Introduction/Background There is no consensus on the most reliable procedure to determine remission of cervical cancer after chemoradiotherapy (CRT). Therefore, this study aims to assess the diagnostic performance of two different imaging techniques, MRI and PET-CT, for its ability to determine the presence of locoregional residual disease after CRT in patients with locally advanced cervical cancer.

Methodology Patients diagnosed with locally advanced cervical cancer (FIGO 2009) treated with CRT were retrospectively identified from a regional cohort. The accuracy of MRI and PET-CT was assessed for its ability to detect locoregional residual disease with histology as the reference standard.

Results The negative predictive value (NPV) and positive predictive value (PPV) of MRI and PET-CT combined for locoregional residual disease detection was 84.2% (95% CI 73.2–92.1), and 70.4% (95% CI 51.8–85.2), respectively. The NPV and PPV of MRI alone was 80.2% (95% CI 71.2–87.5) and 47.7% (95% CI 35.8–59.7); and 81.1% (95% CI 72.2–88.3) and 55.8 (95% CI 42.2–68.7) for PET-CT alone.

Conclusion In this study, the reliability of MRI and PET-CT to detect locoregional residual disease is limited. Combining MRI and PET-CT did not improve predictive values. Routine use of both MRI and PET-CT in the follow-up after CRT should be avoided. MRI during follow-up is the advised imaging technique. Pathological confirmation of the presence of locoregional residual disease before performing salvage surgery is warranted.

Disclosures None.