| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 All authors: Department of Radiology, Samsung Medical Center Sungkyunkwan University School of Medicine, 50 Ilwon-dong, Kangnam-ku, Seoul 135-710, Korea.
Received August 17, 1999;
accepted after revision October 12, 1999.
Address correspondence to H. K. Lim.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Thin-section helical CT scans (5-mm collimation) of 19 consecutive patients with proven diverticulitis and 21 consecutive patients with surgically proven carcinoma involving the cecum and ascending colon were reviewed retrospectively. Two radiologists independently analyzed these parameters: degree of pericolic infiltration, mesenteric fluid, vascular engorgement, arrowhead-shaped wall thickening, air-filled diverticula, inflamed diverticula, and preserved enhancement pattern of involved colonic wall. Inter-observer agreement was assessed with a kappa statistical analysis, and the features that most distinguished diverticulitis from colonic carcinoma were selected with a stepwise logistic-regression analysis.
RESULTS. The two CT findings of right-sided colonic diverticulitis
that most distinguished it from colonic carcinoma were inflamed diverticula
and the preservation of an enhancement pattern of the involved colonic wall.
Excellent interobserver agreement (
> 0.60) was obtained for both
findings. Inflamed diverticula ( = 0.80) had a mean sensitivity,
specificity, and accuracy for diverticulitis of 86.8%, 92.9%, and 90.0%,
respectively, in differentiating right-sided colonic diverticulitis from
colonic carcinoma. Preserved wall enhancement pattern ( = 0.70) had a
mean sensitivity, specificity, and accuracy of 89.5%, 95.3%, and 92.5%,
respectively.
CONCLUSION. On thin-section helical CT, an inflamed diverticula and a preserved enhancement pattern of the thickened colonic wall were the two most statistically significant CT findings of acute diverticulitis involving the cecum and ascending colon that distinguished diverticulitis from colonic carcinoma.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The correct preoperative diagnosis of right-sided colonic diverticulitis has rarely been made; instead, the diverticulitis is usually discovered unexpectedly at surgery for suspected appendicitis [1,2,3,4, 12,13,14,15,16,17]. Even after the appendix is discovered to be healthy at surgery, unnecessarily extensive surgery is often performed because differentiating colonic carcinoma from right-sided colonic diverticulitis is difficult [2, 4, 14,15,16,17,18]. Because conservative treatment is recommended for right-sided colonic diverticulitis [3, 12, 14, 15, 18], early accurate diagnosis is important to avoid unnecessary laparotomy and also to prevent potential complications.
Clinically, right-sided colonic diverticulitis has been one of the greatest mimics of acute appendicitis; however, with current thin-section helical CT, most healthy appendixes can be revealed [13, 19] and the differentiation of colonic carcinoma from acute appendicitis is not difficult. In addition to acute appendicitis, colonic carcinoma must be distinguished from acute diverticulitis [5, 16, 17, 20,21,22]. In approximately 10% of patients, diverticulitis is reported to be indistinguishable from carcinoma on CT [5, 22]. Common conventional CT findings of acute diverticulitis are hazy infiltration of pericolic fat, focal thickening of the colonic wall, and diverticula, which also can be found in colonic carcinoma [6, 7, 10, 23]. A recent study [24] with helical CT has reported that an arrow-head-shaped wall thickening of the colon is a specific sign of colonic diverticulitis in regions other than the cecum. With respect to the sigmoid colon, some studies have suggested useful CT findings of diverticulitis in excluding colonic carcinoma [11, 20, 22, 23], but none of these studies addresses right-sided colonic diverticulitis.
Recently, we have observed two distinctive findings of right-sided colonic diverticulitis on thin-section helical CT [13]: inflamed diverticula and preservation of a layered enhancement pattern of colonic wall (Fig. 1A,1B). The purpose of this study was to determine the value of these characteristic findings of right-sided colonic diverticulitis on thin-section helical CT in excluding colonic carcinoma.
|
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The diagnosis of right-sided colonic diverticulitis was surgically proven in six patients. All 13 patients who did not undergo surgery had been followed up for 4-37 months (mean, 14 months) to confirm the absence of any other colonic disease. Among the 13 patients, 10 had undergone 2- or 3-month follow-up CT that showed marked improvement of diverticulitis or normal findings. Barium enema was performed in six patients, including the three patients who were not followed up with CT, and showed typical findings of diverticulitis: eccentric intramural mass effect associated with serosal spiculation, edema and spasm without mucosal destruction, and ruptured or unruptured diverticula. Colonoscopy was performed in three patients who also underwent follow-up CT.
For comparative study, we searched surgical records for patients with right-sided colonic carcinoma surgically proven during the past 2 years. A list of 47 patients with surgically proven cecal and ascending colonic carcinoma was obtained. Among the 47 patients, we found 39 patients who also underwent contrast-enhanced thin-section helical CT of the abdomen and pelvis. Among these 39 patients, those with carcinoma foci within an adenomatous polyp, with no pericolic infiltration at pathology, or with evident metastatic lesions at CT were excluded. Finally, 21 consecutive patients with surgically proven colonic carcinoma whose CT scans showed colonic wall thickening and varying degrees of pericolic fat infiltration were included.
CT Technique
All CT examinations were performed with a helical CT scanner (HiSpeed
Advantage; General Electric Medical Systems, Milwaukee, WI). The upper abdomen
from the level of the hepatic dome to the inferior tip of the liver was
scanned with a helical mode (7-mm collimation at a pitch of 1:1 and 7-mm
reconstruction intervals). The rest of the abdomen and pelvis was scanned with
a clustered data acquisition mode (5-mm collimation and 5-mm intervals). In
our institution, a clustered mode is used for routine CT of the lower abdomen
to reduce tube overload and to improve resolution.
Each patient ingested 600 ml of 2.5% diluted sodium amidotrizoate and meglumine amidotrizoate mixture (Gastrografin; Schering, Berlin, Germany) 50-60 min before CT scanning and an additional 300 ml just before CT scanning. CT scanning began 70 sec after the start of the IV injection of 120 ml of iopromide (Ultravist 300; Schering) at a rate of 2.5 ml/sec. The same techniques were used both for diverticulitis and colonic carcinoma, with the exception of the rectal contrast media. We do not routinely administer contrast media via the rectum, except water, to patients with suspected or proven colonic carcinoma. In this study population, water was given via the rectum for 14 of the 21 patients with colonic carcinoma and one patient with right-sided colonic diverticulitis who was clinically suspected of having colonic carcinoma. However, in five of these 15 patients, water did not extend to the ascending colon or the cecum because of the patients' intolerance. For the remaining patients, no contrast media or air insufflation was administered via the rectum.
Imaging Analysis
All CT scans were reviewed retrospectively and independently by two
experienced abdominal radiologists. The same parameters were applied for
diverticulitis and colonic carcinoma. CT scans of all patients were randomly
distributed and assessed. The interpreters had no knowledge of clinical or
pathologic data other than the age and sex of the patients.
The evaluated parameters were as follows: the relative degree of pericolic infiltration (1, mild pericolic haziness or thin pericolic strands; 2, between 1 and 3; and 3, areas of ill-defined soft tissue or obvious abscess formation), presence or absence of simple diverticula (air-filled diverticula without thickened wall), fluid at the root of the mesentery [22], mesenteric vascular engorgement [22], arrowhead sign [24], inflamed diverticula [13] (enhancement of thickened diverticular wall surrounded by the area of peridiverticular inflammation), and preservation of enhancing pattern of the colonic wall [13] (inner high attenuation, middle low attenuation, and outer high attenuation). The wall thickness was measured at the maximal magnification on a 2K x 2K picture archiving and communication system (PACS) monitor (General Electric Medical Systems Integrated Imaging Solutions, Mt. Prospect, IL) by one radiologist.
Statistical Analysis
The interobserver agreement was assessed with the kappa statistic. A kappa
value of 0.60 was considered to indicate excellent agreement, between 0.40 and
0.60 was considered good agreement, and less than 0.40 was considered poor
agreement, as suggested by Landis and Koch
[25]. Discrepancies were
resolved by the consensus of the two interpreters for each parameter. The
final results based on the consensus interpretations were assessed
individually with regard to the relationship with the final diagnosis
(diverticulitis versus colonic carcinoma) by using a Fisher's exact test or
chi-square test for categoric variables and a Mann-Whitney test for continuous
variables. A p value of less than 0.05 was considered to indicate a
statistically significant difference.
With respect to the parameters that showed individually significant differences, a stepwise logistic-regression analysis [26] was used to identify the findings that allowed the best prediction of diverticulitis versus colonic carcinoma. The sensitivity, specificity, and accuracy for finally selected parameters were assessed in differentiating right-sided colonic diverticulitis from right-sided colonic carcinoma.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
= 0.80) and preservation of enhancing pattern of involved
colonic wall ( = 0.70); good for mesenteric venous engorgement (
= 0.58), simple diverticula ( = 0.55), and fluid at the root of the
mesentery ( = 0.54); and poor for the remaining parameters. The
interobserver discrepancy was finally resolved by the consensus of the two
radiologists. Statistical analysis based on the consensus interpretations of
the individual findings revealed that four parameters—inflamed
diverticula (p = 0.000), preservation of wall enhancing pattern
(p = 0.000), simple diverticula (p = 0.011), and wall
thickness (p = 0.012)—showed significant difference in
prevalence or degree between diverticulitis and colonic carcinoma
(Table 1). The
others—excluding fluid at the root of the mesentery, mesenteric venous
engorgement, arrowhead sign, and the degree of pericolic
infiltration—were not statistically different (p > 0.05),
contrary to previous studies involving the sigmoid colon
[22,23,24].
|
The stepwise logistic regression also revealed that four parameters were significantly different (p < 0.05) between right-sided colonic diverticulitis and colonic carcinoma. In decreasing order of importance, those parameters were preservation of wall enhancement pattern (R = 0.745), inflamed diverticula (R = 0.741), simple diverticula (R = 0.286), and thickness of involved colonic wall (R = 0.280). Inflamed diverticula and preservation of wall enhancement pattern showed far more statistically significant association with diverticulitis.
Inflamed diverticula had a mean sensitivity, specificity, and accuracy for diverticulitis of 86.8%, 92.9%, and 90.0%, respectively, in excluding colonic carcinoma. Preserved wall enhancement pattern had a mean sensitivity, specificity, and accuracy of 89.5%, 95.3%, and 92.5%, respectively.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Our results suggested three differences between the ascending and sigmoid colon that make applications of these findings difficult. First, the sigmoid colon usually runs parallel to the axial scan, whereas the right-sided colon is perpendicular to the axial scan. Therefore, muscular hypertrophy is unlikely to reveal a sawtooth appearance. In addition, the arrowhead sign was rarely seen and was a nonspecific finding in our study. This is probably because the frequency of the arrowhead sign depends on the degree of colonic distention and the orientation of the affected bowel [24]. In the study evaluating the arrowhead sign, only one case of right-sided colonic diverticulitis was included in a total of 47 patients [24]. In right-sided colonic diverticulitis, most cases show circumferential rather than eccentric thickening of the colonic wall [12, 13]. Additionally, it is difficult to distend the proximal right-sided colon as sufficiently as the left-sided colon by enema, especially for elderly patients. Furthermore, even with sufficient luminal distention, the arrowhead sign is not specific for diverticulitis in the cecum because it can also be seen in acute appendicitis [27].
Second, in our results, right-sided colonic carcinoma with pericolic involvement was frequently (71.4-95.2%) accompanied by vascular engorgement as diverticulitis (68.4-73.7%), and fluid at the root of mesentery was rarely found for both diseases, which is in contrast to a previous report [22]. Third, our study also showed that in colonic carcinoma, wall thickening is more severe relative to pericolic infiltration compared with that of diverticulitis. Marked overlap, however, was present between the two diseases (Figs. 3,4,5).
|
|
|
The parameters evaluated in this study were the findings reported to be specific or frequent for acute colonic diverticulitis. Stepwise discriminant analysis revealed that inflamed diverticula and a preserved enhancing pattern of thickened colonic wall were the two most statistically significant findings of diverticulitis that distinguish right-sided colonic diverticulitis from colonic carcinoma.
The recognition of inflamed diverticula is important. As shown in our
patient population, colonic carcinoma can be associated, not infrequently
(14-29%), with simple diverticulosis. Inflamed diverticula, when defined on CT
as diverticula associated with a thickened, enhancing diverticular wall and
peridiverticular inflammatory changes, showed various levels of attenuation
depending on their contents: a fecalith, fecal material, air, or fluid or
soft-tissue attenuation [12,
13]. Nine of the 19 patients
with diverticulitis in our study had inflamed diverticula filled with
soft-tissue attenuation discernible only by the enhancing diverticular wall
(Fig. 6), which could be missed
on CT using thicker collimation. Even with thin-section CT, when performed
without IV contrast material
[24], the inflamed diverticula
was visualized far less frequently (30%) than it was in our study. With
thin-section helical CT with IV contrast enhancement, interpreters in our
study could identify inflamed diverticula in 79-95% of patients with
diverticulitis with excellent interobserver agreement ( = 0.80).
|
Although a preservation of wall enhancement pattern is nonspecific, the
finding is helpful in differentiating diverticulitis from colonic carcinoma
involving the right-sided colon. Because the axis of the right-sided colon is
perpendicular to the CT plane, shouldering of colonic carcinoma, which can be
helpful in excluding benign wall thickening, is rarely seen in the right-sided
colon. Preservation of the wall layer structure is a common sonographic
feature of diverticulitis described in the literature
[12], but it cannot be
frequently seen on conventional or unenhanced CT. With thin-section helical CT
with IV contrast enhancement, each interpreter in our study recognized the
preservation of wall enhancement pattern in 84-95% of patients with
diverticulitis with excellent interobserver agreement ( = 0.70).
The most important limitation in the interpretation of those two findings was proximal obstructive colitis in colonic carcinoma. One case of colonic carcinoma was interpreted as a preserved wall enhancement pattern by one observer and also as having an inflamed diverticulum by both observers (Fig. 7A,7B). Actually, there was a thickened wall replaced by an enhancing soft-tissue mass representing carcinoma at the most distal segment of layered wall thickening; thus, scrutiny of the entire segment of the involved colon is required. In this patient the wall of the diverticulum within the segment of the obstructive colitis was also thickened, but it was not accompanied by peridiverticular inflammation. It would be helpful to be aware that inflamed diverticula are usually located at the level of maximal pericolic inflammation and maximal wall thickening [12].
|
|
The degree of pericolic infiltration was not significantly different by statistical analysis between the two diseases involving the right-sided colon. Care must be taken with interpretation of the results. The patients with colonic carcinoma included in our study had findings similar to that of diverticulitis with conventional CT criteria: wall thickening with pericolic fat involvement. Three of the 21 colonic carcinomas were associated with perforation. In previous reports, the difficulty in differentiating diverticulitis from colonic carcinoma was focused on perforating carcinoma [3, 6, 23]; however, the results of our study showed that colonic carcinomas had a similar degree of pericolic infiltration in diverticulitis, even without perforation. Early use of an imaging examination in our institution for right-sided lower abdominal pain to rule out appendicitis might have affected the population and detection of early diverticulitis of the right-sided colon.
Our study had some limitations. Although the interpreters were not aware of the frequency of both diseases, they knew that the patients had one of the two diseases. Thus, the frequency of inflamed diverticula might be higher than it would be if evaluated in a prospective study. Because more than half of the patients with colonic carcinoma were administered water as a contrast media via the rectum, while no rectal contrast material was used for patients with diverticulitis (except one patient), there could be a bias in measuring the wall thickness of the two diseases. However, as already mentioned, the degree of distention at the level of the right-sided colon seems to be not much different between the two techniques. Another limitation was that most (68%) patients with a final diagnosis of diverticulitis did not have histopathologic confirmation. During a recent 1-year period, no case of right-sided colonic diverticulitis was found at surgery in our institution. It was partly owing to the familiarity with characteristic CT findings of right-sided colonic diverticulitis and also attributable to the early, frequent use of sonography in right-sided abdominal pain that could lower the diagnosis of complicated diverticulitis requiring surgery.
In summary, diverticulitis involving the cecum and ascending colon has differences in anatomic and clinical aspects from that involving the sigmoid colon. Accordingly, those differences may change our perception of the relative frequency and importance of CT findings for diverticulitis in excluding colonic carcinoma. On thin-section helical CT with IV contrast enhancement, inflamed diverticula and a preserved wall enhancement pattern were the two most discriminative findings in excluding the diagnosis of colonic carcinoma involving the right colon. With careful interpretation of CT findings, most cases of right-sided colonic diverticulitis can be reliably distinguished from colonic carcinoma.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  C. Hoeffel, M. D. Crema, A. Belkacem, L. Azizi, M. Lewin, L. Arrive, and J.-M. Tubiana Multi-Detector Row CT: Spectrum of Diseases Involving the Ileocecal Area RadioGraphics, September 1, 2006; 26(5): 1373 - 1390. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Gutman, J.-L. Alberini, M. Wartski, D. Vilain, E. Le Stanc, F. Sarandi, C. Corone, C. Tainturier, and A. P. Pecking Incidental Colonic Focal Lesions Detected by FDG PET/CT Am. J. Roentgenol., August 1, 2005; 185(2): 495 - 500. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. M. Pereira, C. B. Sirlin, P. S. Pinto, R. B. Jeffrey, D. L. Stella, and G. Casola Disproportionate Fat Stranding: A Helpful CT Sign in Patients with Acute Abdominal Pain RadioGraphics, May 1, 2004; 24(3): 703 - 715. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
