partial brain

[nkmiami]

Have a case presently of inpatient with 4-5 symptomatic brain mets (nsclc) lung all superiorly located (in the brain) and decided to go with 300 x10 with lateral opposed partial brain fields. (organizing and starting stereo takes us about a week and difficult for inpts) Haven't seen much literature on partial brain, and wouldnt mind some input on how others are handling this.

It seems like we often have a binary choice between whole brain and stereo, but whole brain has been far from stellar in recent stage 4 nslc trial/lancet.

---

Members don't see this ad.

 

[RadOncDoc21]

Have a case presently of inpatient with 4-5 symptomatic brain mets (nsclc) lung all superiorly located (in the brain) and decided to go with 300 x10 with lateral opposed partial brain fields. (organizing and starting stereo takes us about a week and difficult for inpts) Haven't seen much literature on partial brain, and wouldnt mind some input on how others are handling this.

It seems like we often have a binary choice between whole brain and stereo, but whole brain has been far from stellar in recent stage 4 nslc trial/lancet.

Why not just use protons?

 

[RSAOaky]

Do you think there is something about this patient's cancer that causes him to preferentially seed the superior portion of his brain and no where else? I don't think so either. The reason for treating the whole brain is to treat the lesions we can't see. I'm not sure if you do a lot of stereotactic treatments but if you do consider how often you get a patient with brain mets on MRI and then get a thin slice MRI for treatment planning and find several more. It happens often. If you're not able to do SRS for some reason then this patient needs whole brain. It will save the future radiation oncologist that treats him a bit of a headache in trying to match fields and will save him a weird haircut from partial brain.

 

[nkmiami]

Obviously, I dont think microscopic disease is localized to just that portion of the brain, just like you know that 300 x10 will not provide long term control of the lesions present on mri, so why worry about disease you cant see (prohylactic), when there is such a low chance of long term control of the disease you can see- and just reserve stereo for areas of progression in and out of the field? (by that logic, we should be combining stereo in the case of multiple mets with whole brain. BTW what separates what you can and cant see in terms of log kill is small:-10 exp 8 cells- not visible even with volumetric mri, 10 exp 9- you can see- the point being, there is always plenty you cant see, but prophylactic radiation doesnt seem to help. )

The issue here is that for most of us without a gamma knife, you logistically cant start stereo within several days, so do we just default in those cases to whole brain? ( and even with a gamma knife- i dont think it reimburses much as inpt, so it is very infrequently used for inpts.)

 

[medgator]

The issue here is that for most of us without a gamma knife, you logistically cant start stereo within several days, so do we just default in those cases to whole brain? ( and even with a gamma knife- i dont think it reimburses much as inpt, so it is very infrequently used for inpts.)

The issue with inpatients boils down to the DRG flat reimbursement for the entire admission. The hospital will eat it big if you charge full GK fees on an inpatient.

In terms of partial brain that isn't srs/focal srt, a lot of people talk about doing it but there are no good guidelines. Nccn and astro really don't discuss 3d partial brain xrt.

Despite where the paradigm is going, it is unclear to me what the benefit of not doing wbrt first is in someone with more than 3-4 mets. Memantine is available and can be used in some situations if it is really a concern. Wbrt also makes sense if you don't think pt will be compliant or has a poor PS

 

[nkmiami]

The issue with inpatients boils down to the DRG flat reimbursement for the entire admission. The hospital will eat it big if you charge full GK fees on an inpatient.

In terms of partial brain that isn't srs/focal srt, a lot of people talk about doing it but there are no good guidelines. Nccn and astro really don't discuss 3d partial brain xrt.

Despite where the paradigm is going, it is unclear to me what the benefit of not doing wbrt first is in someone with more than 3-4 mets. Memantine is available and can be used in some situations if it is really a concern. Wbrt also makes sense if you don't think pt will be compliant or has a poor PS

Fair enough- although my experience with poor performance pts is that they are the ones who dont tolerate whole brain well, and if that recent lancet trial is to be believed, dont benefit from it.

scarbtj had a good experience giving just a few treatments to stabilize the lesion and then performing stereo after a delay- would be nice if he could update us on this...

 

[medgator]

Fair enough- although my experience with poor performance pts is that they are the ones who dont tolerate whole brain well, and if that lancet trial is to be believed, dont benefit from it.

Generally, if I am getting called on an inpatient, it's for symptomatic brain mets which are the CC on admission. Sometimes steroids are enough to address sx, sometimes not. If someone has bad Neuro sx, usually the response is better with xrt+dex vs dex alone. I always give pt option of steroids alone, but the more aggressive ones want the chance at possibly improving Neuro sx and extending survival by some weeks or months. 20/5 is not unreasonable in someone with poor PS and significantly symptomatic disease

 

[evilbooyaa]

If it's truly poor PS, end of the line from systemic therapy perspective (EGFR or ALK mutations?), poor extracranial control, etc. , then why treat it at all? You have QUARTZ to back up your decision on this.

If you are going to treat with standard fx I don't see the rationale of NOT doing whole brain, or HS-whole brain if applicable. I've seen posterior fossa RT done only twice, for disease that was confined to the cerebellum, but I've never seen RT to just the cerebrum (which would certainly all be at risk if you're doing standard treatment). I do not think there is any role for focal brain radiation using 3D technique.

How big are they? Are they symptomatic and that's why patient was admitted or were they found incidentally? Did symptoms (if any) improve with Dex? Why does stereo on an inpatient take 1 week? They probably already have an MRI given the diagnosis, so you don't obligatorily have to repeat that (although if there's a lot of edema there may be some shifting given the dex)

If you're doing Linac, then Sim them, contour same day, 1-3 days for planning. Let dosimetry/physics know it's an urgent treatment. Fractionate if necessary so dosimetry isn't struggling to give you an acceptable plan - even focal 3 fraction RT (I believe you personally have shown your plans showing acceptable dosimetry for multiple lesion with single iso on your linac based system) is going to be better for this patient than 3 x 10.

 

[nkmiami]

If it's truly poor PS, end of the line from systemic therapy perspective (EGFR or ALK mutations?), poor extracranial control, etc. , then why treat it at all? You have QUARTZ to back up your decision on this.

If you are going to treat with standard fx I don't see the rationale of NOT doing whole brain, or HS-whole brain if applicable. I've seen posterior fossa RT done only twice, for disease that was confined to the cerebellum, but I've never seen RT to just the cerebrum (which would certainly all be at risk if you're doing standard treatment). I do not think there is any role for focal brain radiation using 3D technique.

How big are they? Are they symptomatic and that's why patient was admitted or were they found incidentally? Did symptoms (if any) improve with Dex? Why does stereo on an inpatient take 1 week? They probably already have an MRI given the diagnosis, so you don't obligatorily have to repeat that (although if there's a lot of edema there may be some shifting given the dex)

If you're doing Linac, then Sim them, contour same day, 1-3 days for planning. Let dosimetry/physics know it's an urgent treatment. Fractionate if necessary so dosimetry isn't struggling to give you an acceptable plan - even focal 3 fraction RT (I believe you personally have shown your plans showing acceptable dosimetry for multiple lesion with single iso on your linac based system) is going to be better for this patient than 3 x 10.

presently a lot goes into our stereo process- special planning mri with 1.5 t mprage, special registration, special immobilization, iv contrast at sim etc so that it is difficult in 1-3 days, not to mention that hospitals don't like inpatient stereo.
I believe ultimately the process will get more streamlined- for example -new brain lab system can autocontour mets and plan very quickly. Hippocampal sparing whole brain is probably even more time consuming from a planning perspective (unless you are doing lateral opposed fields and just bringing in a collimator to block the hippocampus.)

 

[evilbooyaa]

presently a lot goes into our stereo process- special planning mri with 1.5 t mprage, special registration, special immobilization, iv contrast at sim etc so that it is difficult in 1-3 days, not to mention that hospitals don't like inpatient stereo.
I believe ultimately the process will get more streamlined- for example -new brain lab system can autocontour mets and plan very quickly. Hippocampal sparing whole brain is probably even more time consuming from a planning perspective (unless you are doing lateral opposed fields and just bringing in a collimator to block the hippocampus.)

Lol at that HS-WBRT.

Skip the planning MRI (as long as your diagnostic is 1.5mm slices and not 5mm slices or something). Special registration and immobilization? That can't be put together in one day?

I'm not sure what kind of practice setting you are in, but the stuff to do stereo (especially if you can do frameless with a mask or some sort of bite block immobilization) is laying around the department (maybe not the MRI). If the patient is coming down from the hospital he'll have access so IV contrast shouldn't be an issue.

Brain mets don't take that long to contour manually. It's not like a H&N plan or gyn where you're drawing a bunch of nodal volumes or doing differential CTV expansions.

If the hospital doesn't let you do inpatient stereo, then so be it, and all this talk is for nada. I realize and am cognizant of the fact that what I think of as routine at my institution is not be the case for all places that do stereotactic technique.

In that case I would just do whole brain. I don't see a purpose or role for partial 3D-CRT.

 

[RadOncDoc21]

Inpatient protons?

 

[Burt Radnolds]

I see no problem with partial brain. Your logic is plenty sound and our field is way too dogmatic at times. The reasoning of whole brain, stereo, or nothing does not compute for me. If doing nothing per QUARTZ is supposedly acceptable in a symptomatic patient desiring palliation, then partial brain is completely fine. Yes there is a chance you fail at the margin or in the unirradiated brain, but there are plenty of other options at that point, if you happen to get there.

 

[medgator]

Inpatient protons?

is there some inside joke I am missing out on?

 

[RadOncDoc21]

is there some inside joke I am missing out on?

Is this really a thing people are doing? I thought we were just presenting hypothetical extremes.

 

[thecarbonionangle]

Is this really a thing people are doing? I thought we were just presenting hypothetical extremes.

Ohhh people are totally doing it. If only you knew

 

[RadOncDoc21]

Ohhh people are totally doing it. If only you knew

Lol! My bad, carry on... I guess I was trained the “old school” way. Here I thought SRS to 10 mets was fancy!

 

[nkmiami]

Lol at that HS-WBRT.

Skip the planning MRI (as long as your diagnostic is 1.5mm slices and not 5mm slices or something). Special registration and immobilization? That can't be put together in one day?

I'm not sure what kind of practice setting you are in, but the stuff to do stereo (especially if you can do frameless with a mask or some sort of bite block immobilization) is laying around the department (maybe not the MRI). If the patient is coming down from the hospital he'll have access so IV contrast shouldn't be an issue.

Brain mets don't take that long to contour manually. It's not like a H&N plan or gyn where you're drawing a bunch of nodal volumes or doing differential CTV expansions.

If the hospital doesn't let you do inpatient stereo, then so be it, and all this talk is for nada. I realize and am cognizant of the fact that what I think of as routine at my institution is not be the case for all places that do stereotactic technique.

In that case I would just do whole brain. I don't see a purpose or role for partial 3D-CRT.

I am not in private practice, and for example, if I tried stereo without a specific planning MRI- that would rock the physicists world, piss on their fire hydrant so to speak, even though I am sure the differences are neglibile and can be offset with margin. For better or for worse many instiutions, have a whole "ritual" associated with stereo.

BTW, 3d (more than 2 fields) 5 gy x 5 with -2 cm margins- seems to be tolerated fine in the brain, in a population that has longer survival than mets in this randomized trial:
http://ascopubs.org/doi/10.1200/JCO.2015.62.6606
Reasonable to consider for localized mets?

 

[Palex80]

I often wonder, why people like to resort to the extremes.

Sure, you cannot perform stereotactic treatment for this patient. Fine. I do not care about the reasons.
But why give him WBRT for 4-5 brain mets?

Do a "normal" planning CT scan. No iv contrast, no special mask, use 3-5mm slice thickness if the mets aren't tiny. Then fuse your planning CT with the MRI you have.
Contour the 4-5 mets. Add a 8mm or 1cm margin to them.
Plan for 6 x 5 Gy on the PTVs using a fairly plain VMAT.
Treat. Don't use CBCT, just do a simple lateral portal, it should suffice.

You will be spearing probably something like 80% of the patients brain from doses beyond 15 Gy, if those metastases are not huge and not all over the place.

Sure, 2 lateral fields on the simulator are easier, but is what I propose that much harder? The planning CT without contrast and a simple mask, probably take about the same time as a simulation.
Fusing CT and MRI is done pretty fast with the software we have nowadays, if you are lucky to have a good software package you may only need a couple of clicks. Contouring 4-5 mets takes something like 5 minutes.
Sure, planning a VMAT will takes some time and will bother your dosimetrists. But don't plan to hard, you don't need to use OAR constraints.

I understand that you cannot do stereotactic treatment for this patient, for whatever reasons. But why not add just a tiny bit of more effort to spare him from unnecesary toxicity? And 6 x 5 Gy are more efficient than 10 x 3 Gy. You can actually probably durably contol some of these mets with that dose.


I treated a patient with melanoma in the autumn of 2016 with 40+ brain mets using 30 / 3 WBRT. She went on to receive BRAF inhibitors after WBRT and immunotherapy after that. The disease was mostly contolled until a month ago. She presented with headache, the med oncologist did an MRI which showed 32 brain mets, all of them contast enhancing. He asked me if I could treat. Most of my colleagues didn't want to treat. The patient's PS was poor.
I did a CT using a stereotactic mask, fused CT with MRI, contoured the metastases, added a 3mm to them and sent the volumes to my dosimetrist asking him to do a single isocenter plan with a maximum of 3 arcs to deliver 30 / 3 again.
Mean dose to the brain was roughly 10 Gy. Coverage was good. We treated.
One week after treatment she left the hospital without any headaches and doing a lot better. I hope she may live 2-3 months longer because of treatment, she likes to spend time with her grandchildren.

 

[medgator]

.
But why give him WBRT for 4-5 brain mets?

Because the evidence for treatment of 4-5+ mets is sorely lacking and there is no clear answer or evidence than srs/fsrt > wbrt. A symptomatic poor PS pt can prob tolerate wbrt much easier than fsrt

 

[nkmiami]

I often wonder, why people like to resort to the extremes.

Sure, you cannot perform stereotactic treatment for this patient. Fine. I do not care about the reasons.
But why give him WBRT for 4-5 brain mets?

Do a "normal" planning CT scan. No iv contrast, no special mask, use 3-5mm slice thickness if the mets aren't tiny. Then fuse your planning CT with the MRI you have.
Contour the 4-5 mets. Add a 8mm or 1cm margin to them.
Plan for 6 x 5 Gy on the PTVs using a fairly plain VMAT.
Treat. Don't use CBCT, just do a simple lateral portal, it should suffice.

You will be spearing probably something like 80% of the patients brain from doses beyond 15 Gy, if those metastases are not huge and not all over the place.

Sure, 2 lateral fields on the simulator are easier, but is what I propose that much harder? The planning CT without contrast and a simple mask, probably take about the same time as a simulation.
Fusing CT and MRI is done pretty fast with the software we have nowadays, if you are lucky to have a good software package you may only need a couple of clicks. Contouring 4-5 mets takes something like 5 minutes.
Sure, planning a VMAT will takes some time and will bother your dosimetrists. But don't plan to hard, you don't need to use OAR constraints.

I understand that you cannot do stereotactic treatment for this patient, for whatever reasons. But why not add just a tiny bit of more effort to spare him from unnecesary toxicity? And 6 x 5 Gy are more efficient than 10 x 3 Gy. You can actually probably durably contol some of these mets with that dose.


I treated a patient with melanoma in the autumn of 2016 with 40+ brain mets using 30 / 3 WBRT. She went on to receive BRAF inhibitors after WBRT and immunotherapy after that. The disease was mostly contolled until a month ago. She presented with headache, the med oncologist did an MRI which showed 32 brain mets, all of them contast enhancing. He asked me if I could treat. Most of my colleagues didn't want to treat. The patient's PS was poor.
I did a CT using a stereotactic mask, fused CT with MRI, contoured the metastases, added a 3mm to them and sent the volumes to my dosimetrist asking him to do a single isocenter plan with a maximum of 3 arcs to deliver 30 / 3 again.
Mean dose to the brain was roughly 10 Gy. Coverage was good. We treated.
One week after treatment she left the hospital without any headaches and doing a lot better. I hope she may live 2-3 months longer because of treatment, she likes to spend time with her grandchildren.

Agreed, I dont see why we arent seeing more of this.

 

[Palex80]

Because the evidence for treatment of 4-5+ mets is sorely lacking and there is no clear answer or evidence than srs/fsrt > wbrt. A symptomatic poor PS pt can prob tolerate wbrt much easier than fsrt

You don't need "evidence" in the form of a randomized trial here and I am not comparing WBRT with SRS/SFRT.
You already have evidence than WBRT is not better than steroids, actually: QUARTZ.
So, if you still think that you should treat this patient, then do it with the less possible toxic treatment. Focal RT is less toxic than WBRT. If you can't do SRS/SFRT then do something like I proposed. It's less toxic and equally (propably more) efficient when it comes to the macroscopic metastases.

I am not sure, why you think WBRT is better tolerable than SRS/SFRT in a symptomatic, pooroPS patient. I think it all comes down, to how mich effort you are going to put into it. If you use your best immobilization with a bite-block or sonething for making the SRS/SFRT, use iv contrast for CT and put the patient on the treatment couch for 30 min per fraction doing non-coplanar techniques, sure that's less tolerable than WBRT delivered with 2 fields.
But do you have to do all that? This is not a AVM patient, he has multiple brain mets. Adapt your efforts to what is logical. Can you still call it SRS/SFRT then? Probably not, but it's focal therapy and thus will result in less side effects than WBRT.
There are more gears in your car between the first and the fifth. Don't use only the two extremes.

 

[RadOncDoc21]

You don't need "evidence" in the form of a randomized trial here and I am not comparing WBRT with SRS/SFRT.
You already have evidence than WBRT is not better than steroids, actually: QUARTZ.
So, if you still think that you should treat this patient, then do it with the less possible toxic treatment. Focal RT is less toxic than WBRT. If you can't do SRS/SFRT then do something like I proposed. It's less toxic and equally (propably more) efficient when it comes to the macroscopic metastases.

I am not sure, why you think WBRT is better tolerable than SRS/SFRT in a symptomatic, pooroPS patient. I think it all comes down, to how mich effort you are going to put into it. If you use your best immobilization with a bite-block or sonething for making the SRS/SFRT, use iv contrast for CT and put the patient on the treatment couch for 30 min per fraction doing non-coplanar techniques, sure that's less tolerable than WBRT delivered with 2 fields.
But do you have to do all that? This is not a AVM patient, he has multiple brain mets. Adapt your efforts to what is logical. Can you still call it SRS/SFRT then? Probably not, but it's focal therapy and thus will result in less side effects than WBRT.
There are more gears in your car between the first and the fifth. Don't use only the two extremes.

I get what you're saying and "logically" it makes sense. If you have patients that have done well and believe it is a good technique to use, I would just suggest publishing your data. At the end of the day, it is palliative intent and makes sense.

I've seen physicians practice "their way" and sometimes it can be scary!

 

[medgator]

I get what you're saying and "logically" it makes sense. If you have patients that have done well and believe it is a good technique to use, I would just suggest publishing your data. At the end of the day, it is palliative intent and makes sense.

I've seen physicians practice "their way" and sometimes it can be scary!

I don't the buy wbrt argument is so bad for patients, esp when someone can't lay on the table for a vmat tx.

There is zero data for anything above 3-4+ mets in reality. Wbrt is safe effective and QUICK at palliation

 

[scarbrtj]

The rationale for WBRT is becoming less and less and less. Someone said, "it is unclear to me what the benefit of not doing wbrt first is in someone with more than 3-4 mets..." The benefit is that old saw about toxicity being related to treatment volume. Nkmiami's approach of doing reduced-volume opposed laterals is going to treat about 50-75% less normal brain tissue. That's laudable imho. You can see a difference in how patients feel after two weeks of WBRT versus 2 weeks of IBRT (involved brain RT... look there, I just coined a new acronym). The *only* rationale I see in doing WBRT, which would be perhaps best rephrased as elective normal brain radiotherapy, is to treat subclinical metastases ("The reason for treating the whole brain is to treat the lesions we can't see," said someone else above) or if you had mets "all over the place" so to speak. Name me ANY other palliative RT setting wherein we fret so much over subclinical disease. Yet, somehow we have this lingering antediluvian agita over not treating subclinical, invisible brain mets in patients with multiple, visible NSCLC brain mets e.g. Of course we know the in-brain failure rate is higher in WBRT-less regimens; yet close monitoring and focal RT to recurrences make this problem satisfactorily solvable.

 

[evilbooyaa]

I often wonder, why people like to resort to the extremes.

Sure, you cannot perform stereotactic treatment for this patient. Fine. I do not care about the reasons.
But why give him WBRT for 4-5 brain mets?

Do a "normal" planning CT scan. No iv contrast, no special mask, use 3-5mm slice thickness if the mets aren't tiny. Then fuse your planning CT with the MRI you have.
Contour the 4-5 mets. Add a 8mm or 1cm margin to them.
Plan for 6 x 5 Gy on the PTVs using a fairly plain VMAT.
Treat. Don't use CBCT, just do a simple lateral portal, it should suffice.

You will be spearing probably something like 80% of the patients brain from doses beyond 15 Gy, if those metastases are not huge and not all over the place.

What you described above, to me, is basically FSRT. 6Gy x 5 is a FSRT dose. I would still use CBCT, however. I would also use whatever immobilization the patient can tolerate, including your special mask. Skip IV contrast. I'd do 2mm margins 7-8Gy x 3 given uncertainty, if patient's diagnostic MRI had 1.5 or 2mm slices. Treat with 1 or 2 arcs to minimize time on table.

This will give you more durable control than 30/10 to the currently active lesions.

That is what I was advocating in my previous post. Fractionate so as to not make the dosimetrists pissed. Explain to physics that it'll be OK.