Acute Diverticulitis of the Cecum and Ascending Colon: The Value of Thin-Section Helical CT Findings in Excluding Colonic Carcinoma

Transcription

1 Hyun-Jung Jang 1 Hyo K. Lim Soon Jin Lee Won Jae Lee Eung Yeop Kim Seung Hoon Kim Received August 17, 1999; accepted after revision October 12, All authors: Department of Radiology, Samsung Medical Center Sungkyunkwan University School of Medicine, 50 Ilwon-dong, Kangnam-ku, Seoul , Korea. Address correspondence to H. K. Lim. AJR 2000;174: X/00/ American Roentgen Ray Society Acute Diverticulitis of the Cecum and Ascending Colon: The Value of Thin-Section Helical CT Findings in Excluding Colonic Carcinoma OBJECTIVE. The purpose of this study was to assess the value of characteristic thin-section helical CT findings of acute diverticulitis involving the cecum and ascending colon in excluding colonic carcinoma. 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. Interobserver 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 (κ > 0.60) was obtained for both findings. Inflamed diverticula (κ = 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 (κ = 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. R ight-sided colonic diverticulitis is considered to be a rare condition in the Western population [1 4] and radiologic studies of acute colonic diverticulitis have usually been limited to the sigmoid colon [5 11]. Recently, owing to the liberal use of sonography and CT in patients with abdominal pain, right-sided colonic diverticulitis has been reported to be more common [12] than was previously estimated and has received more attention [3, 12, 13]. 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 4, 12 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 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 22]. In approximately 10% of patients, diverticulitis is AJR:174, May

2 Jang et al. 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 arrowhead-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. 1). 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. Materials and Methods Patient Selection We searched our department medical records for the period of September 1994 to May 1999 for patients with a final diagnosis of diverticulitis of the ascending colon or cecum. We obtained a list of 27 patients who had also undergone contrast-enhanced CT of the abdomen and pelvis. Eight of the 27 patients were excluded because they neither underwent surgery to prove the diagnosis nor were followed up radiologically or clinically over 3 months. We included the other 19 consecutive patients with proven acute diverticulitis of the cecum and ascending colon who underwent thin-section helical CT in our institution. The patient sample consisted of 10 men and nine women, ranging from 23 to 81 years old (mean age, 56 years). A 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 B 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) 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 Fig. 1. Diverticulitis of ascending colon in 54-year-old man. A, Thin-section axial CT scan shows marked wall thickening of ascending colon with inflamed diverticulum (arrow), mild peridiverticular inflammation and preservation of layered enhancement pattern as inner hyperattenuating layer, thickened middle layer of low attenuation (m), and outer high attenuation. B, Axial CT scan of contiguous slice cephalad to A more clearly shows preserved enhancement pattern of thickened colonic wall. Also note pericolic infiltration (arrowheads) and venous engorgement (arrows). 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 1398 AJR:174, May 2000

3 Acute Diverticulitis of the Cecum and Ascending Colon 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 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 rightsided colonic diverticulitis from right-sided colonic carcinoma. Results The results of the independent analyses by the two radiologists regarding all parameters for right-sided colonic diverticulitis and rightsided colonic carcinoma are summarized in Table 1. The interobserver agreement was excellent for inflamed diverticula (κ = 0.80) and preservation of enhancing pattern of involved colonic wall (κ = 0.70); good for mesenteric venous engorgement (κ = 0.58), simple diverticula (κ = 0.55), and fluid at the root of the mesentery (κ = 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 TABLE 1 Thin-Section Helical CT Findings in Patients with Right-Sided Colonic Diverticulitis or Colonic Carcinoma Note. For the p value, the statistical analysis was performed with the results of the consensus of the two observers for each parameter. κ = kappa value in observer agreement. a Relative severity of pericolic infiltration: 1 = mild pericolic haziness or thin pericolic strands, 2 = between 1 and 3, 3 = areas of illdefined soft tissue or frank abscess formation. 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 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 Fig. 2. Ascending colonic carcinoma in 61- year-old woman. Axial CT scan shows colonic wall thickening (short straight arrows) with enhancement, pericolic strands, and diverticula (curved arrows), which meet diagnostic criteria of diverticulitis on conventional CT. Note loss of layered enhancement pattern of involved colonic wall. Also note lumen of colon (long thin arrow). Diverticulitis (n = 19) (%) Carcinoma (n = 21) (%) κ p CT Findings Observer 1 Observer 2 Observer 1 Observer 2 Inflamed diverticula 15 (78.9) 18 (94.7) 2 (9.5) 1 (4.8) Preserved wall enhancement pattern 18 (94.7) 16 (84.2) 2 (9.5) 0 (0) Simple diverticula 15 (78.9) 15 (78.9) 6 (28.6) 3 (14.3) Arrowhead sign 4 (21.1) 3 (15.8) 4 (19.0) 1 (4.8) 0.04 >0.05 Mesenteric venous engorgement 13 (68.4) 14 (73.7) 15 (71.4) 20 (95.2) 0.58 >0.05 Fluid at the root of mesentery 1 (5.3) 2 (10.5) 1 (4.8) 1 (4.8) 0.54 >0.05 Degree of pericolic infiltration a 1 5 (26.3) 6 (31.6) 10 (47.6) 13 (61.9) 2 12 (63.2) 4 (21.1) 8 (38.1) 2 (9.5) (5.3) 8 (42.1) 3 (14.3) 6 (28.6) Wall thickness (mm) 14.9 ± ± excluding colonic carcinoma. Preserved wall enhancement pattern had a mean sensitivity, specificity, and accuracy of 89.5%, 95.3%, and 92.5%, respectively. Discussion Diverticulitis is a benign and primarily nonsurgical disease that must be differentiated preoperatively from colonic carcinoma [12, 14, 15, 18, 22]. Several groups of researchers have reported that accurate distinction between these two diseases is not possible with CT and further investigation should be performed to exclude carcinoma [5, 7, 21]. All colonic carcinomas included in this study showed abnormal thickening of the colonic wall and some degree of pericolic infiltration, and some of these carcinomas also had underlying diverticulosis; these findings are indistinguishable from findings generally indicative of sigmoid diverticulitis AJR:174, May

4 Jang et al. (Fig. 2). Some studies, on the other hand, have suggested helpful CT findings of sigmoid diverticulitis that exclude carcinoma: mesenteric inflammation [22], sawtoothlike thickening of colonic haustra [5, 11], and more severe pericolic infiltration relative to the degree of wall thickening [3, 23]. Recently, an arrowheadshaped wall thickening has been reported to be a finding specific to colonic diverticulitis [24]. 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 ( %) accompanied by vascular engorgement as diverticulitis ( %), 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 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. Fig. 3. Acute diverticulitis of ascending colon in 27-year-old woman. Thin-section axial CT scan shows wall thickening of ascending colon (short black arrow) and terminal ileum (white arrow) with layered enhancement pattern clearly as inner hyperattenuating layer, middle layer of low attenuation, and outer hyperattenuating layer. There is inflamed diverticulum (long black arrow) that contains fecalith and peridiverticular infiltration. Fig. 4. Carcinoma of ascending colon in 68-year-old man. Compared with Figure 3, thin-section axial CT scan shows greater thickness of colonic wall involved by enhancing colonic carcinoma and resultant loss of layered enhancement pattern (solid arrow). Also note marked pericolic infiltration (open arrows). Fig. 5. Overlap in wall thickness and pericolic infiltration between diverticulitis and colonic carcinoma. Thin-section axial CT scan of 61-year old man with ascending colonic carcinoma shows similar thickness of involved colonic wall and more pericolic infiltration than that in Figure 3. Note loss of layered enhancement pattern as result of enhancement of tumor involving colonic wall (solid arrows) and engorged mesenteric vessels (open arrow). Fig. 6. Ascending colonic diverticulitis in 37-year-old woman. Thin-section axial CT scan at level of proximal ascending colon shows inflamed diverticulum within surrounding peridiverticular inflammation recognized by enhancing diverticular wall (arrowheads), which may be overlooked on unenhanced imaging AJR:174, May 2000

5 Acute Diverticulitis of the Cecum and Ascending Colon Fig year-old man with obstructive colitis proximal to colonic carcinoma. A, Axial CT scan shows concentric wall thickening (long arrows) with preservation of layered enhancement pattern in ascending colon. Note diverticulum with thickened wall (short arrow), which was interpreted as inflamed diverticulum by both observers. However, diverticulum is not accompanied by peridiverticular inflammation. B, Axial CT scan at level caudal to A shows colonic wall being replaced by enhancing soft tissue, representing carcinoma (arrows) found at most distal segment of layered wall thickening. 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 (κ = 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 thinsection 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 (κ = 0.70). A 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. 7). 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 rightsided 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 rightsided colonic diverticulitis can be reliably distinguished from colonic carcinoma. B AJR:174, May

6 Jang et al. References 1. Balthazar EJ, Megibow AJ, Gordon RB, Hulnick D. Cecal diverticulitis: evaluation with CT. Radiology 1987;162: Gouge TH, Coppa GF, Eng K, Ranson JH, Localio SA. Management of diverticulitis of the ascending colon. Am J Surg 1983;145: Katz DS, Lane MJ, Ross BA, Gold BM, Jeffrey RB Jr, Mindelzun RE. Diverticulitis of the right colon revisited. AJR 1998;171: Lo CY, Chu KW. Acute diverticulitis of the right colon. Am J Surg 1996;171: Balthazar EJ, Megibow A, Schinella RA, Gordon R. Limitation in the CT diagnosis of acute diverticulitis: comparison of CT, contrast enema, and pathologic findings in 16 patients. AJR 1990;154: Cho KC, Morehouse HT, Alterman DD, Thornhill BA. Sigmoid diverticulitis: diagnostic role of CT comparison with barium enema studies. Radiology 1990;176: Pradel JA, Adell J-F, Taourel P, Djafari M, Monnin-Delhom E, Bruel J-M. Acute colonic diverticulitis: prospective comparative evaluation with US and CT. Radiology 1997;205: Johnson CD, Baker ME, Rice RP, Silverman P, Thomson WM. Diagnosis of acute colonic diverticulitis: comparison of barium enema and CT. AJR 1987;148: Lieberman JM, Haaga JR. Computed tomography of diverticulitis. J Comput Assist Tomogr 1983;7: Hulnick DH, Megibow AJ, Balthazar EJ, Naidich DP, Bosniak MA. Computed tomography in the evaluation of diverticulitis. Radiology 1984;152: Rao PM, Rhea JT, Novelline RA, et al. Helical CT with only colonic contrast material for diagnosing diverticulitis: prospective evaluation of 150 patients. AJR 1998;170: Oudenhoven LFIJ, Koumans RKJ, Puylaert JBCM. Right colonic diverticulitis: US and CT findings new insights about frequency and natural history. Radiology 1998;208: Jang H-J, Lim HK, Lee SJ, Choi SH, Lee MH, Choi MH. Acute diverticulitis of the cecum and ascending colon: thin-section helical CT findings. AJR 1999;172: Arrington P, Judd CS Jr. Cecal diverticulitis. Am J Surg 1981;142: Graham SM, Ballantyne GH. Cecal diverticulitis: a review of the American experience. Dis Colon Rectum 1987;30: Wagner DE, Zollinger RW. Diverticulitis of the cecum and ascending colon. Arch Surg 1961;83: Greaney EM, Snyder WH. Acute diverticulitis of the cecum encountered at emergency surgery. Am J Surg 1957;94: Crist DW, Fishman EK, Scatarige JC, Cameron JL. Acute diverticulitis of the cecum and ascending colon diagnosed by computed tomography. Surg Gynecol Obstet 1988;166: Rao PM, Rhea JT, Novelline RA, et al. Helical CT technique for the diagnosis of appendicitis: prospective evaluation of a focused appendix CT examination. Radiology 1997;202: Chintapalli KN, Chopra S, Ghiatas AA, Esola CC, Dodd GD III. Diverticulitis versus colon cancer: differentiation with CT findings (abstr). Radiology 1996;201(P): Ryan JM, Boland GW, Harisinghani MG, Schima W, Mueller PR. Patients with CT diagnosis of acute diverticulitis: frequency of unsuspected neoplasm (abstr). Radiology 1996;201(P): Padidar AM, Jeffrey RB Jr, Mindelzun RE, Dolph JF. Differentiating sigmoid diverticulitis from carcinoma on CT scans: mesenteric inflammation suggests diverticulitis. AJR 1994;163: Birnbaum BA, Balthazar EJ. CT of appendicitis and diverticulitis. Radiol Clin North Am 1994;32: Rao PM, Rhea JT. Colonic diverticulitis: evaluation of the arrowhead sign and the inflamed diverticulum for CT diagnosis. Radiology 1998;209: Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977;33: Kleinbaum DE, Kupper LL. Applied regression analysis and other multivariate methods. North Scituate, RI: Duxbury, 1978: Rao PM, Wittenberg J, McDowell RK, Rhea JT, Novelline RA. Appendicitis: use of arrowhead sign for diagnosis at CT. Radiology 1997;202: AJR:174, May 2000