Protons are blowing Rad Onc's boat out the CMS water

[Palex80]

That graph is ridiculous. Why do photons build up dose in air?

---

Members don't see this ad.

 

[RickyScott]

All this is true but even so at the time frames specified protons were supposed to be superior to photons. Or else what’s their raison d’etre.

Protons in any time frame are supposed to be superior to photons… I’ve never seen it where (again within a present timeframe of reference) people are like “you know these protons are very limited… we need to wait for some technical innovations before they’re superior to photons and we use them on people.”

It’s the Bragg peak, stupid. Or is it the Bragg peak plus PACS and image guidance.

View attachment 361455

Will never be able to compensate for higher bed at the edge of Bragg peak, in oars despite incessantly claiming “robustness” !

 

[madchemist89]

Proton therapy has a long way to go before it reaches the robustness and sophistication of photon treatments.

 

[RickyScott]

Proton therapy has a long way to go before it reaches the robustness and sophistication of photon treatments.

Because of the inherent degrees of freedom in let/bed- diff late and acute toxicities for every type of tissue/ and o2 level at every dose level (and these are just the variables I am aware of), I am not sure it is remotely modelable short of a laplace demon.

 

[thecarbonionangle]

Protonists will tell you arc therapy is the future. This is the new narrative!

 

[Ray D. Ayshun]

If you broke cancer into it's atomic components, it's mostly protons. Cancer loves protons the same way it loves sugar and air. Cancer hates photons.

 

[communitydoc13]

Loma Linda is the longest operating hospital proton center, 32 years now. If we had a linac (or cobalt machine) from 1990, how comfortable would we be practicing with it now? Would our teams even know how to run or QA it? Would the vendor still support it?

No cone beam, no KV/KV pair, no MLC, no IMRT, SBRT or 3DCRT, no reliable isocenter for radiosurgery, lots of analog dials, cerrobend blocks, 2D only plans, and the record and verify system was a trifold piece of paper filled out by the therapists daily (radiation oncologists were also called therapists not that long ago).

The wedge was an actual wedge of metal that was frequently put in backward or not at all by mistake. The simulator was not a CT, MRI or PETCT, but something called a Ximatron or Odelft and its Xray tube would often overheat while taking films (yes, on actual film) that had to be developed in an onsite dark room.

Contours and field shaping were done in less than 2 minutes with a wax pencil on a light box (I sort of miss that part). The "body contour" was acquired with a piece of wire that you bent to sort of match the patient's skin. The CT of the tumor, if you were lucky enough to have one, was from a tiny 2x2 inch image on an actual piece of film from the radiology dept across the street, that is IF you could find the patient's film jacket. No PACS, no internet, few computers, tiny CRT screens - a 15 inch color monitor was a luxury.

Dose calcs were often hand calcs prescribed to Dmax, d 1/2, or 3 or 5 cm deep. No 3-D dose distribution, maybe "2.5-D" and the patient was assumed to be a uniform block of water, no heterogeneity correction for lung tissue or other cavities. Weekly patient management actually required a lot of medication and coaching to stop patients from quitting due to toxicity. Sometimes I'm surprised that we even survived as a field.

Is it important to inquire about the corresponding technical advances in proton therapy since then?

It’s the Bragg peak, stupid. Or is it the Bragg peak plus PACS and image guidance.

Of course it's important to inquire, but the more I dig, the more I believe that that 1D representation of proton/photon dosimetry matters so much less now than it did in 1990, and that the intrinsic dosimetric uncertainties of protons (or any directly ionizing radiation) mean that those remarkable advances in photon treatment (outside of image guidance) are likely not reproducible for ion based treatments.

Photons are so much better now because our high confidence in real dosimetry lets us run through millions of perturbations with regard to treatment planning. And, we remain confident even when our solution means that the machine is doing some sort of absurd MLC dance.

Put your cell culture plates across photon beams of varying energies (the types we use in clinic) and I doubt much magic will happen at certain energy thresholds, along the axis of beam or perpendicular to the axis of the beam.

Do the same with protons?

Procure gives out the 1D diagram in its brochure. Per discussion here, many docs will look at proton plans as though they are photon plans, believing that the dose is real. Imagine what it will be like when we have 200 proton clinics nationally?

 

[IonsAreOurFuture]

You are correct, right now you cannot trust the dose distribution on the proton screen, which is why we see rib fractures after 50 Gy that is more like 70+ RBE. This is an impediment to wide scale adoption. Fortunately Raystation now offers LET based optimization, but that doesn't translate easily to RBE for clinicians.

Range uncertainty adoption is all over the map, literally, as most passive scatter places - mainly opened prior to 2010 - use a margin of +/- 3.5% + 3 mm in addition to their setup margins, which can become quite big. Newer pencil beam centers running Monte Carlo calcs use more like 3% without a fixed mm add-on, which is still not as tight as I'd like it to be.

Passive scatter protons with a brass block and plastic range compensator are the proton equivalent of 3D conformal, in that you cannot make a concave dose distribution or treat very complex shapes.

Even among the centers that do have pencil beam, spot size is usually 1 cm or so in the older facilities beyond 5 years old or so. Newer ones are more like 5 mm, which is like having MLC leaves of similar dimensions - sometimes it matters, sometimes not.

Proton margins tend to be wider than Xray margins in my experience, perhaps due to the lack of cone beam on all but the newer decade of machines, and even then not all. A lot of it though is tradition and user confidence.

I think that doctors and patients are becoming more aware of some of these limitations, but it is definitely true that all proton centers are not created equal. Buyer beware for the time being, unless you or a friend really know what you're doing.

 

[grenz]

You are correct, right now you cannot trust the dose distribution on the proton screen, which is why we see rib fractures after 50 Gy that is more like 70+ RBE. This is an impediment to wide scale adoption. Fortunately Raystation now offers LET based optimization, but that doesn't translate easily to RBE for clinicians.

Range uncertainty adoption is all over the map, literally, as most passive scatter places - mainly opened prior to 2010 - use a margin of +/- 3.5% + 3 mm in addition to their setup margins, which can become quite big. Newer pencil beam centers running Monte Carlo calcs use more like 3% without a fixed mm add-on, which is still not as tight as I'd like it to be.

Passive scatter protons with a brass block and plastic range compensator are the proton equivalent of 3D conformal, in that you cannot make a concave dose distribution or treat very complex shapes.

Even among the centers that do have pencil beam, spot size is usually 1 cm or so in the older facilities beyond 5 years old or so. Newer ones are more like 5 mm, which is like having MLC leaves of similar dimensions - sometimes it matters, sometimes not.

Proton margins tend to be wider than Xray margins in my experience, perhaps due to the lack of cone beam on all but the newer decade of machines, and even then not all. A lot of it though is tradition and user confidence.

I think that doctors and patients are becoming more aware of some of these limitations, but it is definitely true that all proton centers are not created equal. Buyer beware for the time being, unless you or a friend really know what you're doing.

Ok so let’s play a fun game. Which proton centers are doing it the “right” way and which are not?

 

[deleted1111261]

Ok so let’s play a fun game. Which proton centers are doing it the “right” way and which are not?

Obviously the ones that advertise the most !!

Or send letters to us on Doximity to beg for patients.

 

[Doctorer]

Ok so let’s play a fun game. Which proton centers are doing it the “right” way and which are not?

Very few, and there’s debate on what the “right way” is, and what disease sites/situations it matters for.

Source: a misanthrope who occasionally posts on sdn, definitely not a radiation oncologist.

Edit: fyi the physicists at each proton site will justify the way they're treating and call all other physicists wrong.

 

[thesauce]

Ok so let’s play a fun game. Which proton centers are doing it the “right” way and which are not?

Loma Linda is doing it wrong

 

[RickyScott]

But even if you “do it right” - pencil beam and balloon/space oar, a number of pts require rectal coagulation. Maybe not matt spraker can comment? (Kudos to honestly reporting data as I am sure some centers would bury this) I would disagree with their conclusion and counting laser coagulation as G2.

Analysis of Gastrointestinal Toxicity in Patients Receiving Proton Beam Therapy for Prostate Cancer: A Single-Institution Experience - PubMed

PBT was associated with acceptable rates of GR2+ transient GI toxicity, mostly in the form of RB, which correlated with anticoagulation use. High EPIC bowel domain quality of life was maintained in the 2 years after treatment.

pubmed.ncbi.nlm.nih.gov

 

[thecarbonionangle]

I would not get irradiated on a Mevion even if you payed me.

 

[Palex80]

I would not get irradiated on a Mevion even if you payed me.

What's so bad about it? (sorry for my proton-newbiness)

EDIT:

 

[thecarbonionangle]

What's so bad about it? (sorry for my proton-newbiness)

EDIT:
View attachment 361538

It is just a cheaply made machine, 100 percent chinese owned, yes. The company has declared bankruptcy multiple times. You get what you pay for, like getting your proton system at walmart. They don’t have the best energy selection system. Their bragg peak isn’t as sharp. Make rad onc great again!

 

[RickyScott]

It is just a cheaply made machine, 100 percent chinese owned, yes. The company has declared bankruptcy multiple times. You get what you pay for, like getting your proton system at walmart. They don’t have the best energy selection system. Their bragg peak isn’t as sharp. Make rad onc great again!

Do Mevions treat one beam per day?

 

[deleted1111261]

I shop at Walmart for a lot of stuffs!

 

[evilbooyaa]

All this is true but even so at the time frames specified protons were supposed to be superior to photons. Or else what’s their raison d’etre.

Protons in any time frame are supposed to be superior to photons… I’ve never seen it where (again within a present timeframe of reference) people are like “you know these protons are very limited… we need to wait for some technical innovations before they’re superior to photons and we use them on people.”

It’s the Bragg peak, stupid. Or is it the Bragg peak plus PACS and image guidance.

View attachment 361455

Ah yes, put the proton beam directly into the hippocampus so as to effectively sterilize everyone's memories that protons have had decades to show a clinical toxicity benefit and to date have failed to do so at everything beyond "but look at the computer screen!!111"

 

[IonsAreOurFuture]

All this is true but even so at the time frames specified protons were supposed to be superior to photons. Or else what’s their raison d’etre.

Protons in any time frame are supposed to be superior to photons… I’ve never seen it where (again within a present timeframe of reference) people are like “you know these protons are very limited… we need to wait for some technical innovations before they’re superior to photons and we use them on people.”

It’s the Bragg peak, stupid. Or is it the Bragg peak plus PACS and image guidance.

View attachment 361455

I think the proton and photon technological progress is a little bit like the tortoise and the hare race. No consistent leader for the entire time, but fits and starts.

In 1990, when prostate cancer was treated with 2D conformal RT, and the 60-70% isodose going through the full rectal width and bladder, the ability to treat with lateral-only beams was a big advance. I think X-rays pulled ahead during the IMRT era, and especially with SBRT, when on-board CBCT really opened up the possibilities for tight margins and high doses. Protons missed out due to lack of 3D image guidance in that era, and it showed up in the negative MDACC trial of 3D passive scatter protons vs IMRT with CBCT-guidance for lung CA.

Now that CBCT is standard on both, and daily adaptive planning is becoming in vogue, a lot of the concerns about proton uncertainty (range, positioning, calculation) get reduced, especially with a vertical dual-energy CT scanner that is part of the upright chair gantry, you can in theory sim, plan and treat a single shot or 5 shot SBRT with a proton "plan of the day." Auto-contouring and Raystation planning are so quick now, most of the time is me drawing my volumes and physicists debating whether the plan could be even better.

I think it's safe to say that both are here to stay at this point, and that we'll be seeing more proton systems going into smaller and smaller communities and centers, mostly as single room systems and not the big 4 room gambles that were built in the early 2000's (unless you are MDACC or Mayo, etc).

 

[communitydoc13]

Now that CBCT is standard on both, and daily adaptive planning is becoming in vogue, a lot of the concerns about proton uncertainty (range, positioning, calculation) get reduced, especially with a vertical dual-energy CT scanner that is part of the upright chair gantry, you can in theory sim, plan and treat a single shot or 5 shot SBRT with a proton "plan of the day." Auto-contouring and Raystation planning are so quick now, most of the time is me drawing my volumes and physicists debating whether the plan could be even better.

Yeah, that really sounds like a community clinic work flow. Only feasible when you are collecting bank per shot and have massive resources. All for what level of value?

Brutal.

Meanwhile, I need physicists and medoncs now to keep my fairly healthy community practice running.

 

[evilbooyaa]

I think the proton and photon technological progress is a little bit like the tortoise and the hare race. No consistent leader for the entire time, but fits and starts.

In 1990, when prostate cancer was treated with 2D conformal RT, and the 60-70% isodose going through the full rectal width and bladder, the ability to treat with lateral-only beams was a big advance. I think X-rays pulled ahead during the IMRT era, and especially with SBRT, when on-board CBCT really opened up the possibilities for tight margins and high doses. Protons missed out due to lack of 3D image guidance in that era, and it showed up in the negative MDACC trial of 3D passive scatter protons vs IMRT with CBCT-guidance for lung CA.

Now that CBCT is standard on both, and daily adaptive planning is becoming in vogue, a lot of the concerns about proton uncertainty (range, positioning, calculation) get reduced, especially with a vertical dual-energy CT scanner that is part of the upright chair gantry, you can in theory sim, plan and treat a single shot or 5 shot SBRT with a proton "plan of the day." Auto-contouring and Raystation planning are so quick now, most of the time is me drawing my volumes and physicists debating whether the plan could be even better.

I think it's safe to say that both are here to stay at this point, and that we'll be seeing more proton systems going into smaller and smaller communities and centers, mostly as single room systems and not the big 4 room gambles that were built in the early 2000's (unless you are MDACC or Mayo, etc).

CBCT is standard on *NEW* proton machines. Can the proton apologists convince all those folks with *OLD* machines to stop using any machine that doesn't have CBCT? How about stopping the use of *OLD* machines that use passive scatter?

What would be the presumed CLINICAL advantage of 5-fraction proton SBRT done adaptively compared to 5-fraction photon SBRT done adaptively?

It's just a self-fulfilling prophecy. The flowsheet previously posted is exactly correct in the mind of all proton apologists. Here it is again (Credit @Gfunk6 ):

 

[temujim]

CBCT is standard on *NEW* proton machines. Can the proton apologists convince all those folks with *OLD* machines to stop using any machine that doesn't have CBCT? How about stopping the use of *OLD* machines that use passive scatter?

What would be the presumed CLINICAL advantage of 5-fraction proton SBRT done adaptively compared to 5-fraction photon SBRT done adaptively?

It's just a self-fulfilling prophecy. The flowsheet previously posted is exactly correct in the mind of all proton apologists. Here it is again (Credit @Gfunk6 ):

View attachment 361648

Did locums at a place in the mid 2010s where the linac did not have MLCs. People will do whatever they want with their sunk costs.

 

[Palex80]

Did locums at a place in the mid 2010s where the linac did not have MLCs. People will do whatever they want with their sunk costs.

All the RTTs in that place

 

[thecarbonionangle]

CBCT is standard on *NEW* proton machines. Can the proton apologists convince all those folks with *OLD* machines to stop using any machine that doesn't have CBCT? How about stopping the use of *OLD* machines that use passive scatter?

What would be the presumed CLINICAL advantage of 5-fraction proton SBRT done adaptively compared to 5-fraction photon SBRT done adaptively?

It's just a self-fulfilling prophecy. The flowsheet previously posted is exactly correct in the mind of all proton apologists. Here it is again (Credit @Gfunk6 ):

View attachment 361648

I heard Ramesh Rengan say they don’t even have CBCT not so long ago. This is supposedly a “modern” proton center.

 

[evilbooyaa]

Did locums at a place in the mid 2010s where the linac did not have MLCs. People will do whatever they want with their sunk costs.

Sure, I interviewed at a job when I was graduating residency that had a photon machine without CBCT or VMAT capabilities. But I, as a photon enthusiast, was willing to admit that that machine was garbage and that they should probably not do any SBRT that is not the lung and requires CBCT for visualization (unless fiducials) or something.

 

[seper]

Did locums at a place in the mid 2010s where the linac did not have MLCs. People will do whatever they want with their sunk costs.

compensator-based IMRT. Better than MLC from that era

 

[RickyScott]

 

[Fpg1245]

Who cares? You’re gonna use them anyway.

 

[BobbyHeenan]

What is he talking about "why haven't we seen this earlier?"

We have seen it . Of course they're just signals with weak-ish data...but possible signals have been seen-
brainstem changes in peds posterior fossa
rib fractures in breast
skin toxicity in one proton abpi trial
increased reconstruction/implant complications in reconstruction breast
SEER data rectal bleeds in prostate

 

[TheWallnerus]

What is he talking about "why haven't we seen this earlier?"

We have seen it . Of course they're just signals with weak-ish data...but possible signals have been seen-
brainstem changes in peds posterior fossa
rib fractures in breast
skin toxicity in one proton abpi trial
increased reconstruction/implant complications in reconstruction breast
SEER data rectal bleeds in prostate

Would add esophagitis from proton breast data too; some of the percentages are ridiculously high

 

[NotMattSpraker]

This is a great example of why I just cant with "proton" people and their scientific virtue signaling.

Here is an open access narrative of how the MDACC HN trial was designed then re-designed. You can form your own opinion about whether this is the way we should do technology development trials. Comparing Intensity-Modulated Proton Therapy With Intensity-Modulated Photon Therapy for Oropharyngeal Cancer: The Journey From Clinical Trial Concept to Activation - PubMed

 

[BobbyHeenan]

This is a great example of why I just cant with "proton" people and their scientific virtue signaling.

Here is an open access narrative of how the MDACC HN trial was designed then re-designed. You can form your own opinion about whether this is the way we should do technology development trials. Comparing Intensity-Modulated Proton Therapy With Intensity-Modulated Photon Therapy for Oropharyngeal Cancer: The Journey From Clinical Trial Concept to Activation - PubMed

What is the point of a non-inferiority trial of a therapy that is way more expensive? What am I missing here?

I didn't know this about the ongoing trial.

 

[RickyScott]

This is a great example of why I just cant with "proton" people and their scientific virtue signaling.

Here is an open access narrative of how the MDACC HN trial was designed then re-designed. You can form your own opinion about whether this is the way we should do technology development trials. Comparing Intensity-Modulated Proton Therapy With Intensity-Modulated Photon Therapy for Oropharyngeal Cancer: The Journey From Clinical Trial Concept to Activation - PubMed

Russia is very envious of some of the propaganda coming out re protons.

 

[NotMattSpraker]

What is the point of a non-inferiority trial of a therapy that is way more expensive? What am I missing here?

I didn't know this about the ongoing trial.

That's a good question. Ask the NRG influencers. And you know I feel compelled to point out that this isn't personal or the PI's "fault" because everyone takes everything so personally.

"The major point of contention was use of an endpoint based on the CTCAE scale for the phase III portion of the trial, because of a perceived lack of objectivity (owing to its dependence on physician reporting) and insufficient sensitivity to account for the numerous forms of toxicity experienced by a patient, which could thereby dilute potential differences between the two treatments."

I just ughhh... this is our system for better or worse.

 

[BobbyHeenan]

That's a good question. Ask the NRG influencers. And you know I feel compelled to point out that this isn't personal or the PI's "fault" because everyone takes everything so personally.

"The major point of contention was use of an endpoint based on the CTCAE scale for the phase III portion of the trial, because of a perceived lack of objectivity (owing to its dependence on physician reporting) and insufficient sensitivity to account for the numerous forms of toxicity experienced by a patient, which could thereby dilute potential differences between the two treatments."

I just ughhh... this is our system for better or worse.

I agree toxicity will be challenging given patients and docs won't be blinded and there is so much built in subjectivity.

I'm curious to see a harder endpoint like feeding tube dependence at 6-12 months, % weight change, narcotic usage, and global QoL scores though even that can be manipulated.

 

[TheWallnerus]

I agree toxicity will be challenging given patients and docs won't be blinded and there is so much built in subjectivity.

I'm curious to see a harder endpoint like feeding tube dependence at 6-12 months, % weight change, narcotic usage, and global QoL scores though even that can be manipulated.

They mentioned feed tube dependence at 12 months but with low incidence need very big sample sizes; so, infeasible.

 

[BobbyHeenan]

They mentioned feed tube dependence at 12 months but with low incidence need very big sample sizes; so, infeasible.

it's gonna come down to ALARA isn't it?

No differences anywhere but the dvh's look better.

 

[NotMattSpraker]

I agree toxicity will be challenging given patients and docs won't be blinded and there is so much built in subjectivity.

I'm curious to see a harder endpoint like feeding tube dependence at 6-12 months, % weight change, narcotic usage, and global QoL scores though even that can be manipulated.

Right, it is hard, but thats only half the argument. The hypothesis is that toxicity is improved and they are sitting on data that PFS is the same, and maybe PROs are slightly improved with protons (MDADI difference of 1-2). To turn around and make the primary end point non-inferiority PFS is scientifically disingenuous. At least just say the truth, you're worried that it might be negative.

I strongly disagree this is better. A negative trial with noisy data that is hard to interpret would be more beneficial for our field than the current design.

As an aside, it's funny how a faculty from the same institution managed to have a toxicity end point in esophagus and it made it through to the NRG phase III. They even provided an innovative approach to interpreting toxicity data (although not used in phase III). I guess it's not insurmountable for some.

 

[communitydoc13]

As an aside, it's funny how a faculty from the same institution managed to have a toxicity end point in esophagus and it made it through to the NRG phase III. They even provided an innovative approach to interpreting toxicity data (although not used in phase III). I guess it's not insurmountable for some.

When you have to look this hard for positive signal, the signal is not strong (if real at all).

My concern is a "too big to fail phenomenon". There has been so much investment that trials will continue to be run until some marginal positive signal is seen and in turn this will be spun aggressively. I believe we have already seen this type of behavior on the part of the proton people.

Under what circumstance do we, as a field, largely abandon the proton experiment? If the answer is no circumstance, then you have a too big to fail phenomenon.

Four national centers for niche indications never bothered me.

 

[medgator]

.

Under what circumstance do we, as a field, largely abandon the proton experiment? If the answer is no circumstance, then you have a too big to fail phenomenon.

 

[RickyScott]

Right, it is hard, but thats only half the argument. The hypothesis is that toxicity is improved and they are sitting on data that PFS is the same, and maybe PROs are slightly improved with protons (MDADI difference of 1-2). To turn around and make the primary end point non-inferiority PFS is scientifically disingenuous. At least just say the truth, you're worried that it might be negative.

I strongly disagree this is better. A negative trial with noisy data that is hard to interpret would be more beneficial for our field than the current design.

As an aside, it's funny how a faculty from the same institution managed to have a toxicity end point in esophagus and it made it through to the NRG phase III. They even provided an innovative approach to interpreting toxicity data (although not used in phase III). I guess it's not insurmountable for some.

i dont understand. How do they claim success if non inferior pfs is met?

 

[BobbyHeenan]

i dont understand. How do they claim success if non inferior pfs is met?

Med oncs and ENTS will read the headline:

Phase 3 Trial shows proton therapy a standard of care option for oropharynx treatment.

Just gotta get them in the door and show them those dose clouds.

 

[NotMattSpraker]

Under what circumstance do we, as a field, largely abandon the proton experiment? If the answer is no circumstance, then you have a too big to fail phenomenon.

This is where I always chuckle. So many convos slide toward a point where any rational adult would say "well, it's a business". But in medicine you just can't say that without sounding horrible, so say lots of weird things instead.

What is the experiment? That it will help some patients or that it will help centers turn a nice profit?

This is rhetorical but sometimes I look at European proton trials and US proton trials and I think about that question.

i dont understand. How do they claim success if non inferior pfs is met?

You ever see an episode of Dr. Oz's show?

 

[Mandelin Rain]

NMS is putting in some of my favorite work right now.

 

[fiji128]

Even Ralph calling it out.

 

[medgator]

Even Ralph calling it out.View attachment 362365

All hail big rad onc, protons 4 lyfe

 

[BobbyHeenan]

Somebody REALLY needs to fact check that paper /notion from the MDA doc about cost of 5 fraction proton.

Post the medicare numbers for 5 fraction IMRT vs. 5 fraction proton. Let's just take a look. I think there's some "fuzzy math" in that paper and they're hiding behind "peer review."

 

[fiji128]

Somebody REALLY needs to fact check that paper /notion from the MDA doc about cost of 5 fraction proton.

Post the medicare numbers for 5 fraction IMRT vs. 5 fraction proton. Let's just take a look. I think there's some "fuzzy math" in that paper and they're hiding behind "peer review."

 

[BobbyHeenan]

I want to see actual numbers though.
a partial arc VMAT (or 4-6 field IMRT) versus protons.
5 fractions.
If someone gives me the proton cpt codes I can get it myself...

License for Use of Current Procedural Terminology, Fourth Edition ("CPT") | CMS

www.cms.gov