ABSTRACT
Aims
As primary hepatocellular carcinoma (HCC) is a prevalent digestive tract malignancy, identifying novel biomarkers for early diagnosis and prognosis is crucial. This study combined specific miRNAs with the GALAD model to create a new diagnostic model for hepatitis B virus (HBV)-related HCC.
Patients & methods
From 2020 to 2022, 884 patients from Zhongshan Hospital were enrolled, including 430 HBV-related HCC, 275 HBV-related chronic liver disease (CLD), and 179 healthy donors (HD). Stepwise regression selected features, and multivariable logistic regression built the GALADM model. A nomogram integrating age, gender, serum markers, and a score derived from seven plasma cell-free microRNAs (miRNA7) was developed.
Results
MiRNA7 value was higher in HCC patients and rose with disease progression. The GALADM model showed superior diagnostic performance, with AUCs of 0.87, 0.96, and 0.90 when distinguishing HCC from CLD, HD, and CLD+HD, outperforming the GALAD model and single markers. It also excelled in diagnosing early-stage and Alpha-fetoprotein (AFP)-negative HCC. The nomogram had an AUC of 0.977 and proved clinically useful.
Conclusion
The GALADM model, combining miRNA7 with the GALAD model, surpasses the original GALAD model, enabling early-stage and AFP-negative HCC diagnosis.
KEYWORDS: Biomarker, microRNAs, diagnostic model, GALADM, hepatocellular carcinoma
1. Introduction
Primary hepatocellular carcinoma (HCC) is one of the most common malignant tumors of the digestive tract. Primary liver cancer, including HCC, has become the sixth most common tumor and the third leading cause of tumor-related death worldwide. According to data statistics from the World Health Organization in 2022, there were approximately 0.865 million thousand new cases of HCC globally, with about 0.758 million deaths [1]. And HBV infection is an important risk factor for the occurrence, development, and recurrence of HCC in China [2]. Due to the lack of specific diagnostic markers for early HCC, many patients with HCC are already in the advanced stage at the time of initial diagnosis and missing the suitable choice for surgery. The 5-year survival rate for advanced HCC is only 15%, while that for early-stage liver cancer is nearly 75% [3]. Therefore, it is of vital importance to explore new biomarkers for the early diagnosis and prognostic assessment of HCC.
The accuracy of current imaging tests and circulating tumor biomarkers for HCC diagnosis is far from satisfactory. Alpha-fetoprotein (AFP) is usually used as a tumor biomarker for the diagnosis of HCC. However, its diagnostic sensitivity is only 52.9%, especially in the poor performance of the early diagnosis. Combined with imaging, the diagnostic sensitivity can be improved [4,5]. Because of the limitations in clinical applications, it requires novel liquid biopsy tools, which can provide easy and noninvasive diagnosis of HCC, identification of patients with higher risk for HCC development, and establishment of cost-effective surveillance programmers for the early detection of HCC in high-risk populations (e.g., HBV related).
MicroRNA (miRNA), a type of small endogenous non-coding RNA, was first discovered in 1993 [6–9]. Fan and his team found that miRNA7 was a group of novel early diagnostic markers for liver cancer, including miR-21, miR-26a, miR-27a, miR-122, miR-192, miR-223, and miR-801. Based on miRNA7, the sensitivity and specificity were 86.1% and 76.8% separately in diagnosing HCC [10].
The aim of this study is to develop a new multivariate prediction model, named GALADM, which includes miRNA7 and other clinical characteristics. This model is designed to improve the diagnosis of HCC in the Chinese population. It is expected to enhance the efficiency of early HCC diagnosis.
2. Materials & methods
2.1. Study design and population characteristics
From Jan 2020 to December 2022, a total of 884 patients in Zhongshan Hospital Affiliated to Fudan University were enrolled in this study，including 430 HBV-related HCC patients, 106 patients with cirrhosis (CIS), 169 patients with chronic hepatitis B (CHB), and 179 healthy donors (HD). The enrolled individuals were separated into four independent cohorts for clinical evaluation. HCC was defined on the basis of pathologic diagnosis, biopsy, ultrasound, computed tomography (CT), magnetic resonance imaging (MRI) characteristics, and AFP serology according to the American Association for Study of Liver Disease guidelines [11]. In addition, healthy donors and patients with chronic hepatitis B and/or liver cirrhosis but without a history of malignancy were enrolled as negative controls.
The present study conformed to the principles of the Declaration of Helsinki. Approval was obtained from the Research Ethics Committee of the Zhongshan Hospital of Fudan University (Approval number: B2019-059 R).
2.2. Measurement of peripheral blood biomarkers levels
Before receiving treatment, the levels of AFP, Alpha-Fetoprotein Lens Culinaris Agglutinin 3(AFP-L3), Protein Induced by Vitamin K Absence or Antagonist-II(PIVKA-II), and miRNA7 were detected and recorded in patients. The serum AFP level was quantitatively measured using the electrochemiluminescence method (Roche, Cat. No. 07026706190), based on the sandwich principle. The serum AFP-L3 concentration was determined by the immunofluorescence method (FUJIFILM Wako Pure Chemical Corporation, Cat. No. 995–66301). The serum PIVKA-II level was measured using chemiluminescent enzyme immunoassay (FUJIREBIO INC., Cat. No. 233887), based on the two-step sandwich principle. The lower limits of detection for AFP, AFP-L3, and PIVKA-II were 0.9 ng/mL, 0.5%, and 5.0 mAU/mL, respectively. The levels of miRNA7 in plasma were detected using a plasma microRNA detection kit (JUSBIO SCIENCES, Cat. No. HCC9655). miR-1228 was used as a stable internal reference for the quantification of miRNAs in HCC patients [12].
The comprehensive evaluation value expressing the content of miRNA7 was calculated according to the formula Logit = −1.9449 + 0.10633 × dCt miR-21 + 0.10219 × dCt miR-26a − 0.012441 × dCt miR-27a − 0.28902 × dCt miR-122 - 0.32779 × dCt miR-192 + 0.25855 × dCt miR-223 - 0.029515 × dCt miR-801. Values less than −0.5 were considered negative, and values greater than or equal to −0.5 were considered positive.
2.3. Statistical analysis
Statistical analyses were performed using SPSS 19.0 for Windows (IBM) and R version 3.4.1 (R Foundation for Statistical Computing). Baseline characteristics and laboratory measurements are presented as mean ± standard deviation for normally distributed data or median (interquartile range) for skewed distributed data. Student’s t-test was used for comparison between groups, as appropriate. If variances within groups were not homogeneous, the nonparametric Mann–Whitney U test was used. A receiver operating characteristics (ROC) curve was used to evaluate the diagnostic value of serum biomarkers. For further analysis, a nomogram was formulated based on a multivariate logistic regression analysis. The discrimination was quantified with the area under the ROC. The ‘rms’ package was used for the nomogram and calibration curve. Decision curve analysis (DCA) was conducted to determine the clinical usefulness of the nomogram by quantifying the net benefits at different threshold probabilities in the validation dataset. p value <0.050 was considered statistically significant.
3. Results
3.1. Patient characteristics
A total of 884 patients were enrolled in this study from Zhongshan Hospital Affiliated to Fudan University, including 430 patients with HBV-related hepatocellular carcinoma (HCC), 275 patients with HBV-related chronic liver disease (CLD), and 179 healthy donors (HD). The clinical characteristics of these 884 participants are summarized in Table 1.
Table 1.
Clinical characteristics of enrolled individuals.
| CLD |
|||||
|---|---|---|---|---|---|
| Variables | HCC (n = 430) | Total (n = 275) |
CHB (n = 169) |
CIS (n = 106) |
HD (n = 179) |
| Gender, male, n (%) | 360 (83.7) | 166 (60.4) | 100 (59.2) | 66 (62.3) | 108 (60.3) |
| Age, years | 57 (50–65) | 58 (50–65) | 58 (50–66) | 58 (50–65) | 51 (44–57) |
| GALADM | 0. 945 (0. 861-.974) | 0.213 (0.118–0.427) | 0.208 (0.113–0.458) | 0.220 (0.128–0.400) | 0.137 (0.060–0.224) |
| GALAD | 0.865 (0.747-.965) | 0.245 (0.149–0.404) | 0.244 (0.140–0.432) | 0.247 (0.162–0.371) | 0.271 (0.135–0.366) |
| AFP, ng/mL | 11.1 (2.8–166.0) | 2.3 (1.7–3.8) | 2.5 (1.7–3.8) | 2.0 (1.6–3.5) | 3.2 (2.5–4.2) |
| Lg(AFP) | 1.1 (0.5–2.2) | 0.4 (0.2–0.6) | 0.4 (0.2–0.6) | 0.3 (0.2–0.5) | 0.5 (0.4–0.6) |
| AFP-L3, % | 6.2 (0.5–32.2) | 0.5 (0.5–4.8) | 0.5 (0.5–3.4) | 0.5 (0.5–6.4) | 0.5 (0.5–0.5) |
| PIVKA-II, mAU/mL | 38.0 (25.0–166.0) | 22.0 (17.0–30.0) | 23.0 (17.0–29.0) | 21.5 (17.0–34.0) | 24.0 (21.0–30.0) |
| Lg(PIVKA-II) | 1.6 (1.4–2.2) | 1.3 (1.2–1.5) | 1.4 (1.2–1.5) | 1.3 (1.2–1.5) | 1.4 (1.3–1.5) |
| MiRNA7 | –0.1 (–0.5–0.3) | –0.7 (–1.0- –0.2) | –0.7 (–1.0- –0.3) | –0.7 (–1.0–0.0) | –0.9 (–1.2- –0.7) |
| Tumor characteristics | |||||
| Tumor size, cm | |||||
| ≤5 | 288 (67.0%) | N/A | N/A | N/A | N/A |
| >5 | 142 (33.0%) | N/A | N/A | N/A | N/A |
| Tumor number | |||||
| Single | 350 (81.4%) | N/A | N/A | N/A | N/A |
| Multiple | 80 (18.6%) | N/A | N/A | N/A | N/A |
| Satellite lesion | |||||
| No | 398 (92.6%) | N/A | N/A | N/A | N/A |
| Yes | 32 (7.4%) | N/A | N/A | N/A | N/A |
| Tumor encapsulation | |||||
| Complete | 196 (45.6%) | N/A | N/A | N/A | N/A |
| Incomplete | 234 (54.4%) | N/A | N/A | N/A | N/A |
| Vascular invasion | |||||
| No | 252 (58.6%) | N/A | N/A | N/A | N/A |
| Yes | 178 (41.4%) | N/A | N/A | N/A | N/A |
| Edmondson stage | |||||
| I-II | 336 (78.1%) | N/A | N/A | N/A | N/A |
| III-IV | 94 (21.9%) | N/A | N/A | N/A | N/A |
| BCLC stage | |||||
| 0/A | 329 (76.5%) | N/A | N/A | N/A | N/A |
| B/C | 101 (23.5%) | N/A | N/A | N/A | N/A |
Note: All continuous variables are presented as median with interquartile range.
Abbreviations: AFP, α-fetoprotein; AFP-L3, lens culinaris agglutinin A-reactive fraction of α-fetoprotein; BCLC, Barcelona Clinic Liver Cancer; PIVKA-II, protein induced by vitamin K absence or antagonist-II; HCC, hepatocellular carcinoma; CLD, chronic liver disease; CIS, cirrhosis; CHB, chronic hepatitis B; HD, healthy donor.
Several parameters related to liver functions were significantly changed in HCC group comparing to the other three groups, including increased AFP, AFP-L3, PIVKA-II, and miRNAs levels (as shown in Figure 1 and Table 1).
Figure 1.
AFP, AFP-L3, PIVKA-II, and miRNA7 level in peripheral blood of HCC, CIS, CHB, and HD (*, p value < 0.05; **, p value < 0.01; ***, p value < 0.001.).
3.2. Development of a novel diagnostic model named GALADM with miRNA7
We analyzed the correlation between the levels of miRNA7 and the traditional HCC biomarkers AFP and PIVKA-II of HCC patients. The results showed that there was no significant correlation between these biomarkers, suggesting that the miRNA7 could serve as a valuable supplement to traditional HCC markers in the diagnosis of HCC (Figure 2).
Figure 2.
The correlation between miRNA7 value and AFP/PIVKA-II level of HCC patients.
The GALAD model and the GALADM model were developed by multivariate Logistic regression analysis. The GALAD score was calculated as follows: GALAD score = −6.145 + 1.061 × (GENDER) + 0.027 × AGE +1.343 × lg (AFP) + 0.028 × AFP-L3 + 1.603 × lg (PIVKA-II). The GALADM score was calculated as: GALADM score = −6.828 + 0.782×(GENDER) + 0.050×AGE +1.193 × lg (AFP) + 0.030 × AFP-L3 + 1.810 × lg (PIVKA-II) + 1.482×miRNA7.
A logistic regression analysis identified the above features as independent predictors. Based on these independent predictors, we developed a model, which is presented as a nomogram in Figure 3.
Figure 3.
Nomogram of the HCC diagnostic model including miRNA7.
Using a nomogram, the value of a single subject is located on each variable axis, and a line is drawn upward to determine the number of points for each variable value. The sum of these figures is on the total and sub-axis, and a line is drawn down to diagnostic HCC. The area under the ROC curve for the prediction nomogram was 0.977 (95% CI, 0.969–0.984) for the subjects’ dataset (Suppl. Fig S1).
The ROC curve and DCA of the diagnostic nomogram model are shown in Suppl. Fig S1 and Suppl. Fig S2. The area under the curve (AUC) for HCC diagnosis was 0.977 (95% CI, 0.969–0.984), with a sensitivity of 92.10% and specificity of 91.90%. The Youden index was 0.839. Moreover, the model demonstrated a significant positive net benefit, indicating high clinical utility in the diagnosis of HCC.
3.3. The performance of the GALADM model in diagnosing and identifying HCC across various subgroups
In this study, the GALADM model was developed by combining miRNA7 with the GALAD model, and the diagnostic efficacy of the model was evaluated by the ROC curve. We found that between HCC and patients with CLD, the area under the curve (AUC) (95% confidence interval, 95% CI) of the GALADM model was 0.87 (0.84–0.90), with the sensitivity and specificity being 76.28% and 81.82%, respectively; while the AUC (95% CI) of the GALAD model was 0.85 (0.82–0.88), and the sensitivity and specificity were 68.37% and 85.82%, respectively (Figure 4; Table 2).
Figure 4.
The ROC curves of GALAD, GALADM, AFP, PIVKA-II, miRNA7, and AFP+miRNA7 for the differential diagnosis of HCC patients from CLD/HD.
Table 2.
The diagnostic performance of GALAD, GALADM, AFP, PIVKA-II, miRNA7, and AFP+miRNA7 in HCC patients.
| AUC (95% CI) |
p (GALADM vs. others) |
Sensitivity (%) | Specificity (%) | PPV (%) | NPV (%) | PLR | NLR | |
|---|---|---|---|---|---|---|---|---|
| HCC (n = 430) vs. CLD (n = 275) | ||||||||
| GALADM | 0.87 (0.84–0.90) | Reference | 76.28 | 81.82 | 86.80 | 68.80 | 4.20 | 0.29 |
| GALAD | 0.85 (0.82–0.88) | 0.059 | 68.37 | 85.82 | 88.30 | 63.40 | 4.82 | 0.37 |
| AFP+miRNA7 | 0.78 (0.74–0.81) | <0.001 | 78.60 | 68.73 | 79.70 | 67.30 | 2.51 | 0.31 |
| miRNA7 | 0.70 (0.66–0.73) | <0.001 | 71.86 | 66.91 | 77.20 | 60.30 | 2.17 | 0.42 |
| AFP | 0.79 (0.76–0.82) | <0.001 | 57.44 | 89.09 | 89.20 | 57.20 | 5.27 | 0.48 |
| PIVKA-II | 0.74 (0.71–0.78) | <0.001 | 65.81 | 72.00 | 78.60 | 57.40 | 2.35 | 0.47 |
| HCC (n = 430) vs. HD (n = 179) | ||||||||
| GALADM | 0.96 (0.94–0.97) | Reference | 87.21 | 93.30 | 96.90 | 75.20 | 13.01 | 0.14 |
| GALAD | 0.89 0.86–0.92) | <0.001 | 69.30 | 97.21 | 98.30 | 56.90 | 24.81 | 0.32 |
| AFP+miRNA7 | 0.90 (0.88–0.93) | <0.001 | 82.56 | 92.18 | 96.20 | 68.70 | 10.56 | 0.19 |
| miRNA7 | 0.86 (0.83–0.89) | <0.001 | 74.42 | 91.62 | 95.50 | 59.90 | 8.88 | 0.28 |
| AFP | 0.72 (0.68–0.75) | <0.001 | 55.81 | 96.65 | 97.60 | 47.70 | 16.65 | 0.46 |
| PIVKA-II | 0.73 (0.69–0.77) | <0.001 | 55.58 | 92.74 | 94.80 | 46.50 | 7.65 | 0.48 |
| HCC (n = 430) vs. CLD+HD (n = 454) | ||||||||
| GALADM | 0.90 (0.88–0.92) | Reference | 86.98 | 78.19 | 79.10 | 86.40 | 3.99 | 0.17 |
| GALAD | 0.87 (0.84–0.89) | <0.001 | 69.30 | 89.43 | 86.10 | 75.50 | 6.55 | 0.34 |
| AFP+miRNA7 | 0.83 (0.80–0.85) | <0.001 | 80.93 | 76.21 | 76.30 | 80.80 | 3.40 | 0.25 |
| miRNA7 | 0.76 (0.73–0.79) | <0.001 | 73.72 | 75.33 | 73.90 | 75.20 | 2.99 | 0.35 |
| AFP | 0.76 (0.73–0.79) | <0.001 | 56.28 | 92.29 | 87.40 | 69.00 | 7.30 | 0.47 |
| PIVKA-II | 0.74 (0.71–0.77) | <0.001 | 55.58 | 85.90 | 78.90 | 67.10 | 3.94 | 0.52 |
Note: GALAD model consists of gender, sex, lg(AFP), AFP-L3, and lg(PIVKA-II); GALADM model consists of gender, sex, lg(AFP), AFP-L3, lg(PIVKA-II) and miRNA7.
Abbreviations: AUC, area under curve; 95% CI, 95% confidence interval; AFP, α-fetoprotein; PIVKA-II, protein induced by vitamin K absence or antagonist-II; PPV, positive predictive value; NPV, negative predictive value; PLR, positive likelihood ratio; NLR, negative likelihood ratio.
For identifying HCC patients from HD, the AUC (95% CI) of the GALADM model was 0.96 (0.94–0.97), with the sensitivity and specificity being 87.21% and 93.30%, respectively; and the AUC (95% CI) of the GALAD model was 0.89 (0.86–0.92), with the sensitivity and specificity being 69.30% and 97.21%, respectively (Figure 4; Table 2).
For differentiating HCC patients from non-HCC patients, the AUC (95% CI) of the GALADM model was 0.90 (0.88–0.92), with the sensitivity and specificity being 86.98% and 78.19%, respectively; and the AUC (95% CI) of the GALAD model was 0.87 (0.84–0.89), with the sensitivity and specificity being 69.30% and 89.43%, respectively (Figure 4; Table 2).
These results suggested that the GALADM model, which optimizes the GALAD model by incorporating miRNA7 indicators, demonstrates superior diagnostic efficacy in diagnosing HCC.
When AFP, PIVKA-II, and miRNAs were used individually to diagnose HCC, their AUCs were all significantly lower than that of the GALADM model (p < 0.001). This indicates that the combined application of multiple markers is more effective than any single marker in diagnosing HCC (Figure 4; Table 2).
3.4. Evaluation the performance of GALADM model in different barcelona clinic liver cancer (BCLC) stages
To evaluate the performance of the GALADM model in diagnosing early-stage HCC, ROC curve analysis was performed to compare HCC patients at BCLC stage 0 with those with chronic liver disease (CLD), healthy donors (HD), and the combined group of CLD and HD. The area under the curve (AUC) (95% confidence interval, 95% CI) was 0.84 (0.79–0.87) for CLD, 0.94 (0.91–0.97) for HD, and 0.88 (0.85–0.91) for CLD + HD, respectively.
When distinguishing between HCC patients at BCLC stage A and those with CLD, HD, and CLD + HD, the AUC (95% CI) was 0.86 (0.83–0.89) for CLD, 0.96 (0.93–0.97) for HD, and 0.90 (0.87–0.92) for CLD + HD, respectively. When distinguishing between HCC patients at BCLC stages B + C and those with CLD, HD, and CLD + HD, the AUC (95% CI) was 0.92 (0.88–0.94) for CLD, 0.97 (0.94–0.99) for HD, and 0.94 (0.91–0.96) for CLD + HD, respectively.
These results indicate that the diagnostic efficacy of the GALADM model is superior to that of the GALAD model in HCC patients at different BCLC stages and is significantly higher than that of any single marker (Figure 5; Table 3). The above results suggest that the establishment of the GALADM diagnostic model can not only be used for the identification of early-stage HCC patients, helping clinicians implement corresponding treatment measures in a timely manner and improve clinical efficacy, but also be used for the differential diagnosis of advanced HCC patients, thus benefiting HCC patients at different stages.
Figure 5.
The ROC curves of GALAD, GALADM, AFP, PIVKA-II, miRNA7, and AFP+miRNA7 for distinguishing different BCLC stages HCC patients from CLD/HD.
Table 3.
The diagnostic performance of GALAD, GALADM, AFP, PIVKA-II, miRNA7, and AFP+miRNA7 in different BCLC stages HCC patients.
| AUC (95% CI) |
p (GALADM vs. others) |
Sensitivity (%) | Specificity (%) | PPV (%) | NPV (%) | PLR | NLR | |
|---|---|---|---|---|---|---|---|---|
| HCC (BCLC 0, n = 77) vs. CLD (n = 275) | ||||||||
| GALADM | 0.84 (0.79–0.87) | Reference | 83.12 | 70.18 | 43.80 | 93.70 | 2.79 | 0.24 |
| GALAD | 0.80 (0.75–0.84) | 0.016 | 61.04 | 86.18 | 55.30 | 88.80 | 4.42 | 0.45 |
| AFP+miRNA7 | 0.80 (0.75–0.84) | 0.097 | 80.52 | 68.73 | 41.90 | 92.60 | 2.57 | 0.28 |
| miRNA7 | 0.73 (0.68–0.77) | <0.001 | 72.73 | 69.09 | 39.70 | 90.00 | 2.35 | 0.39 |
| AFP | 0.76 (0.72–0.81) | 0.006 | 59.74 | 91.27 | 65.70 | 89.00 | 6.85 | 0.44 |
| PIVKA-II | 0.62 (0.57–0.68) | <0.001 | 62.34 | 60.00 | 30.40 | 85.10 | 1.56 | 0.63 |
| HCC (BCLC 0, n = 77) vs. HD (n = 179) | ||||||||
| GALADM | 0.94 (0.91–0.97) | Reference | 84.42 | 93.30 | 84.40 | 93.30 | 12.59 | 0.17 |
| GALAD | 0.83 (0.78–0.88) | <0.001 | 61.04 | 97.77 | 92.20 | 85.40 | 27.31 | 0.40 |
| AFP+miRNA7 | 0.93 (0.89–0.96) | 0.589 | 85.71 | 92.18 | 82.50 | 93.70 | 10.96 | 0.15 |
| miRNA7 | 0.88 (0.84–0.92) | 0.022 | 81.82 | 87.15 | 73.30 | 91.80 | 6.37 | 0.21 |
| AFP | 0.70 (0.64–0.75) | <0.001 | 59.74 | 96.65 | 88.50 | 84.80 | 17.82 | 0.42 |
| PIVKA-II | 0.58 (0.51–0.64) | <0.001 | 37.66 | 92.74 | 69.00 | 77.60 | 5.19 | 0.67 |
| HCC (BCLC 0, n = 77) vs. CLD+HD (n = 454) | ||||||||
| GALADM | 0.88 (0.85–0.91) | Reference | 83.12 | 79.52 | 40.80 | 96.50 | 4.06 | 0.21 |
| GALAD | 0.81 (0.78–0.84) | <0.001 | 61.04 | 90.75 | 52.80 | 93.20 | 6.60 | 0.43 |
| AFP+miRNA7 | 0.85 (0.82–0.88) | 0.175 | 84.42 | 75.55 | 36.90 | 96.60 | 3.45 | 0.21 |
| miRNA7 | 0.79 (0.75–0.82) | 0.001 | 80.52 | 71.15 | 32.10 | 95.60 | 2.79 | 0.27 |
| AFP | 0.74 (0.70–0.77) | <0.001 | 59.74 | 93.39 | 60.50 | 93.20 | 9.04 | 0.43 |
| PIVKA-II | 0.61 (0.56–0.65) | <0.001 | 37.66 | 85.90 | 31.20 | 89.00 | 2.67 | 0.73 |
| HCC (BCLC A, n = 252) vs. CLD (n = 275) | ||||||||
| GALADM | 0.86 (0.83–0.89) | Reference | 75.00 | 80.73 | 78.10 | 77.90 | 3.89 | 0.31 |
| GALAD | 0.85 (0.81–0.88) | 0.156 | 80.95 | 73.45 | 73.60 | 80.80 | 3.05 | 0.26 |
| AFP+miRNA7 | 0.76 (0.72–0.79) | <0.001 | 75.79 | 68.36 | 68.70 | 75.50 | 2.40 | 0.35 |
| miRNA7 | 0.68 (0.64–0.72) | <0.001 | 71.43 | 64.73 | 65.00 | 71.20 | 2.03 | 0.44 |
| AFP | 0.77 (0.73–0.80) | <0.001 | 53.57 | 89.09 | 81.80 | 67.70 | 4.91 | 0.52 |
| PIVKA-II | 0.76 (0.72–0.80) | <0.001 | 68.25 | 72.00 | 69.10 | 71.20 | 2.44 | 0.44 |
| HCC (BCLC A, n = 252) vs. HD (n = 179) | ||||||||
| GALADM | 0.96 (0.93–0.97) | Reference | 86.11 | 93.30 | 94.80 | 82.70 | 12.84 | 0.15 |
| GALAD | 0.89 (0.86–0.92) | <0.001 | 70.24 | 93.30 | 93.70 | 69.00 | 10.48 | 0.32 |
| AFP+miRNA7 | 0.89 (0.86–0.92) | <0.001 | 80.16 | 92.18 | 93.50 | 76.70 | 10.25 | 0.22 |
| miRNA7 | 0.85 (0.81–0.88) | <0.001 | 73.41 | 90.50 | 91.60 | 70.70 | 7.73 | 0.29 |
| AFP | 0.69 (0.64–0.73) | <0.001 | 51.59 | 96.65 | 95.60 | 58.60 | 15.39 | 0.50 |
| PIVKA-II | 0.76 (0.71–0.80) | <0.001 | 56.35 | 92.74 | 91.60 | 60.10 | 7.76 | 0.47 |
| HCC (BCLC A, n = 252) vs. CLD+HD (n = 454) | ||||||||
| GALADM | 0.90 (0.87–0.92) | Reference | 86.11 | 78.19 | 68.70 | 91.00 | 3.95 | 0.18 |
| GALAD | 0.86 (0.84–0.89) | <0.001 | 78.97 | 77.31 | 65.90 | 86.90 | 3.48 | 0.27 |
| AFP+miRNA7 | 0.81 (0.78–0.84) | <0.001 | 80.16 | 74.23 | 63.30 | 87.10 | 3.11 | 0.27 |
| miRNA7 | 0.75 (0.71–0.78) | <0.001 | 71.43 | 75.33 | 61.60 | 82.60 | 2.90 | 0.38 |
| AFP | 0.74 (0.70–0.77) | <0.001 | 51.59 | 92.73 | 79.80 | 77.50 | 7.10 | 0.52 |
| PIVKA-II | 0.76 (0.73–0.79) | <0.001 | 56.35 | 85.90 | 68.90 | 78.00 | 4.00 | 0.51 |
| HCC (BCLC B+C, n = 101) vs. CLD (n = 275) | ||||||||
| GALADM | 0.92 (0.88–0.94) | Reference | 88.12 | 81.82 | 64.00 | 94.90 | 4.85 | 0.15 |
| GALAD | 0.91 (0.87–0.94) | 0.600 | 83.17 | 86.55 | 69.40 | 93.30 | 6.18 | 0.19 |
| AFP+miRNA7 | 0.81 (0.77–0.85) | <0.001 | 85.15 | 68.73 | 50.00 | 92.60 | 2.72 | 0.22 |
| miRNA7 | 0.72 (0.67–0.77) | <0.001 | 74.26 | 70.55 | 48.10 | 88.20 | 2.52 | 0.36 |
| AFP | 0.85 (0.81–0.89) | 0.002 | 86.14 | 70.55 | 51.80 | 93.30 | 2.92 | 0.20 |
| PIVKA-II | 0.80 (0.75–0.84) | <0.001 | 60.40 | 89.82 | 68.50 | 86.10 | 5.93 | 0.44 |
| HCC (BCLC B+C, n = 101) vs. HD (n = 179) | ||||||||
| GALADM | 0.97 (0.94–0.99) | Reference | 91.09 | 94.97 | 91.10 | 95.00 | 18.12 | 0.094 |
| GALAD | 0.94 (0.90–0.96) | 0.003 | 83.17 | 97.77 | 95.50 | 91.10 | 37.22 | 0.17 |
| AFP+miRNA7 | 0.91 (0.87–0.94) | 0.001 | 86.14 | 92.74 | 87.00 | 92.20 | 11.86 | 0.15 |
| miRNA7 | 0.87 (0.82–0.91) | <0.001 | 78.22 | 92.18 | 84.90 | 88.20 | 10.00 | 0.24 |
| AFP | 0.81 (0.76–0.86) | <0.001 | 64.36 | 96.09 | 90.30 | 82.70 | 16.46 | 0.37 |
| PIVKA-II | 0.79 (0.73–0.83) | <0.001 | 64.36 | 97.77 | 94.20 | 82.90 | 28.80 | 0.36 |
| HCC (BCLC B+C, n = 101) vs. CLD+HD (n = 454) | ||||||||
| GALADM | 0.94 (0.91–0.96) | Reference | 88.12 | 88.11 | 62.20 | 97.10 | 7.41 | 0.13 |
| GALAD | 0.92 (0.90–0.94) | 0.148 | 83.17 | 90.97 | 67.20 | 96.00 | 9.21 | 0.19 |
| AFP+miRNA7 | 0.85 (0.82–0.88) | <0.001 | 85.15 | 78.41 | 46.70 | 96.00 | 3.94 | 0.19 |
| miRNA7 | 0.78 (0.74–0.81) | <0.001 | 78.22 | 76.21 | 42.20 | 94.00 | 3.29 | 0.29 |
| AFP | 0.84 (0.80–0.87) | <0.001 | 65.35 | 91.63 | 63.50 | 92.20 | 7.81 | 0.38 |
| PIVKA-II | 0.79 (0.76–0.83) | <0.001 | 62.38 | 92.29 | 64.30 | 91.70 | 8.09 | 0.41 |
Note: GALAD model consists of gender, sex, lg(AFP), AFP-L3, and lg(PIVKA-II); GALADM model consists of gender, sex, lg(AFP), AFP-L3, lg(PIVKA-II) and miRNA7.
Abbreviations: AUC, area under curve; 95% CI, 95% confidence interval; AFP, α-fetoprotein; PIVKA-II, protein induced by vitamin K absence or antagonist-II; PPV, positive predictive value; NPV, negative predictive value; PLR, positive likelihood ratio; NLR, negative likelihood ratio.
3.5. Clinical value of the GALADM model in AFP-negative HCC patients
We further evaluated the diagnostic value of the GALADM model in HCC patients with negative AFP (< 20 ng/mL). The AUC (95% CI) of the GALADM model for distinguishing AFP-negative HCC patients from CLD was 0.79 (0.75–0.82). For differentiating AFP-negative HCC patients from HD, the AUC (95% CI) was 0.92 (0.89–0.95). For distinguishing AFP-negative HCC patients from non-HCC patients, the AUC (95% CI) was 0.84 (0.81–0.87). In all these comparisons, the diagnostic efficacy of the GALADM model was superior to that of the GALAD model (Suppl. Fig S3; Table 4).
Table 4.
The diagnostic performance of GALAD and GALADM in AFP-negative (AFP < 20 ng/mL) HCC patients.
| AUC (95% CI) |
p (GALADM vs. others) |
Cut-off | Sensitivity (%) | Specificity (%) | PPV (%) | NPV (%) | PLR | NLR | |
|---|---|---|---|---|---|---|---|---|---|
| HCC (AFP < 20 ng/mL, n = 244) vs. CLD (n = 275) | |||||||||
| GALADM | 0.79 (0.75–0.82) | Reference | 0.266 | 87.30 | 60.36 | 66.10 | 84.30 | 2.20 | 0.21 |
| GALAD | 0.76 (0.72–0.80) | 0.086 | 0.279 | 82.38 | 59.64 | 64.40 | 79.20 | 2.04 | 0.30 |
| HCC (AFP < 20 ng/mL, n = 244) vs. HD (n = 179) | |||||||||
| GALADM | 0.92 (0.89–0.95) | Reference | 0.266 | 87.30 | 84.92 | 88.70 | 83.10 | 5.79 | 0.15 |
| GALAD | 0.81 (0.77–0.85) | <0.001 | 0.455 | 52.87 | 93.30 | 91.50 | 59.20 | 7.89 | 0.51 |
| HCC (AFP < 20 ng/mL, n = 244) vs. CLD+HD (n = 454) | |||||||||
| GALADM | 0.84 (0.81–0.87) | Reference | 0.268 | 87.30 | 70.04 | 61.00 | 91.10 | 2.91 | 0.18 |
| GALAD | 0.78 (0.75–0.81) | <0.001 | 0.388 | 63.52 | 77.31 | 60.10 | 79.80 | 2.80 | 0.47 |
Note: GALAD model consists of gender, sex, lg(AFP), AFP-L3, and lg(PIVKA-II); GALADM model consists of gender, sex, lg(AFP), AFP-L3, lg(PIVKA-II) and miRNA7.
Abbreviations: AUC, area under curve; 95% CI, 95% confidence interval; PPV, positive predictive value; NPV, negative predictive value; PLR, positive likelihood ratio; NLR, negative likelihood ratio.
These results indicate that the GALADM model can effectively identify AFP-negative HCC patients. Moreover, it outperforms the GALAD model in differentiating AFP-negative HCC patients from HD, thereby reducing the risk of missed diagnoses in high-risk individuals and improving the overall detection rate of HCC.
3.6. Clinical value of the GALADM model in AFP-negative and BCLC 0 stage HCC patients
We further evaluated the diagnostic value of the GALADM model in HCC patients with negative AFP (< 20 ng/mL) and BCLC 0 stage. The AUC (95% CI) of the GALADM model for distinguishing AFP-negative HCC and BCLC 0 Stage patients from CLD was 0.959 (0.937–0.981). For differentiating AFP-negative HCC and BCLC0 stage patients from HD, the AUC (95% CI) was 0.998 (0.995–1.00). For distinguishing AFP-negative and BCLC0 stage HCC patients from non-HCC patients, the AUC (95% CI) was 0.975 (0.960–0.989). However, there was no statistically significant difference in diagnostic and differential diagnostic performance between the GALADM and GALAD models in this patient cohort (Table 5).
Table 5.
The diagnostic performance of GALAD and GALADM in AFP-negative (AFP < 20 ng/mL) and BCLC 0 stage HCC patients.
| AUC (95% CI) |
p (GALADM vs. others) |
Cut-off | Sensitivity (%) | Specificity (%) | PPV (%) | NPV (%) | PLR | NLR | |
|---|---|---|---|---|---|---|---|---|---|
| HCC (AFP < 20 ng/mL, BCLC 0 Stage, n = 37) vs. CLD (n = 275) | |||||||||
| GALADM | 0.959 (0.937-0.981) | Reference | 0.807 | 94.60 | 87.60 | 61.70 | 96.98 | 7.63 | 0.06 |
| GALAD | 0.942 (0.913-0.71) | 0.142 | 0.554 | 97.30 | 86.50 | 48.61 | 99.17 | 7.21 | 0.03 |
| HCC (AFP < 20 ng/mL, BCLC 0 Stage, n = 37) vs. HD (n = 179) | |||||||||
| GALADM | 0.998 (0.995-1.000) | Reference | 0.443 | 100.00 | 95.50 | 51.39 | 100.00 | 22.22 | 0.00 |
| GALAD | 0.986 (0.960–1.000) | 0.346 | 0.554 | 97.30 | 98.30 | 48.61 | 95.61 | 57.24 | 0.03 |
| HCC (AFP < 20 ng/mL, BCLC 0 Stage, n = 37) vs. CLD+HD (n = 454) | |||||||||
| GALADM | 0.975 (0.960–0.989) | Reference | 0.679 | 94.60 | 92.50 | 50.00 | 99.29 | 12.61 | 0.06 |
| GALAD | 0.959 (0.934–0.985) | 0.178 | 0.554 | 97.30 | 91.20 | 44.67 | 99.52 | 11.06 | 0.03 |
Note: GALAD model consists of gender, sex, lg(AFP), AFP-L3, and lg(PIVKA-II); GALADM model consists of gender, sex, lg(AFP), AFP-L3, lg(PIVKA-II) and miRNA7.
Abbreviations: AUC, area under curve; 95% CI, 95% confidence interval; PPV, positive predictive value; NPV, negative predictive value; PLR, positive likelihood ratio; NLR, negative likelihood ratio.
These findings indicate that the GALADM model exhibits superior capability in detecting AFP-negative, BCLC 0 stage HCC, thereby mitigating diagnostic oversight in high-risk cohorts while augmenting overall detection efficacy.
4. Discussion
China is a high-indene region for HCC, with HBV infection being a major risk factor for the development, progression, and recurrence of HCC [13]. Currently, traditional screening methods, which rely on ultrasound and AFP, are not sufficiently sensitive for detecting early-stage HCC. This limitation often leads to patients being diagnosed at an advanced stage, thereby missing the optimal window for surgical intervention [14–16]. Therefore, there is an urgent need to develop a more efficient, objective, and accurate diagnostic method for HCC that is specifically tailored to the Chinese population.
In recent years, with the advancement of diagnostic technologies, several novel diagnostic models and biomarkers have been proposed for the early detection of HCC. For instance, plasma miRNAs detection has been incorporated into the Guidelines for the Diagnosis and Treatment of Primary Liver Cancer (2019 Edition) [17], and it can serve as an early diagnostic biomarker for HCC, especially in patients with serum AFP negativity [18,19]. These advancements underscore the importance of integrating novel biomarkers into clinical practice to improve the early detection and management of HCC.
The GALAD model [20] was developed in 2014, integrating patients’ gender and age with serum tumor markers (AFP, AFP-L3, and PIVKA-II) to form a comprehensive diagnostic scoring model. While primarily used for HCC diagnosis in high-risk populations, it may be particularly suitable for screening among HCC patients with CLD. However, for HCC screening in healthy individuals, where age and gender lack discriminative power, alternative biomarkers with higher specificity or fundamentally different model constructs are required. In this study, we endeavored to integrate miRNA7 into existing diagnostic models to enhance the performance of the GALAD model and develop a novel diagnostic model, termed GALADM, specifically for HBV-related hepatocellular carcinoma (HCC). This advancement is designed to improve the detection of early-stage HCC, particularly in patients with AFP-negative status, thereby enhancing the overall diagnostic efficacy of HCC in the Chinese population.
4.1. Development of a novel diagnostic model named GALADM, incorporating miRNA7
The integration of miRNA7 into the GALAD model has significantly enhanced its diagnostic performance, particularly in diagnosing HCC among healthy populations. The optimized model, termed GALADM, demonstrated superior diagnostic efficacy compared to the original GALAD model, with an AUC (95% CI) exceeding 0.95 and both sensitivity and specificity greater than 90%. These metrics highlight a marked improvement in diagnostic capability.
Notably, in distinguishing BCLC stage 0 HCC patients from healthy individuals, the combination of miRNA7 and AFP achieved better diagnostic performance than the traditional GALAD model. This suggests that the incorporation of miRNA7 not only enhances the early detection of HCC but also provides a more reliable screening option for high-risk populations, potentially reducing the rate of missed diagnoses and improving overall survival rates [10,17].
The GALADM model represents a significant advancement in the diagnosis of HCC, particularly in high-risk populations and early-stage disease. This improvement aligns with recent studies that have identified immune-related miRNAs signatures and combined miRNAs models as promising diagnostic tools for HCC. Compared to traditional biomarkers like AFP, the GALADM model addresses the limitations of detecting early-stage HCC, especially in AFP-negative patients. Future work should focus on validating the GALADM model in larger cohorts and exploring its potential for personalized medicine and targeted therapies.
4.2. The performance of GALADM model in diagnosis and identify HCC
The development and evaluation of the GALADM model represent a significant advancement in the diagnosis and identification of HCC. In this study, we constructed a nomogram based on the GALADM model to further assess its diagnostic performance. The results demonstrate that the GALADM model exhibits exceptional discriminative ability, which indicates a strong alignment between the model’s predictive capabilities and actual outcomes, suggesting both high accuracy and reliability.
The calibration curve of the model shows that the actual predicted probabilities closely approach 1, aligning well with the ideal curve. This close approximation indicates minimal deviation from actual results, further confirming the robustness and effectiveness of the GALADM model. Moreover, DCA was employed to evaluate the clinical utility of the diagnostic model. The results reveal that the GALADM model achieves high net benefit across a wide range of threshold probabilities, highlighting its significant clinical value. This suggests that the model not only provides accurate diagnostic predictions but also offers substantial benefits in clinical decision-making, potentially improving patient outcomes and resource allocation in healthcare settings.
4.3. Evaluation the performance of GALADM model in different BCLC stages
The GALADM model, an extension of the well-established GALAD model, represents a significant advancement in the diagnosis of HCC across different BCLC stages. The original GALAD model [17,21,22], which integrates gender, age, and serum biomarkers such as AFP, AFP-L3%, and PIVKA-II, has been shown to outperform traditional biomarkers like AFP alone, with an AUC(ROC) of up to 0.97–0.98 in validation studies [23]. In our study, the GALADM model further enhances this performance by incorporating additional miRNA signatures, particularly miRNA7, which has been identified as a promising biomarker in early-stage HCC detection. This improvement is crucial for high-risk populations and early-stage disease, where traditional biomarkers often fall short, especially in AFP-negative patients.
When evaluated across different BCLC stages, the GALADM model demonstrates robust diagnostic efficacy. In early-stage HCC (BCLC 0 and A), the model’s ability to distinguish between malignant and benign liver lesions is significantly enhanced compared to the original GALAD model. Additionally, the GALADM model shows strong performance in distinguishing BCLC stage 0 HCC patients from healthy individuals, highlighting its potential for early detection and screening in high-risk cohorts. Future work should focus on validating the GALADM model in larger, diverse cohorts to confirm its clinical utility and explore its potential for personalized medicine and targeted therapies.
4.4. Clinical value of the GALADM model in AFP-negative HCC patients
The GALADM model demonstrates superior diagnostic performance across various BCLC stages of HCC, significantly enhancing early detection and differentiation from CLD and healthy states. In very early-stage HCC (BCLC 0), the model shows enhanced sensitivity, particularly among AFP-negative patients, outperforming traditional biomarker models like GALAD. For early-stage HCC (BCLC A), GALADM maintains high accuracy with both sensitivity and specificity exceeding conventional thresholds, highlighting its potential for timely intervention and improved patient outcomes. In intermediate and advanced stages (BCLC B+C), the model continues to exhibit robust performance, comparable to or better than other multimarker algorithms. Furthermore, the GALADM model also demonstrates superior diagnostic and differential diagnostic performance in AFP-negative, BCLC 0 stage patients. Its high diagnostic accuracy across stages underscores GALADM’s utility in guiding treatment decisions and monitoring disease progression. This comprehensive diagnostic capability positions GALADM as a promising tool for enhancing HCC management in clinical settings.
The study acknowledges several limitations, including the sample size and the representativeness of the patient cohort. The current study cohort may not fully capture the diversity of HCC etiologies and patient populations, potentially limiting the generalizability of the findings. Additionally, potential biases in biomarker measurement and patient selection could influence the results. Future research should focus on validating the GALADM model in larger, multicenter cohorts to confirm its robustness across diverse populations.
Furthermore, the development and progression of HCC is a multifactorial process in which immune dysregulation is critical. Our study lacked comprehensive data on the immunological landscape. Future work will investigate CD4+CD25+Foxp3+ regulatory T cells (Tregs), which are the key immunosuppressive players in the HCC microenvironment and prime therapeutic targets [24]. Moreover, microRNAs, the core GALADM components, might regulate both Treg function and EMT progression by targeting key signaling pathways [25]. We will therefore develop integrated diagnostic-therapeutic frameworks for personalized combination therapies involving immune checkpoint inhibitors and anti-angiogenic agents, aiming to optimize treatment and improve outcomes.
5. Conclusions
In summary, the GALADM model represents a significant advancement in HCC diagnosis, offering improved accuracy and clinical utility. Future research should focus on optimizing the miRNAs combination within the model to further enhance diagnostic precision. Additionally, exploring the integration of GALADM with advanced imaging and AI-based tools could provide a more comprehensive diagnostic approach. Long-term studies are needed to assess the model’s impact on patient outcomes and its potential for prognostic and therapeutic applications.
Supplementary Material
Acknowledgments
The authors thank the patients who participated in this study.
Funding Statement
This work was supported by the National Natural Science Foundation of China [82172348, 82473063], the Baoshan District Health Commission Key Subject Construction Project [BSZK-2023-A18], the Key Disciplines of Shanghai Municipality’s Health System [2024ZDXK0067] and the Science Foundation of Zhongshan Hospital, Fudan University [2024ZSQN01]. We declare no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Article highlights
The GALADM model, a novel diagnostic tool integrating miRNA7 with GALAD, was successfully validated.
The miRNA7 could serve as a valuable supplement to traditional markers in HCC diagnosis.
The diagnostic nomogram model demonstrated significant positive net benefit, indicating high clinical utility for HCC diagnosis.
The GALADM model showed robust diagnostic performance for HCC identification.
GALADM model could be used for the identification of early-stage HCC patients (with BCLC are 0+A) from advanced ones (with BCLC are B+C).
GALADM model performed better in distinguishing HCC patients from healthy controls than the GALAD.
The GALADM model was prior to be a potential diagnostic biomarker for HCC patients.
The diagnostic efficacy of the GALADM model was superior to that of the GALAD model in AFP-Negative HCC Patients.
Author contribution
Yihui Yang; Yan Zhou: Conceived and designed the experiments; Performed the experiments; Analyzed and interpreted the data; Wrote the paper; Anli Jin: Analyzed and interpreted the data.
Wenhao Wu; Chunyan Zhang: collect the clinic data; Beili Wang; Baishen Pan: Conceived and designed the experiments; Yunfan Sun; Wei Guo: Conceived and designed the experiments; Contributed reagents, materials, analysis tools or data.
All authors read and approved the final manuscript.
Disclosure statement
The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.
Reviewer disclosure
Peer reviewers on this manuscript have no relevant financial or other relationships to disclose
Writing disclosure
No writing assistance was utilized in the production of this manuscript.
Ethical conduct of research
The authors state that they have obtained appropriate institutional review board approval (Ethics Committee of Zhongshan Hospital of Fudan University, Approval number: B2019-059R) and/or have followed the principles outlined in the Declaration of Helsinki for all human or animal experimental investigations. In addition, for investigations involving human subjects, informed consent has been obtained from the participants involved.
Supplementary material
Supplemental data for this article can be accessed online at https://doi.org/10.1080/17520363.2025.2600248
References
Papers of special note have been highlighted as either of interest (•) or of considerable interest (••) to readers.
- 1.Bray F, Laversanne M, Sung H, et al. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2024;74(3):229–263. doi: 10.3322/caac.21834 [DOI] [PubMed] [Google Scholar]
- 2.Xi L, Zhu J, Zhang H, et al. Epidemiological trends in gastrointestinal cancers in China: an ecological study. Dig Dis Sci. 2019;64(2):532–543. doi: 10.1007/s10620-018-5335-6 [DOI] [PubMed] [Google Scholar]
- 3.Piñero F, Dirchwolf M, Pessôa MG.. Biomarkers in hepatocellular carcinoma: diagnosis, prognosis and treatment response assessment. Cells. 2020;9(6):1370. doi: 10.3390/cells9061370 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Tzartzeva K, Obi J, Rich NE, et al. Surveillance imaging and alpha fetoprotein for early detection of hepatocellular carcinoma in patients with cirrhosis: a meta-analysis. Gastroenterology. 2018;154(6):1706–1718.e1. doi: 10.1053/j.gastro.2018.01.064 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Becker-Assmann J, Fard-Aghaie MH, Kantas A, et al. Diagnostic and prognostic significance of α-fetoprotein in hepatocellular carcinoma. Chirurg. 2020;91(9):769–777. doi: 10.1007/s00104-020-01118-6 [DOI] [PubMed] [Google Scholar]
- 6.Lee RC, Feinbaum RL, Ambros V. The C. elegans heterochronic gene lin-4 encodes small RNAs with antisense complementarity to lin-14. Cell. 1993;75(5):843–854. doi: 10.1016/0092-8674(93)90529-y [DOI] [PubMed] [Google Scholar]
- 7.Khan P, Siddiqui JA, Kshirsagar PG, et al. MicroRNA-1 attenuates the growth and metastasis of small cell lung cancer through CXCR4/FOXM1/RRM2 axis. Mol Cancer. 2023;22(1):1. doi: 10.1186/s12943-022-01695-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Zhao Z, Xue L, Zheng L, et al. Tumor-derived miR-20b-5p promotes lymphatic metastasis of esophageal squamous cell carcinoma by remodeling the tumor microenvironment. Sig Transduct Target Ther. 2023;8(1):29. doi: 10.1038/s41392-022-01242-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Hussen BM, Hidayat HJ, Salihi A, et al. MicroRNA: a signature for cancer progression. Biomed Pharmacother. 2021;138:111528. doi: 10.1016/j.biopha.2021.111528 [DOI] [PubMed] [Google Scholar]
- 10.Zhou J, Yu L, Gao X, et al. Plasma microRNA panel to diagnose hepatitis B virus-related hepatocellular carcinoma. J Clin Oncol. 2011;29(36):4781–4788. doi: 10.1200/JCO.2011.38.2697 [DOI] [PubMed] [Google Scholar]; •• A plasma microRNA panel that has considerable clinical value in diagnosing early-stage hepatocellular carcinoma (HCC).
- 11.Singal AG, Llovet JM, Yarchoan M, et al. AASLD practice guidance on prevention, diagnosis, and treatment of hepatocellular carcinoma. Hepatology. 2023;78(6):1922–1965. doi: 10.1097/HEP.0000000000000466 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Hu J, Wang Z, Liao BY, et al. Human miR-1228 as a stable endogenous control for the quantification of circulating microRNAs in cancer patients. Int J Cancer. 2014;135(5):1187–1194. doi: 10.1002/ijc.28757 [DOI] [PubMed] [Google Scholar]
- 13.Wen B, Te L, Bai C, et al. Relative contribution of hepatitis B and C viruses in primary liver cancer in China: a systematic review and meta-analysis. J Infect. 2024;89(6):106298. doi: 10.1016/j.jinf.2024.106298 [DOI] [PubMed] [Google Scholar]
- 14.Heimbach JK, Kulik LM, Finn RS, et al. AASLD guidelines for the treatment of hepatocellular carcinoma. Hepatology. 2018;67(1):358–380. doi: 10.1002/hep.29086 [DOI] [PubMed] [Google Scholar]
- 15.Yang JD, Hainaut P, Gores GJ, et al. A global view of hepatocellular carcinoma: trends, risk, prevention and management. Nat Rev Gastroenterol Hepatol. 2019;16(10):589–604. doi: 10.1038/s41575-019-0186-y [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Galle PR, Forner A, Llovet JM, et al. EASL clinical practice guidelines: management of hepatocellular carcinoma. J Hepatol. 2018;69(1):182–236. doi: 10.1016/j.jhep.2018.03.019 [DOI] [PubMed] [Google Scholar]
- 17.Department of Medical Administration, National Health Commission of the People′s Republic of China . Guideline for diagnosis and treatment of primary liver cancer (2024 version). Chin J Hepatol. 2024;32:581–630. doi: 10.3760/cma.j.cn501113-20240611-00290 [DOI] [Google Scholar]; •• Serological molecular markers (such as plasma microRNA, combined with high-risk factors for hepatocarcinogenesis and imaging features, enable clinical diagnosis of hepatocellular carcinoma (HCC) according to a stepwise diagnostic roadmap.
- 18.Solhi R, Pourhamzeh M, Zarrabi A, et al. Novel biomarkers for monitoring and management of hepatocellular carcinoma. Cancer Cell Int. 2024;24(1):428. doi: 10.1186/s12935-024-03600-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Zhang B, Ma X, Zhou Y, et al. Diagnostic value of circulating microRNAs for hepatocellular carcinoma: results of a meta-analysis and validation. Biochem Genet. [cited 2025 Jan 3]. doi: 10.1007/s10528-024-11001-2 [DOI] [PubMed] [Google Scholar]
- 20.Johnson PJ, Pirrie SJ, Cox TF, et al. The detection of hepatocellular carcinoma using a prospectively developed and validated model based on serological biomarkers. Cancer Epidemiol Biomarker Prev. 2014. Jan;23(1):144–153. doi: 10.1158/1055-9965.EPI-13-0870 [DOI] [PubMed] [Google Scholar]; • An entirely objective serum biomarker-based model (GALAD) may facilitate the detection and diagnosis of hepatocellular carcinoma and form the basis for a prospective study comparing this approach with the standard radiological approaches.
- 21.Huang CF, Kroeniger K, Wang CW, et al. Surveillance imaging and GAAD/GALAD scores for detection of hepatocellular carcinoma in patients with chronic hepatitis. J Clin Transl Hepatol. 2024;12(11):907–916. doi: 10.14218/JCTH.2024.00172 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Guan MC, Zhang SY, Ding Q, et al. The performance of GALAD score for diagnosing hepatocellular carcinoma in patients with chronic liver diseases: a systematic review and meta-analysis. J Clin Med. 2023;12(3):949. doi: 10.3390/jcm12030949 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Huang C, Fang M, Xiao X, et al. Validation of the GALAD model for early diagnosis and monitoring of hepatocellular carcinoma in Chinese multicenter study. Liver Int. 2022;42(1):210–223. doi: 10.1111/liv.15082 [DOI] [PubMed] [Google Scholar]
- 24.Granito A, Muratori L, Lalanne C, et al. Hepatocellular carcinoma in viral and autoimmune liver diseases: role of CD4+ CD25+ Foxp3+ regulatory T cells in the immune microenvironment. World J Gastroenterol. 2021;27(22):2994–3009. doi: 10.3748/wjg.v27.i22.2994 [DOI] [PMC free article] [PubMed] [Google Scholar]; • Risk of hepatocellular carcinoma (HCC) in viral and autoimmune liver diseases, and immunotherapy targeting CD4+CD25+ regulatory T cells (Tregs) in the HCC microenvironment.
- 25.Santoro A, Simonelli M, Rodriguez-Lope C, et al. A phase-1b study of tivantinib (ARQ 197) in adult patients with hepatocellular carcinoma and cirrhosis. Br J Cancer. 2013;108(1):21–24. doi: 10.1038/bjc.2012.556 [DOI] [PMC free article] [PubMed] [Google Scholar]; • The mesenchymal-epithelial transition factor (MET) receptor tyrosine kinase pathway is frequently dysregulated in human cancers and plays a critical role in the pathophysiology of HCC.
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.