SRS margins for brain mets

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

Palex80

RAD ON
15+ Year Member
Joined
Dec 17, 2007
Messages
3,384
Reaction score
4,390
So...
I had a discussion with a colleague about margins for SRS and how maybe we should treat according to histology of the primary, so I was wondering what you guys think...

Do you use a GTV-CTV margin for SRS? Some do, some don't.
What kind of GTV-PTV margin (if you dont delineate a CTV) do you use?

I've seen all kind of different approaches for this.
We are treating LINAC-based with CBCT.

Members don't see this ad.
 
If using frameless SRS, I always add 3mm for setup error.
 
Members don't see this ad :)
I use both a Gamma Knife (zero margin, GTV = PTV; for cavity cases I contour a bit generously, so effectively similar to the GTV plus 2mm concept published above) and linac frameless based techniques.

One thing that is difficult to account for in linac systems and publications is fusion error. On the linac I usually use 2 mm margin from GTV to PTV. I don't do a GTV to CTV expansion. I feel like that covers me for fusion error and set up errors. If using a multi-target single isocenter linac based technique, then for a very small met far away from the isocenter then I may use 3mm.

If you're really concerned about margins/set up, it is possible to give IV contrast before the CBCT. we did that rarely in training (I don't do it in practice now) and it worked well.
 
If you are giving single fraction SRS, adding margin increases toxicity significantly per study (it may have been randomized) from Duke last year or so. For fractionated SRS, I also use 3mm margins. Be careful with the fusion- this is an overlooked source of error.
Eclipse auto fusion is crap. There can be significant intruder variability in fusion. If you are lucky enough to have brain lab or mimvista software, the auto fusion is usually dead on.

I try to do sims with IV contrast, just to make sure lesions are fusing dead on. In miami, there can be delays between MRI and CT sim further introducing error, and sometimes outside MRI do not have mprage (volumetric scans) sequence, and we can have trouble from insurance getting and additional mri for srs planning.

Lastly, if you are using one isocenter to treat many lesions, rotational corrections can become important.
 
  • Like
Reactions: 1 user
Tx volume relates to toxicity risk in SRS. Error is everywhere: fusion, contouring, setup, movement, IGRT match, etc. It's a little weird to me re: 3mm PTV margins. A 1cm diameter tumor has a volume of ~0.5 cc. A 1cm tumor with a 0,3cm margin becomes a 1.6cm diameter target with a ~2cm volume. So your 3mm margins QUADRUPLE the treatment volume. Seems like doing a whole-house fumigation to kill a single cockroach.

The proof's in the clinical pudding. I do frameless SRS. Have used 1mm expansion margins for years and years (PTV margins, not block or cone margins... and this PTV always gets its dose minimum as the prescribed Gy dose, usually Rx'ing to the ~75-85% where 100% is the global max dose in the plan). Last LC analysis we did was 94% actuarial LC at 1 year which is A-OK IMHO.
 
There are so many factors at play that make it difficult to make these comparisons. Ok, you use 1mm vs. 2mm vs. 3mm, but what is your rx dose? How rapid is your lateral fall off? The lateral fall off in srs plans in the high dose area is what like 10% per mm? 22gy to 1mm may be the same as 20 gy to 2mm, etc. In the end, we may all be doing the same thing, and I think your safest bet is to do what you were trained to do.
 
  • Like
Reactions: 1 user
Members don't see this ad :)
I don't do a CTV expansion. My PTV expansion is 2-3 mm depending on which particular system I'm using.

This - depends on the stabilization device that is being used. Pretty much everyone gets FSRT instead of SRS anyways here, so a small margin doesn't seem to matter as bad with 6 x 4 or 6 x 5 compared to 24 x 1.
 
Pretty much everyone gets FSRT instead of SRS anyways here
Sounds like you have the capacity to both single fraction or FSRT--have you had good luck selling your NSGs on fractionated approach, and if so how/when do you incorporate them in planning? At contour, at plan review, or cut them out?
 
Sounds like you have the capacity to both single fraction or FSRT--have you had good luck selling your NSGs on fractionated approach, and if so how/when do you incorporate them in planning? At contour, at plan review, or cut them out?

Yeah, we do. In general we seem to lean towards FSRT.

At contour, if it's an intact met, they usually quickly eyeball it just to say 'sure, go ahead'. If it's post-op we invite them to come down, review the contour areas and have a discussion on any areas of concern they have.
At plan review (and even dose selection) they warn us about any problematic areas in terms of brain anatomy so we know where to avoid hot spots but otherwise they don't significantly grumble over dose selection.

My n is low (<5) currently, but haven't had significant push back as of yet doing fractionated. Maybe it'll be different when I'm on a more neuro-related service. I see other attendings' patients getting Talon screws (placed by NSG) more often for what it's worth.
 
CNS Oncology
July 2016 ,Vol. 5, No. 3, Pages 111-113 , DOI 10.2217/cns-2016-0025
(doi:10.2217/cns-2016-0025)



Editorial
free.gif

Why hypofractionate stereotactic radiosurgery for brain metastases?
Sten Myrehaug*,1, Hany Soliman1, Chia-Lin Tseng1, Mark Ruschin2, David Larson3 &Arjun Sahgal1

http://www.futuremedicine.com/doi/a...sref.org&rfr_dat=cr_pub=www.ncbi.nlm.nih.gov&
 
  • Like
Reactions: 1 user
I don't think this totally fits here but wanted to ask about a case.

68 year old guy who had a stage I nsclc lobectomy 3.5 years ago presented with some confusion. Found to have a 2.5 cm cerebellar lesion w edema. PET doesn't show any distant uptake, nothing in the lung that is concerning either. What would be your management here?
 
Brain surgery + SRT/SRS to cavity
 
  • Like
Reactions: 1 user
I don't think this totally fits here but wanted to ask about a case.

68 year old guy who had a stage I nsclc lobectomy 3.5 years ago presented with some confusion. Found to have a 2.5 cm cerebellar lesion w edema. PET doesn't show any distant uptake, nothing in the lung that is concerning either. What would be your management here?

Couple factors - Medical co-morbidities? Cerebellum can be a hard place for a GTR. Is it resectable?

I'd be more inclined for fractionated SRS alone (7Gy x 3 or something along those lines). I'm assuming he's on steroids now and his symptoms have improved. If not, then surgery if feasible.

Another thing to consider is pre-op SRS - https://www.ncbi.nlm.nih.gov/pubmed/26528673
 
Sorry guys was in a rush usually would do better w details. He is not resection candidate bc of age and comorbidities, he's 71. Had headaches but stable on steroids. If the primary tx will be RT do you request tissue or just treat? And if treat you do SRS? There is a good amount of edema associated here.
 
Sorry guys was in a rush usually would do better w details. He is not resection candidate bc of age and comorbidities, he's 71. Had headaches but stable on steroids. If the primary tx will be RT do you request tissue or just treat? And if treat you do SRS? There is a good amount of edema associated here.

Tissue is kinda dealer's choice, IMO. I'd have a discussion about the possibility that without tissue we can't be 100% sure this is cancer, but a NSCLC (which has a propensity for single brain mets) diagnosis within the past 5 years likely gives me the go ahead to treat.

I'm assuming that since his MRI (done at time of worst symptoms), he has likely been on relatively high-dose steroids. We don't treat the edema in SRS anyways. Given that he's had additional scans (PET/CT likely took some time), the edema has likely changed compared to the most recent MRI. What's time frame from previous MRI to now?

I'd ideally do a repeat MRI on a stable steroid dose, for treatment planning only (no need for full range of sequences, just maybe T2 FLAIR and definitely T1 Post-Con). Ideally do it immediately after CT Sim (assuming Linac-based SRS) and within the same immobilization, although you can deal with differences on fusion. Given size, I treat with fractionation either 7 x 3 QOD or 6 x 4 (maybe 6 x 5), but again, dealer's choice on actual dosing.

I don't see WBRT in the US for this lesion in this day and age. He's actually kind of in the inclusion criteria for QUARTZ if he's deemed to be not a SRS candidate.
 
Tissue is kinda dealer's choice, IMO. I'd have a discussion about the possibility that without tissue we can't be 100% sure this is cancer, but a NSCLC (which has a propensity for single brain mets) diagnosis within the past 5 years likely gives me the go ahead to treat.

I'm assuming that since his MRI (done at time of worst symptoms), he has likely been on relatively high-dose steroids. We don't treat the edema in SRS anyways. Given that he's had additional scans (PET/CT likely took some time), the edema has likely changed compared to the most recent MRI. What's time frame from previous MRI to now?

I'd ideally do a repeat MRI on a stable steroid dose, for treatment planning only (no need for full range of sequences, just maybe T2 FLAIR and definitely T1 Post-Con). Ideally do it immediately after CT Sim (assuming Linac-based SRS) and within the same immobilization, although you can deal with differences on fusion. Given size, I treat with fractionation either 7 x 3 QOD or 6 x 4 (maybe 6 x 5), but again, dealer's choice on actual dosing.

I don't see WBRT in the US for this lesion in this day and age. He's actually kind of in the inclusion criteria for QUARTZ if he's deemed to be not a SRS candidate.


Thank you very helpful. People use gamma in this situation too? Are there consensus guidelines now, seems like things changed pretty quick on this subject. Is there anything on tissue that would make you do something different (Assuming obviously if it's not benign/infection etc). Thanks
 
seper's answer is right on. The standard management here is surgery followed by RT (SRS single/fractionated). Of course a bunch of dexmaethasone mixed in throughout.

With the additional information of distant cancer history, no other disease, and unresectable cancer, I'd have to look at the MRI to make a decision here.

The posterior fossa is not a place with a lot of room for edema. If there is a lot of edema, the 4th ventricle could be threatened with resultant hydrocephalus, clinical decline, and death. Even if he is not a surgical candidate, would he be a candidate for a VP shunt in the case that the edema becomes a big problem?

Whether to treat empirically depends on the MRI and a discussion with the patient. In a patient that age without a history of a rheumatologic/neurologic condition like MS who otherwise isn't sick, review with a neuroradiologist can help to see if there is anything else on the differential here. Certainly if it looks like a met, I'd treat the patient without tissue with clear consent that this may not be metastasis. The other issue is that you could have a transient increase in edema after the therapy that could trigger the hydrocephalus scenario I mentioned earlier. I'd have a neurosurgeon on my team aware and already discussed this with the patient.

Then I'd go ahead and deliver single fraction SRS. In my current position, I have a GK so based on size I'd probably go with somewhere in the neighborhood of 18 Gy to 50% IDL to the gross disease without margin.

To add to the original discussion, GammaPlan doesn't even have an option to add a margin. So the stalwarts who contour in GammaPlan never added margins to their cases. The argument is that the penumbra is basically the margin. With GK I do add a margin depending on the case, particularly post-op cavities or where the imaging isn't perfect, and import the contours from another contouring system.
 
In the really edema-ish cases I do a partial brain (tumor+edema+1 cm) fractionated approach to ~15 Gy/6 fx. This usually brings the edema, and the lesion, down pretty nicely on a followup MRI with gentle improvement in neuro symptoms if they exist. Then I SRS it. (Where did I get this? A bit of trial-and-error, and some reverence/hybridization of "old school" approaches & trials.)
 
In the really edema-ish cases I do a partial brain (tumor+edema+1 cm) fractionated approach to ~15 Gy/6 fx. This usually brings the edema, and the lesion, down pretty nicely on a followup MRI with gentle improvement in neuro symptoms if they exist. Then I SRS it. (Where did I get this? A bit of trial-and-error, and some reverence/hybridization of "old school" approaches & trials.)
Interesting.
How long do you wait until the followup MRI?
What dose do you then use for SRS?
 
My issue with partial brain fractionated approaches is what to do in 3-6 months when they end up with more brain mets and you need to consider whole brain. We know SRS at our standard doses before or after WBRT are reasonably safe. We don't know the safety of SRS + fractionated RT + WBRT all in the same place. Symptomatic radiation necrosis in a sensitive area like the posterior fossa would be a big problem.

I suppose you could just radiate the posterior fossa now and then radiate the supratentorial brain later if you needed to. But, that's a tough field junction to perform.
 
  • Like
Reactions: 2 users
My issue with partial brain fractionated approaches is what to do in 3-6 months when they end up with more brain mets and you need to consider whole brain. We know SRS at our standard doses before or after WBRT are reasonably safe. We don't know the safety of SRS + fractionated RT + WBRT all in the same place. Symptomatic radiation necrosis in a sensitive area like the posterior fossa would be a big problem.

I suppose you could just radiate the posterior fossa now and then radiate the supratentorial brain later if you needed to. But, that's a tough field junction to perform.

Fair points raised here.

Like you said, we know that WBRT->SRS is safe, so why shouldn't SRS->WBRT or SFRT->WBRT be safe as well?
In some cases you may have overlapping doses due to multiple SRS and/or SFRT treatments having taken place, so you may have some difficulties in performing WBRT later on, from the safety point of view.
What we generally due is to fuse CTs and trasfer isodoses from older plans to the new CT dataset. If we have a patient with multiple treatments in the past, we may select to place a part of the PTV of the intended WBRT outside the 95% isodose and give only 80% to that.
You can do that quite well with IMRT/VMAT.
Another point is fractionation. You could always opt for 2.5 Gy or 2 Gy per day and not 3 Gy or 4 Gy when performing WBRT in previously irradiated patients. This should decrease the risk of severe complications.
 
Interesting.
How long do you wait until the followup MRI?
What dose do you then use for SRS?

I generally plan for the MRI, which is going to be an SRS planning MRI, anywhere from 1-5 days after 15 Gy is complete. If I were to look back at my SRS doses, I would imagine 95% of them fall in the 18-22 Gy range. Probably 20 Gy is thus my most common Rx.

My issue with partial brain fractionated approaches is what to do in 3-6 months when they end up with more brain mets and you need to consider whole brain. We know SRS at our standard doses before or after WBRT are reasonably safe. We don't know the safety of SRS + fractionated RT + WBRT all in the same place. Symptomatic radiation necrosis in a sensitive area like the posterior fossa would be a big problem.

I suppose you could just radiate the posterior fossa now and then radiate the supratentorial brain later if you needed to. But, that's a tough field junction to perform.

THEORETICALLY... there's some forgiveness of the 15 Gy/6 fx partial brain dose in 6 months. Rhesus monkey data would say 1/2 forgiveness; if so, (0.5*15Gy partial brain)+(30Gy WBRT repeat tx in 12 fx) equals 37.5 Gy sum over 18 total fractions. Recall in RTOG 95-08 the WBRT dose was 37.5 Gy/15 fx. So more chance of CNS problems, technically, in the RTOG 95-08 approach than my approach. Patchell #1 gave WBRT doses of 36/12; Patchell #2, 50.4Gy/28 fx. There was also an Australian WBRT trial of 40/20 BID. All of these would also have higher BED-Gy3 than my approach. And I don't remember mention of radiation necrosis in any of these studies (may have happened, IDK). I say all that to say: I am not worried about radiation necrosis by adding ~15 Gy partial brain before SRS, even given possible future SRS or WBRT retreatments. Obviously would prefer a retreat SRS in a different part of CNS :)

Shaw also at one time presented some data about WBRT re-treatment that looked pretty safe, all things considered.
 
Thank you very helpful. People use gamma in this situation too? Are there consensus guidelines now, seems like things changed pretty quick on this subject. Is there anything on tissue that would make you do something different (Assuming obviously if it's not benign/infection etc). Thanks

I think this is somewhat of a debate topic in SRS nowadays. From a radiobiological perspective, it doesn't make sense to hit the edge of tumor with 18Gy if it's larger (versus a higher dose to the edge if it's smaller) and hope that the LC is going to remain the same. GK-SRS does single fraction because it's too much in terms of logistics to consider fractionating, which isn't nearly as big of an issue with linac based-SRS. I think 2.5cm is kind of on the border (IMO) and something that people with a GK are still hitting with single fraction routinely. I will admit that I'm biased towards Linac-based FSRT as we don't have GammaKnife here.

Don't think there are consensus guidelines for single shot vs fractionated in this setting. Given that it's still an active research topic it'll likely be a few years before we see something consensus. Additional issue that if fractionated SRS is deemed better above a certain size, there's a ton of people with GammaKnife centers that either lose a lot of business and/or may do fractionated GK-SRS.

If it does get biopsied and it's something like a glioma or SCLC, then yes, it'll change your management for sure, but it's a question of how likely this is to be anything else.
 
seper's answer is right on. The standard management here is surgery followed by RT (SRS single/fractionated). Of course a bunch of dexmaethasone mixed in throughout.

With the additional information of distant cancer history, no other disease, and unresectable cancer, I'd have to look at the MRI to make a decision here.

The posterior fossa is not a place with a lot of room for edema. If there is a lot of edema, the 4th ventricle could be threatened with resultant hydrocephalus, clinical decline, and death. Even if he is not a surgical candidate, would he be a candidate for a VP shunt in the case that the edema becomes a big problem?

Whether to treat empirically depends on the MRI and a discussion with the patient. In a patient that age without a history of a rheumatologic/neurologic condition like MS who otherwise isn't sick, review with a neuroradiologist can help to see if there is anything else on the differential here. Certainly if it looks like a met, I'd treat the patient without tissue with clear consent that this may not be metastasis. The other issue is that you could have a transient increase in edema after the therapy that could trigger the hydrocephalus scenario I mentioned earlier. I'd have a neurosurgeon on my team aware and already discussed this with the patient.

Then I'd go ahead and deliver single fraction SRS. In my current position, I have a GK so based on size I'd probably go with somewhere in the neighborhood of 18 Gy to 50% IDL to the gross disease without margin.

To add to the original discussion, GammaPlan doesn't even have an option to add a margin. So the stalwarts who contour in GammaPlan never added margins to their cases. The argument is that the penumbra is basically the margin. With GK I do add a margin depending on the case, particularly post-op cavities or where the imaging isn't perfect, and import the contours from another contouring system.

I agree that if you're going to single fraction SRS this (Linac or GK-based) it would be prudent to have a neurosurgeon aware of the patient in case the edema outpaces any steroids the patient is already on. I still maintain that if there was significant edema at diagnosis (but before steroids) he should go for a repeat MRI, just the post-con T1 and T2 Flair to re-evaluate within a few days of planning, hopefully with him on a stable steroid dose (which shouldn't change throughout treatment).

For your other post, Neuronix, I wouldn't like posterior fossa RT in this situation, especially not for one lesion. Agree about difficult to field match for eventual whole brain RT.

@seper - IMRT is feasible to do whole brain RT in the future even in this setting, but IMO, one of the best ways to not get into overly complicated plans in the future is to plan for the future in the present.

In the really edema-ish cases I do a partial brain (tumor+edema+1 cm) fractionated approach to ~15 Gy/6 fx. This usually brings the edema, and the lesion, down pretty nicely on a followup MRI with gentle improvement in neuro symptoms if they exist. Then I SRS it. (Where did I get this? A bit of trial-and-error, and some reverence/hybridization of "old school" approaches & trials.)

Do you SRS every lesion that gets this 15Gy/6Fx course? I can't imagine that dose and fractionation is eradicating anything but the smallest metastases.

I generally plan for the MRI, which is going to be an SRS planning MRI, anywhere from 1-5 days after 15 Gy is complete. If I were to look back at my SRS doses, I would imagine 95% of them fall in the 18-22 Gy range. Probably 20 Gy is thus my most common Rx.

Are you saying that you treat 15Gy/6Fx, then repeat an MRI 1-5 after that is completed? Or you make two separate plans on an SRS planning MRI and do something along the lines of a cone-down? Your previous post made it sound like you do 15/6Fx, then do a follow-up MRI (1 - 3 months after?) then SRS whatever's left.


THEORETICALLY... there's some forgiveness of the 15 Gy/6 fx partial brain dose in 6 months. Rhesus monkey data would say 1/2 forgiveness; if so, (0.5*15Gy partial brain)+(30Gy WBRT repeat tx in 12 fx) equals 37.5 Gy sum over 18 total fractions. Recall in RTOG 95-08 the WBRT dose was 37.5 Gy/15 fx. So more chance of CNS problems, technically, in the RTOG 95-08 approach than my approach. Patchell #1 gave WBRT doses of 36/12; Patchell #2, 50.4Gy/28 fx. There was also an Australian WBRT trial of 40/20 BID. All of these would also have higher BED-Gy3 than my approach. And I don't remember mention of radiation necrosis in any of these studies (may have happened, IDK). I say all that to say: I am not worried about radiation necrosis by adding ~15 Gy partial brain before SRS, even given possible future SRS or WBRT retreatments. Obviously would prefer a retreat SRS in a different part of CNS :)

Shaw also at one time presented some data about WBRT re-treatment that looked pretty safe, all things considered.

But the discussion here is that you give 15Gy partial brain, followed by SRS (which I'm assuming is 20-22Gy NOT including the 15 you just gave) to a focal area, as your 'first' treatment. What's the safety of WBRT (30/12, 30/15, 30/10, pick your favorite fractionation) on top of that?

I think the reason you're getting a lot of questions is because people know that SRS (20-30Gy) + WBRT is safe. But when you add in an extra 15Gy to that combination (rather than comparing 15Gy partial + WBRT), that's where the concern lies. I agree that to the area that didn't get SRS, you're likely going to be fine. The question would be, what's the increased risk of sequellae in the area that gets hit with the 15Gy 'pretreatment', 20Gy SRS, then 30Gy WBRT?
 
The problem is the V10/V12.
If you dont opt for a very complicated plan with couch rotations, etc (in my personal experience) mets over 2cm often become difficult to hit with 20 Gy on the 80% isodose when using a 1mm GTV-CTV + 1mm CTV-PTV delineation procedure.
Both the V10 and the V12 generally increase in these cases beyond the 12ccm/10ccm barrier, meaning you are risking more brain necrosis.

In patients like these we usually opt for multi-fraction SRS.
Results with the 3x9 Gy schedule to the 80% isodose were recently published by Minniti and looked very good.
https://www.ncbi.nlm.nih.gov/pubmed/27209508
These guys are experts in SRS.

And in terms of couch time, if you weigh in the extra quality assurance you need to perform when doing SRS with lots of couch rotations plus the extra time which is often necessary for these procedures (especially if you are not using a FFF-Linac), then I think the 3 x 9 Gy approach is a very good compromise.
 
I also give 8-9 Gy x 3 with 2 mm margin and calculate the "effective" v12. The bed of 7 gy x 3 is kind of low. I have also been giving contrast at the time of ct simulation as it helps delineate lesions and sometimes alerts to fusion misregistration. I have found that eclipse fusion simply sucks compared with mimvista and brainlab (neither of which I have right now).
 
Do you SRS every lesion that gets this 15Gy/6Fx course? I can't imagine that dose and fractionation is eradicating anything but the smallest metastases.
Oh yeah absolutely. 15 Gy eliminates nothing visible on a scan. But it shrinks almost everything, a little; helps edema a lot, quite often, too.
Are you saying that you treat 15Gy/6Fx, then repeat an MRI 1-5 after that is completed? Or you make two separate plans on an SRS planning MRI and do something along the lines of a cone-down? Your previous post made it sound like you do 15/6Fx, then do a follow-up MRI (1 - 3 months after?) then SRS whatever's left.
Generally my first MRI is not a high-res, thin slice SRS MRI (never is really); ie, the diagnostic MRI that showed the met(s). If it's a patient with a bigger lesion, a lot of edema, some neuro sx's, or a combination thereof, I get straight to some partial brain that day at 2.5 Gy a day planning it to ~15 Gy.... knowing that SRS and SRS planning MRI will come soon. The partial fractionated brain and SRS have extreme temporal propinquity.
but the discussion here is that you give 15Gy partial brain, followed by SRS (which I'm assuming is 20-22Gy NOT including the 15 you just gave) to a focal area, as your 'first' treatment. What's the safety of WBRT (30/12, 30/15, 30/10, pick your favorite fractionation) on top of that?
That's correct, it's "first." WBRT at any reasonable fractionation months later is going to be quite safe IMHO (I would favor 30/12); the 15 Gy months prior to a small-ish portion of brain is essentially clinically inconsequential, the lack of consequence being proportional to the time delta of SRS and salvage WBRT (and inconsequential also especially in light of a relative lack of concern about late effects in a patient with multiple bouts of brain mets). As I mentioned, if you follow my math, (0.5*15)+(30/12)=37.5 Gy equivalent lifetime brain dose with 0.5 correction factor on the 15 Gy after a 6 month interval between salvage 30 Gy WBRT and the initial partial brain/SRS. Did that make sense?
 
That's correct, it's "first." WBRT at any reasonable fractionation months later is going to be quite safe IMHO (I would favor 30/12); the 15 Gy months prior to a small-ish portion of brain is essentially clinically inconsequential, the lack of consequence being proportional to the time delta of SRS and salvage WBRT (and inconsequential also especially in light of a relative lack of concern about late effects in a patient with multiple bouts of brain mets). As I mentioned, if you follow my math, (0.5*15)+(30/12)=37.5 Gy equivalent lifetime brain dose with 0.5 correction factor on the 15 Gy after a 6 month interval between salvage 30 Gy WBRT and the initial partial brain/SRS. Did that make sense?

I understand the first two paragraphs so I left those out.

I suppose what you're saying is that your 3 part regimen should be equivalent to 37.5/15, and therefore adding SRS at the initial treatment would be safe, as salvage 37.5/15 WBRT is generally considered safe after SRS. Thanks for the more in-depth explanation.

I guess my final questions are what is the general time frame for the SRS planning MRI (and subsequent initiation of SRS) after the initial 15Gy? I guess my concern is that you will be seeing some tumor shrinkage and possibly shift of it as the (large volume) edema improves with the 15Gy; the time frame between SRS planning MRI and initiation of SRS treatment then becomes important. Are you fractionating these SRS treatments or going single-fraction?
Is this something you do routinely (or I suppose, uncommonly) or is this more like 1 - 3 cases/year?

I don't mean to give you the nth degree, I'm just genuinely curious about your technique.
 
I understand the first two paragraphs so I left those out.

I suppose what you're saying is that your 3 part regimen should be equivalent to 37.5/15, and therefore adding SRS at the initial treatment would be safe, as salvage 37.5/15 WBRT is generally considered safe after SRS. Thanks for the more in-depth explanation.

I guess my final questions are what is the general time frame for the SRS planning MRI (and subsequent initiation of SRS) after the initial 15Gy? I guess my concern is that you will be seeing some tumor shrinkage and possibly shift of it as the (large volume) edema improves with the 15Gy; the time frame between SRS planning MRI and initiation of SRS treatment then becomes important. Are you fractionating these SRS treatments or going single-fraction?
Is this something you do routinely (or I suppose, uncommonly) or is this more like 1 - 3 cases/year?

I don't mean to give you the nth degree, I'm just genuinely curious about your technique.

My most recent patient, 70yo WF ~6 mos out from chemoRT for NSCLC. Disease-free in chest, body.

12/13/16:
XXXXX came to the hospital today with what appeared to be stroke-like symptoms and a quivering of her right jaw. I had previously noticed a little weakness in her right hand as well. Workup ensued including a CT of the head and MRI of the brain. There looks to be a high left parietal lesion. It’s about 1.5 to 2 cm in size and is ring enhancing and gives off a couple of centimeters of edema but without any significant mass effect in that left parietal region. She is about to be admitted to the hospital and is started on IV Decadron. Right now she is completely mentally alert and is really not having any neurological problems at all. Of course she is recently s/p chemoradiation for a locally advanced lung cancer. She had been s/p neoadjuvant chemotherapy before that. Her disease status appears to be stable at the present time at the very least. As mentioned she’s being admitted here to the hospital and I wanted to see her to get some radiation therapy started to the brain.
Labs/X-rays/Pathology:
FINDINGS:
BRAIN AND VENTRICLES: A 1.2 x 1.3 x 1.5 cm ring-enhancing lesion within the left parietal lobe
demonstrates restricted diffusion with moderate adjacent vasogenic edema compatible with
intra-axial metastasis. No intracranial hemorrhage or mass effect. Ventricles are unremarkable. Diffuse
age-appropriate atrophy. Scattered periventricular and subcortical white matter changes are stable.
Posterior fossa is unremarkable.
ASSESSMENT: Oligometastatic non-small cell lung cancer.
PLAN: Right now I want to start with some fractionated partial brain radiation therapy to bring down the swelling. After a few fractions of this we will come in with stereotactic radiosurgery. The risks, benefits, side effects, logistics and indications were all discussed with her today. We’ll get her treatment started likely tomorrow and simulate today.

12/14 to 12/21, patient received 15 Gy/6 fx using the 12/13 diagnostic MRI to plan. We used a 5-field conformal 3D plan, lesion/edema+1 cm PTV margin. The 50% isovolume was ~400cc's, the brain volume was ~1400cc's.

Due to the holidays, we put off the SRS until a few days later. (Normally I would do it just 1-2 days later.) We got a tx planning MRI on 12/27 and there was about 50% less edema and the lesion was down to ~1.2 cm max diameter. We performed SRS on 12/29, 21 Gy minimal lesional dose @81% line using static beams. Frameless. Actually we used a 2 mm PTV margin and that was the lesion. CTV ~1.1 cc's, PTV ~2.6 cc's, 50% isovolume ~7.6 cc's.
 
Few questions fr the group:

Is there good data that shows worse control with 0 mm PTV margin vs 1 mm? Likewise, good data that shows 1 mm results in more radiation necrosis than 0 mm?

If patient had prior wbrt, what max sigle fraction SRS dose do you accept to brainstem or chiasm?
 
unless you only have a gamma knife, who would give single fraction srs next to the optic chiasm at this point? Historically, we did not really account for wbrt when deciding on stereo doses in the recurrent setting. In the upfront setting, some would decrease the prescription dose by 2 Gy.
 
Is there good data that shows worse control with 0 mm PTV margin vs 1 mm?
Some data, yes. Radiosurgery for brain metastasis: impact of CTV on local control. - PubMed - NCBI

Likewise, good data that shows 1 mm results in more radiation necrosis than 0 mm?
Not any I am aware off, but 3mm cause more RN (no wonder...)
Defining the optimal planning target volume in image-guided stereotactic radiosurgery of brain metastases: results of a randomized trial. - PubMed - NCBI

If patient had prior wbrt, what max sigle fraction SRS dose do you accept to brainstem or chiasm?
I'd fractionate
I'd accept Dmax to chiasm of 3x6 Gy and 3x7Gy to brainstem.
 
Top