Diagnostic accuracy of CT and MRI for mandibular invasion in oral cavity carcinoma: a histopathology-correlated study using Brown's and Mcgregor's criteria.

INTRODUCTION

Ginigivobuccal cancers often involve the mandible, impacting surgical planning and outcomes. Clinical assessment of bone invasion is unreliable, making imaging essential. CT excels at detecting cortical erosion but may miss early medullary involvement, while MRI offers greater sensitivity for marrow changes but may overestimate invasion. Diagnostic performance varies depending on the interpretive criteria, Brown's Favors early marrow detection, McGregor's prioritizes cortical specificity. Few studies have directly compared CT and MRI using both criteria in the same cohort. This study addresses that gap by evaluating CT and MRI accuracy against histopathology in gingivobuccal carcinoma using both diagnostic frameworks.

MATERIALS AND

METHODS

This retrospective diagnostic accuracy study included patients with biopsy-proven gingivobuccal squamous cell carcinoma who underwent preoperative CT and MRI, followed by mandibulectomy. Imaging was independently assessed using both Brown's and McGregor's criteria by a blinded radiologist. Histopathological confirmation of cortical or medullary invasion served as the reference standard. Diagnostic metrics (sensitivity, specificity, PPV, NPV, accuracy) were calculated for each modality-criterion pair. McNemar's test and ROC curves were used for comparison. A total enumeration approach was adopted without formal sample size calculation.

RESULTS

Among 287 patients (mean age 54.2 years; 75% male), mandibular invasion was confirmed histologically in 158 cases (55.1%): 64 cortical-only, 48 marrow-only, and 46 both. MRI with Brown's criteria showed highest accuracy (90.3%), sensitivity (89.2%), and specificity (91.7%). CT-Brown followed (accuracy 86.8%). CT-McGregor detected 100% of cortical-only cases with 88.4% specificity. MRI-Brown identified all marrow-only cases (100%), while CT-Brown detected 54.9%; both CT- and MRI-McGregor missed all marrow-only disease. Brown's criteria significantly outperformed McGregor's for medullary detection (p < 0.01). MRI specificity exceeded CT (p ≈ 0.06). Results suggest MRI-Brown is ideal for detecting early marrow invasion, while CT-McGregor provides superior cortical evaluation. A modality-criterion matched approach enhances diagnostic precision and supports individualized surgical planning in gingivobuccal carcinoma.

DISCUSSION: MRI with Brown's criteria achieved the highest accuracy (90.3%), with sensitivity 89.2% and specificity 91.7%. CT-McGregor showed 100% detection of cortical-only invasion and 88.4% specificity. MRI-Brown detected all marrow-only cases (100%), while CT-Brown detected 54.9%, and McGregor criteria missed all. Brown's criteria outperformed McGregor's for medullary disease (p < 0.01). While MRI showed higher specificity than CT (p ≈ 0.06), results support a modality-specific approach tailored to suspected invasion type for optimal surgical planning.

CONCLUSION

This study uniquely compares Brown's and McGregor's criteria across CT and MRI, revealing MRI-Brown's superiority for medullary invasion and CT-McGregor's accuracy for cortical erosion. By aligning interpretive criteria with modality and invasion type, it fills a key diagnostic gap and offers a practical, evidence-based framework for surgical planning in GBSCC. Diagnostic Accuracy of CT and MRI for Mandibular Invasion in Oral Cavity Carcinoma: A Histopathology-Correlated Study Using Brown's and McGregor's Criteria.