British Journal of Radiology 75 (2002),417-421 © 2002 The British Institute of Radiology
Portal vein pulsatility and spectral width changes in patients with portal hypertension: relation to the severity of liver disease
M Barakat, MD
Department of Tropical Medicine and Gastroenterology, Faculty of Medicine, Assiut University, Assiut, Egypt
 |
Abstract
|
|---|
Using Doppler ultrasound in patients with chronic liver disease (CLD) and portal hypertension, portal vein pulsatility pattern and spectral width changes were assessed and evaluated in relation to the severity of liver disease as determined according to the Child–Pugh score. The pulsatility index (PI) was significantly lower in CLD patients (mean±SD: 0.23±0.08) compared with healthy subjects (0.39±0.1) (p<0.001) and lower in Child–Pugh class C compared with Child–Pugh class A patients (0.21±0.07 vs 0.25±0.08, respectively) (p<0.05). The spectral width index was significantly higher in CLD patients vs healthy subjects (0.91±0.16 vs 0.60±0.12, respectively) (p<0.001). The difference was also noted in the early stage (Child–Pugh A patients) when compared with healthy subjects (0.88±0.17). In conclusion, portal vein pulsatility and spectral width indices can reflect the early haemodynamic changes in CLD patients. These changes become more pronounced with the progression of liver disease.
|
Introduction
|
|---|
Doppler ultrasound has been used to assess various portal vein haemodynamic parameters including flow direction, velocity and volume, in patients with portal hypertension [1–4]. However, detailed information regarding the flow waveform and its pulsatility in such conditions is lacking in the literature. It has been noted that the portal vein waveform normally demonstrates a variable fluctuation over time for which the term pulsatility has been adopted [5]. Most previous studies have been concerned with the increase in portal venous pulsatility in association with certain situations, particularly those related to cardiac diseases, e.g. tricuspid regurgitation [6], heart failure [7, 8] and constrictive pericarditis [9]. The aims of this work were to describe the portal vein pulsatility pattern in patients with chronic liver disease (CLD) and portal hypertension as compared with healthy adults and to evaluate the relation of the pulsatility pattern to the severity of liver disease.
|
Subjects and methods
|
|---|
The study included 157 patients with CLD; 93 men and 64 women with an age range of 19–65 years (mean±standard deviation (SD): 45.1±11.2 years). The diagnosis of CLD was based on a combination of clinical data, e.g. jaundice, ascites, muscle wasting, cutaneous spider angiomas, ecchymosis, palmar erythema and flapping tremors, laboratory data, e.g. decreased serum albumin and prolonged prothrombin time, and ultrasound data, e.g. coarsened bright liver echopattern and nodular liver surface, in addition to liver biopsy whenever possible. Grading of the severity of chronic liver disease was assessed according to the Child classification modified by Pugh [10]. This classification system is a score of clinical and laboratory findings including presence of ascites, grade of encephalopathy, serum bilirubin, serum albumin and prothrombin time. Accordingly, 60 patients were allocated into Child–Pugh class A, 45 into Child–Pugh class B and 52 into Child–Pugh class C, where Child–Pugh class C represents the most pronounced liver dysfunction. The aetiology of the liver disease was hepatitis B, hepatitis C, bilharzias (Schistosoma mansoni or S. haematobium) or mixed. No patients were alcoholic. All patients included in the study had oesophageal varices demonstrated on upper endoscopy examination evidencing the presence of portal hypertension. Exclusion criteria were (1) portal vein thrombosis, (2) reversed portal vein flow, (3) previous sclerotherapy, band ligation or any surgical intervention to avoid alteration of the portal haemodynamics, and (4) grade 3 or 4 encephalopathy (the patient would be uncooperative during Doppler examination). In addition, 58 healthy subjects were examined as a control group; 33 men and 25 women with an age range of 18–57 years (mean±SD: 32.9±8.9 years).
The portal vein flow waveform was recorded for patients and healthy subjects using Doppler ultrasound (Acuson 128 XP10; Acuson Corporation, USA) with a vector transducer operating at 3.5 MHz. The transducer was oriented along the longitudinal axis of the main portal vein using a paramediam or a slightly oblique plane. The point of measurement was midway between the confluence of the splenic and superior mesenteric veins and the bifurcation of the portal vein. The Doppler angle, between the axis of the Doppler beam and that of the portal vein, was always <60°. The sample volume was adjusted to include as much of the lumen as possible without including the vessel wall. All measurements were performed on fasting and under respiratory suspension in expiration. Calculations were obtained from a frozen spectral strip (Figure 1
). Portal vein pulsatility was expressed as the pulsatility index (PI) where measurements were taken at the wave envelope and calculated as
|
View larger version (8K):
[in this window]
[in a new window]
|
Figure 1. Simplified representation of a portal vein flow waveform as seen on spectral Doppler display. a, maximum peak velocity at the wave envelope; b, minimum peak velocity; c, minimum velocity at the wave base; w, window area underneath the flow wave.
|
|
Also, the spectral width or the flow wave thickness representing the range of velocities across the vessel lumen was expressed as the spectral width index (SWI) and calculated as
As spectral width increases, window size (area below the waveform that is devoid of flow) decreases or even disappears. With maximum spectral broadening, no window is seen and the SWI would be equal to 1.
Statistical analysis
Results were expressed as mean±SD. Differences in the PI and SWI values between groups were analysed using Kruskal–Wallis and Mann–Whitney tests. Differences in the distribution of marked spectral broadening were analysed with the 
2 test. Probability values <0.05 were considered significant.
|
Results
|
|---|
In healthy subjects, inspection of the portal vein wave envelope has shown slight fluctuation measured between the maximum and minimum velocities (Figure 2). The maximum velocity range was 13–30 cm s-1 (20.5±4.2), whereas the minimum velocity range was 6–20 cm s-1 (12.5±3.6). PI varied between 0.21 and 0.58 (0.39±0.1). Pulsatility was pronounced (PI
0.5) in 13 (22.4%) subjects. SWI was 0.60±0.12, range 0.35–1. SWI was equal to 1 in only one (1.7%) healthy subject, which means that the spectrum of velocities covered the whole area from the wave envelope to the baseline (broad spectrum). In other words, a window was seen underneath the flow wave in all healthy subjects except one. However, in this one individual the waveform was still pulsatile (Figure 3).
View larger version (40K):
[in this window]
[in a new window]
|
Figure 2. Portal vein flow waveform in a healthy subject with a pulsatile pattern, narrow spectrum and window area underneath the flow wave.
|
|
View larger version (50K):
[in this window]
[in a new window]
|
Figure 3. Portal vein flow wave in a healthy subject with a pulsatile pattern, broad spectrum and absent window.
|
|
In patients with CLD, the PI was significantly lower than in healthy subjects (0.23±0.08 vs 0.39±0.1, respectively) (p<0.001). This difference was noted in each of the three subgroups, i.e. Child–Pugh class A, B and C patients, separetely as compared with healthy subjects (p<0.001). Furthermore, the PI was lower in Child–Pugh class C compared with Child–Pugh class A patients, i.e. with progression of liver dysfunction, (p<0.05) (Table 1).
View this table:
[in this window]
[in a new window]
|
Table 1. Portal vein pulsatility index (PI) and spectral width index (SWI) in healthy subjects and chronic liver disease (CLD) patients
|
|
PI was less than 0.2 (almost non-pulsatile flow) in 60 (38.2%) of the CLD patients but in none of the healthy subjects (Figure 4). Conversely, PI was equal to 0.5 in only 3 (1.9%) of the CLD patients and never exceeded this value in any patient.
View larger version (39K):
[in this window]
[in a new window]
|
Figure 4. Portal vein flow wave in a chronic liverdisease patient with almost flat wave envelope (non-pulsatile flow) and absent window. L, liver; A, ascites; GB, gall bladder.
|
|
The SWI was significantly higher in CLD patients in comparison with healthy adults (0.91±0.16 vs 0.60±0.12, respectively) (p<0.001). The difference was still significant between each of the Child–Pugh grades (Child–Pugh class A, B and C separately) and healthy subjects (p<0.001). However, no statistical difference was present between the different Child–Pugh classes (p>0.05), i.e. within the three groups (Table 1
). Complete spectral broadening (no visible window with a SWI=1) was seen in 113 (71.9%) of the CLD patients compared with only 1 (1.7%) of the healthy adults. These 113 cases were distributed as follows: 39 of 60 (65%) cases in Child–Pugh class A patients, 32 of 45 (71.1%) cases in Child–Pugh class B patients and 42 of 52 (80.8%) cases in Child–Pugh class C patients. Although the prevalence of complete spectral broadening was shown to increase with the progression of liver disease, the difference in distribution did not reach statistical significance between the three Child–Pugh grade groups (p>0.05).
|
Discussion
|
|---|
Portal vein flow has been described in earlier reports as being non-pulsatile or continuous [11, 12]. This was probably a general description for the portal vein waveform in comparison to the characteristic prominent pulsatile flow in arteries, e.g. the hepatic artery. With recent interest in analysis of the portal vein waveform, a more precise description has emerged concerning the variable undulant fluctuation of flow, and hence the term pulsatility has been used by different authors [1, 3, 5]. However, various terms have been used in the literature to express pulsatility measures. For example, the terms pulsatility ratio [8] and pulsatility score [13] have been used to describe the same measure that is quantified as the minimum peak velocity/maximum peak velocity. Conversely, the term pulsatility index [5] is calculated as in equation (1). It is clear from these two equations that the PI and pulsatility ratio are two inversely related measurements, i.e. an increase in one of them corresponds to a decrease in the other. In this study, PI was preferable as a high or a low index directly expresses the corresponding situation, i.e. a high or a low portal vein pulsatility.
A precise description of the portal vein waveform in healthy individuals is required to gather more information about the normal patterns. Results collected through this study indicate that, despite wide variation, collectively there are two pulsatility patterns in normal conditions; (a) the more common slight fluctuation pattern with a PI >0.2 and <0.5 (77.6% of cases), and (b) the less common pronounced pulsatility pattern with a PI 0.5 (22.4% of cases). Whilst pulsatility was never severe, a situation defined as systolic flow interruption or reversal [9], i.e. PI is 1, the index was never below 0.2, a situation considered to represent almost non-pulsatile or flat wave envelope [5].
It should be noted that there are varying pulsatility values reported in the literature, particularly regarding the lower and upper limits of normal. For example, in the study of Gallix et al [5], the PI was found to be less than 0.2 in 6 of 23 (26%) normal individuals, which was suggested by the authors to be related to the effect of obesity with increased body mass index and increased intraabdominal pressure. However, Wachsberg et al [13] recorded a PI 1 (expressed as pulsatility ratio ![{els]](/math/els.gif)
0) in a small number of inviduals who had no evidence of heart disease. In order to settle this variation, it can be assumed that the normal range of PI extends between 0.2 and 0.5 (usually less than 1) and that the states of being very low (<0.2) or very high (1) are of limited occurrence. These extreme patterns can be accepted as normal in the absence of any evidence of liver or cardiac diseases.
The other character of the portal vein waveform is the width of the spectrum of the velocities. This is the first study to take this character into consideration when analysing portal vein flow waveform. A window was seen in all healthy subjects except one, in whom the flow wave showed complete spectral broadening (no window; SWI=1). Therefore, it can be considered that the existence of a window is a character of the flow wave in healthy subjects and that complete spectral broadening can occur but with a PI lying in the normal range and in the absence of liver disease.
As shown in the results, portal vein flow wave in patients with CLD becomes less pulsatile than in normal individuals, together with a spectral broadening (increase in SWI). These two changes are evident even in the earlier stages of liver disease or in the compensated Child–Pugh class A patients. Change, particularly in PI, becomes more pronounced with increased severity of liver disease or liver dysfunction, i.e. progression from Child–Pugh class A to Child–Pugh class C state. Haemodynamically, decrease in pulsatility can be attributed, at least in part, to decreased transmission of atrial pressure changes through the hepatic veins as a result of pathological fibrotic changes in the liver [14, 15].
The occurrence of spectral broadening in CLD patients implies the presence of very slow flow streams in the portal vein leading to a decrease in the window area underneath the flow wave or even the complete absence of the window. The portal vein mean flow velocity has previously been shown to decrease in patients with CLD [4]. However, this lowering has been more evident in later stages (Child–Pugh class C patients). Spectral broadening may represent an earlier step before the net mean flow velocity becomes lowered and hence spectral broadening can be noticed even in Child–Pugh class A patients, as can be demonstrated in this study. This is supported by the finding that there was no statistical difference in the prevalence of complete spectral broadening (no window) between Child–Pugh class A and C patients, because once the spectrum shows complete broadening at one stage, e.g. Child–Pugh class A, it cannot subsequently be any broader, i.e. with progression of liver dysfunction, as it has already reached its maximum.
In conclusion, a decrease in PI (<0.2) or an increase in SWI (particularly complete spectral broadening or absent window; SWI=1) can be adjunctive signs to monitor and identify the early haemodynamic changes in patients with CLD.
Received for publication June 12, 2001.
Revision received November 19, 2001.
Accepted for publication February 14, 2002.
|
References
|
|---|
-
Partiquin H, Lafortune M, Burns P, Dauzat M. Duplex Doppler examination in portal hypertension: technique and anatomy. AJR 1987;149:71–6.[Abstract/Free Full Text]
-
Ralls PW. Colour Doppler sonography of the hepatic artery and portal venous system. AJR 1990;155:517–25.[Abstract/Free Full Text]
-
Koslin B, Mulligan SA, Berland LL. Duplex assessement of the portal venous system. Semin Ultrasound CT MR 1992;13:22–33.[Medline]
-
Chawla Y, Santa N, Dhiman RK, Dilawari JB. Portal haemodynamics by duplex Doppler sonography in different grades of cirrhosis. Dig Dis Sci 1998;43:354–7.[Medline]
-
Gallix BP, Taourel P, Dauzat M, Bruel J-M, Lafortune M. Flow pulsatility in the portal venous system: a study of Doppler sonography in healthy adults. AJR 1997;169:141–4.[Abstract/Free Full Text]
-
Abu-Yousef MM, Milam SG, Farner RM. Pulsatile portal vein flow: a sign of tricuspid regurgitation on duplex Doppler sonography. AJR 1990;155:785–8.[Abstract/Free Full Text]
-
Duerinckx AJ, Grant EG, Perrella RR, Szeto A, Tessler FN. The pulsatile portal vein in cases of congestive heart failure: correlation of duplex Doppler findings with right atrial pressure. Radiology 1990;176:655–8.[Abstract/Free Full Text]
-
Catalano D, Caruso G, DiFazzio S, Carpinteri G, Scalisi N, Trovato GM. Portal vein pulsatility ratio and heart failure. J Clin Ultrasound 1998;26:27–31.[Medline]
-
Gorka T, Gorka W. Doppler sonographic diagnosis of severe portal vein pulsatility in constrictive pericarditis: flow normalization after pericardiectomy. J Clin Ultrasound 1999;27:84–8.[Medline]
-
Pugh RNH, Murray-Lyon IM, Dawson JL, Pietroni MC, Williams R. Transection of the oesophagus for bleeding oesophageal varices. Br J Surg 1973;60:646–9.[Medline]
-
Taylor KJW, Burns PN, Woodcock JP, Wells PNT. Blood flow in deep abdominal and pelvic vessels: ultrasonic pulsed-Doppler analysis. Radiology 1985;154:487–93.[Abstract/Free Full Text]
-
Keller MS, Taylor KJW, Riely CA. Pseudoportal Doppler signal in the partially obstructed inferior vena cava. Radiology 1989;170:475–7.[Abstract/Free Full Text]
-
Wachsberg RH, Needleman L, Wilson DJ. Portal vein pulsatility in normal and cirrhotic adults without cardiac disease. J Clin Ultrasound 1995;23:3–15.[Medline]
-
Bolondi L, Bassi S, Gaiani S, Zironi G, Benzi G, Santi V, et al. Liver cirrhosis: changes of Doppler waveform of hepatic veins. Radiology 1991;178:513–6.[Abstract/Free Full Text]
-
Ohta M, Mashizume M, Kawanaka H, Akazawa K, Tomikawa M, Higashi H, et al. Prognostic significance of hepatic vein waveform by Doppler ultrasonography in cirrhotic patients with portal hypertension. Am J Gastroenterol 1995;90:1853–7.[Medline]
This article has been cited by other articles:
|
|
|
|
 
M Barakat
Non-pulsatile hepatic and portal vein waveforms in patients with liver cirrhosis: concordant and discordant relationships
Br. J. Radiol.,
July 1, 2004;
77(919):
547 - 550.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
