Downloaded from www.ajronline.org by 177.241.47.175 on 10/20/15 from IP address 177.241.47.175. Copyright ARRS. For personal use only; all rights reserved

1217

Rupture of the Bowel After Blunt Abdominal Trauma: Diagnosis with CT

Stuart E. Mirvis1 David R. Gens2 Shanmuganathan1

Kathirkamanathan

OBJECTIVE. blunt

trauma

resulting

The accuracy is

well

from blunt trauma

sensitivity

of CT in the detection

established,

SUBJECTS

AND

the

is controversial.

of CT in diagnosing

were obtained

but

METHODS.

Dunng

of injuries of

CT

This study

posttraumatic

in 16 patients

value

of the solid viscera

in

diagnosing

bowel

was conducted

after rupture

to determine

the

bowel rupture.

a 51-month

who subsequently

period,

17 preoperative

CT scans

had bowel ruptures verified

surgically.

Both preoperative (prospective) and retrospective CT findings were analyzed in these patients. Retrospective interpretation was made by consensus of two radiologists. RESULTS. Surgically confirmed bowel ruptures occurred in the duodenum (five), ileum (four), jejunum (four), colon (four), and stomach (two). CT findings considered diagnostic of bowel perforation were detected prospectively on 10 (59%) of 17 scans; these included pneumoperitoneum without prior peritoneal lavage (six), mesenteric, intramural, or retropentoneal

or extravasation tive of bowel intraperitoneal gross anterior

free

air (six),

or direct

visualization

of discontinuity

of the

bowel

wall

of luminal contents (four). Prospective CT findings considered suggesrupture were present on five (29%) of the 17 scans; these included fluid of unknown source (three), thickened (>4-5 mm) bowel wall (two), pararenal fluid without a recognized source (one), and a mesenteric-

bowel wall hematoma

(one). On two of 17 scans, findings were seen in retrospect

only;

these included free intraperitoneal blood without a source (findings on a second CT scan were diagnostic) and pneumopentoneum. CT findings diagnostic or suggestive of bowel injury were detected prospectively on 15 (88%) of 17 scans and were noted in all retrospectively. CONCLUSION. CT is sensitive for the diagnosis of bowel rupture resulting from blunt

trauma, but careful inspection AJR

Received April 6, 1992; May 22, 1992.

accepted

Presented at the third annual American Society of Emergency Antonio, TX, March 1992.

after

revision

meeting of the Radiology, San

I Department of Diagnostic Radiology, University of Maryland Medical Center, 22 5. Greene St., Baltimore, MD 21 201 . Address reprint requests to

S. E. Mirvis. 2

Shock

Medical

Trauma

Center,

Center,

University

22 5. Greene

St.,

21201. 0361 -803X/92/1 596-1217 © American Roentgen Ray Society

of Maryland Baltimore,

159:1217-1221,

and technique

December

are required to detect often subtle findings.

1992

The sensitivity of CT in the detection of bowel rupture remains to be established. Numerous earlier studies [1 -6] found poor sensitivity of CT for detection of bowel injuries; others [7, 8] have claimed high sensitivity of CT for diagnosing bowel injuries and for distinguishing those bowel injuries that are likely to require surgical intervention. The discrepancy in these observations may arise from the type of study (prospective vs retrospective interpretation), the sophistication of the CT equipment used, the expertise of the interpreters, the types of bowel or mesenteric injuries evaluated in the study (e.g., bowel rupture, serosal tear, mesenteric injury, ischemic injury, bowel wall contusion), the oral administration of contrast material or the lack thereof, and the performance of diagnostic peritoneal lavage before CT. In light of this controversy, we analyzed both prospective and retrospective interpretation of CT scans of patients with surgically proved bowel rupture identified from our trauma registry [9].

MD

Subjects During confirmed

and Methods a 51 -month bowel

rupture

period,

from

resulting

July

1987

from blunt

to

October

abdominal

1991

trauma

,

48

patients

that required

had

surgically

surgical

resec-

MIRVIS

Downloaded from www.ajronline.org by 177.241.47.175 on 10/20/15 from IP address 177.241.47.175. Copyright ARRS. For personal use only; all rights reserved

1218

ET

AL.

AJR:159, December

1992

Fig. 1.-Jejunal perforation in 17year-old involved in motorcycle accident. A, CT scan shows intraperitoneal fluid below diaphragm on both sides and minimal pneumoperitoneum (arrow). B, CT scan through mid abdomen shows intraperitoneal free fluid and markedly thickened loops of proximal portion of small bowel (arrows). Jejunal transection was confirmed at surgery. (Reprinted with permission from Mirvis and Shanmuganathan [10].)

tion.

Patients

with

of the serosa),

ischemic

bowel,

degloving

bowel

injuries

(avulsion

or isolated

mesenteric injuries were not included in this study. Among these 48 patients, 1 6 (33%) had preoperative CT, which was performed before diagnostic peritoneal lavage in all cases.

These 16 included

14 men and two women

16-50

years old (mean,

26 years). Mechanisms of injury included automobile accidents (1 1), motorcycle accidents (two), and crushing injuries (three). CT scans were obtained by using a Siemens DRH (July 1 987 to March 1 990) with a routine 3-sec scan time or a Siemens Hi-Q (March 1 990 to October 1991) with a routine 2-sec scan time (Siemens Medical

Systems,

were obtained generalized

Iselin,

increase

trauma during this to the area where contrast material Pharmaceuticals, (Squibb Diagnostics,

in the

the scan

they arrived

administered

used, contrast ml of 60% ml).

When

of the

and

rectally

material

Renografin Hi-Q

of

CT

a second

dose

at the scanning in these

cases.

was given (Squibb

by a drip infusion the

use

preoperative

CT

later in the study for

scans

(14)

because

evaluation

of

of a

abdominal

and because of the proximity of the scanner patients were admitted. All patients received oral consisting of 5 g of Hypaque powder (Winthrop New York, NY) or 1 0 ml of 37% Gastrografin Princeton, NJ) in 10 oz (300 ml) of water 30-45

tube when

followed

Most

period

mm before not

NJ).

with the Hi-Q scanner

scanner

either

or by nasogastric

Contrast

material

When

the

scanner

DRH

IV as a hand-injected

Diagnostics)

of 30% was

orally

suite.

before

contrast

used,

contrast

bolus

scanning;

material

was

of 100 this

(average,

material

was

was

was

200 given

IV as a power-injected bolus of 100 ml of 60% Hypaque (at 2 mI/sec) before scanning; this was followed by an infusion of an additional 50-1 00 ml of 60% contrast material at a rate of 0.7-1 .0 mI/sec. Preoperative interpretations of the CT scans were compared with surgical findings. In addition, CT scans were reviewed by two of the authors in all surgically proved cases of ruptured bowel, and CT findings were compared with the initial interpretations and surgical findings.

wall discontinuity in three (Fig. 5), with extraluminal extravasation of contrast material (Fig. 2) or feces (Fig. 4). Many additional supporting findings were observed in these 10 patients. These included bowel wall thickening in four (Fig. 1), free intraperitoneal fluid in four (Figs. 1 and 3), and anterior pararenal or intramesenteric fluid in three (Figs. 2 and 5). CT findings regarded as suggestive of bowel rupture were noted in another five of the 1 7 scans on prospective interpretation, including diffuse thickening of the bowel wall in two (Fig. 6), gross anterior pararenal fluid without a known source in two (Fig. 7), and transmural bowel wall hematoma in one (Fig. 8). Two findings were not made on prospective interpretation but were found retrospectively. One patient had minimal free pelvic fluid of low attenuation without a recognized source; a subsequent preoperative CT scan obtained 44 hr after the initial study, because of the patient’s increasing abdominal pain and fever, showed findings diagnostic of bowel injury, including pneumoperitoneum and free intraperitoneal fluid (Fig. 9). Initially missed CT findings in another patient included pneumoperitoneum and perigastric omental gas. This patient had several obvious major injuries of the solid viscera that necessitated urgent laparotomy. Overall, diagnostic or suggestive findings of bowel rupture were observed prospectively in 1 5 (88%) of 17 CT studies and in all cases retrospectively. None of the patients who were exammed by CT for blunt abdominal trauma during the study (2237 scans obtained) and who had CT findings that were diagnostic of bowel rupture were successfully managed without surgery, and in all these cases bowel rupture was verified surgically.

Discussion Results Nineteen surgically confirmed bowel ruptures were detected: five to the duodenum, four to the jejunum, four to the ileum, four to the colon, and two to the stomach. Preoperative CT findings considered diagnostic of bowel rupture were noted in 1 0 (63%) of these 1 6 patients. These included pneumoperitoneum without an intrathoracic source or previous peritoneal lavage in six (Fig. 1) [1 0]; extraperitoneal gas in the mesentery (Fig. 2), bowel wall (Fig. 3) [1 1 ], or retroperitoneum (Fig. 4) in six; and direct visualization of bowel

Several studies on the accuracy of CT for detecting bowel rupture in patients undergoing exploratory laparotomy have reported poor sensitivity. Cook et al. [6] evaluated CT results in 83 patients with upper abdominal trauma. In three patients with surgically proved small-bowel perforation, the injury was missed on CT, although in one who had duodenal injury, CT findings indicating perforation were seen retrospectively. Two of these 83 patients, in whom CT findings suggested duodenal rupture, had no injury found during surgery (falsepositive). In a subsequent report, Hofer and Cohen [12]

Downloaded from www.ajronline.org by 177.241.47.175 on 10/20/15 from IP address 177.241.47.175. Copyright ARRS. For personal use only; all rights reserved

AJA:159,

December

Fig. 2.-Duodenal

CT

1992

perforation

in 28-year-old

OF

BOWEL

1219

RUPTURE

woman after motor vehicle accident.

A, CT scan through mid abdomen shows extravasation of contrast material posterior to third portion of duodenum (arrows). B, A more caudal CT scan shows one large and several small irregular gas collections in mesentery, fluid dissecting into mesentery, distended small bowel. C, CT scan through upper pelvis shows still further caudal dissection of mesenteric gas (arrow).

Fig. 3.-Duodenal rupture in a 16-year-old with blunt trauma. CT scan shows thickening of duodenal wall and a dot of gas in wall (arrow). Gas extends into wall of gallbladder. Gross intraperitoneal free fluid and pneumoperitoneum are seen. (Reprinted with permission from

Mirvis and Dunham [11]. © 1992, the Williams Co., Baltimore.)

Fig 5.-Duodenal

disruption

in 18-

year-old man involved in motor vehicle collision. A, CT scan through mid abdomen shows perforation of posterior duodenal wall (arrow) and obvious fluid in anterior pararenal space. B, More caudal CT scan shows marked extravasation of fluid in ante-

nor pararenal

space and abrupt

inter-

ruption in continuity of third portion of duodenum (arrow). Perforations of second and third portions of duodenum

were surgically confirmed.

& Wilkins

Fig. 4.-Cecal

perforation

in a 40-year-old

who sustained

crushing

and fluid-filled

injury.

A, CT scan through upper abdomen shows retroperitoneal gas (arrowheads) and pneumoperitoneum (arrows). B, CT scan through lower abdomen shows disruption of cecum, with extrusion of feces into right pericolic gutter (arrows). Cecal perforation was confirmed surgically. (Reprinted with permission from Mirvis and Dunham [11]. © 1992, the Williams & Wilkins Co., Baltimore.)

Downloaded from www.ajronline.org by 177.241.47.175 on 10/20/15 from IP address 177.241.47.175. Copyright ARRS. For personal use only; all rights reserved

1220

Fig. 6.-Possible jejunal rupture in a 17-yearold after crushing injury. CT scan shows marked thickening of proximal jejunum (arrowheads). Some free intraperitoneal fluid (not shown) also was observed. Proximal jejunal rupture was found at surgery. (Reprinted with permission from Mirvis and Dunham [1 1]. © 1992, the Williams & Wilkins Co., Baltimore.)

MIRVIS

ET AL.

AJR:159,

December

1992

Fig. 7.-Possible duodenal rupture in 32-year-old woman involved in motor vehicle collision. A, CT scan shows marked retroperitoneal (anterior pararenal) fluid and narrowing of third portion of duodenum (arrows). Some intraperitoneal fluid is present. B, More caudal CT scan suggests abrupt interruption of third portion of duodenum (arrow). At surgery, ruptures of second and third portions of duodenum were found. (Reprinted with permission from Mirvis and Dunham [11]. © 1992, the Williams & Wilkins Co., Baltimore.)

Fig. 8.-Possible bowel rupture in a 20-year-old after motor vehicle collision. CT scan shows large heterogeneous mass in right lower quadrant of abdomen. Diagnostic considerations included bowel or mesenteric hematoma. At surgery, a large cecal transmural hematoma and perforation of ascending colon were found, which required segmental resection. (Reprinted with permission from Mirvis and Dunham [11]. © 1992, the Williams & Wilkins Co., Baltimore.)

Fig. 9.-Delayed

diagnosis

of bowel

perforation

in 50-year-old

injured

in a motorcycle

accident. A, Initially findings on CT scan were interpreted as normal. On retrospective review, a small amount of free intraperitoneal fluid (arrow) was recognized in pelvis. B, CT scan obtained 2 days after A, because patient had abdominal pain and fever, shows marked pneumoperitoneum and free intraperitoneal fluid. At surgery, a ruptured jejunum was diagnosed. (Reprinted with permission from Mirvis and Shanmuganathan [10].)

described CT findings in two patients with duodenal perforation, including free intraperitoneal or extraperitoneal gas, extravasation of opacified oral contrast material into the right pararenal space, focal thickening of the duodenum in its midportion that prevented passage of contrast material, and high-attenuation intramural duodenal hematoma. Two retrospective series on the accuracy of CT in detecting bowel injuries have been reported. Both series were derived from the experience at the San Francisco General Hospital and include data from overlapping periods [7, 8]. The larger series by Rizzo et al. [8] includes a retrospective review of 51 patients in whom CT findings suggested bowel or mesenteric injuries and compares CT findings with surgical outcome. In this series, CT showed evidence of significant bowel injury in 26 (93%) of 28 patients whose injuries were con-

firmed at surgery. These authors compared CT findings indicative of bowel injury between 32 patients who underwent celiotomy and 1 9 who were managed by observation. Although some overlap of CT findings occurred, patients undergoing laparotomy had a higher prevalence of pneumoperitoneum (32%), free intraperitoneal fluid (96%), and associated injuries (43%) than did patients managed nonoperatively (0%, 21 %, and 5%, respectively). Rizzo et al. [8] emphasized the value of CT in detecting signs of bowel and mesenteric injury and in distinguishing injuries requiring surgery from those that do not. Unfortunately, that study was a retrospective review, and the influence of the CT findings of potential bowel injury on the decision to proceed to laparotomy cannot be ascertained. Although our study included fewer patients, it was a pro-

Downloaded from www.ajronline.org by 177.241.47.175 on 10/20/15 from IP address 177.241.47.175. Copyright ARRS. For personal use only; all rights reserved

AJR:159,

December

1992

CT

OF

BOWEL

spective evaluation of CT for the detection of bowel rupture. We compared CT findings with surgical results specifically In patients with bowel rupture and did not assess the accuracy of CT for diagnosing bowel ischemia related to blunt trauma, avulsion of the serosa, or bowel wall contusion. Our results support those of Donahue et al. [7] and Rizzo et al. [8] in that CT accurately showed bowel rupture in 1 0 of 1 6 patients and suggested major bowel injury in five others. Significant findings on two CT scans obtained at admission were missed prospectively but were observed retrospectively. Our results support the value of CT for showing findings diagnostic of bowel rupture, including pneumoperitoneum without prior diagnostic peritoneal lavage; discontinuity of the bowel wall; extravasation of contrast materials; and ectopic gas in the retroperitoneum, bowel wall, and mesentery in patients with bowel transection from blunt impact. We believe the diagnosis of bowel rupture can be aided considerably by oral administration of contrast material, and we encourage its use for CT evaluation of blunt abdominal trauma. We did not encounter any complications related to oral administration of contrast material during this study. CT findings suggestive of bowel rupture, including thickening of the bowel wall, free fluid in the anterior pararenal space, and intraperitoneal blood or fluid without an obvious source should mandate close scrutiny of findings on physical examination, possibly combined with interval peritoneal Iavage or additional CT scans. In many of our patients, the CT findings in bowel rupture were subtle, and detection required careful inspection of scans. Meticulous attention to scanning technique, including reduction of artifacts, optimal bowel con-

RUPTURE

1221

trast, and appropriate review of images optimized for detection of minimal pneumoperitoneum will improve sensitivity of CT for diagnosis of bowel rupture. REFERENCES 1 . Marx JA, Morre EE, Jorden AC, Eule J. Limitations of computed raphy in the evaluation of acute abdominal trauma: a prospective ison with diagnostic peritoneal lavage. J Trauma 1985;25:933-946

2. Meyer DM, ThaI ER, Weigelt JA, Aedman

3.

4, 5. 6. 7.

HC. Evaluation

tomogcompar-

of computed

tomography and diagnostic peritoneal lavage in blunt abdominal trauma. J Trauma 1989;29: 1168-1172 Kearney PA, vahey T, Burney RE, Glazer G. Computed tomography and diagnostic peritoneal lavage in blunt abdominal trauma. Arch Surg 1989;124:344-347 Keamey PA. Blunt trauma to the abdomen. Ann Emerg Med 1989;18: 1322-1325 Sherck JP, Oakes DD. Intestinal injuries missed by computed tomography. J Trauma 1990;30:1-7 Cook DE, Walsh JW, Vick CW, Brewer WH. Upper abdominal trauma: pitfalls in CT diagnosis. Radiology 1986;159:65-69 Donahue JH, Federle MP, Griffiths BG, Trunkey DD. Computed tomography in the diagnosis of blunt intestinal and mesenteric injury. J Trauma 1987:27:11-17

8. Rizzo MJ, Federle MP, Griffiths BG. Bowel and mesenteric blunt

abdominal

trauma:

evaluation

with

CT.

Radiology

injury following 1989;173:

143-148 9, Dunham CM, Cowley RA, Gens DR, et al. A methodologic approach for a large trauma registry. Md Med J 1989:38:227-233 10, Mirvis SE, Shanmuganathan K. Abdominal CT in blunt trauma. Semin Roentgenol 1992:27: 150-1 83 1 1 . Mirvis SE, Dunham CM. Abdominal/pelvic trauma. In: Mirvis SE, Young JWA, eds. Imaging in trauma and critical care. Baltimore: Williams & Wilkins, 1992:145-242 12. Hofer GA, Cohen AJ. CT signs of duodenal perforation secondary to blunt abdominal trauma. J comput Assist Tomogr 1989; 13 : 430-432

Rupture of the bowel after blunt abdominal trauma: diagnosis with CT.

The accuracy of CT in the detection of injuries of the solid viscera after blunt trauma is well established, but the value of CT in diagnosing bowel r...
877KB Sizes 0 Downloads 0 Views