Photons vs. Protons for Esophageal Cancer

[Palex80]

Hey, I wanna hear your feedback on this one.

I just found out that this trial is running

protecttrial.eu

protecttrial.eu

A randomized Phase III trial on photons vs. protons for neoadjuvant chemorads in esophageal cancer.

The trial will cost roughly 5 million dollars

IMI Innovative Medicines Initiative | PROTECT-trial | Proton versus photon therapy for esophageal cancer - a trimodality strategy

Radiotherapy plays a key role in the treatment of many cancers. However, as conventional radiotherapy can cause side effects in surrounding organs, the dose is limited meaning that treatment takes longer and can be less effective. Proton therapy is an innovative form of radiotherapy. As the...

www.imi.europa.eu

The primary endpoint of the trial is pneumonitis >G1 and the trial is powered to detect a 10% difference.

Protocol

protecttrial.eu

Bearing in mind that pneumonitis was not a major issue in the completed, randomized phase II trial from the US
(Randomized Phase IIB Trial of Proton Beam Therapy Versus Intensity-Modulated Radiation Therapy for Locally Advanced Esophageal Cancer - PubMed), I wonder if the Phase III trial is worth the 5 millions spent?
Here are the pneumonitis events in the US trial

Accrual was 72 patients with IMRT, 73 with protons

And frankly, even if it comes out positive and there's a 12% difference in pneumonitis, but it's mainly grade 2 pneumonitis...

Thoughts?

---

Members don't see this ad.

 

[thecarbonionangle]

I agree with you. I have never seen “pmeumonitis” going to 50gy in esophagus with IMRT. It is likely to be negative for the primary endpoint.

 

[RickyScott]

I vaguely remember the mdacc trial of photons vs protons had more toxicity (basket of toxicity Endpt) .

 

[ramsesthenice]

This has been a classic "lets get this square peg in a round hole" from my perspective. MDACC published retrospective data suggesting protons had superior overall survival. This was particularly questionable even for retrospective data because they did not see a corresponding improvement in locoregional control which is almost a necessity if you are attributing improvements in survival to a local therapy. Then they did a phase 2B trial (Steven Lin was the first author) and saw not even a hint of difference in PFS or OS. No individual toxicities were improved but if they summed them all up to create a new total toxicity score, protons were better. At this point, the magnitude of clinical benefit achievable by just swapping photons with protons appears fairly limited.

 

[communitydoc13]

Signal searching. It's terrible.

Under any normal circumstance, the pithy "statistically significant but not clinically significant" conclusion would be the best they could hope for regarding a trial like this.

Of course, when there are billions in outlays, any signal will be used to justify the intervention.

edit: you do enough of these (small trials with dubious endpoints), you will get get some fortuitous Type I errors. This is the environment we are in at present.

 

[Gfunk6]

Hey, I wanna hear your feedback on this one.

I just found out that this trial is running

protecttrial.eu

protecttrial.eu

A randomized Phase III trial on photons vs. protons for neoadjuvant chemorads in esophageal cancer.

The trial will cost roughly 5 million dollars

IMI Innovative Medicines Initiative | PROTECT-trial | Proton versus photon therapy for esophageal cancer - a trimodality strategy

Radiotherapy plays a key role in the treatment of many cancers. However, as conventional radiotherapy can cause side effects in surrounding organs, the dose is limited meaning that treatment takes longer and can be less effective. Proton therapy is an innovative form of radiotherapy. As the...

www.imi.europa.eu

The primary endpoint of the trial is pneumonitis >G1 and the trial is powered to detect a 10% difference.

Protocol

protecttrial.eu

Bearing in mind that pneumonitis was not a major issue in the completed, randomized phase II trial from the US
(Randomized Phase IIB Trial of Proton Beam Therapy Versus Intensity-Modulated Radiation Therapy for Locally Advanced Esophageal Cancer - PubMed), I wonder if the Phase III trial is worth the 5 millions spent?
Here are the pneumonitis events in the US trial
View attachment 375997
Accrual was 72 patients with IMRT, 73 with protons

And frankly, even if it comes out positive and there's a 12% difference in pneumonitis, but it's mainly grade 2 pneumonitis...

Thoughts?

Thanks for posting the actual data. SDN likes to **** all over protons, but I have to admit the data between protons and photons is quite sobering re: pneumonitis. Photons increase the risk of Gr1 - Gr2 events by 100% and G3 events by ∞%.

If I had a cyclotron and was counseling patients with esophageal cancer, I would tell them "sure you don't have to travel to MDACC at your expense and find a local place in Houston to stay for 6 weeks, but if you didn't do that, your chance of getting G3 pneumonitis at podunk cancer community center is infinitely higher!"

 

[SneakyBooger]

I rarely see pneumonitis but I don't treat with large elective fields anymore either.

 

[RickyScott]

I rarely see pneumonitis but I don't treat with large elective fields anymore either.

I treat to 4140, but do keep fields large. Never had pneumonitis.

 

[ramsesthenice]

I don't think I have ever seen symptomatic pneumonitis either although admittedly it could pretty easily be masked by other more common side effects. Thing is, clinical pneumonitis isn't really the lung issue protons are really primed to help in my mind. The volume of lung irradiated does appear to correlate with post-op complications (like weaning off the vent). I have an easy time buying that protons could be useful in this regard. It would just be exceptionally hard to show since we have not done a great job quantifying any of this in unselected patients before.

 

[evilbooyaa]

I wonder if by pneumonitis they mean more along lines of post-op ARDS which is where the initial V5 data in esophagus < 50% or whatever came from. True radiation induced pneumonitis is very unlikely. Would be in line with reducing the low dose bath and trying to prevent post-op ARDS... but reducing a radiographic sign alone in an otherwise asymptomatic patient seems.... a bit silly.

 

[Palex80]

I wonder if by pneumonitis they mean more along lines of post-op ARDS which is where the initial V5 data in esophagus < 50% or whatever came from. True radiation induced pneumonitis is very unlikely. Would be in line with reducing the low dose bath and trying to prevent post-op ARDS... but reducing a radiographic sign alone in an otherwise asymptomatic patient seems.... a bit silly.

You are absolutely correct, they seem to be looking also at postoperative lung complications
The primary outcome is the incidence of pulmonary complications during and following nCPT or nCXT and surgery. The proportion of patients with pulmonary complications will be compared between the arms. Complications are scored following CTCAE v5.0, grade ≥ 2 from nCPT or nCXT start until surgery & following Clavien-Dindo, grade ≥ 2 from surgery to 90 days after surgery.

So, indeed, pneumonitis may not be the sole difference seen between protons and photons, but if patients get a postooperative pneumonia, that would count too.

 

[evilbooyaa]

Reasonable endpoint then. G2+ lung complications after surgery. Small subset of my esophageal patients end up going for surgery though....

 

[deleted1111261]

Then that’s worth exploring
More and more seem to get surgery now

 

[Palex80]

Reasonable endpoint then. G2+ lung complications after surgery. Small subset of my esophageal patients end up going for surgery though....

I am still not convinced.

Clavien Dindo Classification

Grades	Definition
Grade I	Any deviation from the normal postoperative course without the need for pharmacological treatment or surgical, endoscopic and radiological interventions Allowed therapeutic regimens are: drugs as antiemetics, antipyretics, analgetics, diuretics and electrolytes and physiotherapy. This grade also includes wound infections opened at the bedside.
Grade II	Requiring pharmacological treatment with drugs other than such allowed for grade I complications. Blood transfusionsand total parenteral nutritionare also included.
Grade III	Requiring surgical, endoscopic or radiological intervention
- IIIa	Intervention not under general anesthesia
- IIIb	Intervention under general anesthesia
Grade IV	Life-threatening complication (including CNS complications)* requiring IC/ICU-management
- IVa	single organ dysfunction (including dialysis)
- IVb	multiorgandysfunction
Grade V	Death of a patient

A grade 2 Clavien-Dindo pneumonia is a pneumonia that has to be treated with antibiotics.
I bet pretty much any pneumonia up to 90d post esophagectomy get‘s treated with antibiotics.

Would a 10% higher rate of any pneumonia be enough to push away from photons to protons?

I would need a harder and more patient-centric endpoint to convince me, for instance:
- admission to ICU
- prolongation of hospital stay
…


The chosen endpoint is so prone to bias. The patient coughs once on the 90d follow-up, I prescribe him a symbicort inhaler and it‘s an event! Thinking about it, it´s even more prone to bias than pneumonitis (which I would have to prove first)!

 

[ramsesthenice]

Reasonable endpoint then. G2+ lung complications after surgery. Small subset of my esophageal patients end up going for surgery though....

How are they managing the curative patients in your neck of the woods then? I bet 80%+ of my folks get surgery.

 

[communitydoc13]

I looked at the trial. I believe the accrual goals are reasonable. We will have to look at the outcomes. I believe all pulm toxicity >G1 is included. This can include really soft things, like PNA tx with abx. Rare but severe late effects however, like fistulas (which I believe are more likely to occur with protons) are unlikely to be included due to the short window of toxicity assessment (90 days from CRT or Surgery).

Better than I thought it was. At least it's a trial.

 

[RickyScott]

Depends on ge junction vs other sites. Almost all ge junction get surgery in my experience

How are they managing the curative patients in your neck of the woods then? I bet 80%+ of my folks get surge

 

[ramsesthenice]

Depends on ge junction vs other sites. Almost all ge junction get surgery in my experience

Me too. Our med oncs (wisely) agree that perioperative FLOT is more toxic for most people than esophageal radiation and are in no hurry to drop radiation.

 

[communitydoc13]

The chosen endpoint is so prone to bias. The patient coughs once on the 90d follow-up, I prescribe him a symbicort inhaler and it‘s an event! Thinking about it, it´s even more prone to bias than pneumonitis (which I would have to prove first)!

Low grade (Grade 2 on down) toxicity is one of the hardest things to grade, much less study as an endpoint. In non-inferiority trials (often for hypofractionation in our field) there are clearly differences in low grade toxicity that are ignored or buried in Supplementary materials in support of the conclusion that two interventions are "meaningfully equivalent".

I am hopeful that this group will publish their comprehensive raw data.

 

[RickyScott]

I am very skeptical of protons because up until 2018 or so we really didn’t hear of side effect, even though they almost certainly were occurring with increased frequency given proton technology at the time (and lack of igrt). Upenn and mdacc and Uf all started around 2007.

 

[Gfunk6]

I am still not convinced.

Clavien Dindo Classification

Grades	Definition
Grade I	Any deviation from the normal postoperative course without the need for pharmacological treatment or surgical, endoscopic and radiological interventions Allowed therapeutic regimens are: drugs as antiemetics, antipyretics, analgetics, diuretics and electrolytes and physiotherapy. This grade also includes wound infections opened at the bedside.
Grade II	Requiring pharmacological treatment with drugs other than such allowed for grade I complications. Blood transfusionsand total parenteral nutritionare also included.
Grade III	Requiring surgical, endoscopic or radiological intervention
- IIIa	Intervention not under general anesthesia
- IIIb	Intervention under general anesthesia
Grade IV	Life-threatening complication (including CNS complications)* requiring IC/ICU-management
- IVa	single organ dysfunction (including dialysis)
- IVb	multiorgandysfunction
Grade V	Death of a patient

The chosen endpoint is so prone to bias. The patient coughs once on the 90d follow-up, I prescribe him a symbicort inhaler and it‘s an event! Thinking about it, it´s even more prone to bias than pneumonitis (which I would have to prove first)!

The grade I endpoint makes me really scratch my head. Apparently if you prescribe a patient ondansetron, acetominophen, furosemide or "physiotherapy" that is considerd grade I pneumonitis . . .

 

[NotMattSpraker]

I looked at the trial. I believe the accrual goals are reasonable. We will have to look at the outcomes. I believe all pulm toxicity >G1 is included. This can include really soft things, like PNA tx with abx. Rare but severe late effects however, like fistulas (which I believe are more likely to occur with protons) are unlikely to be included due to the short window of toxicity assessment (90 days from CRT or Surgery).

Better than I thought it was. At least it's a trial.

I like the end point. It's a tough space to design a trial. FWIW, I like the NRG GI 006 design too.

From the IMI website

Certainly we could all debate the merits of the design, but I wish the US did stuff this way. It's so... professionally scientific.

 

[deleted1111261]

I have faith in the European study

 

[NotMattSpraker]

I have faith in the European study

Im a total toxicity burden "believer", but would be pleasantly surprised if the european trial is positive. We don't know a lot about intermediate lung dose other than V20/mean are related to observed radiation pneumonitis. GI 006 has cardiopulmonary toxicity as a co-primary. We will learn a lot from these trials and everyone should have relative equipoise compared to many questions, it's a shame they aren't enrolling like gangbusters.

 

[evilbooyaa]

How are they managing the curative patients in your neck of the woods then? I bet 80%+ of my folks get surgery.

I don't see a ton of esophagus, but the ones I saw (all adenoCA) always had a surgeon involved who told them the pros/cons and a ton of folks didn't want the post esophagectomy lifestyle or weren't candidates due to extensiveness of say nodal disease.

 

[evilbooyaa]

I am still not convinced.

Clavien Dindo Classification

Grades	Definition
Grade I	Any deviation from the normal postoperative course without the need for pharmacological treatment or surgical, endoscopic and radiological interventions Allowed therapeutic regimens are: drugs as antiemetics, antipyretics, analgetics, diuretics and electrolytes and physiotherapy. This grade also includes wound infections opened at the bedside.
Grade II	Requiring pharmacological treatment with drugs other than such allowed for grade I complications. Blood transfusionsand total parenteral nutritionare also included.
Grade III	Requiring surgical, endoscopic or radiological intervention
- IIIa	Intervention not under general anesthesia
- IIIb	Intervention under general anesthesia
Grade IV	Life-threatening complication (including CNS complications)* requiring IC/ICU-management
- IVa	single organ dysfunction (including dialysis)
- IVb	multiorgandysfunction
Grade V	Death of a patient

A grade 2 Clavien-Dindo pneumonia is a pneumonia that has to be treated with antibiotics.
I bet pretty much any pneumonia up to 90d post esophagectomy get‘s treated with antibiotics.

Would a 10% higher rate of any pneumonia be enough to push away from photons to protons?

I would need a harder and more patient-centric endpoint to convince me, for instance:
- admission to ICU
- prolongation of hospital stay
…


The chosen endpoint is so prone to bias. The patient coughs once on the 90d follow-up, I prescribe him a symbicort inhaler and it‘s an event! Thinking about it, it´s even more prone to bias than pneumonitis (which I would have to prove first)!

I mean are the surgeons/inpatient team managing that post-op pneumonia going to a) know and b) care about who got protons vs photons in neoadjuvant setting to 'bias' the results?

I think G2 vs G3+ is going to be an interesting evaluation. The tough part about G2 is that even that is on a spectrum from needing an inhaler once to needing long-term Abx and BiPAP. I agree that the former is not that big of a deal, while the latter might be. Certainly anyone needing ICU level care being a grade IV will help parse things out a little bit.

 

[fwillison]

I've seen grade 2 pneumonitis (NCI-CTC) after Esophageal irradiation using photon IMRT, but only very occasionally. Never had a Grade 3 or 4 that I am aware of. Since the grade 2 patients are relatively easily and effectively managed, it is hard to imagine this trial moving the needle in this setting.

 

[deleted1111261]

I’m getting a little more convinced by Proton friends about value here. The heart is very spared. There is a significant risk of cardiac toxicity. It is probably one of the sites where I feel no issue if they seek proton eval

 

[communitydoc13]

I’m getting a little more convinced by Proton friends about value here. The heart is very spared. There is a significant risk of cardiac toxicity. It is probably one of the sites where I feel no issue if they seek proton eval

Let's see. There is a downside to 140% hotspots that you don't know are there in an organ like the esophagus.

Now the liver?

 

[evilbooyaa]

Let's see. There is a downside to 140% hotspots that you don't know are there in an organ like the esophagus.

Now the liver?

Potential 140% hot spot in the esophagus doesn't matter if you immediately resect the affected area afterwards

I am a proton skeptic at heart. But agree with Simul. Let's see what the data shows.

 

[communitydoc13]

Potential 140% hot spot in the esophagus doesn't matter if you immediately resect the affected area afterwards

Agree, follow the data.

Upper esophagus within mm of trachea, all esophagus is pretty intimate to the pleura. Necrosis and fistula post-up can happen.

If the data is good, it's good.

 

[fwillison]

Regarding protons, building treatment facilities first, based to a large extent on economics, and then running trials to demonstrate the value of that treatment, is likely to lead to unreliable results.
However, I do agree that the heart is the issue of concern, rather than lung, esophagus or other OAR. It is certainly possible that late cardiac toxicity is improved with proton techniques. In fact, given what we know about the dosimetry of photons vs proton radiation, and the long term sensitivity of the heart to relatively low doses of radiation, I would expect protons to be superior on this aspect. Clinically, the issue will be the magnitude of the effect, especially given the late nature of the complication in a disease with historically low rates of long term survival. But still, a reasonable thing to study.

 

[NotMattSpraker]

Potential 140% hot spot in the esophagus doesn't matter if you immediately resect the affected area afterwards

I am a proton skeptic at heart. But agree with Simul. Let's see what the data shows.

Same, and Im least skeptical about it in the pre-op esophagus setting.

 

[fwillison]

Same, and Im least skeptical about it in the pre-op esophagus setting.

True, but I think we are doing esophagectomies on many patients who should not be having them.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8591817/pdf/13014_2021_Article_1947.pdf

 

[communitydoc13]

Same, and Im least skeptical about it in the pre-op esophagus setting.

I think we need to look at global outcomes here.

I would reference perioperative toxicity outcomes in CROSS for some context. Very little difference in toxicity between neoadjuvant chemorads and surgery alone. So I'm guessing we are looking at a population in the Lin trial that is not representative of most triple modality esophageal patients (or else they looked a lot harder at toxicity).

In the Lin trial (small trial), most patients were viewed as eligible for triple modality therapy and only about half got surgery (seems a little weird, Cross trial got folks overwhelmingly to surgery). Target XRT dose was 50.4.

The total toxicity burden was greater in the photon group, and this was driven by roughly 1/3 of photon patients (ten patients) who got operated on having significant peri-operative toxicity and almost none in the proton group having significant peri-operative toxicity (again a result not reflective of CROSS data as I know it).

But, pFS the same, OS the same (and the photon curve looked better), QOL the same (despite a lower fraction of proton patients going for surgery).

So despite 1/3 of photon patients having a really tough surgical course, there was no manifestation of this in terms of longitudinal outcomes.

I wouldn't be surprised if they get some positive results from this trial. However, I want to look at the long term and total outcomes and then I want to address what endorsing Protons in this setting really means.

Does it matter at 41.4 Gy?
Does it matter if no surgery?

 

[NotMattSpraker]

I think we need to look at global outcomes here.

I would reference perioperative toxicity outcomes in CROSS for some context. Very little difference in toxicity between neoadjuvant chemorads and surgery alone. So I'm guessing we are looking at a population in the Lin trial that is not representative of most triple modality esophageal patients (or else they looked a lot harder at toxicity).

In the Lin trial (small trial), most patients were viewed as eligible for triple modality therapy and only about half got surgery (seems a little weird, Cross trial got folks overwhelmingly to surgery). Target XRT dose was 50.4.

The total toxicity burden was greater in the photon group, and this was driven by roughly 1/3 of photon patients (ten patients) who got operated on having significant peri-operative toxicity and almost none in the proton group having significant peri-operative toxicity (again a result not reflective of CROSS data as I know it).

But, pFS the same, OS the same (and the photon curve looked better), QOL the same (despite a lower fraction of proton patients going for surgery).

So despite 1/3 of photon patients having a really tough surgical course, there was no manifestation of this in terms of longitudinal outcomes.

I wouldn't be surprised if they get some positive results from this trial. However, I want to look at the long term and total outcomes and then I want to address what endorsing Protons in this setting really means.

Does it matter at 41.4 Gy?
Does it matter if no surgery?

Great questions. I'd add too, especially for GE junction, does the field size matter. I used to work with a surgeon that would often ask me to truncate the PTV so that it did not extend past the carina in lower esoph cases. He thought this mattered. I would do it as long as it fell within the standard range of volume extent.

I could also see a subtle finding like TTB at a single highly skilled, high volume institution converting to more easily understandable positive finding in the phase III multi-center setting.

Good thoracic RT (photon and proton) seems to modulate cardiothoracic risk in a pretty subtle way for deadly advanced diseases such as esophagus and NSCLC. This is just my gut feeling based on my read of the literature. I view TTB not as a gamed result, but as enhancing our ability to detect subtle toxicity. It possible we are missing this stuff with individual toxicity or QoL measures, which are quite subjective and noisy. It's possible it's a false positive too of course, that's why we need phase III trials.

If the patients are a bit sicker, surgeons or rad oncs a bit less skilled, post-op care a bit worse, we might see a difference. I didn't love it at first, but the way GI006 is set up is really growing on me.

 

[communitydoc13]

One thing about TTB is that it is a little fuzzy (arbitrary?). They assign a numerical value to toxicity on a scale established by investigators. This is not a dichotomous outcome.

Not my favorite, but then neither are using standard tools to asses statistical significance for differences in toxicity. In the Lin paper, the raw data was available and indicated (in that very small group) that acute perioperative toxicity was worse for their photon cohort.

What TTB lets you do is find signal in rare events if they are bad enough and are assigned a high enough value. We could certainly have come up with a TTB scale that would have revealed that 5 fraction whole breast is a bit more toxic!

The irony (to me) is that one could likely come up with a TTB that demonstrates inferior toxicity outcomes for protons in sites like breast (assign breast shrinkage a score of 10 and rib fracture a score of 90) or head and neck (xerostomia scored a 10, acute mucositis a 10 and osteonecrosis or pharyngeal necrosis a 100).

So, I think we need to be careful any time we see this type of analysis. An author could game their toxicity scale in a way to favor expectations regarding differential toxicity profiles.

 

[NotMattSpraker]

One thing about TTB is that it is a little fuzzy (arbitrary?). They assign a numerical value to toxicity on a scale established by investigators. This is not a dichotomous outcome.

Not my favorite, but then neither are using standard tools to asses statistical significance for differences in toxicity. In the Lin paper, the raw data was available and indicated (in that very small group) that acute perioperative toxicity was worse for their photon cohort.

What TTB lets you do is find signal in rare events if they are bad enough and are assigned a high enough value. We could certainly have come up with a TTB scale that would have revealed that 5 fraction whole breast is a bit more toxic!

The irony (to me) is that one could likely come up with a TTB that demonstrates inferior toxicity outcomes for protons in sites like breast (assign breast shrinkage a score of 10 and rib fracture a score of 90) or head and neck (xerostomia scored a 10, acute mucositis a 10 and osteonecrosis or pharyngeal necrosis a 100).

So, I think we need to be careful any time we see this type of analysis. An author could game their toxicity scale in a way to favor expectations regarding differential toxicity profiles.

Agree, but there is a significant difference in the design and context of the Lin and Choi trials and it has little to do with the end points.

Lin states you should enroll people on GI 006 because it might be better than our current standard. Choi states that Proton APBI is safe and effective, ignoring the current standard. The Penn guys essentially call her out in their paper by recommending that you should treat with photon Florence.

Someone said it before somewhere, the biggest harm was opening all these centers for financial reasons and then trying to backfill the data to justify the centers.

 

[fwillison]

Someone said it before somewhere, the biggest harm was opening all these centers for financial reasons and then trying to backfill the data to justify the centers.

Right, the inherent antecedent bias renders any data suspect that is not the most unassailable.

 

[IonsAreOurFuture]

Agree, but there is a significant difference in the design and context of the Lin and Choi trials and it has little to do with the end points.

Lin states you should enroll people on GI 006 because it might be better than our current standard. Choi states that Proton APBI is safe and effective, ignoring the current standard. The Penn guys essentially call her out in their paper by recommending that you should treat with photon Florence.

Someone said it before somewhere, the biggest harm was opening all these centers for financial reasons and then trying to backfill the data to justify the centers.

We have GI 006 open at my center and have begun enrolling. I've corresponded with Dr. Lin about the design and one of the more interesting endpoints they've chosen to focus on is lymphopenia, like grade 3 lymphopenia, which is ALC <500 I believe. He was very generous in explaining it to me, some of which is below.

ALC is potentially a prognostic marker for response to immunotherapy, and overall survival. One study in melanoma showed that in patients who develop any grade of lymphopenia, defined as ALC <1000, while on immunotherapy, their treatment generally stops working, and their 5-year Overall Survival rate drops from ~75% to ~50%. Impressive OS graphs, it's like you undid the benefit of immunotx in the PACIFIC trial.

We routinely hold radiation for patients with ANC of <500, but don't bat an eye at an ALC < 500, yet it may be just as important if not more so. I personally think that lymphopenia is becoming the new neutropenia, except that we don't yet have the equivalent of a neupogen or neulasta to help treat it, and we don't know how or if to modify treatment once we see it.

Based on the results of Checkmate 577 in the NEJM, the standard of care for esophagus has recently changed to include one year of adjuvant nivolumab for patients whose esophagectomy specimen shows less than a CR to chemo-radiation. Adjuvant nivolumab doubled their median DFS from 11 months vs 22.4 mos with Nivo.

If patients are severely lymphopenic during immunotherapy, I think there could be a scenario where PFS and maybe even OS is impacted. Not due to radiation not working well pre-op, but due to immunotherapy not working well postop. Of course, the trial will still be accruing for a couple years but I'm very interested to see how it turns out, as it seems like just about everyone will be on immunotherapy before too long.

 

[RickyScott]

Radiation is one drug away from elimination in esophageal ca.

 

[temujim]

We have GI 006 open at my center and have begun enrolling. I've corresponded with Dr. Lin about the design and one of the more interesting endpoints they've chosen to focus on is lymphopenia, like grade 3 lymphopenia, which is ALC <500 I believe. He was very generous in explaining it to me, some of which is below.

ALC is potentially a prognostic marker for response to immunotherapy, and overall survival. One study in melanoma showed that in patients who develop any grade of lymphopenia, defined as ALC <1000, while on immunotherapy, their treatment generally stops working, and their 5-year Overall Survival rate drops from ~75% to ~50%. Impressive OS graphs, it's like you undid the benefit of immunotx in the PACIFIC trial.

We routinely hold radiation for patients with ANC of <500, but don't bat an eye at an ALC < 500, yet it may be just as important if not more so. I personally think that lymphopenia is becoming the new neutropenia, except that we don't yet have the equivalent of a neupogen or neulasta to help treat it, and we don't know how or if to modify treatment once we see it.

Based on the results of Checkmate 577 in the NEJM, the standard of care for esophagus has recently changed to include one year of adjuvant nivolumab for patients whose esophagectomy specimen shows less than a CR to chemo-radiation. Adjuvant nivolumab doubled their median DFS from 11 months vs 22.4 mos with Nivo.

If patients are severely lymphopenic during immunotherapy, I think there could be a scenario where PFS and maybe even OS is impacted. Not due to radiation not working well pre-op, but due to immunotherapy not working well postop. Of course, the trial will still be accruing for a couple years but I'm very interested to see how it turns out, as it seems like just about everyone will be on immunotherapy before too long.

Was quite surprised to see grade 3-4 lymphopenia of 24% on the RT only arm (60 Gy in 5 weeks) of HN002. Dose to blood vessels was enough? The bone marrow in the C-spine? Started considering this impact more and more particularly in other sites where plan integrates immunotherapy.

 

[RickyScott]

Was quite surprised to see grade 3-4 lymphopenia of 24% on the RT only arm (60 Gy in 5 weeks) of HN002. Dose to blood vessels was enough? The bone marrow in the C-spine? Started considering this impact more and more particularly in other sites where plan integrates immunotherapy.

Hypofract probable decreases lymphopenia

 

[communitydoc13]

Hypofract probable decreases lymphopenia

IMRT probably increases lymphopenia.

It takes close to nothing to kill lymphocytes and they circulate fast.

But,…no consistent clinical outcome correlation (with regards to XRT induced lymphopenia, lymphopenia in general is prognostic). This is not a new narrative and people were looking into this back when Dr. Lin was in residency.

 

[NotMattSpraker]

We routinely hold radiation for patients with ANC of <500, but don't bat an eye at an ALC < 500, yet it may be just as important if not more so.

I loved everything you wrote, but this surprised me. Routinely?

Lymphopenia is interesting but the finding right now is a correlation. I (used to) tell patients that the biggest potential gain we think we know right now is reducing post-op complications.

I look forward to these results!

 

[RickyScott]

pts with lymphopenia during radiation for gbm also do so much worse. I believe there was a trial at Hopkins where lymphocytes were collected prior to radiation and infused to mantain counts. Made 0 difference. Vitamin D levels are also decreased in most cancer and serious illnesses.

 

[Palex80]

Interventions that increase counts/levels may not alter the course of the disease.

Remember the hematopoetin argument for H&N cancer? We know that anemia is linked to bad outcomes in H&N cancer during RT. Some link that to poor oxygenation of cells. When we tried raising the levels, it didn't change prognosis.

 

[evilbooyaa]

Learned the same painful lesson about futility with transfusing cervical cancer patients up to Hb 10...

But let's see what this one shows!

 

[SneakyBooger]

I am still not convinced.

Clavien Dindo Classification

Grades	Definition
Grade I	Any deviation from the normal postoperative course without the need for pharmacological treatment or surgical, endoscopic and radiological interventions Allowed therapeutic regimens are: drugs as antiemetics, antipyretics, analgetics, diuretics and electrolytes and physiotherapy. This grade also includes wound infections opened at the bedside.
Grade II	Requiring pharmacological treatment with drugs other than such allowed for grade I complications. Blood transfusionsand total parenteral nutritionare also included.
Grade III	Requiring surgical, endoscopic or radiological intervention
- IIIa	Intervention not under general anesthesia
- IIIb	Intervention under general anesthesia
Grade IV	Life-threatening complication (including CNS complications)* requiring IC/ICU-management
- IVa	single organ dysfunction (including dialysis)
- IVb	multiorgandysfunction
Grade V	Death of a patient

A grade 2 Clavien-Dindo pneumonia is a pneumonia that has to be treated with antibiotics.
I bet pretty much any pneumonia up to 90d post esophagectomy get‘s treated with antibiotics.

Would a 10% higher rate of any pneumonia be enough to push away from photons to protons?

I would need a harder and more patient-centric endpoint to convince me, for instance:
- admission to ICU
- prolongation of hospital stay
…


The chosen endpoint is so prone to bias. The patient coughs once on the 90d follow-up, I prescribe him a symbicort inhaler and it‘s an event! Thinking about it, it´s even more prone to bias than pneumonitis (which I would have to prove first)!

Agreed.

I too am not convinced. Feel like this is another study trying to pull the wool over my eyes...