropriate.
In this study, we further validated and expanded on Kim’s findings. The sensitivity of NC-MRI was 97%, higher than the 77% reported previously, which may be attributable to the larger tumor size in our cohort. Previous studies have shown that CECT performs better in detecting larger tumors compared with smaller ones [
5]. Although NC-MRI demonstrated markedly higher sensitivity for HCC detection than CECT, its clinical utility still requires validation in studies with higher-level evidence.
The higher performance of NC-MRI may be explained by its superior lesion-to-liver contrast and sensitivity to fat and iron content, which can enhance visualization of small or atypical HCC lesions even without contrast administration [
6,
7]. Park et al. further demonstrated that the diagnostic sensitivity of Gd-BOPTA-enhanced MRI combined with T2-weighted imaging (T2WI), diffusion-weighted imaging (DWI), and dual-echo sequences was comparable to that of gadoxetic acid-enhanced MRI using four combined sequences [
5]. Taken together, these findings suggest that noncontrast sequences provide robust diagnostic information and may substitute for contrast-enhanced protocols in selected patients.
From a clinical perspective, the advantages of NC-MRI extend beyond diagnostic sensitivity. Patients with HCC often require repeated surveillance imaging after curative treatment, and avoiding cumulative exposure to radiation and contrast-related adverse events is a meaningful advantage. NC-MRI, therefore, represents a valuable alternative for patients with contraindications to iodinated contrast or in healthcare settings where repeated intravenous contrast use is impractical due to cost, availability, or patient preference. Although MRI is generally more expensive and less available than CT, NC-MRI could be cost-effective if it reduces the number of contrast-enhanced examinations during long-term follow-up.
It should be noted, however, that improved sensitivity must be balanced against potential trade-offs in specificity. Hyperintense nodules on T2WI or DWI may represent benign lesions such as regenerative or dysplastic nodules, leading to possible false-positive interpretations [
8,
9]. Therefore, while NC-MRI shows promise as a first-line tool for detection, complementary imaging or histological confirmation may still be necessary in indeterminate cases.
Our study has several limitations. The retrospective design and single-center setting may limit generalizability, and the sample size was modest. Furthermore, we did not assess inter-modality differences in lesion characterization or staging, focusing solely on detection. Future studies should assess NC-MRI not only for diagnostic accuracy but also for cost-effectiveness and workflow. In addition, it would be valuable to explore whether using NC-MRI for initial detection followed by CECT for lesion characterization could improve diagnostic practice. Furthermore, integration of NCMRI with radiomics or serum biomarkers such as AFP or ctDNA may further enhance diagnostic performance, particularly for sub-centimeter lesions that remain challenging for both MRI and CT. Lastly, the diagnostic performance of NC-MRI was not influenced by MVI status in our cohort. However, as no patients with gross vascular invasion were included, the utility of NC-MRI for infiltrative tumors with major vascular involvement remains to be investigated in future studies.
In conclusion, our paired analysis shows that NC-MRI achieves higher sensitivity than CECT for detecting HCC, both in preoperative diagnosis and postoperative surveillance. These findings highlight NC-MRI as a viable diagnostic option in selected patients and extend existing evidence through a direct, head-to-head comparison. Prospective multicenter studies are needed to further define its role in future diagnostic algorithms.
FOOTNOTES
-
Authors’ contribution
All authors contributed to this study at different levels. All authors read and approved the final version.
Study concept and design (Laizhu Zhang, Binghua Li, Decai Yu); acquisition of data (Laizhu Zhang, Weiwei Zong, Jialin Gao, Huan Li, Leizhou Xia); statistical analysis and interpretation of data (Laizhu Zhang, Binghua Li); Radiology evaluation of the patients (Xiaoli Mai, Jun Chen); drafting of the manuscript (Laizhu Zhang, Binghua Li, Decai Yu).
-
Acknowledgements
This work was supported by grants from the National Natural Science Foundation of China (No. 82372834, 82173129, and 82373911), Jiangsu Outstanding Youth Foundation (BK20240119), Young Elite Scientists Sponsorship Program of Jiangsu Association for Science and Technology granted by Jiangsu Anti-Cancer Association and the Fundings for Clinical Trials from the Affiliated Drum Tower Hospital, Medical School of Nanjing University (2023-LCYJPY- 32).
-
Conflicts of Interest
The authors have no conflicts to disclose.
SUPPLEMENTARY MATERIAL
Supplementary material is available at Clinical and Molecular Hepatology website (
http://www.e-cmh.org).
Supplementary Figure 1.
The typical gross classification of HCC based on CECT. (A) Type I; a round tumor with a clear margin. (B) Type II, characterized by a round tumor with a well-defined margin and irregular internal enhancement resembling nodular confluence. (C) Type III, tumors with irregular margins. HCC, hepatocellular carcinoma; CECT, contrast-enhanced computed tomography.
Figure 1.Flowchart of patient selection and comparative sensitivity of NC-MRI versus CECT in HCC detection. (A) Patient selection process. (B) Sensitivity of NC-MRI (T2W+DWI) and CECT (arterial-phase hyperenhancement with washout) for HCC detection. (C) Distribution of gross type on CECT according to washout type, showing the proportion of Type I versus Type II/III morphology. NC-MRI, non-contrast magnetic resonance imaging; CECT, contrast-enhanced computed tomography; HCC, hepatocellular carcinoma.
Abbreviations
contrast-enhanced computed tomography
diffusion-weighted imaging
non-contrast magnetic resonance imaging
REFERENCES
- 1. Kim DW, Chang W, Kim SY, Lim YS, Choi J, Cho J, et al. Noncontrast magnetic resonance imaging for detection of late recurrent hepatocellular carcinoma after curative treatment: a prospective multicenter comparison to contrast-enhanced computed tomography. Clin Mol Hepatol 2025;31:1285-1297.
- 2. Korean Liver Cancer Association (KLCA) and National Cancer Center (NCC) Korea. 2022 KLCA-NCC Korea practice guidelines for the management of hepatocellular carcinoma. Clin Mol Hepatol 2022;28:583-705.
- 3. Joo I, Lee JM, Koh YH, Choi SH, Lee S, Chung JW, et al. 2022 Korean Liver Cancer Association-National Cancer Center Korea Practice Guidelines for Imaging Diagnosis of Hepatocellular Carcinoma: What’s New? Korean J Radiol 2023;24:1-5.
- 4. Fu X, Mao L, Tang M, Yan X, Qiu Y, He J, et al. Gross classification of solitary small hepatocellular carcinoma on preoperative computed tomography: Prognostic significance after radiofrequency ablation. Hepatol Res 2016;46:298-305.
- 5. Park SH, Kim B, Kim SY, Choi SJ, Huh J, Kim HJ, et al. Characterizing Computed Tomography-Detected Arterial Hyperenhancing-Only Lesions in Patients at Risk of Hepatocellular Carcinoma: Can Non-Contrast Magnetic Resonance Imaging Be Used for Sequential Imaging? Korean J Radiol 2020;21:280-289.
- 6. Starekova J, Hernando D, Pickhardt PJ, Reeder SB. Quantification of Liver Fat Content with CT and MRI: State of the Art. Radiology 2021;301:250-262.
- 7. Reeder SB, Yokoo T, França M, Hernando D, Alberich-Bayarri Á, Alústiza JM, et al. Quantification of Liver Iron Overload with MRI: Review and Guidelines from the ESGAR and SAR. Radiology 2023;307:e221856.
- 8. Cha DI, Choi GS, Kim YK, Kim JM, Kang TW, Song KD, et al. Extracellular contrast-enhanced MRI with diffusion-weighted imaging for HCC diagnosis: prospective comparison with gadoxetic acid using LI-RADS. Eur Radiol 2020;30:3723-3734.
- 9. Song Q, Guo Y, Yao X, Rao S, Qian C, Ye D, et al. Comparative study of evaluating the microcirculatory function status of primary small HCC between the CE (DCE-MRI) and Non-CE (IVIM-DWI) MR Perfusion Imaging. Abdom Radiol (NY) 2021;46:2575-2583.
Citations
Citations to this article as recorded by
, Decai Yu1
