Retrospective Study
Copyright ©The Author(s) 2017.
World J Gastroenterol. Jan 28, 2017; 23(4): 687-696
Published online Jan 28, 2017. doi: 10.3748/wjg.v23.i4.687
Table 1 Demographic profile of the study population
ParameterValue (n = 326)
Gender
Males263 (81)
Females63 (19)
Age, yr52 (16-90)
BMI, kg/m223 (17-41)
Etiology
NASH/cryptogenic148 (45)
Alcohol110 (34)
Viral (HBV/HCV)48 (15)
Others (including mixed etiology)20 (6)
Ascites
None161 (49)
Mild135 (42)
Moderate to tense30 (9)
Bleeding status
Bleeder118 (36)
Non-bleeder208 (64)
Hemoglobin, g/dL10.3 (4.5-17.0)
Platelets, × 103/cumm90 (13-422)
Bilirubin, mg/dL1.6 (0.2-11.2)
AST, IU/dL53 (16-209)
INR1.3 (0.9-3.2)
Serum albumin, g/dL3.0 (1.2-4.6)
CTP score7 (5-12)
MELD score12 (6-37)
Varices present293 (90)
Esophageal varices280 (86)
Small varices170/280 (61)
Large varices110/280 (39)
Gastric varices79 (24)
Small varices52/79 (66)
Large varices27/79 (34)
HVPG, mmHg16.0 (1.5-30.5)
Transient elastography, kPa36 (3-75)
Table 2 Groups according to hepatic venous pressure gradient
HVPG (mmHg)n (%)Portal hypertension (> 5 mmHg)CSPH (10 mmHg)SPH (> 12 mmHg)VSPH (> 20 mmHg)
≤ 514 (4)No (14, 4%)No (48, 15%)No (78, 24%)No (266, 82%)
> 5 to < 1034 (10)Yes (312, 96%)
≥ 10 to 1230 (9)Yes (278, 85%)
> 12 to ≤ 20188 (58)Yes (248, 76%)
> 2060 (18)Yes (60, 18%)
Table 3 Predictive values of transient elastography for the prediction of clinically significant portal hypertension (HVPG ≥ 10 mmHg)
TE cut-off value (mmHg)CSPH (n)No CSPH (n)Total (n)Sensitivity (95%CI)Specificity (95%CI)PPV (95%CI)NPV (95%CI)Accuracy (95%CI)LR+ (95%CI)LR-(95%CI)
≥ 21.62191623579% (74%-83%)67% (52%-80%)93% (89%-96%)35% (25%-46%)77% (72%-82%)2.4 (1.6-3.5)0.3 (0.2-0.4)
< 21.6593291
Total27848326
Table 4 Various studies of the diagnostic performance of transient elastography for clinically significant portal hypertension
Ref.PlaceYearNumber of patientsCorrelation coefficient (r) or r2 of TE with HVPGAUROC (95%CI)Best cut-off of TE for CSPH
Carrión et al[33]Spain20061240.8400.94-
Vizzutti et al[14]Italy2007610.8100.99 (0.92-0.99)13.6 (PPV 97%)
Lemoine et al[34]France2008920.5300.84 (0.80-0.88)34.9 for alcohol (PPV 98%)
20.5 for HCV (PPV 88%)
Bureau et al[35]France20081500.8580.945 (0.904-0.987)21 (PPV 92%)
Sánchez-Conde et al[36]Spain2011380.4600.80 (0.64-0.97)14 (PPV 84%)
Reiberger et al[13]Austria20125020.7940.817 (0.752-0.891)18 (PPV 86%)
Llop et al[20]Spain2012970.5520.840 (0.748-0.933)21 (PPV 81%)
Berzigotti et al[37]Spain2013117-0.883 (0.824-0.943)21.1 (sensitivity 65%)
Hong et al[38]South Korea2013590.4960.85121.95 (PPV 87%)
Salzl et al[28]Austria2014880.7650.8716.8 (sensitivity 90%)
Augustin et al[39]Spain201440--25
Zykus et al[31]Lithuania20151070.7500.94917.4 (accuracy 89%)
Procopet et al[32]Europe2015202-0.94 (0.89-0.99)21.1 (accuracy 90%)
(Multicentric)
Kitson et al[40]Australia2015950.3800.90 (0.83-0.97)29.0 (PPV 100%)
Elkrief et al[25]France201579-0.78 (0.58-0.98)65.3 (PPV 100%)
Schwabl et al[41]Austria20152260.836 and 0.8460.957 & 0.96216.1 (accuracy 89% & 90%)
Hametner et al[29]Austria2016236-0.92 (0.86-0.96)24.8 (PPV 98%)
This studyIndia20163260.3610.740 (0.662-0.818)21.6 (PPV 93%)
Total2515Weighted mean: 21.8