Introduction/Background In the 2018 FIGO classification of carcinoma of the cervix, assessment of lymph nodes by imaging and/or pathological examination is now integrated. Patients with suspicious lymph nodes on imaging are designated as stage IIIC with the annotation ‘r’. Compared to the 2009 (clinically based) FIGO classification, we expect that more patients will be assigned a higher stage prior to treatment. Because of the increasingly important role of imaging modalities such as MRI and PET-CT in the staging and treatment of cervical cancer, we analyzed the accuracy of PET-CT in patients with early stage cervical carcinoma, with special emphasis on the positive predictive value (PPV).

Methodology Patients with early cervical cancer (FIGO 2009 stage IB1/IB2/IIA) that underwent a preoperative PET-CT for standard clinical care between 2010 and 2018 and subsequently a lymph node assessment (histology) were included in this retrospective study. Sensitivity (sens), specificity (spec), negative predictive value (NPV) and PPV of the PET CT scan for the prediction of pelvic lymph node metastases was calculated.

Results A total of 102 patients was included. Fifty eight percent of this group had positive nodes. Table 1 shows the sens, spec, NPV and PPV. When patients who had a nodal debulking because of bulky suspicious nodes on MRI were excluded (n=37), the remaining 65 patients (with a prevalence of 41% of positive nodes), showed a PPV of 68% (table 2). This would result in erroneously upstaging to stage IIIC in 32% of the patients with a PET-CT scan and suspicious nodes.

Conclusion We argue that, depending on the prevalence of positive nodes, PET-CT has a large likelihood to overestimate the incidence of lymph node metastases, theoretically resulting in a non-appropriate treatment. Pathological confirmation of lymph nodes suspicious for metastases on PET-CT scan is essential to guide therapy decisions.

Disclosure Nothing to disclose.