Transcript
Yes, I would like to report on the treatment planning study we are doing on radiosurgery, trying to obtain a higher tumor dose without getting a higher normal tissue gEUD. And we're doing that with these people, and of these I am the physicist. So, if you would like to take that into account if you formulate your questions. What is the problem? The problem is that dose levels for brain metastases are often based on RTOG 90-05 study, which means that the larger target volumes will receive a lower dose, and as a consequence, as you could also see yesterday in the talk of my colleague, Dr. Vibake [SP], this results in a higher failure rate for this larger metastases. Another observation is in this study that the failure rate seems to be lower for Gamma Knife treatment than they are for Linac treatments, and we think that might be due to the dose prescription. Often Gamma Knife plans are prescribed to the 50% isodose with 100% in the isocenter, while a Linac plans, in our institute, are prescribed to 80% and sometimes set over places even to 90%. So, this will lead to a different dose distribution, obviously.
So, the purpose of our study was to find a prescription isodose and dose level which would result in a higher target dose without compromising the normal tissue, the normal brain tissue around it. The material we used for that, we had five patients for which the CTV was contoured on MRI, and around this CTV we placed a two-millimeter margin to create the PTV to take geometric uncertainties into account. And the resulting PTV sizes ranged from 1 to 15 cc so far. And we also contoured the normal brain, which we defined here as the total brain minus the CTV.
After this, we made treatment plans using the dynamic conformal arc technique using BrainSCAN 5.31, and we made for each patient 4 plans with different prescription isodoses and composing the PTV. Those were either the 50, 60, 70, or 80% isodose, and always the isocenter dose was kept at 100%. After this, we calculated the gEUD for CTV, PTV, and for normal brain. Well, what is this gEUD? That's the biological part of this story. It stands for generalized Equivalent Uniform Dose, and it is actually a model to calculate the uniform dose level that would have the same biological effects on the volume as the actual non-uniform dose distribution. So, you try to calculate an inhomogeneous dose distribution back to homogeneous dose distribution with the same biological effects. This model was developed by Andrzej Niemierko around '95, I think. And as a physicist, I thought I had to include at least one equation, and this is it.
This gEUD model can be used for both tumor and for normal tissue, and there are different model parameters. Actually, there's only one model parameter per tissue for different organs or different tumor types, and the parameters that we used right now is an A of -10 for the target volumes and an A of 4.6 for the normal brain. And that should be related to necrosis. As we make different treatment plans with lower prescription isodose, with lower isodose and composing the PTV that will lead to a higher dose in the normal brain tissue, which is not what we want. So, after the first treatment planning, we set the prescription dose such that the normal brain gEUD would be the same as for the original 80% plan, which is our clinical plan. And after that, we calculate a new isocenter dose and a gEUD for CTV and PTV.
Now, what does this look like? This is the result for one patient, an example, with the 80% plan as was used clinically, and here the 50%. And as you can see here, the 80% isodose encompasses the PTV and here it is, the 50% isodose. In DVHs, this looks like this. So, right now all plans have the same minimum dose, which is the dose which encompasses the PTV, but for the rest of the target volume, the doses differ according to the different prescription isodoses. And you also see some slight differences in the DVHs of the normal brain tissue. This is the same...well, this is the gEUDs calculated from these DVHs, and as you see here, the normal brain gEUD goes up if you go to a lower prescription isodose, and also the gEUD for PTV and CTV go up. And as I said, this is not what we want, so now we renormalize these three plans. So, that's the gEUD will be the same as for the 80% plan. And this is what you get in that case. So, now these are the same. The PTV gEUD goes up at lower prescription isodoses, and then down again at 50%, but the CTV gEUD continues to go up as does the new isocenter dose. These are the resulting DVHs. So, now you see that minimum doses is not equal to 18 gray anymore for all the plans. So, they cross somewhere halfway, and you can also see that most of the normal brain DVHs are very similar only at this very low dose where you see some difference.
These were the results of one patient, as an example. These are the results averaged over the five patients that we did the treatment planning on so far. And as mentioned...well, this is still the same for these five patients. If you go to a lower prescription percentage, then the gEUD for normal brain goes up, but also the gEUD for the PTV goes up. This is not what we wanted, so we renormalized, then the gEUD for the normal brain is constant. And again, you see that the gEUD for the PTV first goes up, and then down again at 50%. If we add the CTV gEUD to that, you see that this continues to go up to the lower percentages, as does the isocenter dose. These are the results of these five patients in table form. These are the original plans with the gEUD of the normal brain going up, as well for the PTV. After renormalization, you see that the gEUD for the PTV goes up and then down again, but gEUD for CTV and isocenter dose continue to go up at lower percentages.
So, the conclusion so far from this study is that we think that the optimal prescription iso percentage would be 60% combined with an 11% lower prescription dose. This would result in a gEUD for the PTV, which is 3% higher than we have right now, at 80% prescription isodose. It would result in a gEUD of 50% higher for the CTV and an isocenter dose 20% higher, and without any increase of the normal brain gEUD. Some discussion points. Well, you might wonder which would be the most relevant gEUD, the one for the PTV or the CTV? Well, we think that probably the most resistant tumor cells because of hypoxia will be in the center of the CTV. And we also think because of the stereotactic positioning, that the most likely position of the CTV will be centrally. So, even though we don't want to treat the PTV worse than we're doing right now, we think that the dose increase to the CTV will be more relevant than the small dose increase to the PTV. And another important question is what to do next, should we start clinically with this new prescription already or should we first start comparative study comparing the two dose prescriptions? But before we have to answer that question, we will at least need more patient data, and especially with the larger target sizes. Thank you for your attention.
So, the purpose of our study was to find a prescription isodose and dose level which would result in a higher target dose without compromising the normal tissue, the normal brain tissue around it. The material we used for that, we had five patients for which the CTV was contoured on MRI, and around this CTV we placed a two-millimeter margin to create the PTV to take geometric uncertainties into account. And the resulting PTV sizes ranged from 1 to 15 cc so far. And we also contoured the normal brain, which we defined here as the total brain minus the CTV.
After this, we made treatment plans using the dynamic conformal arc technique using BrainSCAN 5.31, and we made for each patient 4 plans with different prescription isodoses and composing the PTV. Those were either the 50, 60, 70, or 80% isodose, and always the isocenter dose was kept at 100%. After this, we calculated the gEUD for CTV, PTV, and for normal brain. Well, what is this gEUD? That's the biological part of this story. It stands for generalized Equivalent Uniform Dose, and it is actually a model to calculate the uniform dose level that would have the same biological effects on the volume as the actual non-uniform dose distribution. So, you try to calculate an inhomogeneous dose distribution back to homogeneous dose distribution with the same biological effects. This model was developed by Andrzej Niemierko around '95, I think. And as a physicist, I thought I had to include at least one equation, and this is it.
This gEUD model can be used for both tumor and for normal tissue, and there are different model parameters. Actually, there's only one model parameter per tissue for different organs or different tumor types, and the parameters that we used right now is an A of -10 for the target volumes and an A of 4.6 for the normal brain. And that should be related to necrosis. As we make different treatment plans with lower prescription isodose, with lower isodose and composing the PTV that will lead to a higher dose in the normal brain tissue, which is not what we want. So, after the first treatment planning, we set the prescription dose such that the normal brain gEUD would be the same as for the original 80% plan, which is our clinical plan. And after that, we calculate a new isocenter dose and a gEUD for CTV and PTV.
Now, what does this look like? This is the result for one patient, an example, with the 80% plan as was used clinically, and here the 50%. And as you can see here, the 80% isodose encompasses the PTV and here it is, the 50% isodose. In DVHs, this looks like this. So, right now all plans have the same minimum dose, which is the dose which encompasses the PTV, but for the rest of the target volume, the doses differ according to the different prescription isodoses. And you also see some slight differences in the DVHs of the normal brain tissue. This is the same...well, this is the gEUDs calculated from these DVHs, and as you see here, the normal brain gEUD goes up if you go to a lower prescription isodose, and also the gEUD for PTV and CTV go up. And as I said, this is not what we want, so now we renormalize these three plans. So, that's the gEUD will be the same as for the 80% plan. And this is what you get in that case. So, now these are the same. The PTV gEUD goes up at lower prescription isodoses, and then down again at 50%, but the CTV gEUD continues to go up as does the new isocenter dose. These are the resulting DVHs. So, now you see that minimum doses is not equal to 18 gray anymore for all the plans. So, they cross somewhere halfway, and you can also see that most of the normal brain DVHs are very similar only at this very low dose where you see some difference.
These were the results of one patient, as an example. These are the results averaged over the five patients that we did the treatment planning on so far. And as mentioned...well, this is still the same for these five patients. If you go to a lower prescription percentage, then the gEUD for normal brain goes up, but also the gEUD for the PTV goes up. This is not what we wanted, so we renormalized, then the gEUD for the normal brain is constant. And again, you see that the gEUD for the PTV first goes up, and then down again at 50%. If we add the CTV gEUD to that, you see that this continues to go up to the lower percentages, as does the isocenter dose. These are the results of these five patients in table form. These are the original plans with the gEUD of the normal brain going up, as well for the PTV. After renormalization, you see that the gEUD for the PTV goes up and then down again, but gEUD for CTV and isocenter dose continue to go up at lower percentages.
So, the conclusion so far from this study is that we think that the optimal prescription iso percentage would be 60% combined with an 11% lower prescription dose. This would result in a gEUD for the PTV, which is 3% higher than we have right now, at 80% prescription isodose. It would result in a gEUD of 50% higher for the CTV and an isocenter dose 20% higher, and without any increase of the normal brain gEUD. Some discussion points. Well, you might wonder which would be the most relevant gEUD, the one for the PTV or the CTV? Well, we think that probably the most resistant tumor cells because of hypoxia will be in the center of the CTV. And we also think because of the stereotactic positioning, that the most likely position of the CTV will be centrally. So, even though we don't want to treat the PTV worse than we're doing right now, we think that the dose increase to the CTV will be more relevant than the small dose increase to the PTV. And another important question is what to do next, should we start clinically with this new prescription already or should we first start comparative study comparing the two dose prescriptions? But before we have to answer that question, we will at least need more patient data, and especially with the larger target sizes. Thank you for your attention.