Comparison with early and late dual-time FDG PET/CT scans after radiotherapy in head and neck cancer

Objectives PET/CT is usually performed 3–4 Mon after the end of RTx to avoid false positive findings whereas early evaluation is needed clinically . Our aim is to compare early 1 Mon dual-time PET/CT with 3 Mon later PET/CT.

Methods Between Dec. 2005 and Dec .2008, 46 patients (M:F=32:14, Age=54±14) with biopsy-proven head and neck cancer (Nasopharynx=32, Larynx=5, Tongue=3, Tonsil=2, other 4) were enrolled consecutively. All patients received standard RTx (Dose=6587±231). PET/CT was performed before, and within 1 month and 3 month after completing their course of RTx. 25 patients underwent concurrent chemotherapy between 1 Mon and 3 Mon PET/CT. PET images were scanned 1 and 3 hour after FDG injection for dual time protocol. 1 Mon and 3 Mon PET scans were evaluated visually and compared. Logistic regression analysis and ROC analysis were done.

Results 8 pts and 12 lesions showed residual FDG accumulations on 3 Mon PET (residual groups) whereas 38 pts and 117 lesions did not. The SUV of pre-PET/CT were not different with residual and non-residual groups (8.02 ± 4.80, 8.13 ± 4.8, p=0.99). 1 hour SUV of residual and non-residual groups were 3.47 ± 2.14 and 2.24 ± 1.20 (p=0.021) and 3 hour SUV were 3.93 ± 2.47 and 2.32 ± 1.47 (p=0.04). Retention index (RI) was 0.13 ± 0.13 and 0.01 ± 0.19 (0.01). 1 hour and 3 hour SUV and RI were all significantly different between residual and non-residual groups and RI has the highest significance on logistic regression analysis. AUC of 1 hour, 3 hour SUV and RI were 0.671, 0.714 and 0.745, respectively.

Conclusions One Mon PET after radiotherapy corresponds with 3 Mon PET with modest accuracy.