Noninvasive Assessment of HCV Patients Using Ultrasound Elastography

Abstract

Among patients with chronic hepatitis C (CHC) infection, extensive research showed that fibrosis progression is a proper surrogate marker for advanced liver disease, eventually leading to dramatic endpoints such as cirrhosis and hepatocellular carcinoma. Therefore, there is growing interest in the use of noninvasive methods for fibrosis assessment in order to replace liver biopsy (LB) in clinical practice and provide optimal risk stratification. Elastographic techniques, such as Vibration Controlled Transient Elastography (VCTE), point-shear wave elastography (p-SWE), and 2D-SWE have shown promising results in this regard, with excellent performance in diagnosing hepatic cirrhosis, and great accuracy for steatosis detection through the Controlled Attenuation Parameter embedded on the VCTE device. In addition, the recent introduction of highly efficient direct-acting antivirals (DAAs) led to viral eradication and a significant decrease in liver damage, lowering the risk of hepatic decompensation, and HCC. Therefore, CHC patients need proper noninvasive and repeatable methods for adequate surveillance, even after treatment, as there still remains a risk of portal hypertension and HCC. However, the usefulness for monitoring fibrosis after the sustained virological response (SVR) needs further research.

1. Introduction

Hepatitis C virus (HCV) infection is a major causative agent of chronic liver disease (CLD) and liver-related death worldwide. Approximately 4 out of 5 infected individuals develop chronic hepatitis C (CHC) and nearly 20% of them insidiously progress to cirrhosis, hepatocellular carcinoma (HCC), and end-stage liver disease [1]. It is estimated that 71.1 people were infected in 2015 worldwide, making it a global public health issue due to its substantial prevalence and effect on overall morbidity and mortality [2].

It has been shown that the accumulation of liver fibrosis has a great impact on the evolution of CHC. Fibrosis is the hallmark of progressive disease, eventually leading to cirrhosis and end-stage liver complications [3]. As highlighted by a prospective study conducted by Yano et al. [4], relatively few patients with absent or low-grade fibrosis develop cirrhosis over the next 20 years (25–30%). However, portal and septal fibrosis were followed by cirrhosis in all cases with a progression rate of 18–20 years for portal fibrosis and 8–10 years for septal fibrosis. Furthermore, the advent of direct-acting antivirals (DAAs) has changed the perspective of CHC therapy, being both well-tolerated and highly efficient in achieving sustained virologic response (SVR) [5]. Therefore, staging liver fibrosis as a triage for starting therapy may no longer be as decisive as before. Rather, prompt diagnosis and management of advanced stages of fibrosis can prevent complications and death through comprehensive preventive and management strategies [6].

Liver biopsy (LB) is traditionally considered the gold standard evaluation for necroinflammatory activity, steatosis, and fibrosis in CHC [7]. However, the method has several drawbacks. Firstly, the result of the histopathological examination is significantly affected by the specimen’s quality and the pathologist’s experience [8, 9, 10, 11]. Secondly, it is an invasive procedure, implies high costs, and might lead to several complications. Noninvasive methods are therefore necessary for optimal risk stratification in order to avoid the use of LB. Even if conventional imaging techniques are noninvasive, they require absolute signs of severe fibrosis or cirrhosis. Therefore, the latest studies focused on noninvasive elastographic techniques, which have shown promising results for the appraisal of liver fibrosis and steatosis in CHC patients.

2. Fibrosis assessment in HCV patients using noninvasive elastographic methods: a classification

Elastography-based imaging techniques quantify tissue stiffness, defined as the resistance of a material in response to an applied mechanical force [12]. Fibrosis modifies the elastic properties of liver tissue so that new techniques have been developed in the past two decades to grade liver fibrosis according to tissue stiffness.

Ultrasonographic (US) and magnetic resonance-based elastographic techniques are available, of which we will focus on US methods. Several guidelines classify elastographic techniques in two main categories: quantitative (“Shear Wave Elastography”, SWE) and qualitative (“Strain Elastography) [13, 14, 15]. Regarding CHC, quantitative methods are most frequently used to evaluate liver stiffness (LS). Currently, three main quantitative techniques showed promising results in this pathology: Vibration Controlled Transient Elastography (VCTE; FibroScan®, Echosens, Paris, France) point-shear wave elastography (pSWE), and 2D- shear wave elastography (2DSWE). For integrative purposes, we decided to summarize the specific advantages and limitations of each technique in HCV patients.

	Advantages	Disadvantages
Vibration Controlled Transient Elastography (VCTE)	Widely used, less expensive, easy to learn [24]; Can be easily repeated overtime; Can provide steatosis assessment through CAP measurement; Great reproducibility (>90 interclass correlation coefficients) [15]; Point-of-care method [6]; Good diagnostic accuracy for fibrosis stages and high performance for cirrhosis (AUROC>0.9) [6]	It cannot be performed in subjects with ascites [6]; Obesity increases LSM (the use of XL probe reduces the limits among these subjects) [24]; Affected by acute hepatitis, food intake, liver congestion, cholestasis, and alcohol consumption [6]
Point-shear Wave Elastography (p-SWE)	Can be easily executed on modified commercial US devices (if the machine is provided with adequate software) [6, 14]; Offers the possibility of choosing the ROI in real-time [14, 25]; Enables entire liver parenchyma examination under B-mode visualization [26]; Avoids masses or large vessels [26]; Can evaluate focal liver lesions’ stiffness, discriminating between malignant and benign lesions [24]; Good applicability: practicable among patients with ascites and obesity [6, 15, 25, 27]; Excellent diagnostic accuracy for advanced fibrosis and cirrhosis [6, 25]; Enables spleen stiffness measurement [6]	Narrow range (0.5–4.4 m/s), making it difficult to set proper cut-off values [14]; Affected by acute hepatitis, food intake, liver congestion, cholestasis, and alcohol consumption [6]
2D-shear Wave Elastography (2D-SW)	Can be easily executed on modified commercial US devices (if the machine is provided with adequate software) [6, 14]; Offers the possibility of choosing a large and adjustable ROI in real-time [14, 25, 28]; Good applicability: practicable among patients with ascites and obesity [6, 28]; Excellent diagnostic accuracy for advanced fibrosis and cirrhosis [6, 28]	Affected by acute hepatitis, food intake, liver congestion, cholestasis, and alcohol consumption [6]

Table 1.

Advantages and disadvantages of noninvasive elastographic techniques.

3. Confounders: pathological changes influencing liver stiffness in HCV patients

Several technical and biological factors affect the performance of elastographic techniques due to an increase in LS unrelated to fibrosis. The former includes shear wave frequencies, location and depth of measurements, and device dependencies. The latter include ingestion of food prior to the examination, inflammation, cholestasis, hepatic venous congestion, and amyloid deposits [16, 17]. Inflammation in acute hepatitis might increase LS up to mimicking cirrhosis, returning to normal simultaneously with the decrease of liver transaminases. A study of 112 CHC patients found a higher value for LS in the case of F3-F4 stages of fibrosis and necroinflammatory activity of at least A2 compared to A0-A1 (14.6 and 6.2 kPa, p = 0.04) [18]. Therefore, it is recommended to consider transaminases’ value before interpreting LS in order to avoid overestimation [19]. If ALT levels are 3 times the normal value, there is a risk of overestimating the fibrosis stage and this should be mentioned with the results [20].

Concerning cholestasis and heart failure with hepatic congestion, LS decreases after proper treatment, hence the effect on shear wave propagation [21, 22]. Of note, a study suggests FibroScan as a potential tool to reveal heart decompensation [23]. In addition, waist circumference may lead to both technical failure and higher LS, but studies show various results and are mainly referring to body mass index (BMI). This is common because central obesity is associated with low-grade inflammation, insulin resistance, and liver steatosis, increasing LS. Furthermore, male gender, dyslipidemia and statins are debated in this regard, with different results (Table 1) [29].

4. Vibration controlled transient elastography performance for fibrosis staging in HCV-infected patients

As already mentioned, staging fibrosis is beneficial to determine the prognosis and follow-up of patients with chronic HCV infection. VCTE is the most validated elastographic modality in this regard. Several meta-analysis found an excellent diagnostic performance of VCTE in diagnosing hepatic cirrhosis, with an AUROC exceeding 0.90. However, the technique is less accurate in case of significant fibrosis (≥F2), with an AUROC ranging between 0.80 and 0.90 with overlapping cutoff values, so that the method is facing difficulties in distinguishing different stages of fibrosis [30, 31, 32, 33, 34]. Concerning CHC, two of these meta-analysis reported AUROC values of 0.83–0.85 and 0.96 for diagnosing significant fibrosis and cirrhosis, respectively [31, 33]. EASL guidelines suggest combining VCTE with serological markers for the assessment of moderate fibrosis (F2-F4) in patients with CHC [6].

A recent meta-analysis counting 24 articles evaluated the performance of VCTE for diagnosing liver cirrhosis in CHC patients. It estimated a sensitivity (Se) of 84% and a specificity (Sp) of 90%, with AUROC 0.95 [35]. Ganne-Carrié suggests in a study with 775 patients that VCTE should be particularly used for ruling out cirrhosis, given its high negative predictive value (NPV) (96%), rather than ruling it in, since the positive predictive value (PPV) was only 74%. Nevertheless, the excellent diagnostic performance for cirrhosis is hereby confirmed [36].

A recent study [37] compares the performance of VCTE with conventional B-mode ultrasound (US). VCTE is clearly superior in diagnosing severe fibrosis and subclinical cirrhosis, with an AUROC of 0.95 for severe fibrosis and 0.96 for cirrhosis versus 0.76 and 0.71, respectively, in the case of US (p < 0.001). Furthermore, combining the two methods does not significantly improve diagnostic accuracy compared with VCTE alone. The two would improve Sp (95.7% versus 76.7; p < 0.001) and PPV (94.3% versus 77.1%; p = 0.002) [37]. Another study by Berzigotti et al. [38] suggests that the two methods work complementary so that US is the preferred technique for ruling in cirrhosis, while VCTE should be used for ruling out the disease. Contrary to the first example, Benzigotti claims a better performance when the two are combined.

The diagnostic performance of VCTE in staging fibrosis is exemplified in Table 2, in reliance to our previous research [50]. Cutoff values range from 4.5 to 9.5 kPa for significant fibrosis (≥F2) and from 11.3 to 16.9 kPa for diagnosing cirrhosis (F4). These values vary considerably mainly according to the prevalence of fibrosis in each study group and the expected outcome [51].

Fibrosis stage	≥ F1			≥ F2			≥ F3			= F4
Study	Cut-off (kPa)	AUROC	Se/Sp (%)	Cut-off (kPa)	AUROC	Se/Sp (%)	Cut-off (kPa)	AUROC	Se/Sp (%)	Cut-off (m/s)	AUROC	Se/Sp (%)
Njei (n = 756) [39]*	N/S	N/S	N/S	4.5–7.2	N/S	97/64	N/S	N/S	N/S	11.8–14.6	N/S	90/87
Lupsor (n = 1202) [20]	5.3	0.879	84.99/73.21	7.4	0.889	80.32/83.97	9.1	0.941	88.8/88.3	13.2	0.970	93.75/93.31
Castera (n = 183) [40]	N/S	N/S	N/S	7.1	0.83	67/89	9.5	0.90	73/91	12.5	0.95	87/91
Afdhal (n = 560) [41]^1	N/S	N/S	N/S	8.4	0.73	58/75	9.6	0.83	71.8/80	12.8	0.90	76/85
Degos (n = 913) [42]	N/S	N/S	N/S	5.2	0.75	90/32	N/S	N/S	N/S	12.9	0.90	72/89
Ziol (n = 251) [43]	N/S	N/S	N/S	8.80	0.79/0.81	56/91	9.6	0.91/0.95	86/85	14.6	0.95	87/91
Carrion (n = 169) [44]	8.5	N/S	N/S	N/S	0.90	90/81	N/S	0.93	N/S	12.5	0.95	87/91
Arena (n = 150) [19]	N/S	N/S	N/S	7.8	0.91	83/82	10.8	0.99	91/94	14.8	0.98	94/92
Sporea (n = 191) [45]	N/S	N/S	N/S	6.8	0.733	59.6/93.3	N/S	N/S	NS	N/S	N/S	N/S
Nitta (n = 165) [46]	N/S	N/S	N/S	7.1	0.87	80.8/80.3	9.6	0.91	87.7/82.4	11.6	0.93	62.5–91.7/78.9–91.5
Zarski (n = 382) [47]	N/S	N/S	N/S	5.2	0.82	93.6/34.8	N/S	N/S	N/S	12.9	0.93	76.8/89.6
Wang (n = 214) [48]	6.5	0.86	75/78	9.5	0.82	69/81	N/S	0.87	N/S	12	0.93	89/84
Yoneda (n = 102) [49]^2	N/S	N/S	N/S	7.8	0.91	78/90	10.4	0.95	88/91	11.3	0.91	90/84

Table 2.

Performance of VCTE for detecting different stages of fibrosis in HCV-infected patients.

Abbreviations: N/S – not specified; AUROC – area under ROC curve; Se – sensibility; Sp – specificity

*

meta-analysis.

¹

HCV 92%, HBV 8%.

²

vXL probe.

5. Liver fibrosis assessment through point-shear wave elastography in HCV patients

Point-shear wave elastography (p-SWE) is incorporated into devices such as Virtual Touch Tissue Quantification (VTTQ®) (Siemens Healthcare, Erlangen, Germany) and Elastography Point Quantification (ElastPQ®) (EPIQ7 ultrasound system, Philips Healthcare, Bothell, WE, USA). Under B-mode visualization, p-SWE enables the precise acquisition of shear wave speed (SWS) in a small ROI (around 1 cm³). After 10 valid measurements in the right hepatic lobe, the median of SWS is reported and interpreted [15, 52]. Results are expressed in m/s for VTTQ or in m/s and kPa for ElastPQ [13]. However, its narrow range (0.5–4.4 m/s) restricts the demarcation of proper cutoff values for discriminating between certain fibrosis stages, making management decisions difficult [14].

We identified several studies that evaluate p-SWE in HCV-infected patients [53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76]. A 2011 pooled meta-analysis by Friedrich-Rust et al. [58] with 380 CHC patients, found AUROC values of 0.88, 0.90, and 0.92 for diagnosing moderate fibrosis (≥ F2), severe fibrosis (≥ F3), and cirrhosis (= F4), respectively. Subsequently, an international multicenter study with 911 HCV-infected patients offered cut-off values of 1.19, 1.33, 1.43, respectively 1.55 m/s for ≥ F1, ≥ F2, ≥ F3 respectively F4, with AUROC values of 0.779, 0.792, 0.829 and 0.842, respectively [59]. Off note is the Takaki study [65] which elaborated the VIA index, a formula that increases the diagnostic accuracy of SWV alone, from 0.882, 0.858 and 1.000 to 0.917, 0.906, and 1.000 for moderate fibrosis, severe fibrosis, and cirrhosis, respectively in the validation set. In 2019, Hsu et al. [74] propound different SWV cutoff values in various diseases. For CHC patients, a SWV cut-off value of 1.225, 1.370 and 1.710 m/s predicts fibrosis stages ≥ F2, ≥ F3 and F4 with AUROC values of 0.786, 0.857 and 0.937, respectively. Overall, in our considered studies presented in Table 3, the AUROC ranged from 0.725 to 0.88 for ≥F1, 0.67 to 0.93 for ≥F2, 0.74 to 0.97 for ≥F3 and 0.79 to 1 for F4 prediction. Nonetheless, the EFSUMB Clinical Practice Guidelines suggest that pSWE can be the first-line assessment in HCV-infected patients for fibrosis evaluation, performing best at ruling out cirrhosis [15].

Fibrosis stage	Technology	[≥ F1			≥ F2			≥ F3			= F4
Study		Cut-off (m/s)	ROC	Se/Sp (%)	Cut-off (m/s)	ROC	Se/Sp (%)	Cut-off (m/s)	ROC	Se/Sp (%)	Cut-off (m/s)	ROC	Se/Sp (%)
Lupsor (n = 112) [53]	VTTQ	1.19	0.725	62.07/85.71	1.34	0.869	67.8/92.86	1.61	0.9	79.07/94.83	2.0	0.936	80/95.45
Fierbinteanu-Braticevici (n = 74) [54]	VTTQ	1.185	N/S	89/87	1.215	90.2	100/71	1.54	99.3	97/100	1.94	99.3	100/98.1
Friedrich-Rust (n = 64) [55]	VTTQ	N/S	N/S	N/S	1.35	0.86	72.9/93.8	1.55	0.93	81.5/91.9	1.75	0.95	88.9/89.1
Rizzo (n = 139) [56]	VTTQ	N/S	N/S	N/S	1.3	0.86	81/70	1.7	0.94	91/86	2.0	0.89	83/86
Sporea (n = 274) [57]	VTTQ	1.19	0.880	73/93	1.21	0.893	84/91	1.58	0.908	84/94	1.82	0.937	91/90
Friedrich-Rust (n = 380) [58]*	VTTQ	N/S	N/S	N/S	N/S	0.88	N/S	N/S	0.90	N/S	N/S	0.92	N/S
Sporea (n = 911) [59]	VTTQ	1.19	0.779	69.9/80	1.33	0.792	69.1/79.8	1.43	0.829	74.8/81.5	1.55	0.842	84.3/76.3
Chen (n = 127) [60]	VTTQ	N/S	N/S	N/S	1.55	0.847	74.1/87	1.81	0.902	90.2/89.5	1.98	0.831	88.9/79.8
Zhang (n = 108) [61]	N/S	N/S	N/S	N/S	1.529	0.779	56.9/88.9	1.78	0.863	73.2/92.5	1.797	0.79	78.6/74.5
Nishikawa (n = 108) [62]	VTTQ	1.28	0.810	69.1/85.7	1.28	0.909	81.8/87.1	1.44	0.869	88.9/82.5	1.73	0.885	85.7/86.2
Yamada (n = 124) [63]	VTTQ	N/S	N/S	N/S	1.26	0.890	92.5/76.2	1.46	0.943	84.6/87.8	N/S	N/S	N/S
Li (n = 128) [64]	ElastPQ	N/S	N/S	N/S	1.53	0.775	57.6/89.5	1.79	0.901	76.4/96.5	1.789	0.792	78.9/75.4
Takaki (n = 176)^1 [65]	VTTQ	N/S	N/S	N/S	1.25; 1.205	0.773; 0.882	75/78.1; 75/90.9	1.595; 1.595	0.863; 0.858	84.9/81.5; 94.3/81.8	1.775; 1.775	0.915; 1.000	85.6/88.9; 100/40
Silva Junior (n = 51) [66]	VTTQ	1.19	0.88	88.4/75	1.31	0.90	89.3/87	1.68	0.97	94.4/90.9	1.95	0.98	100/95.2
Tai (n = 83) [67]	VTTQ	N/S	N/S	N/S	N/S	N/S	N/S	N/S	N/S	N/S	1.41	0.802	70.6/80.3
Chen (n = 137) [68]	VTTQ	N/S	N/S	N/S	1.59	0.8434	72.8/79.4	1.73	0.8997	91.4/77.2	1.96	0.9036	100/68.1
Joo (n = 101) [69]	VTTQ	1.190	0.872	84/85.7	1.335	0.853	83.8/75.8	1.645	0.840	79.5/75.8	1.665	0.828	85/69.1
Mare (n = 168)^2,3 [70]	ElastPQ	N/S	N/S	N/S	6.4	0.96	92/100	6.7	0.97	88.4/100	8.9	0.83	90.9
Gani (n = 29) [71]	N/S	N/S	N/S	N/S	1.32	0.802	79/75	1.48	0.802	78/80	1.79	0.802	72/82
Ragazzo (n = 107) [72]	VTTQ	N/S	N/S	N/S	1.22	0.67	64/69	1.41	0.74	57/84	2.37	0.96	100/94
Alem (n = 2103) [73]^3	VTTQ	N/S	N/S	N/S	1.36	0.89	87.5/80.6	1.45	0.94	97.5/90.3	1.7	0.95	90.3/90.9
Hsu (n = 63) [74]	N/S	N/S	N/S	N/S	1.225	0.786	65/70.9	1.370	0.857	73.1/78.4	1.710	0.937	81.8/86.5
Ueda (n = 108) [75]	VTTQ	N/S	N/S	N/S	1.26	0.93	N/S	1.78	0.83	N/S	1.94	0.86	N/S
Baldea (N = 176) [76]^2,3	ElastPQ	N/S	N/S	N/S	6.51	0.92	96.6/76.4	8.73	0.94	88/85.4	11.1	0.95	86.8/96.7

Table 3.

Performance of p-SWE for detecting different stages of fibrosis in HCV-infected patients.

Abbreviations: N/S – not specified; ROC – receiver operating characteristics curve; Se – sensibility; Sp – specificity

*

meta-analysis.

¹

The Takaki study used two groups: the training set (n = 120) the validation set of VIA index (n = 56); for integrative purposes, we presented in our table only the values of p-SWE measurement (SWV), omitting the VIA index’ ones.

²

The Mare and Baldea studies used kPa for ElastPQ cut-off values.

³

VCTE as reference method.

6. Appraisal of liver fibrosis by 2D shear wave elastography in HCV-infected patients

2D shear wave elastography, a novel US-based technique, allows the estimation of tissue dynamics using focused ultrasonic beams in a certain ROI. This technique has the advantage of displaying a real-time color-coded map overlayed on a B-mode image. Furthermore, 2D-SWE estimates LS expressed in kPa or m/s [28, 77]. It should be executed in a well-visualized area of the right hepatic lobe, clear of large vessels, ligaments, gallbladder, and the liver capsule, with the patient situated in a supine position with breathing suspension [15].

As exemplified in Table 4, several studies reported the diagnostic accuracy of 2D-SWE for fibrosis assessment among HCV-infected patients [76, 78, 79, 80, 81, 82, 83, 84, 85, 86, 88]. In 2017, Herrmann et al. [81] performed a meta-analysis including 13 studies, gathering 379 patients with CHC, that evaluated the diagnostic performance of 2D-SWE for the noninvasive staging of liver fibrosis. They found AUROC values of 0.863, 0.915, and 0.929 for diagnosing significant fibrosis, severe fibrosis, and cirrhosis, respectively. In our analysis, the AUROC values range from 0.82 to 0.888 for ≥ F1, 0.783 to 0.97 for ≥ F2, 0.877 to 0.97 for ≥ F3 and 0.893 to 0.98 for ≥ F4, respectively (Table 4).

Fibrosis stage	≥ F1			≥ F2			≥ F3			= F4
Study	Cut-off (kPa)	ROC	Se/Sp (%)	Cut-off (kPa)	ROC	Se/Sp (%)	Cut-off (kPa)	ROC	Se/Sp (%)	Cut-off (kPa)	ROC	Se/Sp (%)
Bavu (n=113) [78]	N/S	N/S	N/S	9.12	0.948	72/81	10.08	0.962	0.968	13.3	0.968	87/80
Ferraioli (n = 121) [79]	N/S	N/S	N/S	7.1	0.92	90/87.5	8.7	0.99	97.3/95.1	10.4	0.98	87.5/96.8
Tada (n = 55) [80]	N/S	N/S	N/S	8.8	0.940	88.9/91.9	N/S	N/S	N/S	N/S	N/S	N/S
Herrmann (n = 379) [81]*	N/S	N/S	N/S	7.1	0.864	94.7/52	9.2	0.915	90.3/76.8	13.0	0.929	85.8/87.8
Abe (n = 233)^1 [82]	1.480	0.888	75.9/88.2	1.560	0.915	85.3/85.5	1.720	0.940	88.8/83.8	1.930	0.949	91.4/90.8
Serra (n = 51) [83]	N/S	N/S	N/S	9.225	0.783	59.1/86.2	10.695	0.877	72.7/90	11.525	0.893	85.7/88.6
Villani (n = 178) [84]^2,3	N/S	N/S	N/S	8.15	0.899	87.1/73	10.31	0.900	77.2/85.4	12.65	0.899	73.3/88.5
Baldea (n = 176) [76]	N/S	N/S	N/S	6.5	0.92	84.1/88.2	8.19	0.93	96.7/77.4	11.3	0.96	95.7/92.7
Baldea (n = 208) [85]^3,4	N/S	N/S	N/S	7.7; 8.5	0.97; 0.96	86.3/96.9; 87.6/100	8.3; 11.1	0.97; 0.95	94.8/90.3; 89.7/95.1	9.7; 12.3	0.97; 0.96	92.6/91.6; 92.5/94
Aksakal (n = 43) [86]	6.09	0.82	86/70	7.81	0.97	84/96	9.00	0.97	90/97	12.47	0.98	85/98
Numao (n = 141) [87]	N/S	N/S	N/S	N/S	0.86	N/S	N/S	0.97	N/S	N/S	0.91	N/S

Table 4.

Performance of 2D-SWE for detecting different stages of fibrosis in patients with CHC.

Abbreviations: N/S – not specified; ROC – receiver operating characteristics curve; Se – sensibility; Sp – specificity.

*

meta-analysis.

¹

The Abe study used SWV, expressed in m/s.

²

The Villani study used the novel EPIQ 7 US system (ElastQ).

³

These studies used VCTE as reference standard for fibrosis evaluation.

⁴

The Baldea study offered result two 2D-SWE techniques: General Electric and SuperSonic Imagine, respectively.

7. HCV post-sustained virological response/antiviral therapy appraised by liver stiffness

Due to their potency, ease of use, tolerability, and safety, DAA regiments are the recommended choice in subjects with compensated advanced chronic liver disease (cACLD). Their introduction resulted in increasing rates of SVR, reducing LS among these patients. Nonetheless, most of the studies concerning interferon-free treatments are retrospective, with small sample sizes, with short follow-up after SVR, and lacking post-SVR LB [6, 89, 90]. A recent prospective longitudinal study by Knop et al. [91] sought to elucidate the dynamics of liver and spleen stiffness in cirrhotic patients through VCTE and p-SWE, 3 years post-treatment. Even if their analysis showed that LS decreases in a significant proportion of patients with CHC, spleen stiffness, a non-invasive marker for portal hypertension, remained unchanged. Similarly, other research found lower LS values by p-SWE (VTTQ) in HCV-infected patients who achieved SVR [92, 93].

In addition, the diagnostic accuracy of VCTE for SVR prediction remains controversial, since the improvement of LS post-DAA treatment may be overrated by elastography in contrast with histological staging [94]. In fact, a recent prospective multicenter study comprising of 746 HCV-infected patients with CHC with SVR evaluated 3 years post-DAA therapy, discovered cirrhosis by LB in more than half of cACLD patients, in spite of normal VCTE values or liver function parameters. Due to its poor diagnostic accuracy (AUROC = 0.75), VCTE turned out to be an unreliable method for the accurate identification of the fibrosis stage in HCV-infected patients who acquired SVR [95].

Latest EASL guidelines conclude that neither noninvasive elastographic techniques are appropriate enough to detect fibrosis regression after SVR in CHC patients. Additionally, cut-off values of LS by VCTE used in untreated HCV patients should not be utilized for liver fibrosis staging after SVR. Therefore, the appraisal of liver disease severity and prognosis remains an unmet need in this field, requiring larger cohort sizes and extended follow-up in order to establish the role of noninvasive techniques in treating HCV-infected patients [6].

8. The prognostic value of elastography for the prediction of clinical outcomes (decompensation; HCC) in patients with HCV-related cACLD who achieved sustained virological response

In HCV-infected patients, the risk of all-cause mortality and the incidence of HCC diminished in subjects who achieved SVR after interferon-based antiviral therapy, regardless of the grade of fibrosis [96, 97]. In addition, the introduction of novel highly efficient DAAs improved the capability of decreasing the HCC risk, even among patients with advanced liver disease [50]. Nonetheless, a relevant risk of 1.5% remains, requiring proper and cost-effective surveillance methods for these patients [98, 99]. Evidence shows that clinically significant portal hypertension (CSPH), defined as hepatic venous pressure gradient (HVPG) ≥10 mmHg, is the strongest predictor for hepatic decompensation [6]. For these patients, compensated advanced CLD (cACLD) is the proposed term by the Baveno VI consensus [100]. However, HVPG is an invasive and expensive method, requiring reliable noninvasive alternatives [50].

Being a noninvasive, low-cost, and easy to perform method, VCTE turned out to be an outstanding diagnostic instrument for CSPH, with a hierarchical summary ROC of 0.93 [101]. A recent multicenter study of 5648 patients, found that lowering the dual-threshold to <7 kPa and > 12 kPa, provided excellent Se of 91% for excluding and great Sp of 92% for diagnosing cACLD, respectively [100]. In addition, elastography might enable the dynamic appraisal of the HCC risk, especially before and after antiviral treatment. Several studies aimed to elucidate whether VCTE, p-SWE, and 2D-SWE may facilitate HCC surveillance in HCV-infected patients [102, 103, 104, 105]. A recent meta-analysis, comprising 3398 patients, found a pooled HR for HCC development of 3.43 (95% CI, 1.63–7.19) between positive and negative LSM, indicating that VCTE is a trustworthy procedure for HCC prediction in CHC patients treated with DAAs [106]. In a multicenter cohort study, Alonso et al. [107] provided two easy and broadly applicable models for the estimation of HCC risk after SVR. Their model, including baseline albumin (≥ or < 4.2 g/dl), baseline LS (> or ≤ 17.3 KPa), and LS after 1 year (≥ or < 25.5%), increased HCC surveillance efforts (Harrell’s C: 0.77). In 2018, Ioannou et al. [108] internally validated models that calculate the HCC risk following antiviral treatment. However, current EASL clinical practice guidelines recommend that patients with cACLD before antiviral therapy should be continuously supervised for HCC and portal hypertensions, regardless of measurement values of noninvasive tests post-SVR.

9. Controlled attenuation parameter (CAP) for the noninvasive estimation of steatosis in HCV-infected patients

Steatosis is a common histological feature and has an important role in the evolution of CHC, in particular in HCV genotype 3 infections [109, 110]. According to a meta-analysis counting 25 studies and 6400 patients, the prevalence of steatosis in CHC patients is estimated at 55.54%, in most cases affecting less than 33% of hepatocytes [111, 112].

Steatosis seems to accelerate fibrosis in the early stages of the disease, reduce treatment response and promote oncogenesis [2, 113, 114]. A recent retrospective study with 515 CHC patients undergoing DAA treatment found a significant correlation between the grade of steatosis and mortality of any cause or HCC development. Furthermore, steatosis surpasses advanced fibrosis regarding the prediction of a poor response to treatment [115]. Steatosis is thus a simple and important predictor of progression in chronic HCV patients.

US is the commonest imaging technique used in clinical practice to diagnose steatosis due to its high accessibility and low cost. However, it is operator- and machine-dependent and the performance is questionable [116, 117]. There is an increasing interest in developing novel tools for steatosis evaluation. At present, the non-invasive parameter, called controlled attenuation parameter (CAP), available on the FibroScan system, is the most validated one. Using the postulate that fat content is directly related to US beam attenuation, CAP enables the diagnosis and quantification of steatosis [118]. Results are expressed in decibels per meter (dB/m), with values ranging from 100 to 400 dB/m.

Several meta-analysis assessed the CAP performance for detecting and grading hepatic steatosis using LB as reference standard [119, 120]. One of the most important meta-analysis dates from 2017 and includes 2735 patients (36.5% with HCV infection) [119]. Results are consistent, so that CAP provided an AUROC of 0.823 (Se = 68.8%, Sp = 82.2%) for detecting mild steatosis (≥S1), 0.865 (Se = 77.3%, Sp = 81.2%) for moderate steatosis (≥S2) and 0.882 (Se = 88.2%, Sp = 77.6%) for severe steatosis (≥S3) [119].

Concerning CHC, we cite 3 studies, one with 854 CHC patients, the other with 115 patients with chronic hepatitis, 76% of them being infected with HCV, and the latter with 201 patients with 118 (58.7%) subjects with HCV infection [121, 122, 123]. CAP had good diagnostic accuracy for detecting steatosis and for differentiating between different grades at least two grades apart, independently of fibrosis stage or activity grade. Optimal cutoff values were similar and are presented in Table 5. Further validation in large cohorts is however needed in order to validate proper cutoff values. CAP could be ideal as a screening test, as the NPV was high, 0.89–0.87 for ≥S1.

		Sasso et al. [121]	Ferraioli [122]	Lupsor-Platon [123]
	Patients number	854	115; (76% HCV pts)	201; (58.7% HCV pts)
≥S1	Optimal cutoff (dB/m)	222	219	260
	AUROC	0.80	0.76	0.813
≥S2	Optimal cutoff (dB/m)	233	296	285
	AUROC	0.86	0.82	0.822
≥S3	Optimal cutoff (dB/m)	290	N/S	294
	AUROC	0.88	N/S	0.838

Table 5.

Diagnostic performance of CAP in HCV-infected patients.

10. Perspectives

Since shear wave propagation spectroscopy can also provide additional mechanical information on soft tissues, such as viscosity, it might be possible to achieve additional data regarding the utility of 2D-SWE (SSI) for viscosity quantification, a potential marker for necroinflammatory activity [124]. Nonetheless, large cohort prospective studies are required in order to assess the performance of such parameters in biopsied HCV-infected patients.

11. Conclusions

Elastography-based imaging methods are of high interest nowadays. HCV patients can greatly benefit from VCTE due to its numerous qualities- rapid, noninvasive, repeatable for longitudinal evaluation, and cost-effectiveness. It has great discriminative power for fibrosis assessment, performing better at ruling out cirrhosis rather than diagnosing it, because of high specificity and negative predictive value. In addition, CAP is a precious tool for the noninvasive quantification of steatosis. Further validation in large cohorts is still needed in order to validate cutoff values in CHC patients. Among other elastographic techniques, pSWE and 2D-SWE proved to have the similar diagnostic performance to VCTE for the prediction of fibrosis severity in HCV-infected patients. One of the main advantages of non-invasive techniques is that they opened a new era in HCV management, since it can be easily executed when deemed necessary before antiviral therapy and after HCV eradication, as a repeatable surveillance method. Since the introduction of DAAs in HCV therapy, many patients achieve SVR, which is associated with a reduction in fibrosis. However, clinical practice guidelines do not currently recommend using elastography for the assessment of fibrosis decrease after treatment. Moreover, patients should continue surveillance for decompensation and HCC after SVR, regardless of the result of noninvasive methods.

It is essential that further studies focus on establishing standardized cutoff values of LS for adequate prediction of HCC risk in HCV patients, which is considered to be of great importance in current clinical practice.

Conflict of interest

The authors declare no conflict of interest.