Contrast-enhanced ultrasound features of hepatic sarcomatoid carcinoma different from hepatocellular carcinoma

Abstract

BACKGROUND:

Hepatic sarcomatoid carcinoma (HSC) is a rare malignancy of the liver. The ultrasound and clinical features of HSC have not been determined.

OBJECTIVE:

To investigate and compare the ultrasound and clinical features of HSC and hepatocellular carcinoma (HCC), and to reveal the valuable features of HSC.

METHODS:

The ultrasound features and clinical data of pathologically proven HSC (n = 37) were compared with HCC (n = 92) in a matching ratio of 1:4 using the propensity score (age, gender and tumor size).

RESULTS:

The HSC patients were more likely to accompany with clinical symptoms and vascular invasion than HCC patients (40.5% vs 17.4%, 24.3% vs 6.5%, P < 0.05). The size of HSCs was significantly larger than that of HCCs (P < 0.05). The proportion of patients with elevated alpha-fetoprotein was significantly lower in HSC (35.1% vs 54.3%, P < 0.05). On gray-scale ultrasound images, the HSCs were more likely to demonstrate as indistinct margin and irregular shape lesions compared to HCCs (78.4% vs 48.8%; 70.3% vs 23.9%, P < 0.05). Under color Doppler flow imaging (CDFI), the blood flow signals were more frequently detected in HSC lesions (75.7% vs 56.5%, P < 0.05). Resistance index (RI) was higher in HSCs than in HCCs [0.78 (0.70,0.82) vs 0.70 (0.62,0.76), P < 0.05]. On contrast-enhanced ultrasound (CEUS), HSCs mainly showed entirety heterogeneous hyper-enhancement (48.6%), entirety homogeneous enhancement (18.9%), peripheral and internal septal enhancement (18.9%). The incidence of non-enhanced areas inside HSC lesions was higher than that inside HCC lesions (56.8% vs 31.5%, P < 0.05). During the portal venous and late phases, most of the lesions revealed hypo-enhancement in both groups, whereas earlier washout was observed in HSCs [43.0 s (30.5,58.0) vs 60.0 s (46.3,100.0), P < 0.05].

CONCLUSIONS:

CEUS features are useful in preoperative and non-invasive differentiation of hepatic sarcomatoid carcinoma and hepatocellular carcinoma.

Keywords

Hepatic sarcomatoid carcinoma (HSC)hepatocellular carcinoma (HCC)gray-scale ultrasound color Doppler flow imaging (CDFI)contrast-enhanced ultrasound (CEUS)

1 Introduction

Hepatic sarcomatoid carcinoma (HSC) is a rare malignancy of the liver that consists of both carcinomatous and sarcomatous components. The carcinomatous components can be hepatocellular carcinoma (HCC) and intrahepatic cholangiocarcinoma (ICC). In the World Health Organization classification of digestive system tumors, sarcomatous change in HCC is defined as “sarcomatoid HCC (S-HCC)", while sarcomatous change in ICC is defined as “sarcomatoid ICC (S-ICC)” [1, 2]. The pathogenesis of HSC has not been determined. The most widely accepted theory may be the conversion theory, which postulates that sarcomatous components originate from carcinoma transition [3 –7]. Several studies indicated that epithelial-mesenchymal transition was an important mechanism underlying the sarcomatous change [3 , 8–10], which verified the conversion theory. HSC is usually highly invasive and has a poor prognosis. Several studies showed that recurrence-free survival and overall survival of S-HCC and S-ICC were significantly lower than those of HCC and ICC [3 , 12]. Early detection of HSC and radical resection may improve the prognosis of patients [13].

Gray-scale ultrasound (US) and contrast-enhanced ultrasound (CEUS) are real-time, safe and convenient, which have been widely applied in the diagnosis of liver tumors [14 –16]. At present, only few reports focus on gray-scale US and CEUS features of HSC, and the relevant literature is almost all case reports. The ultrasound findings of HSC remain undefined. In this study, US and CEUS features were compared between HSC and HCC to improve the diagnostic efficacy and benefit to clinical practice.

2 Material and methods

2.1 Patients

The clinical medical data of HSC cases meeting the following criteria were retrospectively analyzed, in our hospital from May 2014 to March 2023. The inclusion criteria of the HSC cases were as follows: 1) the lesions were confirmed as HSC by liver resection or biopsy; 2) CEUS was performed within one month before surgery. The exclusion criteria were as follows: 1) metastatic cases; 2) lack of clinical information; 3) poor image quality. A total of 38 HSC cases were included. 135 HCC cases in the same period were selected. Propensity score matching was then performed between HSCs and HCCs at a ratio of 1:4 by age, sex and tumor size. As some cases were not successfully matched, 37 HSC cases and 92 HCC cases were finally included. Among the 37 HSC cases, 36 had a single lesion and one had multiple lesions. Among the 92 HSC cases, 78 had a single lesion and 14 had multiple lesions. The largest lesion was selected, and a total of 129 lesions were included. This study was approved by the ethics committee of our hospital (B2022-223 R).

2.2 Ultrasound image acquisition

Ultrasound systems (Canon Aplio i900, Philips EPIQ7, Siemens Acuson Sequoia) with curved-array transducers at a frequency of 1–8 MHz were used to perform the CEUS examination. A complete ultrasound examination of the whole liver was performed first. Then the color Doppler mode was used to observe the blood flow signals inside and around the lesions. Finally, the CEUS examination was performed when optimal visualization of the target lesion was achieved. A dose of 1.5–2.4 ml of SonoVue (Bracco, Milan, Italy) was injected intravenously and immediately followed by 5 ml of 0.9% NaCl solution. Images were recorded continuously for at least 5 minutes after injection and stored as digital cine clips.

2.3 Ultrasound image assessment

The gray-scale US, CDFI and CEUS images were analyzed by radiologists with more than 10 years of experience. They were blinded to the patients’ clinical and pathological information. The following gray-scale US and CDFI features were evaluated, including the number (single or multiple), location (right, left, junction of left and right or caudate lobe), size, echogenicity (hyperechoic, isoechoic, hypoechoic or mixed echoic), shape (regular or irregular), margin (sharp or indistinct) of lesions, presence of halo sign, intrahepatic cholangiectasis, tumor embolus inside the vessel, abnormal lymph nodes, presence of blood flow signals, resistance index (RI). The CEUS features included the time that the lesion start to enhance, enhancement patterns, the presence of non-enhanced areas inside the lesions, the washout time, the enhancement intensity of the lesions compared to the surrounding parenchyma during the arterial phase (10–45 s), portal venous phase (45–120 s) and late phase (120–300 s) [17] (hyper-enhancement, iso-enhancement, slightly hypo-enhancement or hypo-enhancement).

Enhancement patterns were defined as follows: 1) entirety homogeneous enhancement: the contrast agent rapidly filled the entire lesion with uniform echo inside the lesion; 2) entirety heterogeneous enhancement: the echo intensity varied in the lesion; 3) centripetal enhancement: the lesion enhanced from the periphery and rapidly filled towards the center; 4) dendritic enhancement: several dendritic blood vessels in the lesion enhanced first, after which the contrast agent gradually filled the entire lesion; 5) rim-like enhancement: only the periphery of the lesion was enhanced; 6) peripheral and internal septal enhancement: the periphery of the lesion was enhanced, with large non-enhanced areas inside the lesion, and the separations between non-enhanced areas were enhanced.

The size of non-enhanced areas was classified according to whether their area was larger than 50% of the lesion area. The shape of the non-enhanced areas was divided into regular (the boundary was round or oval, without protrusions) and irregular (the boundary had protrusions).

2.4 Statistical analysis

SPSS 26.0 software (IBM Armonk, NY, USA) was used for data analysis. Continuous variables were expressed as median (25th, 75th) and categorical variables were expressed as frequency (percentage). Mann–Whitney U test, Pearson chi-square test, and Fisher’s exact test were used to compare clinical and ultrasound features between the HSC and HCC groups. P values < 0.05 were considered statistically significant.

3 Results

3.1 Clinical characteristics and pathologic classification of HSC and HCC

Finally, 37 HSC patients and 92 HCC patients were enrolled. Comparisons of clinical characteristics between HSC and HCC patients are demonstrated in Table 1. There were significant differences in the clinical symptoms, tumor size, alpha-fetoprotein (AFP) level and vascular invasion. The most common symptom in HSC patients was abdominal pain or abdominal discomfort (29.7%), other symptoms included right-sided lumbar discomfort (2.7%), fever (2.7%), weight loss (2.7%), and abdominal mass sensation (2.7%). Pathological diagnosis was obtained by liver resection (92 HCCs, 34HSCs) or biopsy (3HSCs). In the HSC group, 28 cases were S-HCC and 9 cases were S-ICC.

Table 1
Clinical characteristics of HSC and HCC

HCC HSC P

Sex Male 71 (77.2%) 27 (73.0%) 0.614

Female 21 (22.8%) 10 (27.0%)

Age (years) 55.0 (49.3,65.0) 56.0 (50.5,62.0) 0.956

Symptoms Symptomatic 16 (17.4%) 15 (40.5%) 0.005

Asymptomatic 76 (82.6%) 22 (59.5%)

Tumor size (mm) 32.5 (20.0,58.8) 50.0 (28.0,80.0) 0.034

Hepatitis B infection 87 (94.6%) 35 (94.6%) 0.999

Liver cirrhosis 73 (79.3%) 25 (67.6%) 0.157

Satellite nodule 5 (5.4%) 3 (8.1%) 0.868

Vascular invasion 6 (6.5%) 9 (24.3%) 0.011

Liver capsule involvement 49 (53.3%) 20 (54.1%) 0.935

Elevated AFP 50 (54.3%) 13 (35.1%) 0.048

Elevated CEA 10 (10.9%) 1 (2.7%) 0.249

Elevated CA19-9 10 (10.9%) 6 (16.2%) 0.591

		HCC	HSC	P
Sex	Male	71 (77.2%)	27 (73.0%)	0.614
	Female	21 (22.8%)	10 (27.0%)
Age (years)		55.0 (49.3,65.0)	56.0 (50.5,62.0)	0.956
Symptoms	Symptomatic	16 (17.4%)	15 (40.5%)	0.005
	Asymptomatic	76 (82.6%)	22 (59.5%)
Tumor size (mm)		32.5 (20.0,58.8)	50.0 (28.0,80.0)	0.034
Hepatitis B infection		87 (94.6%)	35 (94.6%)	0.999
Liver cirrhosis		73 (79.3%)	25 (67.6%)	0.157
Satellite nodule		5 (5.4%)	3 (8.1%)	0.868
Vascular invasion		6 (6.5%)	9 (24.3%)	0.011
Liver capsule involvement		49 (53.3%)	20 (54.1%)	0.935
Elevated AFP		50 (54.3%)	13 (35.1%)	0.048
Elevated CEA		10 (10.9%)	1 (2.7%)	0.249
Elevated CA19-9		10 (10.9%)	6 (16.2%)	0.591

3.2 Gray-scale US, CDFI and CEUS imaging characteristics of HSC and HCC

On gray-scale ultrasound, both HSCs and HCCs most commonly appeared as single lesions (97.3% vs 84.8%). Both HSCs and HCCs were most likely to occur in the right lobe, followed by the left lobe, but 13.5% of HSCs occurred in the junction of left and right while only 1.1% of HCCs occurred there (P < 0.05). Both HSCs and HCCs were mainly hypoechoic (62.2% vs 65.2%). HSCs were more likely to demonstrate as indistinct margins (78.4%) and irregular shapes (70.3%), whereas HCCs were more likely to have sharp margins (51.1%) and regular shapes (76.1%) (P < 0.05). The halo sign was less common in HSCs compared to HCCs (32.4% vs 50.0%), but there was no significant difference between the two. The blood flow signals were easier to detect in HSCs than in HCCs (75.7% vs 56.5%) on CDFI image, and the RI was higher in HSCs compared to HCCs [0.78 (0.70,0.82) vs 0.70 (0.62,0.76), P < 0.05]. The gray-scale US and CDFI imaging features of HSC and HCC are presented in Table 2.

Table 2
Gray-scale US and CDFI imaging characteristics of HSC and HCC

HCC HSC P

Number Single 78 (84.8%) 36 (97.3%) 0.089

Multiple 14 (15.2%) 1 (2.7%)

Tumor location Right lobe 62 (67.4%) 24 (64.9%) 0.027

Left lobe 26 (28.3%) 7 (18.9%)

Junction of left and right 1 (1.1%) 5 (13.5%)

Caudate lobe 3 (3.3%) 1 (2.7%)

Echogenicity of lesions Hypoechoic 60 (65.2%) 23 (62.2%) 0.399

Isoechoic 7 (7.6%) 2 (5.4%)

Hyperechoic 19 (20.7%) 6 (16.2%)

Mixed echoic 6 (6.5%) 6 (16.2%)

Margin Sharp 47 (51.1%) 8 (21.6%) 0.002

Indistinct 45 (48.9%) 29 (78.4%)

Shape Regular 70 (76.1%) 11 (29.7%) <0.001

Irregular 22 (23.9%) 26 (70.3%)

Halo sign 46(50.0%) 12 (32.4%) 0.070

Blood flow signal 52 (56.5%) 28 (75.7%) 0.043

Resistance index 0.70(0.62,0.76) 0.78(0.70,0.82) 0.002

Intrahepatic cholangiectasis 1 (1.1%) 0 (0.0%) 0.999

Tumor embolus inside the vessel 3 (3.3%) 2 (5.4%) 0.947

Abnormal lymph nodes 0 (0.0%) 1 (2.7%) 0.287

		HCC	HSC	P
Number	Single	78 (84.8%)	36 (97.3%)	0.089
	Multiple	14 (15.2%)	1 (2.7%)
Tumor location	Right lobe	62 (67.4%)	24 (64.9%)	0.027
	Left lobe	26 (28.3%)	7 (18.9%)
	Junction of left and right	1 (1.1%)	5 (13.5%)
	Caudate lobe	3 (3.3%)	1 (2.7%)
Echogenicity of lesions	Hypoechoic	60 (65.2%)	23 (62.2%)	0.399
	Isoechoic	7 (7.6%)	2 (5.4%)
	Hyperechoic	19 (20.7%)	6 (16.2%)
	Mixed echoic	6 (6.5%)	6 (16.2%)
Margin	Sharp	47 (51.1%)	8 (21.6%)	0.002
	Indistinct	45 (48.9%)	29 (78.4%)
Shape	Regular	70 (76.1%)	11 (29.7%)	<0.001
	Irregular	22 (23.9%)	26 (70.3%)
Halo sign		46(50.0%)	12 (32.4%)	0.070
Blood flow signal		52 (56.5%)	28 (75.7%)	0.043
Resistance index		0.70(0.62,0.76)	0.78(0.70,0.82)	0.002
Intrahepatic cholangiectasis		1 (1.1%)	0 (0.0%)	0.999
Tumor embolus inside the vessel		3 (3.3%)	2 (5.4%)	0.947
Abnormal lymph nodes		0 (0.0%)	1 (2.7%)	0.287

On contrast-enhanced ultrasound (CEUS), HSCs mainly showed entirety heterogeneous enhancement (48.6%) (Fig. 1), entirety homogeneous enhancement (18.9%), peripheral and internal septal enhancement (18.9%) (Fig. 2), whereas HCCs mainly showed entirety heterogeneous enhancement (55.4%) and entirety homogeneous enhancement (33.7%) (P < 0.05). During the arterial phase of CEUS, most lesions revealed hyper-enhancement (97.3% vs 93.5%) in both HSCs and HCCs. HSCs washed out earlier than HCCs [43.0 s (30.5,58.0) vs 60.0 s (46.3,100.0), P < 0.05], so, hypo-enhancement was more likely to be seen in HSC lesions compared to the HCC lesions during the portal venous phase and the late phase of CEUS. The non-enhanced areas were significantly more frequently observed in HSCs than in HCCs (56.8% vs 31.5%, P < 0.05). The characteristics of the non-enhanced areas in the two groups were investigated. It was found that the size of the non-enhanced areas in both groups was mostly less than 50% of the lesion area (61.9% vs 86.2%, P < 0.05), but there was a significant difference, with HSCs showing larger non-enhanced areas. The shape of the non-enhanced areas was mostly irregular in both HSCs and HCCs (66.7% vs 75.9%). The CEUS imaging characteristics of HSC and HCC are shown in Table 3.

Fig. 1

The CEUS manifestation of sarcomatoid HCC. A 63-year-old man was found to have a liver lesion on physical examination. Gray-scale US revealed a hypoechoic mass in the left lobe of the liver, the mass showed indistinct margin and irregular shape (A). In CEUS arterial phase, the lesion showed entirety heterogeneous enhancement (B). In late phase, the lesion showed hypo-enhancement (C).

Fig. 2

The CEUS manifestation of sarcomatoid ICC. A 54-year-old man, with right abdominal pain. Gray-scale ultrasound revealed a hypoechoic mass in the right lobe of the liver, the mass showed indistinct margin, irregular shape (A). In CEUS arterial phase, the lesion showed peripheral and internal septal enhancement, large non-enhanced areas occurred inside the lesion (B). In portal venous phase, the peripheral and internal septum showed slightly hypo-enhancement (C). In late phase, the lesion showed hypo-enhancement (D).

Table 3

Contrast-enhanced ultrasound imaging characteristics of HSC and HCC

		HCC	HSC	P
The time of starting to enhance (s)		16.0 (14.0,19.0)	16.0 (13.5,18.5)	0.473
Enhancement pattern	Entirety homogeneous enhancement	31 (33.7%)	7 (18.9%)	0.003
	Entirety heterogeneous enhancement	51 (55.4%)	18 (48.6%)
	Centripetal enhancement	1 (1.1%)	2 (5.4%)
	Dendritic enhancement	7 (7.6%)	2 (5.4%)
	Rim-like enhancement	0 (0.0%)	1 (2.7%)
	Peripheral and internal septal enhancement	2 (2.2%)	7 (18.9%)
Contrast-enhanced intensity in arterial phase	Hyperenhancement	86 (93.5%)	36 (97.3%)	0.999
	Iso-enhancement	4 (4.3%)	1 (2.7%)
	Hypo-enhancement	2 (2.2%)	0 (0.0%)
Washout time (s)		60.0 (46.3,100.0)	43.0 (30.5,58.0)	<0.001
Contrast-enhanced intensity in portal venous phase	Iso-enhancement	18 (19.6%)	2 (5.4%)	0.003
	Slightly hypo-enhancement	48 (52.2%)	13 (35.1%)
	Hypo-enhancement	26 (28.3%)	22 (59.5%)
Contrast-enhanced intensity in late phase	Slightly hypo-enhancement	44 (47.8%)	6 (16.2%)	0.001
	Hypo-enhancement	48 (52.2%)	31 (83.8%)
Non-enhanced areas	Present	29 (31.5%)	21 (56.8%)	0.008
	Absent	63 (68.5%)	16 (43.2%)
The size of the non-enhanced areas	Smaller than 50% of the lesion area	25 (86.2%)	13 (61.9%)	0.047
	Larger than 50% of the lesion area	4 (13.8%)	8 (38.1%)
The shape of the non-enhanced areas	Regular	7 (24.1%)	7 (33.3%)	0.475
	Irregular	22 (75.9%)	14 (66.7%)

4 Discussion

Hepatic Sarcomatoid Carcinoma (HSC) contains both malignant epithelial and spindle cell sarcomatoid components [18]. The sarcomatous component is double positive for markers of the epithelium and mesenchyme [11]. HSC is extremely rare, and the morbidity of S-HCC and S-ICC is 1.8–9.4% of HCC and 4.5% of ICC, respectively [13, 19]. Currently, the reports on gray-scale US and CEUS findings of HSC are almost all case reports, and the ultrasound features of HSC remain undefined. In this study, the ultrasound and clinical features of HSC were explored.

HSCs were frequently seen in middle-aged and older men in this study. Compared to HCC, more HSC patients presented with clinical symptoms (40.5% vs 17.4%). The abdominal pain and abdominal discomfort (29.7%) were the most common symptoms of HSCs. Previous researches also showed that HSCs were more likely to present with symptoms than HCCs [3], including abdominal pain, fever, weight loss, and fatigue might occur [13 , 20–23]. Although propensity score matching was performed, the HSC lesions were larger than HCC lesions, which may be explained by the fact that the size difference between HSC and HCC lesions was too large to be eliminated by matching. Multiple previous studies also indicated that the diameter of HSCs was large [3 , 24]. The proportion of elevated alpha-fetoprotein (AFP) was significantly lower in HSC cases than in HCC cases (35.1% vs 54.3%), which is consistent with previous reports [3, 22]. However, previously reported AFP elevations in HSC patients varied widely (7–50%) [18 , 23]. The small sample size of each study may explain this. However, it was also shown that AFP has little diagnostic value for HSC. Similar to the previous studies [1 , 21–23], hepatitis B virus infection and liver cirrhosis were common in both groups, suggesting that hepatitis B is a high-risk factor for HSC and that screening of the high-risk group may facilitate early detection of HSC.

In our study, most of HSCs and HCCs were hypoechoic on US, which is consistent with previous case reports [25 –27]. HSCs were more likely to reveal indistinct margins and irregular shapes. The halo sign was less common in HSCs than in HCCs, but there was no significant difference. As we know, the lesions with halo sign shows sharp margin, which is consistent with the appearance that HCCs were more commonly to show sharp margin. The color Doppler blood flow signals were more common in HSC lesions compared to HCC lesions, indicating that the blood supply of HSC is more abundant and the mass grows faster, tends to be larger, and is more prone to necrosis [28]. The blood flow signals were detected in 75.7% of HSCs, while 24.3% of HSCs did not show blood flow signals due to the location of the lesion close to the heart and the interference of detection. For these lesions, their blood supply needs to be observed using CEUS. In addition, RI was higher in HSCs than HCCs in this study. Yan Wang et al. [29] explained the reason for the high RI of the malignancy. The nodules comprise exclusively arterial tumoral vessels, and hepatocytes may compress the vascular space, producing multiple “stenoses”. Besides, the presence of tumor capsule and cirrhotic parenchyma may affect the venous outflow by compression of peritumoral portal branches. This is the first study to report the higher RI in HSCs than in HCCs. Higher RI may be a characteristic of HSC, which needs further research.

On contrast-enhanced ultrasound (CEUS), HSCs mainly showed entirety heterogeneous enhancement, entirety homogeneous enhancement, peripheral and internal septal enhancement. All those lesions showing entirety homogeneous enhancement were small lesions less than 30 mm, except for one case of 65 mm. Those lesions that showed peripheral and internal septal enhancement were all large lesions larger than 80 mm. This suggests that the enhancement patterns are associated with the tumor size. Because the large lesions grow too fast, the blood vessels are unable to adequately supply the rapidly growing malignant cells, and necrosis appears, resulting in a large non-enhanced area, some tumor cells or fibrous septum exist between the necrotic areas, causing the peripheral and internal septal enhancement. Small lesions have sufficient blood supply, necrotic areas are not common. Unlike HSC, the main patterns of enhancement of HCC were entirety heterogeneous enhancement and entirety homogeneous enhancement, so the pattern of arterial phase enhancement helps to differentiate HSC from HCC. High frame rate contrast enhanced ultrasound (HiFR-CEUS) technology is helpful to better track the movement of microbubbles, which can detect irregular, almost chaotic early hypervascularization in the malignant tumors. More details of the manifestation of arterial phase can be obtained with HiFR-CEUS [30]. Non-enhanced areas occurred in 56.8% of HSCs and 31.5% of HCCs, and these non-enhanced areas were pathologically confirmed as necrotic components. HSC is more poorly differentiated with rapid growth and more necrosis, which is similar to previous studies [1 , 28]. The shape of the non-enhanced areas was mostly irregular in both groups. Large non-enhanced areas (larger than 50% of the lesion area) occurred more frequently in HSCs (38.1% vs 13.8%). We have reason to believe that the larger necrotic area in HSC is related to the factors mentioned above, such as larger tumor diameter and poorer differentiation. Referring to previous case reports, HSCs may show rim-like enhancement with non-enhanced areas or entirety heterogeneous enhancement [26 , 32], which is consistent with the findings of this study. HSCs washed out earlier than HCCs. Guidelines for CEUS in the liver [17] state that early washout (<60 s) usually occurs in poorly differentiated HCC, and HSC are poorly differentiated hepatocellular carcinomas. Therefore, HSC should show early washout, which is consistent with the results of this study.

In order to integrate our findings into clinical routine better, the cut-off values of continuous variables were calculated, the cut-off values of tumor size, washout time and RI were 49 mm, 45.5 s and 0.765, respectively. In clinical practice, if a hypoechoic liver lesion is found to be larger than 49 mm, with indistinct margins, irregular shape, easily detectable blood flow signals and the RI is higher than 0. 765, it should be advised to perform CEUS, and HSC should be highly suspected if it fulfils the conditions that the enhancement pattern is entirety heterogeneous enhancement, or entirety homogeneous enhancement or peripheral and internal septal enhancement, the washout time is earlier than 45 s, and the non-enhanced areas can be detected.

HSC has a poor prognosis, and some researches found that sarcomatoid type was an independent risk factor for patient death [24] and was associated with shorter overall survival [11]. The reason may be that vascular invasion, intrahepatic metastasis and adjacent organ invasion are more frequent in HSC than in HCC [1 , 28]. TANG Y et al. [2] found that patients with vascular invasion had a significantly shorter survival time. Similar to these studies, our study found that HSCs were more likely to have vascular invasion (on pathology) than HCCs (24.3% vs 6.5%). Two of the nine HSCs with vascular invasion (on pathology) presented tumor embolus inside the vessel on ultrasound. The obvious vascular invasion (tumor embolus inside the vessel) can be detected by ultrasound, microvascular invasion requires to be confirmed by pathology. In both groups, more than 50% of cases showed liver capsule involvement, which is also linked to a poor prognosis.

There are some limitations to our study, firstly, due to the rarity of HSC, differences in ultrasound findings based on HSC subtypes were not investigated. Secondly, statistically significant factors were not quantitatively analyzed in this study, and we will develop a diagnostic model and scoring system to further explore the diagnosis of HSC in the future.

5 Conclusion

On CEUS, HSC mainly shows entirety heterogeneous enhancement. It can also show entirety homogeneous enhancement, peripheral and internal septal enhancement. The other valuable features of HSC are large size, hypoechoic, indistinct margin, irregular shape, easily detectable blood flow signals, high RI, early washout, and non-enhanced areas inside the lesion. All the above findings are different from HCC. CEUS is helpful in the diagnosis of HSC and the differential diagnosis of HSC and HCC.

Conflict of interest

The authors declare that there are no conflicts of interest to report.

Funding

This work was supported by the Natural Science Foundation Project of Shanghai “Science and Technology Innovation Action Plan” [Grant No. 20ZR1452800], the National Natural Science Foundation of China [Grant No. 82071942, 82272013], and Shanghai Science and Technology Innovation Action Plan [Grant No.21Y11911200].

References

, Kim

, Min

, Ha

, Jeong

. Imaging features of hepatic sarcomatous carcinoma on computed tomography and gadoxetic acid-enhanced magnetic resonance imaging, Abdom Radiol (NY). 2017;42(5):1424–33.

Tang

, Zhang

, Zhao

, Chen

, Ma

. Development and validation of a comprehensive radiomics nomogram for prognostic prediction of primary hepatic sarcomatoid carcinoma after surgical resection, Int J Med Sci. 2021;18(7):1711–20.

Wang

, Yao

, Sun

, Liu

, Sun

, Lin

, et al. Clinicopathological characteristics and surgical outcomes of sarcomatoid hepatocellular carcinoma, World J Gastroenterol. 2020;26(29):4327–42.

Maeda

, Adachi

, Kajiyama

, Takenaka

, Sugimachi

, Tsuneyoshi

. Spindle cell hepatocellular carcinoma, A clinicopathologic and immunohistochemical analysis of 15 cases. Cancer. 1996;77(1):51–7.

Kakizoe

, Kojiro

, Nakashima

. Hepatocellular carcinoma with sarcomatous change, Clinicopathologic and immunohistochemical studies of 14 autopsy cases. Cancer. 1987;59(2):310–6.

Kim

, Kim

, Jeong

, Baek

. Analysis of intrahepatic sarcomatoid cholangiocarcinoma: Experience from 11 cases within 17 years, World J Gastroenterol. 2019;25(5):608–21.

McCluggage

. Malignant biphasic uterine tumours: carcinosarcomas or metaplastic carcinomas? J Clin Pathol. 2002;55(5):321–5.

Kim

, Kim

, Park

. Sarcomatoid cholangiocarcinoma with osteoclast-like giant cells associated with hepatolithiasis: A case report, Clin Mol Hepatol. 2015;21(3):309–13.

Sung

, Choi

, Lee

, Kim

. Sarcomatoid carcinoma represents a complete phenotype with various pathways of epithelial mesenchymal transition, J Clin Pathol. 2013;66(7):601–6.

10.

Yen

, Lai

, Liao

, Wang

, Liao

, Tung

, et al. Characterization of a new murine cell line of sarcomatoid hepatocellular carcinoma and its application for biomarker/therapy development, Sci Re. 2017;7(1):3052.

11.

Wang

, Kong

, Yang

, Shen

, Zhang

, Wang

. Clinical features of sarcomatoid change in patients with intrahepatic cholangiocarcinoma and prognosis after surgical liver resection: A Propensity Score Matching analysis, J Surg Oncol. 2020;121(3):524–37.

12.

Liao

, Su

, Jeng

, Liang

, Chen

, et al. Clinical Manifestations and Outcomes of Patients with Sarcomatoid Hepatocellular Carcinoma, Hepatology. 2019;69(1):209–21.

13.

Wang

, Cui

, Weng

, Wu

, Qiu

, Lin

. Clinicopathological characteristics and outcome of primary sarcomatoid carcinoma and carcinosarcoma of the liver, J Gastrointest Surg. 2012;16(9):1715–26.

14.

Jung

, Dong

, Jung

. Current aspects of multimodal ultrasound liver diagnostics using contrast-enhanced ultrasonography (CEUS), fat evaluation, fibrosis assessment, and perfusion analysis - An update, Clin Hemorheol Microcirc. 2023;83(2):181–93.

15.

Dong

, Zuo

, Qiu

, Cao

, Wang

, Yu

,et al. Preoperative prediction of microvascular invasion (MVI) in hepatocellular carcinoma based on kupffer phase radiomics features of sonazoid contrast-enhanced ultrasound (SCEUS): A prospective study, Clin Hemorheol Microcirc. 2022;81(1):97–107.

16.

Fan

, Ji

, Cao

, Xu

, Dong

, Wang

. Baseline and contrast-enhanced ultrasound features of hepatic epithelioid angiomyolipoma, Clin Hemorheol Microcirc. 2022;80(4):447–61.

17.

Dietrich

, Nolsøe

, Barr

, Berzigotti

, Burns

, Cantisani

, et al. Guidelines and Good Clinical Practice Recommendations for Contrast-Enhanced Ultrasound (CEUS) in the Liver-Update 2020 WFUMB in Cooperation with EFSUMB, AFSUMB, AIUM, and FLAUS, Ultrasound Med Biol. 2020;46(10):2579–604.

18.

, Xing

, Ma

, Zhao

, Xu

, Zheng

, et al. Primary Liver Sarcomatoid Carcinoma: A Case Series and Literature Review, J Hepatocell Carcinoma. 2021;8 1117–27.

19.

Watanabe

, Uchinami

, Yoshioka

, Nanjo

, Yamamoto

. Prognosis analysis of sarcomatous intrahepatic cholangiocarcinoma from a review of the literature, Int J Clin Oncol. 2014;19(3):490–6.

20.

Wang

, Ming

, Ren

, Qiu

, Zhou

, Yang

, et al. Sarcomatoid intrahepatic cholangiocarcinoma mimicking liver abscess: A case report, World J Clin Cases. 2020;8(1):208–16.

21.

Shi

, Sun

, Ma

, Chang

, Li

. Clinical and imaging characteristics of primary hepatic sarcomatoid carcinoma and sarcoma: a comparative study, BMC Cancer. 2020;20(1):977.

22.

Zhang

, Chai

, Chen

, Wang

, Cheng

, Fang

, et al. MRI Features of Hepatic Sarcomatoid Carcinoma Different From Hepatocellular Carcinoma and Intrahepatic Cholangiocarcinoma, Front Oncol. 2021;11 611738.

23.

Shi

, Ma

, Zhao

, Chang

, Shao

, Qi

, et al. Imaging and clinical features of primary hepatic sarcomatous carcinoma, Cancer Imaging. 2018;18(1):36.

24.

Zhang

, Feng

, Tu

, Sun

, Rui

, Zhang

,et al. Sarcomatoid hepatocellular carcinoma: From clinical features to cancer genome, Cancer Med. 2021;10(18):6227–38.

25.

Xing

, Wang

. Clinical and ultrasound findings of hepatic sarcomatoid carcinoma, Med Ultrason. 2022;24(3):3625.

26.

Yang

, Lv

, Zhao

, Pan

, Zhang

, Wei

. Sarcomatoid hepatocellular carcinoma mimicking hepatic abscess: A case report, Medicine (Baltimore). 2020;99(39):e22489.

27.

Inagaki

, Sugimoto

, Shiraki

, Yoshizawa

, Tameda

, Ogura

, et al. Sarcomatous hepatocellular carcinoma with remittent fever, Intern Med. 2012;51(21):3025–9.

28.

Koo

, Park

, Kim

, Lim

, Yu

, Jin

, et al. Radiological and clinical features of sarcomatoid hepatocellular carcinoma in 11 cases, J Comput Assist Tomogr. 2008;32(5):745–9.

29.

Wang

, Wang

, Ding

, Huang

, Mao

, Xu

. Resistance index in differential diagnosis of liver lesions by color doppler ultrasonography, World J Gastroenterol. 2004;10(7):965–7.

30.

Jung

, Moran

, Engel

, Krüger-Genge

, Stroszczynski

, Jung

. Modified contrast-enhanced ultrasonography with the new high-resolution examination technique of high frame rate contrast-enhanced ultrasound (HiFR-CEUS) for characterization of liver lesions: First results, Clin Hemorheol Microcirc. 2023;83(1):31–46.

31.

Leng

, Xiang

, Tang

, Yang

, Qiu

. Primary hepatic sarcomatoid carcinoma: A case report, Exp Ther Med. 2015;10(3):1145–8.

32.

Watanabe

, Matsumoto

, Ogawa

, Moriyama

, Sugitani

. Sarcomatoid Hepatocellular Carcinoma with Spontaneous Intraperitoneal Bleeding, Intern Med. 2015;54(13):1613–7.