ASCO 2017: interesting stuff

[Palex80]

True, I saw those two abstracts too.

BUT perhaps it's a different biology maybe? A lot of the abscopal data is based on NSCLC, little on HNSCC.
Do you have data on the PD-L1-status?

---

Members don't see this ad.

 

[Palex80]

More interesting stuff:

Meeting Library | Meeting Library

Isn't this the first randomized trial on ANYTHING as neoadjuvant treatment for borerline-resectable pancreatic cancer showing an OS benefit?

We're back in the game!

 

[Neuronix]

In the USA we had decided years ago that no neoadjuvant therapy is a bad idea based on poor margin negative rates for borderline resectables with upfront surgery. The other rationale is that some tumors will progress rapidly despite chemoradiotherapy and those patients are doomed, so why put them through a major surgery that will only increase mortality (a Whipple).

Therefore, NCCN guidelines have stressed for years that some neoadjuvant therapy is strongly recommended without clear data for superiority among the various regimens used. This is being tested in the Alliance trial currently with neoadjuvant chemotherapy +/- SBRT.

We couldn't make a study with a randomization for upfront surgery vs. neoadjuvant therapy for borderline resectables because in the USA at least I'm not sure anyone would find equipoise in the two arms. Numerous retrospective series have looked at that question. In a way I'm glad that the European study was done because at least it proves what we had thought.

 

[BobbyHeenan]

In the USA we had decided years ago that no neoadjuvant therapy is a bad idea based on poor margin negative rates for borderline resectables with upfront surgery. The other rationale is that some tumors will progress rapidly despite chemoradiotherapy and those patients are doomed, so why put them through a major surgery that will only increase mortality (a Whipple).

Therefore, NCCN guidelines have stressed for years that some neoadjuvant therapy is strongly recommended without clear data for superiority among the various regimens used. This is being tested in the Alliance trial currently with neoadjuvant chemotherapy +/- SBRT.

We couldn't make a study with a randomization for upfront surgery vs. neoadjuvant therapy for borderline resectables because in the USA at least I'm not sure anyone would find equipoise in the two arms. Numerous retrospective series have looked at that question. In a way I'm glad that the European study was done because at least it proves what we had thought.

Good points.

I think at my tumor board the new question will be what is the best neoadjuvant regimen. Med onc will push for Folfirinox; in my experience here response rates with it are greater than 5FU/XRT, but the patients are pretty beat up it seems walking into a whipple. I tried to get my cancer center on the Alliance trial as I think it's a good trial...but my surg oncs thought too many defined as "borderline" on the trial were outside of what they were comfortable operating on (I'm in a community setting but with excellent well trained surg oncs I feel).

It's nice to have randomized data now for XRT in the neoadjuvant setting, though 2.4 Gy with gem is something i've never done but I'd be willing to do now.

 

[seper]

I also find FOLFIRINOX a tough regimen. 2.4 Gy X 15 is fast and cheap, glad the Dutch chose it. Do you know if they treated classic nodal fields? I cannot make my mind up whether nodes are worth treating preop.

 

[Neuronix]

FOLFIRINOX vs. Gem+Abraxane is an open question. I find that many institutions pick FOLFIRINOX for younger, healthier patients and Gem+Abraxane for patients over 70 and/or sicker. Some places just do Gem+Abraxane for everyone without clear superiority of FOLFIRINOX because Gem+Abraxane is better tolerated. We had observed and published better results for FOLFIRINOX in the neoadjuvant setting where I trained, but the quality of data was low (few patients, retrospective).

As for radiation regimen, it runs the gamut from SBRT without concurrent chemo to 50.4 Gy with concurrent Xeloda or 5-FU RTOG 0848 style (with slightly modified volumes) and a bunch of hypofractionated regimens inbetween.

I'll re-emphase: I like and use SBRT, but I have a lot of experience with pancreas SBRT. If you're going to do pancreas SBRT, you should know exactly what you're doing or not do it at all. It's a risky technique and not for the faint of heart. There's no clear data that it's better than longer regimens--it's just quicker and I can safely get 50 Gy in 5 fractions to most of the volume in most of the tumors (with unclear benefit). It's not at all a problem to give conventional fractionation because you don't have a training specifically in pancreas SBRT.

 

[BobbyHeenan]

FOLFIRINOX vs. Gem+Abraxane is an open question. I find that many institutions pick FOLFIRINOX for younger, healthier patients and Gem+Abraxane for patients over 70 and/or sicker. Some places just do Gem+Abraxane for everyone without clear superiority of FOLFIRINOX because Gem+Abraxane is better tolerated. We had observed and published better results for FOLFIRINOX in the neoadjuvant setting where I trained, but the quality of data was low (few patients, retrospective).

As for radiation regimen, it runs the gamut from SBRT without concurrent chemo to 50.4 Gy with concurrent Xeloda or 5-FU RTOG 0848 style (with slightly modified volumes) and a bunch of hypofractionated regimens inbetween.

I'll re-emphase: I like and use SBRT, but I have a lot of experience with pancreas SBRT. If you're going to do pancreas SBRT, you should know exactly what you're doing or not do it at all. It's a risky technique and not for the faint of heart. There's no clear data that it's better than longer regimens--it's just quicker and I can safely get 50 Gy in 5 fractions to most of the volume in most of the tumors (with unclear benefit). It's not at all a problem to give conventional fractionation because you don't have a training specifically in pancreas SBRT.

Good input, again.

Off Alliance study are you giving pre-op SBRT after their chemo?

Off study here we typically try to squeeze in XRT (typically 5FU/standard frac) before surgery...but as mentioned often with FOLFIRINOX they're so beat up we've elected to go to surgery then use selective post-op XRT (mainly for nodes + or + or very close margin). I've seen a few patients now after their FOLFIRINOX and sometimes after Gem/abraxane where I felt like if I gave them any further hit to KPS I was going to possibly flip them outside of the chances for a whipple - so I just sent to surgery.

 

[Neuronix]

Off Alliance study are you giving pre-op SBRT after their chemo?

Yes. I would say that this is common within academics and uncommon within private practice at this point.

Off study here we typically try to squeeze in XRT (typically 5FU/standard frac) before surgery...but as mentioned often with FOLFIRINOX they're so beat up we've elected to go to surgery then use selective post-op XRT (mainly for nodes + or + or very close margin). I've seen a few patients now after their FOLFIRINOX and sometimes after Gem/abraxane where I felt like if I gave them any further hit to KPS I was going to possibly flip them outside of the chances for a whipple - so I just sent to surgery.

This is a reasonable approach. I agree that you don't want to make someone medically inoperable with your neoadjuvant treatment. Fractionated plus chemo seems to have some risk of this (though I don't have data to say how much) whereas this almost never happens with the short course of SBRT.

About 60 - 75% will have nodes positive or positive or close margin post-op, and there are database (SEER/NCDB) analyses to back up treatment for those patients. So if you don't get them pre-op they'll probably come back to you post-op if their condition is suitable for more treatment.

 

[evilbooyaa]

ASCO 2018: Tailor RX data confirms that most patients with an intermediate Oncotype DX can skip chemo, notable exception is women <50 with a score >16

Landmark TAILORx Results, Published Today in The New England Journal of Medicine, Demonstrate the Oncotype DX Breast Recurrence Score® Test Definitively Identifies the 70% of Women with Early-stage Breast Cancer Who Receive No Benefit from Chemotherapy, and the 30% of Women for Whom Chemotherapy Benefit Can be Life-saving

Full paper: NEJM - Error

Hopefully the vast majority of these node-negative patients (all that was studied in this trial) can avoid chemotherapy now.

Interesting discussion on the neoadjuvant pre-op RT for pancreas. I think 15 x 2.4Gy with concurrent gem is also scary in terms of GI toxicity, but that's just gestalt not backed by data (with data showing equivalent toxicity outcomes to surgery alone), but I guess if you do a SBRT-like technique it should be well tolerated.

Per their protocol (Preoperative radiochemotherapy versus immediate surgery for resectable and borderline resectable pancreatic cancer (PREOPANC trial): study protocol for a multicentre randomized controlled trial) they didn't electively cover lymph nodes. All GTVs (including clinically positive nodes) + 5mm to CTV, +10mm to PTV. They did do 4DCT and IV Contrast at sim.

Wonder if I could get the GI folks here to get on board with this.

 

[evilbooyaa]

This muddies the picture: Meeting Library | Meeting Library

Word on the street is the arms in the pembro study had an imbalance of pd-l1

I think there's a biologic difference between HNSCC and NSCLC. It seems like definitive treatment in mNSCLC helps, but even in well-selected patients, the same does not pan out for mHNSCC.

If you have a link regarding imbalance of pd-l1 (which most studies don't necessarily show affects outcomes when looking at IT across multiple disease sites) that'd be interesting to know, otherwise we can wait on the paper.

 

[Palex80]

I fully agree with you that borderline-resectable patients should get neoadjuvant treatment. This is standard of care in Europe too, at least in most countries. Our med oncs usually push for FOLFIRINOX.

HOWEVER: The Dutch trial was not done only on borderline-resectable. It was done on both resectable and borderline-resectable tumors. If you want to have a look at the study protocol including resectability criteria, have a look here:
Preoperative radiochemotherapy versus immediate surgery for resectable and borderline resectable pancreatic cancer (PREOPANC trial): study protocol for a multicentre randomized controlled trial

 

[evilbooyaa]

I'm not at the meeting and have no access to all the details that Palex is mentioning. Anyone have a link to something more than just the abstract?

 

[evilbooyaa]

Meeting Library | Meeting Library

PD-L1 expression doesn't matter for activity of immunotherapy. All NSCLC gets Pembro now instead of chemotherapy in the front-line setting.

 

[Krukenberg]

I think there's a biologic difference between HNSCC and NSCLC. It seems like definitive treatment in mNSCLC helps, but even in well-selected patients, the same does not pan out for mHNSCC.

If you have a link regarding imbalance of pd-l1 (which most studies don't necessarily show affects outcomes when looking at IT across multiple disease sites) that'd be interesting to know, otherwise we can wait on the paper.

I heard it from someone at the meeting who heard it from another rad onc working in the immunoRT space. Total hearsay!

 

[Krukenberg]

I heard it from someone at the meeting who heard it from another rad onc working in the immunoRT space. Total hearsay!

Correction. Apparently the imbalance was actually presented at the oral. With RCTs though you can effectively control imbalance with regression so I’d assume their hazard ratio of 0.55 would be adjusted for pd-l1.

 

[Lamount]

I fully agree with you that borderline-resectable patients should get neoadjuvant treatment. This is standard of care in Europe too, at least in most countries. Our med oncs usually push for FOLFIRINOX.

HOWEVER: The Dutch trial was not done only on borderline-resectable. It was done on both resectable and borderline-resectable tumors. If you want to have a look at the study protocol including resectability criteria, have a look here:
Preoperative radiochemotherapy versus immediate surgery for resectable and borderline resectable pancreatic cancer (PREOPANC trial): study protocol for a multicentre randomized controlled trial

Interesting phase II out of MGH supporting total neoadjuvant therapy with FOLFIRINOX followed by SBRT for resolved vascular involvement or long course CRT for persistent vascular involvement. Among resected patients (2/3), median PFS was 48 months and 2 yr OS was 55%.

Neoadjuvant FOLFIRINOX and Chemoradiotherapy for Resectable Pancreatic Adenocarcinoma

 

[seper]

Interesting study, but somewhat misleading number PFS of 48 months that should have been left out. In fact, 2y PFS was 55%. Median follow up 18 months. This not a cure for pancreatic cancer quite yet.

Interesting phase II out of MGH supporting total neoadjuvant therapy with FOLFIRINOX followed by SBRT for resolved vascular involvement or long course CRT for persistent vascular involvement. Among resected patients (2/3), median PFS was 48 months and 2 yr OS was 55%.

Neoadjuvant FOLFIRINOX and Chemoradiotherapy for Resectable Pancreatic Adenocarcinoma

 

[Lamount]

Interesting study, but somewhat misleading number PFS of 48 months that should have been left out. In fact, 2y PFS was 55%. Median follow up 18 months. This not a cure for pancreatic cancer quite yet.

Very good point. Nonetheless, I am happy to see studies combining modern radiotherapy and modern chemotherapy

 

[Palex80]

Statistics, statistics, statistics...

Fifteen-year results of the randomised EORTC trial 22922/10925 investigating internal mammary and medial supraclavicular (IM-MS) lymph node irradiation in stage I-III breast cancer. | 2018 ASCO Annual Meeting Abstracts

15-year follow up of the European trial on irradiation of the internal mammary chain.

The 10-year-paper was published in NEJM along the MA20-paper a couple of years ago.
NEJM - Error

At 15 years follow, some of the effects noted at 10 years seem to fade off or at least the benefit of IM-RT seems not to be that significant any more, although the absolute difference in terms of absolute benefit of IM-RT actually looks bigger.
Possibly lots of patients lost to follow up?

"No difference was observed in the incidence of second malignancies, contralateral breast cancer or cardiovascular deaths"
At 15 years cardiac toxicity should be slowly kicking in and considering the "medieval" techniques used in the trial. I am looking forward into more data on that...

 

[BobbyHeenan]

Very good point. Nonetheless, I am happy to see studies combining modern radiotherapy and modern chemotherapy

It was a little curious to me to see protons were used for the "SBRT" 25 Gy in 5 fraction group. I would think you could extrapolate and use photons for this.

As I recall, most of the centers doing SBRT and/or high dose hyopfractionated treatment (Chris Crane) are using photons for pancreas SBRT, even at facilities that have protons. However, many don't give the 5FU concurrent with the SBRT, so maybe the rationale was to give slightly reduced dose SBRT at 25/5 but with 5FU?

The cynic in me says that on trial they were able to get protons paid for, but if off trial they get insurance denials.

 

[medgator]

25/5 is kinda sbrt lite anyways.... the people who really do a lot of it to go higher that afaik

 

[seper]

Can't wait for CMS or insurance companies to crack down on this.

25/5 is kinda sbrt lite anyways.... the people who really do a lot of it to go high that afaik

 

[medgator]

Can't wait for CMS or insurance companies to crack down on this.

Agreed. Go big/appropriate or go home. Personally, I would refer out for pancreas SBRT to a place where I know they do quite a bit of it.

 

[Lamount]

25/5 is kinda sbrt lite anyways.... the people who really do a lot of it to go higher that afaik

I know there are series using higher doses, but are there data to suggest that dose escalation beyond 25/5 in the neoadjivant setting improves outcomes?... especially after intense chemo like FOLFIRINOX

 

[medgator]

I know there are series using higher doses, but are there data to suggest that dose escalation beyond 25/5 in the neoadjivant setting improves outcomes?... especially after intense chemo like FOLFIRINOX

I'm no expert in pancreatic sbrt/sabr but the general ideal is to dose escalate to "ablative" doses. 25/5 hardly fits the bill considering 20/5 is a short course palliative regimen for a poor PS patient

 

[nkmiami]

I'm no expert in pancreatic sbrt/sabr but the general ideal is to dose escalate to "ablative" doses. 25/5 hardly fits the bill considering 20/5 is a short course palliative regimen for a poor PS patient

Considering in the 1970s most the pelvis including a lot of small bowel would get 5 x5 appa with cobalt in sweden for rectal cancer, 5 x5 is way too low. They must of been overly concerned with adverse effects from protons when they wrote the study. If you are giving 5 x5, you may as well treat all the regional nodes?

 

[Lamount]

Agreed. Go big/appropriate or go home. Personally, I would refer out for pancreas SBRT to a place where I know they do quite a bit of it.

Is “ablation” appropriate as a neoadjuvant treatment of pancreatic cancer? I think this is an interesting question, but as far as I can tell (and correct me if I am wrong), it has yet to be answered. We have data that FOLFIRINOX + 25/5 is as effective as anything else, so what is the argument in dose escalating off study?

 

[Lamount]

I'm no expert in pancreatic sbrt/sabr but the general ideal is to dose escalate to "ablative" doses. 25/5 hardly fits the bill considering 20/5 is a short course palliative regimen for a poor PS patient

Sorry meant to reply to this post, not the other

 

[medgator]

Is “ablation” appropriate as a neoadjuvant treatment of pancreatic cancer? I think this is an interesting question, but as far as I can tell (and correct me if I am wrong), it has yet to be answered. We have data that FOLFIRINOX + 25/5 is as effective as anything else, so what is the argument in dose escalating off study?

Which is fine. I think we were discussing from an insurance company standpoint about whether that truly qualifies as sabr/sbrt. If it does, the Swedish were doing preop sbrt for rectal ca back in the 80s with simple techniques

 

[deleted18755]

Which is fine. I think we were discussing from an insurance company standpoint about whether that truly qualifies as sabr/sbrt. If it does, the Swedish were doing preop sbrt for rectal ca back in the 80s with simple techniques

What do you all (and/or your payers) consider "SBRT" with regards to dose/ratiocination? I always thought of it more with regards to technique and trying to achieve ablation but I don't think I've ever seen a lower threshold of what is generally accepted; however, I would hardly consider 5 Gy x 5 ablative or a reasonable SBRT dose since as mentioned above 4 Gy x 5 is a palliative dose.

AP/PA 4 Gy x 5 with portal imaging is in general very well tolerated and we all would consider palliative so how can adding a single Gy more with each fraction require multiple non-coplanar fields, IGRT, physics support and physician presence for the procedure?

 

[BobbyHeenan]

Is “ablation” appropriate as a neoadjuvant treatment of pancreatic cancer? I think this is an interesting question, but as far as I can tell (and correct me if I am wrong), it has yet to be answered. We have data that FOLFIRINOX + 25/5 is as effective as anything else, so what is the argument in dose escalating off study?

I'm good with that dose - especially as you mention they're headed to surgery, so why go higher.

My comment was more that I felt it curious to use protons for this rather than photons - partially because of the lower dose and partially because in my review of the literature and some of what I consider to be the experts of GI SBRT, they don't use protons (even when available) for pancreas SBRT. Maybe with better on board imaging/gating capabilities some of that may be different now, I'm not sure.

 

[BobbyHeenan]

Which is fine. I think we were discussing from an insurance company standpoint about whether that truly qualifies as sabr/sbrt. If it does, the Swedish were doing preop sbrt for rectal ca back in the 80s with simple techniques

For what it's worth, I've been doing Italian randomized trial style 30 Gy in 5 fraction partial breast in highly selected patients (>60, small tumors, clips, small seroma, etc). I go to the linac every day for treatment like an SBRT case and use daily image guidance (either CBCT or gate-triggered orthogonal KVs) but I don't bill it as SBRT. Not sure if this is the right thing to do, but it just doesn't "feel" SBRT-like to me given the margins and dose.

However, I have done 30/5 for larger brain mets and have definitely billed that as SBRT.

 

[ramsesthenice]

It was a little curious to me to see protons were used for the "SBRT" 25 Gy in 5 fraction group. I would think you could extrapolate and use photons for this.

As I recall, most of the centers doing SBRT and/or high dose hyopfractionated treatment (Chris Crane) are using photons for pancreas SBRT, even at facilities that have protons. However, many don't give the 5FU concurrent with the SBRT, so maybe the rationale was to give slightly reduced dose SBRT at 25/5 but with 5FU?

The cynic in me says that on trial they were able to get protons paid for, but if off trial they get insurance denials.

You are correct. Current pancreas "SBRT" doses are not really SBRT. Chris Crane and a few other places have some very intriguing data using dose escalated SIB. I can't remember his exact numbers off hand but it is in the range of 75 with a BED over 100. He gets away with it because he uses MRI guided photons and can adaptively replan on the treatment table based on normal-tissue positioning (ie bowel avoidance). He presented his data to us recently and showed a few unresectable patients with radiographic CRs and durable control. The MCW group presented their data doing the same thing at ASTRO last year. Their early results were a little too good to be taken at face value but still very encouraging none the less.

 

[Palex80]

"Stereotactic Radiotherapy" is not a matter of dose.
Neither a matter of dose per fraction, nor a matter of total dose.

It's a matter of the technique used.
Technique involves positioning, immobilization, accountability for possible movement, planning, verification of dose and verification of correct delivery. All that is part of stereotactic radiotherapy. And SBRT is stereotactic radiotherapy, just like FSRT and SRS are in the brain.

45/1.8 is not an ablative dose (neither an ablative dose per fraction nor an ablative total dose), but >90% of the people do it in a stereotactic way (and bill for it as well) when they treat a vestibular schwannoma.

Coming back to the issue of pancreatic cancer:
People are mixing up SBRT delivered a sole local treatment in pancreatic cancer with high doses with the aim to induce durable local control and SBRT delivered in the preoperative setting, which has the goal to possibly downstage/downsize surgery and lower the risk of local recurrence after resection. These are two different scenarios.

 

[Lamount]

"Stereotactic Radiotherapy" is not a matter of dose.
Neither a matter of dose per fraction, nor a matter of total dose.

It's a matter of the technique used.
Technique involves positioning, immobilization, accountability for possible movement, planning, verification of dose and verification of correct delivery. All that is part of stereotactic radiotherapy. And SBRT is stereotactic radiotherapy, just like FSRT and SRS are in the brain.

45/1.8 is not an ablative dose (neither an ablative dose per fraction nor an ablative total dose), but >90% of the people do it in a stereotactic way (and bill for it as well) when they treat a vestibular schwannoma.

Coming back to the issue of pancreatic cancer:
People are mixing up SBRT delivered a sole local treatment in pancreatic cancer with high doses with the aim to induce durable local control and SBRT delivered in the preoperative setting, which has the goal to possibly downstage/downsize surgery and lower the risk of local recurrence after resection. These are two different scenarios.

I always thought of modern "SBRT" as a misnomer as we don't actually have a stereotactic frame in most cases. Rather than aligning to a rigid frame with its own coordinate system, we align to imaging (i.e. CBCT). The one exception is with some halo immobilization systems for SRS. Nonetheless, I agree with you that the assumptions made when using the term "SBRT" are the same as would be in a stereotactic system i.e. accurate alignment, immobilization, and appropriate management for organ motion.

 

[seper]

Therefore the situation is ripe for CMS code change. Practitioners have picked up that "body SBRT" is literally raining wRVU and technical revenue, so things like 5X5 SBRT to a bone met is really common in general practice (academic, private).

I always thought of modern "SBRT" as a misnomer as we don't actually have a stereotactic frame in most cases. Rather than aligning to a rigid frame with its own coordinate system, we align to imaging (i.e. CBCT). The one exception is with some halo immobilization systems for SRS. Nonetheless, I agree with you that the assumptions made when using the term "SBRT" are the same as would be in a stereotactic system i.e. accurate alignment, immobilization, and appropriate management for organ motion.

 

[medgator]

Therefore the situation is ripe for CMS code change. Practitioners have picked up that "body SBRT" is literally raining wRVU and technical revenue, so things like 5X5 SBRT to a bone met is really common in general practice (academic, private).

Correct. Without the corresponding dose escalation, the techniques and immobilization really become superfluous and unnecessary.