Predictive value of PET/CT based metabolic information in the modern 3D based radiotherapy treatment of head and neck can-cer patients - single institute study

Abstract

Objective: The aim of the study was to evaluate the predictive value of pretreatment positron emissiontomography (PET) standardized uptake value (SUVmax), standardized uptake value corrected for lean bodymass (SULpeak) value, metabolic tumour volume (MTV) and total lesion glycolysis (TLG) parameters ofthe primary tumour assessed with PET/computed tomography (CT) in the clinical outcome in patientsdiagnosed with histopathologically con􀁁rmed head and neck squamous cell carcinoma. Materials andMethods: Retrospective evaluation was performed using PET/CT image datasets of 52 histologically provenhead and neck cancer patients in 4 weeks' prior receiving de􀁁nitive chemo-radiotherapy (CRT). Positronemission tomography /CT was performed before the CRT and 12 weeks after it for response evaluation.Image data was used for target volume delineation and for specify SUVmax, SULpeak, MTV and TLGparameters of the primary tumour. According to the results of the therapeutic response evaluation twopatient subgroups were created in relation to the presence or absence of viable tumour. Metabolic datafrom pre-treatment PET/CT and therapeutic response were correlated using Kruskal-Wallis test. Results:After completion of the CRT in 24/52 (46%) cases viable residual tumour was detected on restaging PET/CT, while in 28/52 (54%) patients showed complete remission. For the therapeutic success prediction assessment,we could not 􀁁nd any signi􀁁cant correlation with pre-treatment SUVmax and SULpeak values(P>0.44, P>0.33). Total lesion glycolysis provided nearly signi􀁁cant di􀁋erence (P=0.052) and MTV hadshown signi􀁁cant di􀁋erence (P=0.001) between the two patient subgroups statistically. Conclusion: Sim-18 ple metabolic data (SUVmax and SULpeak) from pretreatment 􀁍uorine-18-􀁍uorodeoxyglucose ( F-FDG)PET/CT were unable to predict therapeutic response, while volumetric information containing MTV andTLG parameters proved to be more useful, thus their inclusion to risk strati􀁁cation may also have additionalvalue.