It's an interesting issue but not fully settled as yet!
1)Can GTV delineation be on the basis of PET CT.
Yes, based on the extrapolation of the studies from Lung Cancer, there definitely is decrease in the inter-observer observation. Of course, anything that is "BRIGHT" is not cancer. The issues arise in those areas where you have physiological uptake, e.g. Tonsils and it becomes difficult to differentiate any GTV or nodes in that region.
Further all modalities tend to underestimate the "true GTV" i.e. CT, MRI and PET would not detect say about 10-15% of the superficial mucosal extension. Undue reliance on imaging is not warranted.
The other question is the threshold values that you keep for cut off. For example, is it 2.5 SuV or 40% of SuvMax. Automated systems to overcome this issue are being explored. Another issue (that I have also frequently faced) is the delineation of "edges" of the FDG-PET defined volume alone. How do you know what is the cut off the volume?
This is because of the peritumoral inflammation which makes it "fuzzy" around the edges.
A hierarchical cluster system of contouring based on Geet's et al study is under validation and so is background subtracted relative threshold level which aims to overcome this burning issue. But it needs more robust results before it can be recommended.
2) Any standard guidelines for Target volume delineation for PET CT based contouring?
Not to my knowledge (as standard guidelines) but it has been proposed over the past few years. Primarily the issues have revolved around exact delineation (as I mentioned above).
3) In case of NPC if neo adjuvant chemotherapy is given - the volume delineation for 70Gy can be based on post chemotherapy PET or it has to be pre chemo PET volume only?
Lets assume that a given cancerous mass has 109 cells . Remember that best chemotherapy can at best cause about 2-3 decades of cell kill. So your total mass of cells reduces to say about 106.
For CT/PET thats the minimum threshold of detection. Most of the times this is reported as "good response" but in case of NPC, you are dealing with accelerated population of cells ready to spoil the party. It is difficult to recommend chemoradiation followed by adjuvant chemotherapy over NACT and then chemoradiation but I would prefer the former because there is less likelihood of treatment breaks due to accelerated reactions.
I would also explore non-coplanar IMRT aggressively or in case of residual disease, don't forget Brahytherapy and/or stereotactic options. Investigators have also explored role of Gamma Knife.
Hence the current recommendation is CLEARLY the pre-chemotherapy volume (if it has been delivered) but of course you might have to accept dose constraints of critical structures.
4) PET CT avid nodes to be as GTV nodal volume for 70 Gy and rest as high risk volume or the whole level needs to be included as 70 Gy volume?
I would not rely on PET unduly for nodal delineation. Remember necrotic nodes would not show any uptake; look around for nodal architecture.
Your segmentation based on the FDG PET is as good as the segmentation tool applied. Additional histologic studies and correlation of tumor burden needs to be performed before PET can be recommended as standard for nodal delineation.
Hence broadly the recommendations can be summed as:
1) Indicated for detection of unknown primary but upfront FDG PET is not as sensitive as MRI/CECT.
2) Validity for automatic threshold based systems is awaited.
3) FDG PET is used for boost planning; some investigators have utilized this in adaptive radiotherapy.
Hope this helps :)
