If it's still 6000 cGy (absorbed/physical), nothing's different in the patient, just what you *think* the resulting BED is (6600 classical, something higher, probably with "modern modelling").
With modern modelling (non-constant RBE), and a fixed BED prescription, you're delivering less physical/absorbed dose than you would have with a constant RBE model.
But you mention bumping up by 5% to 10% (increase physical dose for classical model, probably close to previous physical dose for modern model because the RBE would be higher in the modern model). What about the OAR? Part of the idea of variable RBE is that the plan would then put the higher RBE in the tumour and the lower RBE in the OAR (LET optimisation).
Not arguing the value of clinical evidence. Just trying to understand your proposal.
What about going back and recalculating effective dose on already treated patients (especially for PBS) given the physical dose that was used (more or less in-silico trial) and review outcome vs. "modern model" RBE?
Yes - intent was 6000 cGy(RBE) - corrected. And OAR and even margins (distal extra PTV) are places to discuss.
The idea lands (I believe) within your 3rd sentence - "with modern" - we need to verify that the "less can be more" within non-constant RBE scenarios can be "validated." To verify the "less" means reducing toxicity - tough at 5% incidence rate, so the concept is to verify the more via kinetics.
Recalculating prior treatments I think is a given on some level. Realistically, I'd file that closer to "no data, very limited data" approach. And maybe that is a path - a slow transition from one to the other, but I think the ideal lies with clinical validation.
Thanks for reading along - sorry about the typo in dosing which is now corrected. Complex topic
Given that it's unlikely that kinetics will have been acquired during prior clinical treatments... in-silico wouldn't get near to what prospective clinical validation with kinetics data acquisition would offer. Beyond my knowledge what the power of the experiment would need to be (number of patients required to get adequate statistics given the hypothesis and means of measuring it), and the additional cost of gathering the data.
Thanks for the update.
Timely discussion of non-constant RBE and what we'll need to incorporate that in to the clinical environment.
FYI - a more complete version of this concept is being written. I tried to put down more of my thinking and data supporting the path into a much more developed document based on comments and feedback. Likely in the next week it will be pushed.
Look forward to seeing that. I went and re-read AAPM TG-256.
It lists ranges of values up to 1.3 for RBE, and I imagine that for certain approaches to PT Arc treatment (it's coming...) the RBE in the tumour will be closer to 1.3 than 1.1, and the RBE elsewhere will be closer to 1.1. That could result in a physical dose difference of 20% between assuming 1.10 everywhere and factoring in the higher RBE (due to LET and whatever else). While Prostate is the whipping boy for sub-optimal use for PT, it might prove to be a useful benchmark for PT Arc, and due to the very significant difference in RBE, for your proposed comparisons.
did you mean 6000 cGy(RBE)/20 Fractions?
If it's still 6000 cGy (absorbed/physical), nothing's different in the patient, just what you *think* the resulting BED is (6600 classical, something higher, probably with "modern modelling").
With modern modelling (non-constant RBE), and a fixed BED prescription, you're delivering less physical/absorbed dose than you would have with a constant RBE model.
But you mention bumping up by 5% to 10% (increase physical dose for classical model, probably close to previous physical dose for modern model because the RBE would be higher in the modern model). What about the OAR? Part of the idea of variable RBE is that the plan would then put the higher RBE in the tumour and the lower RBE in the OAR (LET optimisation).
Not arguing the value of clinical evidence. Just trying to understand your proposal.
What about going back and recalculating effective dose on already treated patients (especially for PBS) given the physical dose that was used (more or less in-silico trial) and review outcome vs. "modern model" RBE?
Yes - intent was 6000 cGy(RBE) - corrected. And OAR and even margins (distal extra PTV) are places to discuss.
The idea lands (I believe) within your 3rd sentence - "with modern" - we need to verify that the "less can be more" within non-constant RBE scenarios can be "validated." To verify the "less" means reducing toxicity - tough at 5% incidence rate, so the concept is to verify the more via kinetics.
Recalculating prior treatments I think is a given on some level. Realistically, I'd file that closer to "no data, very limited data" approach. And maybe that is a path - a slow transition from one to the other, but I think the ideal lies with clinical validation.
Thanks for reading along - sorry about the typo in dosing which is now corrected. Complex topic
Given that it's unlikely that kinetics will have been acquired during prior clinical treatments... in-silico wouldn't get near to what prospective clinical validation with kinetics data acquisition would offer. Beyond my knowledge what the power of the experiment would need to be (number of patients required to get adequate statistics given the hypothesis and means of measuring it), and the additional cost of gathering the data.
Thanks for the update.
Timely discussion of non-constant RBE and what we'll need to incorporate that in to the clinical environment.
FYI - a more complete version of this concept is being written. I tried to put down more of my thinking and data supporting the path into a much more developed document based on comments and feedback. Likely in the next week it will be pushed.
Look forward to seeing that. I went and re-read AAPM TG-256.
It lists ranges of values up to 1.3 for RBE, and I imagine that for certain approaches to PT Arc treatment (it's coming...) the RBE in the tumour will be closer to 1.3 than 1.1, and the RBE elsewhere will be closer to 1.1. That could result in a physical dose difference of 20% between assuming 1.10 everywhere and factoring in the higher RBE (due to LET and whatever else). While Prostate is the whipping boy for sub-optimal use for PT, it might prove to be a useful benchmark for PT Arc, and due to the very significant difference in RBE, for your proposed comparisons.