The Fundal Pile: Bleeding Gastric Varices

By A.J.W. Millar, R.A. Brown, I.D. Hill, H. Rode, and S. Cywes

Cape Town, South Africa

0 Bleeding from esophageal varices is a common cause of ered. This report describes the management and
major upper gastrointestinal tract blood loss in children with outcome of 5 patients who underwent emergency
portal hypertension but usually ceases spontaneously or is
operation to control continued bleeding from gastric
satisfactorily managed by nonoperative measures. Massive
hemorrhage from gastric fundal varices may be difficult to
fundal varices.
control with compression and sclerotherapy; in these cases,
a direct surgical approach may be indicated. Since 1964.27
MATERIALS AND METHODS
children have undergone aggressive injection sclerotherapy Over the period from 1984 through 1988, 27 patients were
for bleeding esophageal/gastric varices. Nine (6 with portal treated at Red Cross War Memorial Children’s Hospital, Cape
vein thrombosis) bled from gastric fundal varices. In 5 of Town for upper gastrointestinal tract bleeding from varices second-
these, medical management and sclerotherapy failed to ary to portal hypertension (Table 1). Their ages at first presenta-
control the acute bleed. In all 5 there was “rupture” of a large tion ranged from 2 to 11 years. There were 12 boys and 15 girls.
gastric fundal varix or “pile” and bleeding was controlled at Initial management included resuscitation with restoration of
emergency laparotomy by underrunning the varices through blood volume and clotting factors, followed within 24 hours by
a high anterior gastrotomy. Four have subsequently been endoscopic identification of the bleeding site. A total of 5 to 10 mL
successfully managed by continued sclerotherapy and one of ethanolamine oleate was injected at the site of the bleed as well
patient with cirrhosis has died of liver failure. In 3 of the as proximally in an upward spiral fashion for a distance of 10 cm.
survivors both esophageal and gastric fundal varices have Initially the sclerosant was injected into the varices using a rigid
been completely obliterated. No further life-threatening hem- esophagoscope. Since 1985 paravariceal injections using a flexible
orrhage has occurred in any case during a follow-up period of fiberoptic endoscope have been used. Gastric varices were identi-
1 to 6 years. Bleeding from gastric varices is more common fied as the source of bleeding in 2 patients at initial endoscopy and
than previously recorded and more difficult to control by in 7 patients subsequently4 after 1 injection and 3 after 2
nonoperative management, including injection sclerother- injections. In 5 of these 9 cases sclerotherapy and other conserva-
apy. In uncontrolled hemorrhage from gastric varices, surgi- tive measures were unable to control active bleeding (Tables 1 and
cal underrunning offers a means of providing initial control. 2).
Thereafter, the inevitable variceal recurrence may be success- Surgical exploration was performed through an upper midline
fully treated with sclerotherapy. abdominal incision and a high anterior gastrotomy. Blood clot was
Copyright o 7997 by W.B. Saunders Company evacuated, and the gastric varices were identified. In 4 cases a
continuous jet of blood was seen arising from one of approximately
INDEX WORDS: Portal hypertension; gastric varices. 4 large varices that coursed from the gastric fundus toward the
esophagogastric junction. In the fifth patient a fresh blood clot
identified the site of the bleed. Commencing with the actively

E SOPHAGEAL VARICES are a common cause

of serious upper gastrointestinal tract hemor-
rhage in children.‘” Morbidity and mortality are
bleeding varix, each of the varices was underrun throughout its
extent from the esophagogastric junction distally using a 2/O
absorbable suture. In all 5 patients there was complete control of
worse in those with varices secondary to an intrahe- the bleeding.
Table 2 shows the long-term follow-up of the 5 patients. One
patic cause rather than extrahepatic portal vein
patient with cirrhosis of the liver died of liver failure 6 months after
obstruction.4 In most cases the bleeding site in the surgery. A further patient developed an esophageal stricture that
esophagus can be identified at endoscopy and con- responded well to dilatation. The surviving patients are well 26 to
trolled with sclerosant injection.*x’jThereafter, preven- 59 months later, with completely obliterated varices in 3. These 4
tion of further bleeding from esophageal varices is patients have continued to undergo follow-up endoscopy at increas-
ing intervals.
possible with repeated endoscopic injection sclerother-
apy.j,’ DISCUSSION
A small proportion of patients presenting with an
Most acute variceal bleeds in children respond to
acute variceal bleed fail to respond to conservative
nonoperative management and bleeding ceases spon-
measures, including intravenous vasopressin infusion,
balloon tamponade, and injection sclerotherapy. One
reason for failure is bleeding from gastric varices.8,9
From the Departments of Paediam’c Surgery and Paediam’cs,
Not only are they difficult to identify while there is Institute of Child Health and Red Cross War Memorial Children’s
active bleeding, but they may be inadequately com- Hospital, Cape Town, South Africa.
pressed by a Sengstaken tube; are difficult to inject, Date accepted: May 9, 1990.
particularly when they protrude into the lumen of the Address reprint requests to A.J. W. Millar, MD, Department of
Paediatric Surgery, Institute of Child Health, Red Cross War Memorial
gastric fundus at the esophagogastric junction; and Children’s Hospital, Rondebosch 7700, Cape Town, South Africa.
they may be associated with major complications.‘O,ll Copyright o 1991 by W.B. Saunders Company
In these patients emergency surgery must be consid- 0022-3468/91/2606-0015$03.00/0

JournalofPediafricSufgery, Vol26,No6(June),1991: ~~707-709 707

708 MILLAR ET AL

Table 1. Injection Schlerotherapy Management of Bleeding Esophageal and Gastric Varices, Red Cross War Memorial Children’s Hospital
(January lg84-June 1989)

Follow-Up

Total Gastric Mean No. of Outcome

No. of Patients Varices Surgery Injections Until Duration
Etiology (n = 27) (n = 9) (n = 5) Obliteration (mo) Cured Died

Portal vein thrombosis 20 7 4 4.3 20.2 18* 0

(5-59)
Cirrhosis 5 2 1 4.5 33.8 3 2
(45-46)
Hepatic fibrosis 2 0 0 5 41.5 2 0
(3746)

“Two still have a single residual varix after 6 and 7 injections, respectively.

taneously.1’3 Management of those patients who con- strated in 3 of the present patients, attempted gastric
tinue bleeding may include immediate injection scle- variceal compression by tamponade with a Blakemore-
rotherapy or emergency surgical procedures, such as Sengstaken tube may be unsuccessful. In 2 patients it
esophagogastric devascularization or portasystemic did not control the initial hemorrhage; in the third,
shunting.4.‘2 Unfortunately, some postsurgery rebleed- bleeding recommenced immediately on deflation of
ing and development of long-term encephalopathy the balloon.
can be anticipated in shunted patients.4Z13”4Because Recently, Hosking and Johnson proposed a classifi-
of the high morbidity and occasional mortality with cation of gastric varices based on their dominant
emergency surgery, endoscopic sclerotherapy to con- location within the stomach.8 Their classification
trol bleeding from varices has become popular.1X6Z7 divides gastric varices into 3 types. Type II, in which
Howard et al report variceal obliteration by injec- there are gastric varices converging on the cardia, and
tion sclerotherapy in 92% of children with an extrahe- which may be accompanied by smaller esophageal
patic cause of portal hypertension and 75% with an varices above the esophagogastric junction, repre-
intrahepatic cause.’ Of the 108 patients in their sents the bleeding gastric varices encountered in the
series, none had gastric varices identified as the children in this series. The incidence of gastric varices
source of the initial bleed but one child died as a causing variceal hemorrhage in Hosking’s series of
result of uncontrolled hemorrhage from esophagogas- more than 200 patients was 6%. A further 9%
tric varices. Twelve patients subsequently rebled and developed gastric varices during long-term follow-up,
4 were found to be bleeding from gastric varices.5 In many after repeated esophageal variceal sclerother-
contrast, a third (9/27) of children with upper gas- apy. Controversy exists as to whether the gastric
trointestinal bleeding secondary to portal hyperten- varices occur ab initio or secondary to obliterative
sion in this series were found to be bleeding from sclerotherapy. The increased incidence of gastric
gastric varices. The diagnosis of bleeding from gastric varices postsclerotherapy may be due to the develop-
rather than esophageal varices may be difficult in the ment of submucosal collaterals around the fundus. In
acute phase because the field is obliterated by active two of the present patients bleeding from gastric
bleeding and the gastroesophageal cardia is relatively varices occurred before any injection sclerotherapy,
inaccessible even with the flexible endoscope. Once suggesting that they were primary in nature.
identified as the source of bleeding, it may still be The name “fundal pile” was thought appropriate
technically difficult to inject sclerosant into or around because of the resemblance of the dependant bleed-
a varix high up on the gastric fundus. As demon- ing fundal varix to a prolapsed bleeding hemorrhoid.

Table 2. Follow-Up After Surgical Control of Gastric Variceal Hemorrhage in 5 patients

Age at Duration of No. of Re- Further

Varices Operation Follow-Up bleeds From Variceal current
W Diagnosis &no) Varices Injections Status

4.5 PVT 59 0 3 No varices, stricture dilated

4.5 PVT 35 0 3 Grade II varix, injected at 24 mo
12 PVT 29 0 2 No varices
4 PW 26 0 2 No varices
10.9 Hepatitis, cirrhosis 5 5 3 Died; liver failure; varices still present

Abbreviation: PVT. portal vein thrombosis.

BLEEDING GASTRIC VARICES 709

In both conditions there is a submucosal varix protrud- could not be controlled by these methods. In all such
ing into the lumen of the gastrointestinal tract. They cases emergency surgery was lifesaving because imme-
both complicate with fresh bleeding and are related diate control of hemorrhage was obtained. Rather
to alimentary canal sphincters-the lower esophageal than use an extensive technically difficult procedure
sphincter and the internal anal sphincter. It is possi- in a hemodynamically unstable patient, we chose a
ble the pinch-cock action of these sphincters plays a relatively simple surgical procedure. The total opera-
role in development of the varix and precipitation of tion time was short, and control of hemorrhage was
bleeding by impeding venous return and increasing rapidly achieved.
intravariceal pressure. The metaphor is further sus- Although there is control of the acute bleed, these
tained by a recent publication, in which it is suggested patients are still at risk because the underlying portal
that bleeding esophageal varices may be treated by hypertension remains. Therefore, repeated injection
“banding” as advocated for hemorrhoids.‘5 sclerotherapy is advocated after surgery until com-
The authors’ management of children with bleed- plete variceal obliteration is obtained. In the present
ing varices entails resuscitation, followed by pitressin series this took a mean of 4 to 5 injections, which is
and immediate injection sclerotherapy, with balloon similar to the 5- to 6-injection sessions reported in the
tamponade if necessary to temporarily control blood Howard et al series.’ The efficacy of this approach is
loss. Subsequent injections are continued at weekly confirmed in 4 of 5 patients in the present series, in
intervals until there is complete variceal obliteration. whom there has been long-term control of variceal
This has been most successful with bleeding from hemorrhage. In 3 of the 4 there were no varices at
esophageal varices and this experience is similar to most recent endoscopy. The one death that occurred
that reported previously.5,6.‘6However, this approach was 6 months after underrunning of bleeding gastric
is less successful when bleeding is from gastric varices varices and was due to progressive liver failure and
and in 50% of the patients in this series blood loss not to variceal hemorrhage.

REFERENCES
1. Fonkalsrud EW, Myers NA, Robinson MJ: Management of tric varices. An endoscopic study of 50 cases. Arch Surg 92:944-947,
extrahepatic portal hypertension in children. Ann Surg 180:487- 1966
493,1974 10. Yassim YM, Eita MS, Hussein A: Endoscopic sclerotherapy
2. Webb LJ, Sherlock S: The aetiology, presentation and natural for bleeding gastricvarices. Gut 26 A:1105, 1985 (abstr)
history of extrahepatic portal venous obstruction. Q J Med 48:627- 11. Trudeau L, Prindiville T: Sclerosis of bleeding gastric
639.1979 varices. Gastroenterology 84:1338,1983 (abstr)
3. Househam KC, Bowie MD: Extrahepatic portal venous ob- 12. Superina RA, Weber JL, Shandling B: A modified Sugiura
struction-A retrospective analysis. S Afr Med J 64:234-236, 1983 operation for bleeding varices in children. J Pediatr Surg 18:794-
4. Bismuth H, Franc0 D, Alagille D: Portal diversion for portal 799,1983
hypertension in children. Ann Surg 192:18-24,198O 13. Voorhees AB, Chaitman E, Schneider S: Portal-systemic
5. Howard ER, Stringer MD, Mowat AP: Assessment of injec- encephalopathy in the non-cirrhotic patient: Effect of portal-
tion sclerotherapy in the management of 152 children with oesoph- systemic shunting. Arch Surg 107:659-663, 1973
ageal varices. Br J Surg 75404-408, 1988 14. Alagille D, Corlier JC, Chiva M, et al: Long term neuropsy-
6. Lilly JR: Endoscopic sclerosis of esophageal varices in chil- chological outcome in children undergoing portal systemic shunts
dren. Surg Gynecol Obstet 52:513-514,198l for portal vein obstruction without liver disease. J Pediatr Gastro-
7. Paquet KJ: Ten year experience with paravariceal injection enter01 Nutr 5:861-866,1986
sclerotherapy of esophageal varices in children. J Pediatr Surg 15. van Stiegmann G: Advances in surgical technologies. Endo-
20:109-112,1985 scopic ligation of esophageal varices. Am J Surg 156:9A-12B, 1988
8. Hosking SW, Johnson AG: Gastric varices: A proposed 16. Spence RAJ, Johnston GW, Odling-Smee GW, et al: Bleed-
classification leading to management. Br J Surg 75:195-196,1988 ing oesophageal varices with long term follow-up. Arch Dis Child
9. Dogradi AE, Stempien SJ, Owens LK: Bleeding esophagogas- 59:336-340, 1984