ASTRO 2024

[radiation]

MSKCC is picking their patients based on FMISO PET and that seems like the winning strategy. A 10Gy decrease isn't moving the needle far enough to tempt me to exchange PFS in HN.

FMISO PET is unavailable for 99% of the country and the data is based on a non-randomized phase 2 study, which is same level of evidence HN005 was based on.

Nobody should be recommending de-escalation in unselected HPV+ pts based on HN005. However, in the absence of phase 3 data for FMISO PET, which is unlikely to ever happen, what is the information that is going to be needed to pick the right pts for de-escalation? Or is the volume de-escalation going to be the only option for 99% people in this country in the next decade?

A reasonable start for me is to closely look at the pts that progressed on HN005 in the de-escalated arm and see what we can learn biologically and clinically about those pts.

---

Members don't see this ad.

 

[RickyScott]

FMISO PET is unavailable for 99% of the country and the data is based on a non-randomized phase 2 study, which is same level of evidence HN005 was based on.

Nobody should be recommending de-escalation in unselected HPV+ pts based on HN005. However, in the absence of phase 3 data for FMISO PET, which is unlikely to ever happen, what is the information that is going to be needed to pick the right pts for de-escalation? Or is the volume de-escalation going to be the only option for 99% people in this country in the next decade?

A reasonable start for me is to closely look at the pts that progressed on HN005 in the de-escalated arm and see what we can learn biologically and clinically about those pts.

Were deesculated patients permitted to be smokers?

 

[radiation]

Were deesculated patients permitted to be smokers?

10 pack years or less

 

[grenz]

FMISO PET is unavailable for 99% of the country and the data is based on a non-randomized phase 2 study, which is same level of evidence HN005 was based on.

Nobody should be recommending de-escalation in unselected HPV+ pts based on HN005. However, in the absence of phase 3 data for FMISO PET, which is unlikely to ever happen

x.com

x.com

Phase 3 enrolling. I think everyone is so desperate to find a way to de-escalate, I bet it will accrue

 

[RickyScott]

I think the future here is to de-escalate and decrease elective nodal xrt. Overwhelming majority of failures are in the gtv.

 

[radoncftw]

I think the handwaving regarding the results of PARTIQOL is really sad.
I went to the meeting. I listened to the PI, the discussant, and the Red J podcast.
It reminds me of the prostate hypofx studies. Except of course we know how that went, despite the fact that many of the hypofx studies showed increased toxicity.
How you can justify the use of protons for low/int risk pros ca moving forward is beyond me.
Curious to see if a cost effectiveness analysis will be run? Those seemed to be all the rage.

 

[RickyScott]

I think the handwaving regarding the results of PARTIQOL is really sad.
I went to the meeting. I listened to the PI, the discussant, and the Red J podcast.
It reminds me of the prostate hypofx studies. Except of course we know how that went, despite the fact that many of the hypofx studies showed increased toxicity.
How you can justify the use of protons for low/int risk pros ca moving forward is beyond me.
Curious to see if a cost effectiveness analysis will be run? Those seemed to be all the rage.

How can a cost effectiveness analysis be run in the absence of actual prices ie costs? I have yet to see one in this field.

 

[NotMattSpraker]

I think the handwaving regarding the results of PARTIQOL is really sad.
I went to the meeting. I listened to the PI, the discussant, and the Red J podcast.
It reminds me of the prostate hypofx studies. Except of course we know how that went, despite the fact that many of the hypofx studies showed increased toxicity.
How you can justify the use of protons for low/int risk pros ca moving forward is beyond me.
Curious to see if a cost effectiveness analysis will be run? Those seemed to be all the rage.

I thought the podcast was mostly good. I loved the special mention that people overstate the harm of protons… but I guess no mention of how proton centers routinely overstate the benefit. Haha not biased at all 🙄

Dr E had some really good points and insight about things they did to get the trial done. Definitely worth hearing if people haven’t listened.

 

[radoncftw]

I thought the podcast was mostly good. I loved the special mention that people overstate the harm of protons… but I guess no mention of how proton centers routinely overstate the benefit. Haha not biased at all 🙄

Dr E had some really good points and insight about things they did to get the trial done. Definitely worth hearing if people haven’t listened.

I agree - overall the podcast was good. It actually made me go back through and listen to more of them. They talked about other interesting stuff too including the adjuvant bladder trial.

I was going to make an analogy of protons vs. IMRT to drug A vs. Drug B but realized in the context of big pharma maybe that was not the best comparison.

 

[geniusindisguise]

I know this is not ASTRO-related but did want to see what y'all thought about this without creating a brand new post. Of course will need to see full data (likely to be presented at ASCO?), but will it potentially change SOC?

Merck’s KEYTRUDA® (pembrolizumab) Met Primary Endpoint of Event-Free Survival (EFS) as Perioperative Treatment Regimen in Patients With Resected, Locally Advanced Head and Neck Squamous Cell Carcinoma - Merck.com

KEYNOTE-689 is the first Phase 3 trial to demonstrate statistically significant and clinically meaningful improvement in EFS in the intent-to-treat population in the neoadjuvant and adjuvant setting for an anti-PD-1 therapy in earlier stages of head and neck squamous cell carcinoma Merck (NYSE...

www.merck.com

 

[lilshopta]

I know this is not ASTRO-related but did want to see what y'all thought about this without creating a brand new post. Of course will need to see full data (likely to be presented at ASCO?), but will it potentially change SOC?

Merck’s KEYTRUDA® (pembrolizumab) Met Primary Endpoint of Event-Free Survival (EFS) as Perioperative Treatment Regimen in Patients With Resected, Locally Advanced Head and Neck Squamous Cell Carcinoma - Merck.com

KEYNOTE-689 is the first Phase 3 trial to demonstrate statistically significant and clinically meaningful improvement in EFS in the intent-to-treat population in the neoadjuvant and adjuvant setting for an anti-PD-1 therapy in earlier stages of head and neck squamous cell carcinoma Merck (NYSE...

www.merck.com

Everyone in the study received adj RT+/- chemo.

I think the findings are especially relevant for disease sites where surgery is SOC like T4 larynx, oral cavity, and possibly HPV- cancer. But my concerns are 1. are we going to start patients on neoadj IO and delay surgery/risk of disease progression? 2. For p16+ oropharynx, are we really going to treat them with surgery+RT +/- chemo over definitive chemorads? Seems like trimodality therapy would be very toxic.

 

[RickyScott]

Z

Everyone in the study received adj RT+/- chemo.

I think the findings are especially relevant for disease sites where surgery is SOC like T4 larynx, oral cavity, and possibly HPV- cancer. But my concerns are 1. are we going to start patients on neoadj IO and delay surgery/risk of disease progression? 2. For p16+ oropharynx, are we really going to treat them with surgery+RT +/- chemo over definitive chemorads? Seems like trimodality therapy would be very toxic.

maybe you can skip the radiation after the surgery.. thats where this roads leading/

 

[madchemist89]

Med onc will jump on the Keytruda like they did in lung.

 

[OTN]

I'll take anything we can get for oral cavity cancer

 

[RickyScott]

I'll take anything we can get for oral cavity cancer

Good point

 

[evilbooyaa]

I know this is not ASTRO-related but did want to see what y'all thought about this without creating a brand new post. Of course will need to see full data (likely to be presented at ASCO?), but will it potentially change SOC?

Merck’s KEYTRUDA® (pembrolizumab) Met Primary Endpoint of Event-Free Survival (EFS) as Perioperative Treatment Regimen in Patients With Resected, Locally Advanced Head and Neck Squamous Cell Carcinoma - Merck.com

KEYNOTE-689 is the first Phase 3 trial to demonstrate statistically significant and clinically meaningful improvement in EFS in the intent-to-treat population in the neoadjuvant and adjuvant setting for an anti-PD-1 therapy in earlier stages of head and neck squamous cell carcinoma Merck (NYSE...

www.merck.com

Excerpt from the press release:

"The study also showed a statistically significant improvement in major pathological response (mPR), a key secondary endpoint, for patients in the KEYTRUDA arm compared with adjuvant radiotherapy alone."

Am I dumb or is this the above line one of the stupidest things I've read in a while - 'Neoadjuvant therapy showed an improvement in major pathological response compared to going straight to surgery'

Great, so Keytruda did literally anything more significant than the placebo effect.

Will be strongly interested to see what percentage of patients started ICI but then didnot undergo surgery, similar to the Checkmate trial in neochemo + ICI in lung. And whether they played games with definition of PFS.

Will also be interesting to see breakdown by initial site of disease - I really really hope HPV+ Ophx pts were not harmed by being included on this study given the excellent outcomes with bimodality therapy let alone the toxicity of quadmodality therapy, and that this was mostly (exclusively) things like T4 larynx, oral cavity.

If anyone sees a full abstract or presentation or especially a paper in this space, would love to delve deeper.

 

[geniusindisguise]

Excerpt from the press release:

"The study also showed a statistically significant improvement in major pathological response (mPR), a key secondary endpoint, for patients in the KEYTRUDA arm compared with adjuvant radiotherapy alone."

Am I dumb or is this the above line one of the stupidest things I've read in a while - 'Neoadjuvant therapy showed an improvement in major pathological response compared to going straight to surgery'

Great, so Keytruda did literally anything more significant than the placebo effect.

Will be strongly interested to see what percentage of patients started ICI but then didnot undergo surgery, similar to the Checkmate trial in neochemo + ICI in lung. And whether they played games with definition of PFS.

Will also be interesting to see breakdown by initial site of disease - I really really hope HPV+ Ophx pts were not harmed by being included on this study given the excellent outcomes with bimodality therapy let alone the toxicity of quadmodality therapy, and that this was mostly (exclusively) things like T4 larynx, oral cavity.

If anyone sees a full abstract or presentation or especially a paper in this space, would love to delve deeper.

The above table from Multi-D 2024 and a meta-analysis of phase II studies from a while back showed surgical delay rates ~1% to PD-1 alone (10.1001/jamaoto.2021.2191). So overall, surgical delay rates pretty low for a couple doses of single agent PD1.

Still, pCR/MPR rates are pretty low for HNSCC at ~2%/9% based on the phase IIs. Also, for most tumors outside of melanoma, cSCC, and dMMR CRC, there's a chemo backbone for neoadjuvant Tx since the response rates just aren't that great to single agent PD-(L)1. Curious to see how "clinical meaningful" keytruda will be.

 

[temujim]

I know this is not ASTRO-related but did want to see what y'all thought about this without creating a brand new post. Of course will need to see full data (likely to be presented at ASCO?), but will it potentially change SOC?

Merck’s KEYTRUDA® (pembrolizumab) Met Primary Endpoint of Event-Free Survival (EFS) as Perioperative Treatment Regimen in Patients With Resected, Locally Advanced Head and Neck Squamous Cell Carcinoma - Merck.com

KEYNOTE-689 is the first Phase 3 trial to demonstrate statistically significant and clinically meaningful improvement in EFS in the intent-to-treat population in the neoadjuvant and adjuvant setting for an anti-PD-1 therapy in earlier stages of head and neck squamous cell carcinoma Merck (NYSE...

www.merck.com

Most HN SoC driven by OS not EFS. Now if there is evidence the surgery changed (total gloss --> partial gloss) that may be enough.

I kind of like the mPR stuff. The higher it is, the less reason to do surgery, just consolidate with RT.

 

[gmsquid]

Criminal study. A year of anti cancer therapy has an efs benefit over no cancer therapy. No duh. Cooked for the drug company to win. Now ten percent more people are going to be running around with grade 3+ IO toxicities without a clear quality of life or survival benefit.

The ethical study should have had an OS endpoint. With good crossover. Or if efs is an endpoint the study should have been two cycles neoadjuvant with no adjuvant. Outcomes should have also included quality of life. It will be interesting to see the major path response rate. If it’s a high path response rate but everybody still gets the same surgery it’s meaningless.

 

[geniusindisguise]

Most HN SoC driven by OS not EFS. Now if there is evidence the surgery changed (total gloss --> partial gloss) that may be enough.

I kind of like the mPR stuff. The higher it is, the less reason to do surgery, just consolidate with RT.

Criminal study. A year of anti cancer therapy has an efs benefit over no cancer therapy. No duh. Cooked for the drug company to win. Now ten percent more people are going to be running around with grade 3+ IO toxicities without a clear quality of life or survival benefit.

The ethical study should have had an OS endpoint. With good crossover. Or if efs is an endpoint the study should have been two cycles neoadjuvant with no adjuvant. Outcomes should have also included quality of life. It will be interesting to see the major path response rate. If it’s a high path response rate but everybody still gets the same surgery it’s meaningless.

EFS and path response are pretty standard primary endpoints for neoadjuvant ICI studies. No landmark phase III neoadjuvant ICI trials to date has had OS as a primary endpoint including in breast, lung, bladder, melanoma, etc. Given all the numerous failed studies to date in this space, I don't know if I would go so far as to call this a criminal study at least until we see the full data.

 

[gmsquid]

Keynote 412 had a p of .046, but was powered at an alpha .024, similar study design to 689 in hnscc but with crt instead of surgery . So that was considered a failed study because it didn’t meet what it was powered to. However, those curves look pretty good for efs.

Every single adjuvant stage iii study in melanoma has an efs benefit. Only one had a survival benefit versus observation, the oldest: the ipi study. Pembro vs obs, efs no os benefit even on the latest publication. Is ipi better than pembro or was crossover poor?

You are giving an anti cancer therapy that works but just cause you give it earlier doesn’t mean it has an actual impact on real outcomes. Say you cure 60% of people with soc. And IO ultimately controls 25% of people you don’t cure. You will control 70% ultimately. And if efs with soc plus adjuvant IO is 70%, then you really didn’t benefit anybody, you just gave the entire population IO without helping anybody but the drug company.

 

[gmsquid]

EFS and path response are pretty standard primary endpoints for neoadjuvant ICI studies. No landmark phase III neoadjuvant ICI trials to date has had OS as a primary endpoint including in breast, lung, bladder, melanoma, etc. Given all the numerous failed studies to date in this space, I don't know if I would go so far as to call this a criminal study at least until we see the full data.

and this isn’t a neoadjuvant study, this is a neoadjuvant and adjuvant study

 

[gmsquid]

Unlike anti pdl1, we already knew pembro worked in this disease. Moving it up isn’t necessarily better. IO isn’t chemo. It cures people with gross disease. The question is whether you cure more people with IO if they have micrometastatic disease and all the melanoma trials (except the adjuvant ipi one, which nobody uses, I wonder why) have thus far failed to report a survival benefit to adjuvant meaning people are being successfully salvaged (keynote 054 just reported 7 year results!) Despite this everybody is getting adjuvant IO nowadays at the cost of severe toxicities

 

[gmsquid]

The burden for adjuvant studies should be survival or quality of life improvement. A path response is meaningless in a phase iii study if it didn’t show a qol benefit or actual reduction in surgical morbidity.

 

[geniusindisguise]

Keynote 412 had a p of .046, but was powered at an alpha .024, similar study design to 689 in hnscc but with crt instead of surgery . So that was considered a failed study because it didn’t meet what it was powered to. However, those curves look pretty good for efs.

We can't compare apples to oranges. KN 412 is a concurrent IO-RT strategy. Concurrent IO+RT has proven to be a failing strategy time and time again across multiple tumors and trials for example in KN-867, PACIFIC-2, NRG-005, etc.

Every single adjuvant stage iii study in melanoma has an efs benefit. Only one had a survival benefit versus observation, the oldest: the ipi study. Pembro vs obs, efs no os benefit even on the latest publication. Is ipi better than pembro or was crossover poor?

We can't call KN 689 an adjuvant study and compare it to other adjuvant ICI studies. This study as well as perioperative ICI trials is conceptually a neoadjuvant study. SWOG S1801 was a proof of concept that neoadjuvant-adjuvant was superior to adjuvant ICI indicating that the neoadjuvant component makes a difference. Also, there is increasing evidence that EFS as well as pCR/MPR is a good surrogate outcome of OS for neoadjuvant ICI.

and this isn’t a neoadjuvant study, this is a neoadjuvant and adjuvant study

True, but this is nitpicking. Again, while it is technically a perioperative therapy, conceptually speaking it's a neoadjuvant therapy. I'm not sure if the adjuvant portion is needed here. Nobody compares KN 671 and AEGEAN trial to IMpower 010 and PEARLS, but to CM 816. Also, one of the emerging questions in NSCLC is whether the adjuvant component of perioperative therapy is even needed and if so for how long. To that effect, I agree with you that the 1 year of adjuvant component in KN 689 was probably unnecessary, but it's a pharma trial.

Unlike anti pdl1, we already knew pembro worked in this disease. Moving it up isn’t necessarily better. IO isn’t chemo. It cures people with gross disease. The question is whether you cure more people with IO if they have micrometastatic disease and all the melanoma trials (except the adjuvant ipi one, which nobody uses, I wonder why) have thus far failed to report a survival benefit to adjuvant meaning people are being successfully salvaged (keynote 054 just reported 7 year results!) Despite this everybody is getting adjuvant IO nowadays at the cost of severe toxicities

Again, this is the concept of adjuvant therapy, not neoadjuvant ICI, for which the goal is in-situ vaccination: the aim is to generate an optimal immune response prior to removal of the primary tumor and lymphatics.

The burden for adjuvant studies should be survival or quality of life improvement. A path response is meaningless in a phase iii study if it didn’t show a qol benefit or actual reduction in surgical morbidity.

I agree that OS and HRQOL improvement should ultimately be the goals, although de-escalation of surgery or adjuvant (C)RT are worthy goals.

 

[gmsquid]

I understand those points, but I think the reason I’m irate about the study is the adjuvant IO for 12 months. Two reasons it gives cancer therapy to a large group of patients without that benefit for a year. This is criminal design. The question is really whether neoadjuvant does improve os. I mean neoadjuvant vs adjuvant chemo in breast cancer have same ultimate os. And cr is a surrogate for os but it doesn’t mean that they wouldn’t have been long term survivors anyways. I do agree deescalation of rt or surgery are worthy goals. Design the study that way, not a way to expand indications for over a year of adjuvant therapy and not answer the question of whether sequencing actually matters. I work in translational immuno in HNC and understand the rationale for it, though I’m not sure I believe it either based on our work.

Of those studies only keynote 687 used both anti PD1 and a standard of care local therapy and remains unpublished. Is the event rate incredibly low in a properly staged population? It’s uncertain. Hn005 deescalated rt in the IO arm. And also one question about pacific 1 being positive that I still have is whether crossover was appropriate or not. It’s soc now and will always be but that’s also because we didn’t have these ideas back then. Could make the same argument as the adjuvant ipi study showing os benefit in melanoma for pacific 1

 

[gmsquid]

Also when you randomize before surgery unless you restate immediately after surgery and crt before IO you will have early recurrences untreated in the soc arm that are getting treated in the IO arm

 

[gmsquid]

In adjuvant studies patients are restated prior to enrollment, which I believe , though I may be wrong, was also true of pacific

 

[geniusindisguise]

I understand those points, but I think the reason I’m irate about the study is the adjuvant IO for 12 months. Two reasons it gives cancer therapy to a large group of patients without that benefit for a year. This is criminal design. The question is really whether neoadjuvant does improve os. I mean neoadjuvant vs adjuvant chemo in breast cancer have same ultimate os. And cr is a surrogate for os but it doesn’t mean that they wouldn’t have been long term survivors anyways. I do agree deescalation of rt or surgery are worthy goals. Design the study that way, not a way to expand indications for over a year of adjuvant therapy and not answer the question of whether sequencing actually matters. I work in translational immuno in HNC and understand the rationale for it, though I’m not sure I believe it either based on our work.

You and I agree that the adjuvant IO x 12 months is unwarranted and the trial design should have dropped this adjuvant component or at least significantly shortened it. If the KN689 data holds up to scrutiny, at the end of the day, we'll be left wondering whether or not that adjuvant component is necessary without a good way to definitively prove otherwise. I mean it's no surprise that all things being equal CM816 is generally the preferred regimen over KN671 for resectable NSCLC.

Also when you randomize before surgery unless you restate immediately after surgery and crt before IO you will have early recurrences untreated in the soc arm that are getting treated in the IO arm

I agree. Informative censoring and how they dealt with early recurrences are going to be key when we see the full data.

 

[geniusindisguise]

Of those studies only keynote 687 used both anti PD1 and a standard of care local therapy and remains unpublished. Is the event rate incredibly low in a properly staged population? It’s uncertain. Hn005 deescalated rt in the IO arm. And also one question about pacific 1 being positive that I still have is whether crossover was appropriate or not. It’s soc now and will always be but that’s also because we didn’t have these ideas back then. Could make the same argument as the adjuvant ipi study showing os benefit in melanoma for pacific 1

The issue of crossover and whether sufficient patients received post-progression IO plagues plagues not only PACIFIC-1 and ADRIATIC, it also affects KN-412, where only 50% of patients who received post-progression Tx received aPD-1. The difference is that the latter was still negative as were JAVELIN-100 and IMvoke010.

 

[gmsquid]

The issue of crossover and whether sufficient patients received post-progression IO plagues plagues not only PACIFIC-1 and ADRIATIC, it also affects KN-412, where only 50% of patients who received post-progression Tx received aPD-1. The difference is that the latter was still negative as were JAVELIN-100 and IMvoke010.

but it was powered at an exceptionally low p value (.024), showing at least a trend towards benefit (study p of .049) and it’s an efs endpoint so crossover doesn’t help the primary endpoint, could be they underpowered the study expecting a more aggressive patient population

 

[geniusindisguise]

but it was powered at an exceptionally low p value (.024), showing at least a trend towards benefit (study p of .049) and it’s an efs endpoint so crossover doesn’t help the primary endpoint, could be they underpowered the study expecting a more aggressive patient population

KN-412 reported the final analysis of its OS which was stone cold negative - couldn’t fit a needle through the curves. If we had an a priori p value cutoff of 0.05, would we have touted this trial a success with an EFS p value of 0.49 and a negative OS?

Criticizing KN689 for having EFS and not OS as an endpoint and splitting hairs about p values for EFS for KN412 seems to me like applying double standards.

 

[gmsquid]

KN-412 reported the final analysis of its OS which was stone cold negative - couldn’t fit a needle through the curves. If we had an a priori p value cutoff of 0.05, would we have touted this trial a success with an EFS p value of 0.49 and a negative OS?

Criticizing KN689 for having EFS and not OS as an endpoint and splitting hairs about p values for EFS for KN412 seems to me like applying double standards.

.049 not .49. My point is that it practically showed an efs benefit but had a stone cold negative overall survival outcome. which is the same way I view this. Salvage IO is probably the right paradigm. Why do we care about efs if there’s no qol or os benefit

 

[gmsquid]

And the trial was nearly 25% HPV positivie. The HPV negative group nearly had an efs benefit and 689 is primarily going to be HPV negative. Is it really the sequencing or just the fact that there’s an anti cancer therapy on board for an extended amount of time

 

[evilbooyaa]

Unlike anti pdl1, we already knew pembro worked in this disease. Moving it up isn’t necessarily better. IO isn’t chemo. It cures people with gross disease. The question is whether you cure more people with IO if they have micrometastatic disease and all the melanoma trials (except the adjuvant ipi one, which nobody uses, I wonder why) have thus far failed to report a survival benefit to adjuvant meaning people are being successfully salvaged (keynote 054 just reported 7 year results!) Despite this everybody is getting adjuvant IO nowadays at the cost of severe toxicities

?? Are you suggesting tha tpeople with stage IV disease should start being told that they will be cured by their ICI? I'm aware we all have an anecdote bout the exceptional long-term responders, but are we going to start using that 'C' word for pts with stage IV disease just because we sprinkle in IO to whatever they would normally get?

 

[gmsquid]

?? Are you suggesting tha tpeople with stage IV disease should start being told that they will be cured by their ICI? I'm aware we all have an anecdote bout the exceptional long-term responders, but are we going to start using that 'C' word for pts with stage IV disease just because we sprinkle in IO to whatever they would normally get?

I think 5 year os/dfs numbers in the 10-20% range with IO for metastatic disease is about as close as you can get to cure as we have seen for most solid tumors in the metastatic setting. And a relative 10-20% reduction in events is what you get with ici in the adjuvant setting

 

[Palex80]

Melanoma is the exception here. Almost half are likely cure with combo IO (if they can tolerate it).

 

[medgator]

Melanoma is the exception here. Almost half are likely cure with combo IO (if they can tolerate it).

Biggest/most obvious success story for sure. Still remember the IL-2 days pre IO. At least hearing the horror stories about inpatient admission, poor response rates, toxicity etc