- Email: [email protected]
Non-invasive methods can predict oesophageal varices in patients with biliary atresia after a Kasai procedure
Digestive and Liver Disease 43 (2011) 659–663
Contents lists available at ScienceDirect
Digestive and Liver Disease journal homepage: www.elsevier.com/locate/dld
Liver, Pancreas and Biliary Tract
Non-invasive methods can predict oesophageal varices in patients with biliary atresia after a Kasai procedure Antonio Colecchia a,∗ , Anna Rita Di Biase c , Eleonora Scaioli a , Barbara Predieri c , Lorenzo Iughetti c , Maria Letizia Bacchi Reggiani b , Lucia Montrone a , Pier Luca Ceccarelli d , Amanda Vestito a , Laura Viola e , Paolo Paolucci c , Davide Festi a a
Department of Clinical Medicine, University of Bologna, Italy Department of Cardiology, University of Bologna, Italy c Department of Pediatrics, University of Modena, Italy d Department of Pediatric Surgery, University of Modena, Italy e Department of Pediatrics, Hospital of Rimini, Italy b
a r t i c l e
i n f o
Article history: Received 29 October 2010 Accepted 10 April 2011 Available online 19 May 2011 Keywords: Cholestasis Liver fibrosis Liver stiffness Portal hypertension
a b s t r a c t Background: After a Kasai procedure, 70% of patients with biliary atresia develop chronic liver disease with portal hypertension and oesophageal varices. Aims: To investigate the role of new non-invasive parameters in predicting the presence of varices in patients with biliary atresia after a Kasai procedure and to identify the cut-off values of these parameters in predicting the presence of varices. Methods: 31 patients with biliary atresia who had undergone a Kasai portoenterostomy were studied. Clinical, biochemical and abdominal ultrasound examination, liver stiffness measurement (LSM), LSMspleen diameter to platelet ratio score (LSPS) and upper digestive endoscopy were performed. Results: 15 (47%) patients had oesophageal varices (Group A) and 16 had no varices (Group B). Median values of LSM (kPa) and LSPS were significantly higher in Group A than in Group B (LSM: 17.0 vs. 7.5, respectively; p = 0.0001; LSPS: 19.62 vs. 2.94, respectively; p = 0.0001). The optimal cut-offs for predicting oesophageal varices were: LSM > 10.6 kPa (sensitivity: 87%, specificity: 87.5%, PPV: 87%, NPV: 87.5%, and AUC: 0.92) and LSPS ≥9.2 (sensitivity: 91%, specificity: 92%, PPV: 91%, NPV: 92%, and AUC: 0.96). Conclusions: Non-invasive methods can predict the presence of oesophageal varices in patients with biliary atresia; the sequential use of two non-invasive methods improves accuracy. © 2011 Editrice Gastroenterologica Italiana S.r.l. Published by Elsevier Ltd. All rights reserved.
1. Introduction If untreated, biliary atresia (BA), the most common cause of chronic cholestasis in infants, leads to secondary biliary cirrhosis in the first months of life and to death in early childhood [1]. Therapeutic strategy for BA has to be carried out early and sequentially, i.e. within the first three months of life: a Kasai portoenterostomy and, subsequently, in the case of liver failure, orthotopic liver transplantation (OLT) [1]. The survival rate with a native liver after a successful Kasai procedure is >50% at 10 years and 30–40% at 20 years [2,3] whilst OLT has a 10-year survival rate of 86% [1]. Although the Kasai procedure is often successful, 67% of patients will develop chronic liver disease [4] with consequent portal hyper-
∗ Corresponding author at: Dipartimento di Medicina Clinica, Policlinico S. OrsolaMalpighi, Via Massarenti 9, 40138 Bologna, Italy. Tel.: +39 0516364123; fax: +39 0516364123. E-mail address: [email protected] (A. Colecchia).
tension, and 50% of all survivors have oesophageal varices 5 years after the surgical procedure [5], with an increased risk of bleeding. Therefore, early screening for varices is required in order to prevent this complication. In paediatric patients with liver disease, no agreed timing for endoscopic screening exists according to the Baveno IV consensus conference [6]; furthermore, endoscopy has to be performed under general anaesthesia and has a notable complication rate (about 2.5%), related to both the endoscopic procedure and the anaesthesia [7]. Therefore, for this reason, only patients at high risk of having oesophageal varices should undergo endoscopy. Biochemical, clinical and ultrasound parameters, as well as new technology such as liver stiffness measurement (LSM) and video capsule endoscopy, have been used in adults with good results, although none of them have equalled endoscopy in terms of overall accuracy [8]. Recently, two papers dealing with the paediatric population in South America [9,10] and another in Korea [11] have assessed the role of non-invasive parameters in predicting oesophageal varices. In the first two, only physical examination and
1590-8658/$36.00 © 2011 Editrice Gastroenterologica Italiana S.r.l. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.dld.2011.04.006
660
A. Colecchia et al. / Digestive and Liver Disease 43 (2011) 659–663
standard biochemical parameters were used in patients with portal hypertension syndrome, whilst, in the Korean study, only the role of transient elastography was assessed in patients with BA. However, no study using multiple non-invasive parameters has been carried out up to now in western paediatric populations. Furthermore, a new non-invasive method (LSM-spleen diameter to platelet ratio score, LSPS) for the detection of oesophageal varices has recently been developed and validated, but for adults with B-viral liver cirrhosis [12]. Therefore, the aims of the present study were: (a) to investigate the role of new non-invasive parameters in predicting the presence of varices in patients with BA after a Kasai procedure and (b) to identify the cut-off values of these parameters in predicting the presence of varices. 2. Methods 2.1. Patients and study design In the period from 1990 to 2009, 52 patients with BA who had undergone a Kasai portoenterostomy were observed at our units. Nineteen (36%) underwent OLT for end stage liver disease whilst 33 (64%), who still had their native livers, were enrolled in the present study. The study was approved by the Institutional Ethics Committees of both hospitals and written informed consent was obtained from all parents or guardians. A two-day study protocol was used for each patient: on the first day, clinical and biochemical evaluations, abdominal ultrasound and liver stiffness measurements were carried out and, on the second day, an upper digestive endoscopy was carried out. The following biochemical tests were performed: serum aspartate aminotransferase (AST), alanine aminotransferase (ALT), ␥-glutamyl transpeptidase, total bilirubin, platelet count, prothrombin time, albumin; height and weight were also recorded. Using some of these parameters, different non-invasive scores for portal hypertension were later calculated. All endoscopy (AC), ultrasound (DF) and LSM (LM) operators were expert in their respective fields, they were independent of one another and were blinded to the other instrumental, clinical and laboratory data of the patients. 2.2. Upper endoscopic examination A standard endoscopic examination was performed and each endoscopy was performed by the same operator (AC). The endoscopic findings were recorded and graded as follows: grade I when the varices were flattened by insufflation; grade II when they were non-confluent and protruding in the lumen despite insufflation; grade III when the varices were not flattened by insufflation and were confluent. The presence of red signs was also recorded in all patients. According to the criteria proposed at the Baveno I Consensus Conference, patients were considered as having large EV when the grade was ≥II [13]. Endoscopy was performed in the operating room under general anaesthesia only in patients under fourteen years of age. 2.3. Liver stiffness measurement (LSM) Liver stiffness measurement (LSM) was assessed using a FibroScan® (Echosens, Paris, France) after at least 6 h of fasting and after an abdominal ultrasound examination which was also used to guide the probe of the FibroScan® in the most difficult cases. LSM was performed according to the technique proposed and validated by Ziol [14]. Briefly, the probe was placed between the rib bones in proximity of the right lobe of the liver with the child
in a supine position with maximal abduction of the right arm. For each patient, 10 successful measurements were obtained and the median value of these measurements of liver stiffness was calculated automatically by software on a microcomputer installed in the FibroScan® . LSM values were accepted if the success rate was >60% and the interquartile range (IQR) was <30% of the median value; the results were expressed in kiloPascals (kPa). Exclusion criteria for LSM were the presence of ascites and/or patients having an intercostal space too narrow to correctly measure liver stiffness. 2.4. Non-invasive scores of oesophageal varices The following scores were evaluated: a. Platelet count to spleen diameter ratio (Plt/Spl) which was first calculated and validated by Giannini et al. [15]. b. LSM-spleen diameter to platelet ratio score (LSPS) which was calculated according to the formula (LSM × spleen diameter/platelet count) proposed and validated by Kim et al. [12]. c. APRI test (AST to platelet count ratio) which was calculated according to Wai et al. [16]. The maximum spleen bipolar diameter was assessed by ultrasound, measuring the longitudinal diameter of the spleen, expressed in mm and normalized by age using the T-score formula. 2.5. Statistical analyses The results were expressed as median values, and differences between the continuous and categorical variables were examined statistically using the Mann–Whitney test. For the normalization of the spleen diameter, the following formula was used to calculate the T-score: T = 50 + 10(Xi − X)/SD, where Xi is the measured value, X the mean value of subjects matched for both sex and age, and SD is the standard deviation [17]. Receiver operating characteristic (ROC) curves were calculated in order to identify the best cut-off value for each of the diagnostic parameters evaluated, and the area under each ROC curve (AUROC) was computed. Univariate and multivariate logistic regression analyses were carried out to assess the independent predictors of the presence of oesophageal varices. p values less than 0.05 were considered statistically significant. 3. Results 3.1. Patients Of the 33 patients with BA who had native livers and were enrolled in this study, 2 (6%) (3 and 4 years of age) were excluded because their intercostal spaces were too narrow to perform LSM with the FibroScan® . The characteristics of the other 31 patients are summarized in Table 1. Twenty-four (77.4%) were classified as Child Pugh A and 7 (22.6%) as Child Pugh B; all patients without varices, except one (93.7%), were in class A according to the Child-Pugh classification. The mean age at which a Kasai procedure had been performed was 56 ± 12 days. Oesophageal varices were present in 15 patients (48%), of whom 9 (60%) had grade I, 5 (33.3%) had grade II and 1 (6.7%) had grade III; furthermore two of them also had gastric varices. Age, weight, height, ALT, AST, bilirubin, platelet count, prothrombin time, serum albumin, spleen diameter and the ChildPugh classification were significantly different between patients with varices (Group A) and without varices (Group B) (p < 0.05).
A. Colecchia et al. / Digestive and Liver Disease 43 (2011) 659–663
661
Table 1 Baseline characteristics of the study population. Group A, patients with oesophageal varices; Group B, patients without oesophageal varices.
Median age (years) Males (%) Median weight (kg) Median height (cm) Median ALT (IU/l) Median AST (IU/l) Median total bilirubin (mg/dl) Median ␥-GT (IU/l) Platelet count (109 /l) Prothrombin time (%) Serum albumin (g/dl) Spleen diameter (mm)
Total (n = 31)
Group A (n = 15)
Group B (n = 16)
p
12 (4–25) 13 (43) 48.5 (14–80) 140 (90–195) 40 (16–124) 40 (15–209) 0.89 (0.18–3.08) 29 (8–191) 137 (47–290) 90 (60–100) 3.9 (2.7–4.54) 140 (75–220)
13 (8–25) 5 (36) 48 (25–80) 152 (122–172) 52 (16–124) 75 (15–209) 1.40 (0.3–3.08) 32 (16–124) 92 (47–290) 86 (60–100) 3.7 (2.7–4.08) 171 (110–220)
8 (4–24) 8 (50) 29.5 (14–80) 122 (90–195) 34 (22–59) 35 (15–51) 0.50 (0.18–1.95) 24 (8–191) 199 (106–237) 94 (69–100) 4 (3.2–4.54) 100 (75–156)
0.007 ns 0.01 0.003 0.009 0.01 0.005 ns 0.002 0.01 0.01 0.0003
Results are given as median and range values.
Table 2 Median values of non-invasive testing performed in the study population. Group A, patients with oesophageal varices; Group B, patients without oesophageal varices.
LSM (kPa) APRI test Plt /Spl (T-score) LSPS (T-score)
Total (n = 31)
Group A (n = 15)
Group B (n = 16)
p
10.3 (5.65–48) 0.96 (0.34–1.32) 1.95 (0.33–4.63) 6.74 (1.46–130)
17 (7.6–48) 3.18 (0.34–15.2) 0.96 (0.33–2.74) 19.62 (5.63–130)
7.5 (5.65–15.1) 0.73 (0.4–1.32) 3.05 (1.18–4.63) 2.94 (1.46–10.58)
0.0001 0.003 0.0001 0.0001
Results are given as median and range values; LSM, liver stiffness measurement; APRI, AST to platelet count ratio; Plt/Spl, platelet count to spleen diameter ratio; LSPS, LSM-spleen diameter to platelet ratio score.
Table 3 Cut-off of the best non-invasive methods for predicting oesophageal varices. Variable
Cut off
Accuracy
Sensitivity
Specificity
PPV
NPV
+LR
−LR
AUC
APRI LSM Plt/Spl LSPS
>0.96 >10.6 ≤1.06 ≥9.2
83 87 87 92
86 87 72.7 91
81 87.5 93 92
87.5 87 91 91
87 87.5 80 92
4.62 6.9 2.28 12.8
0.16 0.15 0.00 0.21
0.88 0.92 0.90 0.96
APRI, AST to platelet count ratio; LSM, liver stiffness measurement; Plt/Spl, platelet count to spleen diameter ratio; LSPS, LSM-spleen diameter to platelet ratio score; PPV, positive predictive value; NPV, negative predictive value; +LR, positive likelihood ratio; −LR, negative likelihood ration; AUC, area under curve.
3.2. Non-invasive methods for predicting oesophageal varices Liver stiffness measurement was successfully carried out in all 31 patients, even in two very young patients (each 4 years of age); these two patients had severe hepatomegaly and large intercostal spaces, the latter making it possible to successfully carry out LSM. The mean number of valid measurements per patient was 9.2 ± 0.5 (range 6–10) and the success rate was 87 ± 18.1% whilst the median IQR was 1.3. Table 2 shows the results of the non-invasive methods used to predict the presence of oesophageal varices. The median values of the non invasive methods were significantly greater in patients with esophageal varices (Group A) than in those without esophageal varices (Group B).
3.3. Cut-off values of the non-invasive methods for the prediction of oesophageal varices The optimal cut-off values for predicting the presence of varices, obtained by ROC curve analysis, are shown in Table 3. For LSM, a cut-off >10.6 kPa had a total accuracy of 87% with a sensitivity of 87% and a specificity of 87.5%; PPV was 87, NPV 87.5 and AUC 0.92. For the APRI, a cut-off >0.96 had a total accuracy of 83% with a sensitivity of 86% and a specificity of 81%; PPV was 87.5%, NPV 87% and AUC 0.88. For the Plt/Spl ratio (T score), a cut-off ≤1.06 had a total accuracy of 87% with a sensitivity of 72.7% and a specificity of 93%; PPV was 91, NPV 80 and AUC 0.90 and, finally, for the LSPS (T score), a cut-off ≥9.2 had a total accuracy of 92% with a sensitivity
of 91% and a specificity of 92%; PPV was 91, NPV 92 and AUC 0.96. Other parameters (platelet count, prothrombin time and spleen diameter) had AUCs < 0.80. 3.4. Associated factors for predicting oesophageal varices At univariate analysis, age, serum AST, ALT, PT, albumin, bilirubin levels, spleen diameter, APRI, LSM, LSPS and PLT/spleen were associated factors for oesophageal varices (Table 4). At multivariate analysis, only the LSPS was an associated factor (OR = 26; CI 95% = 1.11–604.49; p = 0.042). Table 4 Predictors of oesophageal varices at logistic regression analysis.
Age (years) AST (U/l) ALT (U/l) PT (%) Albumin (g/dl) Spleen diameter (cm) APRI test LSM (kPa) LSPS PLT/spleen
Univariate analysis OR [95% CI]
p
1.15 [0.98–1.35] 1.05 [1.00–1.11] 1.06 [1.00–1.11] 0.90 [0.82–0.99] 0.061 [0.005–0.70] 1.06 [1.02–1.09] 13.7 [1.07–175.98] 1.58 [1.14–2.19] 1.69 [1.11–2.59] 0.17 [0.05–0.54]
0.007 0.009 0.014 0.013 0.013 0.000 0.000 0.000 0.000 0.000
AST, aspartate aminotransferases; PLT, platelet; LSM, liver stiffness measurement; kPa, kiloPascal; SSM, spleen stiffness measurement; LSPS, LSM-spleen diameter to platelet ratio score; PLT/spleen, platelet count to spleen diameter ratio; CI, confidence interval.
662
A. Colecchia et al. / Digestive and Liver Disease 43 (2011) 659–663
4. Discussion This is the first study carried out on a western population of patients having BA who had undergone a Kasai procedure which investigated the usefulness of new, non-invasive indices to detect the presence of oesophageal varices. We found that the liver stiffness measurement, the platelet to spleen ratio and the ratio between liver stiffness measurement and the spleen diameter to platelet ratio accurately predicted the presence of oesophageal varices. At multivariate analysis, only the LSPS was predictive of oesophageal varices. Nevertheless, after a Kasai procedure, 70% of patients develop chronic liver disease with early portal hypertension [4], and 50% of all survivors show oesophago-gastric varices 5 years after the surgical procedure [5]. There is no agreement regarding the proper time at which endoscopic screening of oesophageal varices should take place [6]; furthermore, serious complications have been reported in about 2.5% of the paediatric patients who undergo upper endoscopy [7]. Thus, non-invasive/non-endoscopic tests identifying the presence of varices are needed, mainly in the paediatric population. LSM is a new non-invasive method for predicting liver fibrosis in chronic liver disease and is already being carried out in the adult [18] and paediatric populations [19] with successful results. However, LSM has also been used in the adult population for the prediction of varices with conflicting results [20–22] since it has been suggested [20] that, for a high level of portal pressure (>10/12 mmHg), the correlation between varices and LSM is not linear. In fact, in this condition, the development of portal hypertension becomes independent of the accumulation of fibrillar extracellular matrix which is responsible for the increase in liver tissue stiffness [23,24]. Thus in our study carried out on a particular group of patients (mainly children), we evaluated the accuracy of LSM and of other tests proposed as predictors of varices with the aim of reducing the rate of false positive/negative results involved in any single test. In our study, the LSM correctly identified 27 out of 31 (87%) patients with a sensitivity of 87% and a specificity of 87.5%; these results are similar to those recently reported in a Korean study [11] (90% accuracy, 97% sensitivity and 80% specificity); the LSM cut-off (>10.6 kPa) was also similar in the two studies [11]. However, a comparison between this study and our is difficult since the populations studied are different in terms of mean age (3.8 vs. 11.8 years), prevalence of oesophageal varices (61% vs. 48%) and proposed algorithms. The cut-off values for predicting varices identified by Chang et al. [11] and by us are lower than those found in an adult population [20–22]; this difference could be justified by the particular type of disease. In fact, BA, unlike other chronic liver diseases, is characterized by pre-sinusoidal portal hypertension; as a consequence, varices can be present before liver cirrhosis develops [5,23,24]. In our paper, the LSM results did not seem to be influenced by cholestasis [25] because our population did not have abnormal cholestatic liver enzymes and bilirubin after a Kasai procedure; in addition, ␥-glutamyl transpeptidase was not significantly different in the two groups. As far as the other non-invasive tests are concerned, we carried out not only LSM but also the APRI Test [16], the platelet count to spleen diameter ratio [15] and the LSPS score [12], which are reported to be closely correlated to portal hypertension [12,15,21,26]. The APRI test which we used for the first time in this particular patient population to predict oesophageal varices had a good performance; recently, the APRI test has been used in BA patients but for the assessment of liver fibrosis [27]. The LSPS [12], which was recently validated successfully in a viral cirrhotic population, was used because it integrated both the LSM and the Giannini’s scores into a single ratio. In our study, the Giannini et al. [15] and the LSPS [12] scores gave good results, with an accuracy of 87% and 92%, respectively. These
two scores, used for the first time in our BA patients which were mostly children, had a high level of accuracy, as has been observed in adults [12,15], even if the sensitivity as measured by Giannini was lower (72.7%) than the LSPS (91%). Furthermore, in our study, LSPS was the only predictive factor for the presence of oesophageal varices at multivariate analysis. Two other papers [9,10] have evaluated the presence of oesophageal varices using non-invasive methods with a good level of accuracy; however, in these studies, patients with miscellaneous liver disease/portal vein thrombosis were included. In order to better characterize the clinical usefulness of the observed cut-offs of LSM and LSPS, we hypothesized a diagnostic algorithm for identifying oesophageal varices in BA patients after a Kasai procedure. The clinical purpose of this hypothetical algorithm was to reduce inappropriate upper digestive endoscopies. First, we identified the LSM (≤7.6 kPa) with the best specificity and NPV, and a second cut-off (≥15.4 kPa) with the best sensitivity and PPV; using these values, we correctly diagnosed 19 out of 31 (61%) patients. In the remaining (39%) patients, who fell into an area characterized by a doubtful diagnosis (grey zone), the use of the LSPS enabled us to reach a success rate of 94% as compared to the 87% obtained using LSM only. According to this theoretical exercise, a non-indicated endoscopy would be performed in only two patients without varices; on the contrary, all necessary endoscopies, e.g. in patients with varices, would be performed. Obviously, this hypothesis needs to be validated under these and other clinical conditions. If future studies confirm the clinical usefulness of the proposed algorithm, this result would be particularly important since the correct identification of all varices could reduce the number of patients (in our case 46%) requiring general anaesthesia which is necessary to perform endoscopy in paediatric patients. In conclusion, this study, performed on a rare but clinically important disease, showed that non-invasive methods can successfully predict the presence of oesophageal varices in patients with BA who have undergone a Kasai procedure. Furthermore, the sequential use of two non-invasive methods could improve overall diagnostic accuracy and allow the best detection of oesophageal varices, thus avoiding unnecessary endoscopic examination.
Conflict of interest statement None.
References [1] Hartley JL, Davenport M, Kelly DA. Biliary atresia. Lancet 2009;374:1704–13. [2] Shinkai M, Ohhama Y, Take H, et al. Long-term outcome of children with biliary atresia who were not transplanted after the Kasai operation: >20-year experience at a children’s hospital. J Pediatr Gastroenterol Nutr 2009;48:443–50. [3] Lykavieris P, Chardot C, Sokhn M, et al. Outcome in adulthood of biliary atresia: a study of 63 patients who survived for over 20 years with their native liver. Hepatology 2005;41:366–71. [4] Balistreri WF, Grand R, Hoofnagle JH, et al. Biliary atresia: current concepts and research direction. Summary a symposium. Hepatology 1996;23:1682–92. [5] Mieli-Vergani G, Vergani D. Biliary atresia. Semin Immunopathol 2009;31:371–81. [6] Shneider B, Emre S, Groszmann R, et al. Expert pediatric opinion on the report of the Baveno IV consensus workshop on methodology of diagnosis and therapy in portal hypertension. Pediatr Transplant 2006;10:893–907. [7] Thakkar K, El-Serag HB, Mattek N, et al. Complications of pediatric EGD: a 4-year experience in PEDS-CORI. Gastrointest Endosc 2007;65:213–21. [8] de Franchis R. Non invasive (and minimally invasive) diagnosis of esophageal varices. J Hepatol 2008;49:520–7. [9] Fagundes ED, Ferreira AR, Roquete ML, et al. Clinical and laboratory predictors of esophageal varices in children and adolescents with portal hypertension syndrome. J Pediatr Gastroenterol Nutr 2008;46:178–83. [10] Gana JC, Turner D, Roberts EA, et al. Derivation of a clinical prediction rule for the noninvasive diagnosis of varices in children. J Pediatr Gastroenterol Nutr 2010;50:188–93. [11] Chang HK, Park YJ, Koh H, et al. Hepatic fibrosis scan for liver stiffness score measurement: a useful preendoscopic screening test for the detection of varices
A. Colecchia et al. / Digestive and Liver Disease 43 (2011) 659–663
[12]
[13]
[14]
[15]
[16]
[17] [18] [19]
in postoperative patients with biliary atresia. J Pediatr Gastroenterol Nutr 2009;49:323–8. Kim BK, Han KH, Park JY, et al. A liver stiffness measurement-based noninvasive prediction model for high-risk esophageal varices in B-viral liver cirrhosis. Am J Gastroenterol 2010;105:1382–90. De Franchis R, Pascal JP, Ancona E, et al. Definitions, methodology and therapeutic strategies in portal hypertension. A Consensus Development Workshop Baveno. J Hepatol 1992;15:256–61. Ziol M, Handra-Luca A, Kettaneh A, et al. Noninvasive assessment of liver fibrosis by measurement of stiffness in patients with chronic hepatitis C. Hepatology 2005;41:48–54. 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 esophageal varices in patients with liver cirrhosis. Gut 2003;52:1200–5. 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;38:518–26. Aiken LS, West SG. Multiple regression: testing and interpreting interactions. Newbury Park, CA: Sage; 1991. Stebbing J, Farouk L, Panos G, et al. A meta-analysis of transient elastography for the detection of hepatic fibrosis. J Clin Gastroenterol 2010;44:214–9. de Lédinghen V, Le Bail B, Rebouissoux L, et al. Liver stiffness measurement in children using FibroScan® : feasibility study and comparison with Fibrotest, aspartate transaminase to platelets ratio index, and liver biopsy. J Pediatr Gastroenterol Nutr 2007;45:443–50.
663
[20] Vizzutti F, Arena U, Romanelli RG, et al. Liver stiffness measurement predicts severe portal hypertension in patients with HCV-related cirrhosis. Hepatology 2007;45:1290–7. [21] Castéra L, Le Bail B, Roudot-Thoraval F, et al. Early detection in routine clinical practice of cirrhosis and esophageal varices in chronic hepatitis C: comparison of transient elastography (FibroScan® ) with standard laboratory tests and non invasive scores. J Hepatol 2009;50:59–68. [22] Kazemi F, Kettaneh A, N’kontchou G, et al. Liver stiffness measurement selects patients with cirrhosis at risk of bearing large esophageal varices. J Hepatol 2006;45:230–5. [23] Weinbren K, Hadjis NS, Blumgart LH. Structural aspects of the liver in patients with biliary disease and portal hypertension. J Clin Pathol 1985;38: 1013–20. [24] Pinzani M, Rombouts K. Liver fibrosis: from the bench to clinical targets. Dig Liver Dis 2004;36:231–42. [25] Millonig G, Reimann FM, Friedrich S, et al. Extrahepatic cholestasis increases liver stiffness (FibroScan® ) irrespective of fibrosis. Hepatology 2008;48:1718–23. [26] Bosch J, Garcia-Pagán JC, Berzigotti A, et al. Measurement of portal pressure and its role in the management of chronic liver disease. Semin Liver Dis 2006;26:348–62. [27] Kim SY, Seok JY, Han SJ, et al. Assessment of liver fibrosis and cirrhosis by aspartate aminotransferase-to-platelet ratio index in children with biliary atresia. JPGN 2010;51:198–202.
Contents lists available at ScienceDirect
Digestive and Liver Disease journal homepage: www.elsevier.com/locate/dld
Liver, Pancreas and Biliary Tract
Non-invasive methods can predict oesophageal varices in patients with biliary atresia after a Kasai procedure Antonio Colecchia a,∗ , Anna Rita Di Biase c , Eleonora Scaioli a , Barbara Predieri c , Lorenzo Iughetti c , Maria Letizia Bacchi Reggiani b , Lucia Montrone a , Pier Luca Ceccarelli d , Amanda Vestito a , Laura Viola e , Paolo Paolucci c , Davide Festi a a
Department of Clinical Medicine, University of Bologna, Italy Department of Cardiology, University of Bologna, Italy c Department of Pediatrics, University of Modena, Italy d Department of Pediatric Surgery, University of Modena, Italy e Department of Pediatrics, Hospital of Rimini, Italy b
a r t i c l e
i n f o
Article history: Received 29 October 2010 Accepted 10 April 2011 Available online 19 May 2011 Keywords: Cholestasis Liver fibrosis Liver stiffness Portal hypertension
a b s t r a c t Background: After a Kasai procedure, 70% of patients with biliary atresia develop chronic liver disease with portal hypertension and oesophageal varices. Aims: To investigate the role of new non-invasive parameters in predicting the presence of varices in patients with biliary atresia after a Kasai procedure and to identify the cut-off values of these parameters in predicting the presence of varices. Methods: 31 patients with biliary atresia who had undergone a Kasai portoenterostomy were studied. Clinical, biochemical and abdominal ultrasound examination, liver stiffness measurement (LSM), LSMspleen diameter to platelet ratio score (LSPS) and upper digestive endoscopy were performed. Results: 15 (47%) patients had oesophageal varices (Group A) and 16 had no varices (Group B). Median values of LSM (kPa) and LSPS were significantly higher in Group A than in Group B (LSM: 17.0 vs. 7.5, respectively; p = 0.0001; LSPS: 19.62 vs. 2.94, respectively; p = 0.0001). The optimal cut-offs for predicting oesophageal varices were: LSM > 10.6 kPa (sensitivity: 87%, specificity: 87.5%, PPV: 87%, NPV: 87.5%, and AUC: 0.92) and LSPS ≥9.2 (sensitivity: 91%, specificity: 92%, PPV: 91%, NPV: 92%, and AUC: 0.96). Conclusions: Non-invasive methods can predict the presence of oesophageal varices in patients with biliary atresia; the sequential use of two non-invasive methods improves accuracy. © 2011 Editrice Gastroenterologica Italiana S.r.l. Published by Elsevier Ltd. All rights reserved.
1. Introduction If untreated, biliary atresia (BA), the most common cause of chronic cholestasis in infants, leads to secondary biliary cirrhosis in the first months of life and to death in early childhood [1]. Therapeutic strategy for BA has to be carried out early and sequentially, i.e. within the first three months of life: a Kasai portoenterostomy and, subsequently, in the case of liver failure, orthotopic liver transplantation (OLT) [1]. The survival rate with a native liver after a successful Kasai procedure is >50% at 10 years and 30–40% at 20 years [2,3] whilst OLT has a 10-year survival rate of 86% [1]. Although the Kasai procedure is often successful, 67% of patients will develop chronic liver disease [4] with consequent portal hyper-
∗ Corresponding author at: Dipartimento di Medicina Clinica, Policlinico S. OrsolaMalpighi, Via Massarenti 9, 40138 Bologna, Italy. Tel.: +39 0516364123; fax: +39 0516364123. E-mail address: [email protected] (A. Colecchia).
tension, and 50% of all survivors have oesophageal varices 5 years after the surgical procedure [5], with an increased risk of bleeding. Therefore, early screening for varices is required in order to prevent this complication. In paediatric patients with liver disease, no agreed timing for endoscopic screening exists according to the Baveno IV consensus conference [6]; furthermore, endoscopy has to be performed under general anaesthesia and has a notable complication rate (about 2.5%), related to both the endoscopic procedure and the anaesthesia [7]. Therefore, for this reason, only patients at high risk of having oesophageal varices should undergo endoscopy. Biochemical, clinical and ultrasound parameters, as well as new technology such as liver stiffness measurement (LSM) and video capsule endoscopy, have been used in adults with good results, although none of them have equalled endoscopy in terms of overall accuracy [8]. Recently, two papers dealing with the paediatric population in South America [9,10] and another in Korea [11] have assessed the role of non-invasive parameters in predicting oesophageal varices. In the first two, only physical examination and
1590-8658/$36.00 © 2011 Editrice Gastroenterologica Italiana S.r.l. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.dld.2011.04.006
660
A. Colecchia et al. / Digestive and Liver Disease 43 (2011) 659–663
standard biochemical parameters were used in patients with portal hypertension syndrome, whilst, in the Korean study, only the role of transient elastography was assessed in patients with BA. However, no study using multiple non-invasive parameters has been carried out up to now in western paediatric populations. Furthermore, a new non-invasive method (LSM-spleen diameter to platelet ratio score, LSPS) for the detection of oesophageal varices has recently been developed and validated, but for adults with B-viral liver cirrhosis [12]. Therefore, the aims of the present study were: (a) to investigate the role of new non-invasive parameters in predicting the presence of varices in patients with BA after a Kasai procedure and (b) to identify the cut-off values of these parameters in predicting the presence of varices. 2. Methods 2.1. Patients and study design In the period from 1990 to 2009, 52 patients with BA who had undergone a Kasai portoenterostomy were observed at our units. Nineteen (36%) underwent OLT for end stage liver disease whilst 33 (64%), who still had their native livers, were enrolled in the present study. The study was approved by the Institutional Ethics Committees of both hospitals and written informed consent was obtained from all parents or guardians. A two-day study protocol was used for each patient: on the first day, clinical and biochemical evaluations, abdominal ultrasound and liver stiffness measurements were carried out and, on the second day, an upper digestive endoscopy was carried out. The following biochemical tests were performed: serum aspartate aminotransferase (AST), alanine aminotransferase (ALT), ␥-glutamyl transpeptidase, total bilirubin, platelet count, prothrombin time, albumin; height and weight were also recorded. Using some of these parameters, different non-invasive scores for portal hypertension were later calculated. All endoscopy (AC), ultrasound (DF) and LSM (LM) operators were expert in their respective fields, they were independent of one another and were blinded to the other instrumental, clinical and laboratory data of the patients. 2.2. Upper endoscopic examination A standard endoscopic examination was performed and each endoscopy was performed by the same operator (AC). The endoscopic findings were recorded and graded as follows: grade I when the varices were flattened by insufflation; grade II when they were non-confluent and protruding in the lumen despite insufflation; grade III when the varices were not flattened by insufflation and were confluent. The presence of red signs was also recorded in all patients. According to the criteria proposed at the Baveno I Consensus Conference, patients were considered as having large EV when the grade was ≥II [13]. Endoscopy was performed in the operating room under general anaesthesia only in patients under fourteen years of age. 2.3. Liver stiffness measurement (LSM) Liver stiffness measurement (LSM) was assessed using a FibroScan® (Echosens, Paris, France) after at least 6 h of fasting and after an abdominal ultrasound examination which was also used to guide the probe of the FibroScan® in the most difficult cases. LSM was performed according to the technique proposed and validated by Ziol [14]. Briefly, the probe was placed between the rib bones in proximity of the right lobe of the liver with the child
in a supine position with maximal abduction of the right arm. For each patient, 10 successful measurements were obtained and the median value of these measurements of liver stiffness was calculated automatically by software on a microcomputer installed in the FibroScan® . LSM values were accepted if the success rate was >60% and the interquartile range (IQR) was <30% of the median value; the results were expressed in kiloPascals (kPa). Exclusion criteria for LSM were the presence of ascites and/or patients having an intercostal space too narrow to correctly measure liver stiffness. 2.4. Non-invasive scores of oesophageal varices The following scores were evaluated: a. Platelet count to spleen diameter ratio (Plt/Spl) which was first calculated and validated by Giannini et al. [15]. b. LSM-spleen diameter to platelet ratio score (LSPS) which was calculated according to the formula (LSM × spleen diameter/platelet count) proposed and validated by Kim et al. [12]. c. APRI test (AST to platelet count ratio) which was calculated according to Wai et al. [16]. The maximum spleen bipolar diameter was assessed by ultrasound, measuring the longitudinal diameter of the spleen, expressed in mm and normalized by age using the T-score formula. 2.5. Statistical analyses The results were expressed as median values, and differences between the continuous and categorical variables were examined statistically using the Mann–Whitney test. For the normalization of the spleen diameter, the following formula was used to calculate the T-score: T = 50 + 10(Xi − X)/SD, where Xi is the measured value, X the mean value of subjects matched for both sex and age, and SD is the standard deviation [17]. Receiver operating characteristic (ROC) curves were calculated in order to identify the best cut-off value for each of the diagnostic parameters evaluated, and the area under each ROC curve (AUROC) was computed. Univariate and multivariate logistic regression analyses were carried out to assess the independent predictors of the presence of oesophageal varices. p values less than 0.05 were considered statistically significant. 3. Results 3.1. Patients Of the 33 patients with BA who had native livers and were enrolled in this study, 2 (6%) (3 and 4 years of age) were excluded because their intercostal spaces were too narrow to perform LSM with the FibroScan® . The characteristics of the other 31 patients are summarized in Table 1. Twenty-four (77.4%) were classified as Child Pugh A and 7 (22.6%) as Child Pugh B; all patients without varices, except one (93.7%), were in class A according to the Child-Pugh classification. The mean age at which a Kasai procedure had been performed was 56 ± 12 days. Oesophageal varices were present in 15 patients (48%), of whom 9 (60%) had grade I, 5 (33.3%) had grade II and 1 (6.7%) had grade III; furthermore two of them also had gastric varices. Age, weight, height, ALT, AST, bilirubin, platelet count, prothrombin time, serum albumin, spleen diameter and the ChildPugh classification were significantly different between patients with varices (Group A) and without varices (Group B) (p < 0.05).
A. Colecchia et al. / Digestive and Liver Disease 43 (2011) 659–663
661
Table 1 Baseline characteristics of the study population. Group A, patients with oesophageal varices; Group B, patients without oesophageal varices.
Median age (years) Males (%) Median weight (kg) Median height (cm) Median ALT (IU/l) Median AST (IU/l) Median total bilirubin (mg/dl) Median ␥-GT (IU/l) Platelet count (109 /l) Prothrombin time (%) Serum albumin (g/dl) Spleen diameter (mm)
Total (n = 31)
Group A (n = 15)
Group B (n = 16)
p
12 (4–25) 13 (43) 48.5 (14–80) 140 (90–195) 40 (16–124) 40 (15–209) 0.89 (0.18–3.08) 29 (8–191) 137 (47–290) 90 (60–100) 3.9 (2.7–4.54) 140 (75–220)
13 (8–25) 5 (36) 48 (25–80) 152 (122–172) 52 (16–124) 75 (15–209) 1.40 (0.3–3.08) 32 (16–124) 92 (47–290) 86 (60–100) 3.7 (2.7–4.08) 171 (110–220)
8 (4–24) 8 (50) 29.5 (14–80) 122 (90–195) 34 (22–59) 35 (15–51) 0.50 (0.18–1.95) 24 (8–191) 199 (106–237) 94 (69–100) 4 (3.2–4.54) 100 (75–156)
0.007 ns 0.01 0.003 0.009 0.01 0.005 ns 0.002 0.01 0.01 0.0003
Results are given as median and range values.
Table 2 Median values of non-invasive testing performed in the study population. Group A, patients with oesophageal varices; Group B, patients without oesophageal varices.
LSM (kPa) APRI test Plt /Spl (T-score) LSPS (T-score)
Total (n = 31)
Group A (n = 15)
Group B (n = 16)
p
10.3 (5.65–48) 0.96 (0.34–1.32) 1.95 (0.33–4.63) 6.74 (1.46–130)
17 (7.6–48) 3.18 (0.34–15.2) 0.96 (0.33–2.74) 19.62 (5.63–130)
7.5 (5.65–15.1) 0.73 (0.4–1.32) 3.05 (1.18–4.63) 2.94 (1.46–10.58)
0.0001 0.003 0.0001 0.0001
Results are given as median and range values; LSM, liver stiffness measurement; APRI, AST to platelet count ratio; Plt/Spl, platelet count to spleen diameter ratio; LSPS, LSM-spleen diameter to platelet ratio score.
Table 3 Cut-off of the best non-invasive methods for predicting oesophageal varices. Variable
Cut off
Accuracy
Sensitivity
Specificity
PPV
NPV
+LR
−LR
AUC
APRI LSM Plt/Spl LSPS
>0.96 >10.6 ≤1.06 ≥9.2
83 87 87 92
86 87 72.7 91
81 87.5 93 92
87.5 87 91 91
87 87.5 80 92
4.62 6.9 2.28 12.8
0.16 0.15 0.00 0.21
0.88 0.92 0.90 0.96
APRI, AST to platelet count ratio; LSM, liver stiffness measurement; Plt/Spl, platelet count to spleen diameter ratio; LSPS, LSM-spleen diameter to platelet ratio score; PPV, positive predictive value; NPV, negative predictive value; +LR, positive likelihood ratio; −LR, negative likelihood ration; AUC, area under curve.
3.2. Non-invasive methods for predicting oesophageal varices Liver stiffness measurement was successfully carried out in all 31 patients, even in two very young patients (each 4 years of age); these two patients had severe hepatomegaly and large intercostal spaces, the latter making it possible to successfully carry out LSM. The mean number of valid measurements per patient was 9.2 ± 0.5 (range 6–10) and the success rate was 87 ± 18.1% whilst the median IQR was 1.3. Table 2 shows the results of the non-invasive methods used to predict the presence of oesophageal varices. The median values of the non invasive methods were significantly greater in patients with esophageal varices (Group A) than in those without esophageal varices (Group B).
3.3. Cut-off values of the non-invasive methods for the prediction of oesophageal varices The optimal cut-off values for predicting the presence of varices, obtained by ROC curve analysis, are shown in Table 3. For LSM, a cut-off >10.6 kPa had a total accuracy of 87% with a sensitivity of 87% and a specificity of 87.5%; PPV was 87, NPV 87.5 and AUC 0.92. For the APRI, a cut-off >0.96 had a total accuracy of 83% with a sensitivity of 86% and a specificity of 81%; PPV was 87.5%, NPV 87% and AUC 0.88. For the Plt/Spl ratio (T score), a cut-off ≤1.06 had a total accuracy of 87% with a sensitivity of 72.7% and a specificity of 93%; PPV was 91, NPV 80 and AUC 0.90 and, finally, for the LSPS (T score), a cut-off ≥9.2 had a total accuracy of 92% with a sensitivity
of 91% and a specificity of 92%; PPV was 91, NPV 92 and AUC 0.96. Other parameters (platelet count, prothrombin time and spleen diameter) had AUCs < 0.80. 3.4. Associated factors for predicting oesophageal varices At univariate analysis, age, serum AST, ALT, PT, albumin, bilirubin levels, spleen diameter, APRI, LSM, LSPS and PLT/spleen were associated factors for oesophageal varices (Table 4). At multivariate analysis, only the LSPS was an associated factor (OR = 26; CI 95% = 1.11–604.49; p = 0.042). Table 4 Predictors of oesophageal varices at logistic regression analysis.
Age (years) AST (U/l) ALT (U/l) PT (%) Albumin (g/dl) Spleen diameter (cm) APRI test LSM (kPa) LSPS PLT/spleen
Univariate analysis OR [95% CI]
p
1.15 [0.98–1.35] 1.05 [1.00–1.11] 1.06 [1.00–1.11] 0.90 [0.82–0.99] 0.061 [0.005–0.70] 1.06 [1.02–1.09] 13.7 [1.07–175.98] 1.58 [1.14–2.19] 1.69 [1.11–2.59] 0.17 [0.05–0.54]
0.007 0.009 0.014 0.013 0.013 0.000 0.000 0.000 0.000 0.000
AST, aspartate aminotransferases; PLT, platelet; LSM, liver stiffness measurement; kPa, kiloPascal; SSM, spleen stiffness measurement; LSPS, LSM-spleen diameter to platelet ratio score; PLT/spleen, platelet count to spleen diameter ratio; CI, confidence interval.
662
A. Colecchia et al. / Digestive and Liver Disease 43 (2011) 659–663
4. Discussion This is the first study carried out on a western population of patients having BA who had undergone a Kasai procedure which investigated the usefulness of new, non-invasive indices to detect the presence of oesophageal varices. We found that the liver stiffness measurement, the platelet to spleen ratio and the ratio between liver stiffness measurement and the spleen diameter to platelet ratio accurately predicted the presence of oesophageal varices. At multivariate analysis, only the LSPS was predictive of oesophageal varices. Nevertheless, after a Kasai procedure, 70% of patients develop chronic liver disease with early portal hypertension [4], and 50% of all survivors show oesophago-gastric varices 5 years after the surgical procedure [5]. There is no agreement regarding the proper time at which endoscopic screening of oesophageal varices should take place [6]; furthermore, serious complications have been reported in about 2.5% of the paediatric patients who undergo upper endoscopy [7]. Thus, non-invasive/non-endoscopic tests identifying the presence of varices are needed, mainly in the paediatric population. LSM is a new non-invasive method for predicting liver fibrosis in chronic liver disease and is already being carried out in the adult [18] and paediatric populations [19] with successful results. However, LSM has also been used in the adult population for the prediction of varices with conflicting results [20–22] since it has been suggested [20] that, for a high level of portal pressure (>10/12 mmHg), the correlation between varices and LSM is not linear. In fact, in this condition, the development of portal hypertension becomes independent of the accumulation of fibrillar extracellular matrix which is responsible for the increase in liver tissue stiffness [23,24]. Thus in our study carried out on a particular group of patients (mainly children), we evaluated the accuracy of LSM and of other tests proposed as predictors of varices with the aim of reducing the rate of false positive/negative results involved in any single test. In our study, the LSM correctly identified 27 out of 31 (87%) patients with a sensitivity of 87% and a specificity of 87.5%; these results are similar to those recently reported in a Korean study [11] (90% accuracy, 97% sensitivity and 80% specificity); the LSM cut-off (>10.6 kPa) was also similar in the two studies [11]. However, a comparison between this study and our is difficult since the populations studied are different in terms of mean age (3.8 vs. 11.8 years), prevalence of oesophageal varices (61% vs. 48%) and proposed algorithms. The cut-off values for predicting varices identified by Chang et al. [11] and by us are lower than those found in an adult population [20–22]; this difference could be justified by the particular type of disease. In fact, BA, unlike other chronic liver diseases, is characterized by pre-sinusoidal portal hypertension; as a consequence, varices can be present before liver cirrhosis develops [5,23,24]. In our paper, the LSM results did not seem to be influenced by cholestasis [25] because our population did not have abnormal cholestatic liver enzymes and bilirubin after a Kasai procedure; in addition, ␥-glutamyl transpeptidase was not significantly different in the two groups. As far as the other non-invasive tests are concerned, we carried out not only LSM but also the APRI Test [16], the platelet count to spleen diameter ratio [15] and the LSPS score [12], which are reported to be closely correlated to portal hypertension [12,15,21,26]. The APRI test which we used for the first time in this particular patient population to predict oesophageal varices had a good performance; recently, the APRI test has been used in BA patients but for the assessment of liver fibrosis [27]. The LSPS [12], which was recently validated successfully in a viral cirrhotic population, was used because it integrated both the LSM and the Giannini’s scores into a single ratio. In our study, the Giannini et al. [15] and the LSPS [12] scores gave good results, with an accuracy of 87% and 92%, respectively. These
two scores, used for the first time in our BA patients which were mostly children, had a high level of accuracy, as has been observed in adults [12,15], even if the sensitivity as measured by Giannini was lower (72.7%) than the LSPS (91%). Furthermore, in our study, LSPS was the only predictive factor for the presence of oesophageal varices at multivariate analysis. Two other papers [9,10] have evaluated the presence of oesophageal varices using non-invasive methods with a good level of accuracy; however, in these studies, patients with miscellaneous liver disease/portal vein thrombosis were included. In order to better characterize the clinical usefulness of the observed cut-offs of LSM and LSPS, we hypothesized a diagnostic algorithm for identifying oesophageal varices in BA patients after a Kasai procedure. The clinical purpose of this hypothetical algorithm was to reduce inappropriate upper digestive endoscopies. First, we identified the LSM (≤7.6 kPa) with the best specificity and NPV, and a second cut-off (≥15.4 kPa) with the best sensitivity and PPV; using these values, we correctly diagnosed 19 out of 31 (61%) patients. In the remaining (39%) patients, who fell into an area characterized by a doubtful diagnosis (grey zone), the use of the LSPS enabled us to reach a success rate of 94% as compared to the 87% obtained using LSM only. According to this theoretical exercise, a non-indicated endoscopy would be performed in only two patients without varices; on the contrary, all necessary endoscopies, e.g. in patients with varices, would be performed. Obviously, this hypothesis needs to be validated under these and other clinical conditions. If future studies confirm the clinical usefulness of the proposed algorithm, this result would be particularly important since the correct identification of all varices could reduce the number of patients (in our case 46%) requiring general anaesthesia which is necessary to perform endoscopy in paediatric patients. In conclusion, this study, performed on a rare but clinically important disease, showed that non-invasive methods can successfully predict the presence of oesophageal varices in patients with BA who have undergone a Kasai procedure. Furthermore, the sequential use of two non-invasive methods could improve overall diagnostic accuracy and allow the best detection of oesophageal varices, thus avoiding unnecessary endoscopic examination.
Conflict of interest statement None.
References [1] Hartley JL, Davenport M, Kelly DA. Biliary atresia. Lancet 2009;374:1704–13. [2] Shinkai M, Ohhama Y, Take H, et al. Long-term outcome of children with biliary atresia who were not transplanted after the Kasai operation: >20-year experience at a children’s hospital. J Pediatr Gastroenterol Nutr 2009;48:443–50. [3] Lykavieris P, Chardot C, Sokhn M, et al. Outcome in adulthood of biliary atresia: a study of 63 patients who survived for over 20 years with their native liver. Hepatology 2005;41:366–71. [4] Balistreri WF, Grand R, Hoofnagle JH, et al. Biliary atresia: current concepts and research direction. Summary a symposium. Hepatology 1996;23:1682–92. [5] Mieli-Vergani G, Vergani D. Biliary atresia. Semin Immunopathol 2009;31:371–81. [6] Shneider B, Emre S, Groszmann R, et al. Expert pediatric opinion on the report of the Baveno IV consensus workshop on methodology of diagnosis and therapy in portal hypertension. Pediatr Transplant 2006;10:893–907. [7] Thakkar K, El-Serag HB, Mattek N, et al. Complications of pediatric EGD: a 4-year experience in PEDS-CORI. Gastrointest Endosc 2007;65:213–21. [8] de Franchis R. Non invasive (and minimally invasive) diagnosis of esophageal varices. J Hepatol 2008;49:520–7. [9] Fagundes ED, Ferreira AR, Roquete ML, et al. Clinical and laboratory predictors of esophageal varices in children and adolescents with portal hypertension syndrome. J Pediatr Gastroenterol Nutr 2008;46:178–83. [10] Gana JC, Turner D, Roberts EA, et al. Derivation of a clinical prediction rule for the noninvasive diagnosis of varices in children. J Pediatr Gastroenterol Nutr 2010;50:188–93. [11] Chang HK, Park YJ, Koh H, et al. Hepatic fibrosis scan for liver stiffness score measurement: a useful preendoscopic screening test for the detection of varices
A. Colecchia et al. / Digestive and Liver Disease 43 (2011) 659–663
[12]
[13]
[14]
[15]
[16]
[17] [18] [19]
in postoperative patients with biliary atresia. J Pediatr Gastroenterol Nutr 2009;49:323–8. Kim BK, Han KH, Park JY, et al. A liver stiffness measurement-based noninvasive prediction model for high-risk esophageal varices in B-viral liver cirrhosis. Am J Gastroenterol 2010;105:1382–90. De Franchis R, Pascal JP, Ancona E, et al. Definitions, methodology and therapeutic strategies in portal hypertension. A Consensus Development Workshop Baveno. J Hepatol 1992;15:256–61. Ziol M, Handra-Luca A, Kettaneh A, et al. Noninvasive assessment of liver fibrosis by measurement of stiffness in patients with chronic hepatitis C. Hepatology 2005;41:48–54. 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 esophageal varices in patients with liver cirrhosis. Gut 2003;52:1200–5. 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;38:518–26. Aiken LS, West SG. Multiple regression: testing and interpreting interactions. Newbury Park, CA: Sage; 1991. Stebbing J, Farouk L, Panos G, et al. A meta-analysis of transient elastography for the detection of hepatic fibrosis. J Clin Gastroenterol 2010;44:214–9. de Lédinghen V, Le Bail B, Rebouissoux L, et al. Liver stiffness measurement in children using FibroScan® : feasibility study and comparison with Fibrotest, aspartate transaminase to platelets ratio index, and liver biopsy. J Pediatr Gastroenterol Nutr 2007;45:443–50.
663
[20] Vizzutti F, Arena U, Romanelli RG, et al. Liver stiffness measurement predicts severe portal hypertension in patients with HCV-related cirrhosis. Hepatology 2007;45:1290–7. [21] Castéra L, Le Bail B, Roudot-Thoraval F, et al. Early detection in routine clinical practice of cirrhosis and esophageal varices in chronic hepatitis C: comparison of transient elastography (FibroScan® ) with standard laboratory tests and non invasive scores. J Hepatol 2009;50:59–68. [22] Kazemi F, Kettaneh A, N’kontchou G, et al. Liver stiffness measurement selects patients with cirrhosis at risk of bearing large esophageal varices. J Hepatol 2006;45:230–5. [23] Weinbren K, Hadjis NS, Blumgart LH. Structural aspects of the liver in patients with biliary disease and portal hypertension. J Clin Pathol 1985;38: 1013–20. [24] Pinzani M, Rombouts K. Liver fibrosis: from the bench to clinical targets. Dig Liver Dis 2004;36:231–42. [25] Millonig G, Reimann FM, Friedrich S, et al. Extrahepatic cholestasis increases liver stiffness (FibroScan® ) irrespective of fibrosis. Hepatology 2008;48:1718–23. [26] Bosch J, Garcia-Pagán JC, Berzigotti A, et al. Measurement of portal pressure and its role in the management of chronic liver disease. Semin Liver Dis 2006;26:348–62. [27] Kim SY, Seok JY, Han SJ, et al. Assessment of liver fibrosis and cirrhosis by aspartate aminotransferase-to-platelet ratio index in children with biliary atresia. JPGN 2010;51:198–202.