ProPublica Picks Protons

[grenz]

You are correct, range uncertainty and high LET at end-range are real concerns. They and other considerations (RBE, robust planning, interest in other particles like
Alpha beams and Carbon) are among the reasons why I did a proton fellowship.

I just started using in clinic a variable RBE model, that does not assume RBE =1.1, but rather calculates in the high RBE of 1.3 to 1.5 in the distal tail. I had never seen proton RBE-weighted dose on screen before, and it has already affected my planning.

Case in point, I was recently treating a kid with brainstem tumor. Normal DVH in Eclipse showed max brainstem point dose = 56 Gy, or about 104% of 54 Gy rx dose. The variable RBE plan shows a 61 Gy max dose. Needless to say, we changed the beam arrangement and got it back down around 56 Gy.

I've been waiting a while for that feature, and feel much safer now that it's here.

After the re-plan, did it show the tumor being under covered on the nominal plan you started with (that doesn’t account for RBE)?

---

Members don't see this ad.

 

[A_DeMichele]

We have contributed to all the major trials:

RadComp - node+ breast
COMPPARE - we led the country in Xray prostate patient accruals during many time points
MDACC oropharyngeal trial
NRG GI-006 - esophagus

The biggest problem I have right now is the lack of a major successor to the first 3 trials above. It is a 10 year effort to get a major trial like Radcomp or COMPPARE off the ground and accruing well.

The other big problems in accrual are lack of equipoise in the minds of patients who want to crossover to protons, and insurers, who still refuse to cover patients in a trial

If protons are the experimental arm, why would an insurance company want to cover it?
they wouldn't pay for an experimental drug

 

[IonsAreOurFuture]

After the re-plan, did it show the tumor being under covered on the nominal plan you started with (that doesn’t account for RBE)?

Great question, in both cases, the tumor was well covered.

Unless treating with LET dose painting, the higher RBE will occur after the tumor. This is why we try to never "range-out" into a vital structure. For a brainstem cases, for example, we try not to have more than 1 of 3 beams exiting into brainstem anteriorly.

TG-256 specifically address proton RBE and emphasizes that any new model shall not give a lower dose to tumor or a higher dose to OARS than when operating under the default assumption of RBE=1.1 that we have now.

 

[IonsAreOurFuture]

If protons are the experimental arm, why would an insurance company want to cover it?
they wouldn't pay for an experimental drug

Because protons were FDA approved for cancer in 1988, I don't think it is the same thing as an experimental drug. 300,000+ patients have now been treated with particles, according to PTCOG.

I view the trials as comparing 2 standard of care options, like open vs robotic prostatectomy. Both are FDA approved options, is one better in any way?

1988 is way prior to the invention of most recent advances on a linac, by they way, eg SBRT, IMRT, CBCT, 6DOF couch, MRI linac, IGRT via OBI, etc etc, so are they experimental?

IMRT was adopted as standard of care for parotid sparing in H&N and bladder and rectal sparing in prostate cancer, without any trials, just better looking DVHs.

Most devices are approved without any randomized trials. They need to show safety and efficacy but not always superiority. After that, it's mostly clinical judgment that governs their use. The FDA does not regulate the practice of medicine.

 

[communitydoc13]

1988 is way prior to the invention of most recent advances on a linac, by they way, eg SBRT, IMRT, CBCT, 6DOF couch, MRI linac, IGRT via OBI, etc etc, so are they experimental?

They need to show safety and efficacy

RBE of 1.3 to 1.5 in the distal tail

I'll just leave these thoughts here. All true.

Meanwhile, photon radonc has reaped the downside of high dosimetric certainty, high modulation, high beam orientation degrees of freedom treatment...massively reduced fractionation.

 

[TheWallnerus]

Great question, in both cases, the tumor was well covered.

Unless treating with LET dose painting, the higher RBE will occur after the tumor. This is why we try to never "range-out" into a vital structure. For a brainstem cases, for example, we try not to have more than 1 of 3 beams exiting into brainstem anteriorly.

TG-256 specifically address proton RBE and emphasizes that any new model shall not give a lower dose to tumor or a higher dose to OARS than when operating under the default assumption of RBE=1.1 that we have now.

Wild that I never say I’m operating under a “default assumption” with photons. I mean I guess I do on some things. But never think it.

So RBE=1.1 has the practical effect that a proton gray is less than a photon gray from a pure physics standpoint, yes? Why don’t the proton people just say “we lower the dose by 10% versus photons in all cases?” In all photon cases, we place physics concerns on an equal or higher pedestal than radiobiologic concerns imho. In protons, we start with placing rad bio higher than physics, but the rad bio is an “assumption”… essentially a theory. In science, we should seek to falsify all theories because it’s the only way we can prove them true.

 

[radiaterMike]

Protons are not investigational but rather any potential benefit or detriment over optimally planned IMRT/VMAT is investigational. Protons are costlier, and many CMS coverage criteria mandate cost-effectiveness. And private insurers have cost-effectiveness baked into the coverage criteria. The institutions that have negotiated IMRT rates for proton therapy have only negotiated these rates with select insurers.

 

[grenz]

Wild that I never say I’m operating under a “default assumption” with photons. I mean I guess I do on some things. But never think it.

So RBE=1.1 has the practical effect that a proton gray is less than a photon gray from a pure physics standpoint, yes? Why don’t the proton people just say “we lower the dose by 10% versus photons in all cases?” In all photon cases, we place physics concerns on an equal or higher pedestal than radiobiologic concerns imho. In protons, we start with placing rad bio higher than physics, but the rad bio is an “assumption”… essentially a theory. In science, we should seek to falsify all theories because it’s the only way we can prove them true.

This is why I almost feel like these LET/RBE optimization approaches need a confirmatory clinical trial (standard proton vs advanced optimization). The LET and RBE are all kind of hand wavy in vivo.

 

[communitydoc13]

are all kind of hand wavy

I'm sure @IonsAreOurFuture is conscientious.

However, once you start making adjustments to RBE on the order of 20-40% difference, you are acknowledging remarkable uncertainty in dose within the high dose regions of your plan.

I actually am not sure there is a fix to this. This is an inherent "bug" to protons. There is a lot of biology going on with slow ion interactions.

The feature of high energy photons (not so much the low energy photons that derm uses) is their relative ignorance of biology and even Z. This is what allows for such robust dosimetry.

 

[NotMattSpraker]

Way back, I made this video with PTCOG-NA. I no longer work with them but believe they are still doing this series. The goal of the video was to help teach planning concepts to trainees, but in the first one David Grosshans actually shows this LET optimization in the TPS. It is in the 3rd case, starts at about 15 minutes.

It's pretty interesting stuff. You just turn it on and the DVH changes

If I worked in that department, it would feel like a total black box to me. I'd also want to put people on trial for this reason. I do believe this is being tested on trial at MDACC, at least in CNS tumors.

It's an exciting technique that has a lot of potential for both toxicity sparing and dose escalation if it can become less of a black box.

 

[TheWallnerus]

However, once you start making adjustments to RBE

And again, we somehow got to a point where instead of saying “downward adjustments to dose” we say “upward adjustments to RBE.” It’s a sneaky but seemingly innocuous deceit. Like when Justice Roberts called the individual mandate a tax. Or when congress said the Match is not antitrust. A rad onc physicist gets apoplectic if the delivered dose is off by 10%. Next time there’s a dose misadministration, I need to remember to tell my physicist he didn’t factor in the RBE of the case; then, the dose will be perfect. The RBE of a 6MV photon beam is 0.9, and the RBE of an 18MV beam is 0.7. There is literature to back this… should I start using these RBEs in clinic?

 

[communitydoc13]

should I start using these RBEs in clinic

That would be exceptionally complicated.

RBE is dependent on cell type, endpoint and biological environment. It does not scale with LET in the same way for every cell type or every endpoint.

https://www.sciencedirect.com/science/article/pii/S0167814021067104

Above link to an excellent paper. That this is "emerging" in the 2020s is crazy.

Beware any paper with a TTB (total toxicity burden) endpoint evaluating protons. Always look at the relative risk of really localizable toxicity (brainstem, osteo, pharyngeal necrosis, rib fracture, fistula, rectal bleeding, ulceration).

 

[sirspamalot]

Pro? Tons!

 

[evilbooyaa]

You are correct, range uncertainty and high LET at end-range are real concerns. They and other considerations (RBE, robust planning, interest in other particles like
Alpha beams and Carbon) are among the reasons why I did a proton fellowship.

I just started using in clinic a variable RBE model, that does not assume RBE =1.1, but rather calculates in the high RBE of 1.3 to 1.5 in the distal tail. I had never seen proton RBE-weighted dose on screen before, and it has already affected my planning.

Case in point, I was recently treating a kid with brainstem tumor. Normal DVH in Eclipse showed max brainstem point dose = 56 Gy, or about 104% of 54 Gy rx dose. The variable RBE plan shows a 61 Gy max dose. Needless to say, we changed the beam arrangement and got it back down around 56 Gy.

I've been waiting a while for that feature, and feel much safer now that it's here.

It's always the same story with protons:

1. Protons are better! They just are! Look at the computer screen! Even passive scatter is better than IMRT photons!
2. Huh, passive scatter has maybe more clinical toxicity than IMRT photons in prostate cancer.
3. Well that's because PSPT is garbage! And everyne just needs to use IMPT to treat patients! That's much better
4. Well.... some of these toxicity outcomes are really concerning - here's some data that suggests fixed RBE calcs are putting people at higher risk for toxicity
5. Well that's because fixed RBE planning is garbage! And everyone just needs to use variable RBE calculations to treat patients! That's much better

Even if protons loses in a trial, the answer will be "the protons weren't good enough".

Should all proton centers treating with passive scatter have to close down, since IMPT is much better?

Should all proton centers NOT treating with variable RBE optimization have to close down, since obviously fixed RBE is DANGEROUS for patients to receive near critical OARs? Reminder that that's the main area where protons were supposed to be better.

Is there any published data showing that the variable RBE model is actually 1) valid at calculating dose depositions and 2) robust enough to handle day-to-day variations in set-up in a patient?

Or are folks just experimenting with patients again, and taking the computer screen as gospel, again?

 

[sirspamalot]

Gotts 2 pay for those machines yo.

Keep dem prostate/breast/H&N patients coming in please.. because we need.. I mean to say.. the patients deserve it!

 

[temujim]

It's always the same story with protons:

1. Protons are better! They just are! Look at the computer screen! Even passive scatter is better than IMRT photons!
2. Huh, passive scatter has maybe more clinical toxicity than IMRT photons in prostate cancer.
3. Well that's because PSPT is garbage! And everyne just needs to use IMPT to treat patients! That's much better
4. Well.... some of these toxicity outcomes are really concerning - here's some data that suggests fixed RBE calcs are putting people at higher risk for toxicity
5. Well that's because fixed RBE planning is garbage! And everyone just needs to use variable RBE calculations to treat patients! That's much better

Even if protons loses in a trial, the answer will be "the protons weren't good enough".

Sounds like job security to me. Something we could all use.

 

[IonsAreOurFuture]

This is why I almost feel like these LET/RBE optimization approaches need a confirmatory clinical trial (standard proton vs advanced optimization). The LET and RBE are all kind of hand wavy in vivo.

The RadComp trial of breast CA, n=1,278, and COMPPARE trial, n=2,500, for prostate cancer will have very large comparisons of standard proton (lots of passive scatter, some IMPT). I wish it could have been all IMRT vs all IMPT, instead of largely passive scatter (eg 3D conformal protons).

I hope to see avg 2 year data on side effects and biochemical control for COMPPARE at ASTRO annual mtg, in 1-2 years I suspect. Accrual closed a year ago, Radcomp is about 95% accrued now. Both have taken a long time to get here and will probably not be repeated in our lifetimes.

 

[IonsAreOurFuture]

Way back, I made this video with PTCOG-NA. I no longer work with them but believe they are still doing this series. The goal of the video was to help teach planning concepts to trainees, but in the first one David Grosshans actually shows this LET optimization in the TPS. It is in the 3rd case, starts at about 15 minutes.

It's pretty interesting stuff. You just turn it on and the DVH changes

If I worked in that department, it would feel like a total black box to me. I'd also want to put people on trial for this reason. I do believe this is being tested on trial at MDACC, at least in CNS tumors.

It's an exciting technique that has a lot of potential for both toxicity sparing and dose escalation if it can become less of a black box.

I agree. Nice potential for biological dose escalation by putting the distal Bragg peaks pointing into the GTV core. SIB

 

[grenz]

The RadComp trial of breast CA, n=1,278, and COMPPARE trial, n=2,500, for prostate cancer will have very large comparisons of standard proton (lots of passive scatter, some IMPT). I wish it could have been all IMRT vs all IMPT, instead of largely passive scatter (eg 3D conformal protons).

I hope to see avg 2 year data on side effects and biochemical control for COMPPARE at ASTRO annual mtg, in 1-2 years I suspect. Accrual closed a year ago, Radcomp is about 95% accrued now. Both have taken a long time to get here and will probably not be repeated in our lifetimes.

Why does no one talk about PartiQOL, the actual randomized trial of protons for prostate? Isn’t COMPPARE just a 2-arm cohort study?

 

[IonsAreOurFuture]

Why does no one talk about PartiQOL, the actual randomized trial of protons for prostate? Isn’t COMPPARE just a 2-arm cohort study?

COMPPARE is actually a 4 arm study, with a total of 2500 pts enrolled

1. Xray therapy, typically IMRT, a few centers did SBRT toward the end of recruitment - 1,000 patients in this arm

2. Proton therapy - 900 patients, I think most got passive scatter at the legacy facilities MDACC, UF, Loma Linda

3A Randomized conventional fx protons - 7800 cGy in 39 fx - 300 pts

3B Hypofractionation protons - 6000 cgy in 20 fx like CHiip trial - 300 pts

 

[Chartreuse Wombat]

COMPPARE is actually a 4 arm study, with a total of 2500 pts enrolled

1. Xray therapy, typically IMRT, a few centers did SBRT toward the end of recruitment - 1,000 patients in this arm

2. Proton therapy - 900 patients, I think most got passive scatter at the legacy facilities MDACC, UF, Loma Linda

3A Randomized conventional fx protons - 7800 cGy in 39 fx - 300 pts

3B Hypofractionation protons - 6000 cgy in 20 fx like CHiip trial - 300 pts

Accrual complete and ended 10/31/22.

 

[evilbooyaa]

COMPPARE is actually a 4 arm study, with a total of 2500 pts enrolled

1. Xray therapy, typically IMRT, a few centers did SBRT toward the end of recruitment - 1,000 patients in this arm

2. Proton therapy - 900 patients, I think most got passive scatter at the legacy facilities MDACC, UF, Loma Linda

3A Randomized conventional fx protons - 7800 cGy in 39 fx - 300 pts

3B Hypofractionation protons - 6000 cgy in 20 fx like CHiip trial - 300 pts

Predicted results:
Proton RT was equal or slightly worse than IMRT in terms of clinical toxicity, and non-inferior in terms of oncologic outcomes. 100% of proton patients had a rectal balloon or spacer placed, while 33% of photon patients had a spacer, 0% with daily rectal balloon.

Predicted conclusion:
Obviously IMPT is so much better than PSPT (which is what majority of these poor proton patients received), especially with improvements in variable RBE-based optimization, so we will continue to treat with proton therapy.

 

[sirspamalot]

Predicted results:
Proton RT was equal or slightly worse than IMRT in terms of clinical toxicity, and non-inferior in terms of oncologic outcomes. 100% of proton patients had a rectal balloon or spacer placed, while 33% of photon patients had a spacer, 0% with daily rectal balloon.

Predicted conclusion:
Obviously IMPT is so much better than PSPT (which is what majority of these poor proton patients received), especially with improvements in variable RBE-based optimization, so we will continue to treat with proton therapy.

"Thank you for your support. Please pass the racks this way..."

 

[IonsAreOurFuture]

Predicted results:
Proton RT was equal or slightly worse than IMRT in terms of clinical toxicity, and non-inferior in terms of oncologic outcomes. 100% of proton patients had a rectal balloon or spacer placed, while 33% of photon patients had a spacer, 0% with daily rectal balloon.

Predicted conclusion:
Obviously IMPT is so much better than PSPT (which is what majority of these poor proton patients received), especially with improvements in variable RBE-based optimization, so we will continue to treat with proton therapy.

Actually, SpaceOAR use is about the same in both arms, and higher than I expected, well over half if I recall correctly.

For patients getting prostate-only RT nowadays, it is hard to beat a good IMRT plan, especially with SPACEOAR and a reasonably sized gland. Proton plans really shine for the guys with nodal coverage, because there is no midline dose between the left and right sides of the pelvis. That is a pretty small proportion of guys on the study though.

Rectal balloons are probably doing more harm than good, so I hope neither arm had many.

 

[medgator]

Actually, SpaceOAR use is about the same in both arms, and higher than I expected, well over half if I recall correctly.

For patients getting prostate-only RT nowadays, it is hard to beat a good IMRT plan, especially with SPACEOAR and a reasonably sized gland. Proton plans really shine for the guys with nodal coverage, because there is no midline dose between the left and right sides of the pelvis. That is a pretty small proportion of guys on the study though.

Rectal balloons are probably doing more harm than good, so I hope neither arm had many.

No role for spaceoar routinely imo esp considering the reported toxicities in the literature and the MAUDE database

 

[Upgrayedd]

I agree. Nice potential for biological dose escalation by putting the distal Bragg peaks pointing into the GTV core. SIB

I can’t tell if this is a satire account or not.

A better use of understanding LET characteristics would simply to be able to understand the actual RBE of the plan, and what you see is what you actually get

 

[thesauce]

Predicted results:
Proton RT was equal or slightly worse than IMRT in terms of clinical toxicity, and non-inferior in terms of oncologic outcomes. 100% of proton patients had a rectal balloon or spacer placed, while 33% of photon patients had a spacer, 0% with daily rectal balloon.

Predicted conclusion:
Obviously IMPT is so much better than PSPT (which is what majority of these poor proton patients received), especially with improvements in variable RBE-based optimization, so we will continue to treat with proton therapy.

Agreed. The results section won’t even matter. The conclusion has already been decided.

 

[Doctorer]

I can’t tell if this is a satire account or not.

I can’t figure it out either, and I’m a proton user.

 

[thesauce]

I can’t figure it out either, and I’m a proton user.

I’m curious. Do you tell patients that there is an advantage to protons for prostate cancer? If so, what is it?

 

[deleted1111261]

Ions has an unpopular opinion and is arguing it politely, while people are going ad hominem and pretty rudely. Idk. We say we want contrarian opinions and free / open debate, but I gotta say if I was Ions, I’d just give up trying here. There’s the comment about the Bragg Peak in GTV and it’s theoretical and sorta interesting. This is a specialty that has been fellating the abscopal effect for decades! We are supposedly open to weird ****.

You all know I’m pretty anti-protons. But, if I had to send a patient to someone with protons, it would be someone like them who isn’t overselling. I just read Ions last several posts and can’t see anything that warrants speaking that way.

They are not Frank from MDACC or Nancy Lee. We all have things we say and we all support things that we do, even if not completely 100% evidence based. I hope we all have the confidence to say we are better than the guy 5 miles away - there is no evidence to suggest it, but that’s just natural. And if someone has protons, they aren’t going to just say “screw it, I’m not even going to use them or talk about them any more”. Hopefully, people with protons are judicious, and I get the sense this person is.

I think it would be a loss to this forum to chase them off.

 

[Upgrayedd]

Ions has an unpopular opinion and is arguing it politely, while people are going ad hominem and pretty rudely. Idk. We say we want contrarian opinions and free / open debate, but I gotta say if I was Ions, I’d just give up trying here. There’s the comment about the Bragg Peak in GTV and it’s theoretical and sorta interesting. This is a specialty that has been fellating the abscopal effect for decades! We are supposedly open to weird ****.

You all know I’m pretty anti-protons. But, if I had to send a patient to someone with protons, it would be someone like them who isn’t overselling. I just read Ions last several posts and can’t see anything that warrants speaking that way.

They are not Frank from MDACC or Nancy Lee. We all have things we say and we all support things that we do, even if not completely 100% evidence based. I hope we all have the confidence to say we are better than the guy 5 miles away - there is no evidence to suggest it, but that’s just natural. And if someone has protons, they aren’t going to just say “screw it, I’m not even going to use them or talk about them any more”. Hopefully, people with protons are judicious, and I get the sense this person is.

I think it would be a loss to this forum to chase them off.

Fully agree. I don’t want to chase anyone off. I truly have no idea what the comment about putting the distal bragg peak into the core of the GTV means though. Since the dose deposition over a volume is a series of ranged bragg peaks summed together; ie a spread out bragg peak, there is no distal edge to put in the center of the GTV. The distal bragg peak is no different that the other bragg peaks that built the rest of the dose profile. Except the miscalculations in the bragg peaks that are less distal have less clinical significance. Dose/depth uncertainty probably also has something to do with things. If you want more dose somewhere; push more proton flux to that voxel at whatever energy is needed to range to that depth. The distal bragg peak is distal by definition not convenience.

Its like a loony toons tautology. But I’m open to a proton expert correcting me.

 

[Doctorer]

I’m curious. Do you tell patients that there is an advantage to protons for prostate cancer? If so, what is it?

Thankfully I don’t/don’t have to.

 

[anonymous19027]

Fully agree. I don’t want to chase anyone off. I truly have no idea what the comment about putting the distal bragg peak into the core of the GTV means though. Since the dose deposition over a volume is a series of ranged bragg peaks summed together; ie a spread out bragg peak, there is no distal edge to put in the center of the GTV. The distal bragg peak is no different that the other bragg peaks that built the rest of the dose profile. Except the miscalculations in the bragg peaks that are less distal have less clinical significance. Dose/depth uncertainty probably also has something to do with things. If you want more dose somewhere; push more proton flux to that voxel at whatever energy is needed to range to that depth. The distal bragg peak is distal by definition not convenience.

Its like a loony toons tautology. But I’m open to a proton expert correcting me.

What you are describing with spread out bragg peak is correct but only applies to planning with passive scatter proton therapy. Intensity-modulated protons or pencil beam scanning proton therapy is different and uses singe proton bragg peaks with varying energies to achieve optimal dosimetry. with IMPT you can plan to have distal edge placed in the center of the GTV.

 

[Upgrayedd]

What you are describing with spread out bragg peak is correct but only applies to planning with passive scatter proton therapy. Intensity-modulated protons or pencil beam scanning proton therapy is different and uses singe proton bragg peaks with varying energies to achieve optimal dosimetry. with IMPT you can plan to have distal edge placed in the center of the GTV.

So there is no distal edge on the most distal part if your field, on the most penetrating energy? IMPT redefined physics?

 

[Doctorer]

with IMPT you can plan to have distal edge placed in the center of the GTV.

My guy/girl, read this sentence one more time.

 

[anonymous19027]

So there is no distal edge on the most distal part if your field, on the most penetrating energy? IMPT redefined physics?

No…. Think of it this way, If you are entering patient at a specific gantry angle/patient position with a proton pencil beam where distal edge would be in an undesired OAR after full gtv coverage, you choose to not treat through gtv and place bragg peak it in center of the gtv instead with that individual pencil beam. You would then get coverage of that more distal aspect of gtv beyond the center by doing what you would normally do otherwise with standard 3d planning techniques (rotate collimator, gantry, couch rotation) so this way you minimize the number of pencil beams ranging into oar’s potentially abutting the tumor. This is why treatment time for impt tends to be upward of 30+ min for patients. Does this make sense?

 

[anonymous19027]

My guy/girl, read this sentence one more time.

Clearly you need to read it many more times until it sinks in.

 

[Upgrayedd]

No…. Think of it this way, If you are entering patient at a specific gantry angle/patient position with a proton pencil beam where distal edge would be in an undesired OAR after full gtv coverage, you choose to not treat through gtv and place bragg peak it in center of the gtv instead with that individual pencil beam. You would then get coverage of that more distal aspect of gtv beyond the center by doing what you would normally do otherwise with standard 3d planning techniques (rotate collimator, gantry, couch rotation) so this way you minimize the number of pencil beams ranging into oar’s potentially abutting the tumor. This is why treatment time for impt tends to be upward of 30+ min for patients. Does this make sense?

It does make sense thank you. Still seems a LITTLE silly but makes sense. Like; at that point would photons be better to get that precise dose to the remainder of the distal edge given their lateral edge precision?

It seems that a hybrid photon/proton plan would be the best at that point. Of course I think my point stands if you actually had accurate RBE/LET modeling this would be less of an issue. And that data may be difficult to obtain.

Don’t get me wrong; given your explaination I would certainly feel better about sending patients to people like you as it seems this is much safer. It also looks a lot like a self licking ice cream cone; but I’m not going to try to burn down the entire particle world just for that.

Same thing as our overtraining of residents; we’d be fine with half the resident spots and about one third the number of proton centers.

 

[anonymous19027]

It does make sense thank you. Still seems a LITTLE silly but makes sense. Like; at that point would photons be better to get that precise dose to the remainder of the distal edge given their lateral edge precision?

It seems that a hybrid photon/proton plan would be the best at that point. Of course I think my point stands if you actually had accurate RBE/LET modeling this would be less of an issue. And that data may be difficult to obtain.

Don’t get me wrong; given your explaination I would certainly feel better about sending patients to people like you as it seems this is much safer. It also looks a lot like a self licking ice cream cone; but I’m not going to try to burn down the entire particle world just for that.

Same thing as our overtraining of residents; we’d be fine with half the resident spots and about one third the number of proton centers.

I work at an academic center with IMPT available and I can confidently say I think there is a 'dosimetric' benefit for its use in about 5% of patients. Do I think this has a benefit for prostate cancer? No. Breast? hell no. Pediatric and some rare radioresistant tumors where there is some established dose-response? absolutely. We actually do a fair amount of mixed photons/protons plans, too, so your point is well taken.

 

[Doctorer]

Clearly you need to read it many more times until it sinks in.

You’re still limited by the number of beams that you use/treatment times. You’re going to select beam arrangements/optimize so that you’re not ranging in to the brain stem, for example. Yes impt helps, but it’s not a cure.

 

[anonymous19027]

You’re still limited by the number of beams that you use/treatment times. You’re going to select beam arrangements/optimize so that you’re not ranging in to the brain stem, for example. Yes impt helps, but it’s not a cure.

no one is arguing that. nor is anyone claiming its a cure.

 

[TheWallnerus]

Do I think this has a benefit for prostate cancer? No. Breast? hell no.

And yet breast cancers are proton centers’ most common treatment indication.

All proton center physicians seem to denounce protons for breast. Well not “all.” Tim Williams was recently talking about the life saving benefits of proton therapy for stage one 15 fraction partial breast RT.

 

[communitydoc13]

Does this make sense?

No. (edit: maybe?)

The Bragg peak is just a feature of the dose/depth curve for protons statistically. It does not exist for a single proton. Of course, it's location depends on beam energy.

As mentioned above and in many places, the far side of the Bragg peak seems to be a region where a lot of biology goes on (manifested simply as highly variable and uncertain RBEs in this region). There is a region of uncertainty for any constituent proton beam.

Any pure proton plan is presumably going to achieve coverage through can aggregate of beams of protons, which have dosimetry that includes a bragg peak within the PTV.

Now, are you telling us that the plans with IMPT include beams where the Bragg peak exists outside of the PTV? This would seem weird and to not be taking advantage of what protons bring to the table. Photon plans are made up of beams where Dmax is outside of the PTV almost always, but this not an area of dosimetric uncertainty and we can have confidence in the aggregate dosimetry of an essentially infinite beam number.

My guess is that the solutions that are come upon with IMPT are uniformly made up of beams whose Bragg peak is within the planning target volume. (Mathematically speaking, it would be like saying that the basis set for planning, includes only beams with a Bragg peak within the target volume).

It seems that a hybrid photon/proton plan would be the best at that point.

This makes sense. One could come up with a solution regarding the proton portion of the plan where only a portion of the PTV is covered by protons, and the region of dosimetric uncertainty is within say the GTV. This will be a "proton boost" of uncertain dose but potentially with clear dosimetric advantage regarding nearby OARs (not eating up big 30 Gy volumes in healthy liver or a previously irradiated brainstem for instance).

Probably 4 centers nationally doing this level stuff makes sense. I would not advocate for a boost like this in the overwhelming majority of CNS tumors, prostates, lungs, head and necks or other.

edit: how many gantry positions do you actually use for IMPT? I am not aware of there being anything like RA or even a 6-7 beam IMRT arrangement for IMPT. Usually 2, potentially 3-4 gantry positions? Factoring in range uncertainty as well, it is just not clear to me how you can ensure that the highly uncertain portion of the plan is within the interior portion of the PTV? and it better be uncertain and hot if with the interior.

 

[temujim]

It seems that a hybrid photon/proton plan would be the best at that point. Of course I think my point stands if you actually had accurate RBE/LET modeling this would be less of an issue. And that data may be difficult to obtain.

There are papers out there with hybrid photon/proton plans (eg PMID: 37913953), though this is primarily aimed at NCTP reduction. Photon contribution drops with newer techniques such as arcs. Imagine slow throughput even slower with mixed beam plans would inhibit practicality.

 

[evilbooyaa]

I work at an academic center with IMPT available and I can confidently say I think there is a 'dosimetric' benefit for its use in about 5% of patients. Do I think this has a benefit for prostate cancer? No. Breast? hell no. Pediatric and some rare radioresistant tumors where there is some established dose-response? absolutely. We actually do a fair amount of mixed photons/protons plans, too, so your point is well taken.

I strongly value the input of both you and @IonsAreOurFuture on this forum. I won't pretend to know everything about protons, so I look to folks who aren't as obviously profiting (in some manner) from their use (which is easier when everyone is anonymous) to either agree with my conceptions, or challenge my (perceived) misconceptions.

 

[IonsAreOurFuture]

No. (edit: maybe?)

The Bragg peak is just a feature of the dose/depth curve for protons statistically. It does not exist for a single proton. Of course, it's location depends on beam energy.

As mentioned above and in many places, the far side of the Bragg peak seems to be a region where a lot of biology goes on (manifested simply as highly variable and uncertain RBEs in this region). There is a region of uncertainty for any constituent proton beam.

Any pure proton plan is presumably going to achieve coverage through can aggregate of beams of protons, which have dosimetry that includes a bragg peak within the PTV.

Now, are you telling us that the plans with IMPT include beams where the Bragg peak exists outside of the PTV? This would seem weird and to not be taking advantage of what protons bring to the table. Photon plans are made up of beams where Dmax is outside of the PTV almost always, but this not an area of dosimetric uncertainty and we can have confidence in the aggregate dosimetry of an essentially infinite beam number.

My guess is that the solutions that are come upon with IMPT are uniformly made up of beams whose Bragg peak is within the planning target volume. (Mathematically speaking, it would be like saying that the basis set for planning, includes only beams with a Bragg peak within the target volume).



This makes sense. One could come up with a solution regarding the proton portion of the plan where only a portion of the PTV is covered by protons, and the region of dosimetric uncertainty is within say the GTV. This will be a "proton boost" of uncertain dose but potentially with clear dosimetric advantage regarding nearby OARs (not eating up big 30 Gy volumes in healthy liver or a previously irradiated brainstem for instance).

Probably 4 centers nationally doing this level stuff makes sense. I would not advocate for a boost like this in the overwhelming majority of CNS tumors, prostates, lungs, head and necks or other.

edit: how many gantry positions do you actually use for IMPT? I am not aware of there being anything like RA or even a 6-7 beam IMRT arrangement for IMPT. Usually 2, potentially 3-4 gantry positions? Factoring in range uncertainty as well, it is just not clear to me how you can ensure that the highly uncertain portion of the plan is within the interior portion of the PTV? and it better be uncertain and hot if with the interior.

You bring up several important points! Thank you for thinking about these issues, they also occupy a lot of my free thinking time.

1. Where we put the Bragg peaks and how they are oriented really matters. Especially with passive scatter - this has not been done well historically.

Someone above commented about why does the distal most Bragg peak matter? It is because it is the only "pristine" Bragg peak in a Spread Out Bragg peak. It has full physical dose (100%) because its leading edge does not overlap with the tail of any other peaks, and so its RBE is "undiluted", you get the full 1.3 to 1.5 Gy in the normal anatomy beyond the PTV.

Ideally you want that high RBE only in the GTV, which is possible with IMPT but not passive scatter. I will sometimes try to "trick" the optimizer into doing this by creating an SIB

PS, this is not a fake or satire account.

 

[drowsy12]

can you explain how proton can possibly be advantageous in pancreatic cancer? are you able to achieve as tight of a fall off adjacent to critical structures like the duodenum as you could with VMAT? In my experience with my friends who have proton, this is always a VMAT case.

 

[TheWallnerus]

You bring up several important points! Thank you for thinking about these issues, they also occupy a lot of my free thinking time.

1. Where we put the Bragg peaks and how they are oriented really matters. Especially with passive scatter - this has not been done well historically.

Someone above commented about why does the distal most Bragg peak matter? It is because it is the only "pristine" Bragg peak in a Spread Out Bragg peak. It has full physical dose (100%) because its leading edge does not overlap with the tail of any other peaks, and so its RBE is "undiluted", you get the full 1.3 to 1.5 Gy in the normal anatomy beyond the PTV.

Ideally you want that high RBE only in the GTV, which is possible with IMPT but not passive scatter. I will sometimes try to "trick" the optimizer into doing this by creating an SIB

PS, this is not a fake or satire account.

What do you mean by the full 1.3 to 1.5 Gy
You mean RBE?

But either way. “The full 1.3 to 1.5”… 1.5 is quite more full than 1.3. I would put a “lol” after that, but in all seriousness, we should have way more precision in these discussions than we do. But instead people just shrug and say “well there are a lot of unknowns.” And then the proton therapist hits the beam-on button, “cha ching,” and the anxiety of dealing with all those unknowns slowly subsides.

 

[RickyScott]

can you explain how proton can possibly be advantageous in pancreatic cancer? are you able to achieve as tight of a fall off adjacent to critical structures like the duodenum as you could with VMAT? In my experience with my friends who have proton, this is always a VMAT case.

it seems protons are never a good idea when ptv immediately adjacent/overlap with critical structures (prostate/pancreas).

 

[ramsesthenice]

it seems protons are never a good idea when ptv immediately adjacent/overlap with critical structures (prostate/pancreas).

Im not a proton person and I don't actually think the logic for pancreas will pan out but theoretically it is at least conceptually not that different from what we are doing. The logic is with the higher BED you can try to dose escalate as much of the GTV as possible. For all the reasons stated, you would have to accept lower dosing to the margins but in fairness, that is what we do with most of the MRI-guided SBRT or Crane style dose escalation approaches. They are essentially dose-painted plans to get as much of the GTV as safely possible to a high dose. Its potentially doable with protons but will it be better? I doubt it. I like dose escalated SBRT because its generally very well tolerated and works at least as well if not a little better than conventional fractionation. But people still recur and progress. It has not been as game changing as I had hoped and I don't see protons doing any better. I think PNI and marginal sparing (which is necessary) limit how much we can accomplish with any dose escalation techniques.