Abstract
BACKGROUND: Cervical cancer is a major public health challenge in Nepal, with delayed diagnosis due to limited screening and infrastructure. Accurate staging, critical for treatment planning, integrates MRI with clinical examination (CE) per 2018 FIGO guidelines. This study evaluates MRI-CE concordance in staging cervical cancer at a Nepalese tertiary center, hypothesizing moderate agreement with MRI detecting occult spread missed by CE. METHODS: This retrospective study (2020-2023) included 76 patients with histologically confirmed cervical cancer and complete MRI (1.5T, T2/DWI/post-contrast) and CE (pelvic exam, punch/cone biopsy) data at Purbanchal Cancer Hospital. Staging followed FIGO 2018 guidelines. Univariate analyses summarized patient profiles; MRI sensitivity, PPV, and Cohen's κ assessed concordance. Missing treatment data (48.7%) were categorized as follow-up/remission cases. RESULTS: Of 76 patients (median age 50-59 years), 88.2% had squamous cell carcinoma; Stage IIB predominated (47.4%). MRI-CE concordance was 63.2% (κ = 0.58, 95% CI: 0.42-0.74). MRI upstaged 12 cases (15.8%), detecting parametrial (58.3%) or nodal involvement (41.7%), and downgraded 16 (21.1%), often underestimating stromal invasion (37.5%). MRI sensitivity was 63.2%; PPV was 75%. Treatment data, available for 39 patients, showed lower-than-expected CCRT/ICBT use due to resource constraints. CONCLUSION: MRI complements CE by enhancing staging accuracy, particularly for occult spread, but does not replace it in resource-limited settings. Standardized MRI protocols and prospective studies are needed to optimize diagnostic and therapeutic outcomes in Nepal.