In cervical cancer prognosis factors are size of the tumour, pelvic and para-aortic lymph nodes involvement. The initial treatment taking is determined by lymph node evaluation.

Early stage cervical cancers at their very beginning, with no lymph node involvement, will be treated either by surgery only or by radiotherapy only, or by both at the same time. At advanced stages, from proximal IB to IIB with bad prognosis (tumour larger than 4 cm and pelvic lymph node involvement), as well as for cancers up to distal stages IIB, III and IVA, treatment relies on radio-chemotherapy, either alone or pre-surgery.

The pre-therapeutic complete examination of the cervical cancer includes a clinical examination and a pelvic MRI in order to look for pelvic and para-aortic lymph nodes and to precise the volume and the spreading of the tumour in the pelvic area.

A meta-analysis estimating the interest of the MRI showed a variable sensitivity and a specificity for the detection of such lymph node metastasis with a sensitivity varying from 24 to 75 % and a specificity between 84 and 100 % according to the studies.

As for the use of an 18-FDG PET scan to detect lymph node involvement, the studies realized so far are performed, usually, on a small number of subjects and with heterogeneous populations. However, it seems that such an examination enables a better detection of lymph nodes than the MRI does, especially for para-aortic lymph nodes depending on the studies, sensitivity varies from 57 to 100 %, and specificity between 92 and 100 %.

We propose a prospective, multicentric and multidisciplinary study for the estimation of diagnosis methods. The main aim is to estimate the interest of the 18-FDG PET scan for the detection of pelvic and para-aortic lymph node metastasis in patients with cervical cancer, in comparison with the MRI. This study should enable to include 380 patients within 3 years. An 18-FDG scintigraphy will be performed before any treatment together with an MRI. A comparison between the sensitivity and the specificity of both examinations will be established and a correlation with the histology of the lymph node dissection will be made.

Inclusion Criteria:

• Patient with cervical carcinoma: histological proof of epidermoid carcinoma or of adenocarcinoma
• Stage I to IV (limited to pelvic) (FIGO Classification) (see appendix 3)
• Age ≥ 18 years old
• No contra-indication to surgery
• Dated, informed and signed consent from the patient

Exclusion Criteria:

• Metastatic cervical cancer with extra-pelvic metastasis
• Contra-indication to MRI: implantation of an electronic device, metallic or ferromagnetic foreign body
• Serious co-existing affection with vital prognosis
• Diabetes uncontrolled by a classical treatment: glycaemia > 1.4g/l
• Pregnancy and lactation
• Uncontrolled infection
• Other cancer within the previous 5 years, except skin basal cell carcinoma, or in situ cervical cancer
• Patient with some deficiency preventing her from completely understanding any requirement of the study, which might compromise the patient's consent and/or the observance of the protocol and the consistent participation to the study.