PG was found in 30 (58.8%) patients. Children with cirrhotic PH had the highest rate of PG (64.3%), whereas those with extrahepatic or intrahepatic noncirrhotic PH were similar (50% and 54.5%, respectively); however, the relation between the etiology of PH and PG and PD was not significant (P = 0.38 and P = 0.64, respectively).
All of the patients with PG had scores ≤3 (Table 3). Among them, mucosal mosaic pattern was found in almost all of the patients with cirrhosis (P = 0.041). One patient had gastric antral vascular ectasia (GAVE). Portal hypertensive duodenal lesions (varices at the second portion, isolated red markings, marked vascularisation at bulbus, sessile polyp, and multiple polyps, respectively) were found in 5 patients (9%).
Among 28 patients with chronic liver disease, 18 had Child-Pugh A, 8 Child-Pugh B, and 2 Child-Pugh C cirrhosis. Baveno PG scores were higher in patients with cirrhosis with higher Child scores (P = 0.001) (Fig. 1). The relation between Child scores and PD was not significant (P = 0.53). There was no significant difference in incidence of PG and PD in relation to PELD scores (P = 0.39 and P = 0.53, respectively).
Esophageal varices were found in 46 patients (90%) (grade I, n = 22; grade II, n = 11; grade III, n = 13). Thirteen of them (25%) also had gastric varices (gastroesophageal [GOV] I and GOV II). Seven patients (28%) of 25 intrahepatic cirrhotic, 5 patients (50%) of 10 intrahepatic noncirrhotic PH, and 1 patient (9%) of 11 extrahepatic PH had cherry-red spots on their esophageal varices. No association was found between PG and PD and esophageal varices, number and grade of varices, and cherry-red spots on them. There was no significant difference in incidence of PG and PD in relation to gastric varices (P = 0.33 and P = 0.09, respectively).
Bleeding history was positive in 21 patients. Fifty-seven percent of them bled massively and needed transfusions. The remainder had milder bleeding incidents without transfusion requirement. The total number of bleeding episodes was <5 in most of the patients (12/21). The median number of bleeding episodes during the previous year was 1 (range 0–2). The median time period between endoscopy and the last bleeding episode was 26 weeks (range 0.14–364 weeks). Five episodes of upper gastrointestinal bleeding occurred in 3 patients during the study period. None bled from PG. The bleeding episodes were caused by duodenal ulcer in 1 patient and variceal hemorrhage in others. We did not encounter any iatrogenic bleeding episode in relation to endoscopic forceps biopsies. All of the patients except 1 responded to medical therapy with somatostatin analogs and histamine receptor blockers. We lost 1 patient who bled from esophageal varices and managed with Sengstaken-Blakemore tube for 2 days as a result of sepsis at the 10th day.
No association was found between PG and bleeding history. However, the relation between the bleeding history and PD was significant (P = 0.009). The total number of bleeding, bleeding number during the previous year, and the time period between endoscopy and the last bleeding episode were not significant in relation to PD and PG.
Fifteen patients had not received any treatment for PH. Another 15 patients had undergone 1 to 9 episodes of endoscopic variceal treatment, median 73 weeks (13–312 weeks), before endoscopy along with medical treatment (propranolol). The total number of endoscopic treatments was 56. Five patients had endoscopic band ligation, 2 had sclerotherapy, and 8 had the 2 modalities together. Four patients had surgery (1 splenectomy, 1 splenectomy + distal splenorenal shunt, 2 splenectomy + Siguira operation). The history of endoscopic treatment, the number of sessions, and the time period after the last session showed no correlation with the presence of PG and PD.
Endoscopic diagnosis of PG but not PD showed significant association with the cutaneous findings of chronic liver disease (P = 0.045). No association was found between PG and PD and age of the patient, hepatomegaly, splenomegaly, ascites, laboratory findings, duration and treatment modalities of PH, and presence of H pylori.
Capillary dilation was the only histologic finding showing significant association with the endoscopic diagnosis of PG (P = 0.042) (Fig. 2). None of the histologic findings correlated with endoscopic diagnosis of PD. The addition of scores for each histologic finding either for PG or PD did not make any difference.
Portal hypertensive gastrointestinal mucosal lesions are common complications of PH in children as they are in adults. The prevalence of PG and PD has been reported to be in the range of 4% to 98% and 8% to 69% in the adult population, respectively (2,8–10). Pediatric data on the frequency of PG are limited and vary between 40% and 64% (11–13). There are no published data about the incidence of PD in children. Duodenal polyp seems to be a newly described portal hypertensive duodenal lesion that has not been reported in adults. We found the rate of PG was 58%, similar to previous pediatric studies, and PD was 9%.
Generally, PG is thought to be more common in cirrhotic PH. In a multicenter study among 2720 consecutive adult patients (21% cirrhotic, 79% noncirrhotic) evaluated endoscopically, the rate of PG was found to be 63.3% in children with cirrhosis versus 16.9% in children without cirrhosis (14). In another, larger study, patients were studied for a 2-year period and PG was reported in 61% of children with cirrhosis, 54% of children without cirrhotic PH, and in 20% of children with extrahepatic portal vein obstruction (15). In a pediatric study, no association was found between PG and the cause of PH among 18 children with intrahepatic and 6 children with extrahepatic portal hypertension (13). In our population of 51 patients, children with cirrhotic PH had the highest rate of PG. Additionally, 2 significant findings were the relations between cutaneous findings of chronic liver disease and PG, and between mosaic pattern of gastric mucosa and cirrhosis, supporting the relation between cirrhosis and PG; however, the relation between the etiology of PH and PG and PD was not significant.
Several studies have noted that the likelihood of developing PG was dependent on the severity of liver disease (16–19), whereas the others have not (20–23). In a natural history study, a total of 222 patients with cirrhosis with a mild degree of PH were followed up with upper endoscopy every 12 months for 47 ± 28 months, and Child-Pugh class B or C was found to be associated with the progression from mild to severe PG (17). Similarly, we found that Baveno PG scores were higher in cirrhotics with higher Child-Pugh scores; however, there was no significant difference in incidence of PG and PD in relation to Child-Pugh class and PELD scores. This may be due to the paucity of Child-Pugh class C in our cirrhotic population.
There has been no consensus on the relation of presence and grade of esophageal and/or gastric varices with PG and PD. Some authors found positive correlations between the grade of esophageal varices and PG, whereas some failed to find any association (18–20,24). No correlation between the size of varices and PG in 22 children with PH was reported (11). There was no significant relation between gastric varices and PG in another pediatric study among 40 children with extrahepatic PH (12) similar to some adult studies (15,19). No relation between PD and esophageal (10,24,25) and gastric varices (25) was found. In agreement with these reports, we also failed to find any relation between PG and PD and the presence of esophageal and gastric varices, the size and the number of esophageal varices, and cherry-red spots on them.
Variceal eradication may exacerbate the development of PG. Claims of an association were supported by numerous studies (15,16,23,24). Two publications studying the children with extrahepatic PH during the period 1992 through 2002 emerged from India. PG increased in frequency after eradication of esophageal varices by endoscopic sclerotherapy from 12% to 41% among 163 children and from 24% to 51% among 186 children (26,27). The authors also described an increase in severity of PG. In another pediatric study including 100 consecutive children with cirrhotic and noncirrhotic PH prophylactic, sclerotherapy increased the risk of bleeding from PG (28). On the contrary, it was shown that changes in the severity of PG after endoscopic variceal sclerotherapy or ligation were reversible in adults (29). Similarly, PG developing after variceal eradication was often transitory and less severe (30). In the context of PD, some studies showed an association between variceal eradication and PD (31), whereas some others could not (24,25,32). In our study, no relation among the history of endoscopic treatment, the number of sessions, and the time period after the last session with PG and PD was found. Portal hypertensive gastrointestinal mucosal changes, however, are dynamic conditions that can change during the time of observation. The heterogeneity of our population in terms of endoscopic therapies and retrospective design are limitations of our study.
PG is a well-established source of gastrointestinal bleeding in patients with PH (1,18,33). In a prospective study of 100 adult patients with cirrhotic PH, it was found that the risk of bleeding from PG was accurately reflected by the PG scoring system proposed at Baveno (1). No bleeding was observed in patients who had scores ≤3. In a retrospective study of 22 children with PH, PG was graded using criteria established by McCormack et al, and 0 of 7 patients with mild gastropathy and 2 of 7 patients with severe gastropathy bled from their gastric mucosa (11,23). In another pediatric study of 139 children with biliary atresia, the initial sign of PG was detected at a median age of about 1 year and was shown to be associated with a 5-year actuarial risk of bleeding of 69% compared to a risk of 20% if no PG was present (34). In the present study, all of the patients with PG had scores ≤3 and none bled from PG, similar to previous data. However, GAVE due to PH, which is a component of the Baveno scoring system, is a rare condition in childhood. This makes it less convenient for use in pediatric scoring. We believe that it may be better to report GAVE as a separate positive finding when present in childhood, instead of one of the established parameters of the scoring system.
Histologically, portal hypertensive mucosal lesions correspond to dilated vessels in the mucosa and submucosa in the absence of erosions or inflammation (23,35). In addition, submucosal veins appeared ectatic and irregular with areas of intimal thickening in surgical and autopsy specimens (23). Some studies showed capillary dilation not only in patients with PG but also in controls (36,37) and concluded that there had not been any correlation between endoscopic and histologic findings of PG. Apart from these, the value of endoscopic biopsies in the diagnosis of vascular dilation from the perspective of availability of adequate tissue or sampling error was questioned (38). Large forceps specimens yielded no additional information over standard forceps biopsies when compared to snare biopsy technique (39). Despite these discussions, different studies using different methods, such as silicon injection technique and morphometric analyses with or without endothelial cell markers, supported the relation between capillary dilation and PG (40–42). We found that the only histologic finding showing significant association with the endoscopic diagnosis of PG was capillary dilation.
Edema and capillary dilation were found in the intestinal biopsies of patients with PH (8,10,32); however, statistically significant association between macroscopical PD and histologic findings in portal hypertensive patients was not shown (10). Moreover, similar rates of vascular dilation in patients with PH and control patients with nonulcer dyspepsia were observed (43). Histologic examination of duodenal polyp described in children with PH revealed lobular capillary proliferation in a hemangiomatous pattern in the lamina propria of the duodenal mucosa (44). Similarly, it was concluded that capillary angiogenesis was an important vascular mechanism for adaptation to PH (10). None of the histologic findings correlated with endoscopic diagnosis of PD in the present study; however, we identified a duodenal polyp suggestive of portal hypertensive duodenal mucosal change in 1 of our patients. This lesion had not been noticed in the previous endoscopy session before variceal eradication. Although we did not biopsy the polyp, its development after endoscopic treatment suggested that it was related to PH.
In conclusion, PG and PD are seen in children with extrahepatic and intrahepatic PH at similar rates reported in pediatric and adult studies. Baveno PG scores increased in parallel with Child-Pugh class in children with cirrhosis. Capillary dilation was the only histologic finding showing a significant association with the endoscopic diagnosis of PG in this study. Duodenal polyp seems to be a newly defined mucosal lesion arising in children with PH.
1. Stewart CA, Sanyal AJ. Grading portal gastropathy: validation of a gastropathy scoring system. Am J Gastroenterol 2003; 98:1758–1765.
2. Burak KW, Lee SS, Beck PL. Portal hypertensive gastropathy and gastric antral vascular ectasia (GAVE) syndrome. Gut 2001; 49:866–872.
3. Pesce MA. Reference ranges for laboratory tests and procedures. In: Kliegman RM, Behrman RE, Jenson HB, et al, eds. Nelson Textbook of Pediatrics
. 18th ed. Philadelphia: Saunders Elsevier; 2007;2943–54.
4. McDiarmid SV, Anand R, Lindblad AS. Development of a pediatric end-stage liver disease score to predict poor outcome in children awaiting liver transplantation. Transplantation 2002; 74:173–181.
5. Child CG, Turcotte JG. Surgery in portal hypertension. In: Child CG, editor. Major Problems in Clinical Surgery. The Liver and Portal Hypertension. Philadelphia: WB Saunders; 1964. pp. 1–85.
6. Conn HO, Binder H, Brodoff M. Fiberoptic and conventional esophagoscopy in the diagnosis of esophageal varices. A comparison of techniques and observers. Gastroenterology 1967; 52:810–818.
7. Sarin SK, Kumar A. Gastric varices: profile, classification, and management. Am J Gastroenterol 1989; 84:1244–1249.
8. Shudo R, Yazaki Y, Sakurai S, et al
. Duodenal erosions, a common and distinctive feature of portal hypertensive duodenopathy. Am J Gastroenterol 2002; 97:867–873.
9. Kumar A, Mishra SR, Sharma P, et al
. Clinical, laboratory, and hemodynamic parameters in portal hypertensive gastropathy: a study of 254 cirrhotics. J Clin Gastroenterol 2010; 44:294–300.
10. Barakat M, Mostafa M, Mahran Z, et al
. Portal hypertensive duodenopathy: clinical, endoscopic, and histopathologic profiles. Am J Gastroenterol 2007; 102:2793–2802.
11. Hyams JS, Treem WR. Portal hypertensive gastropathy in children. J Pediatr Gastroenterol Nutr 1993; 17:13–18.
12. Yachha SK, Ghoshal UC, Gupta R, et al
. Portal hypertensive gastropathy in children with extrahepatic portal venous obstruction: role of variceal obliteration by endoscopic sclerotherapy and Helicobacter pylori
infection. J Pediatr Gastroenterol Nutr 1996; 23:20–23.
13. El-Rifai N, Mention K, Guimber D, et al
. Gastropathy and gastritis in children with portal hypertension. J Pediatr Gastroenterol Nutr 2007; 45:137–140.
14. Carpinelli L, Primignani M, Preatoni P, et al
. Portal hypertensive gastropathy: reproducibility of a classification, prevalence of elementary lesions, sensitivity and specificity in the diagnosis of cirrhosis of the liver. A NIEC multicentre study. New Italian Endoscopic Club. Ital J Gastroenterol Hepatol 1997; 29:533–540.
15. Amarapurkar DN, Dhawan PS, Chopra K, et al
. Stomach in portal hypertension. J Assoc Phys India 1993; 41:638–640.
16. Sarin SK, Sreenivas DV, Lahoti D, et al
. Factors influencing development of portal hypertensive gastropathy in patients with portal hypertension. Gastroenterology 1992; 102:994–999.
17. Merli M, Nicolini G, Angeloni S, et al
. The natural history of portal hypertensive gastropathy in patients with liver cirrhosis and mild portal hypertension. Am J Gastroenterol 2004; 99:1959–1965.
18. Parikh SS, Desai SB, Prabhu SR, et al
. Congestive gastropathy: factors influencing development, endoscopic features, Helicobacter pylori
infection, and microvessel changes. Am J Gastroenterol 1994; 89:1036–1042.
19. Bayraktar Y, Balkancı F, Uzunalimoglu B, et al
. Is portal hypertension due to liver cirrhosis a major factor in the development of portal hypertensive gastropathy? Am J Gastroenterol 1996; 9:554–558.
20. Vigneri S, Termini R, Piraino A, et al
. The stomach in liver cirrhosis. Endoscopic, morphological, and clinical correlations. Gastroenterology 1991; 101:472–478.
21. Merkel C, Schipilliti M, Bighin R, et al
. Portal hypertension and portal hypertensive gastropathy in patients with liver cirrhosis: a haemodynamic study. Dig Liver Dis 2003; 35:269–274.
22. Curvêllo LA, Brabosa W, Rhor R, et al
. Underlying mechanism of portal hypertensive gastropathy in cirrhosis: a hemodynamic and morphological approach. J Gastroenterol Hepatol 2009; 24:1541–1546.
23. McCormack TT, Sims J, Eyre-Brook I, et al
. Gastric lesions in portal hypertension: inflammatory gastritis or congestive gastropathy? Gut 1985; 26:1226–1232.
24. Gupta R, Saraswat VA, Kumar M, et al
. Frequency and factors influencing portal hypertensive gastropathy and duodenopathy in cirrhotic portal hypertension. J Gastroenterol Hepatol 1996; 11:728–733.
25. Figueiredo P, Almeida N, Lérias C, et al
. Effect of portal hypertension in the small bowel: an endoscopic approach. Dig Dis Sci 2008; 53:2144–2150.
26. Itha S, Yachha SK. Endoscopic outcome beyond esophageal variceal eradication in children with extrahepatic portal venous obstruction. J Pediatr Gastroenterol Nutr 2006; 42:196–200.
27. Poddar U, Thapa BR, Singh K. Frequency of gastropathy and gastric varices in children with extrahepatic portal venous obstruction treated with sclerotherapy. J Gastroenterol Hepatol 2004; 19:1253–1256.
28. Gonçalves ME, Cardoso SR, Maksoud JG. Prophylactic sclerotherapy in children with esophageal varices: long-term results of a controlled prospective randomized trial. J Pediatr Surg 2000; 35:401–405.
29. Hou MC, Lin HC, Chen CH, et al
. Changes in portal hypertensive gastropathy after endoscopic variceal sclerotherapy or ligation: an endoscopic observation. Gastrointest Endosc 1995; 42:139–144.
30. Sarin SK, Shahi HM, Jain M, et al
. The natural history of portal hypertensive gastropathy: influence of variceal eradication. Am J Gastroenterol 2000; 95:2888–2893.
31. Menchén L, Ripoll C, Marín-Jiménez I, et al
. Prevalence of portal hypertensive duodenopathy in cirrhosis: clinical and haemodynamic features. Eur J Gastroenterol Hepatol 2006; 18:649–653.
32. Nagral AS, Joshi AS, Bhatia SJ, et al
. Congestive jejunopathy in portal hypertension. Gut 1993; 34:694–697.
33. Primignani M, Carpinelli L, Preatoni P, et al
. Natural history of portal hypertensive gastropathy in patients with liver cirrhosis. The New Italian Endoscopic Club for the study and treatment of esophageal varices (NIEC). Gastroenterology 2000; 119:181–187.
34. Duché M, Ducot B, Tournay E, et al
. Prognostic value of endoscopy in children with biliary atresia at risk for early development of varices and bleeding. Gastroenterology 2010; 139:1952–1960.
35. Quintero E, Pique JM, Bombi JA, et al
. Gastric mucosal vascular ectasias causing bleeding in cirrhosis. Gastroenterology 1987; 93:1054–1061.
36. Misra SP, Dwivedi M, Misra V, et al
. Endoscopic and histologic appearance of the gastric mucosa in patients with portal hypertension. Gastrointest Endosc 1990; 36:575–579.
37. Corbishley CM, Saverymuttu SH, Maxwell JD. Use of endoscopic biopsy for diagnosing congestive gastropathy. J Clin Pathol 1988; 41:1187–1190.
38. Viggiano TR, Gostout CJ. Portal hypertensive intestinal vasculopathy: a review of the clinical, endoscopic, and histopathologic features. Am J Gastroenterol 1992; 87:944–954.
39. Saperas E, Pigué JM, Perez-Ayuso R, et al
. Comparison of snare and large forceps biopsies in the histologic diagnosis of gastric vascular ectasia in cirrhosis. Endoscopy 1989; 21:165–167.
40. Hashizume M, Tanaka K, Inokuchi K. Morphology of gastric microcirculation in cirrhosis. Hepatology 1983; 3:1008–1012.
41. Foster PN, Wyatt JI, Bullimore DW, et al
. Gastric mucosa in patients with portal hypertension: prevalence of capillary dilatation and Campylobacter pylori
. J Clin Pathol 1989; 42:919–921.
42. McCormick PA, Sankey EA, Cardin F, et al
. Congestive gastropathy and Helicobacter pylori
: an endoscopic and morphometric study. Gut 1991; 32:351–354.
43. Desai N, Desai D, Pethe V, et al
. Portal hypertensive jejunopathy: a case control study. Indian J Gastroenterol 2004; 23:99–101.
44. Devadason D, Murphy MS, Brown R, et al
. Duodenal capillary hemangiomatous polyps: a novel manifestation of extrahepatic portal hypertension? J Pediatr Gastroenterol Nutr 2007; 45:114–116.
Keywords:Copyright 2011 by ESPGHAN and NASPGHAN
childhood; duodenopathy; gastropathy; histologic findings; portal hypertension