Viewray‘s Demise

This forum made possible through the generous support of SDN members, donors, and sponsors. Thank you.
I don't think that's what he was saying -

Now, @TheWallnerus can correct me if I'm wrong, but I believe that if you took every department that had a Viewray machine, which was what, 50-55?

So you have those 50-55 sites each just pay the equivalent of a 1.0 FTE RadOnc physician salary/compensation package, that would have been enough money to get them out of bankruptcy.

With these high CapEx ventures, which basically every company selling a radiation machine is, you are banking on the service contract as your LTV because you only sell a machine to a department once every 10-15 years if you're lucky.

I don't know what the service contract with Viewray was (and obviously it was not enough), but rather than hoping angels would swoop in during Chapter 11 and save the company, an alternative strategy would have been to have each existing site pay the equivalent of a 1.0 RadOnc package.

Then, the company wouldn't have died, and really, on an institutional level, that would have been a drop in the bucket investment for each site that was able to pay for a machine and keep it going.

Basically - these letters and social media posts are nice, but words ain't cash.
I get it. But like you said, talk is cheap, and it's not clear to me that the company did the proforma necessary to make sure their tech was financially viable in the long term i guess the replan charges were supposed to be the difference, but that seems kinda nebulous imo

It's sad, given the promise of MRI, but this is the environment we are in. Data free interventions that cost more than the current standard aren't going to cut it long term unless you've got great lobbyists

Members don't see this ad.
 
  • Like
Reactions: 1 user
I don't think that's what he was saying -

Now, @TheWallnerus can correct me if I'm wrong, but I believe that if you took every department that had a Viewray machine, which was what, 50-55?

So you have those 50-55 sites each just pay the equivalent of a 1.0 FTE RadOnc physician salary/compensation package, that would have been enough money to get them out of bankruptcy.

With these high CapEx ventures, which basically every company selling a radiation machine is, you are banking on the service contract as your LTV because you only sell a machine to a department once every 10-15 years if you're lucky.

I don't know what the service contract with Viewray was (and obviously it was not enough), but rather than hoping angels would swoop in during Chapter 11 and save the company, an alternative strategy would have been to have each existing site pay the equivalent of a 1.0 RadOnc package.

Then, the company wouldn't have died, and really, on an institutional level, that would have been a drop in the bucket investment for each site that was able to pay for a machine and keep it going.

Basically - these letters and social media posts are nice, but words ain't cash.
Cash is king for a reason.
 
So because viewray couldn't get a CPT code it's the customers fault?

Honestly, both are at fault here and hopefully both sides take something away from this saga for the next time.... No one will trust new tech again the same way unless Medicare recognizes its value.

Prob better that way anyways... Don't need any more data-free proton boondoggles happening
In a manner of speaking yes; not fault per se. But if you buy a machine that you think is so much clinically better, but doesn’t look great on paper financially, when the company comes knocking for more fees you would oblige or accept the consequences. You would pay the fees in the somewhat far-fetched hope you could gather data on the tech long enough to convince payors it needs a code.
 
  • Like
Reactions: 1 user
Members don't see this ad :)
In a manner of speaking yes; not fault per se. But if you buy a machine that you think is so much clinically better, but doesn’t look great on paper financially, when the company comes knocking for more fees you would oblige or accept the consequences. You would pay the fees in the somewhat far-fetched hope you could gather data on the tech long enough to convince payors it needs a code.
Or in the alternative.. Stick around to maintain your giant paperweight..
 
1699598499395.png
 
  • Haha
Reactions: 1 user

Top image UCLA (one of the first units in the world--#2 or #3), bottom image city of hope. They had finished installing the CoH MR-Linac very soon before the bankruptcy and I believe were in the process of commissioning.

Thought Viewray was a $10 million paperweight now? Treating with it, now?

Several institutions are continuing to treat--mostly in Europe, but a few in the USA have decided to move forward as well.

It's not like the company ceasing to exist means the machine just STOPS. If the company who manufacturers your car stops existing, the car still runs... The question is how to service it, and the machines do need frequent service. These institutions are contracting with the former ViewRay service engineers and other third parties to keep them going.
 
  • Like
Reactions: 3 users
It's not like the company ceasing to exist means the machine just STOPS.
I don't have a Viewray machine, but my neighbor does.

I was under the impression you had to stop treating, per the FDA, unless you could demonstrate continued ability to service/QA the device?

Which, per your post, I guess some places were able to nab those techs and satisfy those requirements? Maybe my neighbor couldn't, which would make sense, they're not huge.

But I'm getting my information secondhand from people over there, and those people aren't the ones in charge of this, so my information is filtered through their flawed understanding.
 
  • Like
Reactions: 2 users
Top image UCLA (one of the first units in the world--#2 or #3), bottom image city of hope. They had finished installing the CoH MR-Linac very soon before the bankruptcy and I believe were in the process of commissioning.



Several institutions are continuing to treat--mostly in Europe, but a few in the USA have decided to move forward as well.

It's not like the company ceasing to exist means the machine just STOPS. If the company who manufacturers your car stops existing, the car still runs... The question is how to service it, and the machines do need frequent service. These institutions are contracting with the former ViewRay service engineers and other third parties to keep them going.
Like Yugos!
 
  • Like
Reactions: 2 users
Members don't see this ad :)
I'm sure Varian is working on it. Tomo was able to do some rudimentary version of this nearly 2 decades ago with the lousiest looking MVCTs ever seen. Additionally, if "adaptive planning" is omitted from ROCR, the demand for CBCT based adaptive planning will be so great, every center in America will purchase.
Was recently at a meeting where ROCR and adaptive were discussed, and it has been specifically decided that any future adaptive code will not be bundled as part of ROCR. Like protons, it will remain exempt, to encourage innovation.

So, daily adaptive protons for the win!

Seriously though, online adaptive will be a big help for particles, probably more than for Xrays, because of the potential impact of anatomy changes on range. There was a debate on this subject at PTCOG-NA in Seattle 3 weeks ago, the audience vote was a 50:50 tie for the proposal that protons should have online adaptive planning.

Some sites like H&N do change, others like CNS not so much
 
Last edited:
  • Like
  • Haha
Reactions: 3 users
Was recently at a meeting where ROCR and adaptive were discussed, and it has been specifically decided that any future adaptive code will not be bundled as part of ROCR.

So, daily adaptive protons for the win!
30 plans in 30 days!
 
  • Like
Reactions: 3 users
Was recently at a meeting where ROCR and adaptive were discussed, and it has been specifically decided that any future adaptive code will not be bundled as part of ROCR. Like protons, it will remain exempt, to encourage innovation.

So, daily adaptive protons for the win!

Seriously though, online adaptive will be a big help for particles, probably more than for Xrays, because of the potential impact of anatomy changes on range. There was a debate on this subject at PTCOG-NA in Seattle 3 weeks ago, the audience vote was a 50:50 tie for the proposal that protons should have online adaptive planning.

Some sites like H&N do change, others like CNS not so much

Any machines that have the ability to do daily adaptive protons currently the way Ethos or MR-Linac can?
 
There is no law, rule, or guidance from the government that revokes the ability of someone to use a medical device just because the parent company goes bankrupt. Seems reasonable to me.
Just like there is no law, rule or guidance from the government that revokes the authority of the parent just because their child makes bad life choices. Seems like a reasonable analogy to me!
 
  • Like
  • Okay...
Reactions: 2 users
Any machines that have the ability to do daily adaptive protons currently the way Ethos or MR-Linac can?

Not that that I am aware, but you know its coming. There are academic careers to be made!

I have to be honest that I do not understand the motivation there. In daily adaptation, I am usually using the technique to move around a sharp gradient and achieve better coverage along an OAR interface.

When I need a sharp gradient, I don't like protons due to imaging, motion management, and robustness challenges.

Pushing around dose in a pancreatic case along the duodenum with protons is kinda scary to me.

That's happened to me before though, someone just needs to show me the vision :)
 
  • Like
Reactions: 2 users
Not that that I am aware, but you know its coming. There are academic careers to be made!

I have to be honest that I do not understand the motivation there. In daily adaptation, I am usually using the technique to move around a sharp gradient and achieve better coverage along an OAR interface.

When I need a sharp gradient, I don't like protons due to imaging, motion management, and robustness challenges.

Pushing around dose in a pancreatic case along the duodenum with protons is kinda scary to me.

That's happened to me before though, someone just needs to show me the vision :)
I 100% agree. I really like using adaptation to generate isotoxic plans. Dose range uncertainty makes this particular activity highly questionable. Not saying it would never be useful, but I think it would be comparatively low yield considering the strengths and weaknesses of each. Further, I know a lot of centers run at or close to capacity. Not sure how enthusiastic they would be about lowering patient throughput. I have to assume optimization would be longer for protons given the additional variables needed to consider.
 
  • Like
Reactions: 1 user
Not that that I am aware, but you know its coming. There are academic careers to be made!

I have to be honest that I do not understand the motivation there. In daily adaptation, I am usually using the technique to move around a sharp gradient and achieve better coverage along an OAR interface.

When I need a sharp gradient, I don't like protons due to imaging, motion management, and robustness challenges.

Pushing around dose in a pancreatic case along the duodenum with protons is kinda scary to me.

That's happened to me before though, someone just needs to show me the vision :)

It's going to lead to even greater worship of the computer screen in proton users.... the next logical step in protons - variable RBE without adaptation is OLD protons! Now we adapt daily so we don't have the same range uncertainty margins that we normally would use!!

Until they start necrosing brainstems and putting holes in duodenums, of course.
 
  • Like
  • Haha
Reactions: 2 users
I 100% agree. I really like using adaptation to generate isotoxic plans. Dose range uncertainty makes this particular activity highly questionable. Not saying it would never be useful, but I think it would be comparatively low yield considering the strengths and weaknesses of each. Further, I know a lot of centers run at or close to capacity. Not sure how enthusiastic they would be about lowering patient throughput. I have to assume optimization would be longer for protons given the additional variables needed to consider.

I think additionally people dont always realize that there is still no function to visualize dose accumulation in on table adaptive platforms. I think this impacts your on-table decision making a lot more than people realize.

If you now come to me on fraction 5/5 and say "we dont know how much the duodenum has actually gotten" and layer that over "the RBE could be 1.4!"

Long optimization is a whole other issue that was kind of a non-issue with the real time on treatment imaging of Viewray. So now you want me to do all of the above, but base my entire mental model of the treatment on an image that is 20 minutes old.

Hard pass (but with an open mind for the future :))
 
  • Like
Reactions: 1 user
Until they start necrosing brainstems and putting holes in duodenums, of course.
The beauty of this is that the patients fly out for protons, get treatment done, then fly back home. Then, when they have a sentinel medical event they wind up in the local ER and nobody know how they were treated so proton don't get blamed.
 
  • Like
  • Wow
Reactions: 4 users
I think additionally people dont always realize that there is still no function to visualize dose accumulation in on table adaptive platforms. I think this impacts your on-table decision making a lot more than people realize.
Yeah, this is another issue. I don't know what you did, but we keep a spread sheet for each patient that therapy updates after each treatment. That way, we can see easily what was over and by how much on prior fractions to decide where we are and if we really need to adapt or not. Of course, even that is too simplistic because it doesn't say what specific part of any organ was over/under but its a decent approximation. Kinda like we do in...brachytherapy. We have a physicist who jokes this is my way of trying to turn EBRT into brachytherapy and they are not entirely wrong :)

The replan optimization is one place the Unity smokes the Viewray systems. Its still not instant, but it is much faster and cuts down on time-dependent changes between image acquisition and treatment planning.
 
  • Like
Reactions: 1 user
Yeah, this is another issue. I don't know what you did, but we keep a spread sheet for each patient that therapy updates after each treatment. That way, we can see easily what was over and by how much on prior fractions to decide where we are and if we really need to adapt or not. Of course, even that is too simplistic because it doesn't say what specific part of any organ was over/under but its a decent approximation. Kinda like we do in...brachytherapy. We have a physicist who jokes this is my way of trying to turn EBRT into brachytherapy and they are not entirely wrong :)

The replan optimization is one place the Unity smokes the Viewray systems. Its still not instant, but it is much faster and cuts down on time-dependent changes between image acquisition and treatment planning.

We did not do that as far as I know. I am no longer there and in my current practice were still in the discussion phase about how, if at all, we will implement an adaptive service.

Im also kind of a vibes guy when it comes to plans where a spreadsheet with point dose numbers would not really increase my comfort. I need to see the plan on the patient haha.

In practice, this wasnt a huge deal because adaptation was always an isotoxic comparison of today versus sim planning. I knew sim planning was safe, so if the adaptive plan did not offer improved coverage and/or reduce OAR DVH, I would toss it and go with the sim plan.

I know that centers are doing this part all differently, so certainly accumulation could matter a lot more for other workflows.

FWIW, I think this is the thing that makes people such unabashed believers to say things like it "saves lives" with no data to back that up. I'd see my HR-PTV coverage go from like 50% to 80% on some days and just be like damnnnn thats sweet. It feels like cheating. SMART trial early results are making an argument that this matters. I hope isotoxic on table adaptation approaches dont fizzle and die.

Haha its like rad oncs are addicted to coverage and dose escalation where it causes physical illness if the numbers are kinda low. Theyre all partaking in that new stuff, cant go back!
 
  • Like
Reactions: 1 user
I think additionally people dont always realize that there is still no function to visualize dose accumulation in on table adaptive platforms. I think this impacts your on-table decision making a lot more than people realize.
Ethos does provide that for the so far delivered dose. You cannot visualize the cumulative dose including the fraction you are about to deliver, but knowing what you delivered so far is helpful. There is the issue about the synthetic CT and if it´s 100&% representative, but with the new hardware update this should be solved.
 
  • Like
Reactions: 1 users
Ethos does provide that for the so far delivered dose. You cannot visualize the cumulative dose including the fraction you are about to deliver, but knowing what you delivered so far is helpful. There is the issue about the synthetic CT and if it´s 100&% representative, but with the new hardware update this should be solved.

I haven't seen that but I haven't seen an Ethos is over a year now. That is awesome!

The argument for Ethos is getting pretty strong. I only saw promotional images, but if those new on board CT images are representative of real life... uphill battle for MR in most centers IMO.
 
  • Like
Reactions: 2 users
I haven't seen that but I haven't seen an Ethos is over a year now. That is awesome!

The argument for Ethos is getting pretty strong. I only saw promotional images, but if those new on board CT images are representative of real life... uphill battle for MR in most centers IMO.
I believe the fuctionality is limited to certain "Intents" (=treatment sites) at the moment. It works well for prostate.
 
  • Like
Reactions: 1 user
I haven't seen that but I haven't seen an Ethos is over a year now. That is awesome!

The argument for Ethos is getting pretty strong. I only saw promotional images, but if those new on board CT images are representative of real life... uphill battle for MR in most centers IMO.

The promotional images are always much better than the actual product.

We've had an Ethos for years. I don't think that we've adapted a single patient on it.

Varian has been trying to convince me for almost 15 years that their cone beam CT quality will make MR-guidance unnecessary. Also, before ~2016 there was always my favorite that MRI-guided radiation therapy would never work technically.

The last time we were in San Diego (no not 2023, I can't even remember what year it was), Varian had me in a private room showing me the "diagnostic CT quality of our new CBCTs". This never materialized. Every year I hear the same thing "check out our NEXT GENERATION CBCT", and then in practice there's all the same lack of contrast, and streak and other artifacts we come to expect from CBCT that limits soft tissue adaptation in abdomen and to some extent pelvis. Also, Ethos motion management has not been very good either, so how are you even dealing with abdomen and lung?

Varian has lied to me many times over the years. I should be fair, the vendors lie all the time. But this for me has been one of the biggest ones.
 
  • Like
Reactions: 10 users
The promotional images are always much better than the actual product.

We've had an Ethos for years. I don't think that we've adapted a single patient on it.

Varian has been trying to convince me for almost 15 years that their cone beam CT quality will make MR-guidance unnecessary. Also, before ~2016 there was always my favorite that MRI-guided radiation therapy would never work technically.

The last time we were in San Diego (no not 2023, I can't even remember what year it was), Varian had me in a private room showing me the "diagnostic CT quality of our new CBCTs". This never materialized. Every year I hear the same thing "check out our NEXT GENERATION CBCT", and then in practice there's all the same lack of contrast, and streak and other artifacts we come to expect from CBCT that limits soft tissue adaptation in abdomen and to some extent pelvis. Also, Ethos motion management has not been very good either, so how are you even dealing with abdomen and lung?

Varian has lied to me many times over the years. I should be fair, the vendors lie all the time. But this for me has been one of the biggest ones.

I can imagine for a bladder case it could do a good job with adaptive on variable bladder filling.....but that streak artifact is real. It can be tricky on pelvis/prostate cases....I could see it be a MAJOR problem with high dose "adaptive" GI cases.
 
  • Like
Reactions: 1 users
The promotional images are always much better than the actual product.

We've had an Ethos for years. I don't think that we've adapted a single patient on it.

Varian has been trying to convince me for almost 15 years that their cone beam CT quality will make MR-guidance unnecessary. Also, before ~2016 there was always my favorite that MRI-guided radiation therapy would never work technically.

The last time we were in San Diego (no not 2023, I can't even remember what year it was), Varian had me in a private room showing me the "diagnostic CT quality of our new CBCTs". This never materialized. Every year I hear the same thing "check out our NEXT GENERATION CBCT", and then in practice there's all the same lack of contrast, and streak and other artifacts we come to expect from CBCT that limits soft tissue adaptation in abdomen and to some extent pelvis. Also, Ethos motion management has not been very good either, so how are you even dealing with abdomen and lung?

Varian has lied to me many times over the years. I should be fair, the vendors lie all the time. But this for me has been one of the biggest ones.
Certainly, the CBCT quality will never match the MRI quality.
But scanning, contouring, adapting and treating a patient is now possible in less than 20' on an Ethos.
This is not something one regularly sees with an MRI linac.

We are only adapting in the pelvis / lower abdomen. Everything above that is not good without an MRI linac.
 
Certainly, the CBCT quality will never match the MRI quality.
But scanning, contouring, adapting and treating a patient is now possible in less than 20' on an Ethos.
This is not something one regularly sees with an MRI linac.

We are only adapting in the pelvis / lower abdomen. Everything above that is not good without an MRI linac.

So like definitive bladder cases or SBRT on nodes where there is variable bowel/sigmoid each day for treatment?
 
  • Like
Reactions: 1 user
I think additionally people dont always realize that there is still no function to visualize dose accumulation in on table adaptive platforms. I think this impacts your on-table decision making a lot more than people realize.
I do no on-linac adaptive planning, but I would think this is intrinsic to all adaptive planning? Your cumulative dose is always an approximation and in 3D is referenced to a single image set. I never trust my cumulative DVH very much with sim based adaptive planning for mobile structures for this exact reason. The OARs are in different places on different scans. I would think the value of on-machine adaptive would be pretty much limited to cases with mobile OARs relative to target, but your cumulative dose information should not be taken too seriously in these cases.

In practice, this wasnt a huge deal because adaptation was always an isotoxic comparison of today versus sim planning. I knew sim planning was safe, so if the adaptive plan did not offer improved coverage and/or reduce OAR DVH, I would toss it and go with the sim plan.
Of course you answered my concern. One should approach adaptive as providing an uncertain dosimetric improvement over the course of treatment relative to your sim plan due to the implementation of daily treatment plans that are either isotoxic or better than the sim plan?

I would be squeamish using adaptive for dose escalation above where I could come up with a safe, single sim plan.
 
  • Like
Reactions: 1 users
I do no on-linac adaptive planning, but I would think this is intrinsic to all adaptive planning? Your cumulative dose is always an approximation and in 3D is referenced to a single image set. I never trust my cumulative DVH very much with sim based adaptive planning for mobile structures for this exact reason. The OARs are in different places on different scans. I would think the value of on-machine adaptive would be pretty much limited to cases with mobile OARs relative to target, but your cumulative dose information should not be taken too seriously in these cases.


Of course you answered my concern. One should approach adaptive as providing an uncertain dosimetric improvement over the course of treatment relative to your sim plan due to the implementation of daily treatment plans that are either isotoxic or better than the sim plan?

I would be squeamish using adaptive for dose escalation above where I could come up with a safe, single sim plan.

Yea I mean all points well taken, but I guess Im still optimistic that someone can come up with a dose accumulation model that would be more informative than a single sim plan.

Totally agree you have to understand the model and its limitations though otherwise you eventually get in trouble.
 
  • Like
Reactions: 1 users
I can imagine for a bladder case it could do a good job with adaptive on variable bladder filling.....but that streak artifact is real. It can be tricky on pelvis/prostate cases....I could see it be a MAJOR problem with high dose "adaptive" GI cases.
Maybe Hypersight will be the magic bullet for CBCT-based IGRT in the upper abdomen, but I'll believe it when I see it.

Adaptive bladder is going to be where the money is for Ethos adaptive IMO. I'm not convinced that the slight improvements in dosimetry on the vast majority of prostate cases will be sufficient to turn out for clinical toxicity. If there was a way to pick a high-risk cohort, then maybe. An alternative idea is to have patients try to come in with a full rectum (and full bladder) if doing adaptive in an attempt to reduce the volumetric amount of rectum receiving dose (like an in-vivo rectal balloon)...
 
Maybe Hypersight will be the magic bullet for CBCT-based IGRT in the upper abdomen, but I'll believe it when I see it.

Adaptive bladder is going to be where the money is for Ethos adaptive IMO. I'm not convinced that the slight improvements in dosimetry on the vast majority of prostate cases will be sufficient to turn out for clinical toxicity. If there was a way to pick a high-risk cohort, then maybe. An alternative idea is to have patients try to come in with a full rectum (and full bladder) if doing adaptive in an attempt to reduce the volumetric amount of rectum receiving dose (like an in-vivo rectal balloon)...
The benefit for bladder ca of on line adaptive versus “poor-man” plan of the day adaptive is unlikely imho
 
  • Like
Reactions: 1 user
Maybe Hypersight will be the magic bullet for CBCT-based IGRT in the upper abdomen, but I'll believe it when I see it.

Adaptive bladder is going to be where the money is for Ethos adaptive IMO.

Varian/Siemens funding a daily adaptive bladder trial. Single arm, so doubt any improvements in acute grade 3 tox will convince many.
 
Varian/Siemens funding a daily adaptive bladder trial. Single arm, so doubt any improvements in acute grade 3 tox will convince many.

We get so few definitive bladder cases anyway... That doesn't seem like a very good application that will sell many people on this.
 
  • Like
Reactions: 2 users
We get so few definitive bladder cases anyway... That doesn't seem like a very good application that will sell many people on this.
Cervical cancer is another good indication for adaptive treatment in the pelvis.
 
  • Like
Reactions: 1 user
Cervical cancer is another good indication for adaptive treatment in the pelvis.

Agree, several groups working on this on the MRI-guided side.

Why not do that on Ethos? More potential impact I would think.

I have a feeling that the CBCT imaging and adaptive workflow (i.e. deformable coregistration) is not good enough for cervix, but open to being wrong.
 
Agree, several groups working on this on the MRI-guided side.

Why not do that on Ethos? More potential impact I would think.

I have a feeling that the CBCT imaging and adaptive workflow (i.e. deformable coregistration) is not good enough for cervix, but open to being wrong.
I believe UCSD at least is doing this. If you do giant 1.5cm ant/post margins without an ITV, it'll probably help. In part the issue is that EBRT to 45/25 isn't going to move the needle a ton in terms of toxicity beyond the improvements in IMRT...
 
We get so few definitive bladder cases anyway... That doesn't seem like a very good application that will sell many people on this.

Expanding indications. What everyone wants. Well except maybe the surgeons.
 
The derms claim they do some sort of ultrasound guided adaptive therapy
 
  • Haha
  • Like
  • Angry
Reactions: 4 users
Top