|Year : 2014 | Volume
| Issue : 3 | Page : 83-91
A comparative study of Duplex Doppler ultrasound and blood indices as noninvasive predictors of oesophageal varices in cirrhotic patients
Mona Shehata1, Lobna A AboAli1, Khalid El-Shafey2, Marwa El-Hossary1
1 Department of Tropical Medicine, Faculty of Medicine, Tanta University, Tanta, Egypt
2 Department of Radiology, Faculty of Medicine, Tanta University, Tanta, Egypt
|Date of Submission||13-Apr-2014|
|Date of Acceptance||10-Jun-2014|
|Date of Web Publication||29-Oct-2014|
Lobna A AboAli
MD, 88 Saeed St Almokaoloon Alarab Building 8th Floor, Flat 47, Tanta, 31521
Endoscopic surveillance of oesophageal varices (OV) in cirrhotic patients is expensive and uncomfortable for the patients. Therefore, there is a particular need for noninvasive predictors for OV.
The aim of the present study was to evaluate the accuracy of ultrasound indices and blood indices as noninvasive OV predictors among cirrhotic patients.
Patients and methods
A total of 61 cirrhotic patients were enrolled in this study and were divided into two groups: 21 patients without OV and 40 patients with OV who were further subdivided into 24 patients with small OV and 16 with large oesophageal varices (LOV). P2/MS, serum fibrosis markers (APRI, FIB4, Lok score, and Forns index), abdominal ultrasonography [portal vein diameter (PVD), splenic index], platelet count/spleen diameter ratio (PC/SD), and Doppler ultrasonography [portal vein velocity, splenoportal index, hepatic and splenic impedance indices, and hepatic venous waveform (HVWF)] were assessed in all patients.
P2/MS was the best predictor of OV and LOV [area under the curve (AUC) 0.88 and 0.787, respectively] followed by PC/SD (AUC 0.77 and 0.715, respectively). PVD, serum fibrosis markers, and serum albumin had the least accuracy for OV prediction. For LOV predictions, Lok score had good accuracy (AUC 0.785) followed by serum albumin, PVD, APRI, and Forns index (AUC 0.72, 0.738, 0.734, and 0.738, respectively). Monophasic HVWF showed a good positive predictive value (85%) and specificity (80.95%) for prediction of OV and good sensitivity (81.25%) and negative predictive value (81.25%) for LOV.
P2/MS can identify OV and LOV in cirrhosis with high accuracy followed by PC/SD. Monophasic HVWF is a good noninvasive predictor of OV and LOV in cirrhotic patients.
Keywords: Duplex Doppler, oesophageal varices, fibrosis markers, noninvasive predictors
|How to cite this article:|
Shehata M, AboAli LA, El-Shafey K, El-Hossary M. A comparative study of Duplex Doppler ultrasound and blood indices as noninvasive predictors of oesophageal varices in cirrhotic patients. Tanta Med J 2014;42:83-91
|How to cite this URL:|
Shehata M, AboAli LA, El-Shafey K, El-Hossary M. A comparative study of Duplex Doppler ultrasound and blood indices as noninvasive predictors of oesophageal varices in cirrhotic patients. Tanta Med J [serial online] 2014 [cited 2017 Dec 12];42:83-91. Available from: http://www.tdj.eg.net/text.asp?2014/42/3/83/143552
| Introduction|| |
Oesophageal varices (OV) are the most common clinical manifestations of portal hypertension (PH) in patients with liver cirrhosis. Varices are present in 30-40% of patients with compensated cirrhosis and in 60-80% of patients with decompensate cirrhosis . Bleeding from OV is the most clinically relevant complication of cirrhosis and carries a high risk for morbidity and mortality .
Therefore, periodic screening endoscopy for OV and prophylactic treatment for high-risk OV is currently recommended to all patients with established cirrhosis. However, many patients screened either do not have varices or have nonrisky varices that do not require prophylactic therapy. Hence, endoscopic screening in these patients appears to carry a needless burden of stress and expense on patients and medical facilities. In addition, compliance will be limited in some patients . Accordingly, an ideal noninvasive measure is needed for diagnosing OV before invasive endoscopy.
The estimation of portal blood flow volume with Duplex Doppler ultrasonography (US) is noninvasive and is used to explore the relationship between OV hemodynamics associated with PH and liver cirrhosis . The main characteristics of PH such as a decrease in portal vein velocity (PVV) or an increase in portal vein diameter (PVD) are detectable also by these means .
A new index P2/MS, based on complete blood count, is specifically designed to predict OV in chronic liver disease . However, P2/MS is a simple and economic test; it was not found to be studied previously among Egyptian patients. In addition, serum fibrosis markers, APRI , FIB4 , Lok score , and Forns index , have been tested as noninvasive predictors of OV in cirrhotic patients with controversial results . This is based on the concept that the development of PH is due to liver fibrosis as the most important factor contributing to increased hepatic resistance .
This work was a comparative study to evaluate the accuracy of Duplex Doppler, conventional US indices, and blood indices (P2/MS, APRI, FIB4, Forns index, and Lok score) as noninvasive predictors of OV in Egyptian cirrhotic patients.
| Patients and methods|| |
This study was carried out on 61 cirrhotic Egyptian patients who were enrolled from Tropical Medicine and Radiology Departments, Tanta University Hospital. They were divided into the following groups:
Group 1 that included 21 cirrhotic patients without OV.
Group 2 that included 40 cirrhotic patients with OV who were further subdivided according to variceal size into 24 patients with small OV and 16 with large oesophageal varices (LOV).
Patients with past history of previous variceal bleeding, previous endoscopic treatment for OV such as band ligation or sclerotherapy or surgical intervention for PH, patients on β-blockers, and patients with hepatocellular carcinoma, portal or splenic vein thrombosis, or active infection were excluded from the study.
All participants in the study were subjected to the following:
(1) Detailed history taking and full clinical examination.
(2) Laboratory investigations including complete blood count, liver function tests, blood urea and serum creatinine, serum cholesterol, hepatitis C virus antibodies, HBsAg, and antibilharzial antibodies by indirect hemagglutination test.
(3) Child-Pugh score classification.
(4) Diagnosis of liver cirrhosis or mixed liver disease was based on the clinical, laboratory, and US data. Liver biopsy defined the diagnosis in individual patients who were enrolled for hepatitis C virus interferon therapy.
(5) Calculation of P2/MS index :
P2/MSindex = (platelet count) 2 /(monocyte fraction % × segmented neutrophil fraction %).
(6) Calculation of serum liver fibrosis markers :
Aspartate aminotrans ferase (AST) to platelets ratio index (APRI) = [(AST/ULN) ×100]/platelet count (10 9 /l), where ULN is the upper limit of normal.
FIB 4 = [age (years) × AST (IU/l)]/[platelet count (10 9 /l) × alanine aminotrans ferase (ALT) (IU/l)/2].
Forns index = 7.811 − 3.131 × ln [platelet count (10 9 /l)] + 0.781 × ln [γ-glutamyl transpeptidase (IU/l)] + 3.467 × ln [age (years)] -0.014 [cholesterol (mg/dl)].
Lok score: log odds = 5.56-0.0089 × platelet count (10 3 /mm 3 ) + 1.26 × (AST/ALT) + 5.27 × international normalized ratio.
Lok = [exp (log odds)]/[1 + exp (log odds)].
(7) Abdominal US: It evaluated liver and spleen size, the presence of cirrhosis, periportal fibrosis, ascites, or focal lesions. PVD was detected, with calculation of splenic index = splenic length × splenic width  and platelet count/spleen diameter ratio (PC/SD) .
(8) Doppler US study: It was performed for calculation of PVV, splenoportal index (SPI), and hepatic and splenic impedance indices with detection of hepatic venous waveform (HVWF).
(9) Upper endoscopy: The varices were classified according to Comar and Sanyal  into: grade I, varix that is flush with the wall of the esophagus; grade II, protrusion of the varix but not more than half way to the center of the lumen; grade III, protrusion of the varix more than half way to the lumen; and grade IV, the varices are so large that they meet at the midline. OV were defined risky according to size (grade III and IV), the presence of red signs, and advanced Child score . The presence of risky red signs was noted: red wale marks, cherry-red spots, and hematocystic diffuse spots .
All patients gave their informed written consents, and the study was approved by the Ethical and Research Committee, Tanta Faculty of Medicine, Tanta, Egypt.
The collected data were tabulated and statistically analyzed using statistical package for social studies (version 19, SPSS Inc., Chicago, Illinois, USA). Categorical variables were represented by number and percentage, and observed differences were tested using the Monte-Carlo exact test. Numerical variables were presented as number and percentage. The mean differences were tested by Student's t-test when comparing between two variables. Receiver-operator characteristic (ROC) curve was used to determine the cutoff value for diagnosis. Thereafter, sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy of each cutoff variable were calculated. The level of significance was adopted at P value less than 0.05.
| Results|| |
The demographic, laboratory, and endoscopic profile of the studied groups are summarized in [Table 1]. Serum albumin was significantly decreased in cirrhotic patients with OV (P = 0.001).
|Table 1: Demographic, laboratory, and endoscopic profi le of the studied groups|
Click here to view
In cirrhotic patients with OV, P2/MS and PC/SD were significantly decreased, whereas APRI, Forns index, Lok score, and PVD showed a significant increase when compared with cirrhotic patients without varices (P < 0.05) [Table 2].
|Table 2: P2/MS, serum fi brosis markers, and ultrasound parameters in the studied groups|
Click here to view
Duplex Doppler US parameters in the studied groups [Table 2] and [Figure 1],[Figure 2] and [Figure 3], such as PVV, SPI, hepatic artery resistive index (HARI), hepatic artery pulsatility index, splenic artery resistive index, and splenic artery pulsatility index, did not show significant difference between the groups (P > 0.05).
Regarding HVWF [Table 3] and [Figure 4], in cirrhotic patients without varices, 19% of the patients had monophasic waveform, 66.7% had biphasic waveform, and 14.3% had triphasic waveform, whereas in cirrhotic patients with varices, 60% of the patients had monophasic waveform, 27.5% had biphasic waveform, and 12.5% had triphasic waveform. Comparison between the two groups showed a statistically significant difference (P = 0.001). The sensitivity of monophasic waveform in the prediction of the presence of OV equals 60%, specificity 80.95%, PPV 85.7%, NPV 51.5%, and accuracy 67.21%.
The comparative evaluation of the studied significant parameters [Table 4] showed that the predictors for the presence of OV were P2/MS (cutoff 32.67), PC/SD (cutoff 603), monophasic HVWF, PVD (cutoff 13.45), serum fibrosis markers (APRI, Lok score, and Forns index), and serum albumin. The sensitivity, specificity, accuracy, PPV, NPV, cutoff value, and area under the receiver-operator characteristic curve (AUROC) of these parameters were presented in [Table 4].
|Table 4: Receiver– operator characteristic curve results for different diagnostic tests for the presence of oesophageal varices|
Click here to view
When the patients were classified according to the size of varices into patients with small OV (grade I, II) and patients with LOV (grade III, IV) [Table 5], in cirrhotic patients with large varices, serum albumin, P2/MS, and PC/SD were significantly decreased, whereas APRI, Forns index, Lok score, and PVD showed a significant increase when compared with cirrhotic patients with small varices (P < 0.05).
|Table 5: Serum albumin, P2/MS, serum fibrosis markers, and ultrasound indices (portal vein diameter and platelet count/spleen diameter ratio) in cirrhotic patients with small-sized varices and cirrhotic patients with large varices|
Click here to view
HVWF [Table 3] and [Figure 4] showed a significant difference (P = 0.048) with respect to OV size. In cirrhotic patients with small varices, 45.8% had monophasic waveform, 33.3% had biphasic waveform, and 20.8% had triphasic waveform, whereas in cirrhotic patients with large varices, 81.25% had monophasic waveform, 18.75% had biphasic waveform, and none had triphasic waveform. Regarding HVWF, the sensitivity of monophasic waveform in the prediction of LOV equals 81.25%, specificity 54.16%, PPV 59.16%, NPV 81.25%, and accuracy 65%.
The comparative evaluation of the studied significant parameters [Table 6] showed that the predictors for the presence of LOV were P2/MS (cutoff 22.6), PC/SD (cutoff 500), Lok score (cutoff 0.72), serum albumin (cutoff 2.9), PVD (cutoff 14.2), APRI, Forns index, and monophasic HVWF. The sensitivity, specificity, accuracy, PPV, NPV, cutoff value, and AUROC of these parameters are presented in [Table 6].
|Table 6: Receiver– operator characteristic curve results for different diagnostic tests for large oesophageal varices|
Click here to view
| Discussion|| |
Bleeding from OV is the most clinically relevant complication of cirrhosis and still carries a mortality of up to 20% within 6 weeks of the bleeding episode. The risk of bleeding is related to the size of varices, presence of red signs, and decompensated cirrhosis. The American Association for the Study of Liver Diseases and the Baveno IV Consensus Conference on PH recommended that all cirrhotic patients should undergo endoscopy to assess the presence, the size, and the aspect of varices at the time of liver cirrhosis diagnosis. If no varices are present at index endoscopy, this procedure should be repeated at 2-3 years in compensated cirrhosis and annually in decompensated cirrhosis . However, a generalized screening program of periodical upper endoscopy in cirrhotic patients may lead to high cost and low compliance, as the procedure is invasive and may be poorly accepted by the patients if repeatedly required . Moreover, sedation of cirrhotic patients to perform upper endoscopy may be hazardous, and diagnostic upper endoscopy may contribute to bacterial infections due to disruption of the normal barriers . For these reasons, the selection of patients who may be at risk of having OV, especially those at risk for rupture, would be highly beneficial and cost effective . The need for noninvasive diagnosis for detection of OV and assessing the effect of therapy will benefit in high-risk situations .
The use of noninvasive predictive model would help to evaluate the presence of OV in a selective group of patients who are candidates for primary prophylaxis. Results from trials on primary prophylaxis for variceal bleeding have shown that giving β-blocker therapy without documenting OV has no advantage, causes high cost, and leads to complications .
The aim of this study was to evaluate the accuracy of Duplex Doppler, conventional US indices, and blood indices (P2/MS, APRI, FIB4, Forns index, and Lok score) as noninvasive predictors of OV in Egyptian cirrhotic patients.
This work revealed that the best predictive result was obtained by P2/MS as a newly emerged simple blood index. It could predict the presence of OV at a cutoff of 32.67 or less (AUROC 0.883), with PPV of 98.61% and high accuracy (89.65%). In addition, P2/MS could predict LOV at a cutoff of 22.62 or less (AUROC 0.787), with PPV of 91.82% and high accuracy (82.16%). These findings are in agreement with the findings of Lee et al.  who found that, at a cutoff value less than 13.0, P2/MS could detect OV (AUROC 0.916) with specificity of 92.6% and at a value greater than 39.0, P2/MS ruled out OV with a sensitivity of 94.2%. In addition, Kim et al.  reported a good performance of P2/MS (AUROC 0.915) in predicting OV, and P2/MS at a cutoff of 25 identified patients as having no varices.
The difference in the cutoff values in Lee et al.  and Kim et al.  studies from the present one may be explained by difference in the etiology of cirrhosis; all their cases had postviral hepatitis and none of them was additionally bilharzial. In addition, the present study did not include any alcoholics.
The validity of P2/MS as a surrogate for the progression of PH and OV can be explained; as fibrosis and PH increase, the spleen enlarges with increased sequestration and destruction of platelets together with decreased thrombopoietin production. PH may also promote sequestration of granulocytes and red blood cells, in addition to platelets in the spleen, and the decrease in circulating granulocytes induces a compensatory increase in serum granulocyte monocyte-colony stimulating factor. The proportion of neutrophils and monocytes may then rise, as granulocyte monocyte-colony stimulating factor stimulates the production of these cells more actively than the production of lymphocytes .
P2/MS is a parameter with unique features; first, it is a noninvasive index that could reflect both the presence of OV and the degree of liver fibrosis with a high specificity according to Lee et al. . Second, P2/MS can be obtained from complete blood count data alone, which is simple, inexpensive, and performed routinely. Third, P2/MS uses the differential white blood cell count profiles, which have never been used in this field. Fourth, P2/MS was first evaluated, in this study, on Egyptian patients with greatly promising and interesting results.
In our study, PC/SD was reported to be the second best positive predictor for both OV presence (PPV 93.77%) and LOV (PPV 83.58%) after P2/MS. It could predict the presence of OV at a cutoff value of 603 with AUROC 0.774 and the presence of LOV at cutoff value of 500 with AUROC 0.715.
The present PC/SD findings agree with the Egyptian study conducted by Said et al.  who reported similar results of PC/SD at cutoff value of 415 for prediction of OV.
Giannini et al. , Nashaat et al. , and Agha et al.  showed a cutoff value of 909, 820, and 885, respectively, for prediction of OV. Arulprakash et al.  and Alassan et al.  found a cutoff of PC/SD of 909 and 897 or less, respectively, for diagnosis of LOV. In addition, Berzigotti et al.  stated PC/SD as a predictor for clinically significant PH with AUROC 0.811 and used PH risk score by combination of liver stiffness measurement, the sex of the patient, and the PC/SD for better accuracy (AUROC 0.935).
The difference in the cutoff values between these studies and that of the present work can be explained by their patient sample that included only cirrhotic patients, and none of their patients had evidence of bilharziasis, whereas most of the patients included in the present study had mixed disease etiology: bilharzial and postviral hepatitis C cirrhosis. Both had their insult on the platelets, besides bilharziasis that specifically produces larger even huge spleen. None of these studies included these types of patients that are characteristically prevalent in the Delta region of Egypt. Hence, it is sure to have our different and specific cutoff that needs further larger scale studies in Egypt, being the second best predictive parameter in this study after P2/MS.
PC/SD is a parameter linking splenomegaly implicated in thrombocytopenia of cirrhosis with spleen size being inversely proportional to platelet count. The use of this ratio normalizes platelet count to splenic sequestration, as platelet count alone may be misleading as it cannot be solely attributed to PH .
In this work, PVD could predict the presence of OV at a cutoff value greater than 13.45 (AUROC 0.733) with sensitivity 73.3%, specificity 61.9 %, and PPV 84.1%. PVD results of this study agree with Nashaat et al. , Muhammad et al. , and Bintintan et al.  studies in which PVD with a cutoff of 13.5, greater than 13, and greater than 13 mm, respectively, could predict OV.
In addition, PVD could predict the presence of LOV at a cutoff value greater than 14.2 with AUROC 0.738, sensitivity 72.7%, and specificity 69%. PVD was the third best positive predictor of LOV (PPV 86.85%) with close accuracy to serum albumin, which is reported in this study as the best negative predictor of LOV; hence, both could have complementary results. The combination of two different complementary parameters provides theoretical advantages over the use of a single one, as false results may be overcome leading to more accurate prediction .
In agreement with these results, Hong et al.  and Arulprakash et al.  reported that PVD greater than 11.75 mm and greater than 13.9 mm, respectively, could predict LOV. In addition, Berzigotti et al.  stated that PVD greater than 13 mm and inversion of flow within the portal system are 100% specific for clinically significant PH with strong association with variceal formation and growth.
In the present study, HVWF (monophasic) could predict the presence of OV. The monophasic waveform showed a high PPV of 85.7%, close to PVD, as the third best positive predictor of OV after P2/MS and PC/SD. Both PVD and HVWF, similar to US data, are suggested to have complementary predictive values for OV presence, as PVD had a higher sensitivity and HVWF had a higher specificity.
Consistently, Baik et al.  found a correlation between hepatic venous pressure gradient (HVPG) and HVWF; when HVPG increased, the waveform became flat. Monophasic waveform in hepatic vein (HV) was associated with severe PH (HVPG > 15 mmHg) with high sensitivity and specificity. In addition, Khalil et al.  found that monophasic and biphasic patterns of hepatic veins significantly exist in patients with OV. Furthermore, in agreement with the present study, Thomas et al.  and Payangappad et al.  found a significant correlation between loss of normal HVWF (triphasic pattern) and LOV and concluded that the monophasic HVWF was a good noninvasive predictor of OV. Abdul Rabb et al.  found no significant correlation between HVWF changes and the grading of OV in cirrhotic patients.
HVWF use depends on the fact that normal HVWF shows a triphasic pattern and loss of this pattern in cirrhosis is mainly due to decreased liver compliance . However, Doppler US has some limitations such as massive ascites, severe liver steatosis, and interobserver variability . Monophasic HVWF could predict LOV existence; HVWF appeared close to serum albumin as the relatively best negative predictors for LOV recorded in this study with NPV 81.25%, being higher than its PPV, with a good sensitivity 81.25%.
In our study, serum albumin could predict the presence of OV at a cutoff of 3.8 or less (AUROC 0.661) and LOV at a cutoff of 2.9 or less (AUROC 0.72), being the best negative predictor for LOV among the studied parameters. These data agree with the study by Muhammad et al.  who found that serum albumin of 2.8 g/dl or less had very high sensitivity and specificity in predicting OV and with the study by Galal et al.  who found that serum albumin with cutoff of 3.2 was predictive for the presence of OV and LOV but with poor performance (AUROC 0.381 and 0.382, respectively).
Hypoalbuminemia in cirrhosis is multifactorial and may be due to reduced albumin production in cirrhotic liver or increased loss through the gut (portal hypertensive gastropathy/enteropathy), all related to PH, and this explains the significant decrease in serum albumin in cirrhotic patients with varices compared with the nonvariceal group found in this study.
With respect to serum fibrosis markers, our study showed that APRI at a cutoff greater than 1.26 could predict the presence of OV (AUROC 0.695) with PPV of 81.42%. APRI at a cutoff greater than 1.47 could predict LOV (AUROC 0.734). These findings are in agreement with the studies by Castéra et al. , Tafarel et al. , and Adami et al.  who proposed APRI at a cutoff 1.3, 1.64, and greater than 1.4, respectively, for prediction of OV. In addition, Sebastiani et al.  reported APRI at a cutoff of 1.4 for prediction of OV and at a cutoff of 1.5 for detection of LOV. However, Stefanescu et al.  suggested APRI at a cutoff more than 2.201 with AUROC 0.538 for detection of LOV with least accuracy among other serum fibrosis markers, and Galal et al.  suggested a cutoff greater than 0.16 for detection of OV and LOV existence with AUROC 0.642. These different cutoff results indicate the need for further studies in large-scaled studies.
Serum fibrosis markers use as noninvasive predictors of OV is based on the concept that they indicate more severe hepatic parenchyma architectural distortion (represented by fibrosis and sinusoidal capillarization) and increased intrahepatic circulatory resistance, resulting in PH .
The present study also reported that Lok score at a cutoff greater than 0.61 could predict the presence of OV (AUROC 0.675) and at a cutoff value of 0.72 could predict LOV (AUROC 0.785). Lok score had the best accuracy and PPV for LOV prediction among the studied serum fibrosis markers. These findings agree with the studies by Castéra et al.  and Stefanescu et al.  who proposed a cutoff of at least 0.6 and greater than 0.62 for detection of OV and a cutoff of 0.62 and greater than 0.796, respectively, for detection of LOV. However, Sebastiani et al.  proposed a Lok score cutoff of 0.9 (AUROC 0.77) with best accuracy among serum fibrosis markers for detection of OV and proposed a cutoff of 1.5 for detection of LOV.
In this work, Forns index at a cutoff greater than 8.15 could predict the presence of OV (AUROC 0.681) and at a cutoff value 9.13, it could predict LOV (AUROC 0.738). The present findings of Forns index agree with that detected by Sebastiani et al.  and Stefanescu et al.  who proposed a cutoff for Forns index of 8.5 and greater than 7.279 for detecting OV and a cutoff of 8.8 and greater than 8.538, respectively, for detection of LOV. Both APRI and Forns index were good positive predictors of LOV (PPV > 85%). However, as being the least reported negative predictors (NPV ∼45%), this regress their relative accuracy.
FIB4 could not predict the presence of OV in the present study. This finding agree with the study conducted by Galal and colleagues who found that FIB4 was insignificant for diagnosis of OV. However, FIB4 findings disagree with the study conducted by Sebastiani and colleagues and Stefanescu and colleagues, who concluded that FIB4 could detect the presence of OV at a cutoff of 3.5 and greater than 3.98 and LOV at a cutoff of 4.3 and greater than 6.74, respectively. These different results indicate the need for further large-scale studies.
Furthermore, in this work, splenic index was found insignificant regarding the detection of OV or LOV. Splenic index findings disagree with the study by Liu et al.  who found that splenic index was independently associated with the presence of OV. This can be explained in the setting of our study sample, where most of the patients had evidence of mixed bilharzial and postviral hepatitis cirrhosis compared with the study conducted by Liu et al. , as all their patients were Child A and most of them either were alcoholic or had postviral hepatitis cirrhosis.
In addition, in this study, PVV was found insignificant in diagnosis of OV or LOV. This finding agree with the study conducted by Khalil et al.  and Puneet et al.  who found insignificant correlation between the mean PVV and the presence of OV. However, PVV findings in this study disagree with the findings of Liu et al.  and Tarzamni et al.  who reported PVV to be predictive for the presence of OV. These different results could be explained by the concept that PVV is subjective to interobserver and intraobserver variability and is affected by respiration and prandial state of the patient .
In the present study, SPI was also found insignificant regarding presence and absence of OV and LOV. These findings disagree with the results of the studies conducted by Reichen  and Liu et al.  who found that SPI could predict the presence of OV. This can be explained by difference in the study sample. As SPI depends, in its calculation, on splenic index and PVV, both of these parameters were reported as nonsignificant predictors for OV in the present work.
Moreover, in this study, both hepatic and splenic impedance indices were found insignificant in the diagnosis of OV. These findings agree with the studies by Piscaglia et al.  and Tarzamni et al.  who stated that portal vein mean velocity, congestion index, liver vascular index, and HARI are of no value in the diagnosis of PH. These results also agree with multivariate analysis conducted by Liu et al.  who found that HARI, hepatic artery pulsatility index, splenic artery resistive index, and splenic artery pulsatility index were not predictive for the presence of OV. However, Sacerdoti et al.  concluded that hepatic arterial resistance indices directly correlated with portal resistance, congestion index, and the degree of OV.
In conclusion, this work stated that P2/MS was the best positive predictor of OV and LOV in cirrhotic patients with high accuracy followed by PC/SD. P2/MS could represent a newly emerged, simple, accurate, and economical method for prediction of OV and LOV, which was first studied on Egyptian cirrhotic patients. Monophasic HVWF is a good noninvasive predictor of OV and LOV in cirrhotic patients. Application of these noninvasive predictors may reduce the need for endoscopic screening in cirrhotic patients.
| Acknowledgements|| |
Conflicts of interest
There are no conflicts of interest.
| References|| |
Garcia Taso G. Current management of complications of cirrhosis and portal hypertension: variceal haemorrhage, ascites and spontaneous bacterial peritonitis. Gastroenterology 2001; 120:726-748.
Brandenburger LA, Regenstein FG. Variceal haemorrhage. Curr Treat Options Gastroenterol 2002; 5:73-80.
De Franchis R. Evolving consensus in portal hypertension report of the Baveno IV Consensus Workshop on methodology of diagnosis and therapy in portal hypertension. J Hepatol 2005; 43:167-176.
Li FH, Hao J, Xia JG, et al.
Haemodynamic analysis of oesophageal varices in patients with liver cirrhosis using color Doppler ultrasound. World J Gastroenterol 2008; 11:4560-4565.
Gorg C, Riera-Knorrenschild J, Dietrich J. Colour Doppler ultrasound flow patterns in the portal venous system. Br J Radiol 2002; 75:919-929.
Kim BK, Han K, Park JY, et al.
Prospective validation of P2/MS noninvasive index using complete blood counts for detecting oesophageal varices in B-viral cirrhosis. Liver Int 2010; 30:860-866.
Wai CT, Greenson JK, Fontana RJ, et al.
A simple noninvasive index can predict both significant fibrosis and cirrhosis in patients with chronic hepatitis C. Hepatology 2003; 38:518-526.
Vallet-Pichard A, Mallet V, Naples B, et al.
FIB-4: an inexpensive and accurate marker of fibrosis in HCV Infection. Comparison with liver biopsy and fibrotest. Hepatology 2007; 46:32-36.
Lok AS, Ghany MG, Goodman ZD, et al.
Predicting cirrhosis in patients with hepatitis C based on standard laboratory tests: results of the HALT-C cohort. Hepatology 2005; 42:282-292.
Forns X, Ampurdanes S, Liovet JM, et al.
Identification of chronic hepatitis C patients without hepatic fibrosis by a simple predictive model. Hepatology 2002; 36:986-992.
Castéra L, Le Bail B, Roudot-Thoraval F, et al.
Early detection in routine clinical practice of cirrhosis and oesophageal varices in chronic hepatitis C: comparison of transient elastography (FibroScan) with standard laboratory tests and non-invasive scores. J Hepatol 2009; 50:59-68.
Stefanescu H, Grigorescu M, Lupsor M, et al.
A new and simple algorithm for the noninvasive assessment of oesophageal varices in cirrhotic patients using serum fibrosis markers and transient elastography. J Gastrointestin Liver Dis 2011; 20:57-64.
Liu C, Hsu S, Liang C, et al.
Oesophageal varices: noninvasive diagnosis with duplex doppler US in patients with compensated cirrhosis. Radiology 2008; 248:132-139.
Barrera F, Riquelme A, Soza A, et al.
Platelet count/spleen diameter ratio for non-invasive prediction of high risk oesophageal varices in cirrhotic patients. Ann Hepatol 2009; 8:325-330.
Comar KM, Sanyal AJ. Portal hypertensive bleeding. Gastroenterol Clin North Am 2003; 32:1079-1105.
Madhotra R, Molcahy HE, Willnar I, et al.
Prediction of oesophageal varices in patients with cirrhosis. J Clin Gastroenterol 2002; 34:81-85.
Silva G. New serum markers for predicting oesophageal varices: is it a reality? J Gastroenterol Hepatol 2013; 28:112-121.
De Franchis R. Non-invasive (and minimally invasive) diagnosis of oesophageal varices. J Hepatol 2008; 49:520-527.
Khalil F, Khalil K, Khalil T et al.
Evaluation of clinical, biochemical and ultrasound parameters in diagnosis of oesophageal varices. Med J Cairo Univ 2010; 78:105-109.
Berzigotti A, Seijo S, Reverter E, et al.
Assessing portal hypertension in liver diseases. Expert Rev Gastroenterol Hepatol 2013; 7:141-155.
Gill ML, Atiq M, Sattar S, et al.
Non-endoscopic parameters for the identification of oesophageal varices in patients with chronic hepatitis. J Pak Med Assoc 2004; 54:575.
Lee JH, Yoon JH, Lee CH, et al.
Complete blood count reflects the degree of oesophageal varices and liver fibrosis in virus-related chronic liver disease patients. J Viral Hepatol 2009; 16:444-452.
Said HE, Elsayed EY, Ameen A, et al.
Cytopenia as a predictor of oesophageal varices in patients with liver cirrhosis. Rep Opin 2010; 2: 35-41.
Giannini E, Botta F, Borro P et al.
Platelet count/spleen diameter ratio: proposal and validation of a non-invasive parameter to predict the presence of oesophageal varices in patients with liver cirrhosis. Gut 2003; 52:1200-1205.
Nashaat EH, Abd-Elaziz H, Sabry M, et al.
Non-endoscopic predictors of oesophageal varices and portal hypertensive gastropathy. Nat Sci 2010; 8:43-50.
Agha A, Anwar E, Bashir K, et al.
External validation of the platelet count/spleen diameter ratio for the diagnosis of oesophageal varices in hepatitis C virus-related cirrhosis. Dig Dis Sci 2011; 54: 654-660.
Arulprakash S, Chitra S, Muthukumaran K, et al.
Non-invasive prediction of large oesophageal varices in chronic liver disease patients. Saudi J Gastroenterol 2010; 16:38-42.
Alassan KM, Fulgence YB, Constant A, et al.
Usefulness of noninvasive predictors of oesophageal varices in black African cirrhotic patients in Cˆote d′Ivoire (West Africa). Gastroenterol Res Pract 2012; 2012:216390.
Berzigotti A, Seijo S, Arena U, et al.
Elastography, spleen size, and platelet count identify portal hypertension in patients with compensated cirrhosis. Gastroenterology 2013; 144:102-111.
Muhammad SK, Shaikh MA, Shaikh BA. Sensitivity, specificity and predictive values of noninvasive markers of oesophageal varices in cirrhosis of liver. Asian J Med Res 2012; 1:98-102.
Bintintan A, Chira RI, Mircea PA. Non-invasive ultrasound based diagnosis and staging of oesophageal varices in liver cirrhosis. A systematic review of literature published in the third millennium. Med Ultrason 2013; 15: 116-124.
Hong WD, Zhu QH, Huang ZM, et al.
Predictors of oesophageal varices in patients with HBV-related cirrhosis: a retrospective study. BMC Gastroenterol 2009; 9:11.
Baik SK, Kim JW, Kim HS, et al.
Recent variceal bleeding: doppler US hepatic vein waveform in assessment of severity of portal hypertension and vasoactive drug response. Radiology 2006; 240:574-580.
Thomas J, Madhavan M, Devadas K, et al.
Doppler assessment of hepatic venous waves for predicting large varices in cirrhotic patients. Saudi J Gastroenterol 2011; 17:36-39.
Payangappad PK, Yerol PK, Haridas A, et al.
Monophasic hepatic venous waveform: a non-invasive predictor of large oesophageal varices. J Clin Exp Hepatol 2012; 2:6-39.
Abdul Rabb B, Amanullah A, Nazish B, et al.
Hepatic vein waveform in liver cirrhosis: correlation with Child′s class and size of varices. J Pak Med Assoc 2012; 62:794-797.
Galal G, Ghweil A, Muhammad EM, et al.
Clinical utility of simple fibrosis markers in prediction of oesophageal varices in chronic hepatitis C patients with advanced cirrhosis. Med J Cairo Univ 2012; 80:85-93.
Tafarel JR, Tolentino LH, Correa LM, et al.
Prediction of oesophageal varices in hepatic cirrhosis by noninvasive markers. Eur J Gastroenterol Hepatol 2011; 23:754-758.
Adami MR, Ferreira CT, Kielig CO, et al.
Non-invasive methods for prediction of oesophageal varices in pediatric patients with portal hypertension. World J Gastroenterol 2013; 19:2053-2059.
Sebastiani G, Tempesta D, Fattovich G, et al.
Prediction of oesophageal varices in hepatic cirrhosis by simple serum non-invasive markers: results of a multicenter, large-scale study. J Hepatol 2010; 53:630-638.
Puneet M, Ranjana G, Gaurav M, et al.
Association between portal vein color Doppler findings and the severity of disease in cirrhotic patients with portal hypertension. Iran J Radiol 2011; 8:211-217.
Tarzamni MK, Somi MH, Farhang S, et al.
Portal hemodynamics as predictors of high risk oesophageal varices in cirrhotic patients. World J Gastroenterol 2008; 14:1898-1902.
Reichen J. The Splenoportal Index accurately predicted the presence of oesophageal varices in compensated cirrhosis. Evid Based Med 2006; 14:27.
Piscaglia F, Danati G, Ceciliani L, et al.
Influence of the spleen on portal haemodynamics: a non-invasive study with Doppler ultrasound in chronic liver disease and haematological disorders. Scand J Gastroenterol 2002; 37:1220-1227.
Sacerdoti D, Merkel C, Bolognesi M, et al.
Hepatic arterial resistance in cirrhosis with and without portal vein thrombosis: relationships with portal hemodynamics. Gasteroenterology 1995; 108:1152.
[Figure 1], [Figure 2], [Figure 3], [Figure 4]
[Table 1], [Table 2], [Table 3], [Table 4], [Table 5], [Table 6]