Correlation of Clinical, Magnetic Resonance Imaging, and Intraoperative Findings in the Preoperative Assessment of Fistula-in-Ano: A Prospective Observational Study

Abstract

Background: Accurate preoperative delineation of the fistulous tract, internal opening, and any secondary extensions is essential for successful surgical management of fistula-in-ano, since incomplete identification is a major cause of persistence and recurrence. Magnetic resonance imaging (MRI) has increasingly been adopted as the preoperative imaging modality of choice, but its diagnostic performance must be validated against intraoperative findings in the population in which it is used.

Objective: To assess and correlate clinical examination, MRI, and intraoperative findings in patients with fistula-in-ano, and to determine the diagnostic accuracy of MRI, using intraoperative findings as the reference standard, for the internal opening, tract complexity, secondary tracts, abscess, horseshoe extension, and supralevator extension.

Methods: This prospective observational study enrolled 30 adults with a clinical diagnosis of fistula-in-ano at a tertiary-care teaching hospital over an 18-month period. All patients underwent digital rectal examination and proctoscopy, followed by pelvic MRI (1.0 Tesla, phased-array coil; T1 and T2-weighted turbo spin-echo sequences in axial, coronal, and optional sagittal-oblique planes) prior to surgery. Fistulas were classified using both the Parks classification and the St. James's University Hospital MRI-based grading system. All patients subsequently underwent examination under anesthesia and definitive surgery (fistulotomy or fistulectomy, with seton placement for high tracts), which served as the diagnostic reference standard. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of MRI were calculated for each finding.

Results: The cohort comprised 26 men and 4 women (mean age 35.6 years; most common age group 31–40 years). All patients had a palpable external opening on examination, most frequently at the 3 o'clock position (20%), while the internal opening was clinically palpable in only 4 of 30 patients (13.3%). MRI identified the internal opening in all 30 patients (100%), most commonly at the 6 o'clock position (26.7%). By Parks classification, intersphincteric fistulas predominated (63.3%), followed by transsphincteric (26.7%), suprasphincteric (6.7%), and extrasphincteric (3.3%) tracts. Intraoperatively, simple tracts were confirmed in 17 patients (56.7%) and complex tracts in 13 (43.3%); of the complex tracts, abscess was present in 5 (38.5%), secondary tracts in 4 (30.8%), horseshoe extension in 3 (23.1%), and supralevator extension in 1 (7.7%). Overall MRI–surgical concordance was 86.7% (26/30). MRI showed a sensitivity/specificity/PPV/NPV of 100%/84.6%/89.4%/100% for simple tracts, 100%/89.5%/100%/84.6% for complex tracts, 75%/88.8%/75%/88.9% for secondary tracts, 80%/100%/100%/88.8% for abscess, and 33.3%/90%/50%/81.8% for horseshoe extension.

Conclusion: MRI is a highly accurate, noninvasive preoperative tool for delineating fistula anatomy and shows excellent diagnostic performance for the internal opening, tract complexity, and abscess, but more limited sensitivity for horseshoe extension in this cohort. Intraoperative findings remain the diagnostic reference standard, and MRI is best used as a complementary road map that informs, rather than replaces, careful intraoperative assessment.