punctate brain mets

[nkmiami]

Saw asymptomatic melanoma pt with several 1-2cm brain mets and 8-9 punctate lesions. In the past I used to treat everything, but I am really considering treating just the "overt" disease and closely following the rest. Would be interested to know how others are approaching this.

---

Members don't see this ad.

 

[BobbyHeenan]

I kind of take it on a case by case basis.

For reference, I do both gamma knife and TrueBeam frameless 3-5 fraction (roughly speaking, gamma knife for smaller lesions, truebeam treatment for larger). My neurosurgeons are fantastic and very involved and we review all follow up scans together and make the decisions together.

If the punctate lesions are noted on a follow up MRI after their radiosurgery we typically just repeat the scan in ~2 months - especially if they haven't started their systemic therapy and/or they are about to start new systemic therapy which may have CNS activity; certainly with melanoma I've seen them go away or be stable for quite a while on BRAF/MEK or immune therapy.

If they've already got their gamma knife frame on then I often just go ahead and treat the punctate mets.

With your patient with that many I would fault no one for doing any of the following - only SBRT/SRS the big one(s), radiosurgery to all of them, or whole brain.

If it was me with the brain mets I'd probably sign up for SRS on the mets > 1 cm, observe the rest and hope I get some immune therapy or BRAF activity in the brain and hold off on the other ones.

 

[evilbooyaa]

If the punctate ones are going to push you from SRS to whole brain, then I wouldn't treat them.

Punctate to me means < 1-2mm in size, for which the accuracy of treating stuff at that size isn't great anyways. Anything bigger than like 3-4mm gets treated. Linac-based at my institution.

If patient is treatment naive I don't think it's wrong to do shared decision making on observing the non-punctate ones. That being said, intracranial response rates are not nearly as good for systemic therapy as they are for good ol' radiation. However, any lesions in close proximity to one(s) that you're definitely treating now are getting treated in my book to avoid having to come back to that area in the future.

 

[medgator]

If braf/pdl1 negative, why not just start with wbrt and then srs boost the residual disease.

I guess melanoma is why some here wouldn't do wbrt?

 

[BobbyHeenan]

If braf/pdl1 negative, why not just start with wbrt and then srs boost the residual disease.

I guess melanoma is why some here wouldn't do wbrt?

Yes - melanoma is what is swaying my opinion here.

With NSCLC for example I'm giving whole brain here.

 

[evilbooyaa]

Yes - melanoma is what is swaying my opinion here.

With NSCLC for example I'm giving whole brain here.

Assuming non-EGFR/ALK-mutated, then maybe. If it is EGFR or ALK mutated I'm SRSing the big ones and getting them to TKI. I won't skip out on SRS for a TKI, but I will skip out on WBRT for TKI with close surveillance and spot SRS as needed. Yes, it's more expensive than WBRT for all, but it's also less toxic.

Even if it wasn't mutated, I'd consider the same treatment plan but SRS everything if patient was motivated to avoid whole brain. Minimize toxicity if they're understanding of need for repeat trips in future.

This would be regardless of histology (outside of something like small cell) in this range and distribution of mets.

 

[Radiator20]

I am really considering treating just the "overt" disease and closely following the rest.

I'm not sure I see the point in treating bigger lesions but not treating punctate lesions, unless there's genuine uncertainty around whether they're mets. Maybe they'll be temporized for a while with systemic therapy but they're going to be the highest risk sites for CNS failure moving forward so would favor just treating them now - there shouldn't, in general, be that much added toxicity for SRS to a few more tiny lesions, and there's no additional logistical burden since they're having to get SRS anyway. But if all the patient had was punctate mets, and they had a systemic therapy option with good CNS activity, I don't think I'd pull the trigger on SRS just for the punctate mets - would agree with close observation in that circumstance.

With NSCLC for example I'm giving whole brain here.

I certainly don't think WBRT is wrong and would present it as an option. But I would also offer the option of SRS to all lesions, if pt was on board with close f/u. (Especially if they had a targetable mutation or other systemic option with good CNS activity.)

why not just start with wbrt and then srs boost the residual disease

Curious--what SRS dose do you use as an immediate post-WBRT boost? Have you seen radionecrosis in this setting? To me it seems that there's no way to know which if any of these mets may fail first after WBRT, and most/all might be durably controlled, so SRS boost to all up front might end up being quite a bit of premature prophylactic tx.

 

[medgator]

Curious--what SRS dose do you use as an immediate post-WBRT boost? Have you seen radionecrosis in this setting? To me it seems that there's no way to know which if any of these mets may fail first after WBRT, and most/all might be durably controlled, so SRS boost to all up front might end up being quite a bit of premature prophylactic tx.

I would wait at least a few months post wbrt before doing anything, and would re image, should've clarified.

If I ended up needing to treat, Then I would stick with the standard rtog single fx dosing based on size

 

[Radiator20]

I would wait at least a few months post wbrt before doing anything, and would re image, should've clarified.

If I ended up needing to treat, Then I would stick with the standard rtog single fx dosing based on size

Oh, gotcha, that makes much more sense. Thanks!

 

[Neuronix]

Curious--what SRS dose do you use as an immediate post-WBRT boost? Have you seen radionecrosis in this setting?

See RTOG 9508: Whole brain radiation therapy with or without stereotactic radiosurgery boost for patients with one to three brain metastases: phase III results of... - PubMed - NCBI

SRS was 1 week after WBRT. Grade 3 toxicity was something like 3%.

 

[evilbooyaa]

The citation is welcome Neuronix, but no one I know follows that treatment paradigm because single unresectable brain met will get SRS alone in this day and age. I wouldn't routinely add SRS as a boost to WBRT without assessing response first, like medgator said.

 

[MDACCRules]

The citation is welcome Neuronix, but no one I know follows that treatment paradigm because single unresectable brain met will get SRS alone in this day and age. I wouldn't routinely add SRS as a boost to WBRT without assessing response first, like medgator said.

I think he posted it because of the dose used on top of WBRT and the relative safety of it. But, I think people use that same dose even without whole brain.

Some of those punctate lesions just stay stable or disappear. And if you put a 1 mm margin on a 2 mm spot, you're treating almost 7x more normal brain than tumor (4 mm3 of tumor, 29 mm3 of normal brain). I don't know ... seems reasonable to follow, especially if on immunotherapy. Whole brain is barbaric.

 

[medgator]

Whole brain is barbaric.

So is recurrent/progressive metastatic disease to the brain. Pick your poison

 

[MDACCRules]

So is recurrent/progressive metastatic disease to the brain. Pick your poison

Um, I think that’s exactly what I did by saying would avoid whole brain. With a few trials showing detriment, I’d rather hold off than kill them.

 

[medgator]

Um, I think that’s exactly what I did by saying would avoid whole brain. With a few trials showing detriment, I’d rather hold off than kill them.

In the wrong patient, you're probably killing them with untreated disease

 

[Neuronix]

Whole brain is barbaric.

Tell that to NIA and Evicore

 

[Palex80]

WBRT for asymptomatic lesions has never shown a survival benefit in a trial...

 

[MDACCRules]

WBRT for asymptomatic lesions has never shown a survival benefit in a trial...

Yep. And a few showed significant detriments. But don’t let the wholey brain people tell you otherwise.

 

[emt409]

A point that no one has underscored here is that waiting for them to grow comes with significant increase in volume at risk for radionecrosis. The increase in V12 and mean brain dose in multi-target radiosurgery is not strictly linear with the size of the mets, because there is co-variance between the number and the total volume and the distribution, whether it's linac or GK.

If I'm sure it's a met, and I'm doing SRS, I treat it. Like what medgator said above, if there's doubt, then observe.

 

[medgator]

WBRT for asymptomatic lesions has never shown a survival benefit in a trial...

Neurologic death increased without wbrt in a randomized trial. Many things are done without an OS benefit. Before the update of the post-prostatectomy data, there was no OS benefit documented but men were avoiding long term ADT with its use, which is worth it to some

 

[Palex80]

Neurologic death increased without wbrt in a randomized trial. Many things are done without an OS benefit. Before the update of the post-prostatectomy data, there was no OS benefit documented but men were avoiding long term ADT with its use, which is worth it to some

I presume you are referring to Patchell.
Patchell is however another scenario. RESECTED solitary metastasis followed by WBRT, without evidence of macroscopic residual tumor in the brain at the timepoint of irradiation.

 

[MDACCRules]

No, he is correct that EORTC trial showed a benefit for WBRT in terms of reducing neurologic death, but the primary endpoint was negative so it is a negative trial (PS deterioration to 2). Japanese trial showed a benefit if you look at just the good prognosis patients, but not overall, and definitely not in poor prognosis patients. Primary endpoint (overall survival) was negative. So, it's a negative trial. MDACC showed overall survival detriment with addition of WBRT. The primary endpoint was met (neurocognitive function). It's a POSITIVE trial.

So, can consider adding WBRT based on 2 negative trials (who makes decisions like this?). Or omitting based on 1 positive trial.

Poll- who makes affirmative treatment decisions based on negative trials?

 

[evilbooyaa]

Are you insinuating that we should never do WBRT? Or just in the 1-3 or 1-4 met group?

I'm not a fan of it just like you, but are you overarchingly claiming that because the primary end point in a trial is negative, all data from that trial should be ignored, since it's a negative trial?

Regardless, the Paul Brown trial did NOT show a detriment to overall survival with addition of WBRT in all comers. the P-value was 0.92 for that survival endpoint. In patients with the BEST GPA, it showed a survival benefit to addition of WBRT.
"Significantly better OS was observed in the DS-GPA 2.5-4.0 group in WBRT + SRS vs the SRS alone, with a median survival time of 16.7 (95% CI, 7.5-72.9) months vs 10.6 (95% CI, 7.7-15.5) months (P = .04) (hazard ratio

---

, 1.92; 95% CI, 1.01-3.78)." Stereotactic Radiosurgery With or Without Whole-Brain Radiotherapy for Brain Metastases: Secondary Analysis of the JROSG 99-1 Randomized Clinical T... - PubMed - NCBI
It did show increased neurocognitive issues, as expected.

Listen, I agree with you that doing WBRT for 1-3 CNS mets routinely is not the right answer for me. How high you can take that 3 is currently a matter of debate. Some say 10. Some say it's volume dependent. Some others say more than 10 especially if radioresistant histology. I agree that SRS (either to everything, or to largest lesion with close follow-up if planned systemic therapy has a chance for intracranial presentation) would likely be my personal preference in this situation. However, if the patient got WBRT, it wouldn't be unreasonable.

As long as you MRI this patient q3 months you'll most likely catch additional stuff before it becomes symptomatic.

*EDIT* - Realized I linked the wrong trial from a window I had open - Effect of Radiosurgery Alone vs Radiosurgery With Whole Brain Radiation Therapy on Cognitive Function in Patients With 1 to 3 Brain Metastases: A R... - PubMed - NCBI is the Paul Brown trial

 

[MDACCRules]

Are you insinuating that we should never do WBRT? Or just in the 1-3 or 1-4 met group?

I'm not a fan of it just like you, but are you overarchingly claiming that because the primary end point in a trial is negative, all data from that trial should be ignored, since it's a negative trial?

Regardless, the Paul Brown trial did NOT show a detriment to overall survival with addition of WBRT in all comers. the P-value was 0.92 for that survival endpoint. In patients with the BEST GPA, it showed a survival benefit to addition of WBRT.
"Significantly better OS was observed in the DS-GPA 2.5-4.0 group in WBRT + SRS vs the SRS alone, with a median survival time of 16.7 (95% CI, 7.5-72.9) months vs 10.6 (95% CI, 7.7-15.5) months (P = .04) (hazard ratio

---

, 1.92; 95% CI, 1.01-3.78)." Stereotactic Radiosurgery With or Without Whole-Brain Radiotherapy for Brain Metastases: Secondary Analysis of the JROSG 99-1 Randomized Clinical T... - PubMed - NCBI
It did show increased neurocognitive issues, as expected.

Listen, I agree with you that doing WBRT for 1-3 CNS mets routinely is not the right answer for me. How high you can take that 3 is currently a matter of debate. Some say 10. Some say it's volume dependent. Some others say more than 10 especially if radioresistant histology. I agree that SRS (either to everything, or to largest lesion with close follow-up if planned systemic therapy has a chance for intracranial presentation) would likely be my personal preference in this situation. However, if the patient got WBRT, it wouldn't be unreasonable.

As long as you MRI this patient q3 months you'll most likely catch additional stuff before it becomes symptomatic.

---

Do you think a secondary trial analysis by post-stratification by DS-GPA is a good way to statistically analyze a result? I think this is where @scarbrtj can chime in about good science ...

 

[Gfunk6]

Poll- who makes affirmative treatment decisions based on negative trials?

This is a very narrow-minded view. As you know, there are many Radiation Oncology clinical trials that do not meet their primary outcomes due to statistical insignificance. As you also know, the response by the PIs is typically not, "Oh well, let's through the data in the trash and work on something else," but rather "let's look at the secondary end-points and perform unplanned subset analyses to see what we can extract of value."

These are all negative trials, statistically speaking. And we still use them to justify what we do.

 

[MDACCRules]

This is a very narrow-minded view. As you know, there are many Radiation Oncology clinical trials that do not meet their primary outcomes due to statistical insignificance. As you also know, the response by the PIs is typically not, "Oh well, let's through the data in the trash and work on something else," but rather "let's look at the secondary end-points and perform unplanned subset analyses to see what we can extract of value."

These are all negative trials, statistically speaking. And we still use them to justify what we do.

In the new age of meta-science, where people are finally analyzing methods and results, you're pushing the idea of chasing p-values to find something publishable? We went through an era of not publishing negative trials, publishing negative ones that are made positive by ignoring the primary endpoint, and just throwing a bunch of spaghetti on the wall to see if it sticks. Great! Now, all we have is p-value chasing, SEER/NCDB database analyses, and overwhelmingly publishing positive trials. Essentially, medical science is snorkeling through fatbergs - collections of debris, sewage, and used condoms to find a dull 1993 quarter. GO SCIENCE!

 

[Gfunk6]

In the new age of meta-science, where people are finally analyzing methods and results, you're pushing the idea of chasing p-values to find something publishable? We went through an era of not publishing negative trials, publishing negative ones that are made positive by ignoring the primary endpoint, and just throwing a bunch of spaghetti on the wall to see if it sticks. Great! Now, all we have is p-value chasing, SEER/NCDB database analyses, and overwhelmingly publishing positive trials. Essentially, medical science is snorkeling through fatbergs - collections of debris, sewage, and used condoms to find a dull 1993 quarter. GO SCIENCE!

I see. Do you happen to agree with your esteemed institutions position that protons are so much better than photons that it would be unethical to perform a clinical trial?

I guess we can take that on faith rather than look at SEER analysis of protons vs photons showing (at best) equivalence.

 

[MDACCRules]

I see. Do you happen to agree with your esteemed institutions position that protons are so much better than photons that it would be unethical to perform a clinical trial?

I guess we can take that on faith rather than look at SEER analysis of protons vs photons showing (at best) equivalence.

I vehemently disagree that it would be unethical. In fact, it would be ethical to perform a clinical trial. Don't ascribe all the Church's dogma to the individual members.

 

[evilbooyaa]

Do you think a secondary trial analysis by post-stratification by DS-GPA is a good way to statistically analyze a result? I think this is where @scarbrtj can chime in about good science ...

S

No, it's not ideal but it's better than no evidence. The enemy of good is perfect. However, it's not like I'm having that level of evidence dictate my clinical practice, given that I wouldn't routinely add WBRT to SRS in 1-3 mets.

EDIT* - Edited for the sake of harmony in the face of misunderstanding.

As discussed below, we are talking past each other by discussing two separate papers.

 

[MDACCRules]

No, it's not ideal but it's better than no evidence. The enemy of good is perfect. However, it's not like I'm having that level of evidence dictate my clinical practice, given that I wouldn't routinely add WBRT to SRS in 1-3 mets.

Regardless, you stated there was a survival detriment of WBRT in your previous post (the reason I went and looked up the Paul Brown paper again). I'm here to tell you that your knowledge of that data is incorrect. If you want to continue being defensive and/or throwing out non-sequiturs for the sake of it rather than just admitting that your statement was incorrect, then so be it.

ARE YOU JOKING? DID YOU NOT READ THE MDACC PAPER? THERE WAS A STATISICALLY SIGNIFICANT SURVIVAL DERTRIMENT.

Neurocognition in patients with brain metastases treated with radiosurgery or radiosurgery plus whole-brain irradiation: a randomised controlled tr... - PubMed - NCBI

Excuse you! I don't know if you could be more wrong. I explained succinctly above that 2 papers were negative and one showed a detriment. Chang presented this at ASTRO with a smirk on his face, when you were in the infancy of medical school, not knowing the difference between a photon and a protein. Not being defensive. Not throwing non-sequiturs.

"rather than just admitting that your statement was incorrect" - do you plan on addressing this?

There are 3 papers. EORTC. MDACC. Japan. Please read all 3. You're going to fail your boards if you don't know the landmark trials. Wow. The quality of the residents may well be slipping ..

 

[Gfunk6]

ARE YOU JOKING? DID YOU NOT READ THE MDACC PAPER? THERE WAS A STATISICALLY SIGNIFICANT SURVIVAL DERTRIMENT.

Neurocognition in patients with brain metastases treated with radiosurgery or radiosurgery plus whole-brain irradiation: a randomised controlled tr... - PubMed - NCBI

Excuse you! I don't know if you could be more wrong. I explained succinctly above that 2 papers were negative and one showed a detriment. Chang presented this at ASTRO with a smirk on his face, when you were in the infancy of medical school, not knowing the difference between a photon and a protein. Not being defensive. Not throwing non-sequiturs.

"rather than just admitting that your statement was incorrect" - do you plan on addressing this?

There are 3 papers. EORTC. MDACC. Japan. Please read all 3. You're going to fail your boards if you don't know the landmark trials. Wow. The quality of the residents may well be slipping ..

Overall survival was a secondary endpoint of this trial. Neurocognitive function was the primary endpoint. It closed early after enrolling a mere fraction of patients because the statisticians said that there was a 96% probability that the primary endpoint (e.g. neurocognitive function) was met. Therefore, while you can claim that SRS > WBRT + SRS improves neurocognitive outcomes, you cannot extrapolate to survival.

Also, the WBRT was combined with SRS and not delivered as monotherapy (WBRT) alone.

In my own practice I favor SRS > WBRT in the vast majority of situations. However, there are cases of patients with innumerable mets or leptomeningeal disease. In those cases I treat with hippocampal sparing WBRT or WBRT with memantine.

 

[evilbooyaa]

When you previously said MDACC trial about whole brain, I was assuming you're talking about the multi-institutional phase III trial with the lead author who is from MDACC. The one (to me) that's clinically relevant now, published in 2016, has 213 patients and makes the statements I cited.
But you would rather hang your hat instead on a single-institution paper that's 7 years older, has 1/4 the number of patients, and say there's a difference in survival across 40 patients, in a trial that was stopped early?

So yeah, I guess that small single institution trial from 10 years ago did show a univariate survival difference in a barebones paper that focused on neurocognitive decline being worse with WBRT (which I agree with) but didn't get into stringest details of why SRS patients would live longer than WBRT. That trial has been repeated in a multi institution setting with 4x the number of patients, see above, and didn't corroborate those findings. Might be worth it to keep on reading and realize that more data comes out after you finish residency.

Anyways, here's the quote from the discussion on why SRS patients magically lived longer than WBRT patients in this small data set:
"Based on a post-hoc analysis of the timing of systemic therapy in our study, patients assigned to the SRS alone group received systemic therapy over 1 month earlier and for a median of two more cycles than patients assigned to SRS plus WBRT. These exploratory findings suggest that more prompt systemic therapy might have contributed to the prolonged survival of patients in the SRS alone group. Another possible explanation for the survival difference is that patients randomly assigned to SRS plus WBRT had a greater systemic disease burden than their counterparts in the SRS alone group at the time of enrollment."

If the authors of the article aren't saying "lol WBRT kills people guys" then why are you saying it?

When I go to boards I'm going to go with the Paul Brown numbers. Who cares about small single institution studies when you have a phase 3 multi-institutional RCT that asks and answers the same question?

I think it's really funny you call these the 'landmark trials' like anybody would let your statement of "WBRT causes more deaths in people with 1-3 mets if given with SRS than SRS alone" slide on oral boards, instead of just saying "WBRT causes neurocognitive decline, based on every study ever that has looked into it"

Regardless, I apologize, it was clearly a misunderstanding. When you said MDACC whole brain paper, I figured you meant the Paul Brown one. You did not. However, I invite you to move past single institution trials and embrace the multi-institutional trials and bigger sample sizes that have asked and answered the question.

 

[MDACCRules]

Yes, I said this. All the trials being discussed are WBRT+SRS vs SRS alone. Nobody is talking about a control arm with WBRT alone for 1-3 mets. That's just ridiculous in modern era. The addition of WBRT to SRS vs SRS alone messes up your brain and possibly kills you. That's the working hypothesis. We should work on improving cognition, and improving survival. I'm not sure we have any evidence that WBRT improves QOL or survival.

I think that the above comment about knowledge was quite uncalled for, and I would ask the moderator to discuss insulting members. It is clear on that study that the primary endpoint was met, and that survival was lower in the WBRT+SRS arm vs the SRS alone arm. Isn't that true? Does it give you pause or do you ignore that finding that there was a several month difference in OS? Maybe their neurocognitive outcome was poor because they were dead. Who knows? That study's outcome, and the other negative studies should concern us, shouldn't it?

Hippocampal sparing brain radiation is by definition not whole brain, and sounds like a reasonable strategy. Memantine is eh, but it's generic, at least.

 

[evilbooyaa]

Yes, I said this. All the trials being discussed are WBRT+SRS vs SRS alone. Nobody is talking about a control arm with WBRT alone for 1-3 mets. That's just ridiculous in modern era. The addition of WBRT to SRS vs SRS alone messes up your brain and possibly kills you. That's the working hypothesis. We should work on improving cognition, and improving survival. I'm not sure we have any evidence that WBRT improves QOL or survival.

I think that the above comment about knowledge was quite uncalled for, and I would ask the moderator to discuss insulting members. It is clear on that study that the primary endpoint was met, and that survival was lower in the WBRT+SRS arm vs the SRS alone arm. Isn't that true? Does it give you pause or do you ignore that finding that there was a several month difference in OS? Maybe their neurocognitive outcome was poor because they were dead. Who knows? That study's outcome, and the other negative studies should concern us, shouldn't it?

Hippocampal sparing brain radiation is by definition not whole brain, and sounds like a reasonable strategy. Memantine is eh, but it's generic, at least.

I have edited my previous post in response to the fact that we were talking past each other, and discussing the outcome of two different trials that could (at least to me) both be considered 'the MDACC trial'.

 

[Cancerdancer]

WBRT for asymptomatic lesions has never shown a survival benefit in a trial...

Neither has radiation for stage III NSCLC. Do you also just default to chemo alone for such patients?

 

[seper]

Not true, there is a study showing 60 Gy improves survival over 40 Gy.

Neither has radiation for stage III NSCLC. Do you also just default to chemo alone for such patients?

 

[Palex80]

Neither has radiation for stage III NSCLC. Do you also just default to chemo alone for such patients?

Role of radiotherapy in combined modality treatment of locally advanced non-small-cell lung cancer. - PubMed - NCBI

 

[RickyScott]

We have many (at least 6-8)large phase III studies in NSLC showing 5 year survival of 15-20% in stage III NSLC. I dont think there is a single one with chemo showing anything in the ball park.

 

[Cancerdancer]

I'm pretty sure Seper is thinking of RTOG 7301, which showed improved local control, not overall survival.

To Palex, who linked the Kubota study (which was powered to assess a difference in response rate, and was negative for that primary endpoint):

"There were no significant differences found between the RT/CT and CT groups in terms of overall survival, (Fig 1). The median survival duration was 461 days in the CT/RT group and 447 days in the CT group. However, the 2- and 3-year survival rates were significantly different, with 35.5% 2-year and 29.0% 3-year survival in the CT/RT group and only 9.4% and 3.1%, respectively, in the CT group (P = .016 and .0049, respectively)."

Pretty funky statistics giving P-values at different timepoints along curves that cross among 63 patients. I think we'd all call that a (-) analysis for OS. Note the abstract says RT "increases the number of long-term survivors" not improves OS.

While I believe RT is beneficial, and yes we have 5-yr OS data of 15-20% and chemo studies don't, my point was we often utilize therapies (and believe in their value) without great evidence.