MRI Features for Predicting Microvascular Invasion and Postoperative Recurrence in Hepatocellular Carcinoma Without Peritumoral Hypointensity

Figures & data

Table 1 Clinical and Pathological Characteristics of MVI-Positive and MVI-Negative Cohorts

Characteristic	MVI Positive (n=63)	MVI Negative (n=67)	P value
Age (year)	55.0±11.3	55.6±11.4	0.805
Gender			0.184
Female	5(7.9%)	11 (16.4%)
Male	58(92.1%)	56 (83.6%)
Etiology of liver disease			0.703
Hepatitis B	52(82.5%)	58(86.6%)
Hepatitis C	7(11.1%)	7(10.4%)
No or other	4(6.3%)	2(3.0%)
WBC (×10^9/L)	6.06(5.17–7.43)	5.71(4.50–6.73)	0.156
RBC (×10^12/L)	4.67±0.60	4.77±0.68	0.384
HB (g/L)	147(130–155)	142(130–152)	0.545
PLT (×10^9/L)	170(138–215)	178(132–220)	0.929
PT (s)	13.2±1.3	12.9±1.3	0.169
INR (R)	1.02(0.96–1.08)	1.03(0.97–1.11)	0.400
APTT(s)	35.8(28.5–40.1)	36.8(32.6–39.7)	0.584
FIB (g/L)	2.92(2.50–3.66)	2.89(2.48–3.39)	0.631
ALT (U/L)	28.0(20.0–42.5)	26.0(18.0–42.0)	0.531
AST (U/L)	29.0(21.0–50.0)	27.0(22.0–41.0)	0.702
ALB (g/L)	43.2±4.4	43.3±4.3	0.945
GLB (g/L)	28.6±4.6	28.9±4.5	0.655
ALP (U/L)	80(65.5–113.0)	84.0(68.5–107.0)	0.718
GGT (U/L)	52.0(31.5–97.0)	50.0(26.5–100.0)	0.730
TBIL (μmol/L)	11.8(8.9–15.4)	11.9(8.4–15.9)	0.716
Cr (μmol/L)	81.0(70.0–89.0)	80.0(67.0–88.0)	0.658
AFP (ng/mL)	14.0(4.3–222.2)	17.0(4.9–375.6)	0.628
CTP	5.0(5.0–6.0)	5.0(5.0–6.0)	0.438
MELD	4.52(3.03–5.70)	4.26(1.79–6.53)	0.757
Edmondson grade			<0.001
I	1 (1.6%)	4 (6.0%)
II	17 (27.0%)	40 (59.7%)
III	37 (58.7%)	22 (32.8%)
IV	8 (12.7%)	1 (1.5%)

Notes: Chi-square test or Fisher exact probability test, as appropriate, were used to compare the difference in categorical variables among different groups. Student t test or Mann–Whitney U-test, as appropriate, were used to compare the difference in continuous variables.

Abbreviations: MVI, microvascular invasion; WBC, white blood cell count; RBC, red blood cell count; HB hemoglobin; PLT, platelets; PT, prothrombin time;INR, international normalized ratio of prothrombin; APTT, activated partial thromboplastin time, FIB, fibrinogen, ALT, alanine aminotransferase; AST, aspartate aminotransaminase; ALB, albumin; GLB, globulin; ALP, alkaline phosphatase; GGT, gamma-glutamyl transferase; TBIL, total bilirubin; Cr, creatinine; AFP, alpha-fetoprotein; CTP, Child-Turcotte-Pugh stage; MELD, model for end-stage liver disease score.

Figure 1 Flowchart showing the patient inclusion and exclusion criteria.

Table 2 The Interobserver Agreement of MRI Features

Characteristics	Weighted Kappa	95% CI
Tumor margin	0.747	0.626–0.867
Capsule	0.609	0.461–0.758
Mosaic sign	0.667	0.536–0.797
Intratumoral necrosis	0.794	0.687–0.900
Intratumoral bleeding	0.757	0.644–0.871
Intratumoral fat	0.530	0.324–0.737
Low vascular composition	0.591	0.469–0.713
Intratumoral artery	0.638	0.516–0.760
Satellite nodule	0.885	0.796–0.974
Corona enhancement	0.621	0.444–0.798
Peritumoral hypointensity on HBP	0.511	0.339–0.682
Increased uptake in HBP	0.431	0.153–0.709

Abbreviations: CI, confidence interval; HBP, hepatobiliary phase.

Table 3 MRI Features of MVI-Positive and MVI-Negative HCCs

Characteristics	Total Patients (n=130)	MVI Positive (n=63)	MVI Negative (n=67)	P value
Tumor diameter (cm)	3.80(2.30–6.30)	4.30(2.65–6.10)	3.30(1.85–5.20)	0.116
Tumor margin				0.802
Smooth	44(33.8%)	22(34.9%)	22(32.8%)
Nonsmooth	86(66.2%)	41(65.1%)	45(67.2%)
Capsule				0.182
Absent	27(20.8%)	10(15.9%)	17(25.4%)
Present	103(79.2%)	53(84.1%)	50(74.6%)
Mosaic sign				0.230
Absent	71(54.6%)	31(49.2%)	40(59.7%)
Present	59(45.4%)	32(50.8%)	27(40.3%)
Intratumoral necrosis				0.781
Absent	80(61.5%)	38(60.3%)	42(62.7%)
Present	50(38.5%)	25(39.7%)	25(37.3%)
Intratumoral bleeding				0.781
Absent	80(61.5%)	38(60.3%)	42(62.7%)
Present	50(38.5%)	25(39.7%)	25(37.3%)
Intratumoral fat				0.946
Absent	107(82.3%)	52(82.5%)	55(82.7%)
Present	23(17.7%)	11(17.5%)	12(17.9%)
Low vascular composition				0.649
<20%	43(33.1%)	21(33.3%)	22(32.8%)
20–50%	50(38.5%)	22(34.9%)	28(41.8%)
>50%	37(28.5%)	20(31.7%)	17(25.4%)
Intratumoral artery				0.057
Absent	73(56.2%)	30(47.6%)	43(64.2%)
Present	57(43.8%)	33(52.4%)	24(35.8%)
Satellite nodule				0.822
Absent	92(70.8%)	44(69.8%)	48(71.6%)
Present	38(29.2%)	19(30.2%)	19(28.4%)
Corona enhancement				0.021
Absent	109(83.8%)	48(76.2%)	61(91.0%)
Present	21(16.2%)	15(23.8%)	6(9.0%)
Peritumoral hypointensity on HBP				0.021
Absent	109(83.8%)	48(76.2%)	61(91.0%)
Present	21(16.2%)	15(23.8%)	6(9.0%)
Increased uptake in HBP				0.719
Absent	122(93.8%)	60(95.2%)	62(92.5%)
Present	8(6.2%)	3(4.8%)	5(7.5%)

Notes: Chi-square test or Fisher exact probability test, as appropriate, were used to compare the difference in categorical variables among different groups. Mann–Whitney U-test was used to compare the difference in continuous variables.

Abbreviations: MVI, microvascular invasion; HBP, hepatobiliary phase.

Figure 2 Representative MRI features associated with histopathological findings. HBP images (a) for a 29-year-old male patient showing peritumoral HBP hypointensity (arrow) and satellite nodule (arrowhead). Hepatocellular carcinoma with MVI-positive was confirmed by histopathology (b). HBP images (c) for a 41-year-old male patient showing a radiological capsule (arrow) without peritumoral HBP hypointensity. Hepatocellular carcinoma with MVI-positive was confirmed by histopathology (d).

Abbreviations: HBP, hepatobiliary phase; MVI, microvascular invasion.

Table 4 MRI Features of MVI-Negative and MVI-Positive HCCs in Subgroup Without Peritumoral HBP Hypointensity

Characteristics	Total Patients (n=109)	MVI Positive (n=48)	MVI Negative (n=61)	P value
Tumor diameter (cm)	3.40(2.20–5.60)	3.45(2.25–5.55)	3.40(2.10–5.60)	0.116
Tumor margin				0.579
Smooth	40(36.7%)	19(39.6%)	21(34.4%)
Nonsmooth	69(63.3%)	29(60.4%)	40(65.6%)
Capsule				0.038
Absent	21(19.5%)	5(10.4%)	16(26.2%)
Present	88(80.7%)	43(89.6%)	45(73.8%)
Mosaic sign				0.284
Absent	63(57.8%)	25(52.1%)	38(62.3%)
Present	46(42.2%)	23(47.9%)	23(37.7%)
Intratumoral necrosis				0.255
Absent	70(64.2%)	28(58.3%)	42(68.9%)
Present	39(35.8%)	20(41.7%)	19(31.1%)
Intratumoral bleeding				0.773
Absent	72(66.1%)	31(64.6%)	41(67.2%)
Present	37(33.9%)	17(35.4%)	20(32.8%)
Intratumoral fat				0.406
Absent	107(82.6%)	38(79.2%)	52(85.2%)
Present	23(17.4%)	10(20.8%)	9(14.8%)
Low vascular composition				0.782
<20%	40(36.7%)	18(37.5%)	22(36.1%)
20–50%	40(36.7%)	16(33.3%)	24(39.3%)
>50%	19(26.6%)	14(29.2%)	15(24.6%)
Intratumoral artery				0.241
Absent	68(62.4%)	27(56.3%)	41(62.4%)
Present	41(37.6%)	21(43.8%)	20(37.6%)
Satellite nodule				0.884
Absent	81(74.3%)	36(75.0%)	45(73.8%)
Present	28(25.7%)	12(25.0%)	16(26.2%)
Corona enhancement				0.107
Absent	93(85.3%)	38(79.2%)	55(90.2%)
Present	16(14.7%)	10(20.8%)	6(9.8%)
Increased uptake in HBP				0.394
Absent	102(93.6%)	46(95.8%)	56(91.8%)
Present	7(6.4%)	2(4.2%)	5(8.2%)

Notes: Chi-square test or Fisher exact probability test, as appropriate, were used to compare the difference in categorical variables among different groups. Mann–Whitney U-test was used to compare the difference in continuous variables.

Abbreviations: MVI, microvascular invasion; HBP, hepatobiliary phase.

Figure 3 Kaplan–Meier curves showing recurrence-free survival in patients with and without peritumoral hepatobiliary phase hypointensity.

Table 5 Univariate and Multivariate Cox Regression of MRI Features in Predicting Intrahepatic Recurrence in Subgroup Without Peritumoral HBP Hypointensity

Imaging Feature	Univariable Analysis		Multivariable Analysis
	Hazard Ratio (95% CI)	P value	Hazard Ratio (95% CI)	P value
Tumors size > 5 cm	2.928(1.545–5.550)	0.001	–	–
Unsmooth margin	2.186(1.037–4.611)	0.040	–	–
Mosaic sign	1.964(1.032–3.737)	0.040	–	–
Necrosis	2.341(1.242–4.411)	0.009	–	–
Satellite nodule	3.324(1.733–6.378)	<0.001	3.324(1.733–6.378)	<0.001

Abbreviations: HBP, hepatobiliary phase; CI, confidence interval.

Figure 4 Kaplan–Meier curves showing recurrence-free survival in patients without peritumoral HBP hypointensity (PHBPH) and satellite nodule (SN), with SN but without PHBPH and with PHBPH.

Figure 5 Representative MRI features associated with recurrence-free survival. HBP images (a) for a 60-year-old female patient showing an irregularly hypointense tumor (arrow) but without peritumoral HBP hypointensity or a satellite nodule. Tumor recurrence has not been found after 28 months of follow-up. HBP images (b) for a 50-year-old male patient showing multiple satellite nodules (arrow) around the tumor without peritumoral HBP hypointensity. Tumor recurrence was detected 4 months after curative resection.

Figure 6 Two stepwise flowchart incorporating peritumoral hepatobiliary phase hypointensity and satellite nodule for stratification of the risk of intrahepatic recurrence. Recurrence-free survival was significantly shorter in patients with high-risk hepatocellular carcinoma than in those with low-risk hepatocellular carcinoma (hazard ratio 3.511, 95% confidence interval 1.995–6.181, P<0.001).