Jpn J Radiol (2014) 32:6–13 DOI 10.1007/s11604-013-0260-2
ORIGINAL ARTICLE
Value of DW-MRI ADC quantification of colonic wall lesions in differentiation of inflammatory bowel disease and colorectal carcinoma Serhat Avcu • Mehmet Deniz Bulut • Alpaslan Yavuz • Aydın Bora • Mehmet Beyazal
Received: 25 February 2013 / Accepted: 20 October 2013 / Published online: 1 November 2013 Ó Japan Radiological Society 2013
Abstract Purpose The purpose of our study was to assess the efficiency of diffusion-weighted magnetic resonance imaging (DW-MRI) and the quantification of apparent diffusion coefficient (ADC) values in differentiating colorectal carcinoma from colonic inflammatory bowel disease (IBD) in cases with isolated colonic wall lesions and uncertain clinical and radiologic diagnostic criteria. Methods The study comprised 58 patients with segmental or focal isolated colonic wall thickening. All lacked satisfactory clinical–radiological findings for etiology determination. The mean ADC values of the thickened colonic walls were retrospectively compared with final histopathologic diagnoses. Receiver operating characteristic (ROC) curve analysis was used to determine the ADC cutoff value for differentiation. Results Mean ADC value in the colorectal carcinoma group was significantly lower than that in the IBD group: n = 27, 1.02 ± 0.26 9 10-3 mm2/s; and n = 31, 1.53 ± 0.19 9 10-3 mm2/s, respectively (P \ 0.001). Cutoff ADC value for differentiating colorectal carcinoma from IBD was calculated as 1.39 9 10-3 mm2/s, with 83.9 % sensitivity and 85.2 % specificity. Conclusion ADC measurement of the involved colonic wall segments with DW-MRI has the potential to differentiate isolated colonic IBD from colorectal carcinoma in
S. Avcu A. Yavuz (&) A. Bora M. Beyazal Radiology Department, Yuzuncu Yil University, Ercis Yolu, Van, Turkey e-mail: [email protected] M. D. Bulut Van Goverment Education and Research Hospital, Edremit Yolu, Van, Turkey
123
cases in which clinical–radiologic findings are insufficient for a definitive diagnosis. Keywords MRI DW-MRI Colorectal carcinoma Inflammatory bowel disease
Introduction Ulcerative colitis (UC) and Crohn’s disease (CD) are the primary forms of idiopathic inflammatory bowel disease (IBD), and aberrant host responses to enteric environmental agents are the acknowledged etiological factors for individuals who are genetically susceptible to IBD. Most colorectal carcinomas are believed to arise from sporadic adenomas, but some cancers arise from nonpolypoid dysplastic mucosa, such as the types of cancer occurring with IBD [1]. The majority of neoplasms of the gastrointestinal (GI) tract appear as eccentric tumor masses, whereas some spread locally in a plaque-like infiltrative fashion and are characterized by a lack of marked focal elevation of the colonic wall; thus, differentiating between neoplasms of the GI tract can be difficult, especially for benign inflammatory entities. However, inflammatory pathologies, such as UC and CD, occur with segmental or diffuse thickening of the bowel wall. Double-contrast barium enema is highly accurate in revealing early changes, extent, and severity of IBD; however, computed tomography (CT) is superior to barium studies and endoscopy for diagnosing extramural complications. Magnetic resonance imaging (MRI) plays a similar role to that of CT; additionally, pelvic imaging with fat-suppressed techniques combined with sequences that highlight fluid can emphasize collections to a greater degree than CT. Although colorectal carcinoma has typical features, differential diagnosis can be problematic if the
Jpn J Radiol (2014) 32:6–13
radiological morphology is unusual. Inflammation due to CD, amoebiasis, and tuberculosis can also mimic a primary cancer [2, 3], whereas infiltrating colonic carcinomas can mimic inflammatory processes and vice versa [4]. In this study, we aimed to evaluate the effectiveness of the apparent diffusion coefficient (ADC) value obtained by diffusion-weighted (DW)-MRI in a limited group of individuals for differentiating colorectal carcinomas from IBD when conventional CT and/or MRI findings are insufficient for an accurate diagnosis.
Materials and methods Sixty-five patients (31 men and 34 women; mean age 63.18 ± 14.32 years) with segmental colonic wall thickening determined by CT or conventional MRI, were assessed. These patients lacked satisfactory clinical and radiological findings to determine the etiology of their disease. The bowel wall was considered to be thickened when it was [3 mm and the increased wall thickness had not progressed to the next segment of the colon (limited to one segment, such as rectum, sigmoid, ascending, transverse, or descendant). Patients who had initial clinical and imaging findings characteristic of IBD (disease onset between 15 and 40 years, recurrent episodes of diarrhea, low-grade fever, malabsorption, and marked thickening of the small intestine) or of colorectal carcinoma (sudden weight loss, distant metastasis, mesenteric malignant lymphadenopathy, massive invasion of the surrounding tissue, focal irregular colonic thickening) were excluded. Three patients who required transportation to another institution and four who had final diagnoses other than IBD or colorectal carcinoma (three with colonic ischemia, one with polyposis coli) were also excluded. Fifty-eight patients (28 men and 30 women; mean age 62.71 ± 11.23 years) were underwent routine abdominal MRI examination including DW-MRI sequences. All patients with isolated colonic wall thickening and final histopathologic diagnoses of IBD or colorectal carcinoma were reviewed to determine the feasibility of DW-MRI for initial differentiation of these entities. Pathology confirmation for the final diagnosis was obtained in 53 cases by endoscopic biopsy and in five by surgery (four carcinoma, one IBD; endoscopic process was not successful to sample the lesions). Conventional and DW-MR imaging findings and final endoscopic pathology results were assessed. A 1.5-T MRI system (Siemens Magnetom Symphony, Erlangen, Germany) with a phased-array body coil was used for examinations. The gradient force of the superconductive (niobium–titanium) magnet was 30 mT/m, and the maximum field-of-view (FOV) width was 400 mm. Before DW-MRI examination, conventional abdominal
7
MRI was performed with the following settings: T2weighted true fast imaging with a steady-state precession (true-FISP) sequence on the axial plane (TR 4.4 s; TE 2.2 s; average 2; flip angle 80°; matrix 256 9 256; slice number 25; slice thickness 5 mm; slice gap 15 %) and T2weighted True-FISP sequence on the coronal plane (TR 4.3 s; TE 2.15 s; average 1; slice number 35; slice thickness 4 mm; slice gap 15 %), followed by axial and coronal turbo inversion recovery in magnitude (TIRM) (TR 7,660 s; TE 100 s; flip angle 80°; matrix 256 9 256; slice number 35; slice thickness 4 mm; average 1; gap 15 %) with breath holding. DW-MRI was obtained on the axial plane with the single-shot spin-echo-planar imaging (SSEPI) technique with the following settings: TR/TE 3700/76; matrix 128 9 128; slice thickness 5 mm; slice gaps 1 mm; FOV 400 mm; averages 4; acquisition time 156 s. The protocol used in our clinic for echo-planar DWMRI is as follows: 0, 50, 400, 800 s/mm2, and ADC. In all patients, images were acquired on three planes (x, y, z) with three different b values by applying a diffusion-sensitive gradient (b = 50, b = 400, b = 800 s/mm2). Craniocaudal coverage of the DWI sequence was set by referring to the localizer image. No anticholinergic drugs were administered. An independent workstation (Leonardo console, software version 2.0; Siemens AG Medical solutions, Forchheim, Germany) was used to process the DW-MRI data and to remodel ADC maps. Signal-intensity changes of lesions were set to consider b = 800 s/mm2 as valued diffusion-weighted trace images. Lesions with high intensity on DWI and low intensity on ADC maps were considered diffusion restricted. Hyperand hypointensity were determined in comparison with normal colonic wall intensity. Measurements were obtained with circular regions of interest (ROIs), which were placed on the diffusion-restricted, thickened colonic walls by referencing previous MRI images. For ADC measurements, images were magnified, and oval ROIs were placed on the largest possible area covering the bowel wall. At least three samplings from a different part of each lesion were obtained to determine ADC values. ROI areas varied between 10 and 40 mm2 (mean 21.2 mm2). ADC quantifications were performed by two radiologists experienced with abdominopelvic imaging and blinded to the final histopathological diagnoses and previous clinical and radiological findings. DW-MRI results and ADC values were retrospectively compared with the final pathologic diagnoses. Descriptive statistics, in terms of the ADC properties emphasized, were expressed within the context of mean, standard deviation (SD), and minimum and maximum values. Unpaired t test was performed to determine whether there were any differences between patients with IBD and those with colorectal carcinoma. Receiver operating
123
8
Jpn J Radiol (2014) 32:6–13
characteristic (ROC) curve analysis was performed to determine ADC cutoff values for differentiating the inflammation group from the colorectal carcinoma group. The significance level in the measurements was 1 %, and measurements were performed using SPSS (version 13.0).
Results Twenty-seven patients were diagnosed with carcinoma, and 31 were finally diagnosed with inflammatory colonic disease by histopathologic observation. In the pathologically confirmed carcinoma group, lesions comprised nine well-differentiated, 11 intermediate-differentiated, one less-differentiated, and five mucinous adenocarcinomas, as well as one patient with colonic wall thickening due to invasion of a bladder adenocarcinoma. The distribution of 31 IBD cases was 21 UC, four CD, five nonspecific colitis other than UC-CD, and one postinflammatory polyp. In the 27 colorectal carcinoma cases, all masses had clear demarcation with increased intensity on the DW-MRI and were hypointense on ADC mapping (Fig. 1). Mean ADC value of the carcinoma lesions was 1.02 ± 0.26 9 10-3 mm2/s. For subtypes, mean ADC value of adenocarcinomas and mucinous adenocarcinomas was 0.98 9 10-3 and 1.23 9 10-3 mm/s, respectively. Mean ACD value in the 31 cases of inflammatory colonic disease was 1.53 ± 0.19 9 10-3 mm2/s (range 1.09–2.19 9 10-3 mm2/s); in the 21 UC cases 1.55 9 10-3 mm2/s (Fig. 2); in the four CD cases 1.49 9 10-3 mm2/s (Fig. 3), in the five nonspecific colitis cases 1.53 9 10-3 mm2/s (Fig. 4), and in the one case of inflamed colonic polyps 1.63 9 10-3 mm2/s (Fig. 5). Mean ADC value of colorectal carcinoma was lower than that of inflammatory colonic disease, and the difference was statistically significant (P \ 0.001). The results of ADC quantification among patient groups are summarized in Table 1. The box plot graphic evaluation of ADC value distribution among the colorectal carcinoma and IBD groups are shown in Fig. 6. The cutoff value determined by the ROC curve for differentiation of colorectal carcinoma from IBD, according to ADC values, was 1.39 9 10-3 mm2/s; sensitivity was 83.9 % and specificity 85.2 % (Fig. 7).
Discussion Radiologic diagnoses of colorectal pathologies have been made using varying modalities, such as plain films, barium enema studies, CT, and MRI. The main role of abdominal plain films is to diagnose and monitor obstruction or colitis. Barium enema remains the most common radiological
123
Fig. 1 A 65-year-old male patient with circumferential sigmoid wall thickening: a Lesion heterogeneously hypointense on T2-weighted true fast imaging with steady-state precession (FISP) images. b Hyperintensity on diffusion-weighted magnetic resonance imaging (DW-MRI) (restricted diffusion). c Isohypointensity on apparent diffusion coefficient (ADC) mapping with a mean ADC value of 1.39 9 10-3 mm2/s (b = 800 s/mm2). Pathology diagnosis was intermediate differentiated adenocarcinoma
Jpn J Radiol (2014) 32:6–13
Fig. 2 A 32-year-old male patient: a Diffusion restriction of the sigmoid colon was detected with diffusion-weighted magnetic resonance imaging (DW-MRI) (arrow). b Hypointensity on apparent diffusion coefficient (ADC) mapping, with a mean ADC value of 1.67 9 10-3 mm2/s (b = 800 s/mm2). Pathology diagnosis was ulcerative colitis
technique for colonic examination and is the criterion standard for defining fine mucosal detail and general colonic configuration and calibration. These conventional imaging methods are now used to a limited extent due to the development of more sophisticated techniques, such as CT and MRI. Colonic wall lesions with concomitant pathologies, such as adjacent infiltration, lymph node– visceral metastasis, perforation, and penetration, can be evaluated by CT and MRI. Many pathognomonic radiological findings facilitate the diagnosis of malignant colorectal lesions and IBD, both well-known causes of colonic wall thickening; however, for some individuals, only a pathologic examination can provide a final diagnosis. It can be challenging to differentiate inflammatory colonic involvements, such as tuberculosis, CD, and UC, from neoplasms, because primary or metastatic carcinomas can mimic inflammatory involvements. DW-MRI is a sensitive, functional imaging modality for detecting the diffusion process of molecules, mainly water in tissue, in vivo and noninvasively. Tissue cell density can
9
Fig. 3 A 34-year-old female patient: a Diffusion restriction of the descending colon was detected with diffusion-weighted magnetic resonance imaging (DW-MRI) (arrow). b Isointensity on apparent diffusion coefficient (ADC) mapping, with a mean ADC value of 1.50 9 10-3 mm2/s (b = 800 s/mm2). Pathology diagnosis was compatible with Crohn’s disease
be indicative of malignancy and even of tumor aggressiveness; thus, diffusion curves that decay quickly and large ADC values typical indicate healthy tissue and benign pathologic processes with large extracellular space and little cellularity, whereas curves that decay slowly and small ADC values indicate malignancy or hypercellularity [5]. DW-MRI, formerly used in neuroradiology, has recently been used in clinical applications as a new functional imaging method and as the preferred method in abdominal examinations due to the development of ultrafast sequences, such as EPI. Although previous studies using DW-MRI have been conducted using breath holding, Thoeny et al. [6] conducted a study during normal respiration and observed no remarkable movement artifacts. In our study, the SS-EPI technique was used without breath holding during, and appropriate examinations were performed on patients who were not able to hold their breath for long periods of time. Ichikawa et al. [7] reported that in abdominal DW-MRI examinations, using b values [400 mm2/s provides more definitive ADC measurements. The use of a high b value provides more definitive
123
10
Fig. 4 A 77-year-old female patient: a Diffusion restriction of the sigmoid colon was determined with diffusion-weighted magnetic resonance imaging (DW-MRI) (arrow). b Hypointensity on apparent diffusion coefficient (ADC) mapping, with a mean ADC value of 1.65 9 10-3 (b = 800 s/mm2). Pathology diagnosis was nonspecific colitis
information about the molecular diffusion of water by eliminating microcirculatory and perfusion effects of blood in tissue capillary beds. In our study, we set b values at 50, 400, and 800 s/mm2 based on previous findings. Satisfactory diagnostic quality was obtained, and we recommend 800 s/mm2 as a satisfactory b value for colorectal DWMRI examinations. A few published reports evaluated the diagnostic effectiveness of DW-MRI in bowel lesions. Nasu et al. [8] examined 42 colorectal malignant masses using b values of 0, 250, 500, and 1,000 s/mm2. All tumors were hyperintense on DW-MRI and ADC values were compared with urine findings. The mean ADC value of colorectal carcinoma was 1.05 ± 0.1 9 10-3 mm2/s, which was comparable with that in our study. Nasu et al. concluded that the sensitivity-encoding (SENSE) DW-MRI affords not only high detectability of colorectal cancer but also a high
123
Jpn J Radiol (2014) 32:6–13
Fig. 5 A 17-year-old male patient: a Nodular diffusion restriction areas of the rectum with diffusion-weighted magnetic resonance imaging (DW-MRI) (arrow). b Hypointensity on apparent diffusion coefficient (ADC) mapping, with a mean ADC value of 1.51 9 10-3 mm2/s (b = 800 s/mm2). Pathology diagnosis was inflammatory polyps
possibility of reflecting tumor cellularity. A total of 23 patients, consisting of 14 cases of rectosigmoid adenocarcinomas and nine of IBD, were retrospectively reviewed in a preliminary report by Kilickesmez et al. [9], with 30 healthy participants as a control group. Mean ADC values between groups were significantly different (P \ 0.01). The cutoff value for the determination of carcinomas, 1.14 9 10-3 mm2/s, had a sensitivity and specificity of 93.3 % and 93.3 %, respectively. Cutoff and mean ADC values of colorectal carcinomas in our study were similar to those found by Kilickesmez et al. However, control group ADC value quantification was not performed in our study because of the difficulty in generating an ROI that contained only the normal colonic wall, which created the possibility of a partial volume artifact that could have affected ADC values. Kiyru et al. and Oto et al. [10, 11] demonstrated inflammatory changes in the gastrointestinal
Jpn J Radiol (2014) 32:6–13
11
Table 1 Mean apparent diffusion coeficient (ADC) values of the colorectal carcinoma and inflammatory bowel disease (IBD) groups correlated with histopathologic diagnosis Patient group (n = 58)
Histopathologic biopsy results
Mean ADC value (910-3 mm2/s)
Colorectal carcinoma (n = 27)
Colorectal adenocarcinoma (n = 21)
0.98
Colorectal mucinous adenocarcinoma (n = 5)
1.23
Bladder adenocarcinoma (n = 1)
0.98
Ulcerative colitis (n = 21)
1.55
Crohn’s disease (n = 4) Nonspecific colitis other than IBD (n = 5)
1.49 1.53
Postinflammatory polyp (n = 1)
1.63
IBD (n = 31)
1.02 ± 0.26 (0.56–1.55)
1.53 ± 0.19 (1.09–2.19)
Fig. 7 Receiver operating characteristic (ROC) curve for apparent diffusion coefficient (ADC) value–based differentiation of colorectal carcinoma and inflammatory bowel disease (IBD) groups (83.9 % sensitivity, 85.2 % specificity; cutoff value 1.39 9 10-3 mm2/s)
Fig. 6 Apparent diffusion coefficient (ADC) values among colorectal carcinoma and inflammatory bowel disease (IBD) groups
system caused by CD using DW-MRI. They found that in 31 CD patients, sensitivity, specificity, and accuracy for detecting disease-active areas upon visual assessment were 86.0 %, 81.4 %, and 82.4 %, respectively [10]. Mean ADC value in the disease-active area was less than that in the disease-inactive area (P \ 0.001). Sensitivity, specificity, and accuracy rates were 86.4 %, 97.4 %, and 93.3 % in the small intestine compared with 85.7 %, 75.7 %, and 77.0 % in the colon, respectively. Specificity and accuracy rates of ADC values for the small intestine were higher than those for the colon. High signal intensities were obtained by DWMRI in a few normal colonic segments, and this entity had a reduced specificity rate for the colonic wall. Upon visual assessment, sensitivity, specificity, and accuracy for detecting disease-active segments were 86.0 %, 81.4 %, and 82.4 %, respectively. Oto et al. [11] reported that the
mean ADC value was 1.59 ± 0.45 9 10-3 mm2/s in 19 inflamed intestinal segments of 11 patients with CD and 2.74 ± 0.68 9 10-3 mm2/s in normal intestinal segments (P \ 0.0001). Cutoff value was 2.0 9 10-3 mm2/s, with 84 % sensitivity and 91 % specificity. Kilic¸kesmez et al. [12] reported on the reliability of DW-MRI for assessing inflammatory activity in UC and found ADC values for the sigmoid colon were similar among patients in active, subacute, and remission phases of UC (P = 0.472). ADC values of the rectum were different among patients in active, subacute, and remission phases (P = 0.009). Rectal ADC values of patients in remission were greater than rectum ADC values of patients in active (P = 0.009) and subacute (P = 0.007) phases. Based on these findings, the authors concluded that DW-MRI is a useful method for evaluating disease activity in the rectum. The low ADC values of the colonic segments with active disease were related to inflammation, fibrosis, and an increased intestinal-wall cell volume. Kılıc¸kesmez et al. [9] reported preliminary findings of a comparison of DW-MRI and ADC values of colorectal adenocarcinomas, IBD, and normal controls. Mean ADC values of the control and IBD groups were 1.47 ± 0.19 9 10-3 and 1.37 ± 0.12 9 10-3 mm2/s, respectively. Nine IBD patients were included—two in the active and seven in the inactive phase—with mean ADC values of 1.21 ± 0.08 9 10-3 and 1.42 ± 0.09 9 10-3, respectively. ADC values of the rectosigmoid carcinoma and IBD groups were significantly different (P \ 0.01).
123
12
Jpn J Radiol (2014) 32:6–13
Table 2 Summary of results of this and previous studies evaluating the feasibility of diffusion-weighted magnetic resonance imaging (DWMRI) in colorectal carcinoma and/or inflammatory bowel disease (IBD) Mean ADC values (910-3 mm2/s) Colorectal carcinoma
Inflammatory bowel disease Crohn’s disease
Control group Ulcerative colitis
Nasu et al. [8] (n = 42)
1.05 ± 0.1 (rectosigmoid region)
–
3.35 ± 0.49 (measured from patient’s urine)
Kilickesmez et al. [9]
0.97 ± 0.14 (n = 14) (rectosigmoid region)
1.37 ± 0.12 (n = 9) (rectosigmoid region)
1.47 ± 0.19 (n = 30)
Oto et al. [11] (n = 11)
–
1.59 ± 0.45 (proven inflamed segments of terminal ileum and colon)
–
2.74 ± 0.68 (normal segments of terminal ileum and colon)
Kiyru et al. [10] (n = 31)
–
1.61 ± 0.44 (small bowel)
–
Kilickesmez et al. [9] (n = 28)
–
2.56 ± 0.51 (in small bowel)
1.52 ± 0.43 (large bowel)
2.31 ± 0.59 (in large bowel)
–
Rectal region: 1.08 ± 0.14 (active phase)
–
1.13 ± 0.23 (subacute phase) 1.29 ± 0.17 (remission phase) Our study
1.02 ± 0.26 (n = 27)
1.53 ± 0.19 (IBD with isolated colonic wall involvement)
0.98 (n = 22) colorectal adenocarcinoma
1.49 (n = 4) (colonic involvement solely)
–
1.55 (n = 21) (colonic involvement solely)
1.23 (n = 5) colorectal mucinous adenocarcinoma
In our study, the mean ADC value was 1.49 9 10-3 mm2/s in segments with CD, similar to the studies published by Kiyru et al. All IBD cases in our study were in the active or subacute phase. Differently from these previously published studies, our study included patients with nonspecific colitis other than IBD and postinflammatory colonic polyps from among the IBD population. Results of our and previous studies are summarized in Table 2. In our study, the cutoff ADC value for DW-MRI to contribute to the differential diagnosis of colorectal carcinoma from IBD was established. Additional information provided by DW-MRI has the potential to contribute to the diagnostic precision of individual cases, especially when initial disease manifestations and endoscopic biopsy procedures are insufficient for accurate diagnosis. Thus, DWMRI would be beneficial if used in abdominal MRI examinations of such cases with diagnostic difficulty. Despite its merits, the disadvantages of DW-MRI, such as high cost, low spatial image resolutions, and long examination time—as well as the inability of patients with claustrophobia to undergo imaging—must be considered. One limitation of our study was that it was designed retrospectively; thus, results may have been affected by
123
selection bias. Moreover, follow-up data for each patient were not provided, and comparisons between MRI and CT or barium enema study were not performed.
Conclusion We conclude that DW-MRI has the potential to differentiate colorectal carcinomas from IBD, and the cutoff rate for DW-MRI ADC value suggested in this study is an additional indicative figure for more accurate diagnoses in patients for whom clinical and conventional radiological examinations have failed to provide a final diagnosis. Conflict of interest
None.
References 1. Halligan S, Robinson PJA. The large bowel. In: Sutton D, Robinson PJA, Jenkins JPR, Whitehouse RW, Allan PL, Wilde P, Stevens JM, editors. Textbook of radiology and imaging. 7th ed. London: Elsevier; 2003. p. 641. 2. Ng DC, Kwok SY, Cheng Y, Chung CC, Li MK. Colonic amoebic abscess mimicking carcinoma of the colon. Hong Kong Med J. 2006;12(1):71–3.
Jpn J Radiol (2014) 32:6–13 3. Gadwood KA, Bedetti CD, Herbert DL. Colonic tuberculosis mimicking annular carcinoma: report of a case. Dis Colon Rectum. 1981;24(5):395–8. 4. Yu XQ, Wang ZJ, Zhao B, Ma HC, Jin ML. Extensively infiltrating colon adenocarcinoma similar to Crohn’s disease: a case report. Clin Res Hepatol Gastroenterol. 2012;36(4):63–8. 5. Ichikawa T, Erturk SM, Motosugi U, Sou H, Iino H, Araki T, Fujii H. High B-value diffusion-weighted MRI in colorectal cancer. Am J Roentgenol. 2006;187:181–4. 6. Thoeny HC, De Keyzer F, Oyen RH, Peeters RR. Diffusionweighted MR imaging of kidneys in healthy volunteers and patients with parenchymal diseases: initial experience. Radiology. 2005;235:911–7. 7. Ichikawa T, Haradome H, Hachiya J, Nitatori T, Araki T. Diffusion-weighted MR imaging with single-shot echo-planar imaging in the upper abdomen: preliminary clinical experience in 61 patients. Abdom Imaging. 1999;24:456–61. 8. Nasu K, Kuroki Y, Kuroki S, Murakami M, Nawano S, Moriyama N. Diffusion-weighted single shot echo planar imaging of
13
9.
10.
11.
12.
colorectal cancer using a sensitivity-encoding technique. Jpn J Clin Oncol. 2004;34:620–6. Kilickesmez O, Atilla S, Soylu A, Tasdelen N, Bayramoglu S, Cimilli T, Gurmen N. Diffusion-weighted imaging of the rectosigmoid colon: preliminary findings. J Comput Assist Tomogr. 2009;33:863–6. Kiryu S, Dotanuki K, Takao H, Watanabe M, Inoue Y, Takazoe M, et al. Free-breathing diffusion-weighted imaging for the assessment of inflammatory activity in Crohn’s disease. J Magn Reson Imaging. 2009;29:880–6. Oto A, Zhu F, Kulkarni K, Karczmar GS, Turner JR, Rubin D. Evaluation of diffusion-weighted MR for detection of bowel inflammation in patients with Crohn’s disease. Acad Radiol. 2009;16:597–603. Kılıc¸kesmez O, Soylu A, Yas¸ ar N, Demirbas¸ T, Dolapc¸ıog˘lu C, Poturog˘lu S, et al. Is quantitative diffusion-weighted MRI a reliable method in the assessment of the inflammatory activity in ulcerative colitis? Diagn Interv Radiol. 2010;16:293–8.
123
ORIGINAL ARTICLE
Value of DW-MRI ADC quantification of colonic wall lesions in differentiation of inflammatory bowel disease and colorectal carcinoma Serhat Avcu • Mehmet Deniz Bulut • Alpaslan Yavuz • Aydın Bora • Mehmet Beyazal
Received: 25 February 2013 / Accepted: 20 October 2013 / Published online: 1 November 2013 Ó Japan Radiological Society 2013
Abstract Purpose The purpose of our study was to assess the efficiency of diffusion-weighted magnetic resonance imaging (DW-MRI) and the quantification of apparent diffusion coefficient (ADC) values in differentiating colorectal carcinoma from colonic inflammatory bowel disease (IBD) in cases with isolated colonic wall lesions and uncertain clinical and radiologic diagnostic criteria. Methods The study comprised 58 patients with segmental or focal isolated colonic wall thickening. All lacked satisfactory clinical–radiological findings for etiology determination. The mean ADC values of the thickened colonic walls were retrospectively compared with final histopathologic diagnoses. Receiver operating characteristic (ROC) curve analysis was used to determine the ADC cutoff value for differentiation. Results Mean ADC value in the colorectal carcinoma group was significantly lower than that in the IBD group: n = 27, 1.02 ± 0.26 9 10-3 mm2/s; and n = 31, 1.53 ± 0.19 9 10-3 mm2/s, respectively (P \ 0.001). Cutoff ADC value for differentiating colorectal carcinoma from IBD was calculated as 1.39 9 10-3 mm2/s, with 83.9 % sensitivity and 85.2 % specificity. Conclusion ADC measurement of the involved colonic wall segments with DW-MRI has the potential to differentiate isolated colonic IBD from colorectal carcinoma in
S. Avcu A. Yavuz (&) A. Bora M. Beyazal Radiology Department, Yuzuncu Yil University, Ercis Yolu, Van, Turkey e-mail: [email protected] M. D. Bulut Van Goverment Education and Research Hospital, Edremit Yolu, Van, Turkey
123
cases in which clinical–radiologic findings are insufficient for a definitive diagnosis. Keywords MRI DW-MRI Colorectal carcinoma Inflammatory bowel disease
Introduction Ulcerative colitis (UC) and Crohn’s disease (CD) are the primary forms of idiopathic inflammatory bowel disease (IBD), and aberrant host responses to enteric environmental agents are the acknowledged etiological factors for individuals who are genetically susceptible to IBD. Most colorectal carcinomas are believed to arise from sporadic adenomas, but some cancers arise from nonpolypoid dysplastic mucosa, such as the types of cancer occurring with IBD [1]. The majority of neoplasms of the gastrointestinal (GI) tract appear as eccentric tumor masses, whereas some spread locally in a plaque-like infiltrative fashion and are characterized by a lack of marked focal elevation of the colonic wall; thus, differentiating between neoplasms of the GI tract can be difficult, especially for benign inflammatory entities. However, inflammatory pathologies, such as UC and CD, occur with segmental or diffuse thickening of the bowel wall. Double-contrast barium enema is highly accurate in revealing early changes, extent, and severity of IBD; however, computed tomography (CT) is superior to barium studies and endoscopy for diagnosing extramural complications. Magnetic resonance imaging (MRI) plays a similar role to that of CT; additionally, pelvic imaging with fat-suppressed techniques combined with sequences that highlight fluid can emphasize collections to a greater degree than CT. Although colorectal carcinoma has typical features, differential diagnosis can be problematic if the
Jpn J Radiol (2014) 32:6–13
radiological morphology is unusual. Inflammation due to CD, amoebiasis, and tuberculosis can also mimic a primary cancer [2, 3], whereas infiltrating colonic carcinomas can mimic inflammatory processes and vice versa [4]. In this study, we aimed to evaluate the effectiveness of the apparent diffusion coefficient (ADC) value obtained by diffusion-weighted (DW)-MRI in a limited group of individuals for differentiating colorectal carcinomas from IBD when conventional CT and/or MRI findings are insufficient for an accurate diagnosis.
Materials and methods Sixty-five patients (31 men and 34 women; mean age 63.18 ± 14.32 years) with segmental colonic wall thickening determined by CT or conventional MRI, were assessed. These patients lacked satisfactory clinical and radiological findings to determine the etiology of their disease. The bowel wall was considered to be thickened when it was [3 mm and the increased wall thickness had not progressed to the next segment of the colon (limited to one segment, such as rectum, sigmoid, ascending, transverse, or descendant). Patients who had initial clinical and imaging findings characteristic of IBD (disease onset between 15 and 40 years, recurrent episodes of diarrhea, low-grade fever, malabsorption, and marked thickening of the small intestine) or of colorectal carcinoma (sudden weight loss, distant metastasis, mesenteric malignant lymphadenopathy, massive invasion of the surrounding tissue, focal irregular colonic thickening) were excluded. Three patients who required transportation to another institution and four who had final diagnoses other than IBD or colorectal carcinoma (three with colonic ischemia, one with polyposis coli) were also excluded. Fifty-eight patients (28 men and 30 women; mean age 62.71 ± 11.23 years) were underwent routine abdominal MRI examination including DW-MRI sequences. All patients with isolated colonic wall thickening and final histopathologic diagnoses of IBD or colorectal carcinoma were reviewed to determine the feasibility of DW-MRI for initial differentiation of these entities. Pathology confirmation for the final diagnosis was obtained in 53 cases by endoscopic biopsy and in five by surgery (four carcinoma, one IBD; endoscopic process was not successful to sample the lesions). Conventional and DW-MR imaging findings and final endoscopic pathology results were assessed. A 1.5-T MRI system (Siemens Magnetom Symphony, Erlangen, Germany) with a phased-array body coil was used for examinations. The gradient force of the superconductive (niobium–titanium) magnet was 30 mT/m, and the maximum field-of-view (FOV) width was 400 mm. Before DW-MRI examination, conventional abdominal
7
MRI was performed with the following settings: T2weighted true fast imaging with a steady-state precession (true-FISP) sequence on the axial plane (TR 4.4 s; TE 2.2 s; average 2; flip angle 80°; matrix 256 9 256; slice number 25; slice thickness 5 mm; slice gap 15 %) and T2weighted True-FISP sequence on the coronal plane (TR 4.3 s; TE 2.15 s; average 1; slice number 35; slice thickness 4 mm; slice gap 15 %), followed by axial and coronal turbo inversion recovery in magnitude (TIRM) (TR 7,660 s; TE 100 s; flip angle 80°; matrix 256 9 256; slice number 35; slice thickness 4 mm; average 1; gap 15 %) with breath holding. DW-MRI was obtained on the axial plane with the single-shot spin-echo-planar imaging (SSEPI) technique with the following settings: TR/TE 3700/76; matrix 128 9 128; slice thickness 5 mm; slice gaps 1 mm; FOV 400 mm; averages 4; acquisition time 156 s. The protocol used in our clinic for echo-planar DWMRI is as follows: 0, 50, 400, 800 s/mm2, and ADC. In all patients, images were acquired on three planes (x, y, z) with three different b values by applying a diffusion-sensitive gradient (b = 50, b = 400, b = 800 s/mm2). Craniocaudal coverage of the DWI sequence was set by referring to the localizer image. No anticholinergic drugs were administered. An independent workstation (Leonardo console, software version 2.0; Siemens AG Medical solutions, Forchheim, Germany) was used to process the DW-MRI data and to remodel ADC maps. Signal-intensity changes of lesions were set to consider b = 800 s/mm2 as valued diffusion-weighted trace images. Lesions with high intensity on DWI and low intensity on ADC maps were considered diffusion restricted. Hyperand hypointensity were determined in comparison with normal colonic wall intensity. Measurements were obtained with circular regions of interest (ROIs), which were placed on the diffusion-restricted, thickened colonic walls by referencing previous MRI images. For ADC measurements, images were magnified, and oval ROIs were placed on the largest possible area covering the bowel wall. At least three samplings from a different part of each lesion were obtained to determine ADC values. ROI areas varied between 10 and 40 mm2 (mean 21.2 mm2). ADC quantifications were performed by two radiologists experienced with abdominopelvic imaging and blinded to the final histopathological diagnoses and previous clinical and radiological findings. DW-MRI results and ADC values were retrospectively compared with the final pathologic diagnoses. Descriptive statistics, in terms of the ADC properties emphasized, were expressed within the context of mean, standard deviation (SD), and minimum and maximum values. Unpaired t test was performed to determine whether there were any differences between patients with IBD and those with colorectal carcinoma. Receiver operating
123
8
Jpn J Radiol (2014) 32:6–13
characteristic (ROC) curve analysis was performed to determine ADC cutoff values for differentiating the inflammation group from the colorectal carcinoma group. The significance level in the measurements was 1 %, and measurements were performed using SPSS (version 13.0).
Results Twenty-seven patients were diagnosed with carcinoma, and 31 were finally diagnosed with inflammatory colonic disease by histopathologic observation. In the pathologically confirmed carcinoma group, lesions comprised nine well-differentiated, 11 intermediate-differentiated, one less-differentiated, and five mucinous adenocarcinomas, as well as one patient with colonic wall thickening due to invasion of a bladder adenocarcinoma. The distribution of 31 IBD cases was 21 UC, four CD, five nonspecific colitis other than UC-CD, and one postinflammatory polyp. In the 27 colorectal carcinoma cases, all masses had clear demarcation with increased intensity on the DW-MRI and were hypointense on ADC mapping (Fig. 1). Mean ADC value of the carcinoma lesions was 1.02 ± 0.26 9 10-3 mm2/s. For subtypes, mean ADC value of adenocarcinomas and mucinous adenocarcinomas was 0.98 9 10-3 and 1.23 9 10-3 mm/s, respectively. Mean ACD value in the 31 cases of inflammatory colonic disease was 1.53 ± 0.19 9 10-3 mm2/s (range 1.09–2.19 9 10-3 mm2/s); in the 21 UC cases 1.55 9 10-3 mm2/s (Fig. 2); in the four CD cases 1.49 9 10-3 mm2/s (Fig. 3), in the five nonspecific colitis cases 1.53 9 10-3 mm2/s (Fig. 4), and in the one case of inflamed colonic polyps 1.63 9 10-3 mm2/s (Fig. 5). Mean ADC value of colorectal carcinoma was lower than that of inflammatory colonic disease, and the difference was statistically significant (P \ 0.001). The results of ADC quantification among patient groups are summarized in Table 1. The box plot graphic evaluation of ADC value distribution among the colorectal carcinoma and IBD groups are shown in Fig. 6. The cutoff value determined by the ROC curve for differentiation of colorectal carcinoma from IBD, according to ADC values, was 1.39 9 10-3 mm2/s; sensitivity was 83.9 % and specificity 85.2 % (Fig. 7).
Discussion Radiologic diagnoses of colorectal pathologies have been made using varying modalities, such as plain films, barium enema studies, CT, and MRI. The main role of abdominal plain films is to diagnose and monitor obstruction or colitis. Barium enema remains the most common radiological
123
Fig. 1 A 65-year-old male patient with circumferential sigmoid wall thickening: a Lesion heterogeneously hypointense on T2-weighted true fast imaging with steady-state precession (FISP) images. b Hyperintensity on diffusion-weighted magnetic resonance imaging (DW-MRI) (restricted diffusion). c Isohypointensity on apparent diffusion coefficient (ADC) mapping with a mean ADC value of 1.39 9 10-3 mm2/s (b = 800 s/mm2). Pathology diagnosis was intermediate differentiated adenocarcinoma
Jpn J Radiol (2014) 32:6–13
Fig. 2 A 32-year-old male patient: a Diffusion restriction of the sigmoid colon was detected with diffusion-weighted magnetic resonance imaging (DW-MRI) (arrow). b Hypointensity on apparent diffusion coefficient (ADC) mapping, with a mean ADC value of 1.67 9 10-3 mm2/s (b = 800 s/mm2). Pathology diagnosis was ulcerative colitis
technique for colonic examination and is the criterion standard for defining fine mucosal detail and general colonic configuration and calibration. These conventional imaging methods are now used to a limited extent due to the development of more sophisticated techniques, such as CT and MRI. Colonic wall lesions with concomitant pathologies, such as adjacent infiltration, lymph node– visceral metastasis, perforation, and penetration, can be evaluated by CT and MRI. Many pathognomonic radiological findings facilitate the diagnosis of malignant colorectal lesions and IBD, both well-known causes of colonic wall thickening; however, for some individuals, only a pathologic examination can provide a final diagnosis. It can be challenging to differentiate inflammatory colonic involvements, such as tuberculosis, CD, and UC, from neoplasms, because primary or metastatic carcinomas can mimic inflammatory involvements. DW-MRI is a sensitive, functional imaging modality for detecting the diffusion process of molecules, mainly water in tissue, in vivo and noninvasively. Tissue cell density can
9
Fig. 3 A 34-year-old female patient: a Diffusion restriction of the descending colon was detected with diffusion-weighted magnetic resonance imaging (DW-MRI) (arrow). b Isointensity on apparent diffusion coefficient (ADC) mapping, with a mean ADC value of 1.50 9 10-3 mm2/s (b = 800 s/mm2). Pathology diagnosis was compatible with Crohn’s disease
be indicative of malignancy and even of tumor aggressiveness; thus, diffusion curves that decay quickly and large ADC values typical indicate healthy tissue and benign pathologic processes with large extracellular space and little cellularity, whereas curves that decay slowly and small ADC values indicate malignancy or hypercellularity [5]. DW-MRI, formerly used in neuroradiology, has recently been used in clinical applications as a new functional imaging method and as the preferred method in abdominal examinations due to the development of ultrafast sequences, such as EPI. Although previous studies using DW-MRI have been conducted using breath holding, Thoeny et al. [6] conducted a study during normal respiration and observed no remarkable movement artifacts. In our study, the SS-EPI technique was used without breath holding during, and appropriate examinations were performed on patients who were not able to hold their breath for long periods of time. Ichikawa et al. [7] reported that in abdominal DW-MRI examinations, using b values [400 mm2/s provides more definitive ADC measurements. The use of a high b value provides more definitive
123
10
Fig. 4 A 77-year-old female patient: a Diffusion restriction of the sigmoid colon was determined with diffusion-weighted magnetic resonance imaging (DW-MRI) (arrow). b Hypointensity on apparent diffusion coefficient (ADC) mapping, with a mean ADC value of 1.65 9 10-3 (b = 800 s/mm2). Pathology diagnosis was nonspecific colitis
information about the molecular diffusion of water by eliminating microcirculatory and perfusion effects of blood in tissue capillary beds. In our study, we set b values at 50, 400, and 800 s/mm2 based on previous findings. Satisfactory diagnostic quality was obtained, and we recommend 800 s/mm2 as a satisfactory b value for colorectal DWMRI examinations. A few published reports evaluated the diagnostic effectiveness of DW-MRI in bowel lesions. Nasu et al. [8] examined 42 colorectal malignant masses using b values of 0, 250, 500, and 1,000 s/mm2. All tumors were hyperintense on DW-MRI and ADC values were compared with urine findings. The mean ADC value of colorectal carcinoma was 1.05 ± 0.1 9 10-3 mm2/s, which was comparable with that in our study. Nasu et al. concluded that the sensitivity-encoding (SENSE) DW-MRI affords not only high detectability of colorectal cancer but also a high
123
Jpn J Radiol (2014) 32:6–13
Fig. 5 A 17-year-old male patient: a Nodular diffusion restriction areas of the rectum with diffusion-weighted magnetic resonance imaging (DW-MRI) (arrow). b Hypointensity on apparent diffusion coefficient (ADC) mapping, with a mean ADC value of 1.51 9 10-3 mm2/s (b = 800 s/mm2). Pathology diagnosis was inflammatory polyps
possibility of reflecting tumor cellularity. A total of 23 patients, consisting of 14 cases of rectosigmoid adenocarcinomas and nine of IBD, were retrospectively reviewed in a preliminary report by Kilickesmez et al. [9], with 30 healthy participants as a control group. Mean ADC values between groups were significantly different (P \ 0.01). The cutoff value for the determination of carcinomas, 1.14 9 10-3 mm2/s, had a sensitivity and specificity of 93.3 % and 93.3 %, respectively. Cutoff and mean ADC values of colorectal carcinomas in our study were similar to those found by Kilickesmez et al. However, control group ADC value quantification was not performed in our study because of the difficulty in generating an ROI that contained only the normal colonic wall, which created the possibility of a partial volume artifact that could have affected ADC values. Kiyru et al. and Oto et al. [10, 11] demonstrated inflammatory changes in the gastrointestinal
Jpn J Radiol (2014) 32:6–13
11
Table 1 Mean apparent diffusion coeficient (ADC) values of the colorectal carcinoma and inflammatory bowel disease (IBD) groups correlated with histopathologic diagnosis Patient group (n = 58)
Histopathologic biopsy results
Mean ADC value (910-3 mm2/s)
Colorectal carcinoma (n = 27)
Colorectal adenocarcinoma (n = 21)
0.98
Colorectal mucinous adenocarcinoma (n = 5)
1.23
Bladder adenocarcinoma (n = 1)
0.98
Ulcerative colitis (n = 21)
1.55
Crohn’s disease (n = 4) Nonspecific colitis other than IBD (n = 5)
1.49 1.53
Postinflammatory polyp (n = 1)
1.63
IBD (n = 31)
1.02 ± 0.26 (0.56–1.55)
1.53 ± 0.19 (1.09–2.19)
Fig. 7 Receiver operating characteristic (ROC) curve for apparent diffusion coefficient (ADC) value–based differentiation of colorectal carcinoma and inflammatory bowel disease (IBD) groups (83.9 % sensitivity, 85.2 % specificity; cutoff value 1.39 9 10-3 mm2/s)
Fig. 6 Apparent diffusion coefficient (ADC) values among colorectal carcinoma and inflammatory bowel disease (IBD) groups
system caused by CD using DW-MRI. They found that in 31 CD patients, sensitivity, specificity, and accuracy for detecting disease-active areas upon visual assessment were 86.0 %, 81.4 %, and 82.4 %, respectively [10]. Mean ADC value in the disease-active area was less than that in the disease-inactive area (P \ 0.001). Sensitivity, specificity, and accuracy rates were 86.4 %, 97.4 %, and 93.3 % in the small intestine compared with 85.7 %, 75.7 %, and 77.0 % in the colon, respectively. Specificity and accuracy rates of ADC values for the small intestine were higher than those for the colon. High signal intensities were obtained by DWMRI in a few normal colonic segments, and this entity had a reduced specificity rate for the colonic wall. Upon visual assessment, sensitivity, specificity, and accuracy for detecting disease-active segments were 86.0 %, 81.4 %, and 82.4 %, respectively. Oto et al. [11] reported that the
mean ADC value was 1.59 ± 0.45 9 10-3 mm2/s in 19 inflamed intestinal segments of 11 patients with CD and 2.74 ± 0.68 9 10-3 mm2/s in normal intestinal segments (P \ 0.0001). Cutoff value was 2.0 9 10-3 mm2/s, with 84 % sensitivity and 91 % specificity. Kilic¸kesmez et al. [12] reported on the reliability of DW-MRI for assessing inflammatory activity in UC and found ADC values for the sigmoid colon were similar among patients in active, subacute, and remission phases of UC (P = 0.472). ADC values of the rectum were different among patients in active, subacute, and remission phases (P = 0.009). Rectal ADC values of patients in remission were greater than rectum ADC values of patients in active (P = 0.009) and subacute (P = 0.007) phases. Based on these findings, the authors concluded that DW-MRI is a useful method for evaluating disease activity in the rectum. The low ADC values of the colonic segments with active disease were related to inflammation, fibrosis, and an increased intestinal-wall cell volume. Kılıc¸kesmez et al. [9] reported preliminary findings of a comparison of DW-MRI and ADC values of colorectal adenocarcinomas, IBD, and normal controls. Mean ADC values of the control and IBD groups were 1.47 ± 0.19 9 10-3 and 1.37 ± 0.12 9 10-3 mm2/s, respectively. Nine IBD patients were included—two in the active and seven in the inactive phase—with mean ADC values of 1.21 ± 0.08 9 10-3 and 1.42 ± 0.09 9 10-3, respectively. ADC values of the rectosigmoid carcinoma and IBD groups were significantly different (P \ 0.01).
123
12
Jpn J Radiol (2014) 32:6–13
Table 2 Summary of results of this and previous studies evaluating the feasibility of diffusion-weighted magnetic resonance imaging (DWMRI) in colorectal carcinoma and/or inflammatory bowel disease (IBD) Mean ADC values (910-3 mm2/s) Colorectal carcinoma
Inflammatory bowel disease Crohn’s disease
Control group Ulcerative colitis
Nasu et al. [8] (n = 42)
1.05 ± 0.1 (rectosigmoid region)
–
3.35 ± 0.49 (measured from patient’s urine)
Kilickesmez et al. [9]
0.97 ± 0.14 (n = 14) (rectosigmoid region)
1.37 ± 0.12 (n = 9) (rectosigmoid region)
1.47 ± 0.19 (n = 30)
Oto et al. [11] (n = 11)
–
1.59 ± 0.45 (proven inflamed segments of terminal ileum and colon)
–
2.74 ± 0.68 (normal segments of terminal ileum and colon)
Kiyru et al. [10] (n = 31)
–
1.61 ± 0.44 (small bowel)
–
Kilickesmez et al. [9] (n = 28)
–
2.56 ± 0.51 (in small bowel)
1.52 ± 0.43 (large bowel)
2.31 ± 0.59 (in large bowel)
–
Rectal region: 1.08 ± 0.14 (active phase)
–
1.13 ± 0.23 (subacute phase) 1.29 ± 0.17 (remission phase) Our study
1.02 ± 0.26 (n = 27)
1.53 ± 0.19 (IBD with isolated colonic wall involvement)
0.98 (n = 22) colorectal adenocarcinoma
1.49 (n = 4) (colonic involvement solely)
–
1.55 (n = 21) (colonic involvement solely)
1.23 (n = 5) colorectal mucinous adenocarcinoma
In our study, the mean ADC value was 1.49 9 10-3 mm2/s in segments with CD, similar to the studies published by Kiyru et al. All IBD cases in our study were in the active or subacute phase. Differently from these previously published studies, our study included patients with nonspecific colitis other than IBD and postinflammatory colonic polyps from among the IBD population. Results of our and previous studies are summarized in Table 2. In our study, the cutoff ADC value for DW-MRI to contribute to the differential diagnosis of colorectal carcinoma from IBD was established. Additional information provided by DW-MRI has the potential to contribute to the diagnostic precision of individual cases, especially when initial disease manifestations and endoscopic biopsy procedures are insufficient for accurate diagnosis. Thus, DWMRI would be beneficial if used in abdominal MRI examinations of such cases with diagnostic difficulty. Despite its merits, the disadvantages of DW-MRI, such as high cost, low spatial image resolutions, and long examination time—as well as the inability of patients with claustrophobia to undergo imaging—must be considered. One limitation of our study was that it was designed retrospectively; thus, results may have been affected by
123
selection bias. Moreover, follow-up data for each patient were not provided, and comparisons between MRI and CT or barium enema study were not performed.
Conclusion We conclude that DW-MRI has the potential to differentiate colorectal carcinomas from IBD, and the cutoff rate for DW-MRI ADC value suggested in this study is an additional indicative figure for more accurate diagnoses in patients for whom clinical and conventional radiological examinations have failed to provide a final diagnosis. Conflict of interest
None.
References 1. Halligan S, Robinson PJA. The large bowel. In: Sutton D, Robinson PJA, Jenkins JPR, Whitehouse RW, Allan PL, Wilde P, Stevens JM, editors. Textbook of radiology and imaging. 7th ed. London: Elsevier; 2003. p. 641. 2. Ng DC, Kwok SY, Cheng Y, Chung CC, Li MK. Colonic amoebic abscess mimicking carcinoma of the colon. Hong Kong Med J. 2006;12(1):71–3.
Jpn J Radiol (2014) 32:6–13 3. Gadwood KA, Bedetti CD, Herbert DL. Colonic tuberculosis mimicking annular carcinoma: report of a case. Dis Colon Rectum. 1981;24(5):395–8. 4. Yu XQ, Wang ZJ, Zhao B, Ma HC, Jin ML. Extensively infiltrating colon adenocarcinoma similar to Crohn’s disease: a case report. Clin Res Hepatol Gastroenterol. 2012;36(4):63–8. 5. Ichikawa T, Erturk SM, Motosugi U, Sou H, Iino H, Araki T, Fujii H. High B-value diffusion-weighted MRI in colorectal cancer. Am J Roentgenol. 2006;187:181–4. 6. Thoeny HC, De Keyzer F, Oyen RH, Peeters RR. Diffusionweighted MR imaging of kidneys in healthy volunteers and patients with parenchymal diseases: initial experience. Radiology. 2005;235:911–7. 7. Ichikawa T, Haradome H, Hachiya J, Nitatori T, Araki T. Diffusion-weighted MR imaging with single-shot echo-planar imaging in the upper abdomen: preliminary clinical experience in 61 patients. Abdom Imaging. 1999;24:456–61. 8. Nasu K, Kuroki Y, Kuroki S, Murakami M, Nawano S, Moriyama N. Diffusion-weighted single shot echo planar imaging of
13
9.
10.
11.
12.
colorectal cancer using a sensitivity-encoding technique. Jpn J Clin Oncol. 2004;34:620–6. Kilickesmez O, Atilla S, Soylu A, Tasdelen N, Bayramoglu S, Cimilli T, Gurmen N. Diffusion-weighted imaging of the rectosigmoid colon: preliminary findings. J Comput Assist Tomogr. 2009;33:863–6. Kiryu S, Dotanuki K, Takao H, Watanabe M, Inoue Y, Takazoe M, et al. Free-breathing diffusion-weighted imaging for the assessment of inflammatory activity in Crohn’s disease. J Magn Reson Imaging. 2009;29:880–6. Oto A, Zhu F, Kulkarni K, Karczmar GS, Turner JR, Rubin D. Evaluation of diffusion-weighted MR for detection of bowel inflammation in patients with Crohn’s disease. Acad Radiol. 2009;16:597–603. Kılıc¸kesmez O, Soylu A, Yas¸ ar N, Demirbas¸ T, Dolapc¸ıog˘lu C, Poturog˘lu S, et al. Is quantitative diffusion-weighted MRI a reliable method in the assessment of the inflammatory activity in ulcerative colitis? Diagn Interv Radiol. 2010;16:293–8.
123
