Linac new product
Started by RickyScott
Get help with your application
Use all the free resources available to you from SDN: articles, guides, expert advising, forums discussions, and school research.
I remember they were trademarking things around cardiac SBRT and seeking FDA approval for a new device focused on that. Maybe that’s it?
Advertisement - Members don't see this ad
They are acting like Apple.
Also, if you are going to the Exhibit Hall in ASTRO this year, good luck trying to find the name formerly known as Varian. It is all Siemens now except for the tiny plastic lettering taped on the linacs.
Something else. Halcyon + Couch was already announced at ESTRO last month.
We should start an informal betting pool
How do I search this on Polymarket!?
Kalshi
I know what it is so let’s definitely do this 🙂
My gut feeling tells me that it's going to go in the direction of CBCT-(HyperSight)-enabled adaptive therapy on C-arm linacs. It's clear, that's where Varian is betting, since it has a clear advantage over Elekta there. "Adaptive-for-all" (and not only on Ethos) is where they want to bet.
Elekta is also marketting a similar solution, however (from what I understand) it is semi-online and has to go over Monaco. But they seem to have bought MIM (?) autocontouring and somehow integrated it in their adaptive workflow.
I also presume that Varian is developing a more "user-friendly" software solution, away from the overloaded ARIA, into a more graphic-intuitive-Ethos-style platform. Elekta is also trying to do something similar, by moving away from the cumbersome Mosaiq, that still reminds of LANTIS - a product for Siemens Linacs in the 90s (kind of ironic, huh?) - to this web-based Mosaiq-One platform.
Elekta is also marketting a similar solution, however (from what I understand) it is semi-online and has to go over Monaco. But they seem to have bought MIM (?) autocontouring and somehow integrated it in their adaptive workflow.
I also presume that Varian is developing a more "user-friendly" software solution, away from the overloaded ARIA, into a more graphic-intuitive-Ethos-style platform. Elekta is also trying to do something similar, by moving away from the cumbersome Mosaiq, that still reminds of LANTIS - a product for Siemens Linacs in the 90s (kind of ironic, huh?) - to this web-based Mosaiq-One platform.
Last edited:
Why would anyone buy an ethos if truebeam could adapt?
The Ethos-workflow is very specific and allows you to make adaptive therapy very lean. RTTs can basically adapt on their own.
I am not sure a Truebeam-based adaptive workflow would work as lean/fast as on an Ethos, I guess (if that's what Varian is trying to do), we will find out.
The bottlenecks here are:
- a gantry linac can rotate a lot faster --> CBCT will be faster on the Ethos.
- software --> Truebeam still runs on ARIA, making it "complicated".
Ethos is a stand-alone, confined software solution, with some interaction with ARIA. You can basically run an Ethos with a minimal ARIA installation, alone.
ETHOS is a bit like early iOS versions for the iPhone. Very user intuitive, however with minimal customization and totally sealed off. Basically, a black box.
Last edited:
Advertisement - Members don't see this ad
The adaptive RT reimbursement story is a pretty classic rad onc problem: vendors sold the workflow as “the future,” everyone nodded along because ASTRO was expected to carve out payment, and now we are stuck in the awkward middle where the machine can do the thing, the staff work is real, but the billing pathway is still mostly imaginary.
2026 did not create a clean “adaptive add-on” code. Delivery got reshuffled. IGRT professional survived. Complexity buckets changed. But there is still no standalone CPT code for adaptive RT, and at least some payers are already saying: adaptive planning has not been valued by CMS, so don’t try to rebill 77300/77301/77338/etc every time the anatomy changes.
ROCR is the promised land here: transitional modifier/payment now, real CPT someday, eventual valuation later. Fine. But that is an advocacy pathway, not current cash flow. If a vendor pro forma assumes daily adaptive revenue, ask them for payer-specific written policy, not a PowerPoint from the booth at ASTRO. As multiple people have posted before, ROCR is pretty much dead on arrival in this Congress.
Clinically, adaptive is not fake. It clearly has dosimetric value and probably real value in selected situations: pancreas/upper abdomen, bladder, cervix, H&N anatomy change, maybe certain thoracic cases. But the outcomes data are uneven. MIRAGE supports MRI guidance/margin reduction in prostate SBRT, not “daily adaptive billing.” SMART pancreas data are encouraging but largely phase II/single-arm. Bladder adaptive has randomized feasibility/safety/dose-escalation data. A lot of the rest is still dosimetric, workflow, or ongoing-trial territory. As we all know, our beloved field has a nasty habit of conflating favorable dosimetry in treatment planning with phase III randomized clinical outcomes.
So my position is simple: adaptive RT is a legitimate clinical tool. It is not yet a universal clinical mandate or a guaranteed reimbursement arbitrage. Buying a multi-million-dollar platform based on “ASTRO will fix the codes” is not strategy. It is hope with depreciation.
I couldn't agree more. About a third of what I do is adaptive pancreatic SBRT so I am an advocate. That said, I don't see a realistic advantage for 90+% of patients getting radiation. Its not particularly efficient and struggle to see adding an adaptive workflow to the equivalent of a Versa or TrueBeam as particularly high yield.
it will be a very long wait if it ever comes and the story is not that different from protons. You have to critically assess the advantages of the technology and get realistic about what it can and can't do. Adaptive therapy isn't going to be more efficacious on its own (though in select situations it could facilitate meaningful dose escalation). It also has a roughly 0% chance of decreasing common acute grade 2-3 toxicities without evaluating hundreds if not thousands of patients. The most plausible benefit is a reduction in significant overdosing of adjacent OARs but even then getting the data is a challenge for a couple of reasons. Severe toxicities are relatively rare and likely under reported. The bar to showing significant improvement is very high. Second, the null hypothesis has largely been accepted based on first principle and the highest risk settings (which are most likely to show a benefit) have largely skipped over testing to just using it already.
FletchersGarb
Full Member
Planning codes don't require prior auth (7300/77301/77338), so I suppose pushback occurs via claims not being paid?
So...possible dosimetric benefit which has not been shown to translate into any survival benefit, loco-regional control benefit, or toxicity reduction? "Story is not that different from protons" isn't filling me with confidence.
I agree that there could be a benefit in pancreatic SBRT, but the interesting UK data suggesting increased heterogeneity and underdosing of certain tumor areas near critical structures at SBRT dosing levels doesn't seem to lead to worse local control could mean that adapting dose with SBRT doesn't lead to any benefit over careful planning with OAR avoidance. (I'm trying to find this study and can't at this moment.)
Also, I have to push back some on the MIRAGE data. It showed margin reduction leads to decreased side effects as we would expect. Saying the trial supports MRI guidance is too much of a leap for me. If the trial had used identical margins and demonstrated improved outcomes with MRI guidance, that would be one thing. However, one cannot even say that MRI should allow for increased margin reduction compared with CT-based IGRT due to inherent distortion present in MR imaging. Maybe MR technology has improved since 2016, but I don't think we've been able to eliminate the distortion problem.
Last edited:
Agree
Eh, don’t tell that to UTSW and a few other places
Semi guaranteed… Medicare guaranteed for sure
Agree hope is not a strategy (horrible strategy in chess eg)
Bingo
And most payors don’t have a 77301 units limit
They will sell a lot of those in that case. 301 is a beefy charge
Last edited:
The beefiest 77xxx rad onc code
Why do you rob banks? Because that’s where the money is!
Complete waste of time may be an overstatement but there is no denying the motivation behind these things. I worked for someone who hosted site visits and I had to sit through many of them. The company reps talk a lot about pro forma estimates and the most common reason people admitted into looking into these things was to have something unique to offer to be more competitive in their market.
I think the protons comparisons are quite valid. Do they have a place at high volume referral centers with the right patient mix? Yes. But I can confidently say that most adaptive machines being filled up with...prostates.
What do you think is the right patient mix to make adaptive worth the investment?
I think that any radonc department should be able to use daily CBCT imaging to determine whether there has been enough of a change in volumes (either patient size due to wt loss, significant tumor shrinkage, etc) to warrant replanning, then resim that day, replan that day/the next morning, and start the new plan the next day. I have a hard time believing adapting more than that will have any benefit.
Weight loss and tumor shrinkage are two reasons to adapt, but in my opinion they are hardly the most common reason.
Position of the target, esp. in the pelvis, due to bowel / bladder filling is the more important reason, in my opinion.
The rather large margins we use for cervical cancer, to facilitate for the flexion of the uterus are one thing you no longer require if you use adaptive treatment.
The different plans you may need when treating bladder cancer for the different bladder fillings are also not necessary.
Occasionally, I contour bowel loops out of my CTV when treating whole pelvis (in any setting: rectal, cervical, prostate) on a daily basis. It makes sense. You contoured them out when you created your initial CTV, but those loops are highly mobile.
Do we have data that these tighter margins, smaller/more correct volumes, non-irradiated bowel loops actually make any clinical difference? Nope.
Many think too much of the tumor and less about the OARs. Adaptive treatment does not mean you have to recontour the target every day. You may only end up contouring the OARs every day and re-optimize the plan based on their daily position / configuration.
Cervix would be great use for adaptive given the variation we see on daily CBCTs
Unfortunately my group's MRI LINAC is too full with prostate and liver SBRTs to make room for a [gasp] 25fx course!
Unfortunately my group's MRI LINAC is too full with prostate and liver SBRTs to make room for a [gasp] 25fx course!
Advertisement - Members don't see this ad
There are places that are adapting for conventional fractionation anal cancer.
it's about revenue, simple. there is no disagreement about there not being proven benefit to daily adaptive Long course RT. it's an opening in the billing system and people are capturing it. it's going to continue, so you should consider jumping on. many are.
And next thing you know, 301 gets devalued like crazy
90% pancreas?
sdns favorite radonc?
So, depends on what you mean by investment. From an RVU perspective, fill that MF up with prostate SBRTs. At the end of the day, ROI will always come down to volume and complexity and prostate SBRT is undeniably the sweet spot.
Do I think adaptive is necessary for prostate SBRT? Not usually. I treat most of mine on an MRL and boost DILs to 45 Gy and still only adapt maybe 25% of the time (I find the cine imaging far more useful than plan adaptation for most prostate SBRT). Of course, if you recontoured every case, you can almost always find a numerical justification to adapt the plan. But to many peoples point, I’m hard pressed to see the clinical relevance of most of these tweaks. If the target is in the PTV and there are not hot spots over the urethra, I just treat.
I personally think adaptive planning is most likely to have a benefit when you are generating isotoxic plans and your GTV coverage can have marked variation from day to day because of small changes is adjacent critical structures. Pancreas is a great example, especially head lesions. Boosts for cervical patients who can’t do HDR for some reason are another. Abdominal and pelvic retreatments are the other major bucket.
I will concede, we don’t know if getting 90% GTV coverage is better than 70% when dose escalating (and both are probably better than conventional). My experience has convinced me it is (at least for pancreas). My first adaptive machine was a Unity before it had gating. I could get S/I motion under 5 mm but still struggled to get much over 80% GTV coverage for head lesions. I now use a Viewray with excellent BH and rarely get less than 90% coverage and local control has been unquestionably better. I treated 40-50 pancreatic cases with SBRT last year and can’t off hand recall a single local failure or significant GI complication.
But consider the economics: an adaptive prostate SBRT bills the same as an adaptive pancreatic SBRT (around 83 RVUs). One takes 5 minutes of my time and 30 min on the table. The other takes 20 min of my time and 50-60 min on the table. It’s not at all hard to see which one is the economic winner with the highest overall revenue potential.
Using adaptive for conventional fractionation rarely makes sense IMO. With all the tools we have on standard machines, daily variation over 5-7 weeks just doesn’t usually have much potential for clinical relevance. The first thing you learn doing MR guided RT is we are not as precise as we think. Yet patients still by and large do excellent anyway. And daily CBCTs are good enough now if a tumor is shrinking, you can fuse it to your sim imaging and replan for subsequent fractions without having to get additional imaging a decent percentage of the time. I don’t find the “shrinking tumor” argument very compelling.
adapt 25% of the time? If am taken out of clinic to review/delineate volumes and a plan at the ethos, I am adapting 100% of the time.
I think he means that most of the times in prostate SBRT, you do not really need to "adapt" in terms of delineating targets again.
You use the real time imaging to track target / OAR movement, and that's about it.
We had this discussion years ago, when the MIRAGE trial came out. It seems that very little "adaptive" treatment was done in MIRAGE. The investigators tracked targets and OAR, thus being able to shrink margins.
Yea, they're redoing the trial with adaptive now but I'm unsure if we'll see a clinical signal.
What you described is adapting. I’m saying if everything appears subjectively close enough, I don’t contour or optimize anything. It’s no different than looking at a CBCT and deciding the bowel and bladder are within tolerance.
Correct. Mirage didn’t adapt at all. Just used the MRI for visualization and gating to (comfortably) use smaller margins.
"Just used the MRI for visualization and gating to (comfortably) use smaller margins"
Yes but there is no justification for them to do that- in fact, I would argue the opposite.
edit: see below, I am likely incorrect about geometric distortion with current-generation MR imaging.
Last edited:
Shouldn’t you adapt just because ethos does not have 6th degree table.?
That is a good question, but this largely has to do with what you treat on the Ethos.
The Halcyon also did not have a a 6-DOF table for many years (it got one recently, I believe), yet millions of fractions have been delivered on those machines.
(one) reason why it's not really a great SBRT machine. Great prostate/HN VMAT workhorse though, and sigficantly cheaper than a conventional linac.
Last edited:
Advertisement - Members don't see this ad
I linked to a study above, but essentially there is geometric distortion present in all MR imaging which can range from 1-3 mm. It's never been discussed when it comes to MR-guided RT.
edit: see below, I am likely incorrect about geometric distortion with current-generation MR imaging.
Last edited:
It is. This is part of machine QA.
Radiology scanners (as referenced by OTN's post above) are not held to the same standards when it comes to distortion.
Possible MR induced distortions in MR-guided RT are discussed and mitigated constantly. What an odd statement.
