SP-0362: Hypoxia PET imaging for delineation and response assessment during radiotherapy

Purpose: To review the current status of hypoxia PET imaging for delineating hypoxic volumes (HVs) inside the gross target volume (GTV) and for radiotherapy (RT) outcome prediction. Methods: Hypoxia PET tracers currently used in clinical trials are mainly [F]-FMISO, [F]-FAZA and [F]-HX4. Different image acquisition protocols and methods for delineating HVs will be presented in this talk. Furthermore, results of recent clinical trials will be reviewed where the prognostic value of hypoxia PET imaging before and during RT with respect to outcome as investigated. In our institution, a clinical phase II trial is currently carried out where to date n=33 head and neck cancer patients were included. Patients were examined using dynamic FMISO PET imaging plus conventional FDG PET imaging in addition to planning CT and eventually MRI before the start of RT. HVs were segmented based on parameters derived from the pharmaco-kinetic analysis of the dynamic FMISO PET data. Patients were randomized into two treatment arms. Patients in the experimental arm were treated with a 10% dose escalation to the HV whereas patients in the control arm received standard IMRT treatment with 70 Gy in 35 fractions. Results: Different manual and (semi-)automatic methods for HV delineation based on hypoxia PET data have been used in clinical studies, such as thresholding, tumour-to-background ratio (TBR) based methods, advanced automatic methods or delineations based on dynamic PET imaging. Different contouring techniques for HV definition may result in strongly varying volumes. Recently published clinical trials confirm the prognostic character of hypoxia PET imaging. However, no consensus was found yet with regard to the timing of the hypoxia PET examination. Some studies found that pretreatment hypoxia PET information was correlated to outcome whereas others stated that hypoxia PET data acquired two weeks into RT had prognostic value. For our hypoxia dose painting trial, a planned interim analysis was carried out after recruiting n=20 patients. Median followup time for this group was 36 (11 52) months. 5 patients did not show any tumour hypoxia (HV = 0 mL). The mean HV of the hypoxic tumours was 8.6 mL (0.3 – 49.2 mL). Furthermore, data acquired in this study could confirm a prognostic model relating TCP to a measure derived from dynamic FMISO PET imaging acquired before the start of RT. Conclusion: Hypoxia PET imaging is a very promising tool for the stratification of patients into different risk groups and thus a potentially very interesting molecular marker in the advent of biologically adapted, personalized RT. However, a comparison of results from different clinical trials is difficult due to large discrepancies in terms of imaging protocols and data analysis strategies.