Estimating Microvascular Invasion in Patients with Resectable Multinodular Hepatocellular Carcinoma by Using Preoperative Contrast-Enhanced MRI: Establishment and Validation of a Risk Score

Figures & data

Figure 1 The flow chart of participants’ inclusion and exclusion.

Abbreviation: MVI, Microvascular Invasion.

Table 1 Baseline Characteristics of mHCC Patients in the Training and Validation Cohort

Variable	Training Cohort (n=193)			Validation Cohort (n=80)			P value*
	MVI (+)	MVI (-)	P value	MVI (+)	MVI (-)	P value	1.000
Age (years)^a	56.7 ± 11.11	58.5 ± 10.59	0.248	56.6 ± 11.53	57.9 ± 11.21	0.611	0.843
Gender			0.189			0.341	0.117
Male	85 (85.0%)	85 (91.4%)		41 (97.6%)	35 (92.1%)
Female	15 (15.0%)	8 (8.6%)		1 (2.4%)	3 (7.9%)
Etiology of liver disease			0.096			0.290	0.524
Hepatitis B virus	87 (87.0%)	89 (95.7%)		36 (85.7%)	36 (94.7%)
Hepatitis C virus	2 (2.0%)	1 (1.1%)		2 (4.8%)	0 (0.0%)
None or other	11 (11.0%)	3 (3.2%)		4 (9.5%)	2 (5.3%)
Edmondson-Steiner grade			0.001			0.020	0.154
I-II	31 (31.0%)	52 (55.9%)		10 (23.8%)
III-IV	69 (69.0%)	41 (44.1%)		32 (76.2%)	19 (50.0%)
Liver cirrhosis (P)			0.546		19 (50.0%)	0.216	0.397
Presence	63 (63.0%)	63 (67.7%)		27 (64.3%)	30 (78.9%)
Absence	37 (37.0%)	30 (32.3%)		15 (35.7%)	8 (21.1%)
Satellite nodule (P)			0.006			0.436	1.000
Presence	13 (13.0%)	2 (2.2%)		5 (11.9%)	2 (5.3%)
Absence	87 (87.0%)	91 (97.8%)		37 (88.1%)	36 (94.7%)
Tumor number (I)^b	2.0 (2.0, 2.0)	2.0 (2.0, 2.0)	0.815	2.0 (2.0, 2.0)	2.0 (2.0, 2.0)	0.286	0.386
LTD (cm) (I)^a	5.1 ± 2.86	3.5 ± 1.59	< 0.001	5.0 ± 2.05	3.50 ± 1.71	0.001	0.958
TTD (cm) (I)^a	7.9 ± 3.67	5.5 ± 2.45	< 0.001	7.5 ± 2.61	5.5 ± 2.38	0.001	0.642
HBV-DNA load			0.215			1.000	0.886
> 10^4 IU/mL	36 (36.0%)	25 (26.9%)		13 (31.0%)	11 (28.9%)
≤ 10^4 IU/mL	64 (64.0%)	68 (73.1%)		29 (69.0%)	27 (71.1%)
Serum AFP			0.001			0.144	0.671
< 20 ng/mL	34 (34.0%)	46 (49.5%)		11 (26.2%)	18 (47.4%)
20–400 ng/mL	28 (28.0%)	33 (35.5%)		19 (45.2%)	12 (31.6%)
> 400 ng/mL	38 (38.0%)	14 (15.1%)		12 (28.6%)	8 (21.1%)
TBIL			0.817			0.505	0.671
≤ 20.4 μmol/L	89 (89.0%)	84 (90.3%)		38 (90.5%)	32 (84.2%)
> 20.4 μmol/L	11 (11.0%)	9 (9.7%)		4 (9.5%)	6 (15.8%)
DBIL			1.000			0.774	0.737
≤ 6.8 μmol/L	80 (80.0%)	75 (80.6%)		34 (81.0%)	32 (84.2%)
> 6.8 μmol/L	20 (20.0%)	18 (19.4%)		8 (19.0%)	6 (15.8%)
TP			0.870			0.349	0.237
≤ 65 g/L	25 (25.0%)	25 (26.9%)		12 (28.6%)	15 (39.5%)
> 65 g/L	75 (75.0%)	68 (73.1%)		30 (71.4%)	23 (60.5%)
ALB			0.741			1.000	1.000
≤ 35 g/L	4 (4.0%)	5 (5.4%)		2 (4.8%)	2 (5.3%)
> 35 g/L	96 (96.0%)	88 (94.6%)		40 (95.2%)	36 (94.7%)
ALT			0.057			0.596	0.874
> 50 U/L	28 (29.0%)	15 (16.1%)		10 (23.8%)	7 (18.4%)
≤ 50 U/L	72 (71.0%)	78 (83.9%)		32 (76.2%)	31 (81.6%)
AST			0.016			0.132	1.000
> 40 U/L	35 (35.0%)	18 (19.4%)		15 (35.7%)	7 (18.4%)
≤ 40 U/L	65 (65.0%)	75 (80.6%)		27 (64.3%)	31 (81.6%)
AKP			0.486			1.000	0.651
> 125 U/L	12 (12.0%)	8 (8.6%)		3 (7.1%)	3 (7.9%)
≤ 125 U/L	88 (88.0%)	85 (91.4%)		39 (92.9%)	35 (92.1%)
GGT			0.564			0.125	0.791
> 60 U/L	49 (49.0%)	41 (44.1%)		24 (57.1%)	15 (39.5%)
≤ 60 U/L	51 (51.0%)	52 (55.9%)		18 (42.9%)	23 (60.5%)
TBA			0.379			0.350	0.417
> 10 μmol/L	37 (37.0%)	41 (44.1%)		17 (40.5%)	11 (28.9%)
≤ 10 μmol/L	63 (63.0%)	52 (55.9%)		25 (59.5%)	27 (71.1%)
PLT			0.349			0.535	0.723
≤ 100◊10^9/L	15 (15.0%)	19 (20.4%)		5 (11.9%)	7 (18.4%)
> 100◊10^9/L	85 (85.0%)	74 (79.6%)		37 (88.1%)	31 (81.6%)
PT			1.000			0.664	0.267
> 13s	10 (10.0%)	10 (10.8%)		2 (4.8%)	3 (7.9%)
≤ 13s	90 (90.0%)	83 (89.2%)		40 (95.2%)	35 (92.1%)

Notes: Unless otherwise indicated, data are the number of patients, with the percentage in parentheses. *P value for the training cohort and validation cohort. ^aData are presented as mean ± standard deviation. ^bData are presented as median (interquartile range). (P): Identified with postoperative pathological examination. (I): Identified with preoperative MRI imaging.

Abbreviations: MVI, microvascular invasion; LTD, the largest tumor diameter; TTD, total tumor diameter; AFP, α-fetoprotein; TBIL, total bilirubin; DBIL, direct bilirubin; TP, total protein; ALB, albumin; ALT, alanine aminotransferase; AST, aspartate aminotransferase; AKP, alkaline phosphatase; GGT, γ-glutamyltranspetidase; TBA, total bile acid; PLT, platelet; PT, prothrombin time.

Table 2 MR Features of mHCC Patients in the Training and Validation Cohort

Variable	Training Cohort (n=193)			Validation Cohort (n=80)			P value*
	MVI (+)	MVI (-)	P value	MVI (+)	MVI (-)	P value	1.000
Tumor number > 2	14 (14.0%)	14 (15.1%)	0.841	5 (11.9%)	8 (21.1%)	0.366	0.853
Liver cirrhosis			0.146			0.641	0.224
Presence	51 (51.0%)	58 (62.4%)		26 (61.9%)	26 (68.4%)
Absence	49 (49.0%)	35 (37.6%)		16 (38.1%)	12 (31.6%)
Satellite nodule			< 0.001			0.678	0.506
Presence	20 (20.0%)	1 (1.1%)		4 (9.5%)	2 (5.3%)
Absence	80 (80.0%)	92 (98.7%)		38 (90.5%)	36 (94.7%)
Hemorrhage in mass			0.005			0.152	0.237
Presence	34 (34.0%)	15 (16.1%)		17 (40.5%)	9 (23.7%)
Absence	66 (66.0%)	78 (83.9%)		25 (59.5%)	29 (76.3%)
Fat in mass			0.066			0.816	0.103
Presence	19 (19.0%)	29 (31.2%)		14 (33.3%)	14 (36.8%)
Absence	81 (81.0%)	64 (68.8%)		28 (66.7%)	24 (63.2%)
Arterial rim enhancement			0.221			0.200	0.287
Presence	12 (12.0%)	6 (6.5%)		8 (19.0%)	3 (7.9%)
Absence	88 (88.0%)	87 (93.5%)		34 (81.0%)	35 (92.1%)
Incomplete or absent radiological capsule			0.196			0.022	0.421
Presence	40 (40.0%)	46 (49.5%)		11 (26.2%)	20 (52.6%)
Absence	60 (60.0%)	47 (50.5%)		31 (73.8%)	18 (47.4%)
Peritumoral enhancement			0.052			0.329	0.332
Presence	43 (43.0%)	27 (29.0%)		15 (35.7%)	9 (23.7%)
Absence	57 (57.0%)	66 (71.0%)		27 (64.3%)	29 (76.3%)
Mosaic architecture			< 0.001			0.073	0.507
Presence	62 (62.0%)	29 (31.2%)		22 (52.4%)	12 (31.6%)
Absence	38 (38.0%)	64 (68.8%)		20 (47.6%)	26 (68.4%)
Nodule in nodule architecture			1.000			0.495	1.000
Presence	3 (3.0%)	3 (3.2%)		2 (4.8%)	0 (0.0%)
Absence	97 (97.0%)	90 (96.8%)		40 (95.2%)	38 (100.0%)
Non-smooth tumor margin			0.004			0.108	0.640
Presence	84 (84.0%)	61 (65.6%)		30 (71.4%)	33 (86.8%)
Absence	16 (16.0%)	32 (34.4%)		12 (28.6%)	5 (13.2%)
Atypical enhancement pattern			0.483			0.012	0.871
Presence	19 (19.0%)	22 (23.7%)		13 (31.0%)	3 (7.9%)
Absence	81 (81.0%)	71 (76.3%)		29 (69.0%)	35 (92.1%)
TTD (cm)^a	7.9 ± 3.67	5.5 ± 2.45	< 0.001	7.5 ± 2.61	5.5 ±2.38	0.001	0.642
RLSD (cm)^a	2.8 ± 1.88	2.4 ± 1.42	0.079	3.1 ± 2.40	2.8 ± 1.35	0.430	0.167

Notes: Unless otherwise indicated, data are the number of patients, with the percentage in parentheses. *P value for the training cohort and validation cohort. ^aData are presented as mean ± standard deviation.

Abbreviations: MVI, microvascular invasion; TTD, total tumor diameter; RLSD, the ratio of the largest to the smallest tumor diameter.

Table 3 Univariable and Multivariable Analysis of MVI in the Training Cohort

Variable	Univariate Analysis		Multivariate Analysis
	OR (95% CI)	P value	OR (95% CI)	P value	Regression Coefficient
Male	0.533 (0.215–1.324)	0.175
Age	0.985 (0.959–1.011)	0.247
Etiology of liver disease		0.123
None or other	1
Hepatitis B virus	0.267 (0.072–0.998)	0.048
Hepatitis C virus	0.545 (0.036–8.270)	0.662
HBV-DNA load > 10^4 IU/mL	1.530 (0.828–2.827)	0.175
Serum AFP		0.002		0.043
< 20 ng/mL	1		1
20–400 ng/mL	1.148 (0.587–2.245)	0.687	1.363 (0.631–2.944)	0.430
> 400 ng/mL	3.672 (1.724–7.823)	0.001	3.038 (1.273–7.249)	0.012*	1.111
TBIL > 20.4 μmol/L	1.154 (0.455–2.924)	0.763
DBIL > 6.8 μmol/L	1.042 (0.512–2.120)	0.910
TP ≤ 65 g/L	0.907 (0.476–1.727)	0.766
ALB ≤ 35 g/L	0.733 (0.191–2.818)	0.652
ALT > 50 U/L	2.022 (1.000–4.089)	0.050
AST > 40 U/L	2.244 (1.161–4.334)	0.016
AKP > 125 U/L	1.449 (0.564–3.720)	0.441
GGT > 60 U/L	1.219 (0.691–2.148)	0.494
TBA > 10 μmol/L	0.745 (0.419–1.326)	0.317
PLT ≤ 100◊10^9/L	0.687 (0.326–1.448)	0.324
PT > 13s	0.922 (0.365–2.328)	0.864
Tumor number > 2	0.919 (0.412–2.047)	0.835
Liver cirrhosis	0.628 (0.354–1.115)	0.112
Satellite nodule	23.000 (3.019–175.228)	0.002	12.979 (1.396–120.651)	0.024*	2.563
Hemorrhage in mass	2.679 (1.343–5.342)	0.005
Fat in mass	0.518 (0.266–1.006)	0.052	0.298 (0.130–0.683)	0.004*	−1.210
Arterial rim enhancement	1.977 (0.710–5.504)	0.192
Incomplete or absent capsule	0.681 (0.385–1.205)	0.187
Peritumoral enhancement	1.844 (1.014–3.352)	0.045
Mosaic architecture	3.601 (1.984–6.537)	< 0.001	2.052 (1.003–4.198)	0.049*	0.719
Nodule in nodule architecture	0.928 (0.183–4.716)	0.928
Non-smooth tumor margin	2.754 (1.389–5.462)	0.004
Atypical enhancement pattern	0.757 (0.379–1.612)	0.430
LTD	1.502 (1.251–1.802)	< 0.001
TTD	1.326 (1.172–1.500)	< 0.001	1.230 (1.064–1.423)	0.005*	0.207
RLSD	1.170 (0.978–1.399)	0.087

Notes: *Serum AFP > 400 ng/mL, satellite nodule, fat in mass, mosaic architecture and TTD were independent significant factors in multivariate analyses.

Abbreviations: OR, odds ratio; CI, confidence interval; MVI, microvascular invasion; AFP, α-fetoprotein; TBIL, total bilirubin; DBIL, direct bilirubin; TP, total protein; ALB, albumin; ALT, alanine aminotransferase; AST, aspartate aminotransferase; AKP, alkaline phosphatase; GGT, γ-glutamyltranspetidase; TBA, total bile acid; PLT, platelet; PT, prothrombin time; LTD, largest tumor diameter; TTD, total tumor diameter; RLSD, the ratio of the largest to the smallest tumor diameter.

Table 4 Risk Score for Factors Associated with MVI of Multinodular HCC

Parameter	Score
AFP
≤ 400 ng/mL	0
> 400 ng/mL	1
Satellite nodule (I)
Absence	0
Presence	3
Fat in mass
Presence	0
Absence	1
Mosaic architecture
Absence	0
Presence	1
TTD
≤ 4 cm	0
4–8 cm	1
8–12 cm	2
12–16 cm	3
≥ 16 cm	4

Note: (I) means the satellite nodule on MR imaging.

Abbreviations: AFP, α-fetoprotein; TTD, total tumor diameter.

Figure 2 The calibration curve compared predicted and observed probability of microvascular invasion in the (A) training cohort and (B) validation cohort.

Figure 3 47-year-old man without pathological microvascular invasion. The serum α-fetoprotein was < 400 ng/mL. The Axial contrast-enhanced MR demonstrated two HCC lesions in the right liver section with the total diameter of 3.7 cm. Both lesions contained fat (A and B, white arrows) and demonstrated hyperintense on T2-weighted image (C), diffusion-weighted image (D) and hypointense on T1-weighted image (E). After the injection of contrast agent, both lesions showed typical enhancement pattern with enhancement on the arterial phase (F) and washout on the portal phase (G) and delayed phase (H). Both lesions showed compete radiological capsule. Risk score value of 0 score categorized the patient into low-risk group. After the surgery, no recurrence was observed and the recurrence-free survival was 36 months.

Figure 4 56-year-old man with pathological microvascular invasion. The serum α-fetoprotein was > 400 ng/mL. Axial contrast-enhanced MR demonstrated two HCC lesions in the right liver section with the total diameter of 9.5 cm. Both lesions showed hyperintense on T2-weighted image (A1 and A2) and diffusion-weighted image (B1 and B2) and hypointense on T1-weighted image (C1 and C2). Both lesions showed typical enhancement pattern with enhancement on the arterial phase (D1 and D2) and washout on the portal phase (E1 and E2) and delayed phase (F1 and F2). The smaller lesion showed hemorrhage in mass (C2) and mosaic architecture (D2, arrow). The tumor emboli were found in the specimen sampled at the junction of the smaller tumor (A2–F2) and adjacent liver tissues. Risk score value of 5 categorized the patient into high-risk group. After surgery, recurrence was observed and the recurrence-free survival was 9 months.

Figure 5 MVI Risk Score-based Prognostic Stratification. Graphs showed that multinodular HCC patients in the low-risk groups had significantly shorter RFS than those in the high-risk group in the (A) training cohort and (B) validation cohort (Log rank test, both P < 0.05).

Abbreviations: MVI, Microvascular Invasion; RFS, Recurrence-free Survival.