hippocampal sparing for PCI in limited stage SCLC

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

Kroll2013

Full Member
10+ Year Member
Joined
Jan 18, 2013
Messages
152
Reaction score
15
I know that it is not standard yet , but is anyone doing it ?

ty

Members don't see this ad.
 
I do it on study; they had closed the NRG study down for interim analysis but it's back open.

I haven't done it off study yet; can people get it approved off study by insurance?
 
Members don't see this ad :)
I have done it two or three times off study in elderly pts. Recently had a case where medonc is blaming severe dementia on PCI- pt received regular whole brain.
 
CC-003 currently asking the question. I probably wouldn't offer it off protocol as of yet, but we frequently extrapolate the Japanese data to LS-SCLC so we're not generally doing a ton of PCI.
 
If pursuing PCI at all, then in my view, the data is currently stronger for memantine (benefit in RCT) than for HA-WBRT (benefit compared to historical control). So, would at least offer memantine.

Re: PCI: in LS-SCLC, it's still a potentially curative situation, and you're obviously underdosing part of the brain (10%?) in HA-WBRT, so my worry would be that you're compromising the chance of cure. Realistically, though, the only circumstance in which a pt would be harmed is if they had an isolated CNS failure in the underdosed region. Not sure what the rate of isolated CNS failure in LS-SCLC is, but multiply that by the 10% of brain underdosed and it's probably only a few % of pts who would be affected. And even that minority is highly salvageable with SRS. Moreover, if you believe the benefit of HA-WBRT, then most patients probably derive some neurocognitive benefit, which has to be weighed against the few % chance of harm.

So, while I think the proper answer is that HA-PCI should be done only on trial, practically speaking I think it does seem a very attractive proposition.
 
If pursuing PCI at all, then in my view, the data is currently stronger for memantine (benefit in RCT) than for HA-WBRT (benefit compared to historical control). So, would at least offer memantine.

Re: PCI: in LS-SCLC, it's still a potentially curative situation, and you're obviously underdosing part of the brain (10%?) in HA-WBRT, so my worry would be that you're compromising the chance of cure. Realistically, though, the only circumstance in which a pt would be harmed is if they had an isolated CNS failure in the underdosed region. Not sure what the rate of isolated CNS failure in LS-SCLC is, but multiply that by the 10% of brain underdosed and it's probably only a few % of pts who would be affected. And even that minority is highly salvageable with SRS. Moreover, if you believe the benefit of HA-WBRT, then most patients probably derive some neurocognitive benefit, which has to be weighed against the few % chance of harm.

So, while I think the proper answer is that HA-PCI should be done only on trial, practically speaking I think it does seem a very attractive proposition.

CC-001 (randomized trial) presented at ASTRO showing benefit of HA-WBRT compared to WBRT even when all patients receive memantine (Hippocampus-Sparing Radiation a New Option in Brain Mets Patients)

Otherwise, agree with rest of your post.
 
  • Like
Reactions: 2 users
I know that it is not standard yet , but is anyone doing it ?

ty
I do it on study; they had closed the NRG study down for interim analysis but it's back open.

I haven't done it off study yet; can people get it approved off study by insurance?
It is absolutely standard. It's standard like postop chemoRT for gastric became standard after a single study, or definitive chemoRT became standard for OPSCC after a single study, or Tmz became standard after a single study, 8 Gy became standard palliative dose after a single study, etc etc. and ad infinitum.

For Medicare patients as you know there is no approval per se; you bill the IMRT, you get paid the IMRT. There are no IMRT site specific limitations... there is simply medical necessity, or there is not. For private insurance: a different ballgame.


Although not yet published.
Wonder why this is still "a thing." One, the journal vs. ASTRO abstract review process is not obviously more robust. Two, it has technically been published. Three, odds of non-publication of this are zero. Four, you have knowledge of the outcome: you believe it, or you don't. #contrary
 
Last edited:
As an aside, with non-prophylactic hippocampal sparing WBRT, anyone giving a higher dose to gtv/tumors? I have been asking for 10% to gross disease over 1cm. Seems like you may as well since it is IMRT.
 
It is absolutely standard. It's standard like postop chemoRT for gastric became standard after a single study, or definitive chemoRT became standard for OPSCC after a single study, or Tmz became standard after a single study, 8 Gy became standard palliative dose after a single study, etc etc. and ad infinitum.

For Medicare patients as you know there is no approval per se; you bill the IMRT, you get paid the IMRT. There are no IMRT site specific limitations... there is simply medical necessity, or there is not. For private insurance: a different ballgame.



Wonder why this is still "a thing." One, the journal vs. ASTRO abstract review process is not obviously more robust. Two, it has technically been published. Three, odds of non-publication of this are zero. Four, you have knowledge of the outcome: you believe it, or you don't. #contrary

Nope and nope.

1) You may argue that therapeutic HA-WBRT is standard for non-SCLC pts (SCLC was excluded from CC001). HA-PCI for SCLC is pretty clearly not the same thing. Maybe the results will be the same, maybe they won’t, but it’s an open enough question that we’re doing a trial about it.

2) The distinction between abstract and paper is “still a thing” because peer review is still a thing. Your assertion that peer review for publication of a PhIII trial is no more rigorous than that of an ASTRO abstract is... interesting.

As an aside, with non-prophylactic hippocampal sparing WBRT, anyone giving a higher dose to gtv/tumors? I have been asking for 10% to gross disease over 1cm. Seems like you may as well since it is IMRT.

Not doing it myself. On the one hand, giving more dose to gross disease than microscopic disease is certainly appealing. But this means you’re pre-emptively boosting everything, when most lesions will never fail with standard dosing. And if some do fail later, can always salvage w SRS. I do contour the mets and confirm coverage, though.
 
Last edited:
  • Like
Reactions: 1 users
Last edited:
Wonder why this is still "a thing." One, the journal vs. ASTRO abstract review process is not obviously more robust. Two, it has technically been published. Three, odds of non-publication of this are zero. Four, you have knowledge of the outcome: you believe it, or you don't. #contrary[/QUOTE]

I am going to push back against a few points-
I think it is arguable that peer review for publication is more robust than abstract review for a meeting; I am not saying that either process is without error. I have participated actively in both forms of peer review for more than two decades and peer review for a journal is more involved.

Additionally there is some evidence that the "final product" is discordant from the meeting abstract; occasionally very much so.

Presentation of nonfinal results of randomized controlled trials at major oncology meetings. - PubMed - NCBI

Of course this is radiation and we are not regulated in the same manner as drugs but in this case it might be prudent to wait for publication.
 
  • Like
Reactions: 1 users
Members don't see this ad :)
Pretty sure Medicare has a list of icd10 codes they will pay imrt for. If it is not on that list you'll get a denial, but there is an appeal process

https://www.google.com/url?sa=t&sou...FjAAegQIAxAB&usg=AOvVaw34ZxOUAS3TE4EfZXoq5Ezo
That link shows that IMRT is covered for every single cancer ICD10 diagnosis known to man and a few benign conditions too; that said, brain mets are covered on that list. Used to in the "old days" there was a restrictive dx list. Then, there were no codes listed as restrictive. Now I see they're listing every single (cancer) diagnosis code as covered (e.g. LCIS is covered?).
I am going to push back against a few points-
I think it is arguable that peer review for publication is more robust than abstract review for a meeting; I am not saying that either process is without error. I have participated actively in both forms of peer review for more than two decades and peer review for a journal is more involved.

Additionally there is some evidence that the "final product" is discordant from the meeting abstract; occasionally very much so.
I'm glad you're pushing. I was being intentionally provocative. I've been involved in the peer review of ASTRO abstracts, JCO, Red Journal, etc. And yes on one hand journal review is more involved (writing out your thoughts/make a nice reviewer writeup etc vs abstract is just a binary decision). On the other hand, the hurdle to full-throated publication is quite less hurdle-some after you've got a late-breaking abstract at ASTRO of phase III trial with a pretty low p-value finding. If anything, I'm telling people who don't see behind the scenes that there's not some secret sauce that goes into journal peer review MUCH more so that abstract peer review. Perhaps what I should say is a researcher should put as much thought, effort, math, science and ethics into an abstract publication as he or she does into a reviewed journal article so that us poor readers can have equal faith in both; I rely on that being true as a reviewer or reader. I like your pub link about abstract/publication discordance for systemic therapies. If I'm reading it right, there's a ~10% really bad discordance rate. If you go 2013-2016, top ~2 abstracts from ASTRO each year (RTOG 9910, RTOG 0841, DART01/05 GICOR, RTOG 0621, COG ACNS0121*, RTOG 0617, Chinese Stage IV esophageal trial, UK Vortex Sarcoma*), 6 studies have been confirmed w/ journal pubs and 2* are unpublished. So. I'm a practice-changer, more often than not, for sockdolager phase III radiation abstracts. AFAIK, I've not had to retreat to old methods because of an abstract flip-flop. When is a standard a standard? When Whelan's 5y data for breast hypofractionation came out in full publication form, everyone I knew at that time said "Nope, not standard." I said it was standard, started offering it. I started offering it when his abstract (for 5y data) appeared! That was *really* controversial.
1) You may argue that therapeutic HA-WBRT is standard for non-SCLC pts (SCLC was excluded from CC001). HA-PCI for SCLC is pretty clearly not the same thing.
I totally mixed my metaphors and forgot that SCLC patients were excluded from CC-001, a "brain mets" trial. (And forgot OP was asking about SCLC PCI.) So it is acceptable to say it's not a standard for SCLC brain mets *or* SCLC PCI for that reason. But for those who want to "extrapolate"... PCI is not prophylactic and is a misnomer. Radiation doesn't prevent malignant seeding of normal tissue; it can only work by treating a malignancy (or maybe causing abscopal effects). So PCI is really treatment of subclinical brain mets and not "prophylactic" any more than whole breast RT post-lumpectomy is prophylactic, postop XRT for anything is not prophylactic etc.
 
The benefit of HA is expected to be the greatest with longer follow up. So, LD–SCLC patients should in theory benefit the most, since 25% of them should be alive 5 years post PCI.
 
  • Like
Reactions: 1 user
As an aside, with non-prophylactic hippocampal sparing WBRT, anyone giving a higher dose to gtv/tumors? I have been asking for 10% to gross disease over 1cm. Seems like you may as well since it is IMRT.

Yes. But why not boost it more than 3 Gy (10%)? Or did I misunderstand? What are you really accomplishing with 33 Gy vs. 30 Gy?

30 Gy with hippocampal avoidance. 10 Gy SIB to gross dz + 2 mm PTV margin. VMAT. Multiple boost targets fine. Would consider boosting any gross dz, not just 1 cm+. Don't see the logic in boosting a 1.3 cm met, but leaving a 0.9 cm and 0.7 cm met alone. Boost all 3 or none at all.

Agree that if you are going to use VMAT, then make the most of it if circumstances make sense (and will probably make sense a lot with sites like lung going forward in terms of going after more durable control).

Not what I'm going to say on boards, but there's a lot of small series out there looking at various SIB-IMRT techniques with whole brain.
 
Last edited:
That link shows that IMRT is covered for every single cancer ICD10 diagnosis known to man and a few benign conditions too; that said, brain mets are covered on that list. Used to in the "old days" there was a restrictive dx list. Then, there were no codes listed as restrictive. Now I see they're listing every single (cancer) diagnosis code as covered (e.g. LCIS is covered?).
.
Sure but I don't think the pci code is in there. Pci isn't the same code as brain mets
 
But why not boost it more than 3 Gy (10%)?
So, I still don't agree that one should do SIB to gross disease for the reasons I listed above. But if you were going to, I agree that boosting by only 3 Gy seems too little. I don't know what the right number is (not sure anyone really does - grateful for thoughts and the rationale behind them) but more than 3 Gy seems rational. In a sense, this is recapitulating the RTOG 9508 approach without using SRS. Keep in mind that was 37.5 Gy WBRT + 15-24 Gy SRS boost and showed a LC benefit to the treated lesion but not much else (though possibly OS benefit in the best RPA/GPA pts). I think it's unlikely that addition of SIB to HA-WBRT would show more benefit than addition of SRS to conventional WBRT.

2 cm PTV margin
2 cm PTV??
 
Yes. But why not boost it more than 3 Gy (10%)? Or did I misunderstand? What are you really accomplishing with 33 Gy vs. 30 Gy?

30 Gy with hippocampal avoidance. 10 Gy SIB to gross dz + 2 cm PTV margin. VMAT. Multiple boost targets fine. Would consider boosting any gross dz, not just 1 cm+. Don't see the logic in boosting a 1.3 cm met, but leaving a 0.9 cm and 0.7 cm met alone. Boost all 3 or none at all.

Agree that if you are going to use VMAT, then make the most of it if circumstances make sense (and will probably make sense a lot with sites like lung going forward in terms of going after more durable control).

Not what I'm going to say on boards, but there's a lot of small series out there looking at various SIB-IMRT techniques with whole brain.

These plans are very hetreogenous and asking for 10% is mainly to obtain 100% coverage of gtvs, and ends up 3-5% and try to force hotspots into gtv with some rings. i guess the question is, should I be asking for more- I dont know. I wouldnt compare it to standard whole brain trials, because usually selecting pts who you hope to have better prognosis and hopefully wont be dead in 4-5 months, which is the average survival in whole brain.
 
Last edited:
So, I still don't agree that one should do SIB to gross disease for the reasons I listed above. But if you were going to, I agree that boosting by only 3 Gy seems too little. I don't know what the right number is (not sure anyone really does - grateful for thoughts and the rationale behind them) but more than 3 Gy seems rational. In a sense, this is recapitulating the RTOG 9508 approach without using SRS. Keep in mind that was 37.5 Gy WBRT + 15-24 Gy SRS boost and showed a LC benefit to the treated lesion but not much else (though possibly OS benefit in the best RPA/GPA pts). I think it's unlikely that addition of SIB to HA-WBRT would show more benefit than addition of SRS to conventional WBRT.


2 cm PTV??

Sorry, typo meant 2 mm. If you're using an SRS mask instead of an IMRT whole brain mask, you could do 1 mm, but no need to do that. I would use the IMRT reinforced mask instead of the standard whole brain mask though.

Agree that the goal here is LC, not OS. And think LC may be more important as control of extracranial disease improves with new molecular agents. Advantage of doing this is it allows you to boost gross disease with no extra treatments like SRS and if you're already doing VMAT anyway for hippocampal sparing, it makes sense (to me at least). Dosimetrist may not like it.
 
These plans are very hetreogenous and asking for 10% is mainly to obtain 100% coverage of gtvs, and ends up 3-5% and try to force hotspots into gtv with some rings. i guess the question is, should I be asking for more- I dont know. I wouldnt compare it to standard whole brain trials, because usually selecting pts who you hope to have better prognosis and hopefully wont be dead in 4-5 months, which is the average survival in whole brain.

I remember seeing this on a patient. LS-SCLC. Got prior PCI 5 years ago. Did fine (really, excellent) until showed up 4 or 5 new brain mets. No other sites of disease. The decision was to retreat brain to 20 Gy with hippocampal avoidance and SIB to gross dz. If it were one met, maybe would have considered SRS. Don't know in small cell about doing SRS to multiple mets. Anyway, not your typical 4-5 month OS WBRT patient.
 
Sorry, typo meant 2 mm. If you're using an SRS mask instead of an IMRT whole brain mask, you could do 1 mm, but no need to do that. I would use the IMRT reinforced mask instead of the standard whole brain mask though.

Agree that the goal here is LC, not OS. And think LC may be more important as control of extracranial disease improves with new molecular agents. Advantage of doing this is it allows you to boost gross disease with no extra treatments like SRS and if you're already doing VMAT anyway for hippocampal sparing, it makes sense (to me at least). Dosimetrist may not like it.

Yeah, it’s just that you’re treating everything now, when many/most lesions would be controlled without the SIB. Personally, I’d rather wait and SRS later only those lesions that fail. But, no real data here to compare that I’m aware of, so I suppose either is reasonable. I think this would be a great question for a follow up trial: memantine + HA-WBRT, randomize to SIB to gross disease (with dose escalation design, likely graduated by size) vs no SIB and SRS salvage. Personally, I suspect the SIB arm would need less salvage SRS, but no other difference in outcomes since salvage SRS is effective and fairly non-toxic.
 
Whole brain doesnt controls most lesions, the patient just dies. 300 x 10 does not eliminate gross disease.
 
Whole brain doesnt controls most lesions, the patient just dies. 300 x 10 does not eliminate gross disease.

Sure, but 3.3 x 10 as you propose doesn't eliminate gross disease either.

In the SRS era, the pts still getting WBRT are generally getting it b/c they're not eligible for SRS for some reason, which usually means poor prognosis in one way or another. In these pts, I don't think the point of WBRT is to eliminate gross disease. It is just to control it for long enough to avoid neurologic death or disability until--as you say--they die of competing extracranial risks. The "right" dose is the minimal dose necessary to achieve that. If you guess too high and cause toxicity, not like you can take back dose later. If you guess too low, you can generally salvage with SRS.

Like I say, in the absence of real data, I think either approach is reasonable, but I do think there are considerations on both sides.
 
  • Like
Reactions: 1 user
Sure but I don't think the pci code is in there. Pci isn't the same code as brain mets
"PCI" only works if there are mets in the brain. (In the pre-PCI era, ~50-70% of patients at death felt to be brain mets-free actually had brain mets at autopsy.) There is no such thing as prophylactic irradiation (well, verbally there is... "in reality" no). So IMHO you have to code brain mets when you do whole brain RT of any sort. Prophylaxis: prevents a disease from occurring. You can't give a pregnancy prophylaxis to a pregnant woman e.g. Else, one is by way of insisting on the whole brain RT being prophylactic that the brain is totally metastasis-free; ergo, insisting on irradiating a completely normal, non-cancerous brain (bad). Am I heretical? Yes. Also, if a trial shows HA-PCI is good for SCLC, Medicare will (likely) not adjust its diagnosis codes to "allow" for IMRT (ie add in the PCI C-code), at least not right away. So what then?
 
"PCI" only works if there are mets in the brain. (In the pre-PCI era, ~50-70% of patients at death felt to be brain mets-free actually had brain mets at autopsy.) There is no such thing as prophylactic irradiation (well, verbally there is... "in reality" no). So IMHO you have to code brain mets when you do whole brain RT of any sort. Prophylaxis: prevents a disease from occurring. You can't give a pregnancy prophylaxis to a pregnant woman e.g. Else, one is by way of insisting on the whole brain RT being prophylactic that the brain is totally metastasis-free; ergo, insisting on irradiating a completely normal, non-cancerous brain (bad). Am I heretical? Yes. Also, if a trial shows HA-PCI is good for SCLC, Medicare will (likely) not adjust its diagnosis codes to "allow" for IMRT (ie add in the PCI C-code), at least not right away. So what then?

Agree with this. Unfortunately, Slotman study set us back 20 years in terms of managing intracranial disease for SCLC. With the availability of q3-4month MRI surveillance, there is just no reason to do PCI et al. It is bad oncology and bad medicine.

Intracranially stage the patient, if they're negative, surveil them; if they're positive, treat them with WBRT, or in select cases even offer them SRS. Save WBRT for when you need an intracranial reset.
 
  • Like
Reactions: 1 users
If pursuing PCI at all, then in my view, the data is currently stronger for memantine (benefit in RCT) than for HA-WBRT (benefit compared to historical control). So, would at least offer memantine.

Re: PCI: in LS-SCLC, it's still a potentially curative situation, and you're obviously underdosing part of the brain (10%?) in HA-WBRT, so my worry would be that you're compromising the chance of cure. Realistically, though, the only circumstance in which a pt would be harmed is if they had an isolated CNS failure in the underdosed region. Not sure what the rate of isolated CNS failure in LS-SCLC is, but multiply that by the 10% of brain underdosed and it's probably only a few % of pts who would be affected. And even that minority is highly salvageable with SRS. Moreover, if you believe the benefit of HA-WBRT, then most patients probably derive some neurocognitive benefit, which has to be weighed against the few % chance of harm.

So, while I think the proper answer is that HA-PCI should be done only on trial, practically speaking I think it does seem a very attractive proposition.

In the setting of brain mets (not LS-SCLC) I found this data here in the dicussion part of RTOG 0933

"Conformal avoidance of the hippocampus poses the risk of attenuating the benefit of WBRTbecause of the emergence of new brain
metastases within the hippocampal avoidance region. Prior estimates have approximated this risk at 8.6%"

It's a little different situation I know, but can give us a ball park figure.

In regards to doing it off trial, I have been tempted, but also tempted to do close serial MRI observations (have not yet). I have done memantine with bleh results. Memantine trial did fail its main endpoint in its study...

The one thing I am concerned about doing it off trial is this part:

To yield acceptable image quality, the pre-contrast-enhanced should have a resolution of
1 x 1 x 1.2 mm and should follow the protocols established by the Alzheimer’s Disease
Neuroimaging Initiative (ADNI).

I personally think that is unnecessary for hippocampal contouring.

PCI for SCLC is so bizarre and very unsettling to know we are treating people's brain that have no cancer in them and for those that do a very pitiful dose. Yes, I know we do elective treatment volumes and give systemic therapy without knowing, but PCI feels strange (not just me right?).
 
To yield acceptable image quality, the pre-contrast-enhanced should have a resolution of
1 x 1 x 1.2 mm and should follow the protocols established by the Alzheimer’s Disease
Neuroimaging Initiative (ADNI).

I personally think that is unnecessary for hippocampal contouring.

I agree you don't need 1 mm 3D T1 pre sequence for contouring the hippocampus. But I do want a 1 mm 3D T1 post sequence (e.g., MPRAGE) for finding and contouring mets.

Prior estimates have approximated this risk at 8.6%"
Nice find--thanks.

"PCI" only works if there are mets in the brain.

PCI for SCLC is so bizarre and very unsettling to know we are treating people's brain that have no cancer in them and for those that do a very pitiful dose. Yes, I know we do elective treatment volumes and give systemic therapy without knowing, but PCI feels strange (not just me right?).

It's all about your definition of "met." Unless you think PCI is doing something to the brain to inhibit future seeding from systemic sites, then yes, it's clearly treating existing disease in the brain. I don't think it's controversial that there are likely to be occult tumor cells in the brain below the threshold of detection and that that's what PCI is treating. That's presumably also why WBRT has lower rates of distant brain failure than SRS. Like Megatron says, I don't see how this is different from treating elective volumes in any other disease site (where again, there must already be tumor cells for RT to have benefit). If we're parsing words carefully, one could still say PCI is still prophylactic in that it prevents the development of clinically or radiographically evident mets.
 
PCI feels strange (not just me right?).
The very definition of agita. But like Jim Morrison said: people are strange. Heck, to me, the most "prophylactic" of all things we do is give whole breast RT to an older woman with Stage I breast CA post-lumpectomy. We do this for a ~3% absolute local control benefit. In other words, we treat 33 old ladies to help 1. The number needed to treat is not that bad for "prophylactic" cranial irradiation for small cell lung cancer, which also gives a survival benefit. (Again, it is not prophylactic; talk amongst yourselves.)
 
Last edited:
Last edited:
Yet we have data documenting an OS benefit for it, and we know the natural history of sclc is a really high intracranial relapse rate long term

Indeed. I've been thinking about the epistemology of clinical trials a lot lately. Expert opinion, retrospective data, and phase II data should not trump phase III trial results, but this happens a lot in oncology. I still give PCI b/c no matter what my hesitations are due to the aforementioned trials.

RTOG 0617 update with a 8 mo median survival difference (28.7 vs 20.3) and recent RTOG 1016 + European data on cis vs cetux - definitely made me respect phase III randomized trial data.
 
Indeed. I've been thinking about the epistemology of clinical trials a lot lately. Expert opinion, retrospective data, and phase II data should not trump phase III trial results, but this happens a lot in oncology. I still give PCI b/c no matter what my hesitations are due to the aforementioned trials.

In the community, it's borderline malpractice imo to not give it esp in LS disease. Still a cat 1 rec by the nccn. I also think it should be given in ES patients with a CR everywhere else

I've had a few patients refuse and I ended up getting called 6-12 mos later to give "T" CI when they had multiple mets
 
Last edited:
  • Like
Reactions: 1 user
In ES disease, I believe PCI is now optional in view of recent phase III study.

In the community, it's borderline malpractice imo to not give it esp in LS disease. Still a cat 1 rec by the nccn. I also think it should be given in ES patients with a CR everywhere else

I've had a few patients refuse and I ended up getting called 6-12 mos later to give "T" CI when they had multiple mets
 
  • Like
Reactions: 1 user
In the community, it's borderline malpractice imo to not give it esp in LS disease. Still a cat 1 rec by the nccn. I also think it should be given in ES patients with a CR everywhere else

I've had a few patients refuse and I ended up getting called 6-12 mos later to give "T" CI when they had multiple mets

While it is a recommended therapy, NCCN also makes allowances for close surveillance with MRI Brain q3 month (assuming that initial pre-treatment MRI is negative) based on the Japanese data.
 
While it is a recommended therapy, NCCN also makes allowances for close surveillance with MRI Brain q3 month (assuming that initial pre-treatment MRI is negative) based on the Japanese data.

Only in ES disease. Limited remains PCI, category 1 no option for surveillance.

Personally, my gut reaction is to treat good PS patients with a CR and ES disease the same way. Japanese data included patients with any response.
 
Last edited:
Most of my patients and insurers don't give me the choice.

I beg and plead many patients to get WBRT (with or without hippocampal sparing), even when it's strongly indicated. In the past 6 months I have had two patients die from progressive brain metastases rather than receive whole brain radiotherapy. When it comes to WBRT or PCI, many patients go to Google and search whole brain radiation, and that's the end of it. The patient communities have spoken--WBRT will turn you into a vegetable and kill you, avoid at all costs. This has led me to numerous 60+ minute conversations.

Regarding the data: I will be enrolling on NRG CC003 which just reopened for the phase III component. Outside of trial I am not offering hippocampal sparing for PCI.

I suppose if I had a patient who demanded hippocampal sparing PCI or wouldn't get PCI at all, I would consider hippocampal avoidance if insurance lets me. That written, my patients don't come in asking for hippocampal avoidance, and when I try to explain it to them it doesn't really seem to sway them. It's still whole brain, and to be avoided at all costs.

I saw the CC001 data presented in abstract form twice. That's good enough for me. Some insurance reviewers are demanding publication first. It seems to me the private payers are just trying to stall to save a few bucks.
 
  • Like
Reactions: 1 users
Yet we have data documenting an OS benefit for it, and we know the natural history of sclc is a really high intracranial relapse rate long term

Only in ES disease. Limited remains PCI, category 1 no option for surveillance.

This is absolutely correct but with significant caveats. The category 1 recommendation is based on CT era studies without routine CT surveillance for those that did not get PCI.

So the data we have tells us today that if our patient has SCLC and can only have one staging CT head at diagnosis without routine CT surveillance then there is an overall survival advantage to PCI.

The one MRI era study in ES-SCLC was completely negative and shut down early for futility. It was actually trending toward a survival detriment.

Obviously it is still a category 1 recommendation for all LS-SCLC patients. I expect the Japanese to help us end this debate. I think we will look back soon and be embarrassed that we were treating patients without brain metastasis for so long.
 
  • Like
Reactions: 1 users
This is absolutely correct but with significant caveats. The category 1 recommendation is based on CT era studies without routine CT surveillance for those that did not get PCI.

So the data we have tells us today that if our patient has SCLC and can only have one staging CT head at diagnosis without routine CT surveillance then there is an overall survival advantage to PCI.

The one MRI era study in ES-SCLC was completely negative and shut down early for futility. It was actually trending toward a survival detriment.

Obviously it is still a category 1 recommendation for all LS-SCLC patients. I expect the Japanese to help us end this debate. I think we will look back soon and be embarrassed that we were treating patients without brain metastasis for so long.

The question is:

How many LD-SCLC patients in complete remission after thoracic CRT and a clean MRI would develop isolated brain metastases in the next years without PCI?

I am talking about sole relapse with brain mets, not liver mets and brain mets at the same time.
We do not really know the answer to that. Why? Because everybody gives PCI nowadays and because the earlier studies did not perform an MRI after completion of thoracic CRT.

One small study suggests, that a relevant amount of patients may actually develop brain mets during CRT, so patients getting PCI without a prior scan immediately prior to PCI may be actually getting therapeutic WBRT and not PCI.
https://www.sciencedirect.com/science/article/pii/S1556086415314908
In short: 105 patients with LD-SCLC staged with MRI initially (no brain mets) --> CRT --> 40 of them in complete remission in CT --> second MRI prior to planned PCI --> 13 of them found to have brain mets...


However...
If you think that the rate of LD-SCLC in CR after thoracic CRT and with a clean MRI that receive observation and not PCI is >10%, then I think PCI is worth it.
Brain mets in SCLC seldom arise as isolated lesions and it is going to be hard to effectively "salvage" more than half of these relapsing patients with radiosurgery, WBRT +/- systemic treatment. You would have missed your chance to perform PCI and that's it. But the magnitude will not be big, perhaps 5%? But everything additional we do in LD-SCLC is "worth" 5% itself, which results in the ominous 25% OS rate at 5 years...
 
  • Like
Reactions: 1 user
The History of PCI would make for a good Lifetime TV movie. Fraught with controversy, passion, and interesting characters. I remember doing 36-40 Gy whole brain for PCI way back when because it seemed like the "best" dose. Someone wrote above they feel "strange" when doing PCI, ie giving XRT to a "normal" brain. If true we should feel strange all the time. With no hint of irony radiation oncologists do not feel strange when they electively ("prophylactically?") treat ~2000 cc's of normal pelvis in rectal cancer for a ~15cc tumor. Actually being on the receiving end of PCI can make you feel strange in the short and long-term (thank goodness we went to ~25 Gy for PCI instead of 35+ Gy to mitigate that strangeness). But probably what can make one feel strangest of all is growing SCLC tumors in one's brain.
 
  • Like
Reactions: 1 users
How do you guys treat someone who's had complete resolution of 3 small brain mets after completion of chemo and xrt, NED everywhere else, essentially had limited stage disease minus the brain mets which the radiologist hedged on until after the repeat scan was done.

Prophylactic or therapeutic brain dose?
 
How do you guys treat someone who's had complete resolution of 3 small brain mets after completion of chemo and xrt, NED everywhere else, essentially had limited stage disease minus the brain mets which the radiologist hedged on until after the repeat scan was done.

Prophylactic or therapeutic brain dose?
Prophylactic is thera... nevermind :) 25/10
 
  • Like
Reactions: 1 user
How do you guys treat someone who's had complete resolution of 3 small brain mets after completion of chemo and xrt, NED everywhere else, essentially had limited stage disease minus the brain mets which the radiologist hedged on until after the repeat scan was done.

Prophylactic or therapeutic brain dose?
Excellent question.

I would probably give him a PCI dose (25/2.5) as WBRT and boost the former sites with something like 2-3 x 5 Gy (depending on how big they were). Totally NOT evidence-based...
 
  • Like
Reactions: 1 user
PCI is for microscopic disease, right? If they're not visible on MRI then they are microscopic disease. 25/10.

That being said, I am routinely offering MRI surveillance for these patients, and WBRT at progression (not evidence based either).
 
I think I would be okay with surveillance (q 3 month MRI) if patient was going to actively follow-up. But treating with PCI is fine too
 
Top