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

[Tsigolonco noitaidar]

---

Members do not see this ad.

Imrt vs 3d is photons to photons. Some parts of the proton dose distribution can have BEDs up to 1.7+ (Effectively biological hot spots of 150%) near the Bragg peak, hence worse rectal toxicity and brain necrosis of kiddies at under 60gy.

Your analogy would have to include imrt that delivered 150% hot spots next to critical organs at risk!

So you're basically saying that we don't need RCTs for IMRT vs 3D because it's the same particle as 3D and because hotspots are not that high? And if so, how are you justifying the cost of IMRT?

 

[TheWallnerus]

Here is my question: if someone was to tell me that there is a more expensive radiation treatment modality that has limited randomized evidence to support it, I would think of IMRT. So I don't understand how folks so easily adopted IMRT when there was little to no randomized evidence comparing 3D to IMRT (currently, there are a few studies, but not many)? Is it because there is a lot of RWD to support IMRT? Are the existing RCTs sufficient enough to support widespread use of IMRT at every radiation oncology center in the country? Surely 3DCRT could be done at a much cheaper cost, so should we be sticking with 3D?

IMRT satisfied clinicians' OODA loop tests when it first saw use in clinic. People saw prostate patients have less rectal side effects. People saw head/neck plans which were essentially impossible with 3DCRT become possible with IMRT. People saw less side effects in breast patients. IMRT really revolutionized anal ca tx efficiency AND clinical outcomes. TPSs showed lower lung V20s w/IMRT, sometimes dramatic diffs. And planning just in general became quicker and easier in complex cases. On top of all that, the barrier of entry for IMRT in most clinics was low: most times just a software upgrade in the early days. All these were shared experiences. People could get together and agree on a shared reality. Later, IMRT trials appeared. There are several randomized IMRT trials for breast eg, and breast is the #1 indication for RT. AFAIK there is not falsifying data on IMRT; i.e. that IMRT is clinically worse than 3D. There is falsifying data for protons in some sites.

So you're basically saying that we don't need RCTs for IMRT vs 3D because it's the same particle as 3D and because hotspots are not that high? And if so, how are you justifying the cost of IMRT?

IMRT often looks better than 3D in silico. And protons often look better than IMRT in silico. You can justify the cost of IMRT if payors will pay. Proton costs, on both the reimbursement and cost-of-equipment side, are so high that justification can be difficult. I am getting reimbursed less for 15 fraction breast IMRT than 33 fraction 3D breast. Cost and justification arguments are complex. (If proton reimbursement equaled IMRT reimbursement, proton costs would not be justified!)

More important... most important... are protons clinically superior radiotherapy? If so, how? And in what sites?

 

[metallica81788]

This the classical radiation oncology problem when people start going down the "we need randomized evidence to show us that x is better" for absolutely any/everything - you rationalize yourself out of existence.

Excuse me while I go design a 2x2 trial on if I need breathe oxygen to breathe. Although I won't change my practice until at least the 30 year update.

 

[thecarbonionangle]

RTOG 1308 was a missed oppportunity. If goal was to eval isotoxic dose escalation, this could have been an arm still but they should have kept the standard arm from 0617. My frustration with all the resources going into these studies is that they miss the mark on plain obvious things almost like the GOG was designing all proton studies. Look at the margins in 1308, they are on the bigger side, which might negate some benefit, they were clearly worried about missing a moving target.

 

[RickyScott]

So you're basically saying that we don't need RCTs for IMRT vs 3D because it's the same particle as 3D and because hotspots are not that high? And if so, how are you justifying the cost of IMRT?

Cost is based on where you are treated and has less relation to IMRT vs 3D. short course 3D at MDACC/MSKCC/UPENN is more expensive than IMRT in the community. Its not hotpots- you dont know what the hell is going on with protons and range uncertainy

 

[TheWallnerus]

This the classical radiation oncology problem when people start going down the "we need randomized evidence to show us that x is better" for absolutely any/everything - you rationalize yourself out of existence.

Excuse me while I go design a 2x2 trial on if I need breathe oxygen to breathe. Although I won't change my practice until at least the 30 year update.

True but... If I had the money for a proton machine, I wouldn't buy it. Without randomized evidence, I'd be too afraid that my investment might go kaput and I'd spend myself out of existence! With randomized evidence, even just one or two studies, you could better get insurances to reimburse at levels commensurate with the high purchase price and get Medicare to believe in the reimbursement too; which right now, they do not believe.

 

[thecarbonionangle]

True but... If I had the money for a proton machine, I wouldn't buy it. Without randomized evidence, I'd be too afraid that my investment might go kaput and I'd spend myself out of existence! With randomized evidence, even just one or two studies, you could better get insurances to reimburse at levels commensurate with the high purchase price and get Medicare to believe in the reimbursement too; which right now, they do not believe.

The problem is the same companies who demand evidence refuse to pay for patients to be on a RCT to evaluate it. So it becomes a catch 22: where is the evidence? Oh you trying to get that for us? Well we won’t pay as this is not “medically necessary”. It totally stagnates any progress. This is why medicare patients are the meat of these trials and they are older, sicker, which might negate any benefit due to selection bias

 

[communitydoc13]

And if so, how are you justifying the cost of IMRT?

The cost of IMRT has been continuously coming down, and under APM, IMRT will not cost more than other planning strategies for Medicare patients. Regarding dosimetry time, IMRT cost has not been justified for some time (actually easier for many newer dosimetrists). Some added real cost to IMRT QA. Cost of added processing power in today's era is trivial.

IMRT has certainly potentiated hypofractionation (a double edged sword) and in that regard is cost effective.

Under APM, I will probably do more IMRT in settings where I didn't before because of concern for cost.

Agree, you don't need a randomized trial for every incremental move forward. If only protons were only marketed as such.

This is why medicare patients are the meat of these trials and they are older, sicker, which might negate any benefit due to selection bias

These are almost all of my patients. Older, sicker, unlikely to benefit. This is most of cancer care.

 

[Tsigolonco noitaidar]

IMRT satisfied clinicians' OODA loop tests when it first saw use in clinic. People saw prostate patients have less rectal side effects. People saw head/neck plans which were essentially impossible with 3DCRT become possible with IMRT. People saw less side effects in breast patients. TPSs showed lower lung V20s w/IMRT, sometimes dramatic diffs.

So basically anecdotal observations ("people saw") and DVH improvements support the use of IMRT over 3D. Good to know.

There are several randomized IMRT trials for breast eg, and breast is the #1 indication for RT. AFAIK there is not falsifying data on IMRT; i.e. that IMRT is clinically worse than 3D. There is falsifying data for protons in some sites.

Based on this data, are you treating all your breast cases with IMRT?

More important... most important... are protons clinically superior radiotherapy? If so, how? And in what sites?

I will answer this by quoting you--just replace "3DCRT" with "IMRT" and "IMRT" with "IMPT": "People saw prostate patients have less rectal side effects. People saw head/neck plans which were essentially impossible with 3DCRT become possible with IMRT. People saw less side effects in breast patients. TPSs showed lower lung V20s w/IMRT, sometimes dramatic diffs."

 

[TheWallnerus]

Based on this data, are you treating all your breast cases with IMRT?

Yes. There are like 3-to-1 more breast IMRT trials than any other body site. The randomized evidence for IMRT is strongest in breast.

So basically anecdotal observations ("people saw") and DVH improvements support the use of IMRT over 3D. Good to know.

I will answer this by quoting you--just replace "3DCRT" with "IMRT" and "IMRT" with "IMPT": "People saw prostate patients have less rectal side effects. People saw head/neck plans which were essentially impossible with 3DCRT become possible with IMRT. People saw less side effects in breast patients. TPSs showed lower lung V20s w/IMRT, sometimes dramatic diffs."

I feel like you're being facetious lol. I will assume you're a proton user. I have never shepherded a single patient through proton treatment. Much less a breast proton patient. What kind of less side effects do you see in clinic with proton breast radiotherapy. Less skin effects? If so, why has it been so difficult to trial this in protons versus the IMRT users' ability to do so?

 

[medgator]

So basically anecdotal observations ("people saw") and DVH improvements support the use of IMRT over 3D. Good to know.


Based on this data, are you treating all your breast cases with IMRT?


I will answer this by quoting you--just replace "3DCRT" with "IMRT" and "IMRT" with "IMPT": "People saw prostate patients have less rectal side effects. People saw head/neck plans which were essentially impossible with 3DCRT become possible with IMRT. People saw less side effects in breast patients. TPSs showed lower lung V20s w/IMRT, sometimes dramatic diffs."

Clinically we have not seen the same bang for the buck with protons vs imrt esp when you consider the cost differential. The onus is on the proton folks to show that benefit considering they are the ones going out to buy the machines without any data to back it up

 

[thecarbonionangle]

The benefit in breast is not skin toxicity. Skin toxicities may actually be higher, higher entrance dose for e and protons. Believers would argue potential benefits in RNI setting. Data will have to show something.

 

[TheWallnerus]

The problem is the same companies who demand evidence refuse to pay for patients to be on a RCT to evaluate it. So it becomes a catch 22: where is the evidence? Oh you trying to get that for us? Well we won’t pay as this is not “medically necessary”. It totally stagnates any progress. This is why medicare patients are the meat of these trials and they are older, sicker, which might negate any benefit due to selection bias

This is a good point. But how much of a Catch-22 versus just some shortsighted thinking. Places have $100 million sitting around to buy a proton machine, but don't have the money to provide free or reduced price RT to a few hundred patients over ~5 years to gather data to prove protons-are-good theory. Or even better, spread the cost of such an endeavor over multiple proton sites to make the financials less onerous. Breast is a perfect example. Thousands and thousands of patients diagnosed each year. In my state, the main insurance company pays IMRT rates for proton patients when proton tx is denied. That should be enough money to run a breast trial. And why hasn't the NCI helped fund any of this. Has no one written up a compelling fundable proton proposal?

 

[TheWallnerus]

The benefit in breast is not skin toxicity. Skin toxicities may actually be higher, higher entrance dose for e and protons. Believers would argue potential benefits in RNI setting. Data will have to show something.

So this is a point I was trying to make earlier "People saw less side effects with IMRT," and @Tsigolonco noitaidar was saying, if I am not mistaken, that he/she sees less side effects with protons for breast. So anecdotes, personal experience, in silico data, retrospective data, and prospective randomized data support IMRT for breast. What sort of less side effects are seen by docs in the clinic with protons for breast. I am interested to know given that breast patients are the #1 proton patient nationally.

 

[thecarbonionangle]

This is a good point. But how much of a Catch-22 versus just some shortsighted thinking. Places have $100 million sitting around to buy a proton machine, but don't have the money to provide free or reduced price RT to a few hundred patients over ~5 years to gather data to prove protons-are-good theory. Or even better, spread the cost of such an endeavor over multiple proton sites to make the financials less onerous. Breast is a perfect example. Thousands and thousands of patients diagnosed each year. In my state, the main insurance company pays IMRT rates for proton patients when proton tx is denied. That should be enough money to run a breast trial. And why hasn't the NCI helped fund any of this. Has no one written up a compelling fundable proton proposal?

I agree that a compromise would have been to take IMRT rates but many hospital systems refuse to do this arguing that it costs more money to run a proton centre. It takes more physics, more dosimetry, more computing power, a service contract with engineers on site. I have seen argued that it is not viable at IMRT rates but i am not entirely sure this is true. Maybe someone privy to proton centre economics might comment

 

[RickyScott]

The benefit in breast is not skin toxicity. Skin toxicities may actually be higher, higher entrance dose for e and protons. Believers would argue potential benefits in RNI setting. Data will have to show something.

Protons cause rib fractures which are unheard of with photons in breast

 

[TheWallnerus]

I agree that a compromise would have been to take IMRT rates but many hospital systems refuse to do this arguing that it costs more money to run a proton centre. It takes more physics, more dosimetry, more computing power, a service contract with engineers on site. I have seen argued that it is not viable at IMRT rates but i am not entirely sure this is true. Maybe someone privy to proton centre economics might comment

Proton centers' lack of convincing data has been their Achilles heel IMHO. Granted I am a total protonic booboisie, but it would seem to me that proton users' first priority wouldn't be sweating payment rates but instead getting some damned data.

 

[thecarbonionangle]

Protons cause rib fractures which are unheard of with photons in breast

I believe you have seen this but hard for anybody to comment without looking at the plan. Distal bragg peak uncertainty is a real thing but not surprising. A physicist at ASTRO candidly told me once we basically have no idea what the BED is at that distal end.

 

[BobbyHeenan]

I treat 30/5 photon IMRT APBI. It’s magical. I find it really hard to believe unless completely cost neutral that there is meaningful proton benefit for APBI.

Proton APBI goes on out there I assure you.

 

[RickyScott]

I treat 30/5 photon IMRT APBI. It’s magical. I find it really hard to believe unless completely cost neutral that there is meaningful proton benefit for APBI.

Proton APBI goes on out there I assure you.

Proton Apbi is absurd. (Unless you are Ben smith trying to bilk employmers out of tens of millions) Brachy is as well, but would think savi would still be better than protons.

 

[BobbyHeenan]

Apbi is absurd. (Unless you are Ben smith trying to bilk employments out of tens of millions) brachy is as well, but would think savi would still be better than protons.

I do that too, though fell in love with 30/5. The catheters can be cumbersome at times.

 

[ramsesthenice]

This is a good point. But how much of a Catch-22 versus just some shortsighted thinking. Places have $100 million sitting around to buy a proton machine, but don't have the money to provide free or reduced price RT to a few hundred patients over ~5 years to gather data to prove protons-are-good theory. Or even better, spread the cost of such an endeavor over multiple proton sites to make the financials less onerous. Breast is a perfect example. Thousands and thousands of patients diagnosed each year. In my state, the main insurance company pays IMRT rates for proton patients when proton tx is denied. That should be enough money to run a breast trial. And why hasn't the NCI helped fund any of this. Has no one written up a compelling fundable proton proposal?

Its not really a catch22. It only looks that way because medical devices are not regulated the same way as drugs. There is a reason drug companies spend many 10s to hundreds of million dollars to bring a new drug to market. Its on the manufacturer to prove efficacy and safety before there is an approval and hence reimbursement pathway. You absolutely have to prove they work before billing anything for new drugs. If protons would have been limited to a few select centers to prove efficacy with industry/NIH funding before being given broad billing codes do you think they would have still proliferated? Who knows. Either way, we wouldn't be arguing over the benefits every other week if the proper studies had been done upfront/ever.

 

[OTN]

Apbi is absurd. (Unless you are Ben smith trying to bilk employmers out of tens of millions) Brachy is as well, but would think savi would still be better than protons.

Protons for APBI is so absurd it crosses over into an unethical use of medical resources.

 

[jondunn]

Apbi is absurd.

this is hilariously and hopelessly wrong.

proton for APBI not needed and is pretty ridiculous

but APBI, 30 Gy/5 fractions is pretty incredible. once you do it, you won't go back.

 

[jondunn]

i mean honestly if you don't want to do APBI 5 fx Livi for financial reasons, fine, just do IMPORT-LOW then. there is no reason not to in suitable patients. the cosmesis is better.

 

[TheWallnerus]

Protons for APBI is so absurd it crosses over into an unethical use of medical resources.

Not unethical if it paid the exact same as photons and had the same or better side effects/outcomes as photons.

But we know it doesn't get paid the same, and I have no friggin' idea if it's better, same, or worse side effects/outcomes. A 0.5% rib fracture rate and 5% higher rate of Gr 1/2 erythema vs photons, yeah, protons would be unethical.

The heart argument for protons for early breast has a long row to hoe. There was a ~0.1% incidence of "cardiac events" in IMPORT-LOW e.g.

We can talk 'til we are blue in the face though right. Haters gonna hate. Alligators gonna alligate. Protons gonna protonate.

 

[jondunn]

this is hilariously and hopelessly wrong.

proton for APBI not needed and is pretty ridiculous

but APBI, 30 Gy/5 fractions is pretty incredible. once you do it, you won't go back.

This is why many are pushing for re-certification if you are ten years out or more!

 

[ramsesthenice]

Not unethical if it paid the exact same as photons and had the same or better side effects/outcomes as photons.

This is the issue. Money. Ive seen a number of quips about RCTs here ("we don't need an RCT to know we have to breath") but I think that conflates issues. Do you need an RCT to use protons? No. But I for one do think that you should have some evidence before you are given codes to bill for a higher level of service. And I am not just picking on protons. MRiGRT. Daily adaptive therapy codes. I don't think its at all unreasonable for payers to deny extra charges for lack of evidence.

 

[Tsigolonco noitaidar]

This is the issue. Money. Ive seen a number of quips about RCTs here ("we don't need an RCT to know we have to breath") but I think that conflates issues. Do you need an RCT to use protons? No. But I for one do think that you should have some evidence before you are given codes to bill for a higher level of service. And I am not just picking on protons. MRiGRT. Daily adaptive therapy codes. I don't think its at all unreasonable for payers to deny extra charges for lack of evidence.

I guess the question is what is "some evidence" if not an RCT? And the follow up question is whether VMAT meets (or should meet) that standard of "some evidence" or not when compared to 3DCRT?

 

[Tsigolonco noitaidar]

So this is a point I was trying to make earlier "People saw less side effects with IMRT," and @Tsigolonco noitaidar was saying, if I am not mistaken, that he/she sees less side effects with protons for breast. So anecdotes, personal experience, in silico data, retrospective data, and prospective randomized data support IMRT for breast. What sort of less side effects are seen by docs in the clinic with protons for breast. I am interested to know given that breast patients are the #1 proton patient nationally.

This is a good point and I suppose I should clarify that breast treatments requiring RNI, particularly when IMNs are covered, tend to be much better covered with PBS. If you use VMAT in such situations, you end up with very high cardiac doses. In some cases the IMNs are undercovered down to ~80% if needed to help achieve lower heart doses.

I think the concept of saying "we have lots of RCT data for IMRT over 3D" and pointing to breast data is misleading. The "IMRT" used in those trials was primarily forward planned field-in-field--essentially a glorified 3D with the planner pushing a few MLCs around to cool off the corresponding hot spots. And the comparator in some studies was 2D with physical wedges--a low bar for any technique to supercede. So aside from the multitude of breast RCTs using "IMRT" aka field-in-field technique as the experimental arm, there are but a handful of RCTs comparing real IMRT (e.g. VMAT or multiple field step-in-shoot) to 3DCRT despite the fact that every center in every developed country has had real IMRT for 20 years.

 

[deleted1111261]

This is a good point and I suppose I should clarify that breast treatments requiring RNI, particularly when IMNs are covered, tend to be much better covered with PBS. If you use VMAT in such situations, you end up with very high cardiac doses. In some cases the IMNs are undercovered down to ~80% if needed to help achieve lower heart doses.

I think the concept of saying "we have lots of RCT data for IMRT over 3D" and pointing to breast data is misleading. The "IMRT" used in those trials was primarily forward planned field-in-field--essentially a glorified 3D with the planner pushing a few MLCs around to cool off the corresponding hot spots. And the comparator in some studies was 2D with physical wedges--a low bar for any technique to supercede. So aside from the multitude of breast RCTs using "IMRT" aka field-in-field technique as the experimental arm, there are but a handful of RCTs comparing real IMRT (e.g. VMAT or multiple field step-in-shoot) to 3DCRT despite the fact that every center in every developed country has had real IMRT for 20 years.

Bat signal for @TheWallnerus for IMRT vs “IMRT” lecture!

 

[TheWallnerus]

Bat signal for @TheWallnerus for IMRT vs “IMRT” lecture!

Eh. I’m trying to prevent entropy increasing in the universe. Only in rad onc…

Protons: There is no data for IMRT.
IMRT: Quotes verbatim the IMRT data.
Protons: You are being misleading. (Puts quotes around IMRT.) QED

 

[ramsesthenice]

I guess the question is what is "some evidence" if not an RCT? And the follow up question is whether VMAT meets (or should meet) that standard of "some evidence" or not when compared to 3DCRT?

Are you being serious?

Prostate (0126)
Head and neck
Post op pelvis (cervix)
Lung (0617)

All randomized trials using standard toxicity definitions showing reduced toxicity with IMRT compared to 3D. That’s just what I can name off the top of my head. Next question please.

 

[Tsigolonco noitaidar]

Are you being serious?

Prostate (0126)
Head and neck
Post op pelvis (cervix)
Lung (0617)

All randomized trials using standard toxicity definitions showing reduced toxicity with IMRT compared to 3D. That’s just what I can name off the top of my head. Next question please.

I wasn't asking for or questioning the number of photon RCTs. Just having a conversation with you about what level of evidence you consider to be sufficient to justify a modality since you said protons does not need an RCT to justify its use. I am asking you what level of evidence would you consider sufficient to justify protons.

I see your point about IMRT vs 3D RCTs. There are of course a handful (but not many), especially when the pseudo-IMRT breast trials are excluded. My question in that regard is whether we can take the existing RCTs for photon IMRT > 3D, and extrapolate to every disease site? That is to say, do a few RCTs in a few disease sites showing IMRT > 3D justify its widespread use across all/most disease sites? It's a question meant to provoke discussion and I suppose my point is that I think we should have a similar level of standards for the use of protons as we do for IMRT--i.e. if a certain level of evidence is good enough for widely using IMRT then the same level of evidence may be good enough to widely use protons as well.

 

[TheWallnerus]

I wasn't asking for or questioning the number of photon RCTs. Just having a conversation with you about what level of evidence you consider to be sufficient to justify a modality since you said protons does not need an RCT to justify its use. I am asking you what level of evidence would you consider sufficient to justify protons.

Randomized data showing any benefit would be sufficient

 

[medgator]

Are you being serious?

Prostate (0126)
Head and neck
Post op pelvis (cervix)
Lung (0617)

All randomized trials using standard toxicity definitions showing reduced toxicity with IMRT compared to 3D. That’s just what I can name off the top of my head. Next question please.

0617 secondary analysis confirming IMRT as soc was published in JCO and came in very handy during my evilcore b***h sessions with the p2p docs

 

[medgator]

I wasn't asking for or questioning the number of photon RCTs. Just having a conversation with you about what level of evidence you consider to be sufficient to justify a modality since you said protons does not need an RCT to justify its use. I am asking you what level of evidence would you consider sufficient to justify protons.

I see your point about IMRT vs 3D RCTs. There are of course a handful (but not many), especially when the pseudo-IMRT breast trials are excluded. My question in that regard is whether we can take the existing RCTs for photon IMRT > 3D, and extrapolate to every disease site? That is to say, do a few RCTs in a few disease sites showing IMRT > 3D justify its widespread use across all/most disease sites? It's a question meant to provoke discussion and I suppose my point is that I think we should have a similar level of standards for the use of protons as we do for IMRT--i.e. if a certain level of evidence is good enough for widely using IMRT then the same level of evidence may be good enough to widely use protons as well.

Except said evidence doesn't exist for the big indications afaik

 

[TheWallnerus]

It's a question meant to provoke discussion and I suppose my point is that I think we should have a similar level of standards for the use of protons as we do for IMRT--i.e. if a certain level of evidence is good enough for widely using IMRT then the same level of evidence may be good enough to widely use protons as well.

The shift from 3D to IMRT was simply one of dosing photons in a different temporospatial manner. It wasn't a completely different radiation modality at the basic physics and radiobiological levels. IMRT and 3D are analogous; the shift from 2D to 3D was Bayesianistically informative to achieve the shift from 3D to IMRT. Protons are so different. It's literally a different drug than IMRT whereas IMRT was the same drug as 3D. So one can logically and scientifically argue that protons require a higher bar of proof than IMRT did. The needs-RCTs-for-IMRT argument doesn't fly very well. But even if it did, there are RCTs.

 

[ramsesthenice]

I wasn't asking for or questioning the number of photon RCTs. Just having a conversation with you about what level of evidence you consider to be sufficient to justify a modality since you said protons does not need an RCT to justify its use. I am asking you what level of evidence would you consider sufficient to justify protons.

I see your point about IMRT vs 3D RCTs. There are of course a handful (but not many), especially when the pseudo-IMRT breast trials are excluded. My question in that regard is whether we can take the existing RCTs for photon IMRT > 3D, and extrapolate to every disease site? That is to say, do a few RCTs in a few disease sites showing IMRT > 3D justify its widespread use across all/most disease sites? It's a question meant to provoke discussion and I suppose my point is that I think we should have a similar level of standards for the use of protons as we do for IMRT--i.e. if a certain level of evidence is good enough for widely using IMRT then the same level of evidence may be good enough to widely use protons as well.

Ahhh. I see what you are saying. That’s an interesting question. My bias is not necessarily. Once you have level 1 evidence it becomes more of a balance. Take anal cancer. There is no randomized data of IMRT vs 3D but there are a few so few single arm trials which unambiguously show marked improvements in meaningful endpoints compared to prior 3D RCTs. Couple with positive RCT in a similar anatomical site (post op cervix) I would argue the data is sufficient to support its use and think little would be achieved by doing an RCT for this. Oh wouldn’t use positive data for CNS or H&N data to support thoracic or abdominal IMRT. This is where my issue with the rapid expansion of protons to everything comes in.

I like to think I am a thoughtful person. As an example, I don’t usually do IMRT for rectal cancer. The dose is relatively low and unlike GYN tumors you don’t need to cover the anterior pelvic lymphatics and hence there is not a lot to carve out. By the time you add expansions to account for bladder filling etc the volumes are similar.

 

[ramsesthenice]

I wasn't asking for or questioning the number of photon RCTs. Just having a conversation with you about what level of evidence you consider to be sufficient to justify a modality since you said protons does not need an RCT to justify its use. I am asking you what level of evidence would you consider sufficient to justify protons.

I see your point about IMRT vs 3D RCTs. There are of course a handful (but not many), especially when the pseudo-IMRT breast trials are excluded. My question in that regard is whether we can take the existing RCTs for photon IMRT > 3D, and extrapolate to every disease site? That is to say, do a few RCTs in a few disease sites showing IMRT > 3D justify its widespread use across all/most disease sites? It's a question meant to provoke discussion and I suppose my point is that I think we should have a similar level of standards for the use of protons as we do for IMRT--i.e. if a certain level of evidence is good enough for widely using IMRT then the same level of evidence may be good enough to widely use protons as well.

To your other question, to use a modality you really need safety and comparability data and I would argue protons have met that bar. To get a higher level of billing, there needs to be good evidence of superiority with a meaningful endpoint.

 

[Lamount]

The shift from 3D to IMRT was simply one of dosing photons in a different temporospatial manner. It wasn't a completely different radiation modality at the basic physics and radiobiological levels. IMRT and 3D are analogous; the shift from 2D to 3D was Bayesianistically informative to achieve the shift from 3D to IMRT. Protons are so different. It's literally a different drug than IMRT whereas IMRT was the same drug as 3D. So one can logically and scientifically argue that protons require a higher bar of proof than IMRT did. The needs-RCTs-for-IMRT argument doesn't fly very well. But even if it did, there are RCTs.

This argument is reasonable...

One of the major issues that have plagued protons that uncertainty has not been well accounted for, it is far more relevant for protons vs. photons.

As CW has mentioned, protons will frequently appear to have a "better" DVH than photons, but appearances can be deceptive. These DVH's presume that one's assumptions about the physics of the target are correct... and unlike photons, small errors in density or modeling of heterogeneity can have big consequences on dosimetry.

The good news is that with IMPT and robust optimization, proton physics/dosimetry have improved dramatically in the past 5 years... the bad news is that until there are contemporary trials including this new proton technology in a majority of patients, folks like me are left data-less, waving our hands to folks like you.

 

[radiation]

This argument is reasonable...

One of the major issues that have plagued protons that uncertainty has not been well accounted for, it is far more relevant for protons vs. photons.

As CW has mentioned, protons will frequently appear to have a "better" DVH than photons, but appearances can be deceptive. These DVH's presume that one's assumptions about the physics of the target are correct... and unlike photons, small errors in density or modeling of heterogeneity can have big consequences on dosimetry.

The good news is that with IMPT and robust optimization, proton physics/dosimetry have improved dramatically in the past 5 years... the bad news is that until there are contemporary trials including this new proton technology in a majority of patients, folks like me are left data-less, waving our hands to folks like you.

My biggest issue that we have been doing all this "old" proton treatment off protocol unbeknownst to patients that they have essentially been experimented on without consent while we figured out how to optimize proton delivery. Imagine if protons was a drug and it started causing unexpected brainstem necrosis - that drug would been dead on arrival

 

[BobbyHeenan]

My biggest issue that we have been doing all this "old" proton treatment off protocol unbeknownst to patients that they have essentially been experimented on without consent while we figured out how to optimize proton delivery. Imagine if protons was a drug and it started causing unexpected brainstem necrosis - that drug would been dead on arrival

…the corollary to this is they’ll explain how the lung or prostate trial are (or will be) negative because they didn’t use whatever newer tech *insert pencil beam or more robust planning*

*CMS and insurers enter the chat*

“So you’ve been treating all these people telling us protons is better so you want cases approved. Then the trial is negative. So we want to deny coverage. But then you tell us …well, maybe that old way wasn’t better. But this new way is definitely better. So it’s hard to believe you this time.”

 

[evilbooyaa]

This is a good point and I suppose I should clarify that breast treatments requiring RNI, particularly when IMNs are covered, tend to be much better covered with PBS. If you use VMAT in such situations, you end up with very high cardiac doses. In some cases the IMNs are undercovered down to ~80% if needed to help achieve lower heart doses.

I think the concept of saying "we have lots of RCT data for IMRT over 3D" and pointing to breast data is misleading. The "IMRT" used in those trials was primarily forward planned field-in-field--essentially a glorified 3D with the planner pushing a few MLCs around to cool off the corresponding hot spots. And the comparator in some studies was 2D with physical wedges--a low bar for any technique to supercede. So aside from the multitude of breast RCTs using "IMRT" aka field-in-field technique as the experimental arm, there are but a handful of RCTs comparing real IMRT (e.g. VMAT or multiple field step-in-shoot) to 3DCRT despite the fact that every center in every developed country has had real IMRT for 20 years.

Doesn't have to be VMAT, nothing wrong with fixed angle multi-field static beam IMRT... can get you very low heart doses while giving benefits of IMRT.

Dosimetrists, Physicists, and (most importantly) Radiation Oncologists who accept high cardiac doses are the problem with IMRT for RNI breast plans. Main scenario I can think of where it's truly the patient or the technology that's the problem is in someone where you want to treat the whole IMN chain in say a L breast cancer, not just the first 3 intercostal levels. I'm sure there are some true anatomic exceptions where protons are better, but it's just not necessary in the vast majority of RNI patients. Heck, RNI with something more complicated than 3D isn't necessary in most patients on most treatment machines.

I do IMRT for RNI as needed, both sides, and keep mean hearts < 2-3Gy. You want to tell me that going from < 2Gy to 0.5Gy mean heart dose is going to make a big difference (when original Darby paper had range up to mean heart dose of 27Gy), when the average MHD for L-sided cancers was 6.6Gy, at 7.4% relative risk increase per 1Gy?

 

[elementaryschooleconomics]

Doesn't have to be VMAT, nothing wrong with fixed angle multi-field static beam IMRT... can get you very low heart doses while giving benefits of IMRT.

I agree, I have been pleasantly surprised with what "old school" multi-field static IMRT can do in breast re: heart dose.

In some cases the IMNs are undercovered down to ~80% if needed to help achieve lower heart doses.

Can we say for certain that 80% is "undercovered"? I think this is the narrative I've seen in breast proton papers, because the right/left Danish IMN paper (with an OS benefit) specified 90% of the CTV (not PTV) receive at least 90% of the dose, correct? Though MA.20 was 80%, and B51 allows 90% of the PTV receiving at least 80% of the dose.

I just tried looking up what the Korean JAMA Onc paper from ASTRO specified, and they say it's "per protocol", and when you follow that reference it's a 2009 paper with an English abstract...and the full text is Korean. I cannot read Korean. Curse you, American school system!

I'm genuinely curious, I haven't gone down this rabbit hole because I loathe the question of nodal coverage in breast and have mentally internalized "first three intercostal spaces, 1cm around vessels, check 80% IDL". I assume if I was in the breast proton space, I'd have more of an opinion on this.

 

[Tsigolonco noitaidar]

I agree, I have been pleasantly surprised with what "old school" multi-field static IMRT can do in breast re: heart dose.


Can we say for certain that 80% is "undercovered"? I think this is the narrative I've seen in breast proton papers, because the right/left Danish IMN paper (with an OS benefit) specified 90% of the CTV (not PTV) receive at least 90% of the dose, correct? Though MA.20 was 80%, and B51 allows 90% of the PTV receiving at least 80% of the dose.

I just tried looking up what the Korean JAMA Onc paper from ASTRO specified, and they say it's "per protocol", and when you follow that reference it's a 2009 paper with an English abstract...and the full text is Korean. I cannot read Korean. Curse you, American school system!

I'm genuinely curious, I haven't gone down this rabbit hole because I loathe the question of nodal coverage in breast and have mentally internalized "first three intercostal spaces, 1cm around vessels, check 80% IDL". I assume if I was in the breast proton space, I'd have more of an opinion on this.

I don't know the answer to your undercoverage question--is 80% undercovered? Did MA.20 and other studies choose 80% because they knew they couldn't successfully cover the IMNs with 95% dose in the majority of cases? Or was it because they felt that 80% was sufficient for microscopic disease control. If the latter, then why don't we treat to similar isodose lines for the supraclav as well? Conceptually, microscopic disease in the supraclav and in the IMN should require the same dose. I don't have a point here, I am just asking questions.

 

[Tsigolonco noitaidar]

Doesn't have to be VMAT, nothing wrong with fixed angle multi-field static beam IMRT... can get you very low heart doses while giving benefits of IMRT.

Dosimetrists, Physicists, and (most importantly) Radiation Oncologists who accept high cardiac doses are the problem with IMRT for RNI breast plans. Main scenario I can think of where it's truly the patient or the technology that's the problem is in someone where you want to treat the whole IMN chain in say a L breast cancer, not just the first 3 intercostal levels. I'm sure there are some true anatomic exceptions where protons are better, but it's just not necessary in the vast majority of RNI patients. Heck, RNI with something more complicated than 3D isn't necessary in most patients on most treatment machines.

I do IMRT for RNI as needed, both sides, and keep mean hearts < 2-3Gy. You want to tell me that going from < 2Gy to 0.5Gy mean heart dose is going to make a big difference (when original Darby paper had range up to mean heart dose of 27Gy), when the average MHD for L-sided cancers was 6.6Gy, at 7.4% relative risk increase per 1Gy?

You should publish static field IMRT as a method superior to VMAT or 3D for successfully keeping heart doses low while fully covering IMNs. Would save some folks from protons!

 

[RickyScott]

I don't know the answer to your undercoverage question--is 80% undercovered? Did MA.20 and other studies choose 80% because they knew they couldn't successfully cover the IMNs with 95% dose in the majority of cases? Or was it because they felt that 80% was sufficient for microscopic disease control. If the latter, then why don't we treat to similar isodose lines for the supraclav as well? Conceptually, microscopic disease in the supraclav and in the IMN should require the same dose. I don't have a point here, I am just asking questions.

Xrt for microscopic disease is an s shape curve per radonc urban legend or was it Fletcher? whereby 30 gy is getting rid of 50% microscopic disease and 45–50 Gy 95% so 40gy “, especially with added systemic therapy, would be controlling microscopic disease most of the time.

 

[Tsigolonco noitaidar]

Xrt for microscopic disease is an s shape curve per radonc urban legend or was it Fletcher? whereby 30 gy is getting rid of 50% microscopic disease and 45–50 Gy 95% so 40gy “, especially with added systemic therapy, would be controlling microscopic disease most of the time.

So why don't we treat PMRT/RNI to 40 Gy then? Would certainly save the patient at least 5 fractions. Or does that not fit the SDN narrative of avoiding cutting fractions at all cost because there goes our bonus.

 

[RickyScott]

So why don't we treat PMRT/RNI to 40 Gy then? Would certainly save the patient at least 5 fractions. Or does that not fit the SDN narrative of avoiding cutting fractions at all cost because there goes our bonus.

Because it has been done that way for 50+ years and at least theoretically raising local control by a few percent points without downside.