Finally, it's here! LET-based Arc proton therapy

[IonsAreOurFuture]

To the credit of everyone here and elsewhere in our field who has been rightly concerned about the previously invisible proton hotspots of high LET resulting in rib fractures, brainstem damage or mandible ORN, I'm glad to see that one vendor actually cares about it and has introduced LET-based proton optimization:

"RayStation 2023B also comes with two innovative features within particle therapy: support for discrete proton arcs and robust optimization based on linear energy transfer (LET). LET optimization can be used in proton and ion therapy to lower LET in risk organs, which could reduce potential side effects of radiotherapy. Additionally, for treatments with heavy ions, such as carbon ions, particles with high LET can be focused inside the tumor, increasing the probability of tumor control."

RaySearch releases RayStation 2023B | RaySearch Laboratories

<p><span><span><span>Radiation therapy becomes more and more tailored to each patient with highly conformal dose delivery to the tumor and reduced exp

www.raysearchlabs.com

It's not a cure-all but at least there is now official vendor support for LET-dose painting to take high LET out of the organs at risk and put it into the GTV where it belongs.

The simultaneous launch of proton arc planning is a huge deal as well, considering that most proton patients have historically been treated with just 2 or 3 beams a day.

Arc is synergistic with LET-dose painting, since it lets you bring in the "paintbrush" from any angle you want and drop the high LET where you need it to go.

---

Members don't see this ad.

 

[IonsAreOurFuture]

The implementation of arc-based IMPT and LET dose painting could be a game changer for our field in actually lowering the integral dose rather than just pushing it around inside the body as IMRT and VMAT did.

The pristine (undiluted) Bragg peak at the leading edge of a SOBP has been a major obstacle to OAR sparing with protons, because it has always dumped the highest LET beyond the tumor and not inside it. Now we should be able to pack the GTV with those pristine Bragg peaks of RBE 1.2 to 1.4 and keep the RBE outside to 1.0-1.1, if practice follows theory. (I know, I know, it doesn't always work out that way)

It's like a biological boost of 25% to the core of a tumor with potentially little downside.

 

[TheWallnerus]

The implementation of arc-based IMPT and LET dose painting could be a game changer for our field

After the game is changed a thousand and one times... does it just flip right 'round again to the same game?

Riffing on my SDN "Custom Title"...

What does e^(i*1001*pi)+1 equal

 

[NotMattSpraker]

The implementation of arc-based IMPT and LET dose painting could be a game changer for our field in actually lowering the integral dose rather than just pushing it around inside the body as IMRT and VMAT did.

The pristine (undiluted) Bragg peak at the leading edge of a SOBP has been a major obstacle to OAR sparing with protons, because it has always dumped the highest LET beyond the tumor and not inside it. Now we should be able to pack the GTV with those pristine Bragg peaks of RBE 1.2 to 1.4 and keep the RBE outside to 1.0-1.1, if practice follows theory. (I know, I know, it doesn't always work out that way)

It's like a biological boost of 25% to the core of a tumor with potentially little downside.

Always appreciate the update.

Im curious how this is being tested and "validated" clinically. I heard of at least one trial at MDACC. Can you share thoughts there?

 

[IonsAreOurFuture]

Always appreciate the update.

Im curious how this is being tested and "validated" clinically. I heard of at least one trial at MDACC. Can you share thoughts there?

I don't think any human has yet been treated with a moving particle arc, except for electron arc therapy of post mastectomy chestwalls at the University of Utah before year 2000 or so. I think it petered out once IMRT came around.

There is this review article from the Green Journal, free open access, but if anyone is aware of patient use, I would love to learn more:

https://www.thegreenjournal.com/article/S0167-8140(24)00704-7/fulltext

 

[radiation]

how does arc based protons make any sense? The main advantage of protons is reducing low/med dose bath - the more entrance beams you have the more you lose that benefit. making the tumor 200% hot inside is easy to do with photons and is standard for sbrt/srs.

 

[OTN]

The implementation of arc-based IMPT and LET dose painting could be a game changer for our field in actually lowering the integral dose rather than just pushing it around inside the body as IMRT and VMAT did.

The pristine (undiluted) Bragg peak at the leading edge of a SOBP has been a major obstacle to OAR sparing with protons, because it has always dumped the highest LET beyond the tumor and not inside it. Now we should be able to pack the GTV with those pristine Bragg peaks of RBE 1.2 to 1.4 and keep the RBE outside to 1.0-1.1, if practice follows theory. (I know, I know, it doesn't always work out that way)

It's like a biological boost of 25% to the core of a tumor with potentially little downside.

Wait, so with current proton planning the highest LET is delivered outside of the target area?

 

[drowsy12]

how does arc based protons make any sense? The main advantage of protons is reducing low/med dose bath - the more entrance beams you have the more you lose that benefit. making the tumor 200% hot inside is easy to do with photons and is standard for sbrt/srs.

There is a trade off to a bit more low dose bath to having more conformity of the high dose target

Currently if a target is next to a OAR and a high dose gradient is needed, vmat beats proton all day every day. The Rx dose volume is bigger with proton than photon.

Proton arc therapy if it ever actually materializes will change the game for what proton can do

 

[deleted1180461]

how does arc based protons make any sense? The main advantage of protons is reducing low/med dose bath - the more entrance beams you have the more you lose that benefit. making the tumor 200% hot inside is easy to do with photons and is standard for sbrt/srs.

Entrance beams still have much lower dose deposition on the way to the target and no exit. So yeah, there would still be some dose bath from the entrance beams all around, but it would be half of the IMRT plan and no exit, so altogether having never looked at these plans before I'd still guess, what 70-80% lower dose bath than VMAT?

 

[thecarbonionangle]

Entrance beams still have much lower dose deposition on the way to the target and no exit. So yeah, there would still be some dose bath from the entrance beams all around, but it would be half of the IMRT plan and no exit, so altogether having never looked at these plans before I'd still guess, what 70-80% lower dose bath than VMAT?

Yes, exactly. It would not be the exact same as VMAT. It would be better. Of course you can still have the same argument if this is “clinically significant”. I believe it will be with more immunotherapy in the future, more re-treatments, more polymetastatic patients getting multiple courses, etc.

 

[communitydoc13]

Press release over a year ago. Uncertainties can of course compound.

A couple points.

Wait, so with current proton planning the highest LET is delivered outside of the target area?

Highest LET is on the "backside of the Bragg Peak". This does not mean the highest dose. This is a low dose area with a very high LET and anticipated increase in RBE and uncertainty. Presently (per my understanding as I'm not a user) protonists use intuition about where to let this area fall relative to adjacent OARs because it makes them squeamish (appropriately so).

While LET scales with RBE, there is much more to this.

Investigating Dependencies of Relative Biological Effectiveness for Proton Therapy in Cancer Cells

Relative biological effectiveness (RBE) accounts for the differences in biological effect from different radiation types. The RBE for proton therapy remains uncertain, as it has been shown to vary from the clinically used value of 1.1. In this work we ...

www.ncbi.nlm.nih.gov

I've linked to this paper before. All proton dosimetry is based on photon dosimetry. RBE in the limited cell work out there is all over the place (like really out there...factors of 5 type stuff not 1.1 vs 1.4 type stuff). Protons are discriminatory chemical entities in ways that photons are not.

I am unconvinced that meaningful uncertainties are accounted for in any of these newly marketed products.

But damning to our community in general is the marked difference in plans generated with LET optimization vs just dosimetric plans (what most protonists have been doing for most of the time). There are lots of papers out there...just google them. Holy cow. It's not that we know that the LET optimized plans are correct (we don't know that), it's that the physics community has known about LET dependence along the path of the photon beam forever and the accounting of this one uncertainty alone results in wildly different plans.

IMO, no doc should have ever been looking at "dosimetric" proton plans ever and been taking them seriously.

 

[TheWallnerus]

Press release over a year ago. Uncertainties can of course compound.

A couple points.



Highest LET is on the "backside of the Bragg Peak". This does not mean the highest dose. This is a low dose area with a very high LET and anticipated increase in RBE and uncertainty. Presently (per my understanding as I'm not a user) protonists use intuition about where to let this area fall relative to adjacent OARs because it makes them squeamish (appropriately so).

While LET scales with RBE, there is much more to this.

Investigating Dependencies of Relative Biological Effectiveness for Proton Therapy in Cancer Cells

Relative biological effectiveness (RBE) accounts for the differences in biological effect from different radiation types. The RBE for proton therapy remains uncertain, as it has been shown to vary from the clinically used value of 1.1. In this work we ...

www.ncbi.nlm.nih.gov

I've linked to this paper before. All proton dosimetry is based on photon dosimetry. RBE in the limited cell work out there is all over the place (like really out there...factors of 5 type stuff not 1.1 vs 1.4 type stuff). Protons are discriminatory chemical entities in ways that photons are not.

I am unconvinced that meaningful uncertainties are accounted for in any of these newly marketed products.

But damning to our community in general is the marked difference in plans generated with LET optimization vs just dosimetric plans (what most protonists have been doing for most of the time). There are lots of papers out there...just google them. Holy cow. It's not that we know that the LET optimized plans are correct (we don't know that), it's that the physics community has known about LET dependence along the path of the photon beam forever and the accounting of this one uncertainty alone results in wildly different plans.

IMO, no doc should have ever been looking at "dosimetric" proton plans ever and been taking them seriously.

Sounds great

Meanwhile my physicist rejects an IMRT plan if the QA measures more than a couple percent off from expected

 

[NotMattSpraker]

I am unconvinced that meaningful uncertainties are accounted for in any of these newly marketed products.

I agree. I was asking about how they are testing these LTE op strategies clinically, I dont really care about arc therapy.

Yes, exactly. It would not be the exact same as VMAT. It would be better. Of course you can still have the same argument if this is “clinically significant”. I believe it will be with more immunotherapy in the future, more re-treatments, more polymetastatic patients getting multiple courses, etc.

The Reflexion guys had a much stronger argument in this realm than the proton guys do in my opinion. I know when Im delivering treatment in the context of high toxicity risk (i.e. reirradiation), I want my therapy to be as wildly physically and biologically unpredictable as possible.

That said, the proton guys have distorted reality much stronger than everyone else. They are the original "Trust me" bros, disregarding the need for randomized evidence for literal decades. Protons has never really been shown to be clinically better than VMAT but its just better. Arcing them... EVEN better.

Frankly once you get over the disappointment in the field, its pretty entertaining to watch all the Rad Oncs compete to overstate the benefits of whatever technology theyve decided to make in to their identity

On that note, did you see this advertisement thoughtful editorial? I hope Eleckta is crafting a LTE to remind everyone that they too make MRgRT machines!

MRI-Linac–Based Radiotherapy—Promising or Hype?

This Viewpoint discusses the value of magnetic resonance imaging with linear accelerator (MRI-linac) technology, given the unmet needs in its development.

jamanetwork.com

 

[NotMattSpraker]

Woops I am sorry, here is the unusually written editorial. The other paper is kind of a rebuttal I think.

When Best Care Takes a Back Seat to the Bottom Line

This Viewpoint discusses the clinical vs economic success of magnetic resonance imaging combined with linear accelerator technology and its impact on patients.

jamanetwork.com

 

[BobbyHeenan]

Woops I am sorry, here is the unusually written editorial. The other paper is kind of a rebuttal I think.

When Best Care Takes a Back Seat to the Bottom Line

This Viewpoint discusses the clinical vs economic success of magnetic resonance imaging combined with linear accelerator technology and its impact on patients.

jamanetwork.com

THe MRI linac team and then the rebuttal written by authors with "particle therapy" program titles

 

[NotMattSpraker]

THe MRI linac team and then the rebuttal written by authors with "particle therapy" program titles

View attachment 390487

Don't miss the part where the chair of Dana Farber and MGH declares ownership interest in a mystery company with a tele-forward name.

 

[IonsAreOurFuture]

Wait, so with current proton planning the highest LET is delivered outside of the target area?

Yes, as the protons decelerate at their end of range, they deposit more physical dose (manifest as a Bragg peak) but also more biological dose - higher RBE.

The increase in LET/RBE at end of range is because at low proton energies, like alpha particles, they become more densely ionizing and cause some DNA DSB instead of almost exclusively SSB.

Because we always treat with a margin, the protons will ultimately stop beyond the distal edge of the tumor, i.e. stop in normal tissue. That is one reason why you try not to terminate your beam aiming at the spinal cord or brainstem; or if you must treat the brainstem, like for a medulloblastoma boost, use at least 3 beams like a PA and 2 laterals. That way you can "dilute" the high RBE down to roughly 1/3rd in any one direction; so instead of a 30% RBE hotspot in one area from just a single beam, it's more like 10% in 3 areas.

With LET dose painting, placement of these hotspots should be modifiable, so we will be able to both see and move the LET hotspots into the tumor, but more importantly, out of the OARs.

 

[IonsAreOurFuture]

THe MRI linac team and then the rebuttal written by authors with "particle therapy" program titles

View attachment 390487

I'm going to present a Big Tent platform that everyone can get on board with:

"MRI-guided VMAT Proton SBRT with daily online adaptive!"

That should check most people's boxes, or at least, most medical physicists.

"Sorry brachy and PETCT linac, you didn't make the cut; we'll have to lump you in with, umm, theranostics...so head on down to Nuc Med now."

 

[ramsesthenice]

Yes, as the protons decelerate at their end of range, they deposit more physical dose (manifest as a Bragg peak) but also more biological dose - higher RBE.

The increase in LET/RBE at end of range is because at low proton energies, like alpha particles, they become more densely ionizing and cause some DNA DSB instead of almost exclusively SSB.

Because we always treat with a margin, the protons will ultimately stop beyond the distal edge of the tumor, i.e. stop in normal tissue. That is one reason why you try not to terminate your beam aiming at the spinal cord or brainstem; or if you must treat the brainstem, like for a medulloblastoma boost, use at least 3 beams like a PA and 2 laterals. That way you can "dilute" the high RBE down to roughly 1/3rd in any one direction; so instead of a 30% RBE hotspot in one area from just a single beam, it's more like 10% in 3 areas.

With LET dose painting, placement of these hotspots should be modifiable, so we will be able to both see and move the LET hotspots into the tumor, but more importantly, out of the OARs.

You make plausible arguments and I sincerely hope something positive comes from this. The reality is that most plausible arguments are wrong. I’m not trying to be a naysayer. I work with MRLinacs (as in, we have one so I use it…I didn’t set out to be an expert in this or make it a major focus of my research) and while I think most of what people are proposing to do with them is total BS, isotoxic adaptive dose escalation could actually offer some meaningful benefit in the right cases. It’s pretty much what you are proposing. Get as much dose to as much of the target as you can and forgo the conventional 95% coverage paradigm. It’s basically Brachy with EBRT. The question is, does it work? I’ll be honest, I am not getting the local control that people have reported for pancreatic adenos with 50/5. Keep in mind, these mofos are super conformal and hence, very hot. Dose escalation was supposed to be game changing in this setting, but I’m not seeing it yet. Will a higher RBE to a portion of the target do any better? Maybe. I personally wouldn’t wager much money on it. Fortunately for you, I’m a terrible gambler.

 

[IonsAreOurFuture]

You make plausible arguments and I sincerely hope something positive comes from this. The reality is that most plausible arguments are wrong. I’m not trying to be a naysayer. I work with MRLinacs (as in, we have one so I use it…I didn’t set out to be an expert in this or make it a major focus of my research) and while I think most of what people are proposing to do with them is total BS, isotoxic adaptive dose escalation could actually offer some meaningful benefit in the right cases. It’s pretty much what you are proposing. Get as much dose to as much of the target as you can and forgo the conventional 95% coverage paradigm. It’s basically Brachy with EBRT. The question is, does it work? I’ll be honest, I am not getting the local control that people have reported for pancreatic adenos with 50/5. Keep in mind, these mofos are super conformal and hence, very hot. Dose escalation was supposed to be game changing in this setting, but I’m not seeing it yet. Will a higher RBE to a portion of the target do any better? Maybe. I personally wouldn’t wager much money on it. Fortunately for you, I’m a terrible gambler.

One of the nice things about LET-based optimization is that it is a relatively cheap advance - just a software upgrade, much like VMAT was to step and shoot IMRT - but even more impactful because Xray VMAT didn't improve the biological dose distribution, just the physical. I don't see an obvious downside if it means we can get rid of rib fractures and other LET related complications.

We can't design a phase III trial to validate every software release, but adoption does have to follow some common sense rules. For instance, TG-256 points out that we shouldn't risk de-escalating the dose to a GTV, so the nominal physical dose would still need to be standard prescription dose, and the LET contribution is a "bonus." The greater impact, I think, is not so much boosting the GTV but rather getting the high LET out of distal OARS where it can do known harm.

 

[IonsAreOurFuture]

You make plausible arguments and I sincerely hope something positive comes from this. The reality is that most plausible arguments are wrong. I’m not trying to be a naysayer. I work with MRLinacs (as in, we have one so I use it…I didn’t set out to be an expert in this or make it a major focus of my research) and while I think most of what people are proposing to do with them is total BS, isotoxic adaptive dose escalation could actually offer some meaningful benefit in the right cases. It’s pretty much what you are proposing. Get as much dose to as much of the target as you can and forgo the conventional 95% coverage paradigm. It’s basically Brachy with EBRT. The question is, does it work? I’ll be honest, I am not getting the local control that people have reported for pancreatic adenos with 50/5. Keep in mind, these mofos are super conformal and hence, very hot. Dose escalation was supposed to be game changing in this setting, but I’m not seeing it yet. Will a higher RBE to a portion of the target do any better? Maybe. I personally wouldn’t wager much money on it. Fortunately for you, I’m a terrible gambler.

I do think isotoxic dose escalation shows some promise, the FLAME trial is a good one. The days of escalating dose to the entire PTV (eg RTOG 0617) are probably over.

 

[Palex80]

Highest LET is on the "backside of the Bragg Peak". This does not mean the highest dose. This is a low dose area with a very high LET and anticipated increase in RBE and uncertainty. Presently (per my understanding as I'm not a user) protonists use intuition about where to let this area fall relative to adjacent OARs because it makes them squeamish (appropriately so).

What biologically happens behind the Bragg Peak somehow resembles what´s on the other side of a black hole, IMHO.

 

[NotMattSpraker]

We can't design a phase III trial to validate every software release, but adoption does have to follow some common sense rules. For instance, TG-256 points out that we shouldn't risk de-escalating the dose to a GTV, so the nominal physical dose would still need to be standard prescription dose, and the LET contribution is a "bonus." The greater impact, I think, is not so much boosting the GTV but rather getting the high LET out of distal OARS where it can do known harm.

Here is the MDACC trial. ClinicalTrials.gov

Very reasonable design for something that is hard to study.

LET op is a model of a model. I am not asking for a phase III trial, I just want to know how it works in a real human being in clinic.

 

[Palex80]

Here is the MDACC trial. ClinicalTrials.gov

Very reasonable design for something that is hard to study.

LET op is a model of a model. I am not asking for a phase III trial, I just want to know how it works in a real human being in clinic.

It makes one wonder, why would they start with ependymoma for that?
Looking at the trial design, it looks like all possible scenarios (GII-III, GTR or not).
Why not go for something more easy, like GBM?

 

[NotMattSpraker]

It makes one wonder, why would they start with ependymoma for that?
Looking at the trial design, it looks like all possible scenarios (GII-III, GTR or not).
Why not go for something more easy, like GBM?

Trials are super political. I did not really appreciate this until I was designing them and engaging these folks.

I'd guess GBM has a lot of available trials at MDACC. Institutions prioritize enrolling to these trials in different ways and the good ones have a "tree" that communicates to clinicians the priority for enrolling. Everywhere prioritizes industry because you make a lot of money by enrolling. A lot of places prioritize IITs because they want to their academics to succeed.

It could be that there just wasn't room for this and they had a lot of ependymoma patients going off trial.

The brain is a good choice to study this IMO. You need some kind of non-invasive biomarker of toxicity.

 

[drowsy12]

It makes one wonder, why would they start with ependymoma for that?
Looking at the trial design, it looks like all possible scenarios (GII-III, GTR or not).
Why not go for something more easy, like GBM?

agree with Not Matt and also makes sense to do it in a disease setting where proton is already SOC versus GBM where less time to understand the point.

 

[Palex80]

I understand the point, but if we want to study LET and potential excessive radionecrosis rates, I‘d rather risk those in a (doomed) 75 year old with a GBM, than in a 4-year old post–GTR with a G2 ependymoma.

 

[drowsy12]

I understand the point, but if we want to study LET and potential excessive radionecrosis rates, I‘d rather risk those in a (doomed) 75 year old with a GBM, than in a 4-year old post–GTR with a G2 ependymoma.

I guess I'm confused. if the standard of care is non-LET optimized IMPT for these patients, what would be the downside in trying LET-optimized?

 

[communitydoc13]

I guess I'm confused. if the standard of care is non-LET optimized IMPT for these patients, what would be the downside in trying LET-optimized?

This ISH is not precision work...although it is very expensive and nuanced...different things. Imagine if you took your dose engine regarding a 4 MeV electron plan seriously.

Say they optimize LET to minimize high LET areas in the brainstem, while achieving adequate "dose" to the target volume. This may change the LET distribution within the target volume, functionally changing effective dose within the target, or even require a different beam angle solution that increases lower doses to larger volumes of normal brain.

LET mapping in on the right in the below figure from A Systematic Review of LET-Guided Treatment Plan Optimisation in Proton Therapy: Identifying the Current State and Future Needs

 

[NotMattSpraker]

I understand the point, but if we want to study LET and potential excessive radionecrosis rates, I‘d rather risk those in a (doomed) 75 year old with a GBM, than in a 4-year old post–GTR with a G2 ependymoma.

Right, so maybe they let you open it. Sweet. Unfortunately, maybe your medical oncologist gets direct pharma patients to enroll to that other trial that is recruiting the same population. Good luck convincing him/her to let you enroll to yours instead. When IITs dont enroll, they close.

Your point is correct scientifically but that doesnt really matter in real life. Id argue it matters less than ever. This is why I prefer to be a spectator now.

I guess I'm confused. if the standard of care is non-LET optimized IMPT for these patients, what would be the downside in trying LET-optimized?

Good question. Is the standard even IMPT and not passive scatter? Do we know?

Im, like, very often confused by this field. Why do we use ezFluence and pretend that its not IMRT? Or not not IMRT?

Why do people do 6-fraction "SBRT" haha.

 

[IonsAreOurFuture]

Here is the MDACC trial. ClinicalTrials.gov

Very reasonable design for something that is hard to study.

LET op is a model of a model. I am not asking for a phase III trial, I just want to know how it works in a real human being in clinic.

Thanks for finding that, it's very interesting. I'm somewhat surprised they started in peds, what with the rarity of cases and long term risks if doing harm, but I suppose that also means that kids have the most to gain by avoiding toxicity.

While waiting for them to accrue 48 ependymoma patients willing to do a phase I trial, then gather 2 years of followup data, analyze and publish, I wonder what should the other 44 US centers do in the meantime. If it's a positive trial (i.e. worse than historical controls), we will definitely learn something. I think it has a high chance of being negative statistically speaking; brain necrosis rates are typically single digit percent events, so one might not see any difference in rare events, but if no difference is seen, would that mean one shouldn't bother with LET optimization?

I suppose the PCG (Proton Collaborative Group) Registry trial also exists for this sort of thing, for real-world comparison of different treatment techniques, which is sort of what MDACC is doing in a narrower disease-site specific way. I don't think MGH or MDACC are PCG members and they probably have enough volume to accrue from 2019 to 2027, but that really feels like a long time to wait before broader implementation.

It probably wouldn't take that long to accrue for a more common disease site. Rib fracture rates for breast, and mandible ORN for H&N might get more adequate numbers quicker. How many patients does it require to detect a 5% difference (eg 1% vs 6% event rate) in a toxicity outcome, with 80% power at an alpha=0.05? The Radcomp trial has about 600 proton breast patients and could serve as a historical control for rib fracture rates. Or is it "an" historical control?

Would that kind of an analysis be of interest to this group?

 

[IonsAreOurFuture]

Good question. Is the standard even IMPT and not passive scatter? Do we know?

The standard, or at least gold standard for now, is pencil beam scanning with a 5 mm wide beam (not 1 cm plus) and cone beam CT at isocenter, and treatment plans made using Monte Carlo. Any newer system from the past 7 years should be able to do that.

Everyone who can afford to is upgrading from passive scatter to pencil beam, even if it means taking their entire system down for months, like Harvard is doing and UPENN has announced they will be doing.

 

[NotMattSpraker]

While waiting for them to accrue 48 ependymoma patients willing to do a phase I trial, then gather 2 years of followup data, analyze and publish, I wonder what should the other 44 US centers do in the meantime.

How about literally anything other than treat prostate and breast off trial? Because that is what most proton centers are doing most of the time. I guess more a little more head and neck and esophagus now.

Im not even arguing that LET Op needs a randomized trial to "prove" its value. Im arguing that I would like to understand how it works, and it's effect on our usual clinical outcomes and biomarkers.

How many patients does it require to detect a 5% difference (eg 1% vs 6% event rate) in a toxicity outcome, with 80% power at an alpha=0.05?

422 I think?

BUT. You could compare it to historical controls for only 62 patients.

OR. You could just draw a picture of how much protons spares and as long as you don't draw the heart on the wrong side, no one will care. 0 patients.

 

[IonsAreOurFuture]

This ISH is not precision work...although it is very expensive and nuanced...different things. Imagine if you took your dose engine regarding a 4 MeV electron plan seriously.

Say they optimize LET to minimize high LET areas in the brainstem, while achieving adequate "dose" to the target volume. This may change the LET distribution within the target volume, functionally changing effective dose within the target, or even require a different beam angle solution that increases lower doses to larger volumes of normal brain.

LET mapping in on the right in the below figure from A Systematic Review of LET-Guided Treatment Plan Optimisation in Proton Therapy: Identifying the Current State and Future Needs

View attachment 390511

Thanks for the graphic, I like this one from a Physics World article:

Proton therapy planning: how to minimize LET in organs-at-risk – Physics World

Researchers identify guidelines for proton therapy planning that help avoid regions of high linear energy transfer outside of target volumes

physicsworld.com

It's probably important to keep in mind that high LET by itself isn't enough to cause injury, you need both high physical dose and high LET/RBE.

 

[Palex80]

I guess I'm confused. if the standard of care is non-LET optimized IMPT for these patients, what would be the downside in trying LET-optimized?

The overall uncertainties, I guess?

 

[TheWallnerus]

Thanks for the graphic, I like this one from a Physics World article:

Proton therapy planning: how to minimize LET in organs-at-risk – Physics World

Researchers identify guidelines for proton therapy planning that help avoid regions of high linear energy transfer outside of target volumes

physicsworld.com

It's probably important to keep in mind that high LET by itself isn't enough to cause injury, you need both high physical dose and high LET/RBE.

View attachment 390521

Allow me a dumb question

On first glance which one of these three do we prefer