RO may have a role in this COVID crisis, but we need a clinical trial.....

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

If you observe zero events in N trials, you can be 95% sure that the true rate is less than 3/N. Five patients, zero "instant deaths" 95%CI 0-60%. DO i think LD-RT is safe? Yes. Beware the "Law of Small Numbers" Faulty generalization - Wikipedia

Nine patients evaluated. One died before (selection), one was not sick enough (selection). Seven enrolled but two were intubated and did not get treated (selecting out bad actors again).

It's like the surgeon that says all of his patients that had no complications did well.

Science involves presecified hypotheses; it helps keep humans from engaging in cognitive biases

Members don't see this ad.
 
  • Like
Reactions: 1 user
I am not opposed to investigating the effects of LD-RT on COVID19. My point was that the data provided in the manuscript was insufficient to draw any conclusions.

It was enough to say "go ahead and study it some more" which is what a lot of people (myself included) were ridiculing them for prior to this study.
 
  • Like
Reactions: 3 users
If this was a drug, thousands would be getting it on the basis of that report. I'm not sure what anyone was expecting for the initial report of what amounts to a small Phase I/II. Rock on Khan.
 
  • Like
  • Haha
Reactions: 2 users
Members don't see this ad :)

If you observe zero events in N trials, you can be 95% sure that the true rate is less than 3/N. Five patients, zero "instant deaths" 95%CI 0-60%. DO i think LD-RT is safe? Yes. Beware the "Law of Small Numbers" Faulty generalization - Wikipedia

Nine patients evaluated. One died before (selection), one was not sick enough (selection). Seven enrolled but two were intubated and did not get treated (selecting out bad actors again).

It's like the surgeon that says all of his patients that had no complications did well.

Science involves presecified hypotheses; it helps keep humans from engaging in cognitive biases
I mean, is it really. The bias was that in this "declining," as they put it, patient group, there would be further decline. But there was an intervention. And then clinical improvement. Which appears coincidental with the intervention. And that improvement should now be challenging the cognitive biases.

This is not a "no complications, they did well" thing. It's a "they were real sick; you thought they'd stay sick or die" thing. Help me see my bias.
 
  • Like
Reactions: 1 user
I mean, is it really. The bias was that in this "declining," as they put it, patient group, there would be further decline. But there was an intervention. And then clinical improvement. Which appears coincidental with the intervention. And that improvement should now be challenging the cognitive biases.

This is not a "no complications, they did well" thing. It's a "they were real sick; you thought they'd stay sick or die" thing. Help me see my bias.
The bias is right before your eyes. you thought they would stay sick. Some percentage (the majority?) of hospitalized COVID patients recover (without LD-RT) exactly what % depends on lots of things of course.

What I would like to see-

Disease (Hospitalized COVID19)
Exposure (LD-RT)
Outcome (death, ventilator, something objective)

Research Hypothesis: In this group of COVID + patients (based on modeling from previous patients) we estimate that xx% will die (or some objective outcome). LD-RT will reduce the risk of death (outcome) by xx%. In this phase II study we will treat XX patients which will allow to determine if LD-RT influences mortality outcome).

One could also come up with a Bayesian method.

The present manuscript provides no quantitative information beyond we treated five patients and four got better (after excluding three who either died or got too sick). Why did they decide to report now? It is analogous to leaving the roulette wheel when you are ahead recognizing that in the long run the house always wins.
 
Last edited:
  • Like
Reactions: 1 user
The bias is right before your eyes. you thought they would stay sick. Some percentage (the majority?) of hospitalized COVID patients recover (without LD-RT) exactly what % depends on lots of things of course.

What I would like to see-

Disease (Hospitalized COVID19)
Exposure (LD-RT)
Outcome (death, ventilator, something objective)

Research Hypothesis: In this group of COVID + patients (based on modeling from previous patients) we estimate that xx% will die (or some objective outcome). LD-RT will reduce the risk of death (outcome) by xx%. In this phase II study we will treat XX patients which will allow to determine if LD-RT influences mortality outcome).

One could also come up with a Bayesian method.

The present manuscript provides no quantitative information beyond we treated five patients and four got better (after excluding three who either died or got too sick). Why did they decide to report now? It is analogous to leaving the roulette wheel when you are ahead recognizing that in the long run the house always wins.

This seems relevant. Atlanta study.

Among the 305 hospitalized patients, the median duration of hospitalization was 8.5 days and duration increased with age (Table 2).

Case fatality among patients aged 18–49 years, 50–64 years, and ≥65 years was 3.4%, 9.8%, and 35.6%, respectively (p<0.001).

A day later, 60% of 90 year olds off of O2. Seems good to me.
 
What I would like to see-
Research Hypothesis: In this group of COVID + patients (based on modeling from previous patients) we estimate that xx% will die (or some objective outcome). LD-RT will reduce the risk of death (outcome) by xx%. In this phase II study we will treat XX patients which will allow to determine if LD-RT influences mortality.
Among the 305 hospitalized patients, the median duration of hospitalization was 8.5 days and duration increased with age (Table 2).
Case fatality among patients aged 18–49 years, 50–64 years, and ≥65 years was 3.4%, 9.8%, and 35.6%, respectively (p<0.001).
A day later, 60% of 90 year olds off of O2. Seems good to me.
It will be nice perhaps to get some actuarial, Kaplan-Meier analysis of low dose RT (with Cox regression on 1 Gy vs 1.5 Gy and so on and so forth) vs placebo in severe COVID-associated pneumonia. We would never likely get that sort of data e.g. for narcan vs placebo in case of over-narcotization. Why? 'Cause it's just so plainly evident what the narcan does. If more and more clinicians used to the natural unirradiated COVID clinical course feel some similar way about the XRT (as people do for narcan, ie "boom, you're better!"), we may never get that actuarial data though.

Imagine a rad "onc" journal article though where the X-axis on a Kaplan-Meier graph runs from 0-~14 days. Imagine XRT being associated with acute life or death outcomes. XRT+COVID causes thoughts to course across uncommonly used clinical circuits in the rad onc's brain.
 
Last edited:
  • Love
  • Like
Reactions: 1 users
I know a few proton centers that will treat anything with protons.

Patient has a few weeks to live and this is purely palliative? No problem.

Is there a benefit to protons? No way. But, the patient wants protons, protons are "not worse" than photons, and insurance will pay for it, so protons are given.

Anything is good enough justification as long as the center can get paid (20% medicare co-pay, cash pay, or attempt lengthy series of arguments with insurance).
"Treat anything with protons"
Like breast?
Which, when using protons, causes more rib fractures. A 7% incidence of rib fractures!
And when comparing my personal anecdotal incidence of rib fractures using XRT for br CA, rib fracture incidence from protons is at least [undefined]* higher.
(thanks for raising the proton antennae today)

* division by zero
 
  • Like
  • Haha
Reactions: 2 users
Those debt payments won't service themselves!
I know a few proton centers that will treat anything with protons.

Patient has a few weeks to live and this is purely palliative? No problem.

Is there a benefit to protons? No way. But, the patient wants protons, protons are "not worse" than photons, and insurance will pay for it, so protons are given.

Anything is good enough justification as long as the center can get paid (20% medicare co-pay, cash pay, or attempt lengthy series of arguments with insurance).
center in nj perhaps?
 
Just grandstanding. But really: try XRT on the lungs in medically deteriorating ~nonagenarians with lung problems and see many getting better rapidly... and rad oncs say "anything could have caused that improvement besides the XRT." Or "I need a comparator to know if XRT is working." Rad oncs would be for Phineas Gage keeping a railroad spike in his skull for fret of removing it wouldn't have any pre-clinical data to back up spike removal, or there are no trials for spike removal to support spike removal, etc. Only in rad onc, folks, as carbonionangle would say.

Ralph Weichselbaum probably still thinks this is "nutty" and "dangerous". I believe these were his words.
 
Ralph Weichselbaum probably still thinks this is "nutty" and "dangerous". I believe these were his words.
Yes they were. Now he’s trying to take up the 1.5 Gy single fraction to lungs is a scary, dark boogeyman approach. (As if radiation oncologists worldwide aren’t irradiating totally normal lungs all the time in non-OS improving, but curative-seeking, scenarios.)

 
  • Haha
  • Angry
Reactions: 3 users
Mo Khan is the man in the arena after all!!!!
I think a real uncovered story here is that a world “expert” in radiation therapy says it’s “hard for him to say” what a small shot of radiation to lungs will do. Ralph only did a three year rad onc residency though. Since I did four, I find it “easy” to say: everyone will be OK.
 
  • Haha
  • Like
Reactions: 4 users
Members don't see this ad :)
Yes they were. Now he’s trying to take up the 1.5 Gy single fraction to lungs is a scary, dark boogeyman approach. (As if radiation oncologists worldwide aren’t irradiating totally normal lungs all the time in non-OS improving, but curative-seeking, scenarios.)



There's so much to unpack in these Tweets! Khan saying his risked his life, Thomas swooping in with his standard pomposity, Ralph hitting back with a full-body Weichselbaum-suplex...I love it.
 
  • Haha
  • Like
Reactions: 5 users
There's so much to unpack in these Tweets! Khan saying his risked his life, Thomas swooping in with his standard pomposity, Ralph hitting back with a full-body Weichselbaum-suplex...I love it.
Well Khan’s vaingloriousness isn’t making any points that’s for sure
 
  • Like
Reactions: 4 users
Well Khan’s vaingloriousness isn’t making any points that’s for sure

I was hoping English was his second language - that would explain a lot of these over-the-top Tweets. Alas, I just watched a video of him on YouTube...American as apple pie.
 
  • Haha
Reactions: 1 user
There's so much to unpack in these Tweets! Khan saying his risked his life, Thomas swooping in with his standard pomposity, Ralph hitting back with a full-body Weichselbaum-suplex...I love it.
Such is the state of academic rad onc.

Everyone looks a bit silly in the end.
 
  • Haha
Reactions: 1 users
LD-LR + Decadron vs Decadron alone trial? or forget LD-LR; just take the Decadron alone benefit and be happy with it.......
 
  • Like
Reactions: 1 users
LD-LR + Decadron vs Decadron alone trial? or forget LD-LR; just take the Decadron alone benefit and be happy with it.......
Given your skin in the game (but I mean really how much skin is truly in the game for rad onc), would like to know if you're serious or facetious here. Me, I think XRT+Dex vs Dex would "promise to be" a great, question-answering trial.
 
  • Like
Reactions: 1 user
Given your skin in the game (but I mean really how much skin is truly in the game for rad onc), would like to know if you're serious or facetious here. Me, I think XRT+Dex vs Dex would "promise to be" a great, question-answering trial.
I would love to see LD-LR + Deca vs Deca trial, don't get me wrong.....
but the inner Ralph W. (and we all have a little Ralph in us) just awoke and I do wonder if the world is ready to push the envelope further given there is a positive PIII study showing 30% survival improvement in the most high-risk pt.....
 
I would love to see LD-LR + Deca vs Deca trial, don't get me wrong.....
but the inner Ralph W. (and we all have a little Ralph in us) just awoke and I do wonder if the world is ready to push the envelope further given there is a positive PIII study showing 30% survival improvement in the most high-risk pt.....
Gotcha. Very very true. What with 30% surv improvement and NNT=8 for dex vs std of care, we are talking like N=500 vs 500 (total N=1000, or more?) patient trials more than likely of RT+dex vs dex to tease out an RT benefit.
RAAAALLLLLPPPPPPPHHHHHHHHH
 
  • Haha
Reactions: 1 user
Full disclosure: I work at a center that has opened two *higher* powered studies to evaluate the safety and efficacy of LDRT for COVID patients. @Chartreuse Wombat, I would hope that most of us are well aware that the Emory study lacks the power to conclusively argue that LDRT is helpful... what it did do is allow us to make a more convincing case to floor/ICU teams to collaborate with us.

Perhaps we should next analyze the data and statistical analysis that inform Ralph W's assertions.
 
  • Like
  • Haha
Reactions: 4 users
Full disclosure: I work at a center that has opened two *higher* powered studies to evaluate the safety and efficacy of LDRT for COVID patients. @Chartreuse Wombat, I would hope that most of us are well aware that the Emory study lacks the power to conclusively argue that LDRT is helpful... what it did do is allow us to make a more convincing case to floor/ICU teams to collaborate with us.

Perhaps we should next analyze the data and statistical analysis that inform Ralph W's assertions.
I hope the investigators have the chutzpah (and ability to tune out RECOVERY "noise") to complete accrual and analyze the data. Newton discovered the calculus but Leibniz is still worthy of mention!
 
  • Like
Reactions: 1 user
I hope the investigators have the chutzpah (and ability to tune out RECOVERY "noise") to complete accrual and analyze the data. Newton discovered the calculus but Leibniz is still worthy of mention!
The man makes a mean cookie

1592326838788.png
 
  • Love
Reactions: 1 users
Emory just updated their study- 10 patients treated, in MedRxiv of course....

 
  • Like
Reactions: 1 users
Emory just updated their study- 10 patients treated, in MedRxiv of course....

For the sake of accuracy, medrxiv isn't a journal, just a place to quickly put out your data prior to peer review. Good and bad papers are on there, some of which may never get published. I don't think I understand why it exists.
 
Emory just updated their study- 10 patients treated, in MedRxiv of course....

"Liars." - Ralph W
 
  • Haha
Reactions: 1 user
Emory just updated their study- 10 patients treated, in MedRxiv of course....


Did the first paper they pre-printed on medrxiv get actually published somewhere?
 
I can only thank God that Ralph et al weren't once again saying that 1 Gy of radiation will produce an extra 4 lung cancers in 100 people and an extra 7 coronary disease episodes out of 100 people. That was, like, bad.
 
  • Like
Reactions: 2 users
Radiation possibly helpful... decadron helpful... and now according to BMJ add hydroxychloroquine to list of helpful medications? I am very confused. Weren't the doctors in DC nuts? Isn't the drug censored or banned or cancelled or something.

"Hydroxychloroquine might be the most effective in reducing the duration of symptoms. Remdesivir has intermediate effectiveness and lopinavir-ritonavir could have intermediate effectiveness. The main limitations of the evidence are risk of bias and imprecision."
ikKxplq.png
 
Going to be a LBA at ASTRO as well. The irony is: "CANCER." In between cardiac SABR for v-tach and LD-RT for pneumonia, we may have to change our journal focus.

bSmFr28.png
 
Going to be a LBA at ASTRO as well. The irony is: "CANCER." In between cardiac SABR for v-tach and LD-RT for pneumonia, we may have to change our journal focus.

bSmFr28.png
9 patients screened; 5 treated 12 co-authors-sounds about right for RadOnc in 2020 #radoncrocks
 
  • Haha
  • Like
Reactions: 4 users
There is an informal/implicit pipeline between the Emory group and that journal (a journal of the ACS, based guess where?) (and guess where that journal's EIC has an appointment at?). So not surprised it was that journal that accepted the paper as they have done with many of their junk NCDB papers from the same institution.
 
  • Haha
  • Angry
Reactions: 3 users
While a lot of rad oncs were on twitter clutching their pearls or writing editorials over an RT treatment that has a ~zero in one billion chance of treatment-associated one year mortality, the surgeons went ahead and removed COVID patients' lungs. Don't know all the details, but 1) I bet the surgeons won't be clamoring for a randomized trial, and 2) there won't be any editorials from the rad onc side about testing lung transplant in animal models, or 3) how lung transplant is riskier than ~1 Gy to the lungs so let's halt the "unacceptable" lung transplanting.

 
  • Like
  • Love
  • Haha
Reactions: 4 users
While a lot of rad oncs were on twitter clutching their pearls or writing editorials over an RT treatment that has a ~zero in one billion chance of treatment-associated one year mortality, the surgeons went ahead and removed COVID patients' lungs. Don't know all the details, but 1) I bet the surgeons won't be clamoring for a randomized trial, and 2) there won't be any editorials from the rad onc side about testing lung transplant in animal models, or 3) how lung transplant is riskier than ~1 Gy to the lungs so let's halt the "unacceptable" lung transplanting.


a surgeon and rad onc have their own room with a box of lightbulbs and there is a bulb that is out in each room. The surgeon changes the bulb and walks out. The rad onc thinks about for a while inspecting each lighbulb, thinking about energy efficiency. Then he puts together a committee of old white men to discuss the light bulb. The committee is deadlocked and cannot agree that the lightbulb must be changed. Tension is palpable in the air. This has to be studied folks!, says one of the learned men. We need more evidence, of course, says another one as be rubs his belly. Eventually, the room goes dark, all bulbs go out. One of the men lights a match and the room catches on fire. They walk out the backdoor with the money. Stay tuned!
 
  • Haha
  • Like
  • Love
Reactions: 8 users
a surgeon and rad onc have their own room with a box of lightbulbs and there is a bulb that is out in each room. The surgeon changes the bulb and walks out. The rad onc thinks about for a while inspecting each lighbulb, thinking about energy efficiency. Then he puts together a committee of old white men to discuss the light bulb. The committee is deadlocked and cannot agree that the lightbulb must be changed. Tension is palpable in the air. This has to be studied folks!, says one of the learned men. We need more evidence, of course, says another one as be rubs his belly. Eventually, the room goes dark, all bulbs go out. One of the men lights a match and the room catches on fire. They walk out the backdoor with the money. Stay tuned!

Some residents show up, being instructed to put out the fire. They deftly extinguish the flames, clean the debris, rescue the orphans, and start rebuilding the room.

The old white men return, to supervise.

"Carbon tetrachloride was formerly used in fire extinguishers for electrical fires. It is no longer used for this purpose because of the formation of the toxic gas phosgene, Cl2CO. Write the Lewis structures for carbon tetrachloride and phosgene!"

The residents struggle with this task, as it is totally unrelated to rebuilding the room.

"IDIOTS. God, the quality of residents sure has dropped since my day..."
 
  • Haha
  • Like
  • Love
Reactions: 5 users
Some residents show up, being instructed to put out the fire. They deftly extinguish the flames, clean the debris, rescue the orphans, and start rebuilding the room.

The old white men return, to supervise.

"Carbon tetrachloride was formerly used in fire extinguishers for electrical fires. It is no longer used for this purpose because of the formation of the toxic gas phosgene, Cl2CO. Write the Lewis structures for carbon tetrachloride and phosgene!"

The residents struggle with this task, as it is totally unrelated to rebuilding the room.

"IDIOTS. God, the quality of residents sure has dropped since my day..."
You probably know this but since we are all geeks, the inventor of phosgene was Fritz Haber. His is a story of ammonia synthesis and fertilizer (and thus a massive alleviation of human poverty), phosgene, nitrogen mustard (the progenitor of chemotherapies), a wife who committed suicide because of Haber's contributions to war, and a lifelong friendship (and some say intellectual input with him too) with Einstein... without whom we wouldn't have modern rad onc. I'd get distracted with the room rebuilding while talking about this; and in the meantime the surgeons have changed lightbulbs in many, many rooms.
 
  • Love
  • Like
Reactions: 3 users
a surgeon and rad onc have their own room with a box of lightbulbs and there is a bulb that is out in each room. The surgeon changes the bulb and walks out. The rad onc thinks about for a while inspecting each lighbulb, thinking about energy efficiency. Then he puts together a committee of old white men to discuss the light bulb. The committee is deadlocked and cannot agree that the lightbulb must be changed. Tension is palpable in the air. This has to be studied folks!, says one of the learned men. We need more evidence, of course, says another one as be rubs his belly. Eventually, the room goes dark, all bulbs go out. One of the men lights a match and the room catches on fire. They walk out the backdoor with the money. Stay tuned!
The committee also agrees that more young radiation oncologists are necessary since there is an endless demand for changing lightbulbs...
 
  • Like
  • Haha
Reactions: 2 users
New paper in the Red Journal


Now, I don't really comprehend this trial...

1. The primary endpoint of the trial was radiologic response. Ok, that's an endpoint you can measure, but do we know if it has any clinical impact for the patients? Why would they choose that?

2. I don't really understand WHAT KIND of patients they actually treated.
Inclusion criteria were either patients who required oxygen OR hyperinflammation, so far so good.
When you look at the results however it seems that these were patients that were in the hospital because of COVID for a long time before study inclusion,, so I have serious doubts how much RT is supposed to help, since the acute phase of the disease was probably over.
All patients were on tapering steroid dose (prednisone or methylprednisolone, 112 median dose 40mg/24h) while on RT. Median time to receive RT from the date of admission was 52 days (ranging 17-85) and from the last antiCovid treatment (Hydroxychloroquine, Lopinavir/Ritonavir, Tocilizumab, Remdesivir) was 25 days (10-75).
So these are patients that were 1.5 months in the hospital already, pretreated with various drugs, tapering steroids (rather contradictory to say you are including patients with hyperinflammation, yet all your patients are actually tapering steroids?!?).
We also don't really know if these patients actually had COVID at the time of treatment or simply radiologic findings because they had COVID in the past. Bear in mind that at least one of them was admitted almost 3 months prior to RT, one would think that COVID would be gone in the mean time, right?

3. Needless to say, it took them 3 months to recruit 9 patients in a single-arm trial in a big Madrid hospital in spring. Rather disappointing, considering what Spain went through earlier this year.


My take home messages:
1. Another trial showing RT is safe (but we knew that already)
2. There seems to have been some clinical improvement in those patients, yet I am not sure if this can be attributed to RT only, since there is no control group.
3. A lot of details are still missing and I have serious doubts if the authos chose the correct patient group (-->heavily pretreated patients who probably no longer have much of an active inflammation).
 
Last edited:
  • Like
Reactions: 1 users
New paper in the Red Journal


Now, I don't really comprehend this trial...

1. The primary endpoint of the trial was radiologic response. Ok, that's an endpoint you can measure, but do we know if it has any clinical impact for the patients? Why would they choose that?

2. I don't really understand WHAT KIND of patients they actually treated.
Inclusion criteria were either patients who required oxygen OR hyperinflammation, so far so good.
When you look at the results however it seems that these were patients that were in the hospital because of COVID for a long time before study inclusion,, so I have serious doubts how much RT is supposed to help, since the acute phase of the disease was probably over.
All patients were on tapering steroid dose (prednisone or methylprednisolone, 112 median dose 40mg/24h) while on RT. Median time to receive RT from the date of admission was 52 days (ranging 17-85) and from the last antiCovid treatment (Hydroxychloroquine, Lopinavir/Ritonavir, Tocilizumab, Remdesivir) was 25 days (10-75).
So these are patients that were 1.5 months in the hospital already, pretreated with various drugs, tapering steroids (rather contradictory to say you are including patients with hyperinflammation, yet all your patients are actually tapering steroids?!?).
We also don't really know if these patients actually had COVID at the time of treatment or simply radiologic findings because they had COVID in the past. Bear in mind that at least one of them was admitted almost 3 months prior to RT, one would think that COVID would be gone in the mean time, right?

3. Needless to say, it took them 3 months to recruit 9 patients in a single-arm trial in a big Madrid hospital in spring. Rather disappointing, considering what Spain went through earlier this year.


My take home messages:
1. Another trial showing RT is safe (but we knew that already)
2. There seems to have been some clinical improvement in those patients, yet I am not sure if this can be attributed to RT only, since there is no control group.
3. A lot of details are still missing and I have serious doubts if the authos chose the correct patient group (-->heavily pretreated patients who probably no longer have much of an active inflammation).

Great walk through of study results, like Vinay Prasad except one fifth as annoying and one tenth as arrogant.

CT response is completely useless. This is like using MRI response for Avastin as your primary endpoint, "expensive steroid" and all that.
 
  • Like
  • Haha
Reactions: 1 users
Top