Med. J. Cairo Univ., Vol. 62, No. 2, June : 317-325, 1994

Anorectal Varices in

ALIGABALLA, M.D.; HANAA EL DEGWY, M.D; MAZEN NAGA, M.D.; NADIA EL FIKI, M.D. and HUSSEIN KHAYRI, M.D.

The Internal Medicine and Surgical Departments, Faculty of Medicine, Cairo University.

Abstract

The study comprised 135 patients with liver cirrhosis and @rtal . They were classified into three groups: Group A comprising 55 patients that had no history of esophageal , Group B comprising 70 patients that had undergone endoscopic injecrion for bleeding and Group C comprising 10 patients that had undergone splenectomy vasoligation. The prevalence of and has been studied in the three groups. Only anoreetal varices showed significant increase in their incidence after esophageal @ -z O.OOS), but both hemorrhoids and aoorectal varices, increase significantly with the increase in severity of (p < 0.001). Significant increase of hemorrhoids and anorectal varices was found with the increase in grading of esophageal varices with tendency to more significant increase in frequency of anorectal varices (2 = 9.3 p < 0.01). Anorectal varices also showed significant increase with the increase in the number of sessions of sclerotherapy (2 = 9.3 p = 5.601 p = 0.07). Finally it was found that, in all three groups, the presence or absence of hemorrhoids has no relation to the incidence of anorectal varicea (.x? = 1.3~ < 0.05).

Introduction between anorectal varices and hemorrhoids ANOREXTAL varices are portosystem- are the subject of much debate [t] Thomp- ic collaterals which develop in patients son [3J defined hemorrhoids as vascular with and can occur in cushions that had no direct connect&t the , and/or external anal to the portal . Although anorectal va- margin (11. The etiology of and relation rices and hemorrhoids were believed by

317 318 Ali Gaballa, et al.

Goligher [4] to be one and the same dis- lation profile and abdominal ultrasonogra- order, the concept of being separate enti- phy. After establishment of these parame- ties’ has been enunciated in the work of ters, patients were classified in accordance Goenka et al in 1991 [S]. In fact it is with Pugh’s modification of Child’s classi- very important to distinguish between rec- fication to assess the severity of their liver tal variceal bleeding and bleeding from he- disease [T. During upper gastrointestinal morrhoids when they co-exist because , the presence of esophageal va- their treatment differs, and resection of va- rices and their grading [8] as well as the rices mistaken as hemorrhoids can be dis- gastric mucosal changes were recorded. astrous [6$ In the present study we aimed Proctosigmoidoscopic examination was to establish the prevalence of anorectal va- done for all patients after anal inspection rices and hemorrhoids in patients with liv- and per-rectal examination using the fibre- er cirrhosis, and to correlate their presence ,optic sigmoidoscopy. Anorectal varices with the grading of esophageal varices and were identified using the criteria of Hosk- with the severity of liver disease. The ef- ing et al [l] outlined in table (1). fect of different therapeutic modalities viz Combined Peri and intra variceal injec- sclerotherapy and devascularization on the tion technique was used for bleeding incidence of anorectal varices and hemor- esophageal varices. It aimed at achieving rhoids was also studied sclerosis of the varices and the inner Material and Methods esophageal wall [9]. Ethanolamine oleate One hundred and thirty five patients was the sclerosant used in all patients. The with portal hypertension presenting to the treatment was repeated at weekly intervals gastrointestinal endoscopy and surgical until the varices were largely obliterated. departments of Kasr El Eini Hospital were To achieve this 3 to 6 sessions were neces- included in the study. They were 47 fe- sary. males and 88 males with a mean age of Results 40.56 f 8.6. All patients’had hepatic cir- rhosis and portal hypertension. They were Patients were classified into 3 groups: evaluated clinically and biochemically. Group A: 55 patients (22 females & 33 Clinical details of the patients includ- males) mean age 40 f 8.1 yrs. They had no ed history of upper or lower gastrointesti- history of upper gastrointestinal bleeding nal bleeding. Treatment received for gas- but all had esophageal varices on endosco- tro intestinal bleeding or anorectal disease PY. was recorded Group B: 70 patients (23 females &

Laboratory investigations ‘included full 47 males) - mean age 40.52 f 8.96 years. blood picture, liver function tests, coagu- They underwent endoscopic injection Anorectal Varices in Liver Cirrhosis 319 sclerotherapy for bleeding esophageal va- males) mean 32.9 f 4.77 years. They un- rices. derwent splenectomy vasoligation. The re- sults of all patients were shown in tables Group C: 10 patients (2 females & 8 (2, 3, 4, 5 & ) and Fig (1).

Table (1): Criteria for Diagnosing Lower Gastrointestinal Varices.

Site of varix Method of Features examination

External anal Inspection Perianai swelling, comptressible, which refills 2 seconds of pressure release Internal anal Sigmoidoscopy Blue / State-grey coloured or saccular swellings Colorectal Sigmoidoscopy Dilated tortuous vessels.

Table (2): The Prevalence of Hemorrhoids and Anorectal Varices in the three Groups of Patients.

Vascular Group A p GroupB P Group C P Lesions 55 pts 70 pts 10 pts

Nb 22 24 4 Hemorrhoids > 0.05 > 0.05 > 0.05 % 40 34.29 40 42 5 Nb 26 Anorectal < 0.005 > 0.05 > 0.05 varices % 47.27 Sig 60 50

Nb 7 4 1 Free % 12.73 5.71 10 320 AIi Gaballa, et al.

Table (3): The Relation of Hemorrhoids and Anorectal Varices in all Patients to Modified Child’s grade of Liver Disease.

Vascular Lesions Grade A Grade B&C P

Nb 11 39 c 0.001 Hemorrhoids % 22 78 Sig

Anorectal Nb 14 58 < 0.001 Vtiices % 19.44_ 80.56 Sig

Table (4): Relation of Hemorrhoids and Anorectal Varices to Grades of Esophageal Varices.

Vascular Lesions Grade 1 Grade II Grade III Grade IV Varices 5 pts Varices 17 pts Varices 19 pts Varices 14 pts

Nb 1 5 7 9 Hemorrhoids % 20 29.41 36.84 64.29 0 4 ‘10 12 Anorectal Nb VaXkXS % 0 23.35 52.63 85.71

n 2 = 9.3 p < 0.01 Sig.

Table (5): Relation of Hemonhoids and Anorectal Varices to Number of Sclerotherapy Sessions.

Vascular Sessions 4 Sessions 5 Sessions 6 Sessions Lesions 14 pts 34 pts 16 pts 6pts

Nb 5 4 3 2 Hemorrhoids 35.71 % 7 41.18 18.75 33.3 Anorectal Nb 19 10 5 VatkW % 50 55.88 62.5 83.3 n 2 = 5.601 p < 0.01 Sig. Anorectal Varices in Liver Cirrhosis 321

Table (6) The Prevalence of Anorectal Varices in Patients with and without Hemorrhoids Dis- ease.

Patients Vascular Grade A Grade B Grade C With lesions 22 pls 24 pts 4 P& hemorrhoids Prevalence Nb 12 16 1 of anorectal % 54.54 66.67 25 varices

Patients Vascular Grade A Grade B Grade C Without lesions 33 pts 46 pts 6 pts hemorrhoids Prevalence Nb 14 25 4 of anorectal % 42.42 54.35 66.66 varices

L X = 5.601 p > 0.01 Non Sig.

Discussion

Few data are available as to the preva- lence of colonic disease in cirrhotics. The prevalence of rectal varices and hemor- rhoids in cirrhotic subjects is controversial [IO]. Although Jacobs et al [ll] report 0% and 28% prevalences for rectal varices and hemorrhoids respectively, Hosking et al in 1989 [l] report 44% and 63% preva- lences of these two lesions. Other authors report rectal varices in 3% of cirrhotic pa- tients [12] compared with 0.07% of the general population 1131. Rabinovitz et al in 1990 [lo] reported 3.6% and 25.2% for rectal varices and hemorrhoids respectively. In the present study the prevalence of he- Fig. 1: Colorectal varices seen by morrhoids was 40% in patients with no flexible sigmoidoscopy. 322 Ali Gaballa, et al. history of upper gastrointestinal hemor- Other investigators [l, 161 showed that rhage (group A), 34.29% in patients who the prevalence of anorectal varices had no underwent injection sclerotherapy (group relation to Childs Pugh’s grading. B) and 40% in patients who underwent In the present study all patients had splenectomy vasoligation (group C). Re- liver cirrhosis and portal hypertension. The garding the prevalence of anorectal varices, pathogenesis of hemorrhoids and/or rectal : it was 47.27% in group A, 60% in group varices would seem to be dependent on the B & 50% in group C (Table 2). These re- venous drainage of the rectum and anal ca- sults denote higher prevalence of anorectal nal. Varices occur where portosystemic varices in patients with obliterated esopha- are present, however, colonic geal varices following sclerotherapy @ < varices are less common than esophageal UO5). Hosking et al [l] and later Goen- varices [l, 171. The relationship between ka et al [5l found no difference in the in- portal hypertension, esophageal varices, cidence of anorectal varices in patients rectal varices and hemorrhoids has been with freshly diagnosed esophageal varices widely studied. Doberneck and Janovski and those whose varices had been obliter- I181 and later Izsak and Finlay [19] re- ated by sclerotherapy. On the other hand ported that portal hypertension was found Foutch and Sivak [14] and later Keane in three fourths of cases with rectal varices, and Brilton in 1986 [15] reported the oc- however, he showed no association be- currence of massive bleeding from colorec- tween the presence or absence of rectal va- tal varices after endoscopic injection scler- rices and the degree of portal hypertension. osis of esophageal varices. They postulated Also Goenka et al. 151 found no correla- the hypothesis that lower GIT varices are tion between anorectal varices and esopha- more prevalent and may bleed after esopha- geal variceal sore, portal congestive gas- geaI variceal sclerotherapy. Our results tropqthy or history of esophageal variceal support this hypothesis. bleeding. Wang et al. [16] obtained the The results of our study also showed same results. In contrast Hosking et al. significant increase in the prevalence of [l] showed that the prevalence of anorec- both hemorrhoids and anorectal varices ta1 varices increased with the length of his- with the increase in the severity of liver tory of portal hypertension. They stated disease @ < 0.001) as shown in table (3). that in the presence of portal hypertension, These results are different from those of esophageal varices usually but not invaria- Rabinovitz et al [lo] who showed a sig- bly develop before anorectal varices while nificant trend for the presence of hemor- hemorrhoids occur in the absence or pres- rhoids but not for anorectal varices with ence of esophageal varices. The results of the increasing degrees of disease severity. our study showed that the prevalence of Anorectal Varices in Liver Cirrhosis 323

hemorrhoids and anorectal varices in- changing significantly @ c 0.05 table 1) creased significantly with the esophageal as compared to patients of group A who score with tendency to more significant in- had no obliteration of their esophageal va- crease in anorectal varices than in hemor- rices. rhoids (x2 = 9.3 & p value 0.01) as Accurate differentiation between ano- shown in table 4. rectal varices and hemorrhoids is essential When we correlated the hemorrhoids for proper treatment. The results of our and anorectal varices to the number of scle- study showed that whether the patient is rotherapy sessions we found that there is bleeding (group B-C) or not (group A) significant increase of the anorectal varices the incidence of anorectal varices did not with the increase in the number of session differ significantly in patients having he- with _K’value of 5.601 &p value of 0.07 morrhgids from those not having hemor- (as shown in table 5). This indicates that rhoids x2 = 1.3 P value > 0.05 (table 6) the severity of esophageal varices and the i.e. presence or absence of hemorrhoids had completion of ablation is accompanied no relation to the prevalence of anorectal with higher incidence of anorectal varices. varices. Bleeding from anorectal varices This can be explained on hemodynamic although infrequent, has been reported by changes in the portal bed following oblit- many authors [22, 23, 6 & S] and it is eration of the varices. Fergali and Kamel very important to distinguish between rec- I201 studied these hemodynamic changes tal variceal bleeding and bleeding from he- after sclerotherapy by duplex sonography morrhoids when they co-exist. Banding, and reported an increased in size of portal submucosal injection cryosurgery and re- vein, splenic vein and superior mesenteric section are the modalities of treatment used vein with increase cross section, dimi- for hemorrhoids while underrunning su- nished velocity and reverse flow in the su- tures [6J, sclerotherapy 1221 and portos- perior mesenteric vein. This is supported ystemic shunts 1241are used for anorectal by the study on the pathogenesis of hemo- varices. Hosking and Johnson [6] showed dynamic changes in the portal bed that was that resection of varices mistaken as he- suggested by Khayri in 1966 [tl]. He morrhoids can occasionally be disastrous. postulated the possibility of the shift of Recently, Heaton et al in 1992 (251 the portal blood from gastroesophageal showed that injection sclerotherapy is sat- communication to other sites of the portos- isfactory for symptomatic anorectal varices. ystemic collaterals. On the other hand the According to the above findings$ve group of patients studied after surgical ab- can conclude that awareness of the preva- lation (group C) had the incidence of their lence of anorectal varices is important in hemorrhoids and anorectal varices not patients with liver cirrhosis. Their presence 324 Ali Gaballa, et al. can be a useful bedside examination to rhoids. Br. J. Surg., 62: 542-52, 1975. support variceal etiology for upper gas- 4. GOLIGHER J.C.: Surgery of the anus, rec- trointestinal hemorrhage. The relationship tum and colon 5th ed. London: Balliere ’ to esophageal varices, sclerotherapy paren- Tindall, 98-99, 1985. chymal liver disease and possible coexis- tence with hemorrhoids may alert the clini- 5. GOENKA. M. K, KOCHHAR R, NAGI B cian to the proper diagnosis in cirrhotic & MEHTA S.K.: Rectosigmoid Varices patients especially tliose presenting with and other mucosal changes in patients gastrointestinal hemorrhage. with portal hypertension. Am. J. Gastroent g-1185-89, 1991. However, there remains two important 6. HOWKING S.E. & HOHNSON A.G.: questions: c Bleeding anorectal varices - a misunder- (a) The possible variation in the pathogen- stood condition. Surgery, 104-70-73, esis of hemorrhoids and andrectal va- . 1988. rices in portal hypertension. 7. PUGH RNH, MURRAY - LYON IM, DARWSON JL PIETRONI MC, WIL- (b) The low incidence of bleeding from LIAMS R.: Transection of the anorectal varices as compared to the for bleeding esophageal varices. Br. J. high incidence of bleeding from Surg., 60: 646-649, 1973. esophageal varices, possibly due to anatomical variation of the mucosa of 8. BUSET, M., DES MAREZ, B., BAIZE M., both sites, alkalinity versus acidity BOURGEOIS N.anh CREMER M.: Bleed- (erosive theory) respectively or perio- ing esophagogastric varices: An endoscop- dicity of the motility of the rectum ic Study. Am. J. Gastroenterod., 82, NO. 3: versus continuous motility even at 241-244, 1987. the rest of the esophagus. 9. SOEHENDRA, N., DE HEER K., KEMP- References ENERS J., and RIJNGE M.,: Sclerothera-

HOSKING S.W., SMART HI, JOHNSON py of esophageal varices: acute arrest of A. G, & TRIGER D.R.: Anorectal varice, gastrointestinal haemorrhage or long-term hemorrhoids and portal hypertension. therapy. Endoscopy, 15, 136, 1983.

Lancet, I-349-52, 1989 10. RABINOVITZ M, SCHADE R. R., WATTS G.T.: Piles are caused by varicose DINDZANSVJ, BELLE S, THIEL D & veins in the anal canal. Lancet-I-144-45, GOVALER J.: Colonic disease in cirho- 1986. sis. , 99-195-99, 1990.

THOMSON W. H.: The nature of hemor- 11. JACOBS D.M., BUBPICK MP, ONSTAD Anorectal Varices in Liver Cirrhosis 325

G, HITCHCOCK R.: The relationship of Dis. 15: 834-841, 1970. hemorrhoids to portal hypertension. Dis. 19. IZSAK E. M. and FINLAY J.M.: Colonic Colon Rectum, 23: 567-59, 1980. Varices. Three case reports and review of 12. BRILTON R. C.: Influence of portal sys- the literature. Am J. Gastroenterol., 73: temic collateral patterns and distribution 131-136, 1980. of varices on results of surgical treatment of bleeding esophageal varices. Surgery, 20. FERGALI A and KAMEL N.: Effect pf 53: 567-574, 1963. sclerotherapy on the portal ciculation us- ing Duplex scanning in portal hyperten- 13. FELDMAN M, SMITH VM, WARNER sion. Monufia Medical Journal, Vol. 2-3 CG.: Varices of the colon. Report of three 189-194, 1992. cases J.A.M.A., 179: 729-30, 1962. 21. KHAYRI M.: Splenotomy devascularisa- 14.. FOUTCH PC and SIVAK MV.: Colonic tion operation for the treatment hepatic variceal hemorrhage after endoscopic in- fibrosis. Kasr El Eini J. surg., 7:37-42, jection sclerosis of esophageal varices: A 1966. report of three case. Am. J. Gastroenterol., 79: 756.60, 1985. 22. WANG M., DESIGAN G, DUNN D.: En- doscopic sclerotherapy for bleeding rec- 15. KEANE RM and BRILTON DC.: Massive tal varices. A case report. Am. J. Gastroen- bleeding from rectal varices following re- terol., 80: 779-80, 1985. peated injection sclerotherapy of esopha-

geal varices. Br. J. Surg., 73-120, 1986. 23. WELNSHEL E, CHEN W, FAL KEN-

16. WANG TP, LEE FY, TSAI Y T, LEE SD STEIN D. B . et al.: Hemonhoids or rectal WANG S. S., HSIA H. C. et at..: Relation- varices: defining the cause of massive he- ship of portal pressure, anorectal varices morrhage e portal hypertension. Gas- and hemorrhoids in cirrhotic patients. J.