Incorporating PET information in radiation therapy planning

PET/CT-guided treatment planning for paediatric cancer patients: a simulation study of proton and conventional photon therapy

OBJECTIVE

To investigate the impact of including fluorine-18 fludeoxyglucose ((18)F-FDG) positron emission tomography (PET) scanning in the planning of paediatric radiotherapy (RT).

METHODS

Target volumes were first delineated without and subsequently re-delineated with access to (18)F-FDG PET scan information, on duplicate CT sets. RT plans were generated for three-dimensional conformal photon RT (3DCRT) and intensity-modulated proton therapy (IMPT). The results were evaluated by comparison of target volumes, target dose coverage parameters, normal tissue complication probability (NTCP) and estimated risk of secondary cancer (SC).

RESULTS

Considerable deviations between CT- and PET/CT-guided target volumes were seen in 3 out of the 11 patients studied. However, averaging over the whole cohort, CT or PET/CT guidance introduced no significant difference in the shape or size of the target volumes, target dose coverage, irradiated volumes, estimated NTCP or SC risk, neither for IMPT nor 3DCRT.

CONCLUSION

Our results imply that the inclusion of PET/CT scans in the RT planning process could have considerable impact for individual patients. There were no general trends of increasing or decreasing irradiated volumes, suggesting that the long-term morbidity of RT in childhood would on average remain largely unaffected.

ADVANCES IN KNOWLEDGE

(18)F-FDG PET-based RT planning does not systematically change NTCP or SC risk for paediatric cancer patients compared with CT only. 3 out of 11 patients had a distinct change of target volumes when PET-guided planning was introduced. Dice and mismatch metrics are not sufficient to assess the consequences of target volume differences in the context of RT.

Comparison of positron emission tomography (PET) and computed tomography (CT) for better target volume definition in radiation therapy planning

BACKGROUND

In cancer patients, positron emission tomography/ computed tomography (PET/CT) fused images present less variability in target contouring, respect to use only CT images, respectively. However, the gold standard has not yet been clearly established between radiation oncologists with regard to PET images and the methodology of contouring targets with confidence using PET/CT fused images. The aim of this study was to determine whether integrated PET/CT fused images provide advantages in virtual simulation compared with morphological contouring only with CT.

MATERIAL AND METHODS

Thirty cancer patients were evaluated in an adapted PET/CT hybrid in radiotherapy (RT) setup position, with 20 of them being suitable for RT: 17 were suitable for curative intent, which was the group of interest in this study. All image series were sent to the RT work station (WS) where CT and PET series were automatically fused by Digital Imaging and Communications in Medicine (DICOM) in each case. PET series were threshold and were subjected to source-to-background contrast algorithms to fi nally redefine the original tumour description. Three different radiotherapy plans (RTP) for each patient were compared after targets were contoured: [1] planning over metabolic (PET) contoured targets, [2] planning over only morphologic (CT) targets, and [3] planning over targets obtained for treatment based on fused PET/CT images.

RESULTS

PET/CT findings altered initial-stage planning in four patients (23.5%) because they had been undergoing chemotherapy. Gross target volume (GTV) and planning target volume (PTV) based only on PET showed more homogeneity to obtain mean doses (p = 0.025) with respect to those based on PET/CT, respectively. However, no percentage differences were observed in median PTV doses between the planning methods, although there was higher variability in PET/CT planning. Morphological (CT) and PET/ CT target volumes were more voluminous than metabolic (PET) volumes. On the other hand, 20% of metabolic (PET) PTV were out of those defined by PET/CT. Thoracic RT plans based on PET preserved better bilateral lung [percentage volume of lung irradiated with a dose of 20 Gy (V20); significance, R(2) = 0.559, p = 0.006].

CONCLUSIONS

For our physicians, PET/CT fused images allowed better contouring of primary tumours in 40% of head and neck cancers and 34% of thoracic cancers. PET/CT provides useful information for virtual simulation therapy. Image treatment and planning in an RT workstation is mandatory.