Familial and Idiopathic Colonic Varices: an Unusual Cause Oflower Gastrointestinal Haemorrhage

Gut 1992; 33:1285-1288 1285

CASE REPORTS Gut: first published as 10.1136/gut.33.9.1285 on 1 September 1992. Downloaded from

Familial and idiopathic colonic varices: an unusual cause oflower gastrointestinal haemorrhage

J P Iredale, P Ridings, F P McGinn, M J P Arthur

Abstract minute and his blood pressure 140/90 mm Hg. A patient is described presenting with an acute Examination ofthe abdomen and rectal examina- lower gastrointestinal haemorrhage as a result tion were both normal. Sigmoidoscopy to 15 cm ofextensive colonic varices. Further investiga- revealed what was thought to be a polyp, this was tion revealed that there were no oesophageal biopsied and histology revealed normal rectal varices or splenomegaly. Liver biopsy showed mucosa only. A barium enema was carried out grade II fatty change only, with no other and was reported as showing 'polyposis coli' specific or significant pathological features. (Figure). Colonoscopy was undertaken and Transhepatic portography showed a raised showed extensive colonic varices. Gastroscopy portal pressure (20 mm/Hg) but the portal showed a normal lower oesophagus. Infrared system was patent throughout. There was an endoscopy excluded dilated submucosal veins in abnormal leash of vessels in the caecum the stomach or oesophagus. Haemoglobin was thought to represent a variceal plexus. This 14-9 g/l. Clotting studies were normal, ferritin, patient was diagnosed as having idiopathic alphafetoprotein, autoantibody profile, and colonic varices. This case is discussed to- immunoglobulins were all normal. Hepatitis gether with nine other reports of idiopathic serology was negative. The liver was bright on colonic varices from the published literature. sonographic scan, consistent with fatty change, Four of these reports describe idiopathic the spleen was normal and three was no evidence colonic varices in more than one member of of an abnormal portal vein or variceal change the same family. Possible modes of inherit- around the cardia and splenic hilum. ance, aetiology of variceal change, natural Superior and inferior mesenteric arterio- http://gut.bmj.com/ history, and prognosis are discussed. graphy was undertaken. The showed a normal (Gut 1992; 33: 1285-1288) arterial anatomy with no evidence of early venous filling, in general the venous anatomy was normal although the veins of the rectum and There are approximately 70 reports in the literature of bleeding colonic varices. This rare cause of lower gastrointestinal haemorrhage is almost on October 1, 2021 by guest. Protected copyright. invariably associated with liver cirrhosis and consequent portal hypertension' or portal venous obstruction.2 We report a patient who presented with bleeding colonic varices. He has been extensively investigated and we conclude that he has idiopathic colonic varices. We describe his case and review the small number of similar cases reported in the literature.

Case report Medicine II, University of Southampton, Mr DM is a 58 year old white man who works as a Southampton labourer at an oil refinery. He presented with a J P Iredale 10 week history of frank rectal bleeding associ- M J P Arthur ated with transient left iliac fossa pain. There was Department of Surgery, no associated change in bowel habit nor any Southampton General weight loss. He had passed blood per rectum 21 Hospital, Southampton years previously at which stage he was diagnosed P Ridings F P McGinn as having haemorrhoids and had been treated Correspondence to: conservatively. Bleeding had ceased spon- Dr J P Iredale, Lecturer in taneously and had not recurred. There was no Medicine, Medicine II, Southampton General relevant family history. He drank 8 units of Hospital, Tremona Road, alcohol a week and smoked a pipe which he did Southampton, Hampshire S09 4XY. not inhale. Barium enema, view oftransverse colon and hepaticflexure Accepted for publication On examination he was mildly obese and showing colonic varices which were misdiagnosed as polvposis 7 January 1992 looked well. His pulse rate was 90 beats per coli. 1286 Iredale, Ridings, McGinn, Arthur

Previous reports ofidiopathic andfamilial colonic varices sigmoid were prominent. Transhepatic porto- Case report and Family graphy and portal pressure studies showed a

reference no Sex Age Presentation history Comment raised portal venous pressure, the values being: Gut: first published as 10.1136/gut.33.9.1285 on 1 September 1992. Downloaded from Hardy, 1967 inferior mesenteric vein 20 mm Hg, distal 5 a M 20 1 Acute Patient (a) had SMV pressure splenic vein 19 mm Hg, portal vein 18 mm Hg. rectal 55 cm H20. All patients Portographic studies showed an abnormal series b M - bleeding ultimately required surgery for abnormal ileocolic artery of tortuous submucosal veins in the caecum, c F 20 10 year history which encircled and thought to represent a caecal variceal plexus. rectal bleeding compressed SMV. Other runs confirmed that the portal system was Solis-1Herruzo, 1977 6 a F 18 7 year history Normal portogram. Portal patent throughout. A liver biopsy undertaken at rectal bleeding pressure 12 cm H20. Liver the time of portography showed the histological function test normal. features of grade II fatty change with no fibrosis Extensively investigated. At laparotomy varices or other specific or significant pathological throughout colon. No features. apparent cause. F 26 15 year history + Liver function tests normal rectal bleeding Parents and 3 siblings of case 7a Discussion underwent colonoscopy - all There are approximately 70 reports in the litera- were normal Weingart, 1982 ture of bleeding from colonic varices. Compared 13 M 70 63 year history Bled profusely after biopsy of with oesophageal varices or other upper gastro- rectal bleeding varix thought to be a polyp. intestinal lesions, colonic varices are an infre- Varices throughout colon. Treated conservatively quent source of blood loss in cirrhotic patients, Lieberman, 198l3 but the majority of patients who present with 11 M 37 19 year history Hepatitis B surface antigen rectal bleeding positive. Cirrhotic portal bleeding from colonic varices have cirrhosis.' hypertension not excluded. Bleeding from colonic varices has been reported Not considered further as a to occur from 2-5% of patients.attending for case of 'idiopathic' colonic varices sclerotherapy to oesophageal varices,3 and Hawkey, 1984 between 3.6% and 56% of cirrhotic patients have 7 a M 66 41 year history + Varices throughout colon. been shown to have rectal varices. We have rectal bleeding Normal barium meal. Portal vein patent and liver normal identified a patient with extensive colonic varices on ultrasound scan. Barium (Figure) which we conclude to be 'idiopathic.' enema reported as polyposis Our investigations of this patient have excluded coli b F 71 50 year history + Varices noted from jejunum to significant liver disease or any evidence of portal rectal bleeding rectum at laparotomy in or mesenteric venous thrombosis. His portal 1944. Liver Bx at operation venous pressure was shown to be above the 'normal'. Both cases (8) treated upper limit ofthe normal range (10 mm Hg). It is http://gut.bmj.com/ conservatively possible that an arteriovenous anomaly may Planuss, 1988 10 M 75 Malaena Idiopathic nature of varices have caused the raised pressure and the develop- duration not 'dubious'. Uneven liver ment of varices but there was no angiographic stated texture at laparoscopy. Portal evidence of such an abnormality. fibrosis on liver biopsy. Oesophageal varices at We have identified nine other reports, com- endoscopy. Wedged hepatic prising 14 patients, of idiopathic colonic varices pressure stated to be similar to

in the literature (Table). Four of these reports on October 1, 2021 by guest. Protected copyright. hepatic venous pressure. Managed conservatively describe an index patient with at least one first Beerman, 1988 degree relative who has evidence of colonic 8 a M 65 Mass felt on + Varices throughout colon. varices. '-'Before concluding that these represent rectal Barium enema reported as examination polyposis coli. Severe rectal true idiopathic colonic varices significant liver bleeding after polyp biopsied disease and portal venous obstruction must b F 65 13 year history + Varices throughout colon. obviously be excluded. In only two of the eight rectal bleeding Dukes's B carcinoma colon resected 1981 at which time no reports6 9 has investigation been adequate to other abnormality noted. exclude portal hypertension, liver disease and Misdiagnosed as polyposis coli oesophageal varices. Two of the other reported on barium enema. Both cases (9) managed patients underwent liver biopsy.7 "' Both were conservatively interpreted as having no significant histological Isbister, 1989 features although that reported by Planus"' was 9 F 18 8 month history Portal pressure 22 cm H20. rectal bleeding Normal portogram. said to show 'portal fibrosis' therefore the Mesenteric arteriography 'idiopathic' nature of varices must be viewed normal. Normal oesophagus. Varices in ascending colon and with some reticence in this case. In addition, sigmoid. At laparotomy three patients were said to have a normal liver at evidence of caecal laparotomy58b9 and in a fourth a normal ultra- arteriovenous malformation. Initially treated conservatively sound scan texture was reported.7a Liver with transfusion. Ultimately function tests and serological markers of required extended right hepatitis are reported in four cases.7; 9'1 12One of hemicolectomy Nikolopoulos, 1990 these patients" was shown to have hepatitis B 12 M 22 Extensive colonic varices showed surface antigenaemia. There is no documenta- at colonoscopy and tion of this patient's portal venous anatomy or angiography. Liver disease not excluded. pressure. Because of the presence of such a Significant variceal haemorrhage significant risk factor for cirrhosis, this report" is at presentation necessitated not considered further in the review ofidiopathic subtotal colectomy colonic varices. Familial and idiopathic colonic varices 1287

In addition to the reports in which portal may be the result of vascular degenerative venous pressure and anatomy was docu- change. The proposition that simple degenera- mented,sa6a912 indirect evidence of portal vein tive ectasia leads to development of extensive patency is presented for patients 5a and 10. colonic varices seems rather tenuous, however. Gut: first published as 10.1136/gut.33.9.1285 on 1 September 1992. Downloaded from Oesophageal varices were excluded in five It is important to note that even in cirrhotic and patients.' la7a8 12 Thus in three cases"" la 13 there is non-cirrhotic portal hypertension it is unusual no documented evidence that might exclude for colonic varices to extend beyond the cirrhosis or portal/mesenteric vein obstruction. anorectal area,2 in contrast with these patients. Lastly, in patient 8b the varices documented to We suggest that all of these patients had a be throughout the colon at presentation had not significant inborn vascular anomaly. This group been noted seven years previously when the were managed conservatively with blood trans- patient underwent resection of a Dukes's B fusion only, no deaths as a result of variceal carcinoma of the descending colon. Colonic bleeding are reported, and although the numbers varices have been reported as a complication of are small, it would seem reasonable to adopt a colonic surgery'4 and as a result of adhesional conservative approach to those patients present- obstruction to,'5 or tumour invasion of,'6 the ing in their fifth decade or later. It is salutary to mesenteric venous system. note that cases 7a, 8a, 8b, and our patient were Case 6a was extensively investigated and diagnosed as having polyposis coli after barium together with her aunt6b appears to represent a enema. Given the familial nature of idiopathic true case of familial idiopathic colonic varices. colonic varices and the apparent similarity ofthe Despite the possibility that case 8b may have radiographic and sigmoidoscopic appearances of resulted from previous colonic surgery, the colonic varices to polyps, misdiagnosis, and occurrence of varices in a first degree relative inappropriate biopsy remain potential pitfalls in with no apparent predisposing factor means that the diagnosis of this condition. The prognosis of the two patients reported by Beerman,8ab idiopathic colonic varices would seem to be good like those reported by Hawkey7ab and Solis- at all ages compared with cirrhotic varices,4 this Herruzo,"a b may also represent true familial may be related to low pressure in the varices as colonic varices, a distinct clinical entity. There is well as the absence of significant hepatocellular some evidence that portal venous pressure is not disease. raised in these patients and we must therefore Idiopathic colonic varices are a very rare cause assume they result from congenital vascular of rectal bleeding which can only be confidently anomalies. The numbers are too small to draw diagnosed after liver disease and portal venous any firm conclusion on possible modes ofinherit- obstruction have been excluded. They may ance. The fact that case 6a's siblings and parents present at any stage from the first decade of life all had normal colonoscopies, however, would onwards. Presentation at a young age is associ-

favour a recessive rather than dominant pattern ated with significant haemorrhage and there may http://gut.bmj.com/ if the inheritance is autosomal. be an identifiable vascular anomaly. Surgical The age of presentation and treatment of treatment to prevent further haemorrhage is patients with idiopathic colonic varices falls into usually required in these patients. Presentation two groups. Those who ultimately undergo in later life may be successfully treated conserva- laparotomy present in their 20s or younger. tively. In this group, the condition may be Patients 5a, Sb, and 5c presented with severe misdiagnosed as polyposis coli particularly if other members of the are affected. In all bleeding and ultimately a vascular anomaly was family on October 1, 2021 by guest. Protected copyright. shown. This is perhaps the most remarkable of age groups there may be a positive family history all the reports, comprising three first degree with a first degree relative affected and this may relatives, all of whom presented with significant relate to an inherited mesenteric vascular rectal bleeding and evidence of mesenteric anomaly. There is no evidence to suggest that venous hypertension, in whom the same annular these patients constitute a homogenous group, aneurysm of the ileocaecocolic artery was found. indeed they may represent a common result of a This bizarre vascular anomaly resulted in com- series of inherited or degenerative vascular pression of a mesenteric vein with consequent anomalies. In all age groups the condition carries variceal change distally. Each required vascular a good prognosis. surgery. Case 6a also underwent exploratory laparotomy but a structural cause for colonic 1 Gudjonsson H, Zeiter D, Gamelli R, Kaye M. Colonic varices: varices was not identified. In case 9 an attempt to report of an unusual case diagnosed by radionuclide scan- manage the patient conservatively with repeated ning, with review ofthe literature. Gastroenterology 1986; 91: 1543-7. transfusion was attempted. Recurrent bleeds 2 Chawla Y, Dilawari JB. Anorectal varices - their frequency in and the need for transfusion, however, meant cirrhotic and non-cirrhotic portal hypertension. Gut 1991; 32: 309-11. that the patient missed significant amounts of 3 Foutch PG, Sirak MV. Colonic variceal haemorrhage after schooling and surgery was therefore undertaken. endoscopic infection sclerosis ofoesophageal varices a report of three cases. AmJ Gastroenterol 1984; 79: 756-60. Case 12 underwent subtotal colectomy for 4 Rabinovitz M, Shade R, Dindzaus VJ, Belle SH, Van Thiel uncontrolled rectal bleeding within 12 days of DH, Gavaler JS. Colonic disease in cirrhosis. Gastro- enterology 1990; 99: 195-9. presentation. 5 Hardy M, Barbin J-Y, Caroli J.Hypertension mesenterique In contrast, the age ofpresentation in cases 7a, familiale. Revue des Maladies de Foie 1967; 6: 236-46. 6 Solis-Herruzo JA. Familial varices of the colon diagnosed by 7b, 8a, 8b, 10, 13 was from the fifth decade colonoscopy. Gastrointest Endosc 1977; 24: 85-6. onwards. It is of interest that a history of 7 Hawkey CJ, Amar SS, Daintith HAM, Toghill PJ. Familial varices of the colon occurring without evidence of portal transient rectal bleeding extended back to the hypertension. BrJZ Radiol 1985; 58: 677-9. third decade in two patients of these7ab and to the 8 Beerman EM, Lagaay MB, Van-Nonhuys JM, Overbosh D. Familial varices of the colon. Endoscopy 1988; 20: 270-2. first decade in one,"3 suggesting the presence of a 9 Isbister WH, Pease CW, Delahunt B. Colonic varices, report congenital vascular anomaly. Late presentation of a case. Dis Col Rectum 1989; 32: 524-7. 1288 Iredale, Ridings, McGinn, Arthur

10 Planus R, Boix J, de Leon R, Pedro-Botet ML, Humbert P, 13 Weingart J, Hochter W, Ottenjann R. Varices of the entire Villagrasa M. Bleeding colonic varices in a patient with colon - an unusual cause of recurrent intestinal bleeding. idiopathic portal hypertension. ] Clin Gastroenterol 1988; 10: Endoscopy 1982; 14: 69-70. 235. 14 Arbona JL, Lichtenstein JE, Spies JB, Hyatt JR Jr. Colonic

11 Lieberman DA, Krippaeline WW, Melnyk CS. Colonic varices as a complication of colonic surgery. Gastrointest Gut: first published as 10.1136/gut.33.9.1285 on 1 September 1992. Downloaded from varices due to intestinal cavernous hemangiomas. DigDis Sci Radiol 1987; 12: 350-2. 1983; 28: 852-8. 15 Manzi D, Samanta AK. Adhesion related colonic varices. 12 Nickolopoulos N, Xynas E, Datnakis K, Kasapidis P, .7 Clin Gastroenterol 1985; 7: 71-5. Vassilakis JS. Varicosis coli totalis: report of a case of 16 Grangvist S. Colonic varices caused bv a carcinoid tumour. idiopathic aetiology. Digestion 1990; 47: 232-5. Gastrointest Radiol 1984; 9: 269-76. http://gut.bmj.com/ on October 1, 2021 by guest. Protected copyright.