73

Downloaded from www.ajronline.org by 200.59.56.91 on 11/13/15 from IP address 200.59.56.91. Copyright ARRS. For personal use only; all rights reserved

lleoanal

Pouches:

Comparison of CT, and Contrast Enemas for Postsurgical Complications

Scintigraphy, Diagnosing

1tk

Ruedi F. Thoeni1 Sean C. FelF1 Barry Engelstad’ Theodore B Schrock2

The value of CT of the pelvis, 1111n-Iabeled leukocyte scintigraphy, and contrast enema (pouchography) for detecting postsurgical complications was assessed in 44 patients with total colectomy, rectal mucosectomy, and ileoanal pouches. lleoanal pouches were created as reservoirs from an ileal loop that was anastomosed to the dentate line of the anus and stayed connected to the remainder of the ileum. This pouch preserves the normal defecatory pathway and eliminates disease-producing mucosa. A total of 57 sets of examinations revealed 22 cases of normal postoperative findings, 22 of pouchitis, 13 of abscess, and three of fistula. Overall sensitivity for detecting complications with pouchography was 60% (18 of 30 findings); with CT, 78% (28 of 36 findings); and with scintigraphy, 79% (23 of 29 findings). Pouchftis was best diagnosed by scintigraphy (sensitivity, 80%), followed by CT (sensitivity, 71%) and pouchography (sensitivity, 53%). Only CT correctly diagnosed all cases of abscess. Fistulas were frequently missed by all three methods. If tests were combined, the overall sensitivity rose to 93% for the combination CT/scintigraphy and to 86% for CT/pouchography, but did not improve for pouchography/scintigraphy (78%). For evaluation of complications in patients with ileoanal pouches, CT should be the initial test. If an abscess is found, no further tests are needed. If CT findings are negative, a scintigram should be obtained. Our data did not establish a clear role for

.

pouchography. AJR 154:73-78,

January

1990

For many years, surgical management of patients with ulcerative colitis or familial polyposis of the colon necessitated total proctocolectomy and creation of a permanent ileostomy [1 ]. However, both physicians and patients often resist construction of an ileostomy because these patients usually are young and physically and sexually active. In addition, there is danger of sexual and bladder dysfunction related to total proctectomy. More recently an innovative surgical procedure

was

introduced

that

eliminates

the

problems

encountered

with

the

ileostomy. This operation consists of construction of an ileoanal reservoir associated with total colectomy, rectal mucosectomy, and ileoanal anastomosis [2-5]. Such a procedure is curative and preserves the normal defecatory pathway [6]. Received July 19, 1989; accepted September 8, 1989.

after

revision

These available

Presented at the annual meeting of the American Roentgen Ray Society, New Orleans, May 1989. 1 Department of Radiology, University of California, San Francisco, P.O. Box 0628, San Francisco, CA 94143-0628. Address repnnt requests to A. F. Thoeni.

two important to patients

goals are not fulfilled by the other surgical options that are with ulcerative colitis or familial polyposis. In ileorectal anas-

tomosis the disease-producing mucosa is left behind; with end-ileostomy nent ileostomy the normal defecatory pathway is not preserved [1]. Major

complications

(inflammation . leostomy,

related

of the pouch), . remainder

of an ileoanal and fistula. Other

to creation

abscess,

or conti-

reservoir include pouchitis complications relate to the

.

San Francisco, San Francisco, CA 94143-0144.

bowel, or functIonal disorders of the pouch [1]. This study was undertaken to determine the respective values of CT of the pelvis, 1111n-Iabeled leukocyte study (scintigraphy), and pouchography for diagnosing

0361-803X/90/1541-0073 © American Roentgen Ray society

postsurgical complications without symptoms related

2

Department

of Surgery,

University

of Califomia,

of the small

associated to pouchitis,

with ileoanal pouches in patients with or abscess, or fistula. The value of the three

THOENI

74

ET AL.

For the 43 scintigrams, blood was drawn from the patient, labeled with 0.5 mCi (18.5 MBq) of 111ln, and reinjected into the patient after

radiographic tests before takedown of ileostomy was assessed also. On the basis of our data, an algorithm for the workup of these patients was developed.

Downloaded from www.ajronline.org by 200.59.56.91 on 11/13/15 from IP address 200.59.56.91. Copyright ARRS. For personal use only; all rights reserved

Subjects

.

and Methods

This study included 57 sets of radiographic examinations (CT scan, scintigram, and pouchogram) in 44 patients who underwent total colectomy and construction of an ileoanal reservoir between January

1983 to June 1 989. J pouches

(Fig. 1) were present

in 20 patients

and S pouches in 24 (Fig. 2). The mean age was 39 years (range, 1 6-65). Women (n = 23) and men (n = 21) were equally represented. In 1 0 of 44 patients, multiple sets of examinations were performed

AJR:154, January 1990

2 hr. One patient was imaged 3.5 hr, 12 were scanned 5-1 5 hr, and 30 were scanned 24 hr after reinjection of the labeled blood. Views were obtained in anterior, posterior, lateral, and/or tail-on-detector projections. Twenty of the 48 pouchograms were obtained in a retrograde fashion through the pouch and 28 in an antegrade fashion through the ileostomy.

Early

in the investigation,

antegrade

approach,

we

could

done before and in 21 instances after takedown of ileostomy. The three studies were performed on the same day in 24 cases and were no more than 3 days apart in the remaining sets. CT was always performed first, followed by pouchography and then scintigraphy.

to the

reservoir

in two.

lateral,

and

anteroposterior

An abscess pouchitis

was associated

was

combined

detected in 22 patients, but in 1 3, and a fistula in three.

with severe pouchitis

with

fistulization

in two

in one patient, and other

patients.

obtained

fashion. Late in the investigation,

at various time intervals; the complete set of three tests was performed in all but 23 instances. In 36 instances, examinations were

No postsurgical complications were pouchitis was found in 22, an abscess

retrograde

pouchograms

were obtained in patients examined before takedown Once we realized that the pouch could be distended

not be examined

at the stoma

site

pouchograms

and

in an antegrade

only three patients with an ileostomy

in an antegrade

in one

of ileostomy. better by the

inability

fashion

because

to cannulate

of infection

the ileal

loop

leading

and Hypaque (diatrizoate sodium, Winthrop-Breon, New York, NY) in 35 studies. All patients who underwent all three tests on the same day received Hypaque during the pouchogram. Three to five spot films in various positions and four overhead films (anteroposterior, posteroanterior, Barium

Before the examination,

was

in 1 3

used

postevacuation

films)

1 mg of glucagon

were

was injected

obtained.

IV.

The

presence

or absence of pathologic changes was proved as shown in Table 1 . One patient with a surgically proved normal pouch had a fistula at the ileostomy site. Rectal and/or pelvic pain was seen in 31 patients,

fever

and leukocytosis

in 1 0, diarrhea

in eight,

rectal

dis-

TABLE

1: Pathologic

charge in six, and nausea and vomiting in one patient. No symptoms related to the pouch were present in 18 of 20 patients who underwent the radiographic examinations before ileostomy takedown. For the 55 pelvic CT studies, axial images were obtained on a GE 9800 unit (140 kVp, 1 20 mA, 3 sec) with oral, rectal, and IV contrast

material. Two patients refused rectal contrast material. Ten milligrams of oral metoclopramide (Raglan, Quad Pharmaceuticals, Indianapolis, IN) was given routinely 45 mm before the CT examination to ensure filling

of distal

small

bowel

loops

[7]. A total

of 1 100 ml of oral and

approximately 150-200 ml of rectal contrast agents was administered. Dynamic scans were used in 1-cm increments starting at a level 3 cm below the anal verge and extending to the iliac crests. This low starting point ensured visualization of low fistula tracts or leaks from the anastomosis at the dentate line.

Fig. 1.-J

pouch.

A, cT scan clearly shows stump (arrows) B, Pouchogram shows typical blind-ending

with surgical stump

staples.

(arrows).

Changes

in Patients

with lleoanal

Reservoirs Pathologic

Proof

None Pouchitis Surgerya Aspiration biopsy with or without drainage Endoscopy of pouch Clinical course uneventful months

Resolution

of symptoms

Changes Abscess

Fistula

20

3

6

2

0 1

0 3

7 0

0 0

1

0

0

0

up to 3

following

0 16 0 1 antibiotic treatment Note-Fifty-seven sets of examinations were performed in 44 patients. a

Careful

analysis of pouch during

takedown

of ileostomy

or corrective

surgery.

Fig. 2.-Round configuration of pouch in absence of blind-ending stump on CT scans above this level suggests presence of S pouch. Inhomogeneity

of pouch

is caused

by retained

stool.

POSTSURGICAL

AJA:154, January 1990

All examinations

were

were

interpreted

several

months

COMPLICATIONS

after the studies

TABLE

performed.

Pouchograms and CT scans were analyzed in conference by two of the authors who were unaware of the final results and clinical histories; the scintigrams were analyzed by one radiologist

with special

Downloaded from www.ajronline.org by 200.59.56.91 on 11/13/15 from IP address 200.59.56.91. Copyright ARRS. For personal use only; all rights reserved

to the other findings.

expertise

in nuclear

Because

medicine

all participating

wall,

and with or without

wispy

densities

in

the region around the pouch; and scintigram-well-defined area of increased uptake posterior to or to the side of pouch. Anterior uptake was

interpreted

as representing

cystitis.

Radiographic criteria for the diagnosis of fistula were as follows: pouchogram-tract of barium or Hypaque extending from the pouch into the soft tissue surrounding the pouch; CT scan-tract of contrast material extending from the pouch, usually associated with increased densities in fat surrounding the tract; and scintigram-area of in-

creased uptake resembling abscess. Results of scintigraphy were accepted as correct if increased uptake of the isotope was present, regardless of whether it was caused by pouchitis, abscess, or fistula. Scintigrams were considered true-negative if no increased uptake was present in the pelvis. In a separate scintigraphic analysis, confidence levels (1 definitely not possible to 6 definitely possible) for =

POUCHES

2: Sensitivity

Pouchitis,

75

of CT, Scintigraphy,

Abscess,

=

distinction between pouchitis and abscess or fistula were determined on the basis of the location of the increased uptake. The sensitivities of the various tests and their combinations were determined. False-positive, true-negative, and false-negative results

also were analyzed.

Results Among 32 sets of examinations in patients with S pouches, only three sets (9.4%), one of which did not include CT, were normal. Among 25 sets of examinations in patients with J pouches, 1 9 (76%) were normal. One patient with a normal S pouch had an anal stricture that needed operative dilatation. This stricture was not seen on CT or scintigraphy, and the patient refused pouchography because of extreme discomfort. The overall sensitivities for detecting any complication and the respective sensitivities for pouchitis, abscess, and fistula are listed in Table 2. In two (1 0%) of 20 sets of studies obtained before ileostomy takedowns, unsuspected complications occurred that prevented takedowns planned for the day after the radiographic examinations. In one of these two patients, a small abscess was present; in the other, a small fistula extended from the anal anastomosis. The overall sensitivity for detecting postsurgical complications in ileoanal pouches increased to 93% if CT and scintigraphic results were combined and to 86% if CT and pouchographic results were combined. However, the combination of the scintigram and pouchogram did not improve the overall

and Pouchography

in

and Fistula

Two-Positive/True-Positive Pathology

were

follows: pouchogram-mass impressing pouch with or without air bubbles, spiculations, and thickening of folds, particularly in the area of the impression; CT scan-soft-tissue mass adjacent to pouch with or without enhancing rim, with or without air bubble(s), with or without of the pouch

ILEOANAL

Detecting

who was blinded

radiologists

involved in the initial patient examination, reinterpretations might not have been completely blinded. Radiographic criteria for the diagnosis of pouchitis were as follows: pouchogram-spiculation, thickening of folds, and spasm of the pouch; CT scan-thickened pouch wall (more than 3 mm) with or without wispy densities in the area around the pouch; and scintigram-increased uptake in the suspected location of the pouch (midline). Radiographic criteria for the diagnosis of abscess were as

thickening

WITH

+ False-Negative (%)

CT

-

Pouchitis

1 5/21

Abscess

12/1 2 (1 00)

Fistula

Total Note.-Numbe

Pouchography

Scintigraphy (71)

1 /3

(33)

28/36

(78)

1 6/20

6/7 1/2

(80)

1 0/1 9 (53)

(86)

7/8

(88)

(50)

1/3

(33)

23/29 (79)

18/30 (60)

rs relate to findings and not studies.

sensitivity of the tests (78%). If increased uptake was present on scintigraphy, CT was helpful to determine whether this abnormality represented an abscess (n = 6). In four patients with pouchitis, scintigraphy alone was positive. Of the 59 CT findings, 22 were true-negative, eight were false-negative (six for pouchitis and two for fistula), and one was false-positive (for pouchitis 9 days after construction of the ileoanal pouch). Of the 48 scintigraphic findings, 1 8 were true-negative and six were false-negative (four for pouchitis, one for abscess, and one for fistula). Only one false-positive finding of pouchitis was found on scintigraphy; this was in a patient with cystitis and posterior position of bladder. Scintigraphy diagnosed six abscesses as pouchitis only, with a confidence level of 5 in five patients (one with abscess and pouchitis) and a confidence level of 4 in one patient. Pouchitis was diagnosed on scintigraphy with confidence levels of 5 and 6 each in five patients (62.5%) (one patient also had a fistula) and with confidence levels of 3 and 4 each in three patients. Fistulas were misdiagnosed as pouchitis with confidence levels of 6 in one patient with pouchitis and fistula and of 2 in one patient with fistula alone and a small amount of increased uptake. Of the 53 pouchographic findings, 20 were true-negative and 1 2 were false-negative (nine for pouchitis, two for fistula, and one for abscess). The three false-positive pouchograms

evidence

were

of abscess,

in one

patient

pouchitis,

each

with

pouchographic

and fistula.

Discussion Creation of an ileoanal reservoir is a two-step procedure at the University of California, San Francisco. The initial step consists of total colectomy, rectal mucosectomy, and creation of an ileoanal pouch and a loop ileostomy. During a second operation (usually after 2 months), the temporary ileostomy is taken down, and the pouch is carefully examined while the patient is under anesthesia. Two types of pouches are used: the J pouch and the S pouch (Fig. 3). Our surgeons incise the rectal mucosa immediately above the dentate line and preserve the internal sphincter. A very small rectal stump is left from which the rectal mucosa is stripped. The J pouch is directly anastomosed to the dentate line (Fig. 3A), whereas the S pouch (Fig. 3B) is connected to the dentate line via a 2-cm spout. To our knowledge, very little information is available in the literature regarding the CT appearance of ileoanal reservoirs

THOENI

Downloaded from www.ajronline.org by 200.59.56.91 on 11/13/15 from IP address 200.59.56.91. Copyright ARRS. For personal use only; all rights reserved

76

ET AL.

AJA:154, January 1990

Fig. 3.-A and B, Schematic drawings of J (A) and 5(B) pouches. In J pouch, ileal loop is sutured (or stapled) directly to dentate line. After surgery, pouch frequently rotates so that stump appears on the left (see Fig. 1). S-shaped ileal loop Is used to fashion reservoir by opposing three segments of terminal ileum (open arrows). 5 pouch has long spout (solid straight arrows), which is sutured to dentate line. Suture lines (curved arrows) on posterior wall are drawn as “see-through” effect.

A

B

[8-1 0], and no report is available investigating the comparative values of CT, scintigraphy, and pouchography for postsurgical complications. Early recognition of these complications is necessary to ensure preservation of the pouch and its functions [1 1]. The S pouch reportedly carries a much higher rate of postoperative complications, probably due to incomplete emptying of the pouch and other problems associated with the long distal segment [1 1]. These are the reasons why, more recently, our surgeons have abandoned the S pouch for the J pouch. Unsuspected complications were found in two asymptomatic patients of 20 patients with J pouches examined before takedown of the ileostomy. This fact shows that radiographic evaluation before surgical takedown of ileostomy can be helpful. Delayed takedown of the ileostomy in these patients prevented a protracted course due to complications. In both patients, complications resolved after antibiotic treatment (and a small incision through the posterior wall of the pouch for drainage of an abscess in one). The postoperative course after takedown of the ileostomy was uneventful in both patients. Pouchitis usually is asymptomatic before takedown of ileostomy, but after takedown, patients with pouchitis experience increased watery bowel movements and cramping pain with or without fever. The cause of pouchitis is unknown; it may be due to stasis in the pouch with bacterial overgrowth and, rarely, to previously undiagnosed Crohn disease. Antibiotic treatment usually is effective, and resolution of inflammation after antibiotic therapy is thought to be diagnostic of pouchitis. If pouchitis remains untreated, the pouch may fail and has to be removed. Abscess and fistula usually are due

to leakage at the anal anastomosis, and immediate with antibiotics and drainage for larger abscesses

treatment are nec-

essary.

Our comparison of the three tests (Table 2) shows that CT and scintigraphy provide similar overall results, but the pouchogram is clearly less sensitive, with many false-negative and false-positive results. Although scintigraphy was the most sensitive test (sensitivity, 80%) for detecting pouchitis (Fig. 4), detection of pouchitis was improved if scintigraphy was combined with CT (combined sensitivity, 95%). CT was the best test for detection of abscess (Fig. 5), whereas fistulas (Fig. 6) were missed frequently by all tests. Study of a larger series of patients with fistulas is needed to determine whether fistulas may be reliably detected by one of the three tests or their combinations. Reasons for false-negative CT findings included absence of detectable wall thickening or increased fat densities in the area around the pouch, incomplete scanning through the anal canal, and absence of contrast material in the pouch due to patient refusal of contrast material. On CT, mild wall thickening of the pouch can be obscured by an incompletely distended pouch or by dense contrast material within the pouch. The one false-positive CT result clearly was caused by acute postsurgical changes. For accurate assessment of postsurgical complications, at least 6-8 weeks should elapse after initial surgery. Some of the false-negative scintigraphic findings in pouchitis and abscess may have been due to the fact that some of our patients were examined after several days of antibiotic treatment (n = 2) or due to the presence of barium from a previous pouchogram (n = 1). In some patients with a com-

Downloaded from www.ajronline.org by 200.59.56.91 on 11/13/15 from IP address 200.59.56.91. Copyright ARRS. For personal use only; all rights reserved

AJR:154, January 1990

POSTSURGICAL

COMPLICATIONS

WITH

ILEOANAL

77

POUCHES

Fig. 4.-Pouchitis. A, Pouchogram shows thickened folds and spiculation (S pouch). B, CT scan shows thickened pouch wall (arrows) and increased density

C, “‘In-labeled

in region around pouch. leukocyte scan shows increased uptake indicative of inflammatory process.

.t$..

,_

-,

#{182}

ib!,

Fig. 5.-Pouch with abscess. A, Pouchogram shows increased B, CT scan shows a low-density pouch wall (curved arrow).

space between S pouch and sacrum and a slight mass impression on pouch (arrows). mass with enhancing rim (straight arrows) posterior to pouch, corresponding to an abscess,

C, 1111n-labeled leukocyte study shows increased

and slightly thickened

uptake in area of pouch.

bination of findings, only one abnormality was identified. The separate analysis of scintigraphic results by using confidence levels for diagnosing pouchitis, abscess, and/or fistula showed that scintigraphy could detect only inflammatory changes and not the type of postsurgical complication. On the basis of our study, there appears to be no role for pouchography in the evaluation of ileoanal pouches. However, the value of this test lies in evaluating the functional abnormality of the pouch and small bowel, which cannot be done with the other tests [101. Our study did not investigate defecatory function, stool patterns, or small bowel abnormalities, all of which are dysfunctions that may be evaluated with pouchography [8-10]. Poor results with the pouchogram may

be due to difficulty in distending an unused pouch or to a pouch that can freely evacuate through a short loop of small bowel into the ileostomy. Nevertheless, it is not surprising that more subtle changes of pouchitis, which cause mild wall thickening, are not well detected by pouchography unless spiculation and thickening of folds become clearly evident. It was astonishing to us that the pouchogram fared so poorly in the detection of fistulas (Fig. 6), even though, in each case of fistula, a retrograde study was performed with the tip close to the anal verge. Previous reports have suggested that the pouchogram is of great value for diagnosing fistulas in patients with ileoanal reservoirs [8]. Our results may have been partially due to the fact that we used water-soluble contrast

A

THOENI

Downloaded from www.ajronline.org by 200.59.56.91 on 11/13/15 from IP address 200.59.56.91. Copyright ARRS. For personal use only; all rights reserved

78

Fig. 6.-Fistula. A, Pouchogram shows small fistula tract (arrow) B, CT scan shows small tract (arrows) extending C, 1111n-labeled leukocyte study shows increased

ET AL.

AJR:154, January 1990

extending from anal anastomosis of this 5 pouch. posteriorly from S pouch. High density of tract is caused uptake in lower pelvis.

material and partially due to the fact that the fistulas were very small. Also, many of our patients with complications related to the pouch had great difficulty in tolerating the pouchogram. Valium or pain medication may be helpful in these instances, but such medication was not used in our study. In conclusion, for the evaluation of postsurgical complications of ileoanal pouches in patients with total colectomy, we recommend that CT of the pelvis be performed initially. If an abscess is demonstrated, no further radiographic test is needed. The abscess can be drained by a surgeon or by a radiologist who, under CT guidance, places a drainage tube percutaneously or transrectally through the wall of the ileoanal pouch. Whether or not symptoms are present, if a CT scan is negative, a scintigram should be obtained to rule out pouchitis or fistula. If pouchitis or fistula is detected with CT or the combination of CT and scintigraphy, antibiotic treatment should be instituted. Takedown of the ileostomy must be delayed if an abscess or fistula is present, but it may proceed in the presence of pouchitis. The role of the pouchogram could not be clearly established from our data. If the inflammation persists or is suspected to be caused by pouch dysfunction or unknown Crohn disease, a pouchogram with a postevacuation film may be advisable. A thorough endoscopic examination of the pouch with the patient under anesthesia may be necessary in these cases.

A

by barium

retained

from pouchogram.

REFERENCES

1 . Schoetz

DJ Jr, Coller

JA, Veidenheimer

MC. Altematives

to conventional

ileostomy in chronic ulcerative colitis. Surg C/in North Am 1985;65:21-33 2. Parks AG, Nicholls RJ, Belliveau P. Proctocolectomy with ileal reservoir and anal anastomosis. Br J Surg 1980;67:533-538 3. Johnston D, Williams NS, Neal DE, Axon ATA. The value of preserving the anal sphincter of 22 mucosal

in operations proctectomies.

for ulcerative colitis and polyposis: Br J Surg 1981;68:874-878

a review

4. Utsunomiya J, Iwama T, Imajo M, et al. Total colectomy, mucosal proctectomy, and ileoanal anastomosis. Dis Co/on Rectum 1980;23:459-466 5. Taylor BM, Beart AW, Dozois AR, Kelly KA, Phillips SF. Straight ileoanal anastomosis

versus

mucosal proctectomy.

ileal

pouch-anal

anastomosis

after

colectomy

and

Arch Surg 1983;1 18:696-701

6. McHugh SM, Diamant NE, McLeod A, J-pouches: a comparison of functional 1987;30:671 -677 7. Thoeni AF, Filson RF. Abdominal and pelvic to enhance bowel opacification. Radio/ogy

Cohen Z. S-pouches outcomes. Dis Co/on CT: use oforal

versus

Rectum

metoclopramide

1988;169:391-393 8. Hennild V, Kjaergard H, Hansen LK. Radiologic evaluation of the continent (S-pouch) ileal reservoir with anal anastomosis. Acta Radio/ (Diagn] (Stockh) 1986;27:301 -304 9. Hillard AE, Mann FA, Becker graphic evaluation. Radio/ogy

JM, Nelson

JA. The ileoanal

J-pouch:

radio-

1985;155:591-594 10. Kremers PW, Scholz FJ, Schoetz DJ Jr, Veidenheimer MC, Coller JA. Radiology of the ileoanal reservoir. AJR 1985;145:559-567 11. Linquist K, Nilsell K, Liljeqvist L. Cuff abscesses and ileoanal anastomotic separations in pelvis pouch surgery: an analysis of possible etiologic factors. Dis Co/on Rectum 1987;30:355-359