MR imaging of Strangulated Small Bowel Obstruction (SBO): "peristalsis gap sign" and "intraluminal bleeding sign"

A review was made in the MR studies including cine MR and surgical and pathological reports of ten patients with surgically confirmed strangulated small bowel obstruction to establish criteria for a definite MR diagnosis and indications for surgery. Akinesis of the closed loop and movement in surrounding bowel loops (peristalsis gap) were noted in all patients. lntraluminal low-intensity on TPwerighted images of the closed loop (intraluminal bleeding sign) was recognized in 5 of 7 patients requiring surgical resection and corresponded to muddy, bloody, bowel content, whereas this abnormal intensity was not present in three patients successfully treated with adhesiolysis alone. I": Small bowel obstruction (SBO) is a relatively common complication in patients who undergo abdominal surgery, and in some patients, emergency surgical treatment is required. Magnetic resonance (MR) imaging has developed into a powerful diagnostic tool because of its high contrast resolution and multi-plane observation ability (1 -4). However, clinical use of this modality is still limited, and its feasibility in emergency situations needs further clarification. We conducted a retrospective study to evaluate the usefulness of MR imaging in the diagnosis and prediction of surgery for bowel resection in patients with SBO. MATERIAL AND METHODS: Ten patients with surgically confirmed strangulated SBO (age range 23-89 years, mean age 56 years) comprised the study group. MR images were retrospectively reviewed and correlated with surgical and pathological findings. All patients had been evaluated with a 1.5T (Magnetom Vision (Siemens, Erlangen, Germany) or VISART Ex (Toshiba, Tokyo, Japan)) MR system. T2-weighted half-Fourier single-shot spin echo (FASE Fast Advanced Spin Echo, and HASTE half-Fourier acquisition single-shot turbo spin echo) and TI-weighted fat-suppressed gradient echo pulse sequences were obtained. The coronal plane was used as the standard imaging plane of its usefulness for surgical procedures. Additional axial and sagittal images were obtained if necessary. Evaluation of peristalsis of the bowel loop was done with cine MR imaging in seven patients with one of the following two techniques: single breath-hold fat-suppressed true FlSP (fast imaging with steady-state precession) sequence, obtained approximately every 1.5 seconds or intermittent breath-hold haif-Fourier single shot T2 weighted sequence, obtained every 6 seconds. Cine MRI duration was 30 seconds in the former and 2 min in the latter. Slice thickness was 10-50 mm and field-of-view was 35-40 cm. Total examination time averaged 20 to 30 minutes. All patients tolerated the procedure. Subtraction images were made from all cine MR images with a dedicated software program called TASK (Technical Automated Subtraction for Kinematic imaging). TASK performs three operations ((1) The image from No.1 is discarded. (2) Frame N0.2 becomes the mask image, which is subtracted from all of the following images. This avoids subtraction error due to saturation effects that can reduce the signal intensity of the image in frame No.2 and frames later. (3) After subtraction, areas of equal signal intensity (SI) are gray, areas of decreased SI are dark, and areas of increased SI are bright). Cine MR imaging was not done in the initial three patients. Therefore, subtraction images obtained from preand post-contrast images were used for evaluation of the peristalsis. We evaluated whether a closed loop was present on the basis of classical signs such as radial distribution, a G or U-shaped dilated loops or other findings. If the closed loop existed, further evaluation was done based on the presence of the "peristalsis gap sign" and the "intraluminal bleeding sign". The peristalsis gap sign was used to describe akinesis in the closed loop, although surrounding loops show evidence of peristalsis. Akinesis in the closed loop was considered present if no wall movement or no luminal signal change was seen on the subtracted image. If there was peristalsis, wall movement was visualized as a misregistration artifacts and the luminal signal became either bright or dark due to the respective increase or decrease in the amount of luminal fluid. The intraluminal bleeding sign was used to describe the presence of a combination of signal intensities, that is, a dark signal on the T2-weighted image and no evidence of a bright signal on the fat-suppressed TI-weighted image. This signal pattern is similar to that generated by deoxyhemoglobin containing hematomas, seen in acute brain hemorrhage. In our early experience, we noted that this signal pattern was consistent with closed loops muddy, bloody contents at surgery and that further corresponded to muddy, bloody contents in surgical pathology specimens. Therefore, we termed this signal pattern the "intraluminal bleeding sign". RESULTS: The closed loop was either detected (n=5) or strongly suspected n=5). All patients had akinesis in the closed loops, and all surrounding loops showed evidence of peristalsis. Therefore, the peristalsis gap sign was seen in all patients resulting in a sensitivity and positive predictive value (PPV) for strangulated obstruction of 100% (10/10) (Fig.1). Adhesiolysis was done in all patients to release the incarcerated loop. During surgery, if the closed loop did not show complete necrosis and appeared salvageable, the surgeon observed the loop for 15 minutes after adhesiolysis and evaluated whether bowel congestion was resolved. As a result, the closed loop in three patients was judged salvageable and bowel resection was avoided. The remaining seven patietns required bowel resection due to bowel necrosis, Le., irreversible strangulation. Luminal contents were sampled in three of these patients, and the presense of muddy, bloody fluid was. confirmed. The intraluminal bleeding sign was noted in five of seven irreversible strangulations. In the three patients in whom bowel was conserved (i.e. reversible strangulation), the intraluminal bleeding sign was negative. Thus, sensitivity of the intraluminal bleeding sign for irreversible strangulation was 71% (5/7) and specificity and PPV were 100%. CONCLUSION: Application of subtraction imaging, especially with cine MR imaging, permitted easy detection of the difference in peristalsis between the closed and surrounding loops (peristalsis gap sign). All patients with the peristalsis gap sign had a strangulated obstruction. lntraluminal bleeding sign suggests "irreversible" strangulated obstruction. Positive peristalsis gap sign and negative intraluminal bleeding sign suggest possibility of "reversible" strangulated obstruction.The combination diagnosis with these signs may be useful in the avoidance of unnecessary bowel resection due to delayed of diagnosis of surgical intervention. REFERENCES: 1. Chou CK., et al., Abdom Imaging 1993; 18: 131-135 2. Beall DP., et al., JCAT 1996; 20: 823-825. 3. Regan F.,elal., AJR 1998; 170: 1465-1469. 4. Umschaden HW., et al., Radiology 2000; 215: 717-725.