Patients with advanced head and neck squamous cell carcinoma (HNSCC) may benefit from organ-preservation treatment based on combination of chemotherapy and radiotherapy without compromising disease-free and overall survival. In patients with initially advanced regional disease, there is controversy about the place of routine planned lymph node neck dissection after chemoradiotherapy, especially in responding patients without clinically invaded residual lymph nodes. There is uncertainty about the lymph nodes status after chemoradiation because the structural imaging modalities (CT, MRI) lack sensitivity and specificity : small positive lymph nodes are not detected, and residual large lymph nodes can be sterilized ( " ghosts nodes " with no sign of viable tumor cells at histopathology). Despite the absence of evidence based on prospective study, in numerous institutions currently, head and neck surgeons are quite reluctant to operate on for neck dissection patients with a complete clinical and radiological response following chemoradiation.

Metabolic imaging of tumors using PET and the glucose analog FDG has proven effective in head and neck SCC, especially after treatment to differentiate disease progression from radiation-induced inflammation.1 Several studies have shown that the metabolic response could predict the presence or absence of residual tumor cells in the primary tumor as well as the probability of relapse .2-4 Conflicting results have been reported on the use of PET to predict the pathological nodal status after chemoradiation, with negative predictive values ranging from 14 % to 100 %.5,6 Discrepancies observed might be due to the fact that PET was performed at variable time points after the end of radiotherapy. Ideally, PET should be performed as late as possible so that tumor regrowth can begin and become detectable, increasing the sensitivity of the procedure.

- The primary objective is to assess the negative predictive value (NPV) of PET as a single examination in correctly predicting the absence of remaining invaded lymph nodes after chemoradiotherapy for advanced HNSCC.

Secondary objectives include :

• The evaluation of the suitability of a wait and see approach without neck dissection in patients considered as complete responders ( based on clinical evaluation and imaging assessment including PET : all these diagnosis tools should be negative to consider a patient as a complete responder); this suitability will be estimated using the negative predictive value of the overall assessment of a complete response including PET-FDG but also the clinical evaluation and imaging.
• The evaluation of the ability of PET-FDG to correctly predict remaining pathologically invaded lymph nodes (PPV) after chemoradiotherapy for advanced HNSCC in patients with a postchemoradiation positive PET a (and who will therefore be considered with less than a complete regional or locoregional response and who will undergo at least neck dissection.

Inclusion Criteria:

• Biopsy-proven HNSCC
• Clinical stage: unpreviously treated T1-T4 oral cavity, oropharynx, hypopharynx, larynx
• Only patients suitable for at least a neck dissection after chemoradiotherapy will be included. Consequently, inclusion criteria are thus : T1-T4 , N1 N2a, N2b N2c N3 M0.
• Patient scheduled for an organ preservation treatment protocol based on concomitant chemoradiation (induction chemotherapy is allowed if this approach is followed by concomitant chemoradiation)