QUARTZ Phase III Trial: WBRT ineffective, ?worse than best supportive care w/NSCLC brain mets

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

scarbrtj

I Don't Like To Bragg
7+ Year Member
Joined
Dec 18, 2015
Messages
3,216
Reaction score
4,930
http://press.thelancet.com/lungcancer.pdf
http://www.thelancet.com/journals/lancet/article/PIIS0140-6736(16)30825-X/abstract

Personally I think the study is well done and the results are difficult with which to argue. And probably if we are honest with our American RTOG selves, the dearth of randomized radiation-or-no-radiation trials here has been a travesty. One less (routine) patient and fractionated RT indication for the rad onc department. I will stop routinely offering WBRT for NSCLC brain mets. (And yet, we need more rad oncs?)

Members don't see this ad.
 
Looking at the OS from both arms, one has to question whether this applies to good PS patients. Many of these pts prob should have gone straight to hospice
 
Looking at the OS from both arms, one has to question whether this applies to good PS patients. Many of these pts prob should have gone straight to hospice


Yep.

I don't know what scarbrtj is talking about. If the patient has a good KPS, this study doesn't really apply to them

Everyone knows RPA class III patients on the other hand have an OS of about 2.3 months. I think for this group it does give us as rad oncs reasons NOT to treat. We may not be doing them any benefit, though pretty much impossible to show a survival difference when the OS is this bad.

would have been interested to see an endpoint as well of decline in neurologic function.
 
Members don't see this ad :)
Are you guys being intentionally obtuse or did you not read the study...
If the patient has a good KPS, this study doesn't really apply to them
What study really applies to these patients? There were good KPS patients in this study.
We may not be doing them any benefit, though pretty much impossible to show a survival difference when the OS is this bad.
"May not" be doing them any benefit? When may we be doing them benefit? What would be a supportable rationale for offering WBRT treatment based on best evidence?

This was a multi-institutional, multi-national study. The patients the rad oncs encountered and the manner in which they initially encountered them match Joe Q. Radonc's average clinic; they certainly match mine based on my read of the materials & methods.

One is slightly depressed but not surprised to see a cavalcade of caveats to the study's findings. But think of how we respond when the findings--i.e. positive instead of negative--are reversed. I remember when Intergroup 0116 for gastric cancer appeared in abstract around 2000, published in NEJM around 2001. Practice patterns in favor of postop chemoRT seemed to change overnight.
 
Obtuse would be extrapolating a British study with historically poor OS figures using a questionable wbrt dose to good PS patients from the US.

Currently have a pt I gave 37.5 Gy wbrt followed by gk srs boost to 5 brain mets, isolated relapse after surgery for stage I lung adenoCA a year prior.

She's reaching her 2 year anniversary this October.

Very obtuse
 
But I would like to know more about these good performance status patients with brain mets from lung cancer deemed not to be candidates for surgery or SRS. They happen, sure, but I just don't think they're that common (as this study showed).

I also have to say I like you're an XRThopeful and that RPA class III patients have a survival of "about 2.3 months." About indeed. That is more hopeful than "about 2 months" ;)
 
Obtuse would be extrapolating a British study with historically poor OS figures using a questionable wbrt dose to good PS patients from the US.

Currently have a pt I gave 37.5 Gy wbrt followed by gk srs boost to 5 brain mets, isolated relapse after surgery for stage I lung adenoCA a year prior.

She's reaching her 2 year anniversary this October.

Very obtuse
But that would not be a patient for this study... which you still haven't read :(
 
We're very down on British data here in the US. I think their food offerings are poor, but the quality of their XRT studies and data are much better than we have here in the U.S., for a START (see what I did there).
 
Taken directly from the abstract "Although the primary outcome measure result includes the prespecified non-inferiority margin", ....

A priori develop an hypothesis and do the study-come up with an answer that you don't like so ignore the prespecified hypothesis...SCIENCE!
 
For those who prefer pictures look at the supplementary material-page 13.




Essentially the results of this study are inconclusive and cannot be used to argue non-inferiority as the authors originally hypothesized as the CI includes the a prior NI margin. It is simply underpowered and the authors should have kept to their original sample size (which was reduced because of poor accrual).
 
Lots of issues with this study, but I think it actually confirms what I've been seeing as a developing standard of care for brain mets. First, I have no idea how they found 81ish patients in each arm who had a single brain met that was NOT amenable to SRS. I can maybe think of a few situations where it couldn't be done, but that raised a red flag for me.

Second, 20 Gy in 5 fractions? Could explain why they couldn't get patients off steroids after WBRT, which has never been an issue in my practice. At all. We're seeing better neurotoxicity data with 37.5 Gy in 15 fx with Namenda, hippocampal-sparing IMRT, etc, so they were way, way off of what I would consider contemporary whole brain RT. It does not matter to me in the least that a trial from 7 years ago in a different disease, asking a different question, used the same dosing schedule in a portion of their patients.

Also, patients with 5+ brain mets DID have a survival advantage in the study. So, what this tells me is that what we've been doing- moving more towards SRS for a few lesions, then transitioning to WBRT for 5+ lesions (who knows what the right number here is)- likely is the right thing to do, especially in light of all the WBRT toxicity data we've been seeing over the last year.

Finally, patients younger than 60 years old also had a survival advantage to the addition of whole brain RT, and there weren't nearly enough patients with RPA class I status to make any meaningful inference whatsoever. It appears, based on all this, that the data suggests that patients with a better prognosis- < 60 years old, maybe RPA class I as well- or patients with a significant intracranial disease burden benefit from whole brain RT.

Therefore, saying "I'm not going to offer WBRT to anyone with NSCLC" really doesn't align with the data this paper is presenting.
 
For those who prefer pictures look at the supplementary material-page 13.




Essentially the results of this study are inconclusive and cannot be used to argue non-inferiority as the authors originally hypothesized as the CI includes the a prior NI margin. It is simply underpowered and the authors should have kept to their original sample size (which was reduced because of poor accrual).

Strictly statistically speaking, I think you're wrong. The power of the study which was pre-spec'd was 80%, and they needed ~534 patients for that power. They accrued 538. The pace of the accrual was somewhat poor, but they met their accrual target. Whether 80% is an adequate power to sate our desires for "conclusivity" (what XRT study is conclusive???) is another debatable matter. As they say, and this has to get through peer review obviously, "The QUARTZ trial is the sole adequately powered RCT specifically addressing the efficacy of supportive care plus WBRT compared with supportive care alone in patients with NSCLC." I agree.
 
Members don't see this ad :)
And this is how the field will be taken down. Not that WBRT is a high value treatment but
1. Design trial to omit RT in resource based system
2. Accrue correct patients by description but make the unsuitable for SRS/surgery designation ambiguous
3. Ignore the pre-specified statistical endpoint of your study
4. Publish for personal gain
5. Trumpet results so other specialties who are less savy about data (which is hilariously sad) don't utilize a tool that helps patients

If this study leads to decreasing WBRT use in approriate patients in the UK, that is a tragedy. If you design a trial with a specified endpoint and accrual, either you accrue the correct population, with enough enrollment to assess that endpoint, or you don't. In this case, they didn't, and are being intellectually dishonest about presenting their conclusions.

It's like a more sinister version of the prostate hyprofractionated data we have so far. But the people who run the NHS in UK are loving it, and the authors just got themselves a lot more recognition. SCIENCE!
 
  • Like
Reactions: 2 users
Therefore, saying "I'm not going to offer WBRT to anyone with NSCLC" really doesn't align with the data this paper is presenting.
Yes, I unartfully stated "I will stop routinely offering WBRT for NSCLC brain mets."
I should have said: "I will stop routinely offering WBRT for NSCLC brain mets patients not candidates for surgery or SRS."
 
Yes, I unartfully stated "I will stop routinely offering WBRT for NSCLC brain mets."
I should have said: "I will stop routinely offering WBRT for NSCLC brain mets patients not candidates for surgery or SRS."

Fair enough, that makes sense to me. I missed the "routinely" to be honest. What I did when this came out is talked to my referring medoncs about the limitations of the data, how we're moving towards SRS anyway, etc. I found it went a long way to getting them to understand our point of view on things.
 
Lots of issues with this study, but I think it actually confirms what I've been seeing as a developing standard of care for brain mets. First, I have no idea how they found 81ish patients in each arm who had a single brain met that was NOT amenable to SRS. I can maybe think of a few situations where it couldn't be done, but that raised a red flag for me.
Would've been nice to know. Sometimes life is about disappointment, especially when trying to piece these things together from afar. But, I assume honesty first, lying last. So maybe big single mets? Close to brainstem mets? Or a pretty low PS patient with a single met?
Second, 20 Gy in 5 fractions? Could explain why they couldn't get patients off steroids after WBRT, which has never been an issue in my practice. At all. We're seeing better neurotoxicity data with 37.5 Gy in 15 fx with Namenda, hippocampal-sparing IMRT, etc, so they were way, way off of what I would consider contemporary whole brain RT. It does not matter to me in the least that a trial from 7 years ago in a different disease, asking a different question, used the same dosing schedule in a portion of their patients.
There is absolutely no data of which I know that suggests 20Gy/5fx 1) puts you at more risk for being "decadron dependent" or 2) is more neurocognitively deleterious than other fractionation regimens. In fact, I'll say there's more data (if you believe alpha/beta data, which is all we got kinda in this arena) that 37.5/15 (BED = 69Gy3) is more neurocognitively deleterious than 20/5 (BED = 53Gy3).
 
Would've been nice to know. Sometimes life is about disappointment, especially when trying to piece these things together from afar. But, I assume honesty first, lying last. So maybe big single mets? Close to brainstem mets? Or a pretty low PS patient with a single met?

There is absolutely no data of which I know that suggests 20Gy/5fx 1) puts you at more risk for being "decadron dependent" or 2) is more neurocognitively deleterious than other fractionation regimens. In fact, I'll say there's more data (if you believe alpha/beta data, which is all we got kinda in this arena) that 37.5/15 (BED = 69Gy3) is more neurocognitively deleterious than 20/5 (BED = 53Gy3).

I hear you about the lack of data with 20 Gy in 5 fx, but in all honesty have you ever had a problem getting patients off steroids after WBRT? I usually start the taper during/right after the course and complete it without any issue.
 
Strictly statistically speaking, I think you're wrong. The power of the study which was pre-spec'd was 80%, and they needed ~534 patients for that power. They accrued 538. The pace of the accrual was somewhat poor, but they met their accrual target. Whether 80% is an adequate power to sate our desires for "conclusivity" (what XRT study is conclusive???) is another debatable matter. As they say, and this has to get through peer review obviously, "The QUARTZ trial is the sole adequately powered RCT specifically addressing the efficacy of supportive care plus WBRT compared with supportive care alone in patients with NSCLC." I agree.
Yes they did meet the accrual goal of 534 but if you read the stats section the original a priori hypothesis that motivated the sample size was >1000 patients. The fact that they could not demonstrate non-inferiority with curves that are superimposable makes my point.

What study is conclusive? In the non-inferiority world a study is conclusive if the non-inferiority margin is completely outside of the observed CI. A helpful resource in this regard is available here. I think you are conflating conclusive with perfect. No study is perfect. My point is that these authors are dressing this up as science. In an experiment you begin with an hypothesis. In this case they could not rule out non-inferiority so scientifically this study cannot be used to support their hypothesis. You can say that in your clinical judgement WBRT offers little but you shouldn't use this inconclusive study to support your argument.

I use the same argument when I criticize the PIVOT study of radical prostatectomy versus nothing. The study rewrote the primary endpoint halfway through and enrolled 700 men instead of 2000 as originally designed. PIVOT is underpowered to answer the question posed, as was made clear by Ian Thompson's accompanying editorial. It was published in NEJM because if fits the narrative that doctors are greedy.

It can be argued that the whole exercise was a bad idea. Non-inferiority trials in clinical scenarios with very poor outcomes (when everyone dies within a short period of time) are problematic.
 
Last edited:
And this is how the field will be taken down. Not that WBRT is a high value treatment but
1. Design trial to omit RT in resource based system
2. Accrue correct patients by description but make the unsuitable for SRS/surgery designation ambiguous
3. Ignore the pre-specified statistical endpoint of your study
4. Publish for personal gain
5. Trumpet results so other specialties who are less savy about data (which is hilariously sad) don't utilize a tool that helps patients

If this study leads to decreasing WBRT use in approriate patients in the UK, that is a tragedy. If you design a trial with a specified endpoint and accrual, either you accrue the correct population, with enough enrollment to assess that endpoint, or you don't. In this case, they didn't, and are being intellectually dishonest about presenting their conclusions.

It's like a more sinister version of the prostate hyprofractionated data we have so far. But the people who run the NHS in UK are loving it, and the authors just got themselves a lot more recognition. SCIENCE!
Please elaborate on how the hypofractionated prostate data is sinister. Seriously, I would like to know.
 
Please elaborate on how the hypofractionated prostate data is sinister. Seriously, I would like to know.

1. CHiP data claiming non-inferiority of treatment in terms of PSA control at 5-years for a disease where 5-year outcomes are not the correct endpoint... and especially with 30% of patients being low risk!!
2. Lack of honest assessment of differences in toxicity burden per study - higher all around in Dutch, but lower GI toxicity in CHiPP (higher GU). Further lack of honesty in downplaying late toxicity risk in a cancer that has 90% specific survival rate at 10 years (lumping low risk and intermediate risk... since these studies do the same:)
3. The duke trial completely accruing the wrong patients (all low risk) and reporting an inadequate time point (follow up was ~7 years?) for non-inferiority. And then further downplaying the long term toxicity differences, which were significant, in a population whose cancer survival rate at 10 years is ~98%?

Overall there is just a lack of honest disclosure that the data are premature, that one study is basically irrelevant today, and that for a cancer with long expected survival a non-inferiority end point should not be in PSA control alone, even with the difficulty of setting a hybrid endpoint (PSA control with < grade 2 long term bowel toxicity). Maybe a more nuanced analysis with PSA control per increased incidence of toxicity? Or, you know, they already have something called QALYs that would have been a difficult, but more honest tool :)'

Breast hypofractionation studies did more than just survival, they looked at cosmesis too (most). Pushing aside the toxicity and taking the wrong group of patients, sometimes with ADT use which is now also a point of contention, and claiming non-inferiority at inadequate time points is sinister, in my book. But if I worked for an insurance company, and could afford anything for my prostate, then it would be great!
 
It's important to know the limitations of the trial, but for what it is worth, it is a very well designed trial and extremely pragmatic. You would be kidding yourself if you thought offering WBRT to a patient population whose survival is measured in weeks would improve their survival. Based on the quality of life data collected, we don't seem to see any improvement in symptoms either. Dose is not an issue - 20Gy/5fx is well established, giving 15fx to this patient population has zero chance of improving OS above and beyond 5fx.

First off, if SRS/Sx is possible, this trial does not apply. There absolutely is a subset of patients who do well with isolated brain mets and brain RT.

Second, determining that SRS/Sx is not possible takes time in many settings, i.e. outpatient - hence the large delay between diagnosis of brain mets and study enrollment, and study enrollment and initiation of treatment. When the delay between diagnosis of brain mets and first radiation treatment is as long as it is on this study, I have a hard time convincing myself that the NSCLC patient who presents to the ED, gets a CT, RPA 1, not a surgical/SRS candidate, and I get called... I would have a hard time saying no WBRT because of this study. Presumably this is partially because of the chemotherapy washout requirements.

Third, I would also love analysis on the RPA 1 patients on this trial - it makes no sense to me that they did so poorly. I would be careful applying the data on primarily bad apples to this entire patient population.

Finally, yes it is a small sample size, and didn't meet their original goals... but lets be honest... how many of you would have enrolled patients on this trial? I think it is fascinating that it was even completed. There are very few long term survivors on the trial, and while I do agree the decision to adjust methodology mid-trial is dicey, it doesn't change my view of the trial. These are important questions to ask, and they bring out the defensive side in people (much like the aforementioned hypofractionation studies for PCa).
 
I noticed the "questionable wbrt dose" comment too. But, recall: "The majority of patients in our study (88 of 143) received only 20 Gy in five fractions," (http://www.nejm.org/doi/full/10.1056/NEJMoa071780#t=article). I heard no one pooh-poohing questionable doses then.

Plenty of other stuff in that study to pooh pooh before you even get to the dose...

Brain imaging was not part of standard staging and follow-up procedures, unless symptoms suggestive of brain metastases were present.

Like.... not even require negative head imaging prior to enrollment of in a "PCI" trial.

Yes, I unartfully stated "I will stop routinely offering WBRT for NSCLC brain mets."
I should have said: "I will stop routinely offering WBRT for NSCLC brain mets patients not candidates for surgery or SRS."
What about the RPA I patient with controlled extracranial disease who come to see you with a 4cm brain met in a bad spot for surgery?
 
Last edited:
1. CHiP data claiming non-inferiority of treatment in terms of PSA control at 5-years for a disease where 5-year outcomes are not the correct endpoint... and especially with 30% of patients being low risk!!
2. Lack of honest assessment of differences in toxicity burden per study - higher all around in Dutch, but lower GI toxicity in CHiPP (higher GU). Further lack of honesty in downplaying late toxicity risk in a cancer that has 90% specific survival rate at 10 years (lumping low risk and intermediate risk... since these studies do the same:)
3. The duke trial completely accruing the wrong patients (all low risk) and reporting an inadequate time point (follow up was ~7 years?) for non-inferiority. And then further downplaying the long term toxicity differences, which were significant, in a population whose cancer survival rate at 10 years is ~98%?

Overall there is just a lack of honest disclosure that the data are premature, that one study is basically irrelevant today, and that for a cancer with long expected survival a non-inferiority end point should not be in PSA control alone, even with the difficulty of setting a hybrid endpoint (PSA control with < grade 2 long term bowel toxicity). Maybe a more nuanced analysis with PSA control per increased incidence of toxicity? Or, you know, they already have something called QALYs that would have been a difficult, but more honest tool :)'

Breast hypofractionation studies did more than just survival, they looked at cosmesis too (most). Pushing aside the toxicity and taking the wrong group of patients, sometimes with ADT use which is now also a point of contention, and claiming non-inferiority at inadequate time points is sinister, in my book. But if I worked for an insurance company, and could afford anything for my prostate, then it would be great!
Thanks for the response. Below is a rebuttal
1) CHHiP and the other non-inferiority trials all compare one form of radiation to another form of radiation. All of the dose escalation trials show an effect on PSA-recurrence with higher doses in the first five years (some with hormones and some not). All three non-inferiority studies CHHiP, NRG Oncology 0415 and PROFIT (presented at ASCO 2016 but not published) were reported because the DMC recommended it because of FUTILITY. This means that it is very, very unlikely (read less than 1:1000) that the hypo arm would do worse. There is no reason to believe that more important endpoints (DM, PCSM, etc) would differ if the biochemical endpoint shows no signal at five years (in fact, the two published trials show that hypo is doing better HR 0.84-0.85 favoring hypo).
2) I assume that you mean HYPRO when you say the Dutch trial. No one (including the authors and the editorialist) has recommended that this regimen be adopted. I don't know how you can say that the reported toxicity results are dishonest. CHHiP found no differences and 0415 observed more Grade 2 toxicity. The published reports both state this unequivocally. Do you think the authors are lying about the data? Furthermore where is the evidence that toxicity will worsen if there is little difference at 5 years. Yes GU toxicity can increase beyond five years but where is the evidence that the late increase will be worse with hypo? All of the disasters with hypo in the 1960's-1980's were observed within the first two years. Furthermore the QOL for CHHiP has been reported and there are no differences. It is expected that 0415 will report QOL at ASTRO 2016
3) NRG Oncology 0415 was not a Duke trial. The PI was at Duke when the paper was published but he was at Wake Forest when the study was designed. You are correct that it included only low risk patients but when the study was written in the early 2000's active surveillance was rarely used as the evidence base was weak. The rationale was to test the "prostate has a low alpha-beta" hypothesis in a group of patients without hormonal therapy. A pure radiation question without confounding. I disagree that long-term toxicity was downplayed. The results are in the paper and acknowledged. Most of the dose-escalation trials reported excess toxicity with higher doses but this didn't seem to temper enthusiasm for the higher doses (which parenthetically have not been demonstrated to effect survival).

If the data are premature (which I disagree with) then how many men followed for how long will address your concern? Ten years? Then what about 15 years? This is a straw man argument in view. Between CHHiP, 0415 and PROFIT more than 5000 men have been followed for a median of 5-6 years-the majority of the evidence shows no increase in sever toxicity and no differences in patient reported QOL.
 
What about the RPA I patient with controlled extracranial disease who come to see you with a 4cm brain met in a bad spot for surgery?
Not a WBRT patient, and not a best supportive care patient. Rx: Hypofractionated high-dose RT. I avoid WBRT (except for PCI!), especially to 30+ Gy doses, like the plague nowadays. Given the QUARTZ, it's essentially antediluvian at this point for NSCLC-associated brain mets. The vast preponderance of the data IMHO indicate limited tumoral irradiation, versus elective normal tissue plus tumoral irradiation, is the preferred treatment strategy for brain mets.

http://www.thelancet.com/journals/lanonc/article/PIIS1470-2045(14)70061-0/abstract
http://www.redjournal.org/article/S0360-3016(14)04294-1/abstract
http://jama.jamanetwork.com/article.aspx?articleid=2536637
 
Not a WBRT patient, and not a best supportive care patient. Rx: Hypofractionated high-dose RT. I avoid WBRT (except for PCI!), especially to 30+ Gy doses, like the plague nowadays. Given the QUARTZ, it's essentially antediluvian at this point for NSCLC-associated brain mets. The vast preponderance of the data IMHO indicate limited tumoral irradiation, versus elective normal tissue plus tumoral irradiation, is the preferred treatment strategy for brain mets.

http://www.thelancet.com/journals/lanonc/article/PIIS1470-2045(14)70061-0/abstract
http://www.redjournal.org/article/S0360-3016(14)04294-1/abstract
http://jama.jamanetwork.com/article.aspx?articleid=2536637

Not sure that wbrt is down for the count, RTOG 0214 did show a decrease in brain met incidence and if I knew someone would be noncompliant with serial MRIs, I would not think twice about giving them wbrt.

It would be nice for nccn to address the role of partial brain xrt, but at this point, the cat 1 recommendation remains adjuvant wbrt after craniotomy for a single met
 
Yes they did meet the accrual goal of 534 but if you read the stats section the original a priori hypothesis that motivated the sample size was >1000 patients. The fact that they could not demonstrate non-inferiority with curves that are superimposable makes my point.

What study is conclusive? In the non-inferiority world a study is conclusive if the non-inferiority margin is completely outside of the observed CI. A helpful resource in this regard is available here. I think you are conflating conclusive with perfect. No study is perfect. My point is that these authors are dressing this up as science. In an experiment you begin with an hypothesis. In this case they could not rule out non-inferiority so scientifically this study cannot be used to support their hypothesis. You can say that in your clinical judgement WBRT offers little but you shouldn't use this inconclusive study to support your argument.
The null hypothesis is that there *is* a -7 QALY days (or more) with the omission of WBRT ("At the time of the design, following discussions within the Trial Management Group and with clinicians in the UK, we decided that a reduction in QALYs of no more than 1 week would convince the clinical and patient communities that WBRT could be omitted from standard practice"). They found a deficit of -4.7 QALY days, 90% CI: -12.7 to 3.3. They failed to reject the null at their pre-specified statistical metrics. But even in trials with non-inferiority margins outside the observed CI, there is still a chance for Type 1 error, which in the case of this non-inferiority trial would be choosing to jettison the standard (ie WBRT) treatment from routine practice. I am doing so, and I realize I might be making a Type 1 error, but the clinical significance and magnitude of this error (in perhaps decreasing QALY days by up to two weeks) I believe is offset by savings in patient side effects, cost to the system, and a lack of survival benefit. Support of the Hypothesis is a continuum, not an absolute. As such, conclusivity is on a continuum as well. You've got to look at the big picture, and in my opinion they did an honest job of doing so: "Although the primary outcome measure result includes the prespecified non-inferiority margin, the combination of the small difference in QALYs and the absence of a difference in survival and quality of life between the two groups suggests that WBRT provides little additional clinically significant benefit for this patient group."

That's science and honesty. This will be the only randomized trial you'll ever get looking at this question. A trial like this sure as hell will not be done in the US.
 
It would be nice for nccn to address the role of partial brain xrt, but at this point, the cat 1 recommendation remains adjuvant wbrt after craniotomy for a single met
Not sure it's "Cat 1"... in the Patchell study there was no survival benefit to WBRT and usually it's those survival-increasing findings that get the Cat 1 recommendation. And the NCCN places partial brain RT (SRS explicitly) on an equal footing with WBRT in the post-craniotomy-for-met setting.

NCCN also says consider best supportive care or WBRT for patients <3 mos estimated survival. Will WBRT still be mentioned by the NCCN for these patients given QUARTZ? Stay tuned.
 
Not sure it's "Cat 1"... in the Patchell study there was no survival benefit to WBRT and usually it's those survival-increasing findings that get the Cat 1 recommendation. And the NCCN places partial brain RT (SRS explicitly) on an equal footing with WBRT in the post-craniotomy-for-met setting.

NCCN also says consider best supportive care or WBRT for patients <3 mos estimated survival. Will WBRT still be mentioned by the NCCN for these patients given QUARTZ? Stay tuned.
You're right, they just removed the cat 1 recommendation for wbrt after surgery with the last update of the guidelines. It was there previously though per their revisions in the beginning of the guidelines. Interestingly, they removed the Cat 2B designation from adjuvant SRS. Now they are all on equal standing.
 
Last edited:
I just skimmed the article and the curves, but at first glance was surprised to see some survivors at like a year after diagnosis that never got whole brain....essentailly an equal number of survivors at 52 weeks. I figured you'd have this tail of survivors in the whole brain arm that was slightly more than that of the steroid alone arm. Shocked to see 6 patients alive at a year with no whole brain ?maybe chemo penetrance or a misdiagnosis?

This is just so different than what anecdotally I see. I just this past week saw two patients that are now 1.2 years out from their whole brain that I started while they were hospitalized and in ROUGH shape, like KPS 50. They're not running marathons but they're doing OK. With steroids alone I have no doubt they would be dead.

I'm not sure this will change my practice a ton, but I think I am going to be more strongly discussing hospice options with some of these patients. Because while I have anecdotal longer term "survivors," I surely have my fair share of patients that pass away within 30-60 days of completing their whole brain and I often wonder if they would have been better served with dex and hospice.

Patients often want to try something though, and it can be very hard to deny them WBRT unless they are just absolute train wrecks. With this article you can maybe say "well, it's not likely to help, so we shouldn't do it," but it's hard to forget your long term survivors that wouldn't be here if you would have not treated them. Tough clinical conundrum, at least for me. I try to be very data driven, but this is tough.
 
  • Like
Reactions: 1 users
I just skimmed the article and the curves, but at first glance was surprised to see some survivors at like a year after diagnosis that never got whole brain....essentailly an equal number of survivors at 52 weeks. I figured you'd have this tail of survivors in the whole brain arm that was slightly more than that of the steroid alone arm. Shocked to see 6 patients alive at a year with no whole brain ?maybe chemo penetrance or a misdiagnosis?

This is just so different than what anecdotally I see. I just this past week saw two patients that are now 1.2 years out from their whole brain that I started while they were hospitalized and in ROUGH shape, like KPS 50. They're not running marathons but they're doing OK. With steroids alone I have no doubt they would be dead.

I'm not sure this will change my practice a ton, but I think I am going to be more strongly discussing hospice options with some of these patients. Because while I have anecdotal longer term "survivors," I surely have my fair share of patients that pass away within 30-60 days of completing their whole brain and I often wonder if they would have been better served with dex and hospice.

Patients often want to try something though, and it can be very hard to deny them WBRT unless they are just absolute train wrecks. With this article you can maybe say "well, it's not likely to help, so we shouldn't do it," but it's hard to forget your long term survivors that wouldn't be here if you would have not treated them. Tough clinical conundrum, at least for me. I try to be very data driven, but this is tough.

Brilliant post my friend, I couldn't have said it better and agree 100%
 
I just skimmed the article and the curves, but at first glance was surprised to see some survivors at like a year after diagnosis that never got whole brain....essentailly an equal number of survivors at 52 weeks. I figured you'd have this tail of survivors in the whole brain arm that was slightly more than that of the steroid alone arm. Shocked to see 6 patients alive at a year with no whole brain ?maybe chemo penetrance or a misdiagnosis?

This is just so different than what anecdotally I see. I just this past week saw two patients that are now 1.2 years out from their whole brain that I started while they were hospitalized and in ROUGH shape, like KPS 50. They're not running marathons but they're doing OK. With steroids alone I have no doubt they would be dead.

I'm not sure this will change my practice a ton, but I think I am going to be more strongly discussing hospice options with some of these patients. Because while I have anecdotal longer term "survivors," I surely have my fair share of patients that pass away within 30-60 days of completing their whole brain and I often wonder if they would have been better served with dex and hospice.

Patients often want to try something though, and it can be very hard to deny them WBRT unless they are just absolute train wrecks. With this article you can maybe say "well, it's not likely to help, so we shouldn't do it," but it's hard to forget your long term survivors that wouldn't be here if you would have not treated them. Tough clinical conundrum, at least for me. I try to be very data driven, but this is tough.

This is a post that either shows exactly why we do randomized trials or why we should *never* do randomized trials. The randomized trials are done in these data-free zones (does omitting WBRT matter? Does omitting the dog-leg matter in testicular ca; are chemo and RT equivalent for testicular ca? Does reducing whole breast RT from 5 weeks to 3 weeks matter? can you reduce total dose and treatment fractions when treating bony mets?) to support or cast doubt on our biases or traditions. But if the trials are never going to matter anyway despite casting doubt on our biases or traditions, why should we even do trials? "Shocked" to see 6 out of ~250 patients alive @ 1 yr with best supportive care? That seems totally possible/reasonable to me. 1 out of 538 patients were known to be alive "at time of analysis"... that's reasonable, too. Even if the number at risk were 6 in the WBRT arm @52 weeks and survival was 0%/number at risk zero (instead of 6 as well) @ 52 weeks in the best supportive care arm, the p-value for the difference in the curves would have still been very high, essentially unchanged.

When you say "With steroids alone I have no doubt they would be dead," I think just the opposite. There's no way you could know this. That statement is as rational as saying "I know if I had a different dealer at the blackjack table, I would've come out on top tonight." In fact, there's an orgy of data that says just the opposite: it's more likely than not that applying WBRT did not change the survival course for your (super rare) long term survivors. It's more likely than not that changing the dealer does not alter your odds at the blackjack table, too (I've tested this extensively). "It's hard to forget your long term survivors that wouldn't be here if you would have not treated them." If I show you a 52-week survivor of brain mets (whom you knew had not had surgery or SRS), and you didn't know whether they had had WBRT... how would you know? Would you say, "Well, they're alive... so they must've had WBRT"? This study is encouraging you not to think that way.
 
This is a post that either shows exactly why we do randomized trials or why we should *never* do randomized trials. The randomized trials are done in these data-free zones (does omitting WBRT matter? Does omitting the dog-leg matter in testicular ca; are chemo and RT equivalent for testicular ca? Does reducing whole breast RT from 5 weeks to 3 weeks matter? can you reduce total dose and treatment fractions when treating bony mets?) to support or cast doubt on our biases or traditions. But if the trials are never going to matter anyway despite casting doubt on our biases or traditions, why should we even do trials? "Shocked" to see 6 out of ~250 patients alive @ 1 yr with best supportive care? That seems totally possible/reasonable to me. 1 out of 538 patients were known to be alive "at time of analysis"... that's reasonable, too. Even if the number at risk were 6 in the WBRT arm @52 weeks and survival was 0%/number at risk zero (instead of 6 as well) @ 52 weeks in the best supportive care arm, the p-value for the difference in the curves would have still been very high, essentially unchanged.

When you say "With steroids alone I have no doubt they would be dead," I think just the opposite. There's no way you could know this. That statement is as rational as saying "I know if I had a different dealer at the blackjack table, I would've come out on top tonight." In fact, there's an orgy of data that says just the opposite: it's more likely than not that applying WBRT did not change the survival course for your (super rare) long term survivors. It's more likely than not that changing the dealer does not alter your odds at the blackjack table, too (I've tested this extensively). "It's hard to forget your long term survivors that wouldn't be here if you would have not treated them." If I show you a 52-week survivor of brain mets (whom you knew had not had surgery or SRS), and you didn't know whether they had had WBRT... how would you know? Would you say, "Well, they're alive... so they must've had WBRT"? This study is encouraging you not to think that way.

I think everyone needs to take a step back and see how this might affect their practice. It might not affect it much at all. Our job as doctors is to take what is published and make our own recommendations for own patients.

It's important to ask yourself if your patient would have eligible for a 'practice changing' study. If not can you extrapolate the results to him/her. If the patient would have been eligible was that patient's clinical situation well represented (i.e. RTOG 9410 included stage ii patients who represented a small minority).

For this study .... Radiation for stage IV nsclc has always been palliative. This study doesn't change that. If a patient has 20 punctate brain metastases and a chest full of disease I'm not doing WBRT. QUARTZ supports this. If a patient has bulky and/or symptomatic brain metastases and/or limited extracranial disease than I'm treating it. QUARTZ didn't change my mind on that. This trial wasn't designed to help me determine who is best suited for WBRT so I need to use clinical judgment. I'll concede that these decisions aren't always straight forward and QUARTZ will certainly impact decision making but you evaluation of the data, evaluation of the patient and discussion with the patient is what will drive your decision.




Sent from my iPhone using SDN mobile
 
Last edited:
  • Like
Reactions: 1 users
I think everyone needs to take a step back and see how this might affect their practice. It might not affect it much at all. Our job as doctors is to take what is published and make our own recommendations for own patients.

It's important to ask yourself if your patient would have eligible for a 'practice changing' study. If not can you extrapolate the results to him/her. If the patient would have been eligible was that patient's clinical situation well represented (i.e. RTOG 9410 included stage ii patients who represented a small minority).

For this study .... Radiation for stage IV nsclc has always been palliative. This study doesn't change that. If a patient has 20 punctate brain metastases and a chest full of disease I'm not doing WBRT. QUARTZ supports this. If a patient has bulky and/or symptomatic brain metastases and/or limited extracranial disease than I'm treating it. QUARTZ didn't change my mind on that. This trial wasn't designed to help me determine who is best suited for WBRT so I need to use clinical judgment. I'll concede that these decisions aren't always straight forward and QUARTZ will certainly impact decision making but you evaluation of the data, evaluation of the patient and discussion with the patient is what will drive your decision.




Sent from my iPhone using SDN mobile

Very well stated and you said what I wanted to much clearer and succinctly.

All NSCLC brain mets patients are not the same, and I find it tough to be dogmatic about no WBRT for patients that fit this trial enrollment criteria. Like you said, limited extracranial disease with symptomatic brain mets are the exact patients I've seen helped by WBRT. This study will make me be more mindful of WBRT for those asymptomatic patients with extensive extracranial disease though.
 
  • Like
Reactions: 1 user
This is a post that either shows exactly why we do randomized trials or why we should *never* do randomized trials. The randomized trials are done in these data-free zones (does omitting WBRT matter? Does omitting the dog-leg matter in testicular ca; are chemo and RT equivalent for testicular ca? Does reducing whole breast RT from 5 weeks to 3 weeks matter? can you reduce total dose and treatment fractions when treating bony mets?) to support or cast doubt on our biases or traditions. But if the trials are never going to matter anyway despite casting doubt on our biases or traditions, why should we even do trials? "Shocked" to see 6 out of ~250 patients alive @ 1 yr with best supportive care? That seems totally possible/reasonable to me. 1 out of 538 patients were known to be alive "at time of analysis"... that's reasonable, too. Even if the number at risk were 6 in the WBRT arm @52 weeks and survival was 0%/number at risk zero (instead of 6 as well) @ 52 weeks in the best supportive care arm, the p-value for the difference in the curves would have still been very high, essentially unchanged.

When you say "With steroids alone I have no doubt they would be dead," I think just the opposite. There's no way you could know this. That statement is as rational as saying "I know if I had a different dealer at the blackjack table, I would've come out on top tonight." In fact, there's an orgy of data that says just the opposite: it's more likely than not that applying WBRT did not change the survival course for your (super rare) long term survivors. It's more likely than not that changing the dealer does not alter your odds at the blackjack table, too (I've tested this extensively). "It's hard to forget your long term survivors that wouldn't be here if you would have not treated them." If I show you a 52-week survivor of brain mets (whom you knew had not had surgery or SRS), and you didn't know whether they had had WBRT... how would you know? Would you say, "Well, they're alive... so they must've had WBRT"? This study is encouraging you not to think that way.

I see what you're saying and agree in part - it is very much speculation on my part. But I don't think the data from QUARTZ is so strong that I will be dogmatic about not offering WBRT. The juxtaposition about the blackjack dealer I don't think it is completely accurate, because that is completely random while some underlying measurable variables that may predict "success" from WBRT may be present while they are not in the blackjack scenario.

I don't think it's unreasonable to take the big cohort data from the trial and look at it and consider that even though a patient may have fit enrollment from the trial, they would be extremely unlikely to have lived more than a few weeks when they're symptomatic from large volume brain mets with minimal extracranial disease.
 
  • Like
Reactions: 1 user
The juxtaposition about the blackjack dealer I don't think it is completely accurate, because that is completely random while some underlying measurable variables that may predict "success" from WBRT may be present while they are not in the blackjack scenario.

We have no compelling data what predicts "success" for WBRT patients (as represented in this trial). Just as we have no data what predicts success in blackjack outside of when to hit, when to stand, when to double, when to split, and when to take insurance... all of which are robustly established by data (in contradistinction to WBRT experiences as outlined in QUARTZ and one previous small randomized trial).
 
Because of the competing causes of death and morbidity in trials like this, the primary outcome should almost always be neuro-related death. If you choose to enroll 2/3 elderly people with uncontrolled systemic disease and want to claim that WBRT didn't somehow salvage their overall situation, you'll get this answer every time. Particularly if you move the goal posts midway through.

It's like saying seat belt vs. no seat belt in a car crashes, when 2/3 of the crashes were over the Grand Canyon.
 
Because of the competing causes of death and morbidity in trials like this, the primary outcome should almost always be neuro-related death. If you choose to enroll 2/3 elderly people with uncontrolled systemic disease and want to claim that WBRT didn't somehow salvage their overall situation, you'll get this answer every time. Particularly if you move the goal posts midway through.

Are you saying we should or should not offer WBRT to "elderly people with uncontrolled systemic disease" with brain mets. Did you already know that we shouldn't? Why didn't the trialists and most everyone else know what you already knew. It seems to me that when you have lung CA-related brain mets and other disease and/or mets outside the brain, death modes are fungible. If I'm just swapping out the mode of death ("neuro-related" versus "liver met-related" for example) with WBRT, who cares? Perhaps you think saving someone from a neuro-related death is better than allowing them to die a neuro-related death--but why is that? What might be easier to measure and less prone to varying opinion is whether or not QALY days were decreased by omitting WBRT. You said by doing this trial "you'll get this answer every time," but I must be a simpleton because I really wouldn't know which way the QALY days would go by omitting WBRT before this trial. Measuring QALY days is a kind of surrogate for measuring the quality of cognition, you could argue, and thereby getting closer perhaps to your idea of measuring for neuro-related death. The QALY day "answer" here was lackluster, as was a secondary analysis re: survival. Even the survival answer I wouldn't have been so sure of before this trial--not that I'm sure now, but I'm far more skeptical re: WBRT in this patient population (given these inclusion/exclusion parameters).
 
Looking at the curve, 25% of the patients died within 4 weeks of randomization (not completion of XRT mind you), 50% died within 8 weeks. This is worse than the worst actors in the classic RPA index.

My human brain and deductive abilities can help me suss out who these poor people may be at time of consultation, when the decision to pursue WBRT vs hospice takes place (as opposed to those with a slightly rosier, though still imperfect prognosis). So no, I didn't need a trial to tell me that WBRT won't save or improve the QALY of people literally actively dying of systemic disease.

Their median patient did about 10% worse than the worst patient group on RTOG trials from 25-40 years ago. I'm sorry, but that's bull****. It's stacking the deck to get the answer they want. Or it's pure selection bias because anyone who clinicians thought may have benefited from WBRT received it off protocol, and thus the accrual issues.
 
Last edited:
The patient groups on previous trials would have potentially been SRS candidates now, and de facto would have better prognostic disease.

The fact this trial was done indicates that these patients were getting WBRT when SRS/NSx was not an option. WBRT does not seem to help the majority of patients. We all have received consults on RPA 3 patients, and this trial definitively states WBRT does not help. I would not offer WBRT for brain mets if SRS/NSx can't be performed and they are KPS <70.

RPA 2, a rather large cohort of patients, also likely does not benefit. I would imagine we would make case by case basis decisions.

I would imagine we will all still offer WBRT to RPA 1 patients.
 
Am I missing a discussion of patients who are EGFR+ who's TKI keeps their systemic disease in check? TKI's have about 1/40th the cns concentration as systemic - likely sub-therapeutic. With controlled primary, and innumerable brain mets these patients will surely die a neurologic death. As a field, it is important to get ahead of this study with the referring docs. We will again fall victim to the bad press that shows no difference in survival of RT or surgery over observation in prostate cancer. Even with clear differences in metastasis - which with longer follow up likely translate into survival. We know from SWOG post op the time frame for survival benefit is around 15 years.
 
Quartz reflects IMHO true practice in the perhaps not so perfect world of the British NHS.
Certainly alot of patients in Quartz got WBRT, when SRS or SFRT would have been viable options.

Have a look at NHS treatment policies for SRS:
https://www.england.nhs.uk/wp-content/uploads/2013/04/d05-p-d.pdf
Page 7-8

Certainly some of those criteria are not met in some of the patients we treat with SRS on daily basis. Is that bad? I do not know.

Among the patients I treated the last couple of weeks, two patients with brain mets have troubled me.

1. A lady, age 70ish with metastatic breast cancer, Her2 positive. She was under 2rd line treatment with chemo/Her2-targetted therapy and developed multiple brain mets (>8). KPS 70+. Extracranial disease limited to bones, but slightly progressive too. I chose to treat with hippocampal-sparing WBRT. Her KPS deteriorated dramatically at about 15 Gy, clearly not related to WBRT. I had to terminate treatment last week and she's in hospice care now. We don't really know what happened.

2. A lady, age 80+, KPS chronically poor (60) mainly because of COPD. She developed an NSCLC with a cT2cN0 primary with two brain mets, no additional extracranial disease. One of the brain mets was in the posterior fossa and was going to cause trouble soon probably. I went for SRS and considered hitting the primary in 2 months from now, if she didn't progress anywhere else. Therapy completed, but the patient passed away one week after SRS in the nursing home. No neurological signs, so it doesn't seem to have been edema due to the SRS in the posterior fossa.

I had rather high hopes for both of these patients, yet they both have benefited zero from the rather complex treatment I applied.
Should I have not treated them? :)
 
Quartz reflects IMHO true practice in the perhaps not so perfect world of the British NHS.
Certainly alot of patients in Quartz got WBRT, when SRS or SFRT would have been viable options.

Have a look at NHS treatment policies for SRS:
https://www.england.nhs.uk/wp-content/uploads/2013/04/d05-p-d.pdf
Page 7-8

Certainly some of those criteria are not met in some of the patients we treat with SRS on daily basis. Is that bad? I do not know.

Among the patients I treated the last couple of weeks, two patients with brain mets have troubled me.

1. A lady, age 70ish with metastatic breast cancer, Her2 positive. She was under 2rd line treatment with chemo/Her2-targetted therapy and developed multiple brain mets (>8). KPS 70+. Extracranial disease limited to bones, but slightly progressive too. I chose to treat with hippocampal-sparing WBRT. Her KPS deteriorated dramatically at about 15 Gy, clearly not related to WBRT. I had to terminate treatment last week and she's in hospice care now. We don't really know what happened.

2. A lady, age 80+, KPS chronically poor (60) mainly because of COPD. She developed an NSCLC with a cT2cN0 primary with two brain mets, no additional extracranial disease. One of the brain mets was in the posterior fossa and was going to cause trouble soon probably. I went for SRS and considered hitting the primary in 2 months from now, if she didn't progress anywhere else. Therapy completed, but the patient passed away one week after SRS in the nursing home. No neurological signs, so it doesn't seem to have been edema due to the SRS in the posterior fossa.

I had rather high hopes for both of these patients, yet they both have benefited zero from the rather complex treatment I applied.
Should I have not treated them? :)

I don't think they are excessively strict, IMO. I get a system having a "guideline" of KPS >= 70 for SRT. I'm not sure how tumor volume > 20cc would push you to WBRT instead of just fractionating.
1. I think if you had to estimate her survival from extracranial disease it'd be more than 6 months with just the bone mets. Yeah, she was on second line chemo but that doesn't mean patient is going to die immediately. You can't use a retrospectoscope to say that you shouldn't have done HS-WBRT because something (non-neurological) happened thus greatly affecting her life.

Her GPA would've been at least 1, which gives her median 6 months survival for metastatic breast cancer.
https://www.ncbi.nlm.nih.gov/pubmed/19942357?dopt=Abstract

2. I think this is where the discussion of poor KPS comes in from the guidelines. I think NHS is worried about scenarios such as this one where if the cranial disease isn't going to kill them, don't bother going guns blazing at it. I think there's a difference between an acute KPS of 60 (say post-op or post-hospitalization) vs a chronic KPS 60 due to chronic, uncontrolled COPD. I still don't see how you offer the patient anything but SRS in this scenario given NO extracranial disease
Her GPA would've been 1.5, giving her median survival approximately 6.5 months, and I think you're justified based off the evidence we have as well at this time.

The wikibook brain mets section has a very succinct guide that I reference frequently for all the new/recurrent brain mets patients we see. (https://en.wikibooks.org/wiki/Radia...erview#Graded_Prognostic_Assessment_.28GPA.29)
 
I don't think they are excessively strict, IMO. I get a system having a "guideline" of KPS >= 70 for SRT. I'm not sure how tumor volume > 20cc would push you to WBRT instead of just fractionating.
I don't agree. I find the quite strict.

Here are the 8 criteria, which all have to be fullfiled for a patient to get SRS in the UK:
1. All patients must have undergone prior assessment by the local multidisciplinary team (MDT). The selection of patients for SRS/SRT must be made by an MDT with an understanding of the systemic and neurological disease processes and must include consideration of surgical treatment if appropriate.
2. In centres where SRS/SRT is delivered, referral may be made directly to the SRS MDT. In centres where there is no local SRS service, referral should be initially to the local neuro-science MDT, who can decide on the appropriateness of onward referral to an agreed SRS centre.
3. All patients being considered for SRS /SRT must be discussed by the specialist MDT at the stereotactic treatment centre and must have both specialist neurosurgery and specialist oncology input. SRS/SRT must not be recommended without the collective agreement of the MDT to ensure that the criteria regarding systemic disease and prognosis are fulfilled and that there is clarity about the place of SRS/SRT in the patient’s overall management plan.
4. Patients must have a Karnofsky Performance Status (KPS) ≥ 70.
5. The diagnosis of cancer must be established and there must be absent or controllable primary disease.
6. Pressure symptoms which would be best relieved by surgery are excluded.
7. Pre-treatment scans must not show a tumour volume of more than 20cc. This will usually mean that no individual tumour has a diameter in excess of 3cm.
8. The MDT has confirmed that the patient’s life expectancy from extracranial disease is expected to be greater than 6 months


1 & 2 & 3 are clear and fully understandable
4 is ok, although there are cases, where I view SRS as more convenient than WBRT. If a patient has a KPS of 60 because of a solitary brain met, does that mean I have to perform WBRT on him and WBRT is not indicated? I can't agree to that. It's more toxic and less effective, why do it?
5 is a matter of debate:
a) I have treated "brain lesions" without knowing the histology. I knew that the patient had a melanoma 10 years ago and now came with a solitary brain lesion. Would I have to do a re-biopsy of it to reconfirm it. Or does my "cancer diagnosis" 10 years ago still count?
b) "controllable" disease is also a very vague term. I treat with SRS in many patients without controlled extracranial disease, disease that will kill them, sometimes within less than a year. Why? Because the brain mets would otherwise kill them before. I wouldn't be allowed to do so according to NHS.¨
6. is clear.
7 is questionable, because I am not familiar how the NHS feels about multi-fraction SRS or FSRT. You don't have to give 1 x 25 Gy. You can also give 3 x 9 or 6 x 6 or whatever else you want for larger tumors or tumors sitting in certain places you don't want to do SRS, if you dont have to (brainstem, chiasm, etc). Do these treatments also count as SRS? Or do the patients have to get WBRT?
8 is interesting, because it seems the MDT can look into the future. I am not sure how you are supposed to measure this? Median OS from the time of the decision for/against SRS? Plus it's becoming alot more complex nowadays with all the immunotherapies and targetted agents to clearly estimate this. Some patients are lucky, some are not.
I saw a patient last week, which was on best supportive care after 4 lines of NSCLC-chemo with brain and liver mets. That was 1.5 years ago. She then went into a compassionate use programme for a new fancy immunotherapy. Back then I would have estimated her survival at around 2 months (after 4 lines of NSCLC-chemo that's where you are more or less). She's alive and doing well now, all manifestations are at partial/complete response.
 
Hmm, good points, Palex.
4. I will concede that having a strict cut-off of KPS > 70 is probably excessive. I would be OK with a 'guideline' of KPS >70 with ability to argue for special cases where KPS may not be that high (again, acute vs chronic KPS <70 comes to mind, KPS<70 DUE to the brain met, or low KPS but good QoL otherwise with assistance). I would not be in favor of WBRT for solitary brain met due to KPS of 60.
5a. My reading was that if you previously had a diagnosis of cancer and had a new lesion pop-up that looked like cancer, it would be reasonable to treat based on that. However, I can see where bureaucrats would want an unnecessary re-biopsy of a lesion that everyone agrees is a brain met. Maybe if the MDT documents that even without repeat histology, this is consistent with cancer and is thus eligible for treatment.
5b. Your point of 'absent/controllable' primary disease is well taken. Again, it's a grey area which could potentially hurt physicians. However, I think the argument from NHS is that if a patient is going to die within a year of extracranial disease, it's a cost-saving measure to do WBRT instead of SRS. I think if WBRT doesn't control the disease and they need SRS/SRT afterwards I would hope NHS is OK with that.

6. I hope they're OK with SRS for patients who are medically inoperable even with pressure symptoms.
7. I think they're OK with FSRT as they mention 'SRS/SRT' in their guidelines. I share your concern that 20cc shouldn't be a hard cut-off to prevent somebody from getting stereotactic radiation, especially fractionated.
8. Yes, the ALK/EGFR mutated NSCLCs with new targeted therapies or other patients (sometimes) with immunotherapies do much better than the numbers would say. It's a hard ask for sure to give a prognosis for an individual patient as I'm sure we've all seen those who outlived their median survival by years and years. I don't have a good answer to your concern about this.

I suppose it is more strict than I thought on initial read. If these were 'guidelines' which would allow some wiggle room for clinical judgement, I'd probably be OK with them. However, I do see where having a hard cut-off for all these measures is likely detrimental. However, NHS has got to cut costs somewhere, and in their mind, making SRS/SRT difficult to give to patients compared to WBRT is likely a short-sighted "cost-effective" option.
 
Top