|Year : 2019 | Volume
| Issue : 3 | Page : 1119-1125
The role of portal congestion index in prediction of esophageal varices in hepatitis C virus-infected patients
Mohamed A Nouh1, Moamena S El-Hammoly1, Safaa A Mohamed2, Maha M Hana1
1 Department of Tropical Medicine, Faculty of Medicine, Menoufia University, Shebeen El-Kom, Egypt
2 Department of Diagnostic Radiology, Faculty of Medicine, Menoufia University, Shebeen El-Kom, Egypt
|Date of Submission||17-Feb-2019|
|Date of Acceptance||14-Apr-2019|
|Date of Web Publication||17-Oct-2019|
Maha M Hana
Beket El-Sabae District, Menoufia
Source of Support: None, Conflict of Interest: None
To assess the role of portal congestion index (PCI) in prediction of esophageal varices (EV) in hepatitis C virus (HCV)-infected patients.
Portal hypertension (PH) often develops in HCV-infected patients. EV rupture leads to variceal bleeding, the most lethal complication of cirrhosis. Duplex Doppler ultrasonography of portal vein is a noninvasive method to assess blood flow velocity and portal vein cross-sectional area as a parameter for PH.
Patients and methods
This study was conducted on 200 HCV-infected patients (116 with EV and 84 without EV). Patients were subjected to complete blood picture, liver and kidney functions, serum electrolytes, viral markers, abdominal ultrasonography, upper gastrointestinal tract endoscopy, and duplex Doppler evaluation of portal vein with calculation of PCI.
There was a highly significant difference between HCV-infected patients with and without EV regarding PCI (P = 0.0001). At cutoff point of PCI 0.11, the sensitivity, specificity, and accuracy to predict EV were 93, 85.7, and 90%, respectively, with area under the curve 0.889. There was a statistically highly significant positive correlation between PCI and presence of fundal varices and PH gastropathy (P = 0.0001). There was a statistically significant positive correlation between PCI and grades of EV (P = 0.012).
PCI is useful as a noninvasive predictor of EV in HCV-infected patients.
Keywords: Doppler ultrasonography, esophageal varices, hepatitis C virus, portal congestion index, upper gastrointestinal tract endoscopy
|How to cite this article:|
Nouh MA, El-Hammoly MS, Mohamed SA, Hana MM. The role of portal congestion index in prediction of esophageal varices in hepatitis C virus-infected patients. Menoufia Med J 2019;32:1119-25
|How to cite this URL:|
Nouh MA, El-Hammoly MS, Mohamed SA, Hana MM. The role of portal congestion index in prediction of esophageal varices in hepatitis C virus-infected patients. Menoufia Med J [serial online] 2019 [cited 2020 Apr 8];32:1119-25. Available from: http://www.mmj.eg.net/text.asp?2019/32/3/1119/268828
| Introduction|| |
Portal hypertension (PH) is a pathological increase in the portal venous pressure between the portal vein and the inferior vena cava to more than 5 mmHg . Development of esophageal varices (EV) and variceal bleeding along with the development of decompensation needs hepatic venous pressure gradient of at least 10 mmHg, which is defined by clinically significant PH . The most lethal complication of cirrhosis associated with mortality of at least 20% at week 6 from bleeding is variceal bleeding, which results from rupture of gastroesophageal varices, the most relevant porto-systemic collaterals . Newly diagnosed hepatitis C virus (HCV) cirrhotic patients need endoscopic screening for EV, because medical treatment must be considered to prevent bleeding in case varices are detected . So, there is a need for a noninvasive method that can predict the presence of EV. The method should be simple, noninvasive, inexpensive, reproducible, accurate, and readily available; have high sensitivity and specificity; follow the natural history; reflect the effect of the treatment accurately; and indicate the prognosis and possibility of success of a treatment . Trial of several methods have been investigated as potential markers of EV, but none of them have proved to be ideal for that purpose . High-risk patients with EV development who can benefit from prophylactic pharmacological and endoscopic therapies need specific noninvasive methods and also low-risk patients need them to avoid unnecessary endoscopy .
Hence, ultrasonographic examination as simple, inexpensive, accurate, and noninvasive technique is strongly suggested. Several ultrasonographic indices including splenic size, portal vein velocity, and portal vein diameter (PVD) have been shown to be predictive of the severity of EV or risks of variceal bleeding in patients with cirrhosis .
This study aimed to assess the role of portal congestion index (PCI) in prediction of EV in HCV-infected patients.
| Patients and Methods|| |
A total of 700 HCV-infected patients were invited to and assessed for participation in the study. Overall, 500 patients were excluded from the study. Finally, 200 HCV-infected patients were enrolled in this study attending the outpatient or inpatient in Menoufia University Hospital in the period between January 2017 and March 2018.
The study included HCV-infected patients. Exclusion criteria were non-HCV cirrhotic patients; patients with PH who were receiving medical, endoscopic or surgical treatment; and patients with hepatocellular carcinoma. This study was approved by the Committee for Ethics of Faculty of Medicine, Menoufia University, Egypt. The aim of the research was explained to each patient, and a written consent was obtained from all participants.
Patients were classified into the following groups after endoscopy:
Group I: 116 patients with EV.
Group II: 84 patients without EV.
All patients were subjected to the following:
- Proper and detailed history taking
- Complete clinical examinations
- Laboratory investigations:complete blood picture, serum bilirubin, serum albumin, alanine transaminase, aspartate transaminase, prothrombin time and concentration, international normalized ratio, blood urea, serum creatinine, and serological tests for viral markers (HBsAg and HCVAb by ELISA). All laboratory investigations were carried out in the Clinical Pathology Department, Faculty of Medicine, Menoufia University
- Child–Pugh classification was calculated depending on patients' clinical and laboratory data
- Upper gastrointestinal tract endoscopy:
All patients were prepared for endoscopy (Olympus endoscope dip0230, Olympus Company,, Tokyo, Japan) by fasting 6 h before endoscopy or underwent emergent endoscopy if bleeding presentation was found. They were positioned in left lateral position and given suitable sedation
Varices were graded according to Westaby et al. 
- Radiological evaluation:
- Abdominal ultrasonography:
- Evaluation of liver regarding its size, echo-pattern, cirrhosis, and focal lesions
- Evaluation of spleen regarding its size, echo-pattern, and focal lesions
- Presence of porto-systemic collaterals
- PVD (mm)
The portal vein was measured at the hepatic hilum, proximal to the bifurcation
- Evaluation of ascites regarding amount and evidence of echoes, adhesions or loculations
- Duplex Doppler study of portal vein and PCI.
All patients underwent color Doppler ultrasonic examination after endoscopy by a single radiologist blinded to the results of endoscopy. All patients were kept on fasting for at least 6 h before the ultrasonography. Moreover, the main Doppler factors were always taken by the same equipment (with a 3.5–5 MHz curve – array transducer, G-50 Siemens) and by the same operator
- Assessment of portal vein patency and blood flow:
It is identified by following the splenic vein to the right until its junction with the superior mesenteric vein. This technique avoids confusing the PV with the bile duct or the inferior vena cava. However, when the PV is difficult to observe in the supine position, the patient is examined in the left lateral decubitus position
- Estimation of portal hemodynamic parameters:
PV cross-sectional area was measured from the B-mode image of the PV while scanning perpendicular to the long axis of the PV
Portal vein blood velocity (cm/s):
Main portal vein blood velocity was measured in its mid-portion, where the hepatic artery crosses the portal vein. The axial size of the sample volume was adjusted to encompass the portal vein lumen in its entirety. The angle between the Doppler beam and the long axis of the vessel was made to be less than 60°. Portal vein blood velocity was automatically calculated on samples of the Doppler signal lasting more than 4 s
PCI: Its radiological range is 0.070 ± 0.029 cm×s
- Platelet count/spleen diameter (PC/SD):
The PC/SD is calculated by dividing number of platelets/ml by the maximum bipolar diameter of spleen in millimeters estimated with abdominal ultrasound.
All data were collected, tabulated, and statistically analyzed using statistical package for social science 19.0 for Windows (SPSS; SPSS Inc., Chicago, Illinois, USA) and MedCalc 13 for Windows (MedCalc Software bvba, Ostend, Belgium). Quantitative data were expressed as mean ± SD or SE. SE = SD/square root of patients' number, which was used in case of big SD. Data were analyzed by independent sample, paired t-test, and one-way analysis of variance followed by Tukey test whenever single-step multiple comparison was needed to maintain integrity of data. However, qualitative data were expressed as number and percentage and were analyzed by χ2-test. The correlation was done using a Pearson correlation test. The receiver operating characteristic curve and 95% confidence interval were performed to determine cutoff values for the studied biomarkers. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were determined. P value was considered significant if less than 0.05 and highly significant if less than 0.001.
| Results|| |
Regarding patients enrolled in this study, their age ranged from 33 to 77 years, with a mean value of 51.07 ± 9.4 in patients without varices and a mean value of 55.91 ± 8.23 in patients with varices. Patients comprised 54% males (108 patient) and 46% females (92 patients). Hemoglobin ranged from 7 to 14 g/dl, with a mean value of 11.48 ± 1.67 in patients without varices and a mean value of 9.74 ± 1.52 in patients with varices. Platelets count ranged from 49 to 216 × 109/l, with mean value of 158.88 ± 50.18 in patients without varices and a mean value of 102.07 ± 31.73 in patients with varices.
Serum albumin in patients enrolled in this study ranged from 1.9 to 4.9 g/dl, with a mean value of 3.67 ± 0.71 in patients without varices and a mean value of 2.81 ± 0.41 in patients with varices. Total serum bilirubin ranged from 0.9 to 4.9 mg/dl, with a mean value of 1.56 ± 1.78 in patients without varices and a mean value of 2.81 ± 1.12 in patients with varices. International normalized ratio ranged from 1 to 1.8, with a mean value of 1.1 ± 0.19 in patients without varices and a mean value of 1.36 ± 0.22 in patients with varices. Serum creatinine ranged from 0.5 to 2 mg/dl, with a mean value of 0.96 ± 0.3 in patients without varices and a mean value of 1.3 ± 0.48 in patients with varices. Serum urea ranged from 22 to 110 mg/dl, with a mean value of 37.17 ± 16.85 in patients without varices and a mean value of 51.78 ± 20.48 in patients with varices.
Regarding the presence of EV, 58% (116 cases) of patients had EV and 42% (84 cases) had no EV. According to Child–Pugh classification, this study revealed that 36% were Child A, 29% were Child B, and 35% were Child C. There was a statistically significant difference between the studied groups in presence of EV regarding Child score (P = 0.001; [Table 1]).
|Table 1: Child classification, abdominal ultrasound, and Doppler of portal vein in studied groups|
Click here to view
There was a higher statistically significant difference between the studied groups (variceal and nonvariceal patients) in prediction of presence of EV regarding PVD, spleen, and PCI and statistically significant difference regarding portal vein velocity in prediction of presence of EV [Table 1].
There was a higher statistically significant difference between the studied groups regarding spleen size and PC/SD in prediction of bleeding varices with a mean value of 16.09 ± 3.2 and 688.59 ± 601.46, respectively (P = 0.008 and 0.003, respectively) and less statistical significant difference regarding PVD (P = 0.01), but there was no statistical significant difference between the studied groups as regarding PCI in prediction of bleeding varices (P = 0.7; [Table 2]).
|Table 2: Portal congestion index in comparison with other predictors in bleeding varices|
Click here to view
Regarding grading of EV, there was a higher statistically significant difference between PC/SD (mean ± SD 618.86 ± 219.9, 549.43 ± 47.53, 359.27 ± 45.62, and 212.45 ± 24.74 in grades 1, 2, 3, and 4, respectively), PVD (11.3 ± 2.3, 13.3 ± 0.5, 13.7 ± 1.9, and 15.2 ± 1.14 in grades 1, 2, 3, and 4, respectively) (P = 0.0001) but less significant difference between PCI (0.14 ± 0.11, 0.24 ± 0.01, 0.26 ± 0.22, and 0.38 ± 0.32 in grades 1, 2, 3, and 4, respectively) and grades of EV (P = 0.012; [Table 3]).
|Table 3: Comparison between different predictors, clinical findings, and grades of esophageal varices|
Click here to view
There was a highly statistical significant difference between grades of EV and child classification, spider nevi, presence of ascites, jaundice, and splenic collaterals by ultrasound (P = 0.0001), and a statistical significant difference between grades of EV and bleeding presentation (P = 0.004), and no significant difference between grades of EV and encephalopathy presentation nor sex (P = 0.07 and 0.3, respectively; [Table 3]).
There was a highly statistically significant difference between portal hypertensive gastropathy (PHG), fundal varices with or without EV, and PC/SD (420.04 ± 141.08 and 312.62 ± 69.95 with EV, respectively), PVD (13.3 ± 1.8 and 15.5 ± 0.7 with EV, respectively) and PCI (0.34 ± 0.3 and 0.16 ± 0.01 with EV, respectively) (P = 0.0001; [Table 4]).
At cutoff point of 11.5, the sensitivity of PVD that can predict EV was 79%, the specificity was 61.9%, and area under the curve (AUC) 0.84, whereas at cutoff point 0.11, the sensitivity of PCI that can predict EV was 93%, the specificity was 85.7%, and AUC = 0.889, and at cutoff point of 605, the sensitivity of PC/SD that can predict EV was 83.4%, the specificity was 70.3%, and AUC 0.796 [Table 5].
|Table 5: Accuracy of parameters in prediction of the presence of esophageal varices|
Click here to view
| Discussion|| |
In this study, there was a statistically significant difference between the studied groups in presence of EV as regarding Child score.
In this study, 48.3% of patients with EV were Child C versus 16.6% without EV. On the contrary, 43.1% of patients with EV were Child B versus 9.5% without EV. This finding is proven by Nashaat et al.  and also by Zaman et al. , who reported that patients in Child B or C are nearly three times more likely to have varices on upper endoscopy than are those in Child A depending on the fact that Child–Pugh class is a well-validated classification for the degree of hepatic function in patients with cirrhosis.
Regarding thrombocytopenia, there was a highly statistical significant difference between the studied groups. These results are in agreement with those reported by Thomopulos et al. , who concluded that thrombocytopenia (PC <118.000) can predict the presence of EV in cirrhotic patient, and Khalil et al. , who reported that a mean value of platelet count was 115.2 ± 57.9 for prediction the presence of EV.
This study found that PCI had no value in the identification of patients with cirrhosis at risk of variceal bleeding, but there was a highly statistical significant difference between spleen size, PC/SD, and patients with or without bleeding varices and also a statistically significant difference regarding PVD and patients with or without bleeding varices.
These results are in agreement with those reported by Feng-Hua et al.  and Hekmatnia et al. , who reported also no significant difference between Doppler indices and prediction of varices at time of bleeding. These results can be explained, as these patients were investigated for Doppler study the next day of admission and even after endoscopy procedure with either band ligation or sclerotherapy of the varices found. So, the PCI was not conclusive at the time of bleeding to predict which patients with PH at the extent to have the risk of bleeding.
Regarding PCI, there was a highly statistically significant difference between the index and presence of EV. These results are in agreement with those reported by Hekmatnia et al.  and Leão et al. , where the index was significant in prediction of the presence of EV.
Regarding portal vein flow velocity, there was a highly significant difference for prediction presence of EV. Portal vein flow velocity was lower in patients with EV when compared with patients without EV. These results agree with Elbarbary et al. , who reported that portal vein flow velocity has significant difference for prediction presence of EV.
Regarding PVD and spleen size, there was a highly significant difference for prediction of the presence of EV. These results agree with Anda et al. , who reported that PVD and spleen size have significant difference for prediction presence of EV.
This study found that platelet count and PC/SD ratio had high significant difference for grading of EV. This is because PH causes splenomegaly and thrombocytopenia. Shekar et al.  and Elhady et al.  also reported that platelet count and PC/SD ratio had highly significant difference for grading of EV.
This study reported that PCI was significant in grading of EV. Anda et al. , agreed with this study in prediction grading of EV. However, Hekmatnia et al.  reported that EV grade had no significant relationship with PCI.
Regarding portal vein flow velocity, PVD, and splenic collaterals, there was a high significant difference for grading EV. These results agree with Sheta et al. , who reported that portal vein flow velocity has significant difference for grading of EV, and Anda et al. , who reported that portal vein flow velocity, PVD, and splenic collaterals have significant difference for prediction grading of EV.
There was a high statistical significant difference between PH gastropathy, fundal varices with or without EV and platelet count, PC/SD, PVD, portal vein velocity, and PCI.
In this study, the best cutoff value of PVD for prediction of EV was 11.5 and area under receiver operating characteristic (AUROC) was 0.84. Berzigotti et al.  found the best cutoff value of PVD for prediction of EV was 12.7 ± 1.8, with AUROC of 0.779. Elbarbary et al.  found the best cutoff value of PVD for prediction of EV was 14.260 ± 1.690, with AUROC 0.887. In contrast, Jaheen et al.  found that PV dilatation is a specific but insensitive indicator of PH; they noted that this sign may also occur in the absence of portal hypertension (e.g. in response to massive splenomegaly or acute PV thrombosis).
The best cutoff value of PC/SD ratio for prediction of EV was less than 605 and AUROC was 0.796.
These results are in agreement with Shekar et al. , who reported that the best cutoff value is less than 608 with sensitivity of 80.77% and specificity of 64%. Sheta et al.  reported that there is a significant relation of PC/SD ratio for the presence of EV at cutoff value of less than 570, sensitivity of 77.1%, specificity of 93.02%, PPV 93.6%, and NPV 75.5%. Elhady et al.  reported the best cutoff value for prediction of OV was less than 750, with sensitivity of 81%, specificity 81%, PPV of 94.1%, NPV of 53.1%, and accuracy of 81%.
The differences between the best cutoff values, sensitivity, specificity, and accuracy in this study and other studies may be attributed to several factors influencing the platelet count, including infection and lower thrombopoietin levels in patients with liver cirrhosis. This finding is proven by Madhotra et al.  who reported that 32% of patients had platelet count less than 68 000/mm without detectable splenomegaly, which may be explained by insufficient synthesis of thrombopoietin.
The best cutoff value for prediction of EV regarding PCI was 0.11 and AUROC was 0.889. This is in agreement with Moriyasu et al. , who found PCI to be 0.11 ± 0.08 in chronic viral hepatitis. Elbarbary et al.  found PCI to be 0.104 ± 0.039 with AUROC 0.900. Berzigotti et al.  found the best cutoff value of PCI for prediction of EV was 0.13 ± 0.06 with AUROC 0.779. Farias et al. , found that the PCI was a sensitive marker of clinically significant PH (average 0.04 ± 0.02). However, it disagrees with Hekmatnia et al. , who found PCI to be 0.05.
This difference could be attributed to the interobserver variability or to the difference in race of the population studied. The interobserver difference in PCI could be attributed to the wide range of velocities obtained when the Doppler angle is changed from 0° to 60° (≤60° of Doppler angle is considered acceptable for clinical usage) .
| Conclusion|| |
PCI, portal vein velocity, PVD, and PC/SD can be helpful to the physicians to be used as noninvasive predictors of EV and grading varices to restrict the use of endoscopic screening. PC/SD can help physicians as a noninvasive predictor of bleeding varices.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
De Franchis R. Non-invasive and minimally invasive diagnosis of oesophagealvarices. J Hepatol 2008; 49
Turon F, Casu S, Hernández-Gea V, Garcia-Pagán JC. Variceal and other portal hypertension related bleeding. Best Pract Res Clin Gastroenterol 2013; 27
Garcia-Tsao G, Goldman L, Schafer AI. Cirrhosis and its sequelae in Chapter 156 Goldman's Cecil Medicine
ed. Philadelphia: Elsevier/Saunders; 2012. pp. 999–1107.
Bratovic I, Lacevic N. Management of esophageal varices. Med Arh 2002; 56
Berzigotti A, Seijo S, Reverter E, Bosch J. Assessing portal hypertension in liver diseases. Expert Rev Gastroenterol Hepatol 2013; 7
Merkel C, Zoli M, Siringo S, van Buuren H, Magalotti D, Angeli P et al
. Prognostic indicators of risk for first variceal bleeding in cirrhosis: a multicenter study in 711 patients to validate and improve the North Italian Endoscopic Club (NIEC) index. Am J Gastroenterol 2009; 95
Giannini EG, Zaman A, Kreil A, Floreani A, Dulbecco P, Testa E et al
. Platelet count/spleen diameter ratio for the noninvasive diagnosis of esophageal varices: results of a multicenter, prospective, validation study. Am J Gastroenterol 2006; 101
Madhotra R, Mulcahy HE, Willner I, Reuben A. Prediction of esophageal varices in patients with cirrhosis. J Clin Gastroentrol 2002; 34
Westaby D, Macdougall BRD, Saunders JB. A study of risk factors in patients with cirrhosis and variceal bleeding. Hepatology (AASLD) 1982; 3
Nashaat EH, Abd-Elaziz H, Sabry M, Ibrahim AA. Non endoscopic predictors of esophageal varices and portal hypertensive gastropathy. Nat Sci 2010; 8
Zaman A, Becker T, Lapidus J, Benner K. Risk factors for the presence of varices in cirrhotic patients without a history of variceal hemorrhage. Arch Intern Med 2001; 161
Thomopoulos KC, Labropoulou-Karatza C, Mimidis KP, Katsakoulis EC, Iconomou G, Nikolopoulou VN. Non-invasive predictors of the presence of large oesophageal varices in patients with cirrhosis. Dig Liver Dis 2003; 35
Khalil FA, Khalil KA, Khalil TH, Hassan AA, Serwah MA. Evaluation of clinical, biochemical and ultrasound parameters in diagnosis of oesophageal varices. Med J Cairo Univ 2010; 78
Li FH, Hao J, Xia JG, Li HL, Fang H. Hemodynamic analysis of esophageal varices in patients with liver cirrhosis using color Doppler ultrasound. World J Gastroenterol 2005; 11
Hekmatnia A, Barikbin R, Farghadani M, Omidifar N, Adibi P. Prediction and screening of esophageal varices in cirrhotic patients using doppler US hemodynamic indices of portal system. Gastroenterol Insights 2011; 3
Leão J, Brock M, Castilho M, Scariot A, Scariot A, Braga W. Non-invasive assessment of fibrosis using color doppler ultrasound in patients with hepatitis c virus in the Amazon Rainforest, Brazil. Am J Trop Med Hyg 2012; 86
Elbarbary AA, Elbedewy MM, Elbadry AM. Hemodynamic analysis of portal hypertension in patients with liver cirrhosis. Tanta Med J 2014; 42
Anda AC, Bordei P, Dumitru E. The role of ultrasonography in the evaluation of portal hemodynamics in healthy adults and pathologic conditions. ARS Medica Tomitana 2016; 2
Shekar CG, Balaji B, Shekar CV, Sammaiah P, Kathyayani R, Kishan B. Study of non-invasive predictors of oesophageal varices in chronic liver disease. Int J Res Pharmacol Pharmacother 2016; 5
Elhady AH, Hammam A, Elnimr A. Evaluation of some non invasive predictors for presence of esophageal varices in patients with compensated HCV positive cirrhosis. Int J Sci Res 2013; 5
Sheta EA, Yousef M, Abd-Elsalam S, Mohamed RE, Ismail A, El-Kalla F, et al
. Non invasive diagnosis of esophageal varices: can it replace screening endoscopy. Int J Curr Microbiol Appl Sci 2016; 5
Berzigotti A, Gilabert R, Abraldes JG, Nicolau C, Bru C, Bosch J, et al
. Noninvasive prediction of clinically significant portal hypertension and esophageal varices in patients with compensated liver cirrhosis. Am J Gastroenterol 2008; 103
Jaheen A. Prevalence of portal hypertensive gastropathy and its predictive factors. Gut Rev 2002; 3
Moriyasu F, Nishida O, Ban N, Nakamura T, Sakai M, Miyake T, et al
. Congestion index of the portal vein. Am J Roentgenol 1986; 146
Farias AQ, Kassab F, Da Rocha EC, Dos Santos Bomfim V, Vezozzo DC, Bittencourt PL, et al
. Propranolol reduces variceal pressure and wall tension in schistosomiasis presinusoidal portal hypertension. J Gastroenterol Hepatol 2009; 24
Park MY, Jung SE, Byun JY, Kim JH, Joo GE. Effect of beam-flow angle on velocity measurements in modern Doppler ultrasound systems. Am J Roentgenol 2012; 198
[Table 1], [Table 2], [Table 3], [Table 4], [Table 5]