Initial staging workup in cervical cancer: clinical examination, imaging, and laboratory assessment - a review from the french cervical cancer guidelines group.

Accurate staging of cervical cancer is essential for guiding treatment selection and determining prognosis. Physical examination, although fundamental for assessing vaginal extension and obtaining biopsies, shows important limitations for evaluating parametrial invasion and overall FIGO staging. Pelvic MRI provides higher diagnostic performance, particularly when diffusion-weighted imaging is incorporated, improving assessment of tumour size, parametrial involvement and residual disease after chemoradiotherapy. Contrast-enhanced sequences offer additional value for detecting small lesions and characterizing stromal invasion. Functional MRI parameters, including ADC changes and dynamic enhancement patterns, show promise for early prediction of treatment response, although standardisation remains insufficient for routine clinical application. After conization, MRI demonstrates good sensitivity and excellent specificity for detecting residual tumour, but postoperative inflammatory changes may reduce accuracy when performed too early. For nodal staging, 18F-FDG PET-CT offers whole-body evaluation, with diagnostic sensitivity strongly influenced by tumour stage and nodal prevalence. PET-CT rarely influences management in IA1-IA2 and has limited impact in IB1 but becomes relevant in IB2-IIB and IIIA-IVA disease, particularly when positive nodal uptake may obviate lymphadenectomy. PET-derived metabolic parameters also correlate with recurrence and survival, though no validated thresholds allow direct clinical use. Serum SCC-Ag is associated with recurrence and survival in observational studies, but its clinical utility remains unproven due to heterogeneity of thresholds and absence of evidence that testing improves outcomes. Future research integrating modern MRI protocols, digital PET-CT technology and validated imaging biomarkers is needed to refine staging algorithms and optimize personalized management.