Prognostic value of 18FDG PET-CT in patients with head and neck cancer candidates to radical radiotherapy

Objectives: To evaluate the predictive role of primary tumor and lymph-node uptake for disease-free survival (DFS) in head and neck cancer (HNC) patients (pts) candidates to radical RT. Methods: 25 HNC pts (median age 59.3 yrs, range: 23-81 yrs) treated with FDG PET/CT guided radical radiotherapy (2005-2011,simultaneous integrated boost (SIB) on PET positive sub-volume intensity-modulated radiotherapy (IMRT), 69Gy, 30 fractions),were retrospectively included. SUV,with and without Partial Volume Effect correction (PVC) and Biological Target Volume (BTV) (60% SUVmax) were calculated for both primary tumor and nodes (SUVT, SUVL and BTVT, BTVL, respectively) on pre-therapy FDG PET/CT, to identify SUV patient clusters.Two-year DFS was calculated for each cluster. Results: FDG PET/CT detected primary tumor in all 25 cases and cervical nodes in 17/25 pts (1 FP and 2 FN). SUVT correlated with SUVL, both with and without PVC (r=0.82, 0.76 respectively). Pts with negative nodes(G1) had SUVT<14 (SUVT<11 without PVC). SUVT did not correlate with BTVT and BTVL. Two separate clusters of SUV thresholds were found for pts with both PET positive tumour and nodes:G2) SUVT<17 and SUVL<13 (SUVT<14.6 and SUVL<11 without PVC) and G3)SUVT>17 and SUVL>13 (SUVT>14.6 and SUVL>13 without PVC). DFS was 55.5%,45.8% and 75% for G1,G2 and G3 when PVC was applied,while 55.5%,50.3%and 66.7% without PVC. Conclusions: Pretreatment SUV has a prognostic value for DFS in HNC pts candidates to RT. PVC is essential for accurate SUV measurement.A high primary tumor uptake (SUVT >17) is always associated with the presence of positive nodes