.

[anonymuncule]

.

---

Members don't see this ad.

 

[Palex80]

I don't like SRS for resection cavity.
The local control rates reported are not good in my opinion.
SRS is a technique were usually a lower isodose is prescribed to the edge of the PTV and a higher dose to the middle of the PTV.
This makes sense if you are treating an intact metastatic lesion, it makes no sense however if you are treating a cavity, since the highest risk for remaining cells is not in the middle of the resection cavity but rather on the circular resection margin.
Furthermore I do have some concerns on morbidity. The normal brain tissue around the cavity may be subjected to less favorable blood circulation and with SRS further damaging endothelium the risk of radionecrosis may rise.

We no longer perform WBRT in the postoperative setting after resection of brain lesions with the exception of resected metastasis of SCLC (which itself is very seldom and only happens if brain surgery was performed as an emergency procedure).
We perform postoperative irradiation of the resection cavity, yet not with SRS but with a fractionated stereotactic treatment schedule, delivering a homogenous dose to the PTV (6 x 5 Gy).
We changed our practice around 4 years ago, after more and more data were pointing into this direction.
Anecdotally some of my colleagues report more patients failing with meningeosis neoplastica after stereotactic treatment and blaim the lack of WBRT for this. I think it's more of a bias. Superior local control in the resection cavity and possibly better distant control with more systemic treatment options becoming available (patients with NSCLC live longer nowadays than 5 years ago and NSCLC is major cause for brain mets) means that more patients are going to fail with meningeosis neoplastic down the road.

Recently a working group published contouring guidelines, suggesting to include a generous margin around the adjacent dura in superficially lying resection cavities. We have not adopted this recommendation.

 

[seper]

Contouring guidelines for brain cavity? Sounds useful, please share.
My reading of this study is that brain control does not impact overall survival of metastatic patients.
This is upsetting to radiation oncologists.
Logically, the next trial should address surgery alone for single brain mets. Some people already advocate it for young patients with breast cancer.

 

[OTN]

I don't like SRS for resection cavity.
The local control rates reported are not good in my opinion.
SRS is a technique were usually a lower isodose is prescribed to the edge of the PTV and a higher dose to the middle of the PTV.
This makes sense if you are treating an intact metastatic lesion, it makes no sense however if you are treating a cavity, since the highest risk for remaining cells is not in the middle of the resection cavity but rather on the circular resection margin.
Furthermore I do have some concerns on morbidity. The normal brain tissue around the cavity may be subjected to less favorable blood circulation and with SRS further damaging endothelium the risk of radionecrosis may rise.

We no longer perform WBRT in the postoperative setting after resection of brain lesions with the exception of resected metastasis of SCLC (which itself is very seldom and only happens if brain surgery was performed as an emergency procedure).
We perform postoperative irradiation of the resection cavity, yet not with SRS but with a fractionated stereotactic treatment schedule, delivering a homogenous dose to the PTV (6 x 5 Gy).
We changed our practice around 4 years ago, after more and more data were pointing into this direction.
Anecdotally some of my colleagues report more patients failing with meningeosis neoplastica after stereotactic treatment and blaim the lack of WBRT for this. I think it's more of a bias. Superior local control in the resection cavity and possibly better distant control with more systemic treatment options becoming available (patients with NSCLC live longer nowadays than 5 years ago and NSCLC is major cause for brain mets) means that more patients are going to fail with meningeosis neoplastic down the road.

Recently a working group published contouring guidelines, suggesting to include a generous margin around the adjacent dura in superficially lying resection cavities. We have not adopted this recommendation.

I've been doing 6 Gy x 5 to the tumor bed for a few years now as well and have been happy with the results. I omit whole brain in these cases to avoid toxicity, but scan q3 months to pick up new mets very early. I do add a 2mm margin to the tumor bed.

 

[maruchan]

This is the article that provides "guidelines" for postop SRS: Consensus Contouring Guidelines for Postoperative Completely Resected Cavity Stereotactic Radiosurgery for Brain Metastases. - PubMed - NCBI. It is more a collection of opinions than a set of guidelines.

Personally, I agree with many of Palex's points. With postoperative SRS, volumes tend to be much bigger than the intact setting, leading to a significant risk of radionecrosis. Furthermore, even with these larger volumes, there is a significant risk of recurrence just outside the treatment field (perhaps due to surgical violation or due to the evolution of surgical cavity shape over time). Anecdotally this holds true even with a 2mm margin. It's unsurprising that SRS has a poor local control rate in the postop setting.
A caveat regarding the NCCTG study in particular though, they make a note in the manuscript that some of the "local failures" may actually have been radionecrosis or pseudoprogression. So the LC rate may not be quite as bad as their paper suggests.

WBRT feels dirty too though given the cognitive side effects. One could argue for close observation with SRS of recurrences as they arise, it's unlikely to change survival and decreases significantly the volume of brain radiated while also decreasing chance of a marginal miss. I'm not aware of any evidence supporting this approach however.

 

[evilbooyaa]

The recent post-op contouring guidelines are not very useful, IMO, because they don't talk about appropriate dosing. Link: Consensus Contouring Guidelines for Postoperative Completely Resected Cavity Stereotactic Radiosurgery for Brain Metastases. - PubMed - NCBI

The most important thing in the post-op scenario, is that you need to have margin to the post-op cavity (I've generally seen at least 1, I would usually do 2mm, and include dural tail if abutting dura initially), and you need to pick a dose that is safe for the size of what you are treating. You can't add 2mm to a large post-op cavity then try to safely and properly do single fraction to it IMO.

The alternative is to not treat and treat at recurrence when you have a target. IMO that's probably OK too as long as you're monitoring.

 

[Palex80]

Logically, the next trial should address surgery alone for single brain mets. Some people already advocate it for young patients with breast cancer.

Well this has been partially tested before in the EORTC trial most recently and was not good.
Adjuvant Whole-Brain Radiotherapy Versus Observation After Radiosurgery or Surgical Resection of One to Three Cerebral Metastases: Results of the EORTC 22952-26001 Study

The local failure rate after resection is significantly higher than after radiosurgery. If you add no adjuvant treatment to surgery, half of the patients are going to have a local recurrence. That's an awful lot...


I add 3mm to the resection cavity when doing 6 x 5 Gy and also include part of the dural tail (without "overdoing" it as in the "guidelines"). It's quite hard to distinguish postoperative dural tail (scar) and possibly contaminated tissue / spread around the cavity.

 

[Palex80]

So maruchan and evilbooyaa, you don't do FSRT of the resection cavity?

Personally I am quite biased against RC even in the primary (non-postoperative) setting.
With the exception of small lesions (<1cm), I have been doing more and more FSRT for brain mets.
The Minniti-schedule for example is excellent in my opinion, I have not seen any toxicity and 3 fractions are not that many.
Single-Fraction Versus Multifraction (3 × 9 Gy) Stereotactic Radiosurgery for Large (>2 cm) Brain Metastases: A Comparative Analysis of Local Contr... - PubMed - NCBI

 

[seper]

Thank you for sharing, it's a very interesting study because it reports the large proportion of control-related neurologic deaths, as opposed to similar US papers.

Well this has been partially tested before in the EORTC trial most recently and was not good.
Adjuvant Whole-Brain Radiotherapy Versus Observation After Radiosurgery or Surgical Resection of One to Three Cerebral Metastases: Results of the EORTC 22952-26001 Study

The local failure rate after resection is significantly higher than after radiosurgery. If you add no adjuvant treatment to surgery, half of the patients are going to have a local recurrence. That's an awful lot...


I add 3mm to the resection cavity when doing 6 x 5 Gy and also include part of the dural tail (without "overdoing" it as in the "guidelines"). It's quite hard to distinguish postoperative dural tail (scar) and possibly contaminated tissue / spread around the cavity.

 

[evilbooyaa]

So maruchan and evilbooyaa, you don't do FSRT of the resection cavity?

Personally I am quite biased against RC even in the primary (non-postoperative) setting.
With the exception of small lesions (<1cm), I have been doing more and more FSRT for brain mets.
The Minniti-schedule for example is excellent in my opinion, I have not seen any toxicity and 3 fractions are not that many.
Single-Fraction Versus Multifraction (3 × 9 Gy) Stereotactic Radiosurgery for Large (>2 cm) Brain Metastases: A Comparative Analysis of Local Contr... - PubMed - NCBI

I'm not sure what RC in this context means.

I nearly always do FSRT after surgical resection. I don't like SF-SRS for post-op. Somewhere in the 8-9Gy x 3 based on the Minniti paper you linked (as well as this one looking at specifically in the post-operative setting: https://www.ncbi.nlm.nih.gov/pubmed/23683828). I try to avoid 5Gy x 5 if I can. I was just stating that I've heard that some people do not treat adjuvantly and instead treat with salvage. The local recurrence rate after surgery alone is 50% - That is high, but if you monitor these folks well enough then perhaps you could avoid treatment in 50% of them (similar to the argument for salvage radiation in prostate vs adjuvant).

I was not aware of that EORTC study. Interesting that intracranial-based survival was better with addition of WBRT but without an OS benefit. Agree with seper that similar US papers haven't explicitly reported that end point. I'm always interested in the follow-up schedules for these patients - they report 3 month MRIs was their plan - I wonder what their compliance to that was. But an interesting end point to be positive. I still think that given the neurocognitive decline seen in WBRT I wouldn't add it routinely in this situation.

 

[Radiator20]

Here's the way I was taught...

If this brain met was the only met (CNS or elsewhere) and got a GTR, and primary has been/will be definitively treated, *consider* WBRT, even to 37.5/15. CNS might be their only site of metastatic disease and you might just sterilize it and cure them. Essentially maximally aggressive locoregional therapy for oligometastatic disease. But only in a non-elderly pt with good baseline neurocog, and with memantine but not HA-WBRT (not very locoregionally aggressive if you're intentionally missing 5-10% of the target). But this is a rare scenario.

If otherwise, seems like many reasonable approaches: 3D partial brain tx (essentially very generous cavity tx, using WBRT fractionation) vs cavity SRT. Brown et al does seem to suggest that SRS is not the best approach (at least, with how they contoured their SRS volumes)--plus the good points Palex and Evilbooya raise. Seems to me that waiting till salvage is going to decrease local control and/or increase radionecrosis if you get enough dose in to overcome that.

If adjuvant SRT, which seems like the vast majority of cases:

Dose: 25/5, with SIB to 30/5 to any areas of enhancing residual disease on POD1 MRI (extrapolating those areas to current MRI), or just 30/5 to whole cavity if there's diffuse concern. Seems like a good balance between minimizing risk of radionecrosis and keeping good local control. Evilbooya, curious why you do not like 25/5?

Contouring similar to the consensus paper linked:
-cavity (no formal margin, but do contour generously, and see PTV below)
-surgical tract to cavity
-any areas of dura, tentorium, falx, or ventricle that tumor touched preop plus 1 cm expansion along those structures. And I suppose if the surgeon somehow got into tentorium, falx, or ventricle intra-op, I'd expand 1 cm from what was violated too even if there wasn't preop contact, as seems like at risk for seeding.
-generally all of the dura underlying the entire craniotomy site plus whatever additional dura looks abnormal postop. This was all exposed and irrigated with tumor cells intraop so seems like it's all going to be at risk of recurrence. This does often result in a generous dural volume but with steep dose falloff, additional dose to underlying brain is usually minimal. Easy to forget, but range lock to stay out of bone on the dural volume.

2mm PTV given multi-fx and cavity may change shape a bit day to day.

 

[evilbooyaa]

25/5 to me is not sufficient dose for long-term local control. The recurrence rates I've personally seen from 25/5 (for both intact intracranial disease and post-op) are unacceptably high. I will do it if necessary it's a gigantic post-op cavity that a 3 fraction regimen can't meet constraints on, but if a post-op cavity can meet brain V24 < 15-17cc (from Minniti post-op with 2% risk of RN) and V18 < 30cc (from Minniti intact) I'm going with a 3 fraction regimen every time. I also disagree that covering the entire surgical tract is necessary, especially for deep seated tumors. That leads to an excessive amount of brain tissue getting prescription dose, IMO. However, it's a land of minimal to no data, so people are free to do whatever they would like, and I would never begrudge somebody for defaulting to 25/5, but that is not going to be my routine practice.

30/5 is more inline with 27/3 but I'm not aware of any published, reasonable dose contraints for a 5 fraction regimen. Anecdotally, I saw one patient with bad necrosis requiring resection 9-12 months after 30Gy/5 given post-operatively.

Also, I don't know that I would ever WBRT someone with a single met. I get the Patchell study, but we've got to come farther than that IMO. However, I am cognizant about my biases against WBRT and my institution's likely under use of it currently.

 

[Radiator20]

Thanks—helpful! 100% agree that 25/5 is not an ideal dose for gross disease (intact met or after STR). 27/3 certainly seems reasonable. Also, I would say we in fact have come farther than Patchell—good MRI means it probably really is a single brain met, and much better systemic tx to control potential systemic micromets. So if anything the benefit of WBRT in this admittedly rare scenario might be larger now. But, again, a rare scenario and surely many reasonable ways to manage it.

 

[Palex80]

Bear in mind that Patchell did not give 30 / 3... So if you are going to do WBRT based on Patchell, you'd have to do 50.4 / 1.8.

 

[Gfunk6]

Bear in mind that Patchell did not give 30 / 3... So if you are going to do WBRT based on Patchell, you'd have to do 50.4 / 1.8.

Excellent point Palex. In the era of SRS and the Eric Chen trial in Lancet Oncology, I personally think it is malpractice to do this much whole brain radiation to a patient with 1-3 brain mets. Even the hippocampal-sparing Phase II study only did 30/10.

 

[evilbooyaa]

Bear in mind that Patchell did not give 30 / 3... So if you are going to do WBRT based on Patchell, you'd have to do 50.4 / 1.8.

Good point. I would LOL and then feel really sad if I saw somebody did this to a patient in the current age.

 

[medgator]

Good point. I would LOL and then feel really sad if I saw somebody did this to a patient in the current age.

But, but, but fractionation.... Late effects...






/s

 

[Radiator20]

Great points as always--thank you! I completely agree that 50.4/1.8 WBRT in this scenario seems inconceivable. That said, clearly even standard WBRT is better than SRS/SRT for distant intracranial control, which might... might... be worthwhile in this rare scenario. I don't necessarily take Patchell as direct support for this option; I'm not sure there is modern data directly supporting this option; nor frankly would I say that it is in any absolute sense the right thing to do; but I do think in the right patient one could at least consider discussing it with them as a possibility.

 

[medgator]

Obviously someone did an RTOG pci trial in nsclc for a reason, even if the difference in recurrence rates didn't pan out into an OS difference.

Personally, I'm still going to give WBRT to a pt with a large 4cm+ adenoca/nsclc met with extensive vasogenic edema and midline shift that is resected, esp if the pt has a less than optimal PS with extensive extracranial mets at dx or relapse.

I wish there was more clarity here regarding techniques, dosing etc, seems like a of heterogeneity out there in terms of who is doing conventional whole, partial brain, fsrt and single fx srs and to what target.

 

[Neuronix]

Personally, I do GK SRS to most post-op cavities (1 mm expansion) and consider fractionation to large cavities with 5 fractions, dose depending on size and amount of gross disease. Then I use a 2 mm PTV for these linac based treatments. I think it's important to cover any dural surfaces the tumor touched prior to resection and do careful fusions and contouring in this regard.

The biggest problem I have with intracranial FSRT is that there is a tremendous heterogeneity to the doses and techniques used and results obtained.

Contouring guidelines for brain cavity? Sounds useful, please share.
My reading of this study is that brain control does not impact overall survival of metastatic patients.
This is upsetting to radiation oncologists.
Logically, the next trial should address surgery alone for single brain mets. Some people already advocate it for young patients with breast cancer.

Lower brain control is not upsetting to this radiation oncologist. The salvage options here are quite good. The neurocognitive outcomes in the population are better if you salvage the failures rather than give WBRT up front to everyone. This was the main point of the Alliance trial.

As for surgery alone vs. SRS for post-op cavities, see: http://www.thelancet.com/journals/lanonc/article/PIIS1470-2045(17)30414-X/abstract

I generally follow the dose schemes in that paper, but 18 Gy for small cavities and try to get everyone to at least 15 Gy, V12 permitting.

A caveat regarding the NCCTG study in particular though, they make a note in the manuscript that some of the "local failures" may actually have been radionecrosis or pseudoprogression. So the LC rate may not be quite as bad as their paper suggests.

This is an important point that confuses all of these studies.

The Alliance trial protocol states: "Disease progression: Development of new nodular contrast enhancement in the surgical bed."

I see plenty of pseudoprogression with SRS, particularly in this immunotherapy era. Many of the local failures reported in the trial with SRS may not have been real.

 

[emt409]

I think is also a great opportunity to plug pre-op SRS whenever possible. It's difficult to sell to the neurosurgeons because they don't want to delay their operation; however, the Emory/Carolina report last year in JNO compared pre-op SRS to post-op WBRT, and found equivalent 2y LC, OS, and rate of leptomeningeal spread. Obviously a single-fraction of reduced dose (20% lower than RTOG 9005) confers much lower risk of NC decline.

Comparing pre-operative stereotactic radiosurgery (SRS) to post-operative whole brain radiation therapy (WBRT) for resectable brain metastases: a multi-institutional analysis

We've really worked hard in our institution to be able to turn around a pre-op SRS plan within 1-2 days, so the neurosurgeon's operations aren't delayed and the results have been excellent.

 

[evilbooyaa]

I actually like the Emory/Carolina report comparing apples to apples better than the above article - Pre-op SRS to Post-op SRS. Equivalent survival and outcomes, but lower rates of leptomeningeal disease and SRN with pre-op RT Comparing Preoperative With Postoperative Stereotactic Radiosurgery for Resectable Brain Metastases: A Multi-institutional Analysis. - PubMed - NCBI

In symptomatic patients it's hard, but IMO if the NSGs are OK to consider it, I would offer that to them.

 

[Radiated_GuineaPig]

Our general practice is to do postoperative SRS to around 15 Gy depending on cavity size. This is based on both Paul Brown's paper (as cited above) and Anita Mahajan's paper (see below: i can't post link because i'm too much of a newb)


Post-operative stereotactic radiosurgery versus observation for completely resected brain metastases: a single-centre, randomised, controlled, phase 3 trial.

 

[RickyScott]

Our general practice is to do postoperative SRS to around 15 Gy depending on cavity size. This is based on both Paul Brown's paper (as cited above) and Anita Mahajan's paper (see below: i can't post link because i'm too much of a newb)


Post-operative stereotactic radiosurgery versus observation for completely resected brain metastases: a single-centre, randomised, controlled, phase 3 trial.

But why one fraction? ( radiologically there is typically only normal tissue in the PTV with 2mm expansion)

 

[Neuronix]

But why one fraction? ( radiologically there is typically only normal tissue in the PTV with 2mm expansion)

I quoted the same paper (Mahajan trial).

This is just one (common) way of doing GK SRS to cavities.

I used to treat cavities with zero expansion to the same doses as RTOG 9005 prescribed to the cavity plus any enhancement or residual.

Now I use the same cavity + enhancement + residual + 1 mm to the doses given in the Mahajan trial (with the caveats I mentioned in my earlier post). That's what they do at my institution so they wanted the expansion.

If you look at the doses on a Gammaknife (i.e. 18 Gy without an expansion vs 16 Gy with 1 mm expansion), it's very similar dosimetrically.

 

[Radiated_GuineaPig]

Oopse sorry neuronix... didn't see that. I scanned the forum a bit too quickly. Yeah i agree, our thinking is that the benefits outweight the risks

 

[evilbooyaa]

But why single fraction? I guess it's a preference thing and since I didn't train on GK I'm not single fraction most of the time in this scenario. We know from Minniti that intact tumors >2cm = more necrosis with single fraction and worse oncologic outcome. If post-op cavities are greater than 2cm (which I've never seen them not be) then why not fractionate that too?

 

[Neuronix]

Oopse sorry neuronix... didn't see that. I scanned the forum a bit too quickly.

GMTA

 

[nkmiami]

But why single fraction? I guess it's a preference thing and since I didn't train on GK I'm not single fraction most of the time in this scenario. We know from Minniti that intact tumors >2cm = more necrosis with single fraction and worse oncologic outcome. If post-op cavities are greater than 2cm (which I've never seen them not be) then why not fractionate that too?

I had several pts on NCCTG trial and I thought it mandated 2mm expansion, probably based on Stanford experience that less conformal plans had better local control. Personally, I have seen a lot more heterogeneity in cavity contours than intact brain mets (I am not talking about contouring the dura), but I see a lot of variability on what is perceived as cavity on the MRI. I also agree that when there is a much larger volume of normal tissue in the ptv than tumor, I would conceptually favor a fractionated approach.

 

[Neuronix]

I had several pts on NCCTG trial and I thought it mandated 2mm expansion.

There is no expansion in the NCCTG trial.

See: SRS With or Without WBRT and Cognitive Function in Patients With Brain Metastases (supplement)