CORRELATION OF PORTAL VEIN PULSATILITY PATTERN AND SEVERITY OF LIVER DISEASE IN PATIENTS WITH CIRRHOSIS AND PORTAL HYPERTENSION

Objective: To correlate the portal vein pulsatility pattern with severity of liver disease in patients with cirrhosis of liver and portal hypertension. Subjects and methods: Doppler signals from the main portal vein of 36 healthy adults and 52 cirrhotic patients with portal hypertension were studied. Severity of liver disease was graded using modified Child-Pugh classification. Pulsatility of portal flow was quantified using portal venous pulsatility index and complete spectral widening was defined as absence of window below the wave base. The Doppler findings were correlated between the study groups. Results: The mean pulsatility index value in control group was 0.37+0.10 and in cirrhotic patients was 0.17+0.03 (Child A0.21+0.02, Child B0.18+0.02, Child C0.14+0.03). The difference between control and cirrhosis group, as well as the difference within different Child classes were statistically significant (P<0.05). None of the patients in control group had complete spectral widening while 76.92% of cirrhotic patients had complete spectral widening (28.5% of Child A, 66.6% of Child B and 100% of Child C). The difference in distribution of complete spectral widening between control and cirrhotic group as well as within the cirrhotic group was statistically significant (P<0.05). Conclusion: Portal vein pulsatility index and spectral widening can reflect the early hemodynamic changes in cirrhotic patients. These changes become more pronounced with increasing severity of liver disease.


Introduction:
Cirrhosis is defined by World Health Organization (WHO) as a diffuse process characterized by fibrosis and the conversion of normal liver architecture into structurally abnormal nodules. 1 Progressive hepatic fibrosis causes several regional hemodynamic changes like hepatic venous outflow obstruction, changes in hepatic artery resistance and development of portal hypertension with increasing sinusoidal resistance.These hemodynamic changes influence the degree of portal hypertension and liver dysfunction. 2,3trasonography is the most commonly used imaging modality for diagnosis and follow up of patients with cirrhosis.The diagnosis usually relies on late findings of volume redistribution, irregularity of liver surface and secondary findings of portal hypertension.However, Bmode sonography is incapable of examining patients of cirrhosis without these late findings. 2oppler sonography is a non-invasive diagnostic modality based on hemodynamic parameters.Hemodynamic changes might have developed even in cases with normal findings on B-mode sonography. 4 Therefore assessment of these alterations has importance for early diagnosis and for close follow up of previously diagnosed cases. 2 Hepatic venous Doppler changes have been described in cirrhosis; however these changes are not specific for cirrhosis and can even be seen in cases of steatosis. 5terations of portal vein hemodynamics in cirrhosis have been observed in various studies.With recent interest in analysis of portal vein pulsatility pattern in relation to cirrhosis of liver, different measures of pulsatility of portal waveform have been studied.Simple parameters like portal vein pulsatility index and spectral widening were found, in some studies, to reflect the early changes in portal hemodynamics in relation to the severity of liver disease. 6Though extensive studies regarding their usefulness is lacking in literature, these early changes can serve as valuable indicators of presence of cirrhosis when other imaging findings are equivocal or negative.These parameters can be of value in assessment of severity of liver disease and the follow up of patient as well. 6spite wide variation, there are in general two pulsatility patterns in normal conditions; (a) the more common slight fluctuation pattern with a PI between 0.2 to 0.5, and (b) the less common pronounced pulsatility pattern with a PI > 0.5.A PI value > 1 (severe pulsatility), a situation defined by systolic flow interruption or reversal, suggests presence of cardiac disease whereas PI value below 0.2 (almost nonpulsatile or flat wave envelope) is considered to represent presence of chronic liver disease. 6pectral width changes, esp.complete spectral widening also suggests presence of chronic liver disease.-Patients with portal vein flow reversal or bidirectional portal flow.
Portal vein pulsed Doppler examination was performed with a 3.5 MHz probe in a commercially available ultrasound system (SONOACE 8000 LIVE) with Doppler facility.
The transducer was kept along the longitudinal axis of main portal vein in an oblique paramedian plane.The point of measurement was midway between the confluence of splenic & superior mesenteric vein and bifurcation of portal vein, the Doppler angle always being <60 degree.The sample volume was adjusted to include as much of the lumen as possible without including the vessel wall.The flow waveform was recorded, for patients as well as healthy subjects.All measurements were performed on fasting with patient breathing quietly.
A matching control group of healthy individuals coming to department of radiology for general health check up examination were also subjected to Doppler study of main portal vein.
Portal vein pulsatility was expressed as the Pulsatility index (PI) which was calculated as: (maximum peak velocity -minimum peak velocity)/ maximum peak velocity.
Complete spectral widening (CSW) was said to be present if no window was visible under the wave base.
The data obtained were compiled and analyzed using standard statistical analysis.SPSS

Discussion:
The normal portal vein demonstrates an undulating hepatopetal flow.Mean portal venous velocity is approximately 15 to 18 cm/sec.Normal portal venous flow rates vary in the same individual: they increase after a meal, decrease after exercise or when the patient is upright.It also varies with respiration.(10,11) As portal hypertension develops, the flow in the portal vein loses its undulatory pattern and becomes progressively flattened.As the severity of portal hypertension increases, flow becomes biphasic and finally hepatofugal. 6,7rious measurements and indices have been used by different authors to study the portal hemodynamics and used it as indicators of hepatic morphological changes.They include measures of portal flow and measures of pulsatility pattern.
There are number of measures of portal flow such as congestion index, modified hepatic index, hepatic vascular index and portal blood flow.However these composite measures of portal flow are difficult to calculate, hindering their widespread use and their reliability too is yet to be proved. 2Unreliability of measures of portal vein flow was also noted in a study by Haktanir A et al. 2 The studies evaluating interobserver difference in the portal flow measurements found that acceptable interobserver agreement was not found between the observers. 8,9lsatility pattern is measured using measures like pulsatility ratio 10 , pulsatility score 11 and pulsatility index.Pulsatility ratio and pulsatility score both describe the same measure that is quantified as the minimum peak velocity/ maximum peak velocity, which increases with progressive loss of pulsatility.Portal vein pulsatility index on the other hand was preferred by other authors as a high or low index directly expresses the corresponding situation, i.e., high or low portal vein pulsatility. 6In this study also portal vein pulsatility index was used for the same reason.

Mean PI values in healthy individuals:
The mean PI value in healthy individuals in our study was 0.377+0.108.The PI values found by different authors are 0.39+0.1 (Barakat M) 6 , 0.48+0.31(Gallix BP et al) 12 , 0.22 (Chou LS et al) 13 .Duerinkx et al 14 and Wachsberg et al 11 found PI values in healthy individuals to be <0.61 and <0.54 respectively.
In our study 91.66% (33 of 36) healthy individuals had PI between 0.2 -0.5 and 8.33% (3 of 36) had pronounced pulsatility >0.5.However none of the individuals had PI >1 or <0.2.Barakat M 6 had found that 77.6% of healthy individuals had PI between 0.2 -0.5 and 22.4% had pronounced PI >0.5, while none had PI >1 or <0.2.These findings are consistent with the general agreement that portal vein pulsatility index values in healthy adults are within 0.2 to 0.5 range.

Mean PI values in cirrhotic patients:
The mean PI value in patients with cirrhosis and PH in this study was 0.17+0.03.The difference of PI value between the control and cirrhosis group was statistically significant.
In a study involving 157 patients, Barakat M 6 had found PI of 0.23+0.08 in cirrhotic patients.The lower PI value in our study was partly due to less number of patients in Child-Pugh class A (13.4% in our study vs. 38.2% in study of Barakat M.).The less number of Child-A patients in our study was due to the fact that liver biopsy was not included as a diagnostic criteria for cirrhosis, which is invaluable in ascertaining the presence of cirrhosis in early disease with minimal clinical, laboratory and morphological changes.
In a study involving 38 children (mean age 3.3 years) with end stage liver disease and portal hypertension, Westra SJ et al 15 found increased pulsatility of portal vein waveform.He had used Resistive Index (RI) as a measure of pulsatility which was calculated using the same equation that has been used in our study.Mean RI value in his study was 0.6+0.33.The reason that had been proposed for increased pulsatility in children was that in children the compliance of hepatic parenchyma is more than in adults to accommodate the arterial pressure variations.

PI in different Child-Pugh classes:
In the present study, the mean pulsatility index value was 0.21 in Child A, 0.18 in Child B and 0.14 in Child C, the differences being statistically significant.The finding that the PI value decreases with increasing severity of liver disease is consistent with the findings of Barakat M 6 who had found the mean PI value in Child A to be 0.25, in Child B 0.23 and in child C to be 0.21.However the lower mean PI value in our study in all three Child classes is difficult to explain.This could in part be related to the etiology of cirrhosis.Barakat M had not included patients with alcoholic cirrhosis in his study while majority of patients in this study had alcoholic cirrhosis.Studies evaluating the relation of portal vein PI with etiology of cirrhosis are not available in literature.

Spectral widening in control and cirrhotic patients:
In this study, none of the patients in control group had complete spectral widening while 76.92% of patients with cirrhosis and PH had complete spectral widening.This is consistent with the study of Barakat M 6 of 157 patients with cirrhosis of liver, where 71.9% of patients with cirrhosis had complete spectral widening.
In our study, complete spectral widening was found in 28 .5% of Child A patients, 66.6% of Child B patients and 100% of Child C patients, the difference being statistically significant.In contrast, Barakat M 6 had not found any statistically significant difference in presence of complete spectral widening between different Child classes.This difference with his findings has to be further evaluated with larger studies; however this could again, in part, be due to significantly less number of patients with Child A cirrhosis (13.46% vs 38.21%) in our study.

Relation of Child-Pugh variables with PI and Spectral widening:
The PI values were lower with increasing grades of all five variables (ascites, encephalopathy, albumin, Bilirubin and PT prolongation), the difference between the lowest and highest grades being statistically significant.However the statistically significant increase in presence of CSW was noted only with increasing values of Bilirubin and PT prolongation.Literature is lacking regarding the relation of PI and spectral widening with the variables of Child's classification.

Conclusion:
Decrease in pulsatility index is a valuable indicator of early hemodynamic changes in cirrhotic patients with portal hypertension, changes being significantly more pronounced with increasing severity of the disease.Similarly presence of complete spectral widening is also a strong predictor of increased severity of the disease.These findings can be used as adjunctive signs to identify and monitor the hemodynamic changes in patients with cirrhosis.
The study had its own limitations.The limitations were small sample size, particularly in Child-Pugh class A, convenient sampling and limited time.TUTH being a referral centre, the patients included may not represent the patients of cirrhosis in general population.
Similarly the healthy subjects included in this study were taken from General Health Check up clinic and may not represent the general population.
Further large-scale studies should be performed evaluating these parameters to establish their roles in patients with cirrhosis and portal hypertension.

Fig 1 .
Fig 1.Normal pulsatile portal vein waveform with presence of window below the wave base.

Fig 2 .
Fig 2. Portal vein wave form in a patient with Cirrhosis (Child C) with portal hypertension: flattened waveform with absent window.