Preoperative MR Imaging of Anal Fistulas: Does It Really Help the Surgeon?
Executive Summary
This document synthesizes findings from a prospective study evaluating the accuracy and clinical utility of high-spatial-resolution magnetic resonance (MR) imaging using a quadrature phased-array coil for the detection of anal fistulas. The analysis demonstrates that while preoperative MR imaging provides marginal benefit for primary simple fistulas, it offers critical diagnostic information for complex cases, specifically those involving Crohn disease or recurrent tracks.
Critical Takeaways:
Clinical Value: MR imaging provided important additional information missed during initial surgical exploration in 21% of all patients.
Target Populations: The benefit is highly stratified: 40% for patients with Crohn disease, 24% for recurrent fistulas, and only 8% for primary simple fistulas.
Diagnostic Superiority: The technique is exceptionally accurate in detecting abscesses (96% sensitivity) and horseshoe fistulas (100% sensitivity), which are frequently overlooked during standard surgical examination under anesthesia.
Recommendation: Preoperative MR imaging is strongly recommended for the work-up of complex (recurrent or Crohn-related) fistulas but is rarely necessary for primary simple fistulas.
Study Parameters and Methodology
The study involved 56 patients (32 men, 24 women; mean age 42) categorized into three distinct clinical groups:
Crohn Disease: 15 patients with associated perianal fistulas.
Recurrent Fistula: 17 patients with a history of previous fistula surgery.
Primary Simple Fistula: 24 patients with no history of Crohn disease or prior fistula surgery.
The "Blinded" Surgical Protocol
To evaluate the true clinical value of MR imaging, surgeons performed initial explorations while blinded to the MR findings. The extent of the disease was established by cannulating tracks with probes. Only after the initial surgery was completed was the MR report disclosed. If the MR imaging suggested more extensive disease, the surgeon returned to verify those findings. Additional MR findings were only considered true-positive if confirmed by further surgical drainage.
Technical Specifications of Imaging
The study utilized a 1.5-T MR system with a quadrature phased-array spine coil. This specific hardware arrangement provides a higher signal-to-noise ratio compared to standard body coils or linear phased-array coils.
Spatial Resolution: The technique achieved a voxel size of 2.6 mm³, significantly smaller than the 11–31 mm³ voxel sizes typical of body-coil techniques.
Sequences: Utilized T1-weighted two-dimensional turbo spin-echo (SE) for anatomical planning and T2-weighted two-dimensional turbo SE in sagittal, coronal, and transverse planes for fistula identification.
Non-Invasive Protocol: Patients required no bowel preparation, no anal catheters, and no contrast material (e.g., gadolinium) or saline instillation (fistulography).
Diagnostic Accuracy and Performance
High-spatial-resolution MR imaging demonstrated high accuracy across various abnormality classifications. The findings were evaluated against the "Final Surgical Findings" (initial surgery corrected by MR-guided exploration) as the reference standard.
Accuracy Metrics for MR Imaging
Observer Agreement
The study found "good to very good" interobserver agreement (Kappa values 0.75 to 0.93) between an experienced radiologist, a radiology resident, and a surgeon trained in reading pelvic MR images. Agreement was highest for abscesses and horseshoe fistulas and lowest for fistula classification and internal opening detection.
Clinical Impact and Value-Add Analysis
The primary value of MR imaging lies in its ability to identify secondary extensions that are easily missed during surgery, particularly those located above the pelvic floor.
Additional Information Provided by Patient Group
The difference in clinical benefit between groups was statistically significant (P < .05):
Crohn Disease Group: 6 of 15 patients (40%) benefited from MR imaging.
Recurrent Group: 4 of 17 patients (24%) benefited.
Primary Simple Group: 2 of 24 patients (8%) benefited.
Specific Lesions Missed at Initial Surgery
MR imaging successfully identified several critical features that surgeons initially overlooked:
Supralevator Abscesses: Seven collections missed at initial surgery were correctly predicted by MR imaging.
Horseshoe Fistulas: Four horseshoe extensions were identified by MR imaging after being missed by surgeons.
Internal Openings: Three openings in the lower rectum were revealed by MR imaging after being missed during exploration.
Comparative Analysis of Diagnostic Modalities
The document highlights the limitations of alternative diagnostic methods compared to high-spatial-resolution MR imaging:
Fistulography: Correct in only 16% of patients.
Computed Tomography (CT): Fails to depict subtle tracks due to low soft-tissue contrast resolution.
Endosonography: Limited by a narrow field of view; reported to be no more accurate than a physical examination under anesthesia.
Endoanal MR (Endocoil): While providing high signal-to-noise ratio near the coil, it fails to visualize secondary extensions beyond the coil's range (surgical concordance rates as low as 64%–68%).
Body-Coil MR: Lacks the spatial resolution necessary for subtle tracks, often requiring invasive contrast agents or fat-suppression techniques that obscure pelvic floor anatomy.
Observations on Accuracy and Limitations
The Challenge of Fibrotic Tracks
While MR imaging had a 100% sensitivity for primary tracks, the specificity was lower (86%). This is attributed to healed fibrotic tracks, which MR imaging occasionally misidentified as active tracks. In retrospect, these fibrotic tracks typically lacked the hyperintense (bright) fluid signal inside the hypointense (dark) fistula track.
Internal Opening Identification
Internal openings are not always directly visualized. Their location is often inferred from the proximity and course of the fistula track relative to the sphincter muscle compartments. This method yields high sensitivity but results in a lower specificity due to false-positive enteric entry site predictions.
Selection Bias
The study population was selected based on a high suspicion of fistulas scheduled for surgery. This led to "overreading" by observers prioritized to identify all possible extensions rather than avoiding false positives. Consequently, sensitivity was prioritized over specificity in the clinical assessment.
Conclusion
High-spatial-resolution MR imaging using a quadrature phased-array coil provides a non-invasive, wide-field-of-view diagnostic tool that significantly improves surgical outcomes for complex anal fistulas. By accurately identifying supralevator abscesses and horseshoe extensions, the technology helps prevent the high recurrence rates associated with overlooked secondary tracks in Crohn disease and recurrent fistula cases. While it is an essential component of the preoperative work-up for complex disease, its use in primary simple fistulas remains limited.
