SRS for tumor near brainstem

[CUBuffsgrad98]

---

Members don't see this ad.

Anyone doing SRS or fractionated SRS for meningioma that is close to, or touching the brainstem? I have seen a lot of people in the community doing it and am constantly getting neurosurg referrals after they have told the patient they will be getting 1x treatment. I have seen 1 publication from U Pitt on this, but no others to suggest it is safe and should be done. I still defer to going 45-50.4 Gy for a benign disease. Anyone doing it different? Any lit that I am missing that suggests its ok to do a single tx or a hypofx treatment?

 

[Gfunk6]

There is a good section in Rad Onc wikibooks on this topic:

Brainstem lesions[edit]

UCSF; 2008 (1991-2005) PMID 17628747 -- "Gamma knife radiosurgery for brainstem metastases: the UCSF experience." (Kased N, J Neurooncol. 2008 Jan;86(2):195-205.)
Retrospective. 42 pts (44 lesions: 7 midbrain, 31 pons, 6 medulla). Median target volume 0.26 cc; median tumor diameter 0.9 cm. Prescribed dose 16 Gy (50% isodose median).
Brainstem lesion FFP 90% at 6 months, 77% at 1 yr. 4 pts had brainstem complications. Poor brainstem outcome associated with melanoma and RCC histology and brainstem lesion volume ≥ 1 mL.
Conclusion: SRS provided excellent local control with low morbidity

Virginia; 2006 (1989-2005) PMID 17219825 -- "Gamma knife surgery for metastatic brainstem tumors." (Yen CP, J Neurosurg. 2006 Aug;105(2):213-9.)
Retrospective. 53 patients (8 midbrain, 42 pons, 3 medulla). Mean volume 2.8 cm3 (0.05-21). Prescription dose mean 17.6 Gy (9-25). 37 patients with follow-up imaging
Conclusion: GKS prolongs survival compared to observation; risks are low

Marseille, 2006 (France)(1992-2001) PMID 16385327 -- "Brainstem metastases: management using gamma knife radiosurgery." (Fuentes S, Neurosurgery. 2006 Jan;58(1):37-42; discussion 37-42.)
Retrospective. 28 patients (9 midbrain, 17 pons, medulla 2). Mean max diameter 17.2 mm (10-30). Mean dose 19.6 Gy (11-30)
Conclusion: GKS valuable modality for safe and effective treatment

Basically, it can be done with low incidence of side effects but follow-up is necessarily limited given patient prognosis. Personally, I have treated brainstem mets with single fraction SRS in such patients.

For benign disease, I would be highly conservative. I would not do single fraction SRS if the point dose to brainstem is anywhere near 12 Gy. In those case I would either use FSRT or conventionally fractionated EBRT.

 

[CUBuffsgrad98]

Thanks for the response, but all of these are for mets within the brainstem. I am looking at benign disease encroaching on the brainstem. A U Pitt review article states they allow up to 15 Gy in a single fx using GK under such circumstances and "higher should the situation call for it, such as with an AVM". To me, that seems rather cavalier without much data. 5Gy x 5 seems reasonable, but is there data for this? QUANTAC states 12.5 Gy has <5% risk of permanent cranial neuropathy or necrosis.

 

[Gfunk6]

In community practice, 15 Gy point dose to brainstem for single fraction SRS is WAAAAY more than I would feel comfortable doing.

Let's do some BED calcs to find alternatives.

Assuming alpha/beta = 2 for brainstem, let's look at 1.8 Gy x 28 fractions = 50.4 Gy
BED =95.76

For 5 Gy x 5 fractions = 25 Gy
BED = 87.5

Therefore I think 5 Gy (or slightly higher) x 5 is very reasonable and I've used it myself for benign disease on occasion.

Some data:

1. Meijer, et. al. IJROBP 2003 - 5 Gy x 5 for vestibular schwannoma
2. Hattangadi, et. al. IJROBP 2012 - 8 Gy x 2 for AVM (protons)

 

[CUBuffsgrad98]

Yeah, 5 x 5 seems ok, but dont have good published clinical data to say it works or it is safe....all basically in theory, which gives me pause.

 

[CUBuffsgrad98]

Found a nice paper from Timmerman in "Frontiers of radation therapy and Oncology" (never heard of it either), from 2011. He has tables of dose limits to structures based on how many fractions. For 5 fractions, he lists brainstem dose to <0.5 cc to be a max of 23 Gy and to a "point" dose max of 31Gy. Not sure where he gets these or how derived, but would suggest that 5 Gy x 5 might be on the aggressive side if this is touching the brainstem. Maybe I will just stick with the standard fraction scheme.

 

[Palex80]

I still defer to going 45-50.4 Gy for a benign disease.

Why? Is it because of the increased costs?

My colleague is doing ACNs with 10 x 3.5 Gy. I find this a good schedule. Would you feel comfortable with that?

 

[CUBuffsgrad98]

Why? Is it because of the increased costs?

My colleague is doing ACNs with 10 x 3.5 Gy. I find this a good schedule. Would you feel comfortable with that?

Thanks for the snarky reply and insinuation that I treat for money, not for patients. No, not because of cost, but rather because it has the most robust and reproducible data for efficacy and safety. Can you say the same for 10 x 3.5 Gy? Btw I believe (but could be wrong) a 5 tx radiosurgery costs the same or more than a 25 tx IMRT. I believe in data, patient safety/convenience, and efficacy over cost, do you? And are you sure this is for acoustic neruomas? Those are pretty uncommon, vestibular schwannomas are a different process, but usually the correct term. Does he know what he is treating?

 

[deleted4401]

I'm pretty sure 25-28 fx IMRT with image guidance would be reimbursed at a much higher rate than 5 fx SRS (at least in Maryland).

I didn't get the sense that you said that because you were treating for more money and I think something may have been lost in translation from that response.

 

[CUBuffsgrad98]

I didn't get the sense that you said that because you were treating for more money and I think something may have been lost in translation from that response.

Fair enough, thanks, still dont have the data for 3.5 Gy x 10 though

 

[Gfunk6]

I never quite understood the point of treating brain tumors with IMRT when you are going to doses <= 50.4 Gy. What are you trying to spare? I like to use non co-planar 3DCRT beam arrangements in these scenarios. Nice homogenous dose distribution to the PTV with reduced treatment time and QA.

My 2 cents.

 

[CUBuffsgrad98]

I never quite understood the point of treating brain tumors with IMRT when you are going to doses <= 50.4 Gy. What are you trying to spare? I like to use non co-planar 3DCRT beam arrangements in these scenarios. Nice homogenous dose distribution to the PTV with reduced treatment time and QA.

My 2 cents.

You might be right, in otherwords, you are saying a fractionated stereotactic treatment, as opposed to IMRT? (If I am understanding correctly)? I will also have to discuss with my physics team what they feel they can accomplish. I think the only advantage of IMRT in this situation is if it is an irregular shape?

As far as the data from Meijer you mentioned, I looked it up, as far as brainstem, the only time it is mentioned is just that when it touched the brainstem they did not add a 1 mm margin (for SRS or SRT). No mention of how many patients, if there was a difference between 1 fx and multiple fx, dose limitation to this, etc. I agree with the article that 5x5 may be a reasonable alternative to 12.5x1, but it would also suggest that I can give 12.5x1 if it is touching the brainstem, which I am reluctant to do.

 

[Gfunk6]

You might be right, in otherwords, you are saying a fractionated stereotactic treatment, as opposed to IMRT? (If I am understanding correctly)? I will also have to discuss with my physics team what they feel they can accomplish. I think the only advantage of IMRT in this situation is if it is an irregular shape?

No, I am suggesting the use of 25-28 fractions of 3DCRT over 25-28 fractions of IMRT. That is if you decide upon the fully fractionated route rather than SRS or FSRT.

 

[CUBuffsgrad98]

No, I am suggesting the use of 25-28 fractions of 3DCRT over 25-28 fractions of IMRT. That is if you decide upon the fully fractionated route rather than SRS or FSRT.

I see, well, I think its still in the best interest to spare as much normal brain tissue as possible with a benign tumor case for both acute and long term side effects. The risk is indeed low, but if it was me, I would want it as low as possible. No reason to expose more normal tissue than necessary with modern radiation technology.

 

[deleted4401]

That logic suggests we should use IMRT for everything...

I see, well, I think its still in the best interest to spare as much normal brain tissue as possible with a benign tumor case for both acute and long term side effects. The risk is indeed low, but if it was me, I would want it as low as possible. No reason to expose more normal tissue than necessary with modern radiation technology.

 

[CUBuffsgrad98]

That logic suggests we should use IMRT for everything...

I disagree (although it would seem that it is going that way). What arent we using IMRT for now......skin.....maybe lymphoma.......off the top of my head cant think of much else. Major US institutions are testing, using, or have published IMRT data for rectal, cervical, endometrial, pancreas, stomach, esophagus, H&N, anal, lung, brain....even breast! Even lymphomas in the mediastinum have publications on this. I dont do this for all of these things, but I wouldnt be surprised if some day we all are, as the standard.

 

[Gfunk6]

I see, well, I think its still in the best interest to spare as much normal brain tissue as possible with a benign tumor case for both acute and long term side effects. The risk is indeed low, but if it was me, I would want it as low as possible. No reason to expose more normal tissue than necessary with modern radiation technology.

Classic "protons and parachutes" argument.

My take: 'Better' dosimetry does not equal better clinical outcomes.

The delta between fractionated 3D vs IMRT in this scenario in terms of cost is probably $20,000. What exactly justify that cost?

"Improved acute and long term side effects?" If all dose constraints are met, how do you justify??

CUBuffsgrad, I'm not picking on you, but this is really a problem with our field in general.

 

[CUBuffsgrad98]

Classic "protons and parachutes" argument.

My take: 'Better' dosimetry does not equal better clinical outcomes.

The delta between fractionated 3D vs IMRT in this scenario in terms of cost is probably $20,000. What exactly justify that cost?

"Improved acute and long term side effects?" If all dose constraints are met, how do you justify??

CUBuffsgrad, I'm not picking on you, but this is really a problem with our field in general.

Im not disagreeing with you and you are correct, very similar to protons argument. Cost and reimbursement will always be an issue, but lets give an example (we are veering way off topic now, but its a good debate 🙂 ), would you send a peds medulloblastoma to a proton facility (if it was available), without data to say it is beneficial and is certainly more expensive? If reimbursement was the same for 3D or IMRT, would you be more inclined to use IMRT, or only use it just as often (in general)? I might use it more, personally.

As far as "constraints" being met, constraints only limit toxicity to levels that are arbitrarily considered "acceptable" Look at lung, a V20 <30% is considered reasonable, but still carries up to a 20% risk of pneumonitis, so why not shoot for lower (if it is possible) to make that a 10% risk? For example, a max dose to the brainstem is considered to carry a risk of "<5%".....basically another way of saying <= 4%. Does that mean I will accept a dose up to 54Gy in all my patients, under all curcumstances....let the dose spread out everywhere because it "meets constraints"? I wouldnt, I want to minimize that risk as low as possible, plus I have to look at the patient, the disease, comorbidities, etc, etc. If they have a brainstem met, they live less than a year, the disease process outweighs the toxicity of going higher. If they are 30 years old, healthy and have a vestibular schwanoma, Im going to be over protective with them. Am I justified in doing so? Maybe not on paper, but I believe I am as a clinician (and I will say, I am a person that prefers science over emotion 100x over, but Im still human).

In all, I believe the most humane medicine is minimal to no acute or long term effects, even if its red skin or loose stool, and maximum efficacy. We will never get there, but the closer we can come, the more humane we are with our treatment.

Unfortunately, cost does play a role. I dont know how to justify side effects over cost, even if acute, but certainly a line must be drawn somewhere. Its difficult to say where that is.

 

[medgator]

I never quite understood the point of treating brain tumors with IMRT when you are going to doses <= 50.4 Gy. What are you trying to spare? I like to use non co-planar 3DCRT beam arrangements in these scenarios. Nice homogenous dose distribution to the PTV with reduced treatment time and QA.

My 2 cents.

Completely agree.

On a side note, anyone use a different brainstem constraint between malignant and benign tumors in terms of brainstem tolerance? The rtog will allow 54-55 Gy for gbms while I use 50.4 for benign conditions

 

[CUBuffsgrad98]

Interestingly enough, I some data for the 3.5 Gy x 10 suggested above. Single institution Belgium:

PMID: 21665381

 

[CUBuffsgrad98]

On a side note, anyone use a different brainstem constraint between malignant and benign tumors in terms of brainstem tolerance? The rtog will allow 54-55 Gy for gbms while I use 50.4 for benign conditions

Latest Phase III GBM study of dasatinib vs placebo with temodar and EBRT allows for 55 Gy to no more than 1% or 1cc of brainstem unless PTV goes into brainstem. Then the limit is confined to brainstem minus PTV. Study chair is Mayo clinic rad onc.

 

[Palex80]

Thanks for the snarky reply and insinuation that I treat for money, not for patients.

I didn't mind to sound rude. And I never wanted to suggest, that you were treating for money and not for patients. I was just not certain, why you wanted to RS for a disease where we have tons of data with normally fractionated treatment.
Where I work in Europe a 6 week 3D-CRT in stereotactic technique for such a disease would be reimbused considerably higher than a radiosurgery. On the other hand, we would lose lots of treatment room time.

No, not because of cost, but rather because it has the most robust and
reproducible data for efficacy and safety. Can you say the same for 10 x 3.5 Gy?

10 x 3.5 Gy is certainly safe. We have tons of data to extrapolate to that assumption.
If its effficient or not is another question, but I'd guess it is. Its certainly equally efficient in terms of BED to 5x5 Gy.

And are you sure this is for acoustic neruomas? Those are pretty uncommon, vestibular schwannomas are a different process, but usually the correct term. Does he know what he is treating?

It's for both, sorry for not being specific enough about that.

 

[Palex80]

Interestingly enough, I some data for the 3.5 Gy x 10 suggested above. Single institution Belgium:

PMID: 21665381

LOL, There's actually data for that! I didn't know that.

It's the same schedule we offer for glioblastoma re-treatment after 60/2.

 

[Palex80]

On a side note, anyone use a different brainstem constraint between malignant and benign tumors in terms of brainstem tolerance? The rtog will allow 54-55 Gy for gbms while I use 50.4 for benign conditions

In my opinion these constraints are set too low.

Quantec calls for 60 Gy to 1-10cc, if I am not mistaken. In daily practice with GBMs we often give 60 Gy to 10ccs as long as we use standard (5mm) margin from CTV-PTV and the CTV goes up to the brainstem surface.

 

[Kroll2013]

Brainstem
Several papers have reported toxicity outcomes after SRS for
brainstem metastases, with a broad range of peripheral doses
(9&#8211;30 Gy), with median peripheral doses on the order of 15&#8211;20
Gy123&#8211;130 While grade 3&#8211;4 toxicity is reportedly uncommon, these
patients suffer from poor survival (median survival 5&#8211;11 months),
and may not have the opportunity to manifest late radiation toxicity.
Also, radiation toxicity is often difficult to discern from symptomatic
progression. Reported toxicity includes hemiparesis,
ataxia, facial palsy and seizures.124&#8211;126 A study from U. Pittsburgh
reported the rate of adverse imaging findings and neurologic deficits
after SRS for 38 patients with benign tumors, followed 6&#8211;84
(median 41) months.131 Not all patients with adverse imaging findings
developed neurologic deficits (defined as new deficits arising
within the brainstem long tracts or adjacent cranial nerves), and
some patients developed neurologic deficits in the absence of
adverse imaging findings. Interestingly, there was no correlation
between marginal dose and adverse imaging findings or neurologic
deficits. In fact, marginal doses >18 Gy were associated with less
neurologic deficits than marginal doses of 15&#8211;17 Gy (16.6% versus
19.1%, not significant), which the authors attribute to differences in
distribution of target types between different dose groups (i.e. the
15&#8211;17 Gy group represented mostly cavernomas and the >18 Gy
represented mostly arteriovenous malformations). For benign tumors
compressing the brainstem, for which patients are already
symptomatic, or at risk of becoming symptomatic, marginal doses
of 13 Gy appear to be well-tolerated.132 The generally accepted
dose constraint to the brainstem is considered to be on the order
of 12 Gy, and efforts to minimize brainstem doses are warranted.
Based upon the published studies, a brainstem maximum dose
of 10&#8211;12 Gy is expected to result in a minimal (<1&#8211;2%) risk of
brainstem toxicity. From the recent QUANTEC review, for doses
of 12.5, 14.2, 16.0 and 17.5 Gy: partial volume irradiation to onethird
of the brainstem results in normal tissue complication probabilities
(NTCP) of 1%, 13%, 61% and 94%, respectively; and brainstem
maximal doses results in NTCPs of 0.2%, 3.2%, 26% and 68%
respectively.133 In general, the brainstem maximum should be
maintained below 10&#8211;12 Gy if feasible, though when therapeutic
dose to the target is compromised with such a constraint, particularly
when the risks of treatment failure outweigh the risks of
treatment toxicity, higher maximal doses to the brainstem shouldbe considered. A notable indication to exceed 12 Gy maximal doses
to the brainstem would be situations in which tumor control
would be compromised (i.e. primary or malignant brain tumors).
Also, in the treatment of trigeminal neuralgia, negligible volumes
of brainstem may receive doses >25&#8211;50 Gy.

Stereotactic radiosurgery and hypofractionated stereotactic radiotherapy:
Normal tissue dose constraints of the central nervous system
Michael T. Milano a,&#8657;,

 

[Kroll2013]

In community practice, 15 Gy point dose to brainstem for single fraction SRS is WAAAAY more than I would feel comfortable doing.


[ Therefore I think 5 Gy (or slightly higher) x 5 is very reasonable and I've used it myself for benign disease on occasion]

unfortunately, a 5*5 regimen isn't sufficient to control an AVM in terms of obliteration rate , which is coorelated with a margin dose of a minimum of 18 Gy.

 

[CUBuffsgrad98]

In community practice, 15 Gy point dose to brainstem for single fraction SRS is WAAAAY more than I would feel comfortable doing.


[ Therefore I think 5 Gy (or slightly higher) x 5 is very reasonable and I've used it myself for benign disease on occasion]

unfortunately, a 5*5 regimen isn't sufficient to control an AVM in terms of obliteration rate , which is coorelated with a margin dose of a minimum of 18 Gy.

5x5 SEEMS fine, and I have seen people use it, but is the long term toxicity (and/or control) data there? Not so sure. For people that are going to be living for several decades, I want to be cautious, but I dont want to waste their time bringing them in 25-30 times if not necessary. I think if it was me, I would go for a more fractionated regimen, but patients dont always see it that way.

 

[Palex80]

I think if it was me, I would go for a more fractionated regimen, but patients dont always see it that way.

Do you mean something like 28x1.8Gy or something between that and the 5x5Gy?