Esophagus dose escalation: ARTDECO trial results

[Gfunk6]

Just published in JCO: https://ascopubs.org/doi/10.1200/JCO.20.03697

260 patients (60% squam, 40% adeno) randomized to 50.4 Gy/1.8 or 61.6 Gy/1.8 with concurrent carbotaxol. No difference in any meaningful outcome.

Looks like Minsky was correct after all . . .

---

Members don't see this ad.

 

[medgator]

I wonder if package time is playing a role here. A few gi folks I've spoke to will try to dose paint the primary and still treat to no longer than 28-30 fx total if patients aren't surgical candidates

 

[TheWallnerus]

I wonder if package time is playing a role here. A few gi folks I've spoke to will try to dose paint the primary and still treat to no longer than 28-30 fx total if patients aren't surgical candidates

Did I just allude in another post that it was weird to take the package time for ENI in head neck cancer from 5 weeks to 7 weeks with no great data and expect the same outcomes? Yes the ship has sailed. But there was a lot of data in head neck carcinoma that time was an important factor. I bet it is in other carcinomas too. Hall gave us a mathematical way to consider time, but time is the red headed stepchild of radiobiology.

 

[Upgrayedd]

So how are you all adapting this for cervical/upper thoracic?

 

[ramsesthenice]

Just published in JCO: https://ascopubs.org/doi/10.1200/JCO.20.03697

260 patients (60% squam, 40% adeno) randomized to 50.4 Gy/1.8 or 61.6 Gy/1.8 with concurrent carbotaxol. No difference in any meaningful outcome.

Looks like Minsky was correct after all . . .

Well if you think about it here is the issue. They are looking at fairly modest dose escalation. Lets play this out for a moment. How much would you guess this would increase pCR in the primary tumor (which should be the best metric of local)? Lets be generous and say 10%. Looking at other disease sites the general rule of thumb is that by the time you factor in regional failures (outside of the primary) then your local regional control would be 60-80% of the advantage gained in the primary so you would predict that the improved LRC would be in the range of 6-7% at best which is pretty close to what they see (though not significant). Now for DFS you have to cut that number in half because of distant failures. Looking at the data I would say that their assumptions are too generous and the study was probably underpowered as a result. Would results be any better if they did an SIB and/or went a little higher? Maybe. But these look pretty modest at best.

 

[StIGMA]

May need BED >100, and invent a way to do that safely.

 

[Ray D. Ayshun]

I cant read it, but what up with the G5 toxicity?

 

[BobbyHeenan]

May need BED >100, and invent a way to do that safely.

Put some amifostine a milkshake just prior to radiation. Make sure it gets only in the normal esophagus but not in the tumor. Godspeed.

 

[Palex80]

Let's take a step back and think of this:

Can you think of any disease where conventionally fractionated dose escalation has shown an OS benefit in the past 30 years in a randomized trial in radiation oncology?

 

[deleted171718]

Let's take a step back and think of this:

Can you think of any disease where conventionally fractionated dose escalation has shown an OS benefit in the past 30 years in a randomized trial in radiation oncology?

Published 29 years and 8 months ago:

A Medical Research Council trial of two radiotherapy doses in the treatment of grades 3 and 4 astrocytoma.

 

[TheWallnerus]

Put some amifostine a milkshake just prior to radiation. Make sure it gets only in the normal esophagus but not in the tumor. Godspeed.

Let's take a step back and think of this:

Can you think of any disease where conventionally fractionated dose escalation has shown an OS benefit in the past 30 years in a randomized trial in radiation oncology?

If you accept that BID radiotherapy is conventional in LS-SCLC, then this "conventionally fractionated" dose escalation showed a big survival benefit. Have I met the conditions of your question? Close? *another wink emoji*

 

[Palex80]

Published 29 years and 8 months ago:

A Medical Research Council trial of two radiotherapy doses in the treatment of grades 3 and 4 astrocytoma.

Oh, you beat me to it by 4 months. Hats off!

 

[Palex80]

If you accept that BID radiotherapy is conventional in LS-SCLC, then this "conventionally fractionated" dose escalation showed a big survival benefit. Have I met the conditions of your question? Close? *another wink emoji*

No, no, no. That's acceleration!

But let's perhaps rephrase my question. Let's think of all the trials of conventionally fractionated dose escalation that failed to improve results.

We have the two negative trials on Grade 2 astrocytoma.
We have RTOG 0617 on NSCLC.
We have negative GBM trials (weren't 90 Gy attempted at some point?)
We have numerous trials on prostate cancer that proved only some bPFS advantage.
There were some attempts on HNSCC as well that also failed, although it gets complicated there due to chemotherapy being used as well.
We have the negative Minsky trial.
There's also the negative French PCI trial in LD-SCLC.
And I do believe it was even attempted in brain mets using dose escalated WBRT.


There's clearly a pattern...

 

[TheWallnerus]

No, no, no. That's acceleration!

But let's perhaps rephrase my question. Let's think of all the trial of conventionally fractionated dose escalation that failed to improve results.

We have the two negative trials on Grade 2 astrocytoma.
We have RTOG 0617 on NSCLC.
We have negative GBM trials (weren't 90 Gy attempted at some point?)
We have numerous trials on prostate cancer that proved only some bPFS advantage.
There were some attempts on HNSCC as well that also failed, although it gets complicated there due to chemotherapy being used as well.
We have the negative Minsky trial.
There's also the negative French PCI trial in LD-SCLC.
And I do believe it was even attempted in brain mets using dose escalated WBRT.


There's clearly a pattern...

Acceleration? The arm that gave improved survival was a 33% longer tx duration as you know. And 33% more dose; no alpha/beta adjustment necessary because the two arms were fractionated exactly equally. It was a very pure dose escalation study using the standardly, and I think conventionally, performed regimen... dose escalated. AGREE TO DISAGREE.

 

[Palex80]

Acceleration? The arm that gave improved survival was a 33% longer tx duration as you know. And 33% more dose; no alpha/beta adjustment necessary because the two arms were fractionated exactly equally. It was a very pure dose escalation study using the standardly, and I think conventionally, performed regimen... dose escalated. AGREE TO DISAGREE.

It's all accelerated. Both arms are accelerated. Disqualified!

Just kidding, I see your point. But I meant a "simple" 1.8-2.0 Gy/d fractionation.

 

[TheWallnerus]

It's all accelerated. Both arms are accelerated. Disqualified!

Just kidding, I see your point. But I meant a "simple" 1.8-2.0 Gy/d fractionation.

Since both arms are PERFECTLY equally accelerated they both exist in the same non-inertial reference frames. And thus relative to the observers in each frame they will appear non-accelerated to each other. And so to call on acceleration you will need fictitious forces. SO YOUR ARGUMENT IS FICTITIOUS.

 

[yesmaster]

conventionally fractionated dose escalation is a dead horse for most disease sites

 

[ramsesthenice]

No, no, no. That's acceleration!

But let's perhaps rephrase my question. Let's think of all the trials of conventionally fractionated dose escalation that failed to improve results.

We have the two negative trials on Grade 2 astrocytoma.
We have RTOG 0617 on NSCLC.
We have negative GBM trials (weren't 90 Gy attempted at some point?)
We have numerous trials on prostate cancer that proved only some bPFS advantage.
There were some attempts on HNSCC as well that also failed, although it gets complicated there due to chemotherapy being used as well.
We have the negative Minsky trial.
There's also the negative French PCI trial in LD-SCLC.
And I do believe it was even attempted in brain mets using dose escalated WBRT.


There's clearly a pattern...

Your observation goes hand in hand with my point earlier. Think about how many criteria need to be met in order for dose escalation to really even have a chance. The primary benefit theoretically is better local control translating into better survival. So:

There have to be a sizable number of local only failures (bye bye lung cancer)

You need to be able to escalate to a high enough dose that the escalation is meaningful (bye bye most large intracranial tumors)

And the disease has to be bad enough that a survival advantage is possible. Prostate is the closest thing we’ve got and still...no. I think the best dose escalation for prostate is EBRT + brachy. In gland control and bPFS look like they are probably better but not survival. Why? Because not enough people are dying of PC? Or because the ones that are dying already had Mets at the time of treatment and better local control didn’t translate into better distant control? Let’s take this exercise a little further. It’s probably safe to assume that surgery followed by postop probably has better local control than either alone (talking true a adjuvant, not salvage). We all know what that showed.

My take is that there are 2 big issues. For a lot of tumors, the problem is not that we don’t give enough dose, it’s that we don’t cover enough disease (and probably can’t by nature). Then there are the GBMs and pancreases of the world for which we don’t give enough (as evidenced by poor local control) but haven’t found a way to give high enough dose to make a difference.

 

[Radonky]

For cervical esophagus our institution isn't changing, we still try to boost due to princess margaret series in the red journal.