Threshold growth has a limited role in differentiating HCC from other focal hepatic lesions

doi:10.21203/rs.3.rs-2597945/v1

Download PDF

Research Article

Threshold growth has a limited role in differentiating HCC from other focal hepatic lesions

https://doi.org/10.21203/rs.3.rs-2597945/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 04 Dec, 2023

Read the published version in BMC Medical Imaging →

You are reading this latest preprint version

Background

The role of threshold growth, as one of the major features (MFs) of HCC in the Liver Imaging Reporting and Data System (LI-RADS) is inconsistent. This study evaluated the LI-RADS diagnostic performance for HCC when threshold growth was removed or replaced by highly frequent ancillary features (AFs).

Methods

This was a retrospective institutional review board-approved study including patients with a high HCC risk with gadoxetic acid-enhanced MRI findings consistent with pathologically proven focal hepatic observations. With pathological results used as the gold standard reference, the observations were divided into three groups: HCC, non-HCC malignancy and benign lesion. The sizes of the lesions with and without threshold growth were compared. The MFs and AFs of each observation were evaluated and compared among the three groups to select the most highly frequent AFs of HCC and was used to replace threshold growth. The LI-RADS categories of observations were categorized as follows: Scheme A, using all MFs and AFs according to LI-RADS v2018; scheme B, using all MFs except threshold growth, with threshold growth treated as an AF favouring malignancy; and scheme C, using the highly frequent AFs inplace of threshold growth as new MFs. The LR-5 (the category of definitely HCC) diagnostic performance for HCC among the 3 schemes was compared.

Results

A total of 379 patients and 426 observations were included. There was no statistically significant difference in the frequency of threshold growth between HCCs and non-HCC malignancies (p = 0.560). Whether HCCs, non-HCC malignancies, or benign lesions, the mean size with threshold growth was smaller than that without threshold growth (all p < 0.05). The nodule-in-nodule feature was a highly frequent AF (p < 0.05) and was used to replace threshold growth as a new MF in scheme C. The LR-5 diagnostic performance values for HCC with schemes A, B, and C were respectively as follows: a sensitivity of 74.4%, 74.0% and 75.6%; specificity of 88.6%, 88.6% and 88.6%; and accuracy of 80.3%, 80.0% and 81.0%. There was no statistically significant difference in diagnostic performance between schemes A and B or between schemes A and C (all p > 0.05).

Conclusion

The threshold growth is removed or replaced by the nodule-in-nodule feature

Hepatocellular carcinoma (HCC)

Magnetic resonance imaging (MRI)

Liver

Diagnosis

With stringent criteria, accurate and reliable non-invasive diagnosis of hepatocellular carcinoma (HCC) can be made with imaging exams, such as computed tomography (CT) and magnetic resonance imaging (MRI) [1], often without the need for further invasive pathological workup. Therefore, imaging studies play a crucial role in the diagnosis of HCC and can be used to guide subsequent treatment [2]. In particular, Gadoxetic acid-enhanced MRI (EOB-MRI), as a first-line diagnostic test, has good diagnostic performance for HCC [1].

The Liver Imaging Reporting and Data System (LI-RADS) published by the American College of Radiology (ACR) is used to assign a category to focal hepatic observations obtained for high-risk HCC patients. Categories of LI-RADS were respectively as follows: LR-1 (definitely benign), LR-2 (probably benign), LR-3 (intermediate probability of malignancy), LR-4 (probably HCC), LR-5 (definitely HCC), LR-M (probably and definitely malignant but not HCC specific), and LR-TIV (definitely tumour-in-vein). Each category reflects a relative probability of benignity, malignancy or HCC in general. The LI-RADS aims to standardize imaging data collection and reporting for HCC to enhance communication, reduce interobserver variability, promote quality assurance and improve diagnostic performance [3].

The CT/MRI LI-RADS describes many major features (MFs) and ancillary features (AFs) of hepatic observations. The MFs of HCC include arterial-phase hyperenhancement (APHE), nonperipheral “washout”, enhancing “capsule” and threshold growth in addition to size. The LI-RADS algorithm uses a combination of MFs to assign initial categories and then adjusts the categories according to AFs [4].

Among the MFs, the contribution of threshold growth to the diagnosis is the most controversial. Chernyak et al. [5] suggested that removing threshold growth as a MF would cause a nonnegligible proportion of LR-5 observations to be downgraded to LR-4. However, Park et al. [6] reported that threshold growth was not a significant diagnostic indicator of HCC. In particular, LI-RADS version 2018 (LI-RADS v2018) simplified the definition of threshold growth to “size increase of a mass by ≥ 50% in ≤ 6 months”, which is based only on expert consensus and consistent with the rules of the Organ Procurement and Transplantation Network (OPTN) in the United States [7]. Therefore, threshold growth as a MF of HCC lacks strong support, particularly since non-HCC malignancies and some benign lesions, such as abscesses, can also achieve the threshold growth defined by LI-RADS, limiting the value of threshold growth in distinguishing HCCs from other lesions. So, whether removing threshold growth as a MF will affect the diagnostic performance of LR-5 for HCC and whether some specific AFs can replace threshold growth will be further researched. In this paper, the above contents will be studied.

Study population

This study was approved by the Ethics Committee of Tianjin Third Central Hospital. Informed consent was waived due to the retrospective study design. Using electronic medical records, data from patients with a high risk of HCC [8] between June 2016 and June 2021 were retrospectively collected. The inclusion criteria were as follows: 1) ≥ 18 years old; 2) received EOB-MRI; and 3) underwent surgery or biopsy pathology. Patients meeting the following criteria were excluded from the study: 1) cirrhosis due to congenital hepatic fibrosis or vascular disorder; 2) interval of more than 1 month between pathological diagnosis and EOB-MRI; 3) diffuse hepatic lesions or nodules > 3; 4) incomplete EOB-MRI or had images that could not be evaluated due to poor quality; 5) received treatment prior to the EOB-MRI; or 6) LR-TIV (tumour-in-vein) (Fig. 1). Patient liver histology, including biopsy and surgical pathology, served as the standard reference.

Clinicopathologic characteristics of patients including age, sex, numbers of cirrhosis, and the causes of liver disease were recorded through the patient records system. A board-certified radiologist with 4 years (J.W.) of experience in abdominal MRI reviewed all images and recorded the number of lesions.

MRI techniques

For all examinations, studies were carried out by using a 3.0T MR system (Magnetom Verio, Siemens Healthcare, Erlangen, Germany), and an 8-channel phased-array torso coil was used for all measurements. The liver MR imaging protocol consisted of in- and out-of-phase T1-weighted imaging acquired with a gradient recalled echo (GRE) sequence, respiratory-triggered axial T2-weighted turbo spin echo (TSE) sequence with fat suppression, free-breathing single-shot echo-planar diffusion-weighted imaging (DWI) with b values of 0, 50, 600, 1000s/mm², and pre- and postcontrast T1-weighted three-dimensional volumetric interpolated breath-hold examination (VIBE) sequences acquired with a GRE sequence in the arterial phase (25 seconds after aortic enhancement using the bolus tracking method), portal venous phase (46–60 seconds), transitional phase (150–180 seconds), and hepatobiliary phase (20 minutes after the contrast injection). Contrast-enhanced dynamic MR images of the liver were obtained after intravenous administration of gadoxetic acid (Primovist; Bayer Healthcare) administered at 0.025 mmol/kg of body weight at a rate of 1.0 ml/s by using a power injector, followed by 25 ml of 0.9% saline chaser at the same rate. The detailed parameters of each acquisition sequence are shown in Table 1.

Table 1

MR imaging protocol
Sequence	TR (ms)	TE (ms)	FA	ST (mm)	SS (mm)	FOV (mm)	Scan matrix
In-phase T1WI	250	2.5	65°	6	1.5	380×285	512×300
Out-of-phase T1WI	250	3.8	60°	6	1.5	380×285	512×300
T2WI	3000–4000	91	90°	6	1.5	380×285	320×180
DWI	6000–7000	68	90°	6	1.5	400×300	128×78
VIBE	4.19	1.47	9°	4	0	400×300	320×168

TR, repetition time; TE, echo time; FA, flip angle; ST, slice thickness; SS, slice spacing; FOV, field of view; T1WI, T1-weighted imaging; T2WI, T2-weighted imaging; DWI, diffusion-weighted imaging; VIBE, volumetric interpolated breath-hold examination.

MRI analysis

All MRI scans were independently reviewed by two board-certified radiologists with 10 years (W.J.H.) and 6 years (D.W.) of experience in abdominal MRI, respectively. All MFs and AFs for each liver observation according to the CT/MRI LI-RADS v2018 were reviewed. All readers were blinded to the pathologic results. Discrepancies between the two readers were resolved by a third radiologist (R.L. with 16 years of experience in abdominal MRI) to reach a final consensus reading. For observations with documented threshold growth, readers determined the qualifying threshold growth criteria (≥ 50% diameter increase in ≤ 6 months) by retrospectively reviewing, measuring and comparing the longest diameter of observation on current and prior exams. The previous imaging studies included EOB-MRI and contrast-enhanced CT (CECT), and no treatment was included between the two examinations. Any phase of MRI except the early arterial phase and DWI were selected to measure the longest diameter on the plane with the largest cross-sectional area of the lesion, and the lesion size was the average of the measurement results of 2 readers.

Comparison of features and size of observations

According to the pathological results, the observations were divided into three groups: HCC, non-HCC malignancy and benign lesion. All major and ancillary features among the three groups were compared, and highly frequent AFs (statistically significant AFs) in HCCs were determined by univariable analysis. The sizes of observations with and without threshold growth were compared in each group.

Category assignment

The LR-1, LR-2 and LR-M observations were assigned according to the LI-RADS v2018 algorithm. Each LR-3 ~ 5 observation was assigned as illustrated in Fig. 2. In scheme A, the categories were assigned using all MFs and AFs according to LI-RADS v2018. In scheme B, the categories were assigned using all MFs except threshold growth, and threshold growth was treated as an AF favouring malignancy, thus, it could upgrade the category up to but not beyond LR-4. In scheme C, the categories were assigned using all MFs except threshold growth, the highly frequent AFs in HCC were considered new MFs, and threshold growth was treated as an AF favouring malignancy. Using pathological results as the gold standard reference, the comparison of diagnostic performance of LR-5 for HCC was assessed among the above 3 schemes, and pairwise comparisons of the distributions of categories assigned using scheme A vs. scheme B or scheme C.

Statistical analysis

Categorical variables are summarized as counts and percentages of patients/observations in corresponding categories. Pearson χ² test, continuity correction test or Fisher’s exact test were used to determine significant differences between groups. Continuous variables are summarized as means and standard deviations and were tested against the null hypothesis by means of Student’s t test. The median (interquartile range) was used for nonnormally distributed data. Diagnostic performance is reported as the sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), accuracy, and Youden index. Diagnostic performance was compared using McNemar’s test. Unless otherwise indicated, all statistical tests were conducted at the 0.05 significance level using 2-tailed tests, and p values are reported. Statistical analyses were performed using SPSS software, version 25.0 (SPSS Inc.).

Patient characteristics and pathologic findings

The study population included 379 patients (265 males and 114 females; mean age, 58 ± 10 years) with 426 observations. There were 314 (82.8%) patients with cirrhosis (diagnosed by typical imaging findings or biopsy). The causes of liver disease were hepatitis B virus (301 patients), hepatitis C virus (38 patients), alcohol (9 patients), autoimmune (4 patients), nonalcoholic steatohepatitis (1 patient), both hepatitis B and C virus (6 patients) and cirrhosis of unknown cause (20 patients). A total of 341 patients had one single observation, 29 patients had two observations, and 9 patients had three observations. Among the 426 observations (median size, 27 mm), there were 250 HCCs, 88 non-HCC malignancies and 88 benign lesions. Surgery was performed for a total of 199 (46.7%) lesions (113 HCCs, 52 non-HCC malignancies, 34 benign lesions), and 227 (53.3%) lesions were confirmed by biopsy (137 HCCs, 36 non-HCC malignancies, 54 benign lesions). The clinicopathologic characteristics of the patients and hepatic observations are shown in Table 2.

Table 2

Clinicopathologic characteristics of patients and hepatic observations
Characteristic	Value (%)
Patient (n = 379)
Mean age (year)^*	58.1 ± 10.4
Sex (%)
Male	265 (69.9)
Female	114 (30.1)
Cirrhosis (%)	314 (82.8%)
Cause of liver disease (%)
Hepatitis B virus	301 (79.4)
Hepatitis C virus	38 (10.0)
Alcohol	9 (2.4)
Autoimmune	4 (1.1)
NASH	1 (0.3)
Hepatitis B and C virus	6 (1.6)
Cirrhosis of unknown cause	20 (5.3)
Number of observations per patient (%)
1	341 (90.0)
2	29 (7.7)
3	9 (2.4)
Observation (n = 426)
Size (mm)
Over all ⁺	27 (16–52)
HCC^*	34.7 ± 27.3
Non-HCC malignancy⁺	45 (27–66)
Benign lesion⁺	22 (14–44)
Final diagnosis (%)
HCC [Surgery/biopsy]	250 (58.7) [113/137]
Non-HCC malignancy [Surgery/biopsy]	88 (20.6) [52/36]
ICC	41 (46.6)
CHC	13 (14.8)
Metastasis	23 (26.1)
Sarcomatoid carcinoma	6 (6.8)
Cystadenocarcinoma	3 (3.4)
Neuroendocrine carcinoma	1 (1.1)
Haemangiosarcoma	1 (1.1)
Benign lesion [Surgery/biopsy]	88 (20.7) [34/54]
Haemangioma	16 (18.2)
DN	34 (38.7)
RN	11 (12.5)
FNH	9 (10.2)
Adenoma or adenomatoid hyperplasia	6 (6.8)
Abscess	4 (4.5)
Angiomyolipoma	2 (2.3)
Epithelioid angiomyolipoma	3 (3.4)
Inflammatory nodule	1 (1.1)
Lipomyoma	1 (1.1)
Ematoma	1 (1.1)

Unless stated otherwise, data are presented as the number of patients or observations with the percentage in parentheses. ^* Data are presented as the mean ± standard deviation; ⁺ Data are presented as the median (interquartile range). NASH, nonalcoholic steatohepatitis; HCC, hepatocellular carcinoma; ICC, intrahepatic cholangiocarcinoma; CHC, combined hepatocellular cholangiocarcinoma; DN, dysplastic nodule; RN, regenerative nodule; FNH, focal nodular hyperplasia.

Comparison of major and ancillary features

Among the MFs, APHE, “washout”, and enhancing “capsule” were significantly more frequently observed in HCCs than in non-HCC malignancies or benign lesions, while the targetoid features were more common in non-HCC malignancies than HCCs (all p < 0.001). Sixty-five of 250 HCCs, 37 of 88 non-HCC malignancies and 32 of 88 benign lesions with previous EOB-MR/CECT within 6 months without treatment (56 EOB-MRI, 78 CECT) were used to determine the change in size. Among those observations, threshold growth was observed in 32 HCCs (Fig. 3) (49.2%), 16 non-HCC malignancies (Fig. 4) (43.2%) and 9 benign lesions (7 dysplastic nodules (DNs), 2 abscesses). There was no significant difference in threshold growth between HCCs and non-HCC malignancies (p = 0.560), but there was a significant difference between HCCs and benign lesions (p = 0.048) (Table 3).

Among the AFs, the nodule-in-nodule appearance was more likely to be present in HCCs than in non-HCC malignancies (p = 0.040) and benign lesions (p = 0.040). Corona enhancement, restricted diffusion, mild-moderate T2 hyperintensity, iron sparing in the solid mass, transitional phase (TP) hypointensity, hepatobiliary phase (HBP) hypointensity and blood products in mass were more likely to be observed in HCCs than in benign lesions (all p < 0.05), while parallel blood pool, marked T2 hyperintensity and hepatobiliary phase isointensity were more likely to be observed in benign lesions than in HCCs (all p < 0.001). The fat-in-mass feature was more likely to be present in HCCs than in non-HCC malignancies (p = 0.027), while cholangiectasis was more likely to be present in non-HCC malignancies than in HCCs (p = 0.024) (Table 3).

Table 3

༎Major and ancillary features in HCCs, non-HCC malignancies and benign lesions
	HCC (n = 250)	Non-HCC malignancy (n = 88)	Benign lesion (n = 88)	p₁ value^*	p₂ value^*
Major features
APHE	220 (88.0)	43 (48.9)	38 (43.2)	< 0.001	< 0.001
Washout	193 (77.2)	14 (15.9)	19 (21.6)	< 0.001	< 0.001
Enhancing capsule	117 (46.8)	7 (8.0)	10 (11.4)	< 0.001	< 0.001
Threshold growth ^a	32/65 (49.2)	16/37 (43.2)	9/32 (28.1)	0.560	0.048
Targetoid features	11 (4.4)	46 (52.3)	4 (4.5)	< 0.001	0.955
Ancillary features
Subthreshold growth^b	37/62 (59.7)	7/13 (53.8)	8/17 (47.1)	0.698	0.352
Corona enhancement	17 (6.8)	10 (11.4)	1 (1.1)	0.174	0.042
Fat sparing in solid mass	4 (1.6)	2 (2.3)	0 (0)	> 0.999	> 0.999
Restricted diffusion	229 (91.6)	82 (93.2)	33 (37.5)	0.638	< 0.001
Mild-moderate T2 hyperintensity	232 (92.8)	84(95.5)	19(21.6)	0.385	< 0.001
Iron sparing in solid mass	0 (0)	0 (0)	0 (0)	-	-
TP hypointensity	222 (88.8)	76 (86.4)	41 (46.6)	0.543	< 0.001
HBP hypointensity	236 (94.4)	80 (90.9)	62 (72.7)	0.254	< 0.001
No-nenhancing capsule	8 (3.2)	2 (2.3)	1 (1.1)	0.940	0.301
Nodule-in-nodule	15 (6.0)	0 (0)	0 (0)	0.040	0.040
Mosaic architecture	12 (4.8)	2 (2.3)	0 (0)	0.476	0.079
Fat-in-mass, more than adjacent liver	19 (7.6)	1 (1.1)	4 (4.5)	0.027	0.328
Blood products in mass	18 (7.2)	2 (2.3)	0 (0)	0.092	0.021
Necrosis	7 (2.8)	5 (5.7)	1 (1.1)	0.357	0.635
Cholangiectasis	1 (0.4)	4 (4.5)	0 (0)	0.024	> 0.999
Parallels blood pool	0 (0)	0 (0)	16 (18.2)	-	< 0.001
Marked T2 hyperintensity	0 (0)	1 (1.1)	24 (27.3)	0.260	< 0.001
Hepatobiliary phase isointensity	14 (5.6)	8 (9.1)	26 (29.6)	0.254	< 0.001

Unless stated otherwise, data are presented as the number of observations with the percentage in parentheses. ^* Calculated using Pearson’s chi-square test, the continuity correction or Fisher’s test; p₁ values are for HCC vs. non-HCC malignancy, and p₂ values are for HCC vs. benign lesion. ^a The data refer to 65 HCCs, 37 non-HCC malignancies and 32 benign lesions that were examined by EOB-MRI/CECT within the previous 6 months and for which threshold growth could be measured; ^b Data refer to 62 HCCs, 13 non-HCC malignancies and 17 benign lesions that were examined by EOB-MRI/CECT within the previous 1 or 2 years and for which the subthreshold growth could be measured. HCC, hepatocellular carcinoma; APHE, Arterial-phase hyperenhancement; TP, transitional phase; HBP, hepatobiliary phase.

Observation size with and without threshold growth

Whether the lesion was HCC, non-HCC malignancy, or benign, the mean sizes of the observed masses with threshold growth were smaller than those of the masses without threshold growth (14.8 vs. 22.6 mm, 22.1 vs. 32.6 mm, 9.7 vs. 17.8 mm, all p < 0.05). The comparison of observation sizes with and without threshold growth are shown in Table 4.

Table 4

Comparison of observation size with and without threshold growth
Observation	With threshold growth (mm)			Without threshold growth (mm)			p value
Observation	Number	Mean size	Range	Number	Mean size	Range	p value
HCC (n = 65)	32	14.8 ± 10.4	5–49	33	22.6 ± 10.8	10–55	0.003
Non-HCC malignancy (n = 37)	16	22.1 ± 16.7	5–70	21	32.6 ± 12.5	15–61	0.036
Benign lesion (n = 32)	9	9.7 ± 4.0	5–19	23	17.8 ± 9.9	8–52	0.026

HCC, hepatocellular carcinoma.

Diagnostic performance

Based on Table 3, nodule-in-nodule architecture was highly frequent AF and was selected instead of threshold growth as a MF of HCC. Based on the criteria for schemes A-C presented in Fig. 2, the observations were assigned to LR categories. There were 36 LR-4 and 185 LR-5 lesions based on the criteria for scheme A. There was one HCC downgraded from LR-5 to LR-4 based on scheme B, and then 4 HCCs upgraded from LR-4 to LR-5 based on scheme C. The results of the categorization of these observations are summarized in Table 5.

Table 5

The categories of observations in schemes A ~ C
Category	Scheme A				Scheme B				Scheme C
Category	HCC	Non-HCC malignancy	Benign lesion	Total	HCC	Non-HCC malignancy	Benign lesion	Total	HCC	Non-HCC malignancy	Benign lesion	Total
LR-1	0	0	20	20	0	0	20	20	0	0	20	20
LR-2	3	2	11	16	3	2	11	16	3	2	11	16
LR-3	18	2	28	48	18	2	28	48	18	2	28	48
LR-4	35	5	16	56	36	5	16	57	32	5	16	53
LR-5	186	11	9	206	185	11	9	205	189	11	9	209
LR-M	8	68	4	80	8	68	4	80	8	68	4	80
Total	250	88	88	426	250	88	88	426	250	88	88	426

Scheme A: The categories were assigned using all MFs and AFs according to the LI-RADS v2018. Scheme B: All MFs except threshold growth were used, and threshold growth was treated as an AF favouring malignancy; thus, it could upgrade the category up to but not beyond LR-4. Scheme C: All MFs except threshold growth were used, nodule-in-nodule architecture was considered a new MF, and threshold growth was treated as an AF favouring malignancy.

The sensitivity and accuracy of LR-5 for HCC with scheme C were slightly higher than those with scheme A (75.6% vs. 74.4%; 81.0% vs. 80.3%), although there was no statistically significant difference (all p > 0.05). The Youden index of scheme C was also slightly higher than that of scheme A (0.642 vs. 0.630). There was no significant difference in the diagnostic performance of LR-5 for HCC between schemes A and B (all p > 0.05). The diagnostic performance is shown in Table 6.

Table 6

Diagnostic performance of LR-5 for HCC with schemes A ~ C
	Sensitivity	Specificity	PPV	NPV	Accuracy	Youden
Scheme A	74.4%	88.6%	90.3%	70.9%	80.3%	0.630
Scheme B	74.0%	88.6%	90.2%	70.6%	80.0%	0.540
Scheme C	75.6%	88.6%	90.4%	71.9%	81.0%	0.642
P_a value (A/B)	> 0.999	> 0.999	—	—	> 0.999	—
P_b value (A/C)	0.375	> 0.999	—	—	0.375	—

Schemes A ~ C are the same as those in Table 5. P_a values are for comparisons of diagnostic performance between scheme A and scheme B using McNemar’s test, and P_b values are for comparisons of diagnostic performance between scheme A and scheme C using McNemar’s test. PPV, positive predictive value; NPV, negative predictive value.

Threshold growth in the LI-RADS is controversial in the diagnostic performance of HCC. In this study, there was no statistically significant difference in the frequency of threshold growth between HCCs and non-HCC malignancies, and certain benign lesions may also show threshold growth. The diagnostic performance of HCC was not affected either by removing threshold growth or by replacing threshold growth with a highly frequent AF.

Among the MFs, APHE and “washout” showed high sensitivity for progressed HCC as the arterial blood supply increased and the portal supply decreased in HCC [9, 10]. In addition, although the sensitivity of enhancing “capsule” was lower than that of APHE and “washout” reported in previous literature, the specificity was higher because it reflects the pseudocapsule composed of compressed fibrous tissue and dilated sinusoids around the HCC [11, 12]. The above three MFs have a definite pathological basis and are recognized by many HCC diagnosis and treatment guidelines [13, 14]. In our study, compared with non-HCC malignancies or benign lesions, HCCs showed significant differences in the APHE, “washout” and enhancing “capsule”, which is consistent with previous literature reports. However, there is no strong pathological support for threshold growth of HCC. In a previous multivariable analysis of CT/MRI LI-RADS, all MFs, except threshold growth, were associated with HCC (OR, 1.6; 95% CI: 0.7, 3.6; P = 0.07) [15]. In our study, there was no significant difference in threshold growth between HCCs and non-HCC malignancies. In general, malignant tumours tend to grow faster and more easily show threshold growth, which means that threshold growth can reflect the degree of the malignancy of lesions but cannot define the pathological types of lesions. Additionally, it is not a specific feature of HCC and appears in lesions from other malignancies. It has been reported that the tumour doubling times of HCC and intrahepatic cholangiocarcinoma (ICC) overlap [16, 17]. In some benign lesions, such as abscesses, threshold growth in LI-RADS can also occur. In our study, there were 2 abscesses with threshold growth. Because the vasculature of some high-grade DNs is similar to that of early HCC, their features tend to be very similar [18], threshold growth also sometimes occurs in DNs. In this study, there were 7 DNs with threshold growth. Overall, these observations indicate that threshold growth has limited value in differentiating HCC from non-HCC malignancies and certain benign lesions.

According to the definition of threshold growth (size increase of a mass by ≥ 50% in ≤ 6 months), when the lesion is small, it is easier to achieve threshold growth. In other words, the initial tumour size can affect the likelihood of threshold growth. In our study, whether the lesion was HCC, non-HCC malignancy, or benign, those with threshold growth were smaller than those without threshold growth (all p < 0.05). It has been previously reported that smaller HCCs increase in size faster than larger HCCs [16, 19]; that is, small HCCs demonstrate shorter doubling times [20] and are more likely to have threshold growth. Our conclusions are consistent with this.

Among the AFs, some specific AFs show good performance in the diagnosis of HCC [21, 22].The nodule-in-nodule feature is a typical manifestation of tumour heterogeneity, and it is a variant of the “mosaic” architecture of HCC [23]. The development of most HCCs in patients with cirrhosis occur via a multistep process of carcinogenesis, from a regenerative nodule (RN) to a DN and finally to HCC [24]. The components of different stages are prone to coexisting in one nodule. Therefore, in LI-RADS v2018, the nodule-in-nodule feature is an AF favouring HCC in particular. Compared with studies from other groups, the nodule-in-nodule feature was more common in HCCs in our study (all p < 0.05). Therefore, it was selected as a highly frequent AF to replace threshold growth as a new MF of HCC.

When the categories were assigned using all MFs except threshold growth and when threshold growth was treated as an AF favouring malignancy, the sensitivity was slightly but non-significantly lower than that of LI-RADS v2018 (74.0% vs. 74.4%), while the specificity remained unchanged. That is, when the threshold growth was removed from the MFs, the diagnostic performance for HCC did not change. When the categories were assigned using all MFs except threshold growth and the nodule-in-nodule characteristic was considered a new MF, the sensitivity and accuracy for HCC were slightly but non-significantly higher than those of LI-RADS v2018 (all p > 0.05), and the specificity remained unchanged. That is, regardless of whether the threshold growth was removed from among the MFs or replaced by other highly frequent AFs, the final diagnostic performance of HCC remained unaffected.

Chernyak et al. [5] reported that when threshold growth is removed, 9% of LR-5 observations will be downgraded to LR-4, which will affect the management of fast-growing LR-5 lesions and cause unnecessary biopsy. However, the aforementioned study was based on LR-RADS v2014, and the concept of threshold growth at that time also included two other items, namely, new observations that are ≥ 10 mm at two years and show a ≥ 100% diameter increase at > 6 months. Therefore, the proportion of lesions with threshold growth was as high as 66.4% in that report. Obviously, this data cannot represent the actual situation after the simplified threshold growth definition of the LI-RADS v2018, and the relatively large number of patients with prior exams may have inflated the importance of threshold growth. In contrast, according to the definition of threshold growth in LI-RADS v2018, only 28.4% of patients that had prior exams and measurements of threshold growth in our study. In addition, the article emphasizes only that removing threshold growth will reduce the number of LR-5 classifications but does not verify the diagnostic performance and cannot explain whether there are other non-HCC lesions misclassified as LR-5 due to the threshold growth.

Another study reported that threshold growth was not a significant diagnostic indicator of HCC and was more common in non-HCC malignancies [6]. The conclusion of this paper was similar to that of our study. Therefore, this calls into question whether threshold growth should be regarded as an HCC-specific MF to guide the assignment of the category of observations, and this remains an unaddressed gap in knowledge. In addition, unlike other MFs, the measurement of threshold growth cannot be done for all observations; specifically, the number of observations for which threshold growth can be measured should be far lower than the total number of observations. In our study, the percentage of lesions that had prior exams and measurements of threshold growth was only 28.4% (121/426); this percentage is lower than the 43.2% reported by Park et al. [6]. It is also worth noting that threshold growth cannot be judged because many of the patients with lesions in our institution were actively treated when the lesions were first discovered and therefore did not meet the criterion for no treatment between the two examinations.

This study has several limitations: 1) There may have been selection bias due to the retrospective study design, the inclusion of hepatic observations confirmed by histopathology and exclusion of HCC confirmed by typical imaging findings. 2) Fewer LR-1 and LR-2 nodules were included in this study due to the lack of pathological diagnosis. 3) The sample size of patients with observations with threshold growth was small because there were few patiens that meet the conditions of no treatment between the two examinations.

In conclusion, threshold growth can occur in both HCC, non-HCC malignancies and certain benign lesions. The sizes of lesions with threshold growth are smaller than those without threshold growth. The diagnostic performance of LR-5 for HCC was not affected after removing the threshold growth or replacing threshold growth with the nodule-in-nodule architecture as the MF. In summary, LI-RADS cannot be used as the single solo diagnostic approach. The purpose of the ACR’s development of this algorithm is to convey confidence in the diagnosis of lesions in patients with a high risk of HCC [25]. Our research suggests that the algorithm can be further improved and made more accurate if it uses all kinds of features.

LI-RADS

Liver Imaging Reporting and Data System

LI-RADS v2018

LI-RADS version 2018

ACR

American College of Radiology

OPTN

Organ Procurement and Transplantation Network

MFs

major features

AFs

ancillary features

HCC

Hepatocellular carcinoma

Computed tomography

CECT

Contrast-enhanced CT

MRI

Magnetic resonance imaging

EOB-MRI

Gadoxetic acid–enhanced MRI

DWI

diffusion-weighted imaging

GRE

gradient recalled echo

TSE

turbo spin echo

VIBE

Volumetric interpolated breath-hold examination

repetition time

echo time

flip angle

slice thickness

slice spacing

FOV

field of view

T1WI

T1-weighted imaging

T2WI

T2-weighted imaging

APHE

Arterial-phase hyperenhancement

TIV

Tumour-in-vein

NASH

Nonalcoholic steatohepatitis

ICC

Intrahepatic cholangiocarcinoma

CHC

Combined hepatocellular cholangiocarcinoma

Dysplastic nodule

Regenerative nodule

FNH

Focal nodular hyperplasia

Transitional phase

HBP

Hepatobiliary phase

PPV

Positive predictive value

NPV

Negative predictive value.

Ethics approval and consent to participate

This study was conducted in accordance with the ethical guidelines of the Helsinki Declaration of the World Medical Association and approved by the Ethics Committee of Tianjin Third Central Hospital (IRB2022-039-01). The need for written informed consent was waived by the Ethics Committee of Tianjin Third Central Hospital (IRB2022-039-01) due to the retrospective design of the study.

Consent for publication

Not applicable.

Availability of data and materials

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Funding

This study was supported by the grants form Tianjin Key Medical Discipline (Specialty) Construction Project (TJYXZDXK-074C), Tianjin Health Science and Technology Project (TJWJ2022QN045), and Tianjin Health Project of Science and Technology (TJWJ2022XK027).

Authors' contributions

Rong Lyu and Di Wang contributed equally to this work.

conception (Rong Lyu, Di Wang), literature research (Rong Lyu, Di Wang, Weijuan Hu,), data curation (Changlu Yu, Zhongsong Gao), imaging analysis (Rong Lyu, Di Wang, Weijuan Hu, Zhongsong Gao), manuscript drafting (Jiao Wang, Mingge Li), manuscript editing (Rong Lyu, Di Wang), final manuscript approval (Rong Lyu, Kefeng Jia). All authors read and approved the final manuscript.

Acknowledgements

Not applicable.

Authors' information

Rong Lyu and Di Wang contributed equally to this work.

* Corresponding author: Kefeng Jia, [email protected]

Department of Radiology, Tianjin Third Central Hospital, Tianjin Key Laboratory of Extracorporeal Life Support for Critical Diseases, Artificial Cell Engineering Technology Research Center, Tianjin Institute of Hepatobiliary Disease, NO.83 Jintang Road, Hedong District, Tianjin 300170, China.

Kim TH, Kim SY, Tang A, Lee JM. Comparison of international guidelines for noninvasive diagnosis of hepatocellular carcinoma 2018 update. Clin Mol Hepatol. 2019;25(3):245–63.
Kim YY, Kim MJ, Kim EH, Roh YH, An C. Hepatocellular carcinoma versus other hepatic malignancy in cirrhosis: performance of LI-RADS version 2018. Radiology. 2019;291(1):72–80.
Cunha GM, Chernyak V, Fowler KJ, Sirlin CB. Up-to-date role of CT/MRI LI-RADS in hepatocellular carcinoma. J Hepatocell Carcinoma. 2021;31(8):513–27.
American College of Radiology. CT/MRI LI-RADS® v2018.Acr.org Web site. 2020. https://www.acr.org/Clinical-Resources/Reporting-and-Data-Systems/LI-RADS/CT-MRI-LI-RADS-v2018.Accessed 6 November 2020.
Chernyak V, Kobi M, Flusberg M, Fruitman KC, Sirlin CB. Effect of threshold growth as a major feature on LI-RADS categorization. Abdom Radiol (NY). 2017;42(8):2089–100.
Park J, Chung YE, Seo N, Choi JY, Park MS, Kim MJ. Should threshold growth be considered a major feature in the diagnosis of hepatocellular carcinoma using LI-RADS? Korean J Radiol. 2021;22(10):1628–39.
Rosenkrantz AB, Campbell N, Wehrli N, Triolo MJ, Kim S. New OPTN/UNOS classification system for nodules in cirrhotic livers detected with MR imaging: effect on hepatocellular carcinoma detection and transplantation allocation. Radiology. 2015;274(2):426–33.
Cho MJ, An C, Aljoqiman KS, Choi JY, Lim JS, Park MS, et al. Diagnostic performance of liver imaging reporting and data system in patients at risk of both hepatocellular carcinoma and metastasis. Abdom Radiol. 2020;45(11):3789–99.
Kim TK, Lee KH, Jang HJ, Haider MA, Jacks LM, Menezes RJ, et al. Analysis of gadobenate dimeglumine-enhanced MR findings for characterizing small (1-2-cm) hepatic nodules in patients at high risk for hepatocellular carcinoma. Radiology. 2011;259(3):730–8.
Shin J, Lee S, Yoon JK, Chung YE, Choi JY, Park MS. LI-RADS Major Features on MRI for Diagnosing Hepatocellular Carcinoma: A Systematic Review and Meta-Analysis. J Magn Reson Imaging. 2021;54(4):518–25.
Lee SY, Kim MJ, Kim SS, Shin H, Kim DY, Choi JY, Park MS, et al. Retrospective comparison of EASL 2018 and LI-RADS 2018 for the noninvasive diagnosis of hepatocellular carcinoma using magnetic resonance imaging. Hepatol Int. 2020;14(1):70–9.
Ishigami K, Yoshimitsu K, Nishihara Y, Irie H, Asayama Y, Tajima T, et al. Hepatocellular carcinoma with a pseudocapsule on gadolinium-enhanced MR images: correlation with histopathologic findings. Radiology. 2009;250(2):435–43.
Korean Liver Cancer Association (KLCA), National Cancer Center (NCC), Goyang K. 2018 korean liver cancer association-national cancer center korea practice guidelines for the management of hepatocellular carcinoma. Korean J Radiol. 2019;20(7):1042–113.
Marrero JA, Kulik LM, Sirlin CB, Zhu AX, Finn RS, Abecassis MM, et al. Diagnosis, staging, and management of hepatocellular carcinoma: 2018 practice guidance by the american association for the atudy of liver diseases. Hepatology. 2018;68(2):723–50.
Pol CBVD, McInnes MDF, Salameh JP, Levis B, Chernyak V, Sirlin CB, et al. CT/MRI and CEUS LI-RADS Major Features Association with Hepatocellular Carcinoma: Individual Patient Data Meta-Analysis. Radiology. 2022;302(2):326–35.
Kim JK, Kim HD, Jun MJ, Yun SC, Shim JH, Lee HC, et al. Tumor volume doubling time as a dynamic prognostic marker for patients with hepatocellular carcinoma. Dig Dis Sci. 2017;62(10):2923–31.
An C, Choi YA, Choi D, Paik YH, Ahn SH, Kim MJ, et al. Growth rate of early-stage hepatocellular carcinoma in patients with chronic liver disease. Clin Mol Hepatol. 2015;21(3):279–86.
Zhong X, Tang H, Lu B, You J, Piao J, Yang P, et al. Differentiation of Small Hepatocellular Carcinoma From Dysplastic Nodules in Cirrhotic Liver: Texture Analysis Based on MRI Improved Performance in Comparison Over Gadoxetic Acid-Enhanced MR and Diffusion-Weighted Imaging. Front Oncol. 2020;10(9):1382.
Nathani P, Gopal P, Rich N, Yopp A, Yokoo T, John B, et al. Hepatocellular carcinoma tumour volume doubling time: a systematic review and meta-analysis. Gut. 2021;70(2):401–7.
Kubota K, Ina H, Okada Y, Irie T. Growth rate of primary single hepatocellular carcinoma: determining optimal screening interval with contrast enhanced computed tomography. Dig Dis Sci. 2003;48(3):581–6.
Li Y, Chen J, Weng S, Yan C, Ye R, Zhu Y, et al. Hepatobiliary phase hypointensity on gadobenate dimeglumine-enhanced magnetic resonance imaging may improve the diagnosis of hepatocellular carcinoma. Ann Transl Med. 2021;9(1):55.
Cerny M, Bergeron C, Billiard JS, Murphy-Lavallée J, Olivié D, Bérubé J. LI-RADS for MR Imaging Diagnosis of Hepatocellular Carcinoma: Performance of Major and Ancillary Features. Radiology. 2018;288(1):118–28.
Giambelluca D, Cannella R, Caruana G, Brancatelli G. “Nodule in nodule” architecture of hepatocellular carcinoma. Abdom Radiol (NY). 2019;44(7):2671–3.
Duan Y, Xie XY, Li Q, Mercaldo N, Samir AE, Kuang M, et al. Differentiation of regenerative nodule, dysplastic nodule, and small hepatocellular carcinoma in cirrhotic patients: a contrast-enhanced ultrasound-based multivariable model analysis. Eur Radiol. 2020;30(9):4741–51.
Cunha GM, Fowler KJ, Roudenko A, Taouli B, Fung AW, Elsayes KM, et al. How to use LI-RADS to report liver CT and MRI observations. Radiographics. 2021;41(5):1352–67.

No competing interests reported.

Download PDF

Journal Publication

published 04 Dec, 2023

Read the published version in BMC Medical Imaging →

Editorial decision: Major revision
07 Aug, 2023
Reviews received at journal
03 Aug, 2023
Reviews received at journal
29 Jul, 2023
Reviewers agreed at journal
21 Jul, 2023
Reviewers invited by journal
06 May, 2023
Editor invited by journal
02 Mar, 2023
Editor assigned by journal
24 Feb, 2023
Submission checks completed at journal
24 Feb, 2023
First submitted to journal
17 Feb, 2023

You are reading this latest preprint version

Threshold growth has a limited role in differentiating HCC from other focal hepatic lesions

Status:

Journal Publication

Version 1

Abstract

Background

Methods

Results

Conclusion

Figures

Introduction

Materials And Methods

Study population

MRI techniques

MRI analysis

Comparison of features and size of observations

Category assignment

Statistical analysis

Results

Patient characteristics and pathologic findings

Comparison of major and ancillary features

Observation size with and without threshold growth

Diagnostic performance

Discussion

Abbreviations

Declarations

References

Additional Declarations

Status:

Journal Publication

Version 1