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

[TheWallnerus]

---

Members do not see this ad.

McLaren Flint Announces $46 Million Expansion of Cancer Center and Region's Only Proton Therapy Center to Grow Comprehensive Oncology Care in Genesee County | McLaren Health Care News

A $46 million project is launching to expand the Karmanos Cancer Institute at McLaren Flint and the McLaren Proton Therapy Center . Once complete, the projects will allow more patients access to proton therapy, the most advanced radiation therapy available, and bring a more comprehensive...

www.mclaren.org

 

[Mandelin Rain]

McLaren Flint Announces $46 Million Expansion of Cancer Center and Region's Only Proton Therapy Center to Grow Comprehensive Oncology Care in Genesee County | McLaren Health Care News

A $46 million project is launching to expand the Karmanos Cancer Institute at McLaren Flint and the McLaren Proton Therapy Center . Once complete, the projects will allow more patients access to proton therapy, the most advanced radiation therapy available, and bring a more comprehensive...

www.mclaren.org

“proton therapy, the most advanced radiation therapy available” [citation needed]

 

[thesauce]

The data are downloadable here.

There's an orgy of data in this spreadsheet. I will come back here from time to time to update, observe, etc. But first things first. This is the most recent data from 2018. Here are the "top 10" in radiation oncology for that year. Name, location, amount paid by Medicare (2 significant digits).

JW.... Provision Knoxville.... $8.5 million
TK.... Naples, FL.... $5.0 million
AM.... Provision Knoxville.... $5.1 million
TS.... TX Center for Proton Therapy.... $3.1 million
AL.... TX Center for Proton Therapy.... $2.9 million
DD.... Port Charlotte, FL.... $4.0 million
BC.... Princeton Radiology Associates.... $2.4 million
MS.... Urology San Antonio.... $2.3 million
HT.... Princeton Radiology Associates.... $2.3 million
AB.... Ft Myers, FL.... $3.4 million
EDIT1:
PROTON THERAPY CENTER HOUSTON LTD... Houston, TX... $8.3 million

Cursory glance I believe 7/11 are protons, 3/11 are (were) 21st Century Oncology, and 1/11 a "urorads."

Hey ASTRO. C'mon man.

I'm curious what this data is actually showing...

These are all the payments to individual providers including both technical and professional components? So if someone is in a freestanding center and billing professional-only then they have no chance of being on the list, but if they are billing technical then they do (even if practice patterns are the same)?

I've heard of some really busy centers where one doc runs 50-70 on beam consistently, but they are not on this list. How can that be - seems like global collections on that setup would be 10M+?

Also the link seems to be dead. I would love to search this for myself: does anyone know how to do this?

 

[jondunn]

 

[BobbyHeenan]

I'm curious what this data is actually showing...

These are all the payments to individual providers including both technical and professional components? So if someone is in a freestanding center and billing professional-only then they have no chance of being on the list, but if they are billing technical then they do (even if practice patterns are the same)?

I've heard of some really busy centers where one doc runs 50-70 on beam consistently, but they are not on this list. How can that be - seems like global collections on that setup would be 10M+?

Also the link seems to be dead. I would love to search this for myself: does anyone know how to do this?

I don’t know for sure but I’ve known people on that list and know another guy that’s close to those numbers. And none of them make more than average.

It has to do with I think the way the global billing is tagged to the doc somehow. One guy I know was working at an old ?21Century ? place . He owned no equipment but at one point was one of the top Medicare people in the state on these lists for charges in his name.

Obviously those proton docs likely don’t own that equipment either.

 

[Reaganite]

can you expand on this idea out of curiosity?

I'm curious what this data is actually showing...

These are all the payments to individual providers including both technical and professional components? So if someone is in a freestanding center and billing professional-only then they have no chance of being on the list, but if they are billing technical then they do (even if practice patterns are the same)?

I've heard of some really busy centers where one doc runs 50-70 on beam consistently, but they are not on this list. How can that be - seems like global collections on that setup would be 10M+?

Also the link seems to be dead. I would love to search this for myself: does anyone know how to do this?

Someone working in a freestanding center on average is gonna bill like 5x of someone billing prof only in a hospital for any given patient. So all the top guys are always gonna be freestanding per the CMS database. Most aren't taking home anything close to what the CMS database shows they collected, and many could be taking home less than the prof only guys with lower absolute numbers. If you want a very, very rough estimate of what someone working in freestanding charges is actually taking home in professional revenue, divide the CMS number by 5. This database also misses every other payor, so you can have very high volume practices that barely register because they treat few true Medicare patients.

 

[TheWallnerus]

I'm curious what this data is actually showing...

These are all the payments to individual providers including both technical and professional components? So if someone is in a freestanding center and billing professional-only then they have no chance of being on the list, but if they are billing technical then they do (even if practice patterns are the same)?

I've heard of some really busy centers where one doc runs 50-70 on beam consistently, but they are not on this list. How can that be - seems like global collections on that setup would be 10M+?

Also the link seems to be dead. I would love to search this for myself: does anyone know how to do this?

Someone working in a freestanding center on average is gonna bill like 5x of someone billing prof only in a hospital for any given patient. So all the top guys are always gonna be freestanding per the CMS database. Most aren't taking home anything close to what the CMS database shows they collected, and many could be taking home less than the prof only guys with lower absolute numbers. If you want a very, very rough estimate of what someone working in freestanding charges is actually taking home in professional revenue, divide the CMS number by 5. This database also misses every other payor, so you can have very high volume practices that barely register because they treat few true Medicare patients.

You both are right imho. Here's what this data is showing:

1) Shows *collected* under a rad onc's NPI from Medicare for that year. If it's a free-standing doc, it will be global. If hospital based, it almost certainly doesn't show the technical (but can't be sure; I think there are occasionally exceptions depending on what NPI the hospital sends to Mcr for -TC charges). For example, I worked in a hospital (solo) in 2014. That year it shows I collected $381K from Medicare. I know the hospital grossed about $8 million collections (all payors) that year. I think I treated a fair number of Medicare patients. My first year in freestanding about same volume of pts my Mcr number went to $895K. HOWEVER... I believe a fair amt of chemo (I practice with a med onc) is in that. Because when I'm in clinic and he's not chemo goes under my NPI. However vice versa; some RO gets billed under his NPI.
2) Many at top of list are proton freestanding docs obv
3) As Reaganite pointed out, keep in mind: this is just Medicare. For some docs, this is a major payor, and many others it's not.
4) I think Mcr advantage plans outlay are in this but not sure.

So is this data useful? Or just a useless mishmash? I know it is useful. It will show trends over time. It gives a great gestalt. It also shows CMS is totally bonkers to attack RO. Why? For example, in 2013, the avg reimbursement per RO per yr was $356K. In 2019, the last year of data, it was $335K. (The median was $150K in 2013 and 2019.)

ASTRO should be mining this data like BASH wants to mine earth killing comets.

I believe this is an updated link. CMS randomly kills links over time.

Centers for Medicare & Medicaid Services Data

www.cms.gov

 

[thesauce]

You both are right imho. Here's what this data is showing:

1) Shows *collected* under a rad onc's NPI from Medicare for that year. If it's a free-standing doc, it will be global. If hospital based, it almost certainly doesn't show the technical (but can't be sure; I think there are occasionally exceptions depending on what NPI the hospital sends to Mcr for -TC charges). For example, I worked in a hospital (solo) in 2014. That year it shows I collected $381K from Medicare. I know the hospital grossed about $8 million collections (all payors) that year. I think I treated a fair number of Medicare patients. My first year in freestanding about same volume of pts my Mcr number went to $895K. HOWEVER... I believe a fair amt of chemo (I practice with a med onc) is in that. Because when I'm in clinic and he's not chemo goes under my NPI. However vice versa; some RO gets billed under his NPI.
2) Many at top of list are proton freestanding docs obv
3) As Reaganite pointed out, keep in mind: this is just Medicare. For some docs, this is a major payor, and many others it's not.
4) I think Mcr advantage plans outlay are in this but not sure.

So is this data useful? Or just a useless mishmash? I know it is useful. It will show trends over time. It gives a great gestalt. It also shows CMS is totally bonkers to attack RO. Why? For example, in 2013, the avg reimbursement per RO per yr was $356K. In 2019, the last year of data, it was $335K. (The median was $150K in 2013 and 2019.)

ASTRO should be mining this data like BASH wants to mine earth killing comets.

I believe this is an updated link. CMS randomly kills links over time.

Centers for Medicare & Medicaid Services Data

www.cms.gov

So in a free-standing center, you only had 895k global that first year? Was it a partial year? How about the years following?

 

[TheWallnerus]

So in a free-standing center, you only had 895k global that first year? Was it a partial year? How about the years following?

It was a full year. I am always +/- 5% of that in subsequent years. But keep in mind "only" $900K/yr reimbursement puts you (a plain old RO physician you) in the 93-94%ile though in Mcr reimbursement. The 50%ile is $150K a year. That said I am in area where I don't treat a lot of Mcr patients; they're only about 1/3 of my payor mix. If you look through the list, top Mcr reimbursers in rad onc are dominated by two things: Florida and proton MDs. (Texas not far behind.) If we could excise Florida and protons from America, there'd need be no APM!

 

[Mandelin Rain]

If we could excise Florida and protons from America, there'd need be no APM!

We might even get [gulp] raises in reimbursement.

 

[elementaryschooleconomics]

Re: the paper from the Tweet @jondunn posted -

This is an interesting paper. I don't know if this endpoint they're using has clinically meaningful significance, at least acutely. However, good Lord did I feel beat over the head with the "RBE 1.1" thing during residency (I think it showed up on both radbio and physics boards).

 

[madchemist89]

The whole 1.1 GyE RBE is a well known over generalization among radiobiologists. Spinal cord RBE for protons has been suggested to potentially be higher than 1.1. I would not be surprised that other tissues may also have an unknown but increased sensitivity. I'm not antiproton, but I believe they should be considered experimental due to such uncertainties.

 

[RickyScott]

The whole 1.1 GyE RBE is a well known over generalization among radiobiologists. Spinal cord RBE for protons has been suggested to potentially be higher than 1.1. I would not be surprised that other tissues may also have an unknown but increased sensitivity. I'm not antiproton, but I believe they should be considered experimental due to such uncertainties.

At the end of the Bragg peak the rbe can be 1.5-1.7 or more.

 

[TheWallnerus]

The whole 1.1 GyE RBE is a well known over generalization among radiobiologists. Spinal cord RBE for protons has been suggested to potentially be higher than 1.1. I would not be surprised that other tissues may also have an unknown but increased sensitivity. I'm not antiproton, but I believe they should be considered experimental due to such uncertainties.

At the end of the Bragg peak the rbe can be 1.5-1.7 or more.

Physicists get anal over >5% dose deviations. They QA the bejesus out of linacs and TPSs to ensure dose predictability and uniformity. Any physicist worth her salt will report a >10% dose deviation as a misadministration. Greater that 20% dose deviation? The physicist might report the MD to the medical board.

Given this, and that physicists clearly work in proton centers... they must not really understand and/or turn a willing blind eye re: radiobiology.

 

[communitydoc13]

This is an interesting paper. I don't know if this endpoint they're using has clinically meaningful significance, at least acutely.

In my recent obsession with protons, it became clear to me that the clinical practice of reviewing "dose" for proton plans is ridiculous. Many papers out of EU and Asia on variable RBE and RBE adjusted plans that when visualized relative to "dose representation" of plans will shock you.

The most important aspect of this is that these are contemporary papers. Not papers from 2000. The real proton researchers realize they are dealing in a very uncertain dosimetric space. Meanwhile, proton centers are popping up like hot cakes. I believe this is even true in places like Germany.

 

[communitydoc13]

Given this, and that physicists clearly work in proton centers... they must not really understand and/or turn a willing blind eye re: radiobiology.

I would like to get feedback from some physicists working with protons routinely. To me, the basic dosimetric/cell culture level work to establish robustness of RBE just doesn't exist for protons. I do wonder how they justify/dismiss the uncertainties demonstrated in very basic treatment setups and how they have any confidence in how these uncertainties are compounded in the clinical setting.

 

[RickyScott]

Physicists get anal over >5% dose deviations. They QA the bejesus out of linacs and TPSs to ensure dose predictability and uniformity. Any physicist worth her salt will report a >10% dose deviation as a misadministration. Greater that 20% dose deviation? The physicist might report the MD to the medical board.

Given this, and that physicists clearly work in proton centers... they must not really understand and/or turn a willing blind eye re: radiobiology.

I would like to get feedback from some physicists working with protons routinely. To me, the basic dosimetric/cell culture level work to establish robustness of RBE just doesn't exist for protons. I do wonder how they justify/dismiss the uncertainties demonstrated in very basic treatment setups and how they have any confidence in how these uncertainties are compounded in the clinical setting.

Would also be interested in examining comparative photon plans that are allowed to have 180% hot spots. I am sure they would be quite conformal

 

[Lamount]

At the end of the Bragg peak the rbe can be 1.5-1.7 or more.

A few important points. The "Bragg peak" is only clinically relevant in passive scatter plans where you have a de facto "spread out Bragg peak". An SOBP has distinct regions. However, in IMPT, the protons are stopping in a much less systematic way, so you rarely get the same sharp gradient of LETd that you do in passive scatter. In passive scatter, the distal Bragg peak is where the greatest dose variations occur, however you really only see anything bigger than 1.2-1.3 in the area of "dose falloff" where the actual dose is very low (e.g. this region may get 10 cGy and 15 cGy (RBE)), so those RBEs of 1.5-1.7 were rarely clinically relevant, even when we used to use passive scatter across the board.

 

[communitydoc13]

However, in IMPT, the protons are stopping in a much less systematic way

Please explain.

Also, there just has to be basic biologic confirmation somewhere right? I mean MDACC did the basic cell culture experiments and were like (we can't model this). This was across a space of cm in a basic plan set-up. Monte-Carlo not fixing that.

I do wonder if anyone, ever, in the recent history of stellar research oriented candidates, has thought it worth while to put together a 3D, cell culture based biological dosimetric experiment for protons (or even photons for that matter) and attempted to correlate with calculated dosimetry.

The onus is of course greater for the proton or other ion folks. I think it is fairly obvious that directly ionizing radiation is going to much more nuanced in terms of real biologic dose than indirectly ionizing radiation. This also holds for electrons, but nobody uses electrons for fine work at this point.

 

[TheWallnerus]

stellar research oriented candidates, has thought it worth while to put together a 3D, cell culture based biological dosimetric experiment for protons (or even photons for that matter) and attempted to correlate with calculated dosimetry.

 

[thesauce]

It was a full year. I am always +/- 5% of that in subsequent years. But keep in mind "only" $900K/yr reimbursement puts you (a plain old RO physician you) in the 93-94%ile though in Mcr reimbursement. The 50%ile is $150K a year. That said I am in area where I don't treat a lot of Mcr patients; they're only about 1/3 of my payor mix. If you look through the list, top Mcr reimbursers in rad onc are dominated by two things: Florida and proton MDs. (Texas not far behind.) If we could excise Florida and protons from America, there'd need be no APM!

Ok so that does kind of make sense. Essentially 1/3rd of your practice (Medicare patients) added up to 900k total for that year including both professional and technical, which would equate to 2.7M total (assuming medicare rates for all) which makes sense for a reasonably-busy PP.

 

[Lamount]

Please explain.

Also, there just has to be basic biologic confirmation somewhere right? I mean MDACC did the basic cell culture experiments and were like (we can't model this). This was across a space of cm in a basic plan set-up. Monte-Carlo not fixing that.

I do wonder if anyone, ever, in the recent history of stellar research oriented candidates, has thought it worth while to put together a 3D, cell culture based biological dosimetric experiment for protons (or even photons for that matter) and attempted to correlate with calculated dosimetry.

The onus is of course greater for the proton or other ion folks. I think it is fairly obvious that directly ionizing radiation is going to much more nuanced in terms of real biologic dose than indirectly ionizing radiation. This also holds for electrons, but nobody uses electrons for fine work at this point.

With Single Field Optimization (SFO) IMPT, the entire target is covered by each field, which would yield a sharp increase of LETd at the distal edge of the target just like passive scatter… so long as you only have one field. However, With both passive scatter and SFO IMPT, using numerous fields blunts this dramatic increase in LETd (as the distal end of the target from one beam’s perspective would be the proximal end of the target from another beam’ perspective). RBE > 1.5 are unlikely to occur if you have multiple beams.

Most IMPT however is some variation of Multifield optimization (MFO), not SFO, meaning only some portion of the target is covered by each field. This, in effect, will further smooth out the LETd curves

The questions regarding RBE are good ones. I have proposed that exact experiment that you were describing. However, RBE is not only dependent upon the biology of the cell line, but is also dependent upon the endpoint. Are we talking tumor control, growth delay, metastasis free survival, normal tissue, complications (acute versus chronic), carcinogenisis ?

RBE is useful to help approximate… But I think the best way to get accurate predictions is the build our knowledge around proton-specific clinical data.

 

[communitydoc13]

RBE is not only dependent upon the biology of the cell line, but is also dependent upon the endpoint.

Certainly true. Sorry that I provided the wrong link for a proton-cell culture paper in an earlier post (although the one I posted does show a wonderfully corrected CNS proton plan based on RBE and relatable to the recently twittered CNS proton paper out of Heidelberg). I'll copy the link again here.

Relative biological effectiveness in proton beam therapy – Current knowledge and future challenges - PMC

www.ncbi.nlm.nih.gov

Now below is the paper regarding cell culture lines and differing proton energies, including nice graphical representations of dose (yd) vs RBE as well as LETd vs RBE. Unfortunately these relationships are all over the place and will clearly be difficult to model.

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

Proton RBE was found to depend on mean proton energy, survival end point, and cell type. Changes in both <math>y‒D</math> and LET_d were also found to impact proton RBE values, but consideration of the energy spectru …

pubmed.ncbi.nlm.nih.gov

The MDACC paper Exploring the advantages of intensity-modulated proton therapy: experimental validation of biological effects using two different beam intensity-modulation patterns - Scientific Reports is explicitly regarding 2 field IMPT plans. They even employed different opposing beam shapes (down-sloping vs flat) to create a more uniform LETd plan. (It should be noted that the technique they employed to try and get relatively uniform LETd within their target region markedly degraded the conformality of the plan, degrading a purported advantage of protons).

What drives me crazy with these contemporary proton dosimetry papers (other than there clearly should have been a massive push to publish this type of work 15 years ago) is the clear admission that absorbed dose is really an inappropriate term for assessing treatment plans using directly ionizing radiation. The fact that these machines are now disseminated with docs likely looking at absorbed dose plans that dosi/physics gives them verges on insanity IMO.

The beauty of photons is that all that water bath dosimetry work really reflects the production of secondary electrons, which in turn are the vehicles for chemical interaction, DNA damage, tumor death and toxicity. This is completely untrue for directly ionizing radiation, which interacts continuously as it goes through the body, depositing relatively low energy (dose) until the bragg peak, but acting differentially on living tissues as it traverses them. By the time you try to do fancy things with IMPT, you are trying to take advantage of contributions in this low energy deposition region of the beams.

That this was not emphasized during my era (early 2010s) training, which shortly preceded the massive expansion of proton facilities is telling.

build our knowledge around proton-specific clinical data

I agree. This should of course be on protocol. But are the docs at the Procure centers even remotely assessing protons plans reasonably?

 

[jondunn]

'By the time you try to do fancy things with IMPT, you are trying to take advantage of contributions in this low energy deposition region of the beams.'

very well said, and true.

 

[IonsAreOurFuture]

With Single Field Optimization (SFO) IMPT, the entire target is covered by each field, which would yield a sharp increase of LETd at the distal edge of the target just like passive scatter… so long as you only have one field. However, With both passive scatter and SFO IMPT, using numerous fields blunts this dramatic increase in LETd (as the distal end of the target from one beam’s perspective would be the proximal end of the target from another beam’ perspective). RBE > 1.5 are unlikely to occur if you have multiple beams.

Most IMPT however is some variation of Multifield optimization (MFO), not SFO, meaning only some portion of the target is covered by each field. This, in effect, will further smooth out the LETd curves

The questions regarding RBE are good ones. I have proposed that exact experiment that you were describing. However, RBE is not only dependent upon the biology of the cell line, but is also dependent upon the endpoint. Are we talking tumor control, growth delay, metastasis free survival, normal tissue, complications (acute versus chronic), carcinogenisis ?

RBE is useful to help approximate… But I think the best way to get accurate predictions is the build our knowledge around proton-specific clinical data.

Thanks Lamount.

For those interested in learning more, the AAPM TG-256 paper on variable RBE from 2 years ago is quite good. One of the 5 or so strategies they discuss to cope with RBE uncertainty is the use of multiple beams, which sort of smooths out the hotspot contributions from any one beam.

It's also a good way to cope with range uncertainty, to have multiple overlapping beams, so that even if one beam falls a little short, the others will still have good coverage.

For those who requested more cell biology information, there are some cool high- res experiments using 96-count well plates for RBE, LET and range experiments. I believe one was published within the past year or so that looked at protons, helium ions (aka alpha particles), and carbon ions:

"Mapping the Relative Biological Effectiveness of Proton, Helium and Carbon Ions with High-Throughput Techniques"

Mapping the Relative Biological Effectiveness of Proton, Helium and Carbon Ions with High-Throughput Techniques - PMC

Although particle therapy using protons and heavier ions has many inherent advantages when compared to x-rays for cancer treatment, numerous unknowns still exist in the radiobiology of particle therapy. Informative high-accuracy biological effects ...

www.ncbi.nlm.nih.gov

It looks from figure 5 like helium ions have RBE similar to carbon ions but helium would be a lot easier, ie cheaper and compact, to accelerate and steer. The ring might even fit in my garage (2 car).

The Heidelberg folks also did a neat comparison of helium and carbon ion arc therapy plans, what they call "SHArc" - basically VMAT for ions. Their acronym stands for "Scanned Hadron Arc" therapy or something like that. Guess which ion produces better dose distributions?

Helium, not carbon.

 

[radiation]

Thanks Lamount.

For those interested in learning more, the AAPM TG-256 paper on variable RBE from 2 years ago is quite good. One of the 5 or so strategies they discuss to cope with RBE uncertainty is the use of multiple beams, which sort of smooths out the hotspot contributions from any one beam.

It's also a good way to cope with range uncertainty, to have multiple overlapping beams, so that even if one beam falls a little short, the others will still have good coverage.

For those who requested more cell biology information, there are some cool high- res experiments using 96-count well plates for RBE, LET and range experiments. I believe one was published within the past year or so that looked at protons, helium ions (aka alpha particles), and carbon ions:

"Mapping the Relative Biological Effectiveness of Proton, Helium and Carbon Ions with High-Throughput Techniques"

Mapping the Relative Biological Effectiveness of Proton, Helium and Carbon Ions with High-Throughput Techniques - PMC

Although particle therapy using protons and heavier ions has many inherent advantages when compared to x-rays for cancer treatment, numerous unknowns still exist in the radiobiology of particle therapy. Informative high-accuracy biological effects ...

www.ncbi.nlm.nih.gov

It looks from figure 5 like helium ions have RBE similar to carbon ions but helium would be a lot easier, ie cheaper and compact, to accelerate and steer. The ring might even fit in my garage (2 car).

The Heidelberg folks also did a neat comparison of helium and carbon ion arc therapy plans, what they call "SHArc" - basically VMAT for ions. Their acronym stands for "Scanned Hadron Arc" therapy or something like that. Guess which ion produces better dose distributions?

Helium, not carbon.

The problem is the more fields you use with charged particles the less benefit you get - the whole point is to reduce low and medium doses. There is myth that continues to be perpetuated that protons are best for an OAR that is next to your target when the reality is that the conformity index will almost always be worse with proton plans because robustness has to account for both PTV and uncertainty

 

[temujim]

With Single Field Optimization (SFO) IMPT, the entire target is covered by each field, which would yield a sharp increase of LETd at the distal edge of the target just like passive scatter… so long as you only have one field. However, With both passive scatter and SFO IMPT, using numerous fields blunts this dramatic increase in LETd (as the distal end of the target from one beam’s perspective would be the proximal end of the target from another beam’ perspective). RBE > 1.5 are unlikely to occur if you have multiple beams.

Most IMPT however is some variation of Multifield optimization (MFO), not SFO, meaning only some portion of the target is covered by each field. This, in effect, will further smooth out the LETd curves

The questions regarding RBE are good ones. I have proposed that exact experiment that you were describing. However, RBE is not only dependent upon the biology of the cell line, but is also dependent upon the endpoint. Are we talking tumor control, growth delay, metastasis free survival, normal tissue, complications (acute versus chronic), carcinogenisis ?

RBE is useful to help approximate… But I think the best way to get accurate predictions is the build our knowledge around proton-specific clinical data.

The good news with the greater understanding of proton or heavy ion RBEs/LET etc, is we can then re-run any negative trials. Get another bite at the apple. Of course any positive trials will stand.

 

[temujim]

The problem is the more fields you use with charged particles the less benefit you get - the whole point is to reduce low and medium doses. There is myth that continues to be perpetuated that protons are best for an OAR that is next to your target when the reality is that the conformity index will almost always be worse with proton plans because robustness has to account for both PTV and uncertainty

I was on a call with an unnamed TPS vendor... the amount of handwaving over robust planning particularly when it came to field matching (most of the less expensive single vault proton solutions have quite small field sizes) was amazing to me. Just point me to some papers guys -- the torrent of words did not inspire confidence.

 

[IonsAreOurFuture]

I was on a call with an unnamed TPS vendor... the amount of handwaving over robust planning particularly when it came to field matching (most of the less expensive single vault proton solutions have quite small field sizes) was amazing to me. Just point me to some papers guys -- the torrent of words did not inspire confidence.

I think this is one of the better papers on that exact topic:

Clinical Implementation of Robust Optimization for Craniospinal Irradiation​

Alexandria Tasson, et al,
Cancers (Basel)
. 2018 Jan 3;10(1):7.

Clinical Implementation of Robust Optimization for Craniospinal Irradiation - PubMed

With robust optimization for spot scanning proton therapy now commercially available, the ability exists to account for setup, range, and interfield uncertainties during optimization. Robust optimization is particularly beneficial for craniospinal irradiation (CSI) where the large target volume...

pubmed.ncbi.nlm.nih.gov

Most of the newer systems are limited to 25 cm field size max, and some are limited to half gantry (220 degrees or less). I would avoid half gantries like the plague.

 

[TheWallnerus]

I think this is one of the better papers on that exact topic:

Clinical Implementation of Robust Optimization for Craniospinal Irradiation​

Alexandria Tasson, et al,
Cancers (Basel)
. 2018 Jan 3;10(1):7.

Clinical Implementation of Robust Optimization for Craniospinal Irradiation - PubMed

With robust optimization for spot scanning proton therapy now commercially available, the ability exists to account for setup, range, and interfield uncertainties during optimization. Robust optimization is particularly beneficial for craniospinal irradiation (CSI) where the large target volume...

pubmed.ncbi.nlm.nih.gov

Most of the newer systems are limited to 25 cm field size max, and some are limited to half gantry (220 degrees or less). I would avoid half gantries like the plague.

Sorry to sound dumb but does this ("some are limited to half gantry 220 degrees or less") mean the gantry can't rotate for a posterior beam. And, can you even shoot a proton beam through a treatment table? Do proton users do that; is it a thing, frowned upon, or forbidden... is this why proton half gantry tech exists. Proton RT going through a table would make the table problematically radioactive over time, so probably at least frowned upon? I see in this paper the beams are coming posteriorly though (through what?) and the pts seem supine not prone... are tx tables "disposable" in proton centers?

 

[temujim]

I think this is one of the better papers on that exact topic:

Clinical Implementation of Robust Optimization for Craniospinal Irradiation​

Alexandria Tasson, et al,
Cancers (Basel)
. 2018 Jan 3;10(1):7.

Clinical Implementation of Robust Optimization for Craniospinal Irradiation - PubMed

With robust optimization for spot scanning proton therapy now commercially available, the ability exists to account for setup, range, and interfield uncertainties during optimization. Robust optimization is particularly beneficial for craniospinal irradiation (CSI) where the large target volume...

pubmed.ncbi.nlm.nih.gov

Most of the newer systems are limited to 25 cm field size max, and some are limited to half gantry (220 degrees or less). I would avoid half gantries like the plague.

That's fine and all, and will go in the files. But what I really want to see is QA with variations in position of phantom and/or summed across imaging from daily set ups. And then I will worry about radioactive tables haha.

Really gone down the proton/heavy ions rabbit hole lately. I feel bad for people pretending they are knowing what they are doing with this tech. You know who you are.

 

[IonsAreOurFuture]

Sorry to sound dumb but does this ("some are limited to half gantry 220 degrees or less") mean the gantry can't rotate for a posterior beam. And, can you even shoot a proton beam through a treatment table? Do proton users do that; is it a thing, frowned upon, or forbidden... is this why proton half gantry tech exists. Proton RT going through a table would make the table problematically radioactive over time, so probably at least frowned upon? I see in this paper the beams are coming posteriorly though (through what?) and the pts seem supine not prone... are tx tables "disposable" in proton centers?

IBA and Mevion sell compact proton systems that are a single room in size and not expandable to multi-room systems. The gantry can rotate in a single direction from AP, to lateral, to PA: 180 degrees, but not back up or down the other side. To treat a left lateral beam and then a right lateral beam on a single patient, the couch has to be kicked 180 degrees, which takes up time, especially if the patient is re-imaged to confirm the new position, so it hurts throughput and patients can get uncomfortable on the table. These half-gantry systems also don't have as good a reputation for reliability in my opinion, but probably cost less than a Varian or Hitachi system.

Hitachi and Varian sell single room systems with a compact gantry, and they rotate 360 degrees and are expandable to serve multiple rooms from a single accelerator (synchrotron for Hitachi, cyclotron for Varian). That's good if you want to start with a single room system and add a 2nd vault down the road. They look and function more like a traditional linac vault would, especially the Varian one, it looks almost like a Truebeam to me and has similar user controls and Eclipse planning system. These 2 vendors also have a better reputation for reliability and uptime.

Treating through the table isn't seen as a big deal unless you treat through a truss support under the table or some part that is significantly thicker in an unpredictable way that would shift the range. I've not heard about table activation being a significant concern in clinic. I think the treated patient would get just as hot if not hotter since there is more dose deposition actually occurring in the patient. My physicists tell me that the biggest source of neutrons in the room is the patient themselves, at least in a pencil-beam scanning system without any blocks or other obstacles in the beamline.

 

[jondunn]

IBA and Mevion sell compact proton systems that are a single room in size and not expandable to multi-room systems. The gantry can rotate in a single direction from AP, to lateral, to PA: 180 degrees, but not back up or down the other side. To treat a left lateral beam and then a right lateral beam on a single patient, the couch has to be kicked 180 degrees, which takes up time, especially if the patient is re-imaged to confirm the new position, so it hurts throughput and patients can get uncomfortable on the table. These half-gantry systems also don't have as good a reputation for reliability in my opinion, but probably cost less than a Varian or Hitachi system.

Hitachi and Varian sell single room systems with a compact gantry, and they rotate 360 degrees and are expandable to serve multiple rooms from a single accelerator (synchrotron for Hitachi, cyclotron for Varian). That's good if you want to start with a single room system and add a 2nd vault down the road. They look and function more like a traditional linac vault would, especially the Varian one, it looks almost like a Truebeam to me and has similar user controls and Eclipse planning system. These 2 vendors also have a better reputation for reliability and uptime.

Treating through the table isn't seen as a big deal unless you treat through a truss support under the table or some part that is significantly thicker in an unpredictable way that would shift the range. I've not heard about table activation being a significant concern in clinic. I think the treated patient would get just as hot if not hotter since there is more dose deposition actually occurring in the patient. My physicists tell me that the biggest source of neutrons in the room is the patient themselves, at least in a pencil-beam scanning system without any blocks or other obstacles in the beamline.

preach, sis!

 

[IonsAreOurFuture]

I just noticed a lot of proton announcements for new centers or expansions, totaling quite a few new rooms at existing centers and single rooms in smaller cities:

MD Anderson expansion - 4 new rooms for 8 total
Mayo Clinic MN expansion - 2 new rooms for 6 total
Mayo Clinic Jacksonville - looks like 4 rooms, some mix of protons and carbon
U PENN - 2 new single room centers in Lancaster PA and Southern NJ
Flint, MI McLaren - 2 new rooms for 4 total
University of Kansas single room IBA
Mercy Hospital St Louis - single room
University of Arkansas - single room
Naperville, IL - single room
Danbury, VT
New Jersey Atlantic Health
FLorida - provision 3 room center near Orlando
Tampa x 2 centers
Florida - Advocate Rad Onc - in Southwest Florida, like Naples
North Carolina - Atrium Health; I would have bet Duke would be first, but nope
Albuquerque, NM - yes, New Mexico, single room IBA

and last but not least, but the most disruptive:

Kansas City Urology - Mevion single room - it has now arrived - Uro Rads with proton therapy

In total, 29 new rooms coming online within the next 5 years that I know of. With the APM possibly dead, I wonder if more will follow.

 

[Wildcat321]

I just noticed a lot of proton announcements for new centers or expansions, totaling quite a few new rooms at existing centers and single rooms in smaller cities:

MD Anderson expansion - 4 new rooms for 8 total
Mayo Clinic MN expansion - 2 new rooms for 6 total
Mayo Clinic Jacksonville - looks like 4 rooms, some mix of protons and carbon
U PENN - 2 new single room centers in Lancaster PA and Southern NJ
Flint, MI McLaren - 2 new rooms for 4 total
University of Kansas single room IBA
Mercy Hospital St Louis - single room
University of Arkansas - single room
Naperville, IL - single room
Danbury, VT
New Jersey Atlantic Health
FLorida - provision 3 room center near Orlando
Tampa x 2 centers
Florida - Advocate Rad Onc - in Southwest Florida, like Naples
North Carolina - Atrium Health; I would have bet Duke would be first, but nope
Albuquerque, NM - yes, New Mexico, single room IBA

and last but not least, but the most disruptive:

Kansas City Urology - Mevion single room - it has now arrived - Uro Rads with proton therapy

In total, 29 new rooms coming online within the next 5 years that I know of. With the APM possibly dead, I wonder if more will follow.

Don't forget Peoria! Looks like they're getting a multi-room setup.

OSF Breaks Ground On New $237 Million Cancer Institute

OSF HealthCare's Cancer Institute will offer proton beam therapy cancer treatment.

www.wcbu.org

Before a groundbreaking ceremony on Tuesday, the bishop for the Catholic Diocese of Peoria, Louis Tylka, led OSF workers in prayer. He encircled a mound of dirt in front of a Caterpillar excavator and blessed the land, flicking holy water at the ground.

Maybe ASTRO should bring on the Bishop as a lobbyist?

 

[communitydoc13]

proton announcements for new centers or expansions

It's inexorable....eff it. Is there physics staffing to account responsibly for uncertainties in all these places? Is there physician on-boarding for understanding how to interpret a proton plan outside of interpreting dose as dose?

What does the industry for proton training look like? Or has it been sold as a natural extension of photon therapy with minimal additional insight needed?

 

[TheWallnerus]

I just noticed a lot of proton announcements for new centers or expansions, totaling quite a few new rooms at existing centers and single rooms in smaller cities:

MD Anderson expansion - 4 new rooms for 8 total
Mayo Clinic MN expansion - 2 new rooms for 6 total
Mayo Clinic Jacksonville - looks like 4 rooms, some mix of protons and carbon
U PENN - 2 new single room centers in Lancaster PA and Southern NJ
Flint, MI McLaren - 2 new rooms for 4 total
University of Kansas single room IBA
Mercy Hospital St Louis - single room
University of Arkansas - single room
Naperville, IL - single room
Danbury, VT
New Jersey Atlantic Health
FLorida - provision 3 room center near Orlando
Tampa x 2 centers
Florida - Advocate Rad Onc - in Southwest Florida, like Naples
North Carolina - Atrium Health; I would have bet Duke would be first, but nope
Albuquerque, NM - yes, New Mexico, single room IBA

and last but not least, but the most disruptive:

Kansas City Urology - Mevion single room - it has now arrived - Uro Rads with proton therapy

In total, 29 new rooms coming online within the next 5 years that I know of. With the APM possibly dead, I wonder if more will follow.

What is sad, APM or no, is that this increases the tension w/ Evicore (and private insurance in general) for every other practicing rad onc in the US in every other disease site. This boom in protons will literally drive down salaries in rad onc because we now seem engaged in a zero sum game, reimbursement-wise. Eventually as proton use gets more "egregious," CMS will have no choice but to 1) do APM, or 2) begin ratcheting down reimbursement in the non-proton space (ie, CPT code 77338 was a response by CMS around 2015 to decrease IMRT reimbursement by letting this code cannibalize the multiple billing of of other codes).

North Carolina - Atrium Health; I would have bet Duke would be first, but nope

Duke was one of last academic centers in US to get IMRT.

 

[Mandelin Rain]

CMS could banish APM AND give us all 10% raises (editor's note: after years of declines in the face of record setting inflation) if it eliminated PPS loophole and paid protons as IMRT

This is a simple, easy, fair, just, and commonsense solution. So it will never happen.

 

[Mandelin Rain]

I hope these business models and proformas build in that as proton therapy becomes widely adopted (hint: we're there), the reimbursement will/must decline. Likely significantly. They're going to squeezing blood from a stone in a few years. (editor's note: unless they got that sweet, sweet PPS exemption)

 

[medgator]

I hope these business models and proformas build in that as proton therapy becomes widely adopted (hint: we're there), the reimbursement will/must decline. Likely significantly. They're going to squeezing blood from a stone in a few years. (editor's note: unless they got that sweet, sweet PPS exemption)

They're not all going to make it.... Probably could scoop up a mevion unit on the cheap by the end of the decade I'm guessing

 

[deleted941485]

They're not all going to make it.... Probably could scoop up a mevion unit on the cheap by the end of the decade I'm guessing

Yeah. Im in the go with the flow crowd on this one. It’s literally the only thing that grabs attention on the RO side. CMS be damned. We need bodies on these machines though which I’m curious how that will work out.

 

[communitydoc13]

It’s literally the only thing that grabs attention on the RO side.

I guess I am just too far removed to know how people really feel about them, how the physicists feel or the docs for that matter. My self-taught understanding of proton dosimetry, as well as the interactions I have had with both leaders who were reluctant to invest 10 years ago and excellent docs who have no remorse leaving protons behind to treat patients with photons, makes me so reluctant to believe that this collective investment, decentralized and removed from serious research, is anything but a collective and hollow money grab.

A bishop praying over a gun, buried deep within the shielded ground to be as close to a permanent human edifice as exists, with proclamations of "a pinpoint-targeted radiation treatment that leaves surrounding healthy tissue untouched" is truly heart of darkness stuff.

Wondering if those radonc leaders, side-eyeing each other during the ritual, knowing darn well that even a perfectly conformal radiation intervention will inevitably be a peripheral part of cancer care, will ever have a "the horror!" moment.

 

[TheWallnerus]

I guess I am just too far removed to know how people really feel about them, how the physicists feel or the docs for that matter. My self-taught understanding of proton dosimetry, as well as the interactions I have had with both leaders who were reluctant to invest 10 years ago and excellent docs who have no remorse leaving protons behind to treat patients with photons, makes me so reluctant to believe that this collective investment, decentralized and removed from serious research, is anything but a collective and hollow money grab.

A bishop praying over a gun, buried deep within the shielded ground to be as close to a permanent human edifice as exists, with proclamations of "a pinpoint-targeted radiation treatment that leaves surrounding healthy tissue untouched" is truly heart of darkness stuff.

Wondering if those radonc leaders, side-eyeing each other during the ritual, knowing darn well that even a perfectly conformal radiation intervention will inevitably be a peripheral part of cancer care, will ever have a "the horror!" moment.

Apocalypse soon

 

[deleted941485]

“this collective investment, decentralized and removed from serious research, is anything but a collective and hollow money grab.”


This is radiation oncology…the specialty I’m going to hell for. There’s no turning back now.

You won’t get a quick death either. Just a slow decline into irrelevance and mediocrity.

 

[TheWallnerus]

“this collective investment, decentralized and removed from serious research, is anything but a collective and hollow money grab.”


This is radiation oncology…the specialty I’m going to hell for. There’s no turning back now.

You won’t get a quick death either. Just a slow decline into irrelevance and mediocrity.

 

[deleted1111261]

Don’t Look Up just nails it for RO.

It doesn’t work for climate change or COVID, as many critics are saying it’s talking about.

Those don’t fit as well as our dumpster fire!

 

[seper]

At least, no new centers announced in DC-Baltimore region

 

[elementaryschooleconomics]

Wow...ok.

Guys - using proton CSI for leptomeningeal carcinomatosis is a bridge too far.

Now, I'm totally on board if we're just saying nodular leptomeningeal disease has better outcomes than carcinomatosis. I don't think literally anyone would disagree with that statement, and that does appear to be the point of this study.

But, to state the obvious, the main benefit of protons over photon is the potential for reduced side effects.

Are we really choosing wisely if we're using protons in a disease where PFS is being measured in months?

Just to put everything together here:
Sloan is PPS-exempt, and their negotiated reimbursements are significantly higher than average
Sloan is in NYC and charges $12 an hour for parking
Per now-deleted Tweets, Sloan obtains routine ECGs on all patients, all of the time
Sloan is, at least occasionally, using proton CSI for patients with leptomeningeal disease which, on average has a very poor prognosis

To be fair to Sloan, perhaps they only did it for this study and will never do it again?

"But ESE", you say, "this is just one institution doing this just a little bit". While true - this helps establish precedent. What's stopping INOVA, with their new proton center, from getting referrals for lepto and treating with proton CSI, using this study as justification? (not singling INOVA out - anyone with protons can join the fun!)

 

[fiji128]

I just noticed a lot of proton announcements for new centers or expansions, totaling quite a few new rooms at existing centers and single rooms in smaller cities:

MD Anderson expansion - 4 new rooms for 8 total
Mayo Clinic MN expansion - 2 new rooms for 6 total
Mayo Clinic Jacksonville - looks like 4 rooms, some mix of protons and carbon
U PENN - 2 new single room centers in Lancaster PA and Southern NJ
Flint, MI McLaren - 2 new rooms for 4 total
University of Kansas single room IBA
Mercy Hospital St Louis - single room
University of Arkansas - single room
Naperville, IL - single room
Danbury, VT
New Jersey Atlantic Health
FLorida - provision 3 room center near Orlando
Tampa x 2 centers
Florida - Advocate Rad Onc - in Southwest Florida, like Naples
North Carolina - Atrium Health; I would have bet Duke would be first, but nope
Albuquerque, NM - yes, New Mexico, single room IBA

and last but not least, but the most disruptive:

Kansas City Urology - Mevion single room - it has now arrived - Uro Rads with proton therapy

In total, 29 new rooms coming online within the next 5 years that I know of. With the APM possibly dead, I wonder if more will follow.

Wow if that Naperville, IL, one is true, that's crazy. Naperville is like 10 miles away from the already operating Northwestern affiliated proton center in Wheaton.

 

[TheWallnerus]

Are we really choosing wisely if we're using protons in a disease where PFS is being measured in months?

No. 327 on the List of Why We Can't Have Nice Things in Rad Onc. The ECG thing is no. 52.