DOI:10.2214/AJR.07.3386
AJR 2008; 191:743-747
© American Roentgen Ray Society
Utility of CT Whirl Sign in Guiding Management of Small-Bowel Obstruction
Jeremy B. Duda1,
Shweta Bhatt and
Vikram S. Dogra
1 All authors: Department of Imaging Sciences, University of Rochester School of
Medicine, 601 Elmwood Ave., Box 648, Rochester, NY 14642.
Received November 6, 2007;
accepted after revision March 5, 2008.
Address correspondence to V. S. Dogra
(Vikram_Dogra{at}urmc.rochester.edu).
Abstract
OBJECTIVE. The purpose of this study was to examine the relation
between the CT whirl sign and outcome among patients with a clinical and
radiologic diagnosis of small-bowel obstruction (SBO).
MATERIALS AND METHODS. The cases of 453 patients who underwent
abdominal CT because of clinical suspicion of SBO were reviewed
retrospectively. Patients with a radiologic diagnosis of SBO were included.
Management with surgery or medical therapy was correlated with the presence of
the whirl sign and other radiologic findings. Statistical calculations were
performed to determine the value of the whirl sign in predicting the type of
management needed for SBO.
RESULTS. According to CT criteria, 194 patients received a diagnosis
of SBO and were included in the study. The whirl sign was identified on the CT
scans of 40 of the 194 patients. Thirty-two of the 40 patients had SBO
necessitating surgery, for a positive predictive value of 80%; 133 of 154
patients did not need surgery, for a negative predictive value of 86%.
Fifty-three of 194 patients either underwent surgery or died of SBO during
conservative therapy. The whirl sign was present on the CT scans of 32 of the
53 patients, for a sensitivity of 60%. One hundred thirty-three of 141
patients did not need surgery and did not have a whirl sign, for a specificity
of 94%. The odds ratio for the whirl sign in predicting the presence of SBO
necessitating surgery was 25.3 (95% CI, 10.3–62.3).
CONCLUSION. A patient with the whirl sign on CT is 25.3 times as
likely as a patient without the sign to have SBO necessitating surgery. The
results suggest an important role of the whirl sign in assessment of treatment
options for patients with clinical and radiologic signs of SBO.
Keywords: CT • management • radiographic sign • small-bowel obstruction • whirl sign
Introduction
Small-bowel obstruction (SBO) accounts for as many as 16% of surgical
admissions for acute abdomen, and it necessitates urgent evaluation to avoid
potentially severe complications
[1]. If the vascular supply to
the bowel is compromised in closed-loop obstruction due to direct compression
or twisting of the mesentery, ischemia or infarction can result
[2–4].
However, clinical signs are not reliably predictive of the need for surgical
intervention in SBO, and delay in treatment is a major prognostic factor for
survival and morbidity [5,
6].
CT can show bowel with compromised blood flow and has been found to improve
the sensitivity of initial evaluations to determine whether a patient needs
surgical intervention [7]. An
array of CT radiologic patterns have been described that correlate with the
nature and severity of obstruction
[2,
8–10].
To our knowledge, however, there have been no efforts to correlate these signs
with the clinical decision to pursue conservative rather than surgical
treatment.
One classic sign of closed-loop obstruction on CT is the whirl sign, a
swirl of mesenteric soft-tissue and fat attenuation with adjacent loops of
bowel surrounding rotated intestinal vessels
[11–13]
(Fig. 1). The whirl sign was
originally reported in a case of intestinal malrotation
[11] but has been subsequently
described in obstruction of the midgut and distal colon
[3,
7,
11,
13,
14]. The tightness of the
whirl pattern reflects the degree to which the mesentery and vessels are
rotated [12]. The presence of
such anatomic features, which represent compression of the vascu lature, may
imply the presence of a more severe clinical manifestation of SBO and may aid
in the treatment of a patient with suspected SBO, especially if other CT signs
also are present [15]. The aim
of this study was to evaluate the utility of the whirl sign in the treatment
of and prediction of outcome among patients with clinical and radiologic signs
of SBO.
View larger version (122K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1 —80-year-old man with surgically confirmed small-bowel
obstruction secondary to internal hernia. Contrast-enhanced CT scan through
abdomen shows whirl sign (thick arrow). Minimal accompanying ascites
and small ventral hernia (thin arrow) are evident.
|
|
Materials and Methods
A retrospective review was performed on a database containing the records
of patients at a tertiary care hospital who were referred for CT because of
signs of SBO during the study period, which lasted from January 2005 to
January 2007. The initial CT diagnosis in all 453 cases in the database was
reviewed by a clinical investigator, and patients with a radiologic diagnosis
of SBO were included in the study. These cases were examined for demographic
data, concurrent diagnoses, radiologic signs of SBO noted by the reading
radiologist including grade of obstruction, and whether clinical follow-up
records were available. Cases were excluded from the study if the initial CT
diagnosis was not SBO or if follow-up infor mation was not available. The
alternative diagnosis and hospital course of excluded patients were noted.
A board-certified radiologist blinded to the clinical data reviewed the
included CT images for signs of SBO and ischemia. A PACS was used to evaluate
scans for the whirl sign, defined as a twisting of bowel visually determined
to be greater than 90° surrounding a central soft-tissue density of
mesentery and vessels, taking on a swirling appearance. The images also were
reviewed for the presence of pneumatosis intestinalis, maximum dilation of the
bowel, maximum thickness of the bowel wall, and ascites.
Outcome was assessed by the clinical investigator on the basis of hospital
management and clinical course. For patients who underwent medi cal treat
ment, hospital records were reviewed for either resolution of the obstruction
or death. The operative notes for patients who underwent lap aro tomy were
reviewed for findings of either SBO and its cause or an alternative diagnosis.
Available pathologic diagnoses also were reviewed.
Patients were grouped according to whether they had SBO necessitating
surgery. Patients undergoing medical management who died of SBO were
considered with the group needing surgery. If SBO was not found at laparotomy,
the patient was considered with the group that did not need surgery.
We evaluated outcome in both groups depending on whether the CT scans
showed the whirl sign. We calculated sensitivity, specificity, positive
predictive value (PPV), negative predictive value (NPV), and odds ratio to
determine the value of the whirl sign for prediction of the presence of SBO
necessitating surgery. The 95% CI was used to determine the level of
statistical significance of the odds ratio and to estimate the precision of
the other metrics.
Results
A total of 453 patients (202 men, 251 women; mean age, 49.8 years; age
range, 10–102 years) were considered in the original sample. In the
cases of 195 (43%) of the patients, the diagnosis of SBO was made with
radiologic criteria. One patient was excluded because of lack of follow-up, so
194 cases were included in the sample. One hundred forty-eight (76%) of the
194 patients were found to have complete obstruction
(Table 1).
Fifty-three (27%) of the 194 patients needed surgery, and 141 (73%) of the
patients were treated conservatively. Among the patients who underwent medical
treatment, 140 were successfully discharged with resolution of symptoms, and
one patient died. SBO was found intraoperatively in 52 (98%) of 53 cases, and
enteritis was the diagnosis in the other case
(Table 2). The most common
causes of obstruction noted by the surgeon were adhesions in 32 (62%), hernia
in 12 (23%), and volvulus in three (6%) of the 52 patients. There was gross
evidence of ischemia or necrosis in 10 (19%) of 52 patients. The reports on 28
pathologic specimens were available; the most common findings were nine cases
of ischemia and necrosis, eight of serosal abnormality, and three of carcinoma
(Table 3).
The most common descriptive findings from the initial radiologic reports of
patients undergoing surgery were transition point with proximal dilatation and
distal collapse in 28 (53%), distended bowel in 12 (23%), and dilatation and
collapse with no obvious transition point in six (11%) of 53 patients. In
surgical patients, the average diameter of the bowel lumen was 4.1 cm at the
point of maximum dilation, and the average wall thickness was 2.7 mm.
Pneumatosis intestinalis was found in four (8%) and ascites in 14 (26%) of the
53 patients (Table 4).
The whirl sign was identified on the CT scans of 40 of the 194 patients in
the study. Thirty-two of the 40 patients had SBO necessitating surgery (Fig.
2A,
2B,
2C), for a PPV of 80%. Eight
(20%) of the 40 patients with a whirl sign underwent successful medical
treatment (Fig. 3A,
3B,
3C). Twenty-one (14%) of the
154 patients without a whirl sign needed surgery, and 133 did not, for an NPV
of 86% (Table 5).
Fifty-three patients either were treated with surgical therapy or died
during conservative therapy for SBO. The whirl sign was present on CT scans of
32 of the 53 patients, for a sensitivity of 60%; 133 of 141 patients had
neither SBO necessitating surgery nor a whirl sign on CT, for a specificity of
94%. The odds ratio between presence and absence of the whirl sign in
prediction of the presence of SBO necessitating surgery in this sample was
25.3 (95% CI, 10.3–62.3) (Table
6).
Among 32 patients with radiographic evidence of high-grade obstruction, 16
under went surgery, accounting for 16 of 53 surgical patients, for a
sensitivity of 30% and PPV of 50%. The finding of high-grade obstruction had a
specificity of 89%, NPV of 77%, and odds ratio of 3.38 (95% CI,
1.54–7.39) for being predictive of the presence of SBO necessitating
surgery (Table 7). In addition,
49 of 148 patients with findings of either complete or high-grade obstruction
needed surgery. The finding of complete or high-grade obstruction had a
sensitivity of 92%, specificity of 30%, PPV of 33%, NPV of 91%, and odds ratio
of 5.19 (95% CI, 1.76–15.3) (Table
8).
Discussion
CT has sensitivity and specificity greater than 90% in the identification
of SBO, but management is varied
[16]. Although there is no
consensus, urgent laparotomy has been advocated for complete SBO because the
obstruction can result in intestinal ischemia and necrosis
[1,
5,
17]. On the other hand, it has
been estimated that in 20–73% of patients, symptoms of SBO may resolve
with a trial of nasogastric decompression, fluid rehydration, and monitoring
[18,
19]; thus the complications of
laparotomy can be avoided. However, delay of surgery when medical management
fails increases morbidity and mortality
[5]. Our study addressed the
need for reliable imaging predictors of the need for surgery for SBO.
The whirl sign depicts anatomic features that can compromise circulation to
the bowel. The sign appears as a twist of bowel wrapping around a single
constrictive focus of mesentery that may contain intestinal blood vessels
[3,
20]. It has been proposed
[9,
21] that the extent of the
swirling pattern correlates with the degree of bowel rotation and thus with
the severity of vascular impairment.
The importance of the whirl sign has been questioned not only because of
findings of a low PPV but also because of low sensitivity in multiple studies
for endpoints such as SBO and volvulus. A retrospective study
[22] with a large number
oncologic patients showed that the whirl sign had a sensitivity of 64% and a
PPV of 21% in the detection of volvulus. In another retrospective study,
investigators found a sensitivity of 21% for surgical findings of closed-loop
obstruction [17]. In studies
[8,
22] with large numbers of
subjects, the whirl sign has been found specific for volvulus and SBO.
However, the authors [23] of
another report suggested the CT finding of mesenteric rotation may be
nonspecific because volvulus was found in only two of six patients with the
whirl sign.
We analyzed outcome among 194 patients with clinical and radiologic
findings of SBO and found that if a whirl sign was present on a CT scan, a
patient was 25.3 times (odds ratio, 25.3) as likely as a patient without the
sign to need surgery. In addition, patients with a whirl sign on CT had a high
probability of having SBO (PPV, 80%), and those without the sign likely did
not have SBO necessitating surgery (NPV, 86%). To our knowledge, PPV and NPV
have not been previously reported in the literature, although it appears that
these values are more clinically useful than earlier reports of low PPV and
high NPV for volvulus [22].
Our findings of low sensitivity (60%) and high specificity (94%) in this, to
our knowledge, largest series of patients with whirl sign in a general sample
in the United States are in line with estimates from prospective studies of
SBO in the surgical literature
[8]. Estimates of the strong
reliability of the whirl sign for predicting the need for surgery may be
especially helpful to physicians faced with equivocal clinical features.
Findings of either high-grade or complete obstruction did not compare
favorably with the presence of the whirl sign in predicting the need for
surgery for SBO. Only one half of the patients with high-grade obstruction
needed surgery (PPV, 50%), and these patients were only three times as likely
as patients without high-grade obstruction to need surgery (odds ratio, 3.38).
Approximately one third of patients identified as having complete or
high-grade SBO needed surgery, accounting for almost all (49 of 53) of the
surgical patients, but identification was incorrect in the other two thirds of
the patients (specificity, 30%; PPV, 33%). Patients with complete obstruction
were five times as likely as patients without complete obstruction to need
surgery (odds ratio, 5.19). Although in the literature reliability values for
complete and high-grade features in prediction of SBO have been found much
higher [16], our results
showed these values may not be accurate for prediction of management. Our
findings suggest that the whirl sign is a highly relevant predictor of
clinical management compared with final radiologic impressions and may be used
to guide formulation of the radiologic report on a patient with SBO.
Aside from the whirl sign, various common CT features of SBO were observed
in this study. The pattern of bowel dilatation proximal to a transition point
and distal collapse was found in 53% of the surgical patients, significantly
less than the 96% in the literature
[17]. Scans of surgical
patients showed increased bowel dilatation and wall thickness on average, and
CT signs of ischemia, such as pneumatosis and ascites, were present in a small
portion of cases.
This study was limited by its retrospective design. We did not have a
method for including patients with the incidental finding of SBO on CT for
another indication, introducing poten tial selection bias. After discharge,
acquisition of follow-up information about patients who did not undergo
surgery was not complete. An attending radiologist identified the study
population from initial nonstandardized readings. Only one blinded radiologist
reviewed the CT scans for the whirl sign, which has been reported to have low
interrater reliability [22].
The exact criteria by which the surgeon decided to operate were difficult to
delineate retrospectively. Because the sample was selected for cases of
clinical and radiologic SBO, it was not determined how often the whirl sign
appeared in excluded patients. A prospective study would help eliminate these
limitations and improve understanding of the role of the whirl sign in the
management of SBO.
A patient with the whirl sign on CT is 25.3 times as likely as a patient
without the sign to have SBO necessitating surgery. Although the sensitivity
and specificity in this study were in line with those in earlier studies, the
reliability estimates of this CT feature were higher than previously reported.
The results of this study suggest an important role for the whirl sign in
assessment of the treatment options for patients with clinical and radiologic
signs of SBO.
References
- Maglinte DD, Heitkamp DE, Howard TJ, Kelvin FM, Lappas JC. Current
concepts in imaging of small bowel obstruction. Radiol Clin North
Am 2003; 41:263
–283[CrossRef][Medline]
- Balthazar EJ. CT of small-bowel obstruction.
AJR 1994; 162:255
–261[Abstract/Free Full Text]
- Balthazar EJ, Birnbaum BA, Megibow AJ, Gordon RB, Whelan CA,
Hulnick DH. Closed-loop and strangulating intestinal obstruction: CT signs.
Radiology 1992;185
: 769–775[Abstract/Free Full Text]
- Frager D, Baer JW, Medwid SW. Detection of intestinal ischemia in
patients with acute small-bowel obstruction due to adhesions or hernia:
efficacy of CT. AJR 1996;166
: 67–71[Abstract/Free Full Text]
- Cheadle WG, Garr EE, Richardson JD. The importance of early
diagnosis of small bowel obstruction. Am Surg1988; 54:565
–569[Medline]
- Frager D, Medwid SW, Baer JW, Mollinelli B, Friedman M. CT of
small-bowel obstruction: value in establishing the diagnosis and determining
the degree and cause. AJR 1994;162
: 37–41[Abstract/Free Full Text]
- Taourel PG, Fabre JM, Pradel EJ, Seneterre EJ, Megibow AJ, Bruel
JM. Value of CT in the diagnosis and management of patients with suspected
acute small-bowel obstruction. AJR 1995;165
:1187
–1192[Abstract/Free Full Text]
- Donckier V, Closset J, Van Gansbeke D, et al. Contribution of
computed tomography to decision making in the management of adhesive small
bowel obstruction. Br J Surg 1998;85
:1071
–1074[CrossRef][Medline]
- Moore CJ, Corl FM, Fishman EK. CT of the cecal volvulus.
AJR 2001; 177:95
–98[Free Full Text]
- Maglinte DD, Balthazar EJ, Kelvin FM, Megibow AJ. The role of
radiology in the diagnosis of small-bowel obstruction.AJR 1997; 168:1171
–1180[Free Full Text]
- Fisher JK. Computed tomographic diagnosis of volvulus in intestinal
malrotation. Radiology 1981;140
: 145–146[Abstract/Free Full Text]
- Shaff MI, Himmelfarb E, Sacks GA, Burks DD, Kulkarni MV. The whirl
sign: a CT finding in volvulus of the large bowel. J Comput Assist
Tomogr 1985; 9:410[CrossRef][Medline]
- Khurana B. The whirl sign. Radiology2003; 226:69
–70[Free Full Text]
- Frank AJ, Goffner LB, Fruauff AA, Losada RA. Cecal volvulus: the CT
whirl sign. Abdom Imaging 1993;18
: 288–289[Medline]
- Ha HK, Kim JS, Lee MS, et al. Differentiation of simple and
strangulated small-bowel obstructions: usefulness of known CT criteria.
Radiology 1997;204
: 507–512[Abstract/Free Full Text]
- Mallo RD, Salem L, Lalani T, Flum DR. Computed tomography diagnosis
of ischemia and complete obstruction in small bowel obstruction: a systematic
review. J Gastrointest Surg 2005;9
: 690–694[CrossRef][Medline]
- Scaglione M, Grassi R, Pinto A, et al. Positive predictive value
and negative predictive value of spiral CT in the diagnosis of closed loop
obstruction complicated by intestinal ischemia. Radiol Med
(Torino) 2004; 107:69
–77[Medline]
- Fevang BT, Jensen D, Svanes K, Viste A. Early operation or
conservative management of patients with small bowel obstruction?
Eur J Surg 2002;168
: 475–481[CrossRef][Medline]
- Seror D, Feigin E, Szold A, et al. How conservatively can
postoperative small bowel obstruction be treated? Am J
Surg 1993; 165:121
–126[CrossRef][Medline]
- Ha HK, Rha SE, Kim JH, et al. CT diagnosis of strangulation in
patients with small-bowel obstruction: current status and future direction.
Emerg Radiol 2000;7
: 47–55[CrossRef]
- Ha HK. CT in the early detection of strangulation in intestinal
obstruction. Semin Ultrasound CT MR 1995;16
: 141–150[CrossRef][Medline]
- Gollub MJ, Yoon S, Smith LM, Moskowitz CS. Does the CT whirl sign
really predict small bowel volvulus? Experience in an oncologic population.
J Comput Assist Tomogr 2006;30
: 25–32[CrossRef][Medline]
- Blake MP, Mendelson RM. The whirl sign: a nonspecific finding of
mesenteric rotation. Australas Radiol1996; 40:136
–139[Medline]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
|
|
 
C.-H. Lin, P.-C. Kao, H.-P. Wang, and W.-C. Lien
Small bowel volvulus and the whirl sign
QJM,
November 1, 2009;
102(11):
815 - 815.
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
