Protons in the news

[yeasterbunny]

https://www.washingtonpost.com/nati...fore-recusing-himself/?utm_term=.73a275599b0b

“It is undisputed among legitimate medical experts that proton radiation therapy is not experimental and causes much less collateral damage than traditional radiation."

I suppose I will have to agree that protons for prostate is not experimental but we dispute about it all the time!

---

Members don't see this ad.

 

[Neuronix]

Agree that it is not experimental, but disagree that it "causes much less collateral damage than traditional radiation" for most prostate cancer cases.

This is where people outside of our field get confused all the time by the hype and the marketing.

 

[Chartreuse Wombat]

Agree that it is not experimental, but disagree that it "causes much less collateral damage than traditional radiation" for most prostate cancer cases.

This is where people outside of our field get confused all the time by the hype and the marketing.

People INSIDE the field get confused as well.

 

[scarbrtj]

“It is undisputed among legitimate medical experts..." is ~99% of the time a red flag for disputed, highly contentious claims that are usually easily shown to be the opposite of undisputed.

What is difficult is that our ears hear words, and these words have meanings for our brains. In modern society and going forward, we need to somehow train our brains to actually hear the opposite of what some people and their words are saying. It is difficult to "think opposite," but doable.

 

[PointA]

People INSIDE the field get confused as well.

Exactly. Just talk to any proton center doc who primarily treats GU to hear about the many evils of IMRT.

 

[radiaterMike]

It becomes an issue of semantics. The treatment is not experimental but the potential benefit (or detriment) of the treatment is investigational (hence the nationally accruing PARTiQOL trial). A reasonable lawyer for the insurer could simply argue that if there was not equipoise in this, such a trial would not have survived through the layers of review to get off the ground.

 

[medgator]

It becomes an issue of semantics. The treatment is not experimental but the potential benefit (or detriment) of the treatment is investigational (hence the nationally accruing PARTiQOL trial). A reasonable lawyer for the insurer could simply argue that if there was not equipoise in this, such a trial would not have survived through the layers of review to get off the ground.

I think this basically would argue to reimbursing at imrt rates until clinical (not dosimetric) superiority is proven to justify the cost differential . Many proton centers will accept imrt reimbursement levels.... Since they really don't have a choice otherwise

 

[radiaterMike]

I think this basically would argue to reimbursing at imrt rates until clinical (not dosimetric) superiority is proven to justify the cost differential . Many proton centers will accept imrt reimbursement levels.... Since they really don't have a choice otherwise

For the insurers, it still boils down to whether the treatment is not investigational and is medically necessary. SBRT for prostate cancer is much less costly than 9 weeks of IMRT, but insurer's were regularly denying it (before the ASTRO model policy and NCCN guidelines addressed SBRT) and some probably still do (since the NCCN guidelines are not so enthusiastic about it). Also - is it necessary for someone to go out of network for proton therapy ? If proton therapy is clinically superior than it is. Without level 1 evidence or compelling non-randomized evidence of superior outcomes (as was the case for IMRT - before there was some randomized evidence for some cancers) insurers are not going to open that Pandora's box.

 

[evilbooyaa]

It's not experimental (it works just fine) but it's more expensive for the same outcomes.

The fact that people can say, with a straight face "it is undisputed that it causes much less collateral damage" is laughable to me. Yes, it looks better on dosimetry, but the rare clinical toxicity studies I've seen either say it worsens clinical toxicity, or at best, that they are equal, focused on things like prostate, lung, H&N, etc.

I think people who push protons for everything not on a clinical trial are snake oil salesmen/women. IJROBP article about protons for partial breast showing worse acute toxicity with protons. https://www.redjournal.org/article/S0360-3016(18)33121-3/fulltext

Grade 2 dermatitis 69% vs 30% for proton vs photon, respectively.

It's blatant false advertising IMO. Unfortunately the lay population eats up anything that is said without verification, and dismisses people poo-pooing protons as 'the enemy'. Just check the comments of the article....

 

[Lamount]

The toxicity profile of protons have improved recently. IMPT better conforms to irregular shapes and facilitates better skin sparing than double scatter. In situations such has H&N, chordoma/chondrosarcoma, base of skull tumors, and that rare breast case, this will likely make things better. All of this being said, I just don't see how protons are helpful for prostate. Protons' hypersensitivity to setup uncertainties mitigate any dosimetric benefit when variable rectal/bladder filling can dramatically change target location day-to-day. Until we have computers fast enough to do adaptive MC optimization based on pre-treatment CBCT, I just don't see how protons adds anything for prostate.

 

[Palex80]

I just don't see how protons adds anything for prostate.

Prostate cancer patients generate a steady income for proton centers. A steady income from a very "easy" therapy. It's prostate proton therapy that pays back the high installation costs and the personell needed to run a proton center. No proton center makes money by irradiating kids or adults with skull base chordomas.

It's the sad reality. 75 year old patients with prostate cancer "fund" the proton CSI of 10 year olds with medulloblastoma.

 

[medgator]

It's the sad reality. 75 year old patients with prostate cancer "fund" the proton CSI of 10 year olds with medulloblastoma.

Honestly protons are an area where it would be nice for some regulation of supply through certificate of need.

Without that though, bankruptcy will be the only thing that works. I imagine we will be seeing a lot more of those as time goes on

 

[deleted941485]

Honestly protons are an area where it would be nice for some regulation of supply through certificate of need.

Without that though, bankruptcy will be the only thing that works. I imagine we will be seeing a lot more of those as time goes on

They’re popping up everywhere these days. I don’t really understand how they are making this economically feasible. It looks like a bubble.

 

[ramsesthenice]

There are plenty of people that think they are more efficacious for diseases like esophageal. MDACC has retrospective studies showing better survival compared to IMRT. These studies should always be viewed with skepticism, especially when you are suggesting that they improve OS without really improving local control...

Be it as it may, the newest NRG esophageal trial is protons vs IMRT. That’s right, we are doing a huge RCT to try to show an effect with a fairly low ceiling that literally no one outside of our field (at best) will care about. It drives me nuts. We could have at least done a 2x2 with a second randomization to adjuvant immunotherapy or something the rest of the community would actually care about and might substantially improve outcomes. If as a field we try to hang our hats on things like protons or MRI guided radiation we are in trouble. These are useful technologies and should be investigated, but they won’t move the needle anywhere near as much as things like immunotherapy. If we keep leaving these things out of our cooperative trials we will get left behind.

 

[Palex80]

There are plenty of people that think they are more efficacious for diseases like esophageal. MDACC has retrospective studies showing better survival compared to IMRT. These studies should always be viewed with skepticism, especially when you are suggesting that they improve OS without really improving local control...

Be it as it may, the newest NRG esophageal trial is protons vs IMRT. That’s right, we are doing a huge RCT to try to show an effect with a fairly low ceiling that literally no one outside of our field (at best) will care about. It drives me nuts. We could have at least done a 2x2 with a second randomization to adjuvant immunotherapy or something the rest of the community would actually care about and might substantially improve outcomes. If as a field we try to hang our hats on things like protons or MRI guided radiation we are in trouble. These are useful technologies and should be investigated, but they won’t move the needle anywhere near as much as things like immunotherapy. If we keep leaving these things out of our cooperative trials we will get left behind.

I have no idea why someone thinks that protons are promising in healing esophageal cancer. What is the rationale?

Putting in 60+ Gy of dose can be done with photons too.
And sparing out any organ won't be the game changer. Surely you can spare the heart and the lungs better with protons. But people don't die because of excessive dose to those organs. They die because of local recurrence and/or metastatic disease. How are protons supposed to change that.
The problem with esophageal cancer is the same as it is with many other cancers we are trying to cure with an organ preservation approach. Putting more dose into the target means more dose to the OAR that you are trying to preserce.
70 Gy of protons to the esophagus are going to cause the same stenosis as 70 Gy of photons...

 

[ramsesthenice]

I have no idea why someone thinks that protons are promising in healing esophageal cancer. What is the rationale?

Putting in 60+ Gy of dose can be done with photons too.
And sparing out any organ won't be the game changer. Surely you can spare the heart and the lungs better with protons. But people don't die because of excessive dose to those organs. They die because of local recurrence and/or metastatic disease. How are protons supposed to change that.
The problem with esophageal cancer is the same as it is with many other cancers we are trying to cure with an organ preservation approach. Putting more dose into the target means more dose to the OAR that you are trying to preserce.
70 Gy of protons to the esophagus are going to cause the same stenosis as 70 Gy of photons...

Their rationale is that their retrospective studies say so. The go on to postulate that because they have a higher mass maybe they are more immunogenic. As I eluded to above, I am not a believer. If that hypothesis were true, there should be some hint of it when looking at LC, PFS, and OS. If your most robust difference (of the three) with a focal therapy is OS, the most likely explanation is bias. Improved survival should be secondary to, not instead of, improved local regional control.

My bigger point was even granting them that they actually manage to show a 5% absolute increase in survival, Very few people will be swayed. Would protons still offer a benefit with adjuvant immunotherapy? Would it be bigger? What about with other novel DNA repair inhibitors?

From a scientific perspective, the possible immune advantages with protons are not particularly moving. If they stimulated a clinically significant increase in immunogenicity it would not be unique to esophageal cancer and we should have seen a hint of this before now.

 

[scarbrtj]

I have data that prayer (the fervent prayer of a righteous man availeth much?) is more effective for esophageal cancer than dose escalation will be...

The go on to postulate that because they have a higher mass maybe they are more immunogenic.

How much more mass does a proton have than a photon? This much more... which is not really saying anything that means anything. IMHO. Billion-plus dollar WAGs... only in rad onc!

 

[ramsesthenice]

Improvement in OS without difference in LC or PFS suggests to me possible toxicity from treatment.

Not sure I follow your logic. How would treatment toxicity realistically mask gains in LC but not OS? If a treatment is so toxic that patients are not finishing or dying as a result that should have as much or more of an impact on OS as it does LC. Unless there are other competing causes of death which is my entire point.

I don’t disagree that the lymphopenia story could be interesting. You have to admit though that if protons improved OS in a robust way by enhancing immune responses we should have seen it by now in the data. The ceiling for this, on its own, appears low. There are more efficient ways of answering important questions to the field than large, single arm RCTs with limited potential.

Think of it another way. If they show a 5% improvement in OS and the alliance trial with adjuvant durvalumab shows a 15% improvement in OS what is everyone going to think? They need protons and immunotherapy? Not likely. The type of RT won’t matter to the Onc community. Or payers. if they had done a 2x2, it would have been much more convincing that actually yes, the type of RT you get actually might matter whether you get immunotherapy or not.

 

[Lamount]

Not sure I follow your logic. How would treatment toxicity realistically mask gains in LC but not OS? If a treatment is so toxic that patients are not finishing or dying as a result that should have as much or more of an impact on OS as it does LC. Unless there are other competing causes of death which is my entire point.

I don’t disagree that the lymphopenia story could be interesting. You have to admit though that if protons improved OS in a robust way by enhancing immune responses we should have seen it by now in the data. The ceiling for this, on its own, appears low. There are more efficient ways of answering important questions to the field than large, single arm RCTs with limited potential.

Think of it another way. If they show a 5% improvement in OS and the alliance trial with adjuvant durvalumab shows a 15% improvement in OS what is everyone going to think? They need protons and immunotherapy? Not likely. The type of RT won’t matter to the Onc community. Or payers. if they had done a 2x2, it would have been much more convincing that actually yes, the type of RT you get actually might matter whether you get immunotherapy or not.

We aren't going to be asking the "onc community" whether or not we should use protons. We will be asking the insurance companies for approval... and they are much more likely if protons vs. XRT shows a OS benefit. Plus, if there is a signal with protons, you better believe that pharma will want to fund a trial... perhaps a competitor.

 

[ramsesthenice]

We aren't going to be asking the "onc community" whether or not we should use protons.

Depends on where you are. If you are at a center with both that’s true. If not and instead you start asking your referring oncologists to give up their revenue for patients needing CRT so that you can refer them all to a distant proton center for a relatively small benefit compared to their fancy new drugs what do you think will happen? We don’t need their approval but we do need their referrals.

At the end of the day we agree with doing proton vs photon trials. Just not on how or why. The simple design is the best way try to efficiently move towards getting approvals. There is no denying that.

 

[evilbooyaa]

I think it's fine to run a proton vs IMRT randomized trial without additional confounders.

Agree that the retrospective stuff on lymphopenia differences potentially driving OS is interesting but needs prospective validation before we start accepting it.

If protons for esophagus actually gave a 5% OS difference in a RCT I'd be happy to refer my few, rare esophageal cancers out to high-volume proton centers. It's not like esophagus is 25%+ of anybody's clinical volume.

 

[scarbrtj]

Protons, IMRT, esophagus.

I can't say I'm exactly enamored with any of the theories as to why equal GyE's of protons would be more curative than equal Gy's of photons. Unless, 1) what we think is a GyE is not really a GyE, or 2) protons have varying RBEs (like it's got 1.1 RBEs at entry but 3 RBE's once it hits the target) upon encountering various "stuff" inside the body. (Maybe these are the same thing.) Again, a theory that photons cause lymphopenia that allows mets to happen or micromets to flourish, versus protons which don't have this "bug," is a little magical-thinking, to me. As Trump says: we will see.

 

[medgator]

Phase IIB?

https://ascopubs.org/doi/abs/10.1200/JCO.19.02503

 

[Krukenberg]

I think it's fine to run a proton vs IMRT randomized trial without additional confounders.

Agree that the retrospective stuff on lymphopenia differences potentially driving OS is interesting but needs prospective validation before we start accepting it.

If protons for esophagus actually gave a 5% OS difference in a RCT I'd be happy to refer my few, rare esophageal cancers out to high-volume proton centers. It's not like esophagus is 25%+ of anybody's clinical volume.

I think there is really something to the Total Toxicity Burden outcome that protons appeared better in for that MDACC esophagus trial (as an outcome measure, can’t say if the difference is real yet). A problem with proton trials where the promised benefit is reduction of toxicity, is our current toxicity outcomes might not be well suited to measuring the difference we’re looking for. It’s unfortunate the NRG trial isn’t using it as a secondary outcome.

 

[ramsesthenice]

I think there is really something to the Total Toxicity Burden outcome that protons appeared better in for that MDACC esophagus trial (as an outcome measure, can’t say if the difference is real yet). A problem with proton trials where the promised benefit is reduction of toxicity, is our current toxicity outcomes might not be well suited to measuring the difference we’re looking for. It’s unfortunate the NRG trial isn’t using it as a secondary outcome.

I’ll admit I don’t know the details of this outcome you are referring to but I think you are on shaky ground here. You are implying that since protons don’t meet traditional toxicity endpoints we need to come up with new ones. Unless you can provide a compelling argument these new endpoints are clinically more meaningful than validated endpoints (and maybe you can), you are bordering on moving the target to confirm your preconceived hypothesis.

 

[thecarbonionangle]

Phase IIB?

https://ascopubs.org/doi/abs/10.1200/JCO.19.02503

attendings have started recommemding protons based on this. They would IMPT a rock if they could. Also used to steal patients. Just yesterday i received instruction to steal a patient. No questions asked!

 

[RadOncMegatron]

attendings have started recommemding protons based on this. They would IMPT a rock if they could. Also used to steal patients. Just yesterday i received instruction to steal a patient. No questions asked!

Just ask them to explain the protocol if they try to steal! Hopefully it's not @Lamount or @FrostyHammer otherwise you know they won't understand.

 

[ramsesthenice]

attendings have started recommemding protons based on this. They would IMPT a rock if they could. Also used to steal patients. Just yesterday i received instruction to steal a patient. No questions asked!

Yuck. This reeks of imprecision. Superficially it appears the suggestion is no individual toxicity is significantly different but if you add the trend for all of them up you get something statistically significant. This probably is statistically valid. What is the probability if you make 10 comparisons that 7-8 of them would trend in the same direction by chance? Pretty low. But where is the clinical relevance? And is it valid to say IMPT decreases the SEVERITY of toxicity based on this metric? it’s a reach at best.

 

[medgator]

Yuck. This reeks of imprecision. Superficially it appears the suggestion is no individual toxicity is significantly different but if you add the trend for all of them up you get something statistically significant. This probably is statistically valid. What is the probability if you make 10 comparisons that 7-8 of them would trend in the same direction by chance? Pretty low. But where is the clinical relevance? And is it valid to say IMPT decreases the SEVERITY of toxicity based on this metric? it’s a reach at best.

Yup...no meaningful disease control or survival endpoint

 

[RadOncMegatron]

Yuck. This reeks of imprecision. Superficially it appears the suggestion is no individual toxicity is significantly different but if you add the trend for all of them up you get something statistically significant. This probably is statistically valid. What is the probability if you make 10 comparisons that 7-8 of them would trend in the same direction by chance? Pretty low. But where is the clinical relevance? And is it valid to say IMPT decreases the SEVERITY of toxicity based on this metric? it’s a reach at best.

You are correct. Design issues aside (I will say on more thought, I am slowly warming up to this Bayes sequential model for prelim/early phase studies so you can have low numbers and push on toward a phase 3 RCT, but not yet for making practice recommendations), the TTB model is subjective and not validated. Again, being harsh if these are being used to change practice and standard of care, but def. agree with this data to recommend moving forward onto do NRG GI 006. The weights from the protocol are below:

 

[scarbrtj]

Most of the published (how many have gone unpublished?) proton studies: LC 'bout equivalent, toxicity less. I don't think these admittedly (seemingly?) repeatable results have been due to decreased RBE for protons in normal tissues/equal RBE in tumors but rather the increased conformality (at all isovolumes) of protons due to their physics. In theory, with post-modern rad onc moving to a very minimal margin or almost margin-less world, the toxicity Δ for protons and XRT could further decrease. But even if the Δ stays the same or if it decreases... toxicity decreases are great and all yet wowsers that's quite the monetary outlay for some arguably gauzy, goalpost-shifting toxicity improvements. As Dolly Parton says, "It costs a lot of money to look this cheap."

 

[thecarbonionangle]

Yuck. This reeks of imprecision. Superficially it appears the suggestion is no individual toxicity is significantly different but if you add the trend for all of them up you get something statistically significant. This probably is statistically valid. What is the probability if you make 10 comparisons that 7-8 of them would trend in the same direction by chance? Pretty low. But where is the clinical relevance? And is it valid to say IMPT decreases the SEVERITY of toxicity based on this metric? it’s a reach at best.

We’ll see what happens! You’re prob right.

“When a man tells you that he knows the exact truth about anything, you are safe in inferring that he is an inexact man…. It is an odd fact that subjective certainty is inversely proportional to objective certainty. The less reason a man has to suppose himself in the right, the more vehemently he asserts that there is no doubt whatever that he is exactly right.” Bertrand Russell.

 

[RadOncMegatron]

Most of the published (how many have gone unpublished?) proton studies: LC 'bout equivalent, toxicity less. I don't think these admittedly (seemingly?) repeatable results have been due to decreased RBE for protons in normal tissues/equal RBE in tumors but rather the increased conformality (at all isovolumes) of protons due to their physics. In theory, with post-modern rad onc moving to a very minimal margin or almost margin-less world, the toxicity Δ for protons and XRT could further decrease. But even if the Δ stays the same or if it decreases... toxicity decreases are great and all yet wowsers that's quite the monetary outlay for some arguably gauzy, goalpost-shifting toxicity improvements. As Dolly Parton says, "It costs a lot of money to look this cheap."

Can't we just use 41.4 Gy and call it a day?

 

[medgator]

Can't we just use 41.4 Gy and call it a day?

Sure, just not at an NCI-designated cancer center

 

[evilbooyaa]

I think there is really something to the Total Toxicity Burden outcome that protons appeared better in for that MDACC esophagus trial (as an outcome measure, can’t say if the difference is real yet). A problem with proton trials where the promised benefit is reduction of toxicity, is our current toxicity outcomes might not be well suited to measuring the difference we’re looking for. It’s unfortunate the NRG trial isn’t using it as a secondary outcome.

Agree with other posters that validating TTB as a more clinically meaningful endpoint compared to standard CTCAE toxicity outcomes should be validated before we go about using it in a phase III clinical trial. Happy to look at literature on that point.

I also agree that it is a reasonable endpoint for 'concept of proof' studies to minimize the sample size.

If you can't meet the current goal posts, it's not appropriate to simply move the goal posts.

 

[PhotonBomb]

Yup...no meaningful disease control or survival endpoint

but no one would expect this with proton.

the goal is toxicity. and I don’t tbink the first trial meaningfully proved it

 

[ramsesthenice]

but no one would expect this with proton.

No one should expect it. But that doesn't stop people at places like MDACC from publishing comparative retrospective studies suggesting improved survival with protons and then proposing that maybe the slight increase in RBE can enhance the immunogenicity of the radiation etc.

Don't get me wrong. For all of my hesitancy towards protons I would love it if they were more effective or less toxic. But the simple fact is even with many years of experience none of the believers have been able to convincingly show it. It is an inescapable fact at this point that any clinical differences are simply not going to be robust. If they were, we would have seen them by now. Can small differences matter? ABSOLUTELY. Especially in conditions where survivor ship is the norm, not the exception. But when proponents keep using words like "indisputable"the only thing they indisputably accomplish is feeding the skeptics.

 

[medgator]

but no one would expect this with proton.

the goal is toxicity. and I don’t tbink the first trial meaningfully proved it

Some of the pro proton people discussed dose escalation/control as a possible endpoint

 

[evilbooyaa]

Some of the pro proton people discussed dose escalation/control as a possible endpoint

IF they could prove that protons is clinically less toxic at equal doses
THEN they could look into dose escalating in proton to deliver an isotoxic treatment and evaluate for efficacy.

Big IF, and no where close enough toh ave data for the THEN as of yet.

That being said, we've tried dose escalation in most relevant things already and have not seen a benefit (Esophagus, Lung).

 

[Krukenberg]

You are correct. Design issues aside (I will say on more thought, I am slowly warming up to this Bayes sequential model for prelim/early phase studies so you can have low numbers and push on toward a phase 3 RCT, but not yet for making practice recommendations), the TTB model is subjective and not validated. Again, being harsh if these are being used to change practice and standard of care, but def. agree with this data to recommend moving forward onto do NRG GI 006. The weights from the protocol are below:


View attachment 305487

use of composite endpoints to increase power are well-accepted in RCTs. I see TTB as a composite endpoint to capture all the toxicities a patient experiences, but weighted by their burden on the patient. With conventional toxicity outcomes, we usually ignore anything less than grade 3, and make it binary such that a patient who has three grade 3 GI toxicities is counted the same as someone who has one grade 3 GI toxicity. I think the conventional outcomes as described have a high risk of measurement bias. TTB could as well, but I think it’s probably better at measuring what we are interested in.

 

[scarbrtj]

use of composite endpoints to increase power are well-accepted in RCTs. I see TTB as a composite endpoint to capture all the toxicities a patient experiences, but weighted by their burden on the patient.

In my simplistic way of thinking, and when I have done prostate toxicity analyses, I never felt the compulsion to come up with new ways "to capture all the toxicities a patient experiences." We have a measuring stick for that: it's known as 1-t where t equals the grade zero toxicity rate (or incidence). That captures all toxicities. (The weighted hand-waving on t I'm less a fan of; it's very subjective.) The TTB to some extent is the same thing. But again, simplistically, the grade zero toxicity rate is "under compared" in radiation oncology.

 

[ramsesthenice]

use of composite endpoints to increase power are well-accepted in RCTs. I see TTB as a composite endpoint to capture all the toxicities a patient experiences, but weighted by their burden on the patient. With conventional toxicity outcomes, we usually ignore anything less than grade 3, and make it binary such that a patient who has three grade 3 GI toxicities is counted the same as someone who has one grade 3 GI toxicity. I think the conventional outcomes as described have a high risk of measurement bias. TTB could as well, but I think it’s probably better at measuring what we are interested in.

To be sure, there certainly are issues with conventional assessments. The grade 3 notion is a great point. We have all seen prostate hypofrac trials where patients clearly have more grade 2 urinary toxicity and we are told its nothing to worry about because there are no differences in grade 3 toxicities. Further than that, how accurate is the grading for a lot of these things? We know patient recall for specific details can be problematic and in many settings they will answer differently when answering a written survey or face to face questioning. All toxicity scales are imperfect systems. But why do you figure that the TTB is better at measuring what you are really interested in? You could argue that the average uncertainty of the whole data set decreases when you input multiple data points although it should be admitted that isn't necessarily the case. Further, in the TTB you asking a huge number of questions and it is not immediately obvious you are appropriately correcting for the relative effect of taking repeated measures. It is worth pursing as a secondary measure for sure, but until it can be tied to something a little more concrete I think it will remain a hard sell for a lot of us.

 

[FrostyHammer]

Ramsesthenice, with all due respect to your concerns, please read my replies in full as written in the Rad Onc Twitter thread starting on April 10, and avoid editorializing without knowing the reality of such issues. I'll summarize for you here. I'm neither a statistician nor work at any proton center but this is based on my own hard work to think and reason things out myself.

I've personally been waiting forever for people to actually think outside the box for new endpoints instead of the same old junk. Making an endpoint like "CTCAE grade 3+ cardiopulmonary events" requires an absurd sample size and does not take into account the duration and severity of those side effects. In other words, if you're enrolled in a trial that has the CTCAE based endpoint, and you happen to get a grade 3 lung toxicity... Followed by a grade 4 cardiac event, grade 4 sepsis, grade 4 anastomotic leak, etc, it's counted the same way as a patient who got grade 3 lung toxicity and nothing else. Is this the right way to conduct a trial? I think not. Only TTB actually encompasses all these things that could happen for x period of time after CRT or surgery. Do NOT rest your hopes on GI 006 because of CTEP as I wrote in that thread. TTB is the only measure that looks at the "overall picture" which CTEP sadly cannot understand. CTCAE only captures one aspect of toxicities and it's woefully inadequate because the "value" of that endpoint to patients is nothing close to TTB. Like I said in that thread, if the patient is someone I know, I'm not even considering IMRT anymore. "Moving the goalposts" is an absolutely inane way to look at this..."moving the goalposts" is what Dan Spratt does with post hoc (retrospective) analyses before convincing you that his findings are real. That trial was NOT moving the goalposts. It's time to finally wisen up to the fact that protons are going to be a clear failure for some sites (prostate, breast), and advantageous for others (esophagus), and the rest TBD; even within a particular site there will be cases which benefit and cases that won't benefit, so there's little point in generalizing that protons are all good or all bad. Same with any oncologic therapy, even immunotherapy. That's life for you.

 

[scarbrtj]

Ramsesthenice, with all due respect to your concerns, please read my replies in full as written in the Rad Onc Twitter thread starting on April 10, and avoid editorializing without knowing the reality of such issues. I'll summarize for you here. I'm neither a statistician nor work at any proton center but this is based on my own hard work to think and reason things out myself.

I've personally been waiting forever for people to actually think outside the box for new endpoints instead of the same old junk. Making an endpoint like "CTCAE grade 3+ cardiopulmonary events" requires an absurd sample size and does not take into account the duration and severity of those side effects. In other words, if you're enrolled in a trial that has the CTCAE based endpoint, and you happen to get a grade 3 lung toxicity... Followed by a grade 4 cardiac event, grade 4 sepsis, grade 4 anastomotic leak, etc, it's counted the same way as a patient who got grade 3 lung toxicity and nothing else. Is this the right way to conduct a trial? I think not. Only TTB actually encompasses all these things that could happen for x period of time after CRT or surgery. Do NOT rest your hopes on GI 006 because of CTEP as I wrote in that thread. TTB is the only measure that looks at the "overall picture" which CTEP sadly cannot understand. CTCAE only captures one aspect of toxicities and it's woefully inadequate because the "value" of that endpoint to patients is nothing close to TTB. Like I said in that thread, if the patient is someone I know, I'm not even considering IMRT anymore. "Moving the goalposts" is an absolutely inane way to look at this..."moving the goalposts" is what Dan Spratt does with post hoc (retrospective) analyses before convincing you that his findings are real. That trial was NOT moving the goalposts. It's time to finally wisen up to the fact that protons are going to be a clear failure for some sites (prostate, breast), and advantageous for others (esophagus), and the rest TBD; even within a particular site there will be cases which benefit and cases that won't benefit, so there's little point in generalizing that protons are all good or all bad. Same with any oncologic therapy, even immunotherapy. That's life for you.

If protons were put on trial in a court of law for assaulting clinical effectiveness they would be found not guilty. If attempted to be put on trial for murdering cost effectiveness the judge would throw the case out on the first day. (Interventions like herceptin are in prison.)

 

[ramsesthenice]

Ramsesthenice, with all due respect to your concerns, please read my replies in full as written in the Rad Onc Twitter thread starting on April 10, and avoid editorializing without knowing the reality of such issues. I'll summarize for you here. I'm neither a statistician nor work at any proton center but this is based on my own hard work to think and reason things out myself.

I've personally been waiting forever for people to actually think outside the box for new endpoints instead of the same old junk. Making an endpoint like "CTCAE grade 3+ cardiopulmonary events" requires an absurd sample size and does not take into account the duration and severity of those side effects. In other words, if you're enrolled in a trial that has the CTCAE based endpoint, and you happen to get a grade 3 lung toxicity... Followed by a grade 4 cardiac event, grade 4 sepsis, grade 4 anastomotic leak, etc, it's counted the same way as a patient who got grade 3 lung toxicity and nothing else. Is this the right way to conduct a trial? I think not. Only TTB actually encompasses all these things that could happen for x period of time after CRT or surgery. Do NOT rest your hopes on GI 006 because of CTEP as I wrote in that thread. TTB is the only measure that looks at the "overall picture" which CTEP sadly cannot understand. CTCAE only captures one aspect of toxicities and it's woefully inadequate because the "value" of that endpoint to patients is nothing close to TTB. Like I said in that thread, if the patient is someone I know, I'm not even considering IMRT anymore. "Moving the goalposts" is an absolutely inane way to look at this..."moving the goalposts" is what Dan Spratt does with post hoc (retrospective) analyses before convincing you that his findings are real. That trial was NOT moving the goalposts. It's time to finally wisen up to the fact that protons are going to be a clear failure for some sites (prostate, breast), and advantageous for others (esophagus), and the rest TBD; even within a particular site there will be cases which benefit and cases that won't benefit, so there's little point in generalizing that protons are all good or all bad. Same with any oncologic therapy, even immunotherapy. That's life for you.

I don’t actually disagree with much of anything you said there . No modality is going to be uniformly good or bad.

Look, TTP clearly moves the goal posts BUT that may not be a bad thing in the end. We fully agree they are not in the right place with the tools we have. It just remains to be seen how far new tools move them. Time will tell.

 

[RadOncMegatron]

Ramsesthenice, with all due respect to your concerns, please read my replies in full as written in the Rad Onc Twitter thread starting on April 10, and avoid editorializing without knowing the reality of such issues. I'll summarize for you here. I'm neither a statistician nor work at any proton center but this is based on my own hard work to think and reason things out myself.

I've personally been waiting forever for people to actually think outside the box for new endpoints instead of the same old junk. Making an endpoint like "CTCAE grade 3+ cardiopulmonary events" requires an absurd sample size and does not take into account the duration and severity of those side effects. In other words, if you're enrolled in a trial that has the CTCAE based endpoint, and you happen to get a grade 3 lung toxicity... Followed by a grade 4 cardiac event, grade 4 sepsis, grade 4 anastomotic leak, etc, it's counted the same way as a patient who got grade 3 lung toxicity and nothing else. Is this the right way to conduct a trial? I think not. Only TTB actually encompasses all these things that could happen for x period of time after CRT or surgery. Do NOT rest your hopes on GI 006 because of CTEP as I wrote in that thread. TTB is the only measure that looks at the "overall picture" which CTEP sadly cannot understand. CTCAE only captures one aspect of toxicities and it's woefully inadequate because the "value" of that endpoint to patients is nothing close to TTB. Like I said in that thread, if the patient is someone I know, I'm not even considering IMRT anymore. "Moving the goalposts" is an absolutely inane way to look at this..."moving the goalposts" is what Dan Spratt does with post hoc (retrospective) analyses before convincing you that his findings are real. That trial was NOT moving the goalposts. It's time to finally wisen up to the fact that protons are going to be a clear failure for some sites (prostate, breast), and advantageous for others (esophagus), and the rest TBD; even within a particular site there will be cases which benefit and cases that won't benefit, so there's little point in generalizing that protons are all good or all bad. Same with any oncologic therapy, even immunotherapy. That's life for you.

You have really softened my approach to TTB, esp. when you compared it to CTCAE grading (which is really a garbage endpoint). I like the idea of the weighting in the TTB, I dislike the way the TTB is for this current trial (too much weight on low grade events ie 9 asymptomatic pericardial effusions =/= 1 pericardial effusion req. surgical intervention) and feel the elephant in the room is the 41.4 Gy regimen if we are talking about toxicity.

I agree with you that CTEP has dropped the ball on this one. Can't they simply do both CTCAE and at the minimum make TTB an exploratory endpoint?

 

[Krukenberg]

Ramsesthenice, with all due respect to your concerns, please read my replies in full as written in the Rad Onc Twitter thread starting on April 10, and avoid editorializing without knowing the reality of such issues. I'll summarize for you here. I'm neither a statistician nor work at any proton center but this is based on my own hard work to think and reason things out myself.

I've personally been waiting forever for people to actually think outside the box for new endpoints instead of the same old junk. Making an endpoint like "CTCAE grade 3+ cardiopulmonary events" requires an absurd sample size and does not take into account the duration and severity of those side effects. In other words, if you're enrolled in a trial that has the CTCAE based endpoint, and you happen to get a grade 3 lung toxicity... Followed by a grade 4 cardiac event, grade 4 sepsis, grade 4 anastomotic leak, etc, it's counted the same way as a patient who got grade 3 lung toxicity and nothing else. Is this the right way to conduct a trial? I think not. Only TTB actually encompasses all these things that could happen for x period of time after CRT or surgery. Do NOT rest your hopes on GI 006 because of CTEP as I wrote in that thread. TTB is the only measure that looks at the "overall picture" which CTEP sadly cannot understand. CTCAE only captures one aspect of toxicities and it's woefully inadequate because the "value" of that endpoint to patients is nothing close to TTB. Like I said in that thread, if the patient is someone I know, I'm not even considering IMRT anymore. "Moving the goalposts" is an absolutely inane way to look at this..."moving the goalposts" is what Dan Spratt does with post hoc (retrospective) analyses before convincing you that his findings are real. That trial was NOT moving the goalposts. It's time to finally wisen up to the fact that protons are going to be a clear failure for some sites (prostate, breast), and advantageous for others (esophagus), and the rest TBD; even within a particular site there will be cases which benefit and cases that won't benefit, so there's little point in generalizing that protons are all good or all bad. Same with any oncologic therapy, even immunotherapy. That's life for you.

agree. I think it’d be great if NRG or NCI commissioned a group to create a TTB for each disease site (brain, pelvis, etc). Many of our Non-proton RT innovations are to reduce toxicity too and if we can have a more sensitive measure than CTCAE then I’m all for it.

 

[FrostyHammer]

If protons were put on trial in a court of law for assaulting clinical effectiveness they would be found not guilty. If attempted to be put on trial for murdering cost effectiveness the judge would throw the case out on the first day. (Interventions like herceptin are in prison.)

Let's look at the big picture. Of all the $ CMS spends on all radiation (including protons), it spends nearly 60% of that on Neulasta alone. Forget chemo and immunotherapy!

Look, TTP clearly moves the goal posts BUT that may not be a bad thing in the end.

As mentioned above, a well thought out prespecified endpoint is not the definition of moving the goalposts. If anything, historical unilateral/binary definitions of an AE are the historically altered goalposts that no one has bothered resetting back to normal. I for one hope that many many trials down the road have novel endpoints, lest we all keep becoming more sheep in the flock.

Can't they simply do both CTCAE and at the minimum make TTB an exploratory endpoint?

They are doing that for 006 but I will say right now that no one will care if a secondary endpoint is impacted if the PEP is negative. One of the PEPs is OS. This is a lock to be negative.

 

[RickyScott]

With conventional toxicity outcomes, we usually ignore anything less than grade 3,

It has been so long and memory is fading, but didn’t many of the original trials justifying imrt over 3D for prostate hinge on improving grade 2, not grade 3 toxicities?

 

[radoncgrad2019]

It has been so long and memory is fading, but didn’t many of the original trials justifying imrt over 3D for prostate hinge on improving grade 2, not grade 3 toxicities?

I don’t think this trial exists?