WBRT dose for diffuse subcentimetric BM secondary to melanoma

This forum made possible through the generous support of SDN members, donors, and sponsors. Thank you.

Kroll2013

Full Member
10+ Year Member
Joined
Jan 18, 2013
Messages
152
Reaction score
15
Dear colleagues, mi have a 22 yo old with metastatic melanoma that presented for symptomatic diffuse subcentimetric BM secondary to his melanoma.
What dose of WBRT do you suggest?
pS: He has local skin recurrence as well as secondary bone mets that responded very well after SBRT


Sent from my iPhone using SDN mobile

Members don't see this ad.
 
If focal treatment is not feasible, I'd do 35 Gy/14 fx
 
Members don't see this ad :)
does BRAF change prognosis of brain mets?
 
Not necessarily, Seper.

It does however provide a meaningful therapy option even for brain mets. Since this young patient has disseminated disease and WBRT has a detrimental effect on neurocognition and quality of life (starting with hair loss), one could advocate to give a BRAF-inhibitor first and in case the mets respond (which they will likely do, if there's a driver mutation), then give WBRT only at a later time point. This might "buy" the patient several more months without side-effects of WBRT.
 
Interesting. surely, there must be published data to support treating melanoma brain mets with systemic therapy, omitting XRT?
 
  • Like
Reactions: 1 user
Where are you guys finding doses/constraints for SRS; Im having a lot of trouble finding something good. All the astro related stuff points to the ASTRO site, not NRG and its all taken down.

Also, I rewatched an SRS lecture from ASTRO annual mtng and I was impressed the number of melanoma lesions they were treating with GK/Linac SRS. Even up to 12 up front, sometimes multiple times months apart in a single patient. Im pretty annoyed there aren't good guidelines right now about what is reasonable and what is not. I would want my 20 year old patient to receive what is best too and if that means SRS for 12 lesions up front ok. Cherry picking cases though to show the good ones in presentations is somewhat unfair too.
 
honestly, I doubt treating > 6-7 lesions with SRS is less toxic than WBRT.
 
honestly, I doubt treating > 6-7 lesions with SRS is less toxic than WBRT.

you should see the cases eric chang was showing. I was pretty amazed, and all the panelist agreed too. These guys are treating 10+ pretty regularly even if it means retreats frequently. He shows treatment schedules for these patients that all included resection of a metastasis that are larger which may be causing symptoms (like how would you even know which of the lesions to resect anyway, what if it is necrosis somewhere else leading to symptoms?). So to me its like wtf is going on in the world of brain mets these days, and I have nothing to fall back on to look it up. WBRT vs SRS is such a significant difference too, I don't want to give WBRT to my pts if others are getting SRS in the exact same situation.

Anyone have constraints/doses? Cant find something good. Thanks
 
you should see the cases eric chang was showing. I was pretty amazed, and all the panelist agreed too. These guys are treating 10+ pretty regularly even if it means retreats frequently. He shows treatment schedules for these patients that all included resection of a metastasis that are larger which may be causing symptoms (like how would you even know which of the lesions to resect anyway, what if it is necrosis somewhere else leading to symptoms?). So to me its like wtf is going on in the world of brain mets these days, and I have nothing to fall back on to look it up. WBRT vs SRS is such a significant difference too, I don't want to give WBRT to my pts if others are getting SRS in the exact same situation.

Anyone have constraints/doses? Cant find something good. Thanks

With multi-lesion SRS, there aren't any good guidelines AFAIK, but at my institution the V12 per lesion should ideally be less than 5-10cc. That's the data for single fraction, but they try to adopt it as well for fractionated (7Gy x 3 qD) as well when possible.

5Gy x 5 to multiple lesions may not work as well, especially in something resistant like melanoma, but obviously carries the lowest risk of SRN.
 
With multi-lesion SRS, there aren't any good guidelines AFAIK, but at my institution the V12 per lesion should ideally be less than 5-10cc. That's the data for single fraction, but they try to adopt it as well for fractionated (7Gy x 3 qD) as well when possible.

5Gy x 5 to multiple lesions may not work as well, especially in something resistant like melanoma, but obviously carries the lowest risk of SRN.

how about single dose brainstem tolerance and volumes; 12 Gy? Conformality index for single fraction Linac SRS etc...stuff like that?
 
Members don't see this ad :)
how about single dose brainstem tolerance and volumes; 12 Gy? Conformality index for single fraction Linac SRS etc...stuff like that?

I mean we try to push brainstem dose further down as possible, something like V10 < 5cc. If it's an issue we fractionate to 5x5 (don't have a ton of experience with melanoma specifically). CI for multi lesion is never as clean. I would personally feel strongly against single fraction regardless of size, especially if you're treating 5+ lesions. Don't have any data to back that up besides the slow transition in most people who don't have a Gamma-Knife slowly transitioning all but the most minute brain mets to a fractionated SRS model.

I know one of the attendings had a melanoma patient who had like 15 lesions or something (and had previously received treatment to another 10+) and gave 5x5 with single iso to all of them. Probably not dosimetrically the cleanest but he seemed to tolerate OK. I mean we'll see about radionecrosis in the future but if you don't do something the intracranial disease will kill them anyways.
 
I hear what you're saying, but I bet you that QOL and cognitive function of GK X 8 lesions is not superior to WBRT at 3 months follow up.
WBRT gives you a fraction of cost and sorrow associated with Gamma Knife X 8.

you should see the cases eric chang was showing. I was pretty amazed, and all the panelist agreed too. These guys are treating 10+ pretty regularly even if it means retreats frequently. He shows treatment schedules for these patients that all included resection of a metastasis that are larger which may be causing symptoms (like how would you even know which of the lesions to resect anyway, what if it is necrosis somewhere else leading to symptoms?). So to me its like wtf is going on in the world of brain mets these days, and I have nothing to fall back on to look it up. WBRT vs SRS is such a significant difference too, I don't want to give WBRT to my pts if others are getting SRS in the exact same situation.

Anyone have constraints/doses? Cant find something good. Thanks
 
I hear what you're saying, but I bet you that QOL and cognitive function of GK X 8 lesions is not superior to WBRT at 3 months follow up.
WBRT gives you a fraction of cost and sorrow associated with Gamma Knife X 8.

That's why you do Linac-based. Having had patients who have gone through both, people are a lot more OK with fractionated linac-based SRS (even if it is with a frame, given how brief the process is) compared to a full day of gamme knife.
 
I would disagree with that as well. LINAC based SRS X 8 probably has integral brain dose >> WBRT.
I've never seen a composite plan of LINAC SRS X 8 lesions but I bet it looks narly.


That's why you do Linac-based. Having had patients who have gone through both, people are a lot more OK with fractionated linac-based SRS (even if it is with a frame, given how brief the process is) compared to a full day of gamme knife.
 
Linac based srs with a single isocenter to multiple metastases can match or exceed gamma knife metrics. UAB and Jeff have good papers on this and Varian has a webinar- i posted the link on an earlier forum. We adopted the technique last year and it works as advertised. You just have to be willing to tolerate gamma knife like hot spots if you want those dose metrics. I have no problem getting the 50% to be within 5 mm of the PTV with multilple mets 1-2 cm in size. Recently did a case with 13.
 
Wow, you actually have a composite plan with 13 LINAC SRS treatments? I know HIPAA and such, but I would really love to see some snapshots. It would make a great resident conference..

Linac based srs with a single isocenter to multiple metastases can match or exceed gamma knife metrics. UAB and Jeff have good papers on this and Varian has a webinar- i posted the link on an earlier forum. We adopted the technique last year and it works as advertised. You just have to be willing to tolerate gamma knife like hot spots if you want those dose metrics. I have no problem getting the 50% to be within 5 mm of the PTV with multilple mets 1-2 cm in size. Recently did a case with 13.
 
That's why you do Linac-based. Having had patients who have gone through both, people are a lot more OK with fractionated linac-based SRS (even if it is with a frame, given how brief the process is) compared to a full day of gamme knife.

Perfexion allows you to treat 10+ mets very quickly depending on source strength, prob 1-2 hours max, it would be an obnoxiously long day trying to do that on any GK before that model
 
Wow, you actually have a composite plan with 13 LINAC SRS treatments? I know HIPAA and such, but I would really love to see some snapshots. It would make a great resident conference..

We had a patient who we are treating with 8 lesions. Reviewing his Sum Plan DVH, Brain V12 is 24cc, V10 is 38.8cc. This is with multiple isocenters and various fractionations depending on lesion size (some are 18 x 1 while others are 7 x 3).

Perfexion allows you to treat 10+ mets very quickly depending on source strength, prob 1-2 hours max, it would be an obnoxiously long day trying to do that on any GK before that model

Is that something for Gamma Knife? I think the painful part about GK (from what I understand, as I've never personally used GK before) is that you get the whole thing affixed, then get your scans, then the plan is made, similar to brachy cases, so there's a lot of waiting around time for the plan to be finalized.

Our machines can treat single iso, 10-15 lesions in about 30-45 minutes as long as the vertical distance of all the lesions isn't too bad. And that's with a pre-made plan.
 
I think with well planned VMAT single iso (there is definitely a learning curve, but lots of publications out there outlining how to plan this well) you can get pretty close or equal to gamma knife plans. You can potentially have slightly larger targets if you need margin to account for fusion error/lack of frame, but whether this is clinically significant is difficult to know. Definitely need to watch rotation if a very small target is at a far distance from iso, but big rotation is rare if you have a well fitting mask and a good therapists/physician understanding about SRS/imaging/alignment. If you're worried about size you can fractionate on the linac.

Good article outlining Gamma knife vs. VMAT SRS: https://www.ncbi.nlm.nih.gov/pubmed/24871143

I have the luxury of a Gamma Knife Perfexion and VMAT/TrueBeam technology. I favor GK for trigeminal neuralgia, acoustics, small meningiomas, and small brain mets. For larger brain mets, some post-op cavities, some larger meningiomas, I often favor fractionated linac radiosurgery (something like 9 Gy X 3 or 6 Gy X5 for 5 Gy X5 depending upon situation).
 
  • Like
Reactions: 1 user
Is that something for Gamma Knife? I think the painful part about GK (from what I understand, as I've never personally used GK before) is that you get the whole thing affixed, then get your scans, then the plan is made, similar to brachy cases, so there's a lot of waiting around time for the plan to be finalized.

Our machines can treat single iso, 10-15 lesions in about 30-45 minutes as long as the vertical distance of all the lesions isn't too bad. And that's with a pre-made plan.
Yes, the older versions of gamma knife required manual changes in the collimator size, there's none of that required in perfexion but yes you still have to fix the frame, do the mri and plan before treating. GK works really well imo for treating a lot of srs cases in higher volume centers.
 
V10 of 38.8 cc is pretty good for 8 lesions, agree. Are you saying this is a single treatment using one isocenter on a Varian? Again, if you'd kindly share some sum plan screenshots for education purposes, I'd massively appreciate...

We had a patient who we are treating with 8 lesions. Reviewing his Sum Plan DVH, Brain V12 is 24cc, V10 is 38.8cc. This is with multiple isocenters and various fractionations depending on lesion size (some are 18 x 1 while others are 7 x 3).



Is that something for Gamma Knife? I think the painful part about GK (from what I understand, as I've never personally used GK before) is that you get the whole thing affixed, then get your scans, then the plan is made, similar to brachy cases, so there's a lot of waiting around time for the plan to be finalized.

Our machines can treat single iso, 10-15 lesions in about 30-45 minutes as long as the vertical distance of all the lesions isn't too bad. And that's with a pre-made plan.
 
17(2).png
17.png
17 lesions. Pt 2 years s/p whole brain for egfr+ nslc. 17 lesions- 8 Gy x 3 on Trilogy. 2mm margin gtv to ptv. Treatment time 15 minutes, frameless. Planning time 3-4 hours. single isocenter, 4 arcs.
 

Attachments

  • 17(2).png
    17(2).png
    503.9 KB · Views: 71
Last edited:
  • Like
Reactions: 1 user
V10 of 38.8 cc is pretty good for 8 lesions, agree. Are you saying this is a single treatment using one isocenter on a Varian? Again, if you'd kindly share some sum plan screenshots for education purposes, I'd massively appreciate...

For this patient it's multi isocenter. Most of the single isocenter plans end up going on TomoTherapy; just institutional preference AFAIK. We rarely do single isocenter on our Varian machine, mainly when the sup-inf distance of the treatments is too large and treatment times for TomoTherapy are breaking 45 minutes.

I'll see what I can do about pictures but nkmiami is probably giving you more of what you want given that their plan is single iso.
 
Yeah can't argue with this, agree, looks better than 30 Gy/10 WBRT. Thank you.
 
  • Like
Reactions: 1 user
I would disagree with that as well. LINAC based SRS X 8 probably has integral brain dose >> WBRT.
I've never seen a composite plan of LINAC SRS X 8 lesions but I bet it looks narly.

Yeah man Linac SRS X8 integral brain dose is actually << WBRT... Depends on location of lesions but mean brain dose can be limbo low... With cyberknife you can have calculated dose of 0 Gy to sensitive structures like eyes, lenses, middle ear, cochlea, etc which get dosed with whole brain.
 
  • Like
Reactions: 1 user
Proper way to endorse a commercial product: Accuray CyberKnife®
 
If you do not have a couch that can compensate for rotation errors, you may miss the target when using a single isocenter for treatment of multiple brain mets.
If you do however habe a Hexapod (or whatever) couch-overlay, then sure, go for it!

We unfortunately do not have a product like this and thus stick to one isocenter pro lesion (we may do one for two, if they are very near each other).
Most brain mets PTVs are more or less spherical, so rotational errors are not a big problem if the isocenter is in the lesion. However if you put the isocenter somewhere between the lesions and have a rotational error you may miss part of the target, depending on how small your PTV-margin is, :)
 
Rotational errors certainly come into play, which is why we use margin and fractionation. I try to repeatedly reposition patients until they are under 1.5-2 degrees with repeat setups and usually dont have a problem.

Having been doing this for a while, i really think the biggest source of error is the fusion (assuming you dont have a good fusion software, and I have been extremely unhappy with eclipse.)
I was able to start using IV contrast last year as part of CT simulations, and it was eye opening how often our MRI fusions were meaningfully off.
 
  • Like
Reactions: 1 users
Rotational errors certainly come into play, which is why we use margin and fractionation. I try to repeatedly reposition patients until they are under 1.5-2 degrees with repeat setups and usually dont have a problem.

Having been doing this for a while, i really think the biggest source of error is the fusion (assuming you dont have a good fusion software, and I have been extremely unhappy with eclipse.)
I was able to start using IV contrast last year as part of CT simulations, and it was eye opening how often our MRI fusions were meaningfully off.

Completely agree with this (ie fusion error being the biggest issue). Lack of fusion and rotational error is one reason I like the Gamma Knife for some of these cases. But I think you get very clinically meaningful improvements with ability to fractionate some cases.

Fusion error is the bigger issue in my opinion. Rotational errors more than 3 degrees are rare, and 1-2 degrees sometimes don't have a huge impact on dose distrubition (depends on distance from target and target size - http://astro2011.abstractsnet.com/acover.wcs?entryid=011534).

For small targets far from iso, I too add margin. Some institutions rarely add margin, so it's institution dependent. You can account for errors with margin, more rigid frame, or a couch that corrects for rotation. It's very difficult however to account for fusion error.
 
By fusion error, do you mean on planning CT to pre-Tx MRI?

You can verify that yourself before you start contouring on the MRI using the eyes and multiple landmarks along multiple levels sup to inf. Not sure what I'm missing here. If fusion is off, fix it before you plan the SRS.

Sometimes if target is extra far from iso (not sure what the cut-off is/was), we'll add one additional mm radially.
 
Thanks for confirming - I thought I was alone here. We have the neurosurgeon and neuroradiologist involved in these cases, and like myself, they almost never used to have issues with a fusion. When we started using IV contrast for the CT simulation, it was really shocking in some cases. Simply put, the fusion component of eclipse sucks, but it is all we have. Other fusion software, such as brain lab and mimvista, and the stryker product that neurosurgeons use in the operating room seem to rarely require adjustments to the automatic alignment and seem to be idiot proof.
 
Fusion error= planning CT to pre tx MRI.
 
By fusion error, do you mean on planning CT to pre-Tx MRI?

You can verify that yourself before you start contouring on the MRI using the eyes and multiple landmarks along multiple levels sup to inf. Not sure what I'm missing here. If fusion is off, fix it before you plan the SRS.

Sometimes if target is extra far from iso (not sure what the cut-off is/was), we'll add one additional mm radially.

IV contrast at time of CT sim helps, but like miami is saying, I too think Eclipse auto-fusion software can be underwhelming, though it can be micro adjusted manually like you're saying... but small fusion error can happen, especially when treating mets that are very small like 3-4 mm. In the rare cases I'd use fractionation for this (ie treating one large met and another smaller met at the same time), I think some of the Velocity software can help a little.

You can get fusion very close, but not perfect some times. When mets are small and far from iso, I just feel more comfortable adding 1-2 mm of margin to account for possible fusion error. It does dramatically increase the volume of PTV (adding 1-2 mm of margin to a 3mm diameter met is a large volumetric increase), but I think clinically it has little if any consequence.
 
Fusion error - basically what I am saying is that you may think the fusion is correct, but it is not. Aligning 2 non discrete structures on differing imaging modalities to each other, like the eye, is open to interpretation and rotational errors can be very difficult to detect. You are not aligning fiducials to each other.
Apparently ,the image overlay function itself can foster perceptual errors with tracking that are exacerbated by the "slider function." of the images.

I very briefly used the brain lab software, and the software seemed to give an automated perfect match each time. I am also told by my physicist that the fusion on cyberknife is excellent as well and rarely require user adjustment.
 
Last edited:
Perfexion allows you to treat 10+ mets very quickly depending on source strength, prob 1-2 hours max, it would be an obnoxiously long day trying to do that on any GK before that model

This is what I do. I'd like to upgrade to frameless (Icon or linac based), but still stuck with frames for the time being.

Tiny mets are about 15-25 minutes each with my current source strength. So total treatment time for 10 mets is at least 200 minutes, though anything larger with multiple shots or need for blocking adds time. So yeah, you're looking at about 3.5-4 hours in the real world. I do this in selected patients, maybe 1-2 a month. I use light anesthesia, so they tolerate the whole thing just fine.

Patients are convinced to do this compared to whole brain when I tell them no hair loss and less/no risk of neurocognitive side effects. Whole brain integral dose is on the order of 2-3 Gy with this approach, which I think has a trivial risk for neurocognitive side effects.

Knock on wood I almost never see symptomatic RT necrosis both in training and in practice.
 
Fusion error - basically what I am saying is that you may think the fusion is correct, but it is not. Aligning 2 non discrete structures on differing imaging modalities to each other, like the eye, is open to interpretation and rotational errors can be very difficult to detect. You are not aligning fiducials to each other.
Apparently ,the image overlay function itself can foster perceptual errors with tracking that are exacerbated by the "slider function." of the images.

I very briefly used the brain lab software, and the software seemed to give an automated perfect match each time. I am also told by my physicist that the fusion on cyberknife is excellent as well and rarely require user adjustment.

I generally evaluate top and bottom of eye as well as brainstem and ventricles, and edge of apex of skull superiorly to feel confident in a fusion. I guess I haven't seen IV Con at CT Sim to determine if that's actually the case.

My experience may be different as at my institution we routinely get a short planning MRI on day of CT Sim for SRS cases if at all possible (especially if MRI is greater than 1 month old). I understand that may not be possible in the real world, and thus even with manual correction of an auto-fusion, there could be differences. Might be worth doing IV Con on day of sim. For us, that would mean doing both CT and MRI contrasts in the same day, so we generally skip the CT contrast.
 
Palex, Brain Lab and others manufacture couch mounts which can rotate the head for SRS without need for costly investment in a robotic couch.

Is that so? I was always under the impression that you need the Hexapod for Elekta products.
 
I recently looked at the brain lab product and was very impressed especially by the software, wanted to buy it but got over ruled.
 
I recently looked at the brain lab product and was very impressed especially by the software, wanted to buy it but got over ruled.

BrainLab has been in the brain MR/CT fusion game for a long time and got a "head start" on everyone. *snare drum* They were doing deformable before anyone.
Who taught you to do such fabulous planning NKMiami?!?! That looks very good :)
I have made peace with Eclipse fusion. Keep in mind that MRI is susceptible to substantial geometric distortion due to perturbations in magnetic field uniformity. Get with your radiologists and/or the MRI vendor rep to make sure your MRI software (on the radiology side obviously) is applying the latest & greatest algorithms to correct for spatial deformity and that your sequence acquisitions are also designed for success. Try to find a real MRI diagnostic physics nerd even. Also try deformable registration in Eclipse. At the end of the day, as you say, this is the main source of error in SRS and persists no matter how much you spend on your fancy proton, Gamma Knife, TruBeam, etc., system.
 
If you do not have a couch that can compensate for rotation errors, you may miss the target when using a single isocenter for treatment of multiple brain mets.
If you do however habe a Hexapod (or whatever) couch-overlay, then sure, go for it!

We unfortunately do not have a product like this and thus stick to one isocenter pro lesion (we may do one for two, if they are very near each other).
Most brain mets PTVs are more or less spherical, so rotational errors are not a big problem if the isocenter is in the lesion. However if you put the isocenter somewhere between the lesions and have a rotational error you may miss part of the target, depending on how small your PTV-margin is, :)

Not to sound dumb, but there is no rotational error if you don't have hardly any rotational error at setup.
 
BrainLab has been in the brain MR/CT fusion game for a long time and got a "head start" on everyone. *snare drum* They were doing deformable before anyone.
Who taught you to do such fabulous planning NKMiami?!?! That looks very good :)
I have made peace with Eclipse fusion. Keep in mind that MRI is susceptible to substantial geometric distortion due to perturbations in magnetic field uniformity. Get with your radiologists and/or the MRI vendor rep to make sure your MRI software (on the radiology side obviously) is applying the latest & greatest algorithms to correct for spatial deformity and that your sequence acquisitions are also designed for success. Try to find a real MRI diagnostic physics nerd even. Also try deformable registration in Eclipse. At the end of the day, as you say, this is the main source of error in SRS and persists no matter how much you spend on your fancy proton, Gamma Knife, TruBeam, etc., system.

Thanks- learned a lot from you guys.
Does eclipse have deformable registration, or do you have to buy velocity? I feel MRI field error distortion may be a bigger problem now that we are treating many lesions with a single isocenter,
 
Last edited:
I recently looked at the brain lab product and was very impressed especially by the software, wanted to buy it but got over ruled.
BrainLab has been in the brain MR/CT fusion game for a long time and got a "head start" on everyone. *snare drum* They were doing deformable before anyone.
Who taught you to do such fabulous planning NKMiami?!?! That looks very good :)
I have made peace with Eclipse fusion. Keep in mind that MRI is susceptible to substantial geometric distortion due to perturbations in magnetic field uniformity. Get with your radiologists and/or the MRI vendor rep to make sure your MRI software (on the radiology side obviously) is applying the latest & greatest algorithms to correct for spatial deformity and that your sequence acquisitions are also designed for success. Try to find a real RI diagnostic physics nerd even. Also try deformable registration in Eclipse. At the end of the day, as you say, this is the main source of error in SRS and persists no matter how much you spend on your fancy proton, Gamma Knife, TruBeam, etc., system.

This a helpful confirmatory reference - 2 mm error, 1 std. And these were facilities taking part in a pediatric brain protocol! Does this mean that 5% of the time someone was making a 4 mm or more error on stereotactic xrt for kiddies?
( fusion is a systematic error- the worst kind- not random, like setup)
 
Last edited:
Not to sound dumb, but there is no rotational error if you don't have hardly any rotational error at setup.
Not to sound dumb, but I was merely pointing to a problem when it comes to treating more than one lesions and placing the isocenter not in the PTV...
 
Not to sound dumb, but I was merely pointing to a problem when it comes to treating more than one lesions and placing the isocenter not in the PTV...
My back of the envelope calcs say that if the iso is 5 inches or less from the target and there is less than 1 degree rotational error (measured), there will be ~2mm or less of targeting error.
****
EDIT: I'm going to amend myself. I don't see a reason any iso should ever be up to 5 inches away from a target. In the case of multicentric disease, or even only two lesions (one in the very front frontal lobe and very back of the occipital lobe), the iso should be ~4 inches or less from any target (unless you're a bad iso placer). Because the average human head (weighs 10 pounds!) has a circumference of 23 inches or less, it approximates a sphere of about 8 inches or less in diameter. The center of a sphere will be the best iso spot for numerous targets within a sphere, and the center will be 4 inches or less therefore from all possible points within a sphere. If the target is on the sphere surface (worst case scenario, ie farthest point from the sphere center/iso) and it's perfectly aligned, and you rotate the sphere one degree, the equator of the sphere has length 8*3.14=25 inches or 635 mm, and rotating one degree means you'll be 635/360=1.8 mm (maximally) off in target accuracy.
 
Last edited:
Speaking of treatment time with GK, does anyone treat to a higher isodose line (80%) to shorten treatment time when dealing with high number of mets? Obviously this will give less heterogeneity inside your PTV and will result in less sharp dose falloff but can save quite a bit of time compared to treating everything to the 50% IDL.
 
Speaking of treatment time with GK, does anyone treat to a higher isodose line (80%) to shorten treatment time when dealing with high number of mets? Obviously this will give less heterogeneity inside your PTV and will result in less sharp dose falloff but can save quite a bit of time compared to treating everything to the 50% IDL.

I dont see how that is ok- wouldnt you be altering your conformality as well? When I used to treat with gamma knife (>7 years ago), I did not always prescribe to the 50%, - sometimes another isodose line would cover the target. You should prescribe to what isodose line best covers the target- this typically is around the 50%, but not always.
 
Last edited:
Top