PELVIC FISTULAS - APPEARANCES ON MR-IMAGES

Citation
Rc. Semelka et al., PELVIC FISTULAS - APPEARANCES ON MR-IMAGES, Abdominal imaging, 22(1), 1997, pp. 91-95
Citations number
21
Categorie Soggetti
Gastroenterology & Hepatology","Radiology,Nuclear Medicine & Medical Imaging
Journal title
ISSN journal
09428925
Volume
22
Issue
1
Year of publication
1997
Pages
91 - 95
Database
ISI
SICI code
0942-8925(1997)22:1<91:PF-AOM>2.0.ZU;2-A
Abstract
Background: This multi-institutional study examines appearances of pel vic fistulas on magnetic resonance (MR) images. Methods: MR images of 46 patients with documented fistulas from five teaching hospitals were retrospectively reviewed. All patients underwent T1-weighted (T1WI), T2-weighted (T2WI), and intravenous gadolinium che late-enhanced T1-we ighted (Gd-T1WI) images. Imaging sequences were separately and then co llectively reviewed. The following determinations were made: fistula d etection, fistula morphology and signal intensity, and the presence of associated abnormalities. Fistulas were classified into two categorie s: (1) fistulas that communicate with the bladder and (2) fistulas tha t do not communicate with the bladder. Fistulas within these two group s were subclassified further. The presence of fistulas was documented by surgery (five patients), endoscopy (six patients), fistulogram (20 patients), or physical exam (15 patients). Results: Among the 46 patie nts, 53 fistulas were documented by means other than MR. Overall T1WI, T2W1 and Gd-T1WI images demonstrated 23, 31, and 39 of 53 fistulas, r espectively. Gd-T1W1 detected significantly more fistulas than T1W1 (p < 0.05). Bladder fistulas were better shown on Gd-T1WI (8/15 fistulas ) than on T1WI and T2WI (2/15 and 3/15) (p < 0.05). Nonbladder fistula s were demonstrated by T1WI, T2WI, and Gd-T1WI images in 21, 28, and 3 1 of 38 fistulas, respectively. Among all fistulas, perianal fistulas (a subcategory of nonbladder fistula) had the highest detection by T1W I, T2WI, and Gd-T1WI in 19, 20 and 22 of 23 fistulas, respectively. On T1WI, 19 of 23 detected fistulas were low in signal intensity. On T2W I, 28 of 39 detected fistulas were high in signal intensity. On Gd-T1W I images, 29 of 40 fistulas were low in signal intensity, with enhance d tract wall. Conclusion: Bladder fistulas were best shown on Gd-T1WI, which was significantly greater than on T1WI or T2WI. Nonbladder fist ulas were comparably shown by all techniques, with all performing mode stly well. Peri anal fistulas were shown equally well by all MR sequen ces and were the fistulas demonstrated with the highest sensitivity on MR images.