Ultrasound in Med. & Biol., Vol. 40, No. 11, pp. 2556–2563, 2014 Copyright Ó 2014 World Federation for Ultrasound in Medicine & Biology Printed in the USA. All rights reserved 0301-5629/$ - see front matter
http://dx.doi.org/10.1016/j.ultrasmedbio.2014.05.011
d
Original Contribution HEPATIC PERFUSION PARAMETERS OF CONTRAST-ENHANCED ULTRASONOGRAPHY CORRELATE WITH THE SEVERITY OF CHRONIC LIVER DISEASE DONG LIU,* LINXUE QIAN,* JINRUI WANG,y XIANGDONG HU,* and LANYAN QIU* * Department of Ultrasound, Beijing Friendship Hospital, Capital Medical University, Beijing, China; and y Department of Ultrasound, Peking University Third Hospital, Beijing, China (Received 15 June 2013; revised 8 May 2014; in final form 14 May 2014)
Abstract—In the study described here, we introduced a new ratio acquired with contrast-enhanced ultrasonography (CEUS): a liver parenchyma blood supply ratio that differentiates arterial and portal phases. Our purpose was to determine whether this ratio and other liver parenchyma perfusion parameters acquired with CEUS can be correlated with the severity of chronic liver disease. Twelve patients with non-cirrhotic chronic liver disease, 35 patients with cirrhosis (child class A: n 5 10; child class B: n 5 13; child class C: n 5 12) and 21 healthy volunteers were examined by CEUS. Time–intensity curves were drawn for regions of interest located in liver parenchyma and right kidney cortex using QLAB quantification software. The arterial and portal phases were differentiated by the time to the maximum enhancement of right kidney and liver parenchyma perfusion data acquired from the time–intensity curves: the intensity of liver parenchyma perfused by hepatic arterial flow (Iap), the intensity of total perfusion of liver parenchyma (Ipeak), the intensity of liver parenchyma perfused by portal venous flow (Ipp) and the ratio of portal perfusion to total perfusion of liver parenchyma expressed by the parameters Ipp/Ipeak, Ipeak, Ipp and Ipp/Ipeak significantly decreased in patients with cirrhosis and in patients with non-cirrhotic chronic liver disease, whereas Iap increased. The parameters Ipp, Ipeak, Ipp/Ipeak and Iap correlated with the severity of chronic liver disease (r 5 20.938, p , 0.001; r 5 20.790, p , 0.001; r 5 20.931 p , 0.001; r 5 0.31, p , 0.05). The diagnostic accuracy rates for cirrhosis expressed as areas under receiver operating characteristic curves were 0.93 for Ipeak, 0.98 for Ipp, 0.98 for Ipp/Ipeak, and 0.69 for Iap. Liver parenchyma perfusion parameters obtained by CEUS were correlated with the severity of chronic liver disease and have the potential to assess cirrhosis noninvasively. (E-mail: [email protected]) Ó 2014 World Federation for Ultrasound in Medicine & Biology. Key Words: Contrast-enhanced ultrasonography, Cirrhosis, Hepatic perfusion.
invasive method for estimation of the severity of chronic liver disease is needed. During the progression of chronic liver disease, alterations involving the microvascular bed of the liver are already evident during the pre-cirrhotic stages of hepatic fibrogenesis (Ridolfi et al. 2012). The increase in intrahepatic vascular resistance decreases the portal fraction of liver perfusion (Rokey and Weisiger 1996). This decrease in portal perfusion is partially compensated by an increase in hepatic arterial flow (Eipel et al. 2010). Many imaging techniques have been used to evaluate hepatic perfusion, including computed tomography (CT) (Nakashige et al. 2004; Tsushima et al. 1999; Van Beers et al. 2001), magnetic resonance imaging (Hagiwara et al. 2008) and isotope scintigraphy (Iwasa et al. 1995; Ziegler et al. 1996). However, these techniques have disadvantages such as high cost and low resolution, which in turn leads to poor precision.
INTRODUCTION Liver cirrhosis is the final stage of progression of chronic liver disease and is common in China. Three percent of patients with compensated cirrhosis progress to decompensation annually. To determine adequate medical therapy and to prevent bleeding from esophageal varices, it is important to diagnose liver cirrhosis accurately and promptly. Needle biopsy of the liver is regarded as the gold standard for the diagnosis of cirrhosis. However, it is invasive and has poor reproducibility, with falsenegative rates ranging from 9.3% to 51% (Pagliaro et al. 1983; Zaitoun et al. 2001). Therefore, a non-
Address correspondence to: Linxue Qian, Department of Ultrasound, Beijing Friendship Hospital, Capital Medical University, 95 Yong’an Road, Beijing 100050, China. E-mail: qianlinxue2002@ 163.com 2556
Hepatic CEUS parameters correlate with disease severity d D. LIU et al.
Because it is non-invasive, inexpensive and reproducible, ultrasound (US) has become the preferred imaging modality for the diagnosis of chronic liver disease. However, conventional US is somewhat limited in the assessment chronic liver disease. As an important supplement to conventional ultrasound, contrast-enhanced ultrasound (CEUS) techniques have great potential in the evaluation of hepatic perfusion and can overcome some of the limitations. Many temporal indices have been found to be helpful in the diagnosis and evaluation of liver fibrosis and cirrhosis (Lim et al. 2005; Staub et al. 2009). However, to assess quantitative changes in different phases of hepatic perfusion in chronic liver disease, temporal indices are not enough. In the present study, we focused on the contrast enhancement phases in the time–intensity curve (TIC) of hepatic parenchyma, and used the intensity indices Iap, Ipp and Ipp/Ipeak and the conventional parameter Ipeak to assess chronic liver disease. METHODS Patients Forty-seven patients with biopsy-proven cirrhotic and non-cirrhotic chronic liver disease who were admitted to our hospital between October 2009 and April 2013 were enrolled in the study. Twenty-one healthy volunteers were enrolled in a normal group. The patients were divided into Child A, Child B and Child C groups using Child–Pugh criteria, which are based on serum bilirubin, serum albumin, prothrombin time (international normalized ratio [INR]) and the presence of encephalopathy, and ascites (Kim and Lee 2013; Pugh et al. 1973). Patients with any known disease that might influence the intra- or extrahepatic or renal circulation, such as focal liver lesions and cardiac, renal or portal diseases; patients who had undergone treatments such as transjugular intrahepatic portosystemic shunt (TIPS) and splenectomy; and patients who had taken antihypertensive drugs such as beta blockers during the preceding 2 wk were excluded from the study. The study protocol conformed to the guidelines outlined in the 1975 Declaration of Helsinki and was approved by the medical ethics committee of Beijing Friendship Hospital Affiliated to Capital Medical University. All patients gave written informed consent before being enrolled in the study. The baseline characteristics of the patients are summarized in Table 1. Materials The ultrasound contrast agent used was SonoVue (Bracco, Milan, Italy). The agent was prepared by mixing the powder with 5 mL 0.9% physiologic saline to form a suspension. Examinations were conducted with a Philips
2557
Table 1. Clinical data of subjects (n 5 68) Datum
Control group (n 5 21)
Non-cirrhotic Cirrhotic group (n 5 12) group (n 5 35)
Age (y) 39.63 6 14.10 48.06 6 7.80 Sex (male/female) 13/8 8/4
52.39 6 11.67 24/11
IU 22 ultrasound system (Philips, Amsterdam, The Netherlands) and a 2- to 5-MHz transducer. The imaging technology was pulse inverse harmonics imaging (PIHI). Procedures To minimize variation, the settings of the scanner such as mechanical index (MI) (the MI used in the present study was 0.05), dynamic range and gain and frequency were kept constant, and the time gain compensation was off. Before the contrast examination, every subject was instructed on breath holding and shallow breathing to minimize the variation caused by motion. With the patient lying with the left side down, the operator selected an appropriate slice that could clearly show part of the liver and part of the right kidney at the same time and then fixed the transducer. The contrast suspension, 2.4 mL, was injected through an antecubital vein using a 20G needle followed by a rapid flush of 5 mL saline (Claudon et al. 2008). A timer was started, and a real-time dynamic image storage system activated. As soon as the liver or kidney began to enhance, the patient was asked to hold his or her breath for at least 30 s and then to breathe softly. Images were recorded for 1.5 min and stored. The film was analyzed separately by two experienced doctors using QLAB quantification software (Philips). The intra-class correlation coefficients of the data obtained by the two doctors were analyzed to evaluate concordance. Data acquisition The liver is supplied by two vessels, the hepatic artery and the portal vein. Therefore, the TIC of liver parenchyma has two components. When the contrast agent from the hepatic artery arrives, the curve begins to rise, and when the contrast agent from the portal vein arrives soon afterward, the curve continues to rise until it peaks (Fig. 1). If we could separate the two components, we could separately quantify the changes in these two components. Because renal perfusion is similar to hepatic artery perfusion, we use the time to maximum enhancement of the right kidney to differentiate the arterial and portal phases (Fig. 2). This method was introduced by Miles et al. (1993) using CT. CEUS analysis We selected two regions of interest (ROIs) in the right kidney and the liver, respectively (Fig. 3), and
2558
Ultrasound in Medicine and Biology
Volume 40, Number 11, 2014
Fig. 3. Regions of interest (ROIs) of liver parenchyma and kidney cortex. The yellow circle represents the ROI of liver parenchyma, and the red circle represents the ROI of kidney cortex. Fig. 1. Schematic representation of the two composites of time–intensity curves of liver parenchyma.
obtained two TICs of the two ROIs in the same onecoordinate system (Fig. 4). The location of the ROI was limited to the midfield of the screen to minimize the variation caused by tissue attenuation. ROIs in the liver were selected in the middle of the liver parenchyma with major veins or arteries absent. ROIs in the kidney were selected in the right kidney cortex and were elliptical, with the long axis parallel to that of the kidney cortex. We analyzed data with the QLAB software (Philips, Amsterdam, The Netherlands) with which the Philips ultrasound system was equipped and drew curves of the original TIC fitted with the gamma variate function. Curves were fit using y(t) 5 A 3 (t–t0) 3 exp(2a 3 (t–t0)) 1 C, where A 5 ascending slope of the curve, a 5 declining slope of
Fig. 2. Schematic representation of the demarcation of the two composites of time–intensity curves of liver parenchyma and kidney cortex. Time–intensity curves of liver parenchyma and kidney cortex were drawn in the same coordinates. The yellow curve represents the liver parenchyma time–intensity curve, and the red curve, the kidney cortex time–intensity curve. The pink line is the vertical across the peak point of the kidney cortex time–intensity curve, and the point where it crosses the time– intensity curve of liver is f the demarcation of hepatic artery perfusion and portal vein perfusion.
the curve, C 5 initial decibel (dB) and t0 5 initial time of ascension. The quality of fit was .75%. The data were extracted from Exal after completing the curve. The point where one of the two curves began to rise was set as zero on time axis. The parameters we investigated and their significance were as follows: hepatic artery perfusion time (Tap), the same time to the peak of the kidney cortex TIC; time to maximum enhancement of hepatic parenchyma (Tpeak), that is, time to peak of liver parenchyma TIC; portal perfusion time (Tpp), that is, time between Tap and Tpeak, Tpp 5 Tpeak–Tap; intensity of hepatic artery perfusion (Iap), that is, intensity at Tap; intensity of total perfusion of hepatic parenchyma (Ipeak), that is, peak of the hepatic TIC; intensity of portal venous perfusion (Ipp), Ipp 5 Ipeak–Iap; relative portal perfusion (Ipp/Ipeak), that is, ratio of portal venous perfusion to total perfusion. Statistical analysis The ROIs were drawn by two investigators and the intra-class correlation coefficients were analyzed. Results are expressed as the mean 6 standard deviation. The liver perfusion parameters of the three patient groups were compared by one-way analysis of variance and then two-by-two by Fisher’s protected least significance test. The correlation between perfusion parameters and severity of chronic liver disease, which was classified into five levels (normal, non-cirrhotic chronic disease, Child classes A–C), was analyzed with Spearman’s rank correlation coefficients. The sensitivity and specificity of the perfusion parameters in diagnosing cirrhosis were analyzed by receiver operating characteristic (ROC) curves. Statistical tests were two-tailed. Results were considered significant at p , 0.05. RESULTS Administration of the ultrasound contrast agent was well tolerated in all participants, and no adverse effects were observed. Time–intensity curves of ROIs in the liver
Hepatic CEUS parameters correlate with disease severity d D. LIU et al.
2559
Fig. 4. Time–intensity curves (TICs) of liver parenchyma and kidney cortex of normal patients. The curves were drawn by QLAB analyzing software on a Philips IU22 instrument and fitted by Gamma-Variate. The yellow curve represents the TIC of liver parenchyma, and the red curve represents the TIC of kidney cortex. The smooth curves are fitting curves, and the unsmooth curves are the original curves.
parenchyma were acquired in all 47 patients and 21 healthy volunteers. In Figures 4 and 5 are typical TICs obtained for normal and cirrhotic patients. The shape of TICs in the cirrhotic group obviously differs from the shape in the control group (Figs. 4 and 5).
Because a good correlation (intra-class correlation coefficients of the measurements obtained by the two doctors were 0.94 for Iap, 0.93 for Ipeak, 0.93 for Ipp and 0.95 for Ipp/Ipeak) was obtained between the two doctors who analyzed the TIC curves, the data were averaged for further analysis.
Fig. 5. Time–intensity curves of liver parenchyma and kidney cortex of a patient with cirrhosis.
2560
Ultrasound in Medicine and Biology
Volume 40, Number 11, 2014
Table 2. Hepatic perfusion parameters Parameter Iap (dB) Ipeak (dB) Ipp (dB) Ipp/Ipeak
Normal group (n 5 21) 21.46 6 4.05 36.58 6 4.16 15.12 6 2.75 0.42 6 0.07
Non-cirrhotic liver disease group (n 5 12) 22.52 6 2.63 33.19 6 2.20y 10.83 6 2.00y 0.32 6 0.05y
Cirrhotic liver disease group (n 5 35) y
24.17 6 3.46 27.50 6 4.09y,z 3.50 6 2.38y,z 0.12 6 0.07y,z
p 0.02 ,0.0001 ,0.0001 ,0.0001
Values are expressed as the mean 6 standard deviation. The p-values in the last column were obtained by comparing three groups in the same row with a one-way analysis of variance. y p , 0.05 when compared with the normal group. z p , 0.05 when compared with the group of non-cirrhotic liver disease.
Hepatic perfusion parameters of the normal, noncirrhotic cirrhotic groups are provided in Table 2 and Figure 6. The Ipeak, Ipp and Ipp/Ipeak values of the cirrhotic group were decreased compared with those of the other groups (p , 0.0001), whereas Iap was increased (p , 0.05). Ipp, Ipeak, Ipp/Ipeak and Iap correlated with
the severity of chronic liver disease (r 5 –0.938, p , 0.001; r 5 –0.790, p , 0.001; r 5 –0.931 p , 0.001; and r 5 0.31, p , 0.05, respectively) (Table 3). In Figure 7 are the ROC curves of Ipeak, Ipp, Ipp/Ipeak and Iap. Areas under the ROC curves for Ipeak, Ipp, Ipp/Ipeak and Iap, were 0.93 6 0.03, 0.98 6 0.01,
Fig. 6. Box plots of perfusion parameters from the groups. The top and bottom of each box represent the 25th and 75th percentiles, giving the interquartile range. The line through the box indicates the median values, and outliers are represented as individual dots. (a–d) Box plots of Iap (a), Ipeak (b), Ipp (c) and Ipp/Ipeak (d). Iap, Ipeak, Ipp and Ipp/Ipeak values of the cirrhotic, normal and non-cirrhotic groups are included.
Hepatic CEUS parameters correlate with disease severity d D. LIU et al.
2561
Table 3. Correlations between liver perfusion parameters and severity of chronic liver disease* Disease severity Perfusion parameter Iap (dB) Ipeak (dB) Ipp (dB) Ipp/Ipeak
Normal (n 5 21)
Non-cirrhotic liver disease (n 5 12)
Child A (n 5 10)
Child B (n 5 13)
Child C (n 5 12)
r
p
21.46 6 4.05 36.58 6 4.16 15.12 6 2.75 0.42 6 0.07
22.52 6 2.63 33.19 6 2.20 10.83 6 2.00 0.32 6 0.05
24.37 6 3.09 30.25 6 3.31 6.46 6 2.40 0.21 6 0.07
24.10 6 3.55 26.84 6 4.03 2.87 6 0.80 0.11 6 0.02
24.22 6 3.89 25.93 6 3.84 1.71 6 0.47 0.07 6 0.02
0.31 20.79 20.94 20.93
0.011 ,0.001 ,0.001 ,0.001
* Each row gives the correlation (r) and significance of the correlation (p) between the five classes of disease severity and the given hepatic perfusion parameter. Data are expressed as the mean 6 standard deviation. Child A, Child B and Child C refer to the Child–Pugh classification.
0.98 6 0.02 and 069 6 0.07 respectively. The cutoff values for the sensitivity and specificity of Ipeak, Ipp, Ipp/ Ipeak and Iap in predicting cirrhosis are listed in Table 4.
DISCUSSION Contrast-enhanced ultrasonography has a unique advantage in measuring blood perfusion. Under certain conditions, the microbubble contrast echo intensity is proportional to the amount of blood perfusion. However, many factors can affect CEUS quantitative analysis including the frequency of the transducer, the mechanical index and the analytical software (Tang et al. 2011). Therefore, in the present study, to avoid these complicating factors, we focused on the investigation of relative portal perfusion, which is a ratio. Use of this ratio can eliminate the contribution of many confounding factors such as differences in device settings and the analytical software used. In this way, the ratio can be used as a normalization method for comparison of results from various institutions.
The liver is an organ rich in vessels that are formed from a series of porous vascular channels. The progression from normal to cirrhotic liver is accompanied by hemodynamic changes, including sinusoidal capillarization, deposition of fibers by activated hepatic stellate or antigen presenting cells, intrahepatic shunts between the branches of the hepatic vessels and so on. These lead to increased resistance of the liver sinusoids (Bauer et al. 1995; Orrego et al. 1981; Rokey and Weisiger 1996). The increase in intrahepatic vascular resistance results in a decrease in portal venous perfusion. Therefore, Ipp was expected to be lower in the cirrhotic group. Normally, the blood from the portal vein constitutes 70%–75% of the total flow, and that from the hepatic artery, 25%–30% (Schenk et al. 1962; Villeneuve et al. 1996). In the physiologic state, hepatic artery blood flow changes inversely according to the change in portal vein blood flow to compensate for changes in portal perfusion and maintain total perfusion. This compensatory mechanism has been termed the hepatic artery buffer reaction (HABR) (Eipel et al. 2010). In
Fig. 7. Receiver operating characteristic (ROC) curves of Ipeak, Ipp, Ipp/Ipeak (a) and Iap (b) predicting liver cirrhosis in the estimation set.
2562
Ultrasound in Medicine and Biology
Table 4. Accuracy in diagnosing cirrhosis Parameter
Cutoff value
Iap Ipeak Ipp Ipp/Ipeak
22.86 32.205 dB 10.245 dB 0.285
Sensitivity (%)
Specificity (%)
77.1 91.4 94.3 97.1
30.3 75.8 78.8 84.8
cirrhotic patients, when portal vein perfusion decreases, hepatic artery blood increases because of the HABR (Aoki et al. 2005). Therefore, Iap was expected to be higher in the cirrhotic group than in the other groups. The total perfusion of hepatic parenchyma is the sum of hepatic artery perfusion and portal venous perfusion, so the change in Ipeak depends on the reduction in portal venous perfusion and the compensation of hepatic artery perfusion. In previous studies of perfusion in cirrhosis, results on changes in total perfusion differed. Kaneko et al. (2005) reported that there was a significant inverse correlation between the gray scale of the liver parenchyma and the hepatic fibrosis index. Orlacchio et al. (2011) also observed that there was a significant inverse correlation between peak signal intensity and fibrosis scores. However, Ridolfi et al. (2012) reported that the peak enhancement in patients with liver cirrhosis was higher than that observed in controls. In the present study, Ipeak was slightly lower in the cirrhotic group than in the normal group, and the difference was significant. That may have been due to the higher proportion of Child B and C patients and lower proportion of Child A patients. Indeed, some studies reported that HABR was associated with Child grading and the level of portal hypertension (Aoki et al. 2005; G€ ulberg et al. 2002; Li et al. 1998; Zhang et al. 2011). The relative portal perfusion parameter Ipp/Ipeak is a ratio influenced by Ipp and Ipeak. The decrease in Ipp was greater than the decrease in Ipeak in cirrhotic patients, so the ratio Ipp/Ipeak was expected to be lower in cirrhotic group. The results of the present study were consistent with the expectations. In the present study, we used a unique method (time to the peak of the kidney time–intensity curve) to differentiate the arterial and portal phases and measure the blood supply ratio (portal vein/[hepatic artery 1 portal vein]). We also introduced a new parameter, Ipp/Ipeak, to assess chronic liver disease. Although Iap, Ipeak, Ipp and Ipp/Ipeak have a good correlation with the severity of chronic liver disease, they do not really reflect arterial and portal venous perfusion. They are only approximately proportional to the amount of blood perfusion and, therefore, are only rough surrogate markers. On the other hand, with the use of Ipp/Ipeak, it is quite simple to eliminate the contribution of many confounding fac-
Volume 40, Number 11, 2014
tors. This parameter has the potential to be useful in clinical practice. CONCLUSIONS The current data indicate that the hepatic CEUS parameters Iap, Ipp, Ipeak and Ipp/Ipeak are significantly correlated with the severity of chronic liver disease. Ipp/Ipeak had high sensitivity and specificity and, thus, was of good diagnostic value. The data suggest that hepatic perfusion parameters acquired with CEUS may be useful in assessing the severity of chronic liver disease and predicting cirrhosis non-invasively. REFERENCES Aoki T, Imamura H, Kaneko J, Sakamoto Y, Matsuyama Y, Kokudo N, Sugawara Y, Makuuchi M. Intraoperative direct measurement of hepatic arterial buffer response in patients with or without cirrhosis. Liver Transpl 2005;11:684–691. Bauer M, Paquette NC, Zhang JX, Bauer I, Pannen BH, Kleeberger SR, Clemens MG. Chronic ethanol consumption increases hepatic sinusoidal contractile response to endothlin-1 in the rat. Hepatology 1995;22:1565–1576. Claudon M, Cosgrove D, Albrecht T, Bolondi L, Bosio M, Calliada F, Correas JM, Darge K, Dietrich C, D’Onofrio M, Evans DH, Filice C, Greiner L, J€ager K, Jong ND, Leen E, Lencioni R, Lindsell D, Martegani A, Meairs S, Nolsøe C, Piscaglia F, Ricci P, Seidel G, Skjoldbye B, Solbiati L, Thorelius L, Tranquart F, Weskott HP, Whittingham T. Guidelines and good clinical practice recommendations for contrast enhanced ultrasound (CEUS)— Update 2008. Ultraschall Med 2008;29:28–44. Eipel C, Abshagen K, Vollmar B. Regulation of hepatic blood flow: The hepatic arterial buffer response revisited. World J Gastroenterol 2010;16:6046–6057. G€ulberg V, Haag K, R€ossle M, Gerbes AL. Hepatic arterial buffer response in patients with advanced cirrhosis. Hepatology 2002;35: 630–634. Hagiwara M, Rusinek H, Lee VS, Losada M, Bannan MA, Krinsky GA, Taouli B. Advanced liver fibrosis: Diagnosis with 3-D whole-liver perfusion MR imaging—Initial experience. Radiology 2008;246: 926–934. Iwasa M, Nakamura K, Nakagawa T, Watanabe S, Katoh H, Kinosada Y, Maeda H, Habara J, Suzuki S. Single photon emission computed tomography to determine effective hepatic blood flow and intrahepatic shunting. Hepatology 1995;21:359–365. Kaneko T, Teshigawara O, Sugimoto H, Hirota M, Inoue S, Takeda S, Nakao A. Signal intensity of the liver parenchyma in microbubble contrast agent in the late liver phase reflects advanced fibrosis of the liver. Liver Int 2005;25:288–293. Kim HJ, Lee HW. Important predictor of mortality in patients with endstage liver disease. Clin Mol Hepatol 2013;19:105–115. Li X, Benjamin IS, Alexander B. The relationship between intrahepatic portal systemic shunts and microsphere induced portal hypertension in the rat liver. Gut 1998;42:276–282. Lim AK, Taylor-Robinson SD, Patel N, Eckersley RJ, Goldin RD, Hamilton G, Foster GR, Thomas HC, Cosgrove DO, Blomley MJK. Hepatic vein transit times using a microbubble agent can predict disease severity non-invasively in patients with hepatitis C. Gut 2005;54:128–133. Miles KA, Hayball MP, Dixon AK. Functional images of hepatic perfusion obtained with dynamic CT. Radiology 1993;188:405–411. Nakashige A, Horiguchi J, Tamura A, Asahara T, Shimamoto F, Ito K. Quantitative measurement of hepatic portal perfusion by multidetector row CT with compensation for respiratory misregistration. Br J Radiol 2004;77:728–734. Orlacchio A, Bolacchi F, Petrella MC, Pastorelli D, Bazzocchi G, Angelico M, Simonetti G. Liver contrast enhanced ultrasound
Hepatic CEUS parameters correlate with disease severity d D. LIU et al. perfusion imaging in the evaluation of chronic hepatitis C fibrosis: Preliminary results. Ultrasound Med Biol 2011;37:1–6. Orrego H, Blendis LM, Crossley IR, Medline A, Macdonald A, Ritchie S, Israel Y. Correlation of intrahepatic pressure with collagen in the Disse space and hepatomegaly in humans and in the rat. Gastroenterology 1981;80:546–556. Pagliaro L, Rinaldi F, Craxı A, Di Piazza S, Filippazzo G, Gatto G, Genova G, Magrin S, Maringhini A, Orsini S, Palazzo U, Spinello M, Vinci M. Percutaneous blind biopsy versus laparoscopy with guided biopsy in diagnosis of cirrhosis: A prospective, randomized trial. Dig Dis Sci 1983;28:39–43. Pugh RN, Murray-Lyon IM, Dawson JL, Pietroni MC, Williams R. Transection of the oesophagus for bleeding oesophageal varices. Br J Surg 1973;60:646–649. Ridolfi F, Abbattista T, Busilacchi P, Brunelli E. Contrast-enhanced ultrasound evaluation of hepatic micro-vascular changes in liver diseases. World J Gastroenterol 2012;18:5225–5230. Rokey DC, Weisiger RA. Endothelin induced contractility of stellate cells from normal and cirrhotic rat liver: Implications for regulation of portal pressure and resistance. Hepatology 1996;24:233–240. Schenk WG Jr, McDonald JC, McDonald K, Drapanas T. Direct measurement of hepatic blood flow in surgical patients: With related observations on hepatic flow dynamics in experimental animals. Ann Surg 1962;156:463–471. Staub F, Tournoux-Facon C, Roumy J, Chaigneau C, Morichaut-Beauchant M, Levillain P, Prevost C, Aube C, Lebigot J, Oberti F, Galtier JB, Laumonier H, Trillaud H, Bernard PH, Blanc JF, Sironneau S, Machet F, Drouillard J, de Ledinghen V, Couzigou P, Foucher P, Castera L, Tranquard F, Bacq Y, de Alteroche L, Ingrand P, Tasu JP. Liver fibrosis staging with contrast-enhanced ultrasonography: Prospective multicenter
2563
study compared with METAVIR scoring. Eur Radiol 2009;19: 1991–1997. Tang MX, Mulvana H, Gauthier T, Lim AK, Cosgrove DO, Eckersley RJ, Stride E. Quantitative contrast-enhanced ultrasound imaging: A review of sources of variability. Interface Focus 2011; 1:520–539. Tsushima Y, Blomley JK, Kusano S, Endo K. The portal component of hepatic perfusion measured by dynamic CT: An indicator of hepatic parenchymal damage. Dig Dis Sci 1999;44:1632–1638. Van Beers BE, Leconte I, Materne R, Smith AM, Jamart J, Horsmans Y. Hepatic perfusion parameters in chronic liver disease: Dynamic CT measurements correlated with disease severity. AJR Am J Roentgenol 2001;176:667–673. Villeneuve JP, Dagenais M, Huet PM, Roy A, Lapointe R, Marleau D. The hepatic microcirculation in the isolated perfused human liver. Hepatology 1996;23:24–31. Zaitoun A, AI M, Mardini H, Awad S, Ukabam S, Makadisi S, Record CO. Quantitative assessment of fibrosis and steatosis in the liver biopsies from patients with chronic hepatitis C. J Clin Pathol 2001;54:461–465. Zhang L, Yin J, Duan Y, Yang Y, Yuan L, Cao T. Assessment of intrahepatic blood flow by Doppler ultrasonography: Relationship between the hepatic vein, portal vein, hepatic artery and portal pressure measured intra-operatively in patients with portal hypertension. BMC Gastroenterol 2011;11:84. Ziegler SI, Haberkorn U, Byrne H, Tong C, Kaja S, Richolt JA, Byrne H, Tong C, Schosser R, Krieter H, Kaja S, Richolt JA, Lammertsma AA, Price P. Measurement of liver blood flow using oxygen-15 labelled water and dynamic positron emission tomography: Limitation of model description. Eur J Nucl Med 1996;23: 169–177.
http://dx.doi.org/10.1016/j.ultrasmedbio.2014.05.011
d
Original Contribution HEPATIC PERFUSION PARAMETERS OF CONTRAST-ENHANCED ULTRASONOGRAPHY CORRELATE WITH THE SEVERITY OF CHRONIC LIVER DISEASE DONG LIU,* LINXUE QIAN,* JINRUI WANG,y XIANGDONG HU,* and LANYAN QIU* * Department of Ultrasound, Beijing Friendship Hospital, Capital Medical University, Beijing, China; and y Department of Ultrasound, Peking University Third Hospital, Beijing, China (Received 15 June 2013; revised 8 May 2014; in final form 14 May 2014)
Abstract—In the study described here, we introduced a new ratio acquired with contrast-enhanced ultrasonography (CEUS): a liver parenchyma blood supply ratio that differentiates arterial and portal phases. Our purpose was to determine whether this ratio and other liver parenchyma perfusion parameters acquired with CEUS can be correlated with the severity of chronic liver disease. Twelve patients with non-cirrhotic chronic liver disease, 35 patients with cirrhosis (child class A: n 5 10; child class B: n 5 13; child class C: n 5 12) and 21 healthy volunteers were examined by CEUS. Time–intensity curves were drawn for regions of interest located in liver parenchyma and right kidney cortex using QLAB quantification software. The arterial and portal phases were differentiated by the time to the maximum enhancement of right kidney and liver parenchyma perfusion data acquired from the time–intensity curves: the intensity of liver parenchyma perfused by hepatic arterial flow (Iap), the intensity of total perfusion of liver parenchyma (Ipeak), the intensity of liver parenchyma perfused by portal venous flow (Ipp) and the ratio of portal perfusion to total perfusion of liver parenchyma expressed by the parameters Ipp/Ipeak, Ipeak, Ipp and Ipp/Ipeak significantly decreased in patients with cirrhosis and in patients with non-cirrhotic chronic liver disease, whereas Iap increased. The parameters Ipp, Ipeak, Ipp/Ipeak and Iap correlated with the severity of chronic liver disease (r 5 20.938, p , 0.001; r 5 20.790, p , 0.001; r 5 20.931 p , 0.001; r 5 0.31, p , 0.05). The diagnostic accuracy rates for cirrhosis expressed as areas under receiver operating characteristic curves were 0.93 for Ipeak, 0.98 for Ipp, 0.98 for Ipp/Ipeak, and 0.69 for Iap. Liver parenchyma perfusion parameters obtained by CEUS were correlated with the severity of chronic liver disease and have the potential to assess cirrhosis noninvasively. (E-mail: [email protected]) Ó 2014 World Federation for Ultrasound in Medicine & Biology. Key Words: Contrast-enhanced ultrasonography, Cirrhosis, Hepatic perfusion.
invasive method for estimation of the severity of chronic liver disease is needed. During the progression of chronic liver disease, alterations involving the microvascular bed of the liver are already evident during the pre-cirrhotic stages of hepatic fibrogenesis (Ridolfi et al. 2012). The increase in intrahepatic vascular resistance decreases the portal fraction of liver perfusion (Rokey and Weisiger 1996). This decrease in portal perfusion is partially compensated by an increase in hepatic arterial flow (Eipel et al. 2010). Many imaging techniques have been used to evaluate hepatic perfusion, including computed tomography (CT) (Nakashige et al. 2004; Tsushima et al. 1999; Van Beers et al. 2001), magnetic resonance imaging (Hagiwara et al. 2008) and isotope scintigraphy (Iwasa et al. 1995; Ziegler et al. 1996). However, these techniques have disadvantages such as high cost and low resolution, which in turn leads to poor precision.
INTRODUCTION Liver cirrhosis is the final stage of progression of chronic liver disease and is common in China. Three percent of patients with compensated cirrhosis progress to decompensation annually. To determine adequate medical therapy and to prevent bleeding from esophageal varices, it is important to diagnose liver cirrhosis accurately and promptly. Needle biopsy of the liver is regarded as the gold standard for the diagnosis of cirrhosis. However, it is invasive and has poor reproducibility, with falsenegative rates ranging from 9.3% to 51% (Pagliaro et al. 1983; Zaitoun et al. 2001). Therefore, a non-
Address correspondence to: Linxue Qian, Department of Ultrasound, Beijing Friendship Hospital, Capital Medical University, 95 Yong’an Road, Beijing 100050, China. E-mail: qianlinxue2002@ 163.com 2556
Hepatic CEUS parameters correlate with disease severity d D. LIU et al.
Because it is non-invasive, inexpensive and reproducible, ultrasound (US) has become the preferred imaging modality for the diagnosis of chronic liver disease. However, conventional US is somewhat limited in the assessment chronic liver disease. As an important supplement to conventional ultrasound, contrast-enhanced ultrasound (CEUS) techniques have great potential in the evaluation of hepatic perfusion and can overcome some of the limitations. Many temporal indices have been found to be helpful in the diagnosis and evaluation of liver fibrosis and cirrhosis (Lim et al. 2005; Staub et al. 2009). However, to assess quantitative changes in different phases of hepatic perfusion in chronic liver disease, temporal indices are not enough. In the present study, we focused on the contrast enhancement phases in the time–intensity curve (TIC) of hepatic parenchyma, and used the intensity indices Iap, Ipp and Ipp/Ipeak and the conventional parameter Ipeak to assess chronic liver disease. METHODS Patients Forty-seven patients with biopsy-proven cirrhotic and non-cirrhotic chronic liver disease who were admitted to our hospital between October 2009 and April 2013 were enrolled in the study. Twenty-one healthy volunteers were enrolled in a normal group. The patients were divided into Child A, Child B and Child C groups using Child–Pugh criteria, which are based on serum bilirubin, serum albumin, prothrombin time (international normalized ratio [INR]) and the presence of encephalopathy, and ascites (Kim and Lee 2013; Pugh et al. 1973). Patients with any known disease that might influence the intra- or extrahepatic or renal circulation, such as focal liver lesions and cardiac, renal or portal diseases; patients who had undergone treatments such as transjugular intrahepatic portosystemic shunt (TIPS) and splenectomy; and patients who had taken antihypertensive drugs such as beta blockers during the preceding 2 wk were excluded from the study. The study protocol conformed to the guidelines outlined in the 1975 Declaration of Helsinki and was approved by the medical ethics committee of Beijing Friendship Hospital Affiliated to Capital Medical University. All patients gave written informed consent before being enrolled in the study. The baseline characteristics of the patients are summarized in Table 1. Materials The ultrasound contrast agent used was SonoVue (Bracco, Milan, Italy). The agent was prepared by mixing the powder with 5 mL 0.9% physiologic saline to form a suspension. Examinations were conducted with a Philips
2557
Table 1. Clinical data of subjects (n 5 68) Datum
Control group (n 5 21)
Non-cirrhotic Cirrhotic group (n 5 12) group (n 5 35)
Age (y) 39.63 6 14.10 48.06 6 7.80 Sex (male/female) 13/8 8/4
52.39 6 11.67 24/11
IU 22 ultrasound system (Philips, Amsterdam, The Netherlands) and a 2- to 5-MHz transducer. The imaging technology was pulse inverse harmonics imaging (PIHI). Procedures To minimize variation, the settings of the scanner such as mechanical index (MI) (the MI used in the present study was 0.05), dynamic range and gain and frequency were kept constant, and the time gain compensation was off. Before the contrast examination, every subject was instructed on breath holding and shallow breathing to minimize the variation caused by motion. With the patient lying with the left side down, the operator selected an appropriate slice that could clearly show part of the liver and part of the right kidney at the same time and then fixed the transducer. The contrast suspension, 2.4 mL, was injected through an antecubital vein using a 20G needle followed by a rapid flush of 5 mL saline (Claudon et al. 2008). A timer was started, and a real-time dynamic image storage system activated. As soon as the liver or kidney began to enhance, the patient was asked to hold his or her breath for at least 30 s and then to breathe softly. Images were recorded for 1.5 min and stored. The film was analyzed separately by two experienced doctors using QLAB quantification software (Philips). The intra-class correlation coefficients of the data obtained by the two doctors were analyzed to evaluate concordance. Data acquisition The liver is supplied by two vessels, the hepatic artery and the portal vein. Therefore, the TIC of liver parenchyma has two components. When the contrast agent from the hepatic artery arrives, the curve begins to rise, and when the contrast agent from the portal vein arrives soon afterward, the curve continues to rise until it peaks (Fig. 1). If we could separate the two components, we could separately quantify the changes in these two components. Because renal perfusion is similar to hepatic artery perfusion, we use the time to maximum enhancement of the right kidney to differentiate the arterial and portal phases (Fig. 2). This method was introduced by Miles et al. (1993) using CT. CEUS analysis We selected two regions of interest (ROIs) in the right kidney and the liver, respectively (Fig. 3), and
2558
Ultrasound in Medicine and Biology
Volume 40, Number 11, 2014
Fig. 3. Regions of interest (ROIs) of liver parenchyma and kidney cortex. The yellow circle represents the ROI of liver parenchyma, and the red circle represents the ROI of kidney cortex. Fig. 1. Schematic representation of the two composites of time–intensity curves of liver parenchyma.
obtained two TICs of the two ROIs in the same onecoordinate system (Fig. 4). The location of the ROI was limited to the midfield of the screen to minimize the variation caused by tissue attenuation. ROIs in the liver were selected in the middle of the liver parenchyma with major veins or arteries absent. ROIs in the kidney were selected in the right kidney cortex and were elliptical, with the long axis parallel to that of the kidney cortex. We analyzed data with the QLAB software (Philips, Amsterdam, The Netherlands) with which the Philips ultrasound system was equipped and drew curves of the original TIC fitted with the gamma variate function. Curves were fit using y(t) 5 A 3 (t–t0) 3 exp(2a 3 (t–t0)) 1 C, where A 5 ascending slope of the curve, a 5 declining slope of
Fig. 2. Schematic representation of the demarcation of the two composites of time–intensity curves of liver parenchyma and kidney cortex. Time–intensity curves of liver parenchyma and kidney cortex were drawn in the same coordinates. The yellow curve represents the liver parenchyma time–intensity curve, and the red curve, the kidney cortex time–intensity curve. The pink line is the vertical across the peak point of the kidney cortex time–intensity curve, and the point where it crosses the time– intensity curve of liver is f the demarcation of hepatic artery perfusion and portal vein perfusion.
the curve, C 5 initial decibel (dB) and t0 5 initial time of ascension. The quality of fit was .75%. The data were extracted from Exal after completing the curve. The point where one of the two curves began to rise was set as zero on time axis. The parameters we investigated and their significance were as follows: hepatic artery perfusion time (Tap), the same time to the peak of the kidney cortex TIC; time to maximum enhancement of hepatic parenchyma (Tpeak), that is, time to peak of liver parenchyma TIC; portal perfusion time (Tpp), that is, time between Tap and Tpeak, Tpp 5 Tpeak–Tap; intensity of hepatic artery perfusion (Iap), that is, intensity at Tap; intensity of total perfusion of hepatic parenchyma (Ipeak), that is, peak of the hepatic TIC; intensity of portal venous perfusion (Ipp), Ipp 5 Ipeak–Iap; relative portal perfusion (Ipp/Ipeak), that is, ratio of portal venous perfusion to total perfusion. Statistical analysis The ROIs were drawn by two investigators and the intra-class correlation coefficients were analyzed. Results are expressed as the mean 6 standard deviation. The liver perfusion parameters of the three patient groups were compared by one-way analysis of variance and then two-by-two by Fisher’s protected least significance test. The correlation between perfusion parameters and severity of chronic liver disease, which was classified into five levels (normal, non-cirrhotic chronic disease, Child classes A–C), was analyzed with Spearman’s rank correlation coefficients. The sensitivity and specificity of the perfusion parameters in diagnosing cirrhosis were analyzed by receiver operating characteristic (ROC) curves. Statistical tests were two-tailed. Results were considered significant at p , 0.05. RESULTS Administration of the ultrasound contrast agent was well tolerated in all participants, and no adverse effects were observed. Time–intensity curves of ROIs in the liver
Hepatic CEUS parameters correlate with disease severity d D. LIU et al.
2559
Fig. 4. Time–intensity curves (TICs) of liver parenchyma and kidney cortex of normal patients. The curves were drawn by QLAB analyzing software on a Philips IU22 instrument and fitted by Gamma-Variate. The yellow curve represents the TIC of liver parenchyma, and the red curve represents the TIC of kidney cortex. The smooth curves are fitting curves, and the unsmooth curves are the original curves.
parenchyma were acquired in all 47 patients and 21 healthy volunteers. In Figures 4 and 5 are typical TICs obtained for normal and cirrhotic patients. The shape of TICs in the cirrhotic group obviously differs from the shape in the control group (Figs. 4 and 5).
Because a good correlation (intra-class correlation coefficients of the measurements obtained by the two doctors were 0.94 for Iap, 0.93 for Ipeak, 0.93 for Ipp and 0.95 for Ipp/Ipeak) was obtained between the two doctors who analyzed the TIC curves, the data were averaged for further analysis.
Fig. 5. Time–intensity curves of liver parenchyma and kidney cortex of a patient with cirrhosis.
2560
Ultrasound in Medicine and Biology
Volume 40, Number 11, 2014
Table 2. Hepatic perfusion parameters Parameter Iap (dB) Ipeak (dB) Ipp (dB) Ipp/Ipeak
Normal group (n 5 21) 21.46 6 4.05 36.58 6 4.16 15.12 6 2.75 0.42 6 0.07
Non-cirrhotic liver disease group (n 5 12) 22.52 6 2.63 33.19 6 2.20y 10.83 6 2.00y 0.32 6 0.05y
Cirrhotic liver disease group (n 5 35) y
24.17 6 3.46 27.50 6 4.09y,z 3.50 6 2.38y,z 0.12 6 0.07y,z
p 0.02 ,0.0001 ,0.0001 ,0.0001
Values are expressed as the mean 6 standard deviation. The p-values in the last column were obtained by comparing three groups in the same row with a one-way analysis of variance. y p , 0.05 when compared with the normal group. z p , 0.05 when compared with the group of non-cirrhotic liver disease.
Hepatic perfusion parameters of the normal, noncirrhotic cirrhotic groups are provided in Table 2 and Figure 6. The Ipeak, Ipp and Ipp/Ipeak values of the cirrhotic group were decreased compared with those of the other groups (p , 0.0001), whereas Iap was increased (p , 0.05). Ipp, Ipeak, Ipp/Ipeak and Iap correlated with
the severity of chronic liver disease (r 5 –0.938, p , 0.001; r 5 –0.790, p , 0.001; r 5 –0.931 p , 0.001; and r 5 0.31, p , 0.05, respectively) (Table 3). In Figure 7 are the ROC curves of Ipeak, Ipp, Ipp/Ipeak and Iap. Areas under the ROC curves for Ipeak, Ipp, Ipp/Ipeak and Iap, were 0.93 6 0.03, 0.98 6 0.01,
Fig. 6. Box plots of perfusion parameters from the groups. The top and bottom of each box represent the 25th and 75th percentiles, giving the interquartile range. The line through the box indicates the median values, and outliers are represented as individual dots. (a–d) Box plots of Iap (a), Ipeak (b), Ipp (c) and Ipp/Ipeak (d). Iap, Ipeak, Ipp and Ipp/Ipeak values of the cirrhotic, normal and non-cirrhotic groups are included.
Hepatic CEUS parameters correlate with disease severity d D. LIU et al.
2561
Table 3. Correlations between liver perfusion parameters and severity of chronic liver disease* Disease severity Perfusion parameter Iap (dB) Ipeak (dB) Ipp (dB) Ipp/Ipeak
Normal (n 5 21)
Non-cirrhotic liver disease (n 5 12)
Child A (n 5 10)
Child B (n 5 13)
Child C (n 5 12)
r
p
21.46 6 4.05 36.58 6 4.16 15.12 6 2.75 0.42 6 0.07
22.52 6 2.63 33.19 6 2.20 10.83 6 2.00 0.32 6 0.05
24.37 6 3.09 30.25 6 3.31 6.46 6 2.40 0.21 6 0.07
24.10 6 3.55 26.84 6 4.03 2.87 6 0.80 0.11 6 0.02
24.22 6 3.89 25.93 6 3.84 1.71 6 0.47 0.07 6 0.02
0.31 20.79 20.94 20.93
0.011 ,0.001 ,0.001 ,0.001
* Each row gives the correlation (r) and significance of the correlation (p) between the five classes of disease severity and the given hepatic perfusion parameter. Data are expressed as the mean 6 standard deviation. Child A, Child B and Child C refer to the Child–Pugh classification.
0.98 6 0.02 and 069 6 0.07 respectively. The cutoff values for the sensitivity and specificity of Ipeak, Ipp, Ipp/ Ipeak and Iap in predicting cirrhosis are listed in Table 4.
DISCUSSION Contrast-enhanced ultrasonography has a unique advantage in measuring blood perfusion. Under certain conditions, the microbubble contrast echo intensity is proportional to the amount of blood perfusion. However, many factors can affect CEUS quantitative analysis including the frequency of the transducer, the mechanical index and the analytical software (Tang et al. 2011). Therefore, in the present study, to avoid these complicating factors, we focused on the investigation of relative portal perfusion, which is a ratio. Use of this ratio can eliminate the contribution of many confounding factors such as differences in device settings and the analytical software used. In this way, the ratio can be used as a normalization method for comparison of results from various institutions.
The liver is an organ rich in vessels that are formed from a series of porous vascular channels. The progression from normal to cirrhotic liver is accompanied by hemodynamic changes, including sinusoidal capillarization, deposition of fibers by activated hepatic stellate or antigen presenting cells, intrahepatic shunts between the branches of the hepatic vessels and so on. These lead to increased resistance of the liver sinusoids (Bauer et al. 1995; Orrego et al. 1981; Rokey and Weisiger 1996). The increase in intrahepatic vascular resistance results in a decrease in portal venous perfusion. Therefore, Ipp was expected to be lower in the cirrhotic group. Normally, the blood from the portal vein constitutes 70%–75% of the total flow, and that from the hepatic artery, 25%–30% (Schenk et al. 1962; Villeneuve et al. 1996). In the physiologic state, hepatic artery blood flow changes inversely according to the change in portal vein blood flow to compensate for changes in portal perfusion and maintain total perfusion. This compensatory mechanism has been termed the hepatic artery buffer reaction (HABR) (Eipel et al. 2010). In
Fig. 7. Receiver operating characteristic (ROC) curves of Ipeak, Ipp, Ipp/Ipeak (a) and Iap (b) predicting liver cirrhosis in the estimation set.
2562
Ultrasound in Medicine and Biology
Table 4. Accuracy in diagnosing cirrhosis Parameter
Cutoff value
Iap Ipeak Ipp Ipp/Ipeak
22.86 32.205 dB 10.245 dB 0.285
Sensitivity (%)
Specificity (%)
77.1 91.4 94.3 97.1
30.3 75.8 78.8 84.8
cirrhotic patients, when portal vein perfusion decreases, hepatic artery blood increases because of the HABR (Aoki et al. 2005). Therefore, Iap was expected to be higher in the cirrhotic group than in the other groups. The total perfusion of hepatic parenchyma is the sum of hepatic artery perfusion and portal venous perfusion, so the change in Ipeak depends on the reduction in portal venous perfusion and the compensation of hepatic artery perfusion. In previous studies of perfusion in cirrhosis, results on changes in total perfusion differed. Kaneko et al. (2005) reported that there was a significant inverse correlation between the gray scale of the liver parenchyma and the hepatic fibrosis index. Orlacchio et al. (2011) also observed that there was a significant inverse correlation between peak signal intensity and fibrosis scores. However, Ridolfi et al. (2012) reported that the peak enhancement in patients with liver cirrhosis was higher than that observed in controls. In the present study, Ipeak was slightly lower in the cirrhotic group than in the normal group, and the difference was significant. That may have been due to the higher proportion of Child B and C patients and lower proportion of Child A patients. Indeed, some studies reported that HABR was associated with Child grading and the level of portal hypertension (Aoki et al. 2005; G€ ulberg et al. 2002; Li et al. 1998; Zhang et al. 2011). The relative portal perfusion parameter Ipp/Ipeak is a ratio influenced by Ipp and Ipeak. The decrease in Ipp was greater than the decrease in Ipeak in cirrhotic patients, so the ratio Ipp/Ipeak was expected to be lower in cirrhotic group. The results of the present study were consistent with the expectations. In the present study, we used a unique method (time to the peak of the kidney time–intensity curve) to differentiate the arterial and portal phases and measure the blood supply ratio (portal vein/[hepatic artery 1 portal vein]). We also introduced a new parameter, Ipp/Ipeak, to assess chronic liver disease. Although Iap, Ipeak, Ipp and Ipp/Ipeak have a good correlation with the severity of chronic liver disease, they do not really reflect arterial and portal venous perfusion. They are only approximately proportional to the amount of blood perfusion and, therefore, are only rough surrogate markers. On the other hand, with the use of Ipp/Ipeak, it is quite simple to eliminate the contribution of many confounding fac-
Volume 40, Number 11, 2014
tors. This parameter has the potential to be useful in clinical practice. CONCLUSIONS The current data indicate that the hepatic CEUS parameters Iap, Ipp, Ipeak and Ipp/Ipeak are significantly correlated with the severity of chronic liver disease. Ipp/Ipeak had high sensitivity and specificity and, thus, was of good diagnostic value. The data suggest that hepatic perfusion parameters acquired with CEUS may be useful in assessing the severity of chronic liver disease and predicting cirrhosis non-invasively. REFERENCES Aoki T, Imamura H, Kaneko J, Sakamoto Y, Matsuyama Y, Kokudo N, Sugawara Y, Makuuchi M. Intraoperative direct measurement of hepatic arterial buffer response in patients with or without cirrhosis. Liver Transpl 2005;11:684–691. Bauer M, Paquette NC, Zhang JX, Bauer I, Pannen BH, Kleeberger SR, Clemens MG. Chronic ethanol consumption increases hepatic sinusoidal contractile response to endothlin-1 in the rat. Hepatology 1995;22:1565–1576. Claudon M, Cosgrove D, Albrecht T, Bolondi L, Bosio M, Calliada F, Correas JM, Darge K, Dietrich C, D’Onofrio M, Evans DH, Filice C, Greiner L, J€ager K, Jong ND, Leen E, Lencioni R, Lindsell D, Martegani A, Meairs S, Nolsøe C, Piscaglia F, Ricci P, Seidel G, Skjoldbye B, Solbiati L, Thorelius L, Tranquart F, Weskott HP, Whittingham T. Guidelines and good clinical practice recommendations for contrast enhanced ultrasound (CEUS)— Update 2008. Ultraschall Med 2008;29:28–44. Eipel C, Abshagen K, Vollmar B. Regulation of hepatic blood flow: The hepatic arterial buffer response revisited. World J Gastroenterol 2010;16:6046–6057. G€ulberg V, Haag K, R€ossle M, Gerbes AL. Hepatic arterial buffer response in patients with advanced cirrhosis. Hepatology 2002;35: 630–634. Hagiwara M, Rusinek H, Lee VS, Losada M, Bannan MA, Krinsky GA, Taouli B. Advanced liver fibrosis: Diagnosis with 3-D whole-liver perfusion MR imaging—Initial experience. Radiology 2008;246: 926–934. Iwasa M, Nakamura K, Nakagawa T, Watanabe S, Katoh H, Kinosada Y, Maeda H, Habara J, Suzuki S. Single photon emission computed tomography to determine effective hepatic blood flow and intrahepatic shunting. Hepatology 1995;21:359–365. Kaneko T, Teshigawara O, Sugimoto H, Hirota M, Inoue S, Takeda S, Nakao A. Signal intensity of the liver parenchyma in microbubble contrast agent in the late liver phase reflects advanced fibrosis of the liver. Liver Int 2005;25:288–293. Kim HJ, Lee HW. Important predictor of mortality in patients with endstage liver disease. Clin Mol Hepatol 2013;19:105–115. Li X, Benjamin IS, Alexander B. The relationship between intrahepatic portal systemic shunts and microsphere induced portal hypertension in the rat liver. Gut 1998;42:276–282. Lim AK, Taylor-Robinson SD, Patel N, Eckersley RJ, Goldin RD, Hamilton G, Foster GR, Thomas HC, Cosgrove DO, Blomley MJK. Hepatic vein transit times using a microbubble agent can predict disease severity non-invasively in patients with hepatitis C. Gut 2005;54:128–133. Miles KA, Hayball MP, Dixon AK. Functional images of hepatic perfusion obtained with dynamic CT. Radiology 1993;188:405–411. Nakashige A, Horiguchi J, Tamura A, Asahara T, Shimamoto F, Ito K. Quantitative measurement of hepatic portal perfusion by multidetector row CT with compensation for respiratory misregistration. Br J Radiol 2004;77:728–734. Orlacchio A, Bolacchi F, Petrella MC, Pastorelli D, Bazzocchi G, Angelico M, Simonetti G. Liver contrast enhanced ultrasound
Hepatic CEUS parameters correlate with disease severity d D. LIU et al. perfusion imaging in the evaluation of chronic hepatitis C fibrosis: Preliminary results. Ultrasound Med Biol 2011;37:1–6. Orrego H, Blendis LM, Crossley IR, Medline A, Macdonald A, Ritchie S, Israel Y. Correlation of intrahepatic pressure with collagen in the Disse space and hepatomegaly in humans and in the rat. Gastroenterology 1981;80:546–556. Pagliaro L, Rinaldi F, Craxı A, Di Piazza S, Filippazzo G, Gatto G, Genova G, Magrin S, Maringhini A, Orsini S, Palazzo U, Spinello M, Vinci M. Percutaneous blind biopsy versus laparoscopy with guided biopsy in diagnosis of cirrhosis: A prospective, randomized trial. Dig Dis Sci 1983;28:39–43. Pugh RN, Murray-Lyon IM, Dawson JL, Pietroni MC, Williams R. Transection of the oesophagus for bleeding oesophageal varices. Br J Surg 1973;60:646–649. Ridolfi F, Abbattista T, Busilacchi P, Brunelli E. Contrast-enhanced ultrasound evaluation of hepatic micro-vascular changes in liver diseases. World J Gastroenterol 2012;18:5225–5230. Rokey DC, Weisiger RA. Endothelin induced contractility of stellate cells from normal and cirrhotic rat liver: Implications for regulation of portal pressure and resistance. Hepatology 1996;24:233–240. Schenk WG Jr, McDonald JC, McDonald K, Drapanas T. Direct measurement of hepatic blood flow in surgical patients: With related observations on hepatic flow dynamics in experimental animals. Ann Surg 1962;156:463–471. Staub F, Tournoux-Facon C, Roumy J, Chaigneau C, Morichaut-Beauchant M, Levillain P, Prevost C, Aube C, Lebigot J, Oberti F, Galtier JB, Laumonier H, Trillaud H, Bernard PH, Blanc JF, Sironneau S, Machet F, Drouillard J, de Ledinghen V, Couzigou P, Foucher P, Castera L, Tranquard F, Bacq Y, de Alteroche L, Ingrand P, Tasu JP. Liver fibrosis staging with contrast-enhanced ultrasonography: Prospective multicenter
2563
study compared with METAVIR scoring. Eur Radiol 2009;19: 1991–1997. Tang MX, Mulvana H, Gauthier T, Lim AK, Cosgrove DO, Eckersley RJ, Stride E. Quantitative contrast-enhanced ultrasound imaging: A review of sources of variability. Interface Focus 2011; 1:520–539. Tsushima Y, Blomley JK, Kusano S, Endo K. The portal component of hepatic perfusion measured by dynamic CT: An indicator of hepatic parenchymal damage. Dig Dis Sci 1999;44:1632–1638. Van Beers BE, Leconte I, Materne R, Smith AM, Jamart J, Horsmans Y. Hepatic perfusion parameters in chronic liver disease: Dynamic CT measurements correlated with disease severity. AJR Am J Roentgenol 2001;176:667–673. Villeneuve JP, Dagenais M, Huet PM, Roy A, Lapointe R, Marleau D. The hepatic microcirculation in the isolated perfused human liver. Hepatology 1996;23:24–31. Zaitoun A, AI M, Mardini H, Awad S, Ukabam S, Makadisi S, Record CO. Quantitative assessment of fibrosis and steatosis in the liver biopsies from patients with chronic hepatitis C. J Clin Pathol 2001;54:461–465. Zhang L, Yin J, Duan Y, Yang Y, Yuan L, Cao T. Assessment of intrahepatic blood flow by Doppler ultrasonography: Relationship between the hepatic vein, portal vein, hepatic artery and portal pressure measured intra-operatively in patients with portal hypertension. BMC Gastroenterol 2011;11:84. Ziegler SI, Haberkorn U, Byrne H, Tong C, Kaja S, Richolt JA, Byrne H, Tong C, Schosser R, Krieter H, Kaja S, Richolt JA, Lammertsma AA, Price P. Measurement of liver blood flow using oxygen-15 labelled water and dynamic positron emission tomography: Limitation of model description. Eur J Nucl Med 1996;23: 169–177.
