RTOG 0617 Update

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

Gfunk6

And to think . . . I hesitated
Moderator Emeritus
Lifetime Donor
15+ Year Member
Joined
Apr 16, 2004
Messages
4,641
Reaction score
5,022
Just heard that the high-dose arms of RTOG 0617 are closing immediately. For those of you that don't know, 0617 is a 2 x 2 study in Stage III NSCLC:

1. 60 Gy + carbo-taxol
2. 74 Gy + carbo-taxol
3. 60 Gy + carbo-taxol + C225
4. 74 Gy + carbo-taxol + C225

Arms 2 and 4 are closed immediately because after interim analysis it was determined that it is statistically impossible for there to be a survival benefit. Further randomization will be between arms 1 and 3 only.

Members don't see this ad.
 
wow. that's a big revelation, and something that probably goes against the core of what everyone has believed about NSCLC and the importance of dose escalation for some time.

Somehow, I doubt people are going to back to treating locally-advanced NSCLC to 60 Gy.
 
Disappointing to those (like me) who buy in to the concept that higher dose should lead to better results in NSCLC. I'm not sure that I'd close the door on the basis of the new broadcast, though. While its true that they've ruled out a "survival" advantage in the interim analysis, it's helpful to keep in mind that the specific advantage they were looking for was a median survival advantage of ~7 months (17.1 to 24 months). I've always viewed the median survival numbers as instructive, but less important in the grand scheme of things than the tail on the right i.e. long term survival.

I will be interested to see whether the local control in 0617 favors the 74 Gy arms. If so, I would remain inclined to push the dose where appropriate. As a field, we have often offered therapy based on a local control rather than a survival benefit, and have occasionally been vindicated by the finding that improving local control will, given time, result in improved survival.

If there is no local control benefit to 74 Gy over 60 Gy, I will simply have to curl up in the fetal position and seriously reconsider my worldview.
 
Members don't see this ad :)
If there is no local control benefit to 74 Gy over 60 Gy, I will simply have to curl up in the fetal position and seriously reconsider my worldview.

Yeah. I wondered why they chose such a wimpy dose (60Gy), even though it was the RTOG standard from 73-01. Then I realized that if they are going to show a difference, it'll probably be easier to show it between 74 and 60, than it will be between 74 and 66.
 
What dose are you guys using for oral boards? Almost everyone I know treats to 66/33fx or 66.6/37.
 
Are people still planning on enrolling patients on 0617 now that it's guaranteed 60 Gy?
 
Trying to answer 2 or 3 questions at once, using an out of practice standard arm, and coming up with a negative trial everytime.

It's the RTOG way.

Maybe they can spin it into a positive using p16 or something. It has worked before....
 
There is a difference between what dose you would like to treat a patient with and what dose you can treat a patient with.

66Gy in 33 fractions is the best answer IMHO. However I have seen many patients, where you just couldn't go over 60 Gy (or even 56 Gy sometimes) because of the lung DVHs. IMRT can be helpful to fulfill your constraints but it's use in NSCLC is not accepted by everyone. I personally have some concerns, when it comes to static IMRT and moving targets.
 
This will definitely hurt further dose escalation studies especially in view of the research funding changes.

Ultimately patient's lose and med oncs win (jk jk!):laugh:
 
Just to review, this was presented at ASTRO

http://www.medpagetoday.com/MeetingCoverage/ASTRO/28862

http://www.oncologyreport.com/news/...improves-lung-cancer-survival/9dd14f2b17.html

MIAMI BEACH – Less turned out to be better in a large clinical trial comparing radiation doses in patients treated with radiation and chemotherapy for stage III non–small cell lung cancer, investigators reported here.

The median overall survival rate at 1 year was 81% for patients treated with standard-dose (60 Gy) radiation, compared with 70.4% for those who received the high dose (74 Gy), according to preliminary findings from the radiation-dose arm of the ongoing phase III Radiation Therapy Oncology Group (RTOG) 0617 trial. The respective median survival rates were 21.7 months and 20.7 months (P = .02).

A planned interim analysis from the trial showed that the radiation comparison had crossed the prespecified boundary for futility, and the high-dose arm was stopped in June 2011, reported Dr. Jeffrey Bradley from Washington University in St. Louis.

"I think this changes practice: If [cancer centers] weren't using 60 Gray before, perhaps they should go back to using 60 Gray, because it does not appear that a higher dose is better," Dr. Bradley commented at the annual meeting of the American Society of Radiation Oncology (ASTRO).

Dr. Benjamin Movsas, chair of radiation oncology at the Henry Ford Health System in Detroit, the invited discussant, said that "as of 2011, level I evidence demonstrates no role for dose escalation in stage III non–small cell lung cancer."

He noted that although there were small differences between the radiation dose groups in terms of tumor histology, gross tumor volume, and other factors, they were not large enough to explain the differences in outcomes.

http://www.medscape.com/viewarticle/750892

The finding that high-dose radiation does not improve survival was unexpected. "We were surprised by the results," he told the ASTRO audience. Most clinicians have believed that higher doses of radiation cured more patients with lung cancer, Dr. Bradley said in a press statement.

But that is not what the study found.

At the initial planned interim analysis for overall survival, the investigators found that, after 90 deaths, the high-dose radiation groups had "crossed the futility boundary." The study was designed to detect a median overall survival improvement of 7 months in the high-dose radiation groups.

Dr. Bradley said that he and his coinvestigators cannot explain the results and that the data are being carefully reviewed.


I wonder what you guys taking the orals this year are going to say on your Lung/sarcoma section ;)
 
Last edited:
63 Gy on RTOG 9410 and 60 Gy in original dose escalation trial were uncorrected ... So, for peripheral lesion, in practice can make argument for ~5-10% higher... for central lesion 60-63 Gy probably okay.

Very disheartening to say the least.

But, it's not that dose escalation doesn't work. It's that 74 Gy isn't high enough. 100 Gy BED works for early stage. Of course 74 Gy won't be enough for stage III. Now, trying to treat safely to that high a dose and to a big volume ... that's the hard part!
 
Members don't see this ad :)
But, it's not that dose escalation doesn't work. It's that 74 Gy isn't high enough. 100 Gy BED works for early stage. Of course 74 Gy won't be enough for stage III. Now, trying to treat safely to that high a dose and to a big volume ... that's the hard part!

The problem is the risk of DM is way higher in IIIA/IIIB than for I-II. So that even if we were to deliver a "curative" dose safely, it might not make a significant dent in survival.
 
That's right ... I guess what I'm saying is that 70-74 just might not have been enough. It's unusual to want to use a higher BED for lower stage disease and lower BED for higher stage disease, unless we are admitting that it's essentially palliative treatment. They should have kept open so we could have assessed impact on LC.

The problem is the risk of DM is way higher in IIIA/IIIB than for I-II. So that even if we were to deliver a "curative" dose safely, it might not make a significant dent in survival.
 
That's right ... I guess what I'm saying is that 70-74 just might not have been enough. It's unusual to want to use a higher BED for lower stage disease and lower BED for higher stage disease, unless we are admitting that it's essentially palliative treatment. They should have kept open so we could have assessed impact on LC.

Since we're using radiosensitizing chemo however, it is possible for lower doses to be curative than used for SBRT. However, your point is well-taken about LC. Now that half the study is negative, I'm sure they will look at LC on subset analysis.
 
We have seen this before guys. Think about the flawed esophagus dose-escalation trial by Minsky.

The Minsky trial showed that dose escalation doesn't always work, although it's still unclear why the patients died in the dose escalated arm, before reaching the escalated dose (maybe the died of fear knowing they are in the dose escalated arm? :confused: just kidding...)
It's because of this trial, that many radiation oncologists still give only 50.4 Gy for esophagial cancer. I personally prefer 54-59.4 Gy. :laugh:



I can also think of two randomized trials in °II astrocytoma, showing no benefit coming out of dose escalation. Not to mention the dozen of futile trials performed in gliobalstoma, trying dose escalation.

Maybe dose escalation is worth, but we are just using the wrong type of radiation to achieve it? Protons? Heavy ions?
Who knows?
 
Correct me if I'm wrong, but the Minsky trial showed no difference when analyzed by dose received.


We have seen this before guys. Think about the flawed esophagus dose-escalation trial by Minsky.

The Minsky trial showed that dose escalation doesn't always work, although it's still unclear why the patients died in the dose escalated arm, before reaching the escalated dose (maybe the died of fear knowing they are in the dose escalated arm? :confused: just kidding...)
It's because of this trial, that many radiation oncologists still give only 50.4 Gy for esophageal cancer. I personally prefer 54-59.4 Gy. :laugh:



I can also think of two randomized trials in °II astrocytoma, showing no benefit coming out of dose escalation. Not to mention the dozen of futile trials performed in gliobalstoma, trying dose escalation.

Maybe dose escalation is worth, but we are just using the wrong type of radiation to achieve it? Protons? Heavy ions?
Who knows?
 
Correct me if I'm wrong, but the Minsky trial showed no difference when analyzed by dose received.
That's what I meant, when I pointed out:
"it's still unclear why the patients died in the dose escalated arm, before reaching the escalated dose"
 
My vote is for simple statistical fluke. Remember, when we base our study design on 95% confidence or power or whatever, theoretically things are allowed to be inexplicably out of whack 5% of the time.
 
My vote is for simple statistical fluke. Remember, when we base our study design on 95% confidence or power or whatever, theoretically things are allowed to be inexplicably out of whack 5% of the time.

I'm inclined to agree but I don't think we'll see this trial repeated any time soon. Also, "statistical anomaly" is a poor justification to prescribe a patient a treatment that was found to be inferior in a multi-institutional randomized trial.
 
I'm inclined to agree but I don't think we'll see this trial repeated any time soon. Also, "statistical anomaly" is a poor justification to prescribe a patient a treatment that was found to be inferior in a multi-institutional randomized trial.

No, I just think it means people will keep doing whatever they've been doing. There are plenty of other trials with arms of 63-66 Gy at least, but yes, it will be hard to justify pushing higher than 70 Gy.
 
No, I just think it means people will keep doing whatever they've been doing. There are plenty of other trials with arms of 63-66 Gy at least, but yes, it will be hard to justify pushing higher than 70 Gy.

Outside of an RTOG trial, I doubt most in practice went higher than that anyways. The whole point of 0617 was to show an obvious difference (hence why 60 Gy was chosen as a comparison to 74 Gy, even though in reality, most people prescribed 63-66 Gy at a minimum)
 
Was it explained in ASTRO what people died from in the 74 Gy arm?
Was there excessive mortality due to side-effects? Did they get more mets? Was the local control better in the 74 Gy arm?
 
Was it explained in ASTRO what people died from in the 74 Gy arm?
Was there excessive mortality due to side-effects? Did they get more mets? Was the local control better in the 74 Gy arm?

Per Dr. Bradley, they are still analyzing the findings but this is what was presented per the news article above:

There were four grade 5 adverse events in the 60-Gy dose group and eight in the 74-Gy group, and two deaths from radiation pneumonitis in each arm.

"We were surprised by the results, and set about looking to explain them," he said.

Multivariate analyses of factors that might contribute to worse survival yielded a hazard ratio of 1.48 (95% CI 1 to 2.22, P=0.38) for radiation dose.

Other contributing factors were non-squamous histology and smaller tumor volume, he said.

When asked why he thought the high-dose radiation had not led to improvements, which has been the case in some other cancers, Bradley noted that lung cancer poses a particular treatment challenge.

"The problem with lung cancer is distant metastases. That remains the dominant problem regardless of the radiation dose in the battle we're fighting here," he said.
 
Back to Fisher's hypothesis. Microscopic distant mets are present at diagnosis, therefore local therapy will not affect overall outcome.

"The problem with lung cancer is distant metastases. That remains the dominant problem regardless of the radiation dose in the battle we're fighting here," he said. :
 
Absolutey untrue however. If local therapy does not affect overall outcome you need to explain innumerable studies for various primary cancers dating back decades showing that local therapy (surgery or radiation) improves survival outcomes.

Also, microscopic mets at diagnosis may or may not be true- just because you can isolate cancer cell in blood or marrow does not mean that they will seed in distant parenchymal sites.


Back to Fisher's hypothesis. Microscopic distant mets are present at diagnosis, therefore local therapy will not affect overall outcome.
 
I think the point being made is that, for example, there was never an advantage seen for post mastectomy RT until there was effective systemic therapy. For lung cancer, similar story. For stage 1-2 NSCLC, distant mets not as big a problem, so surgery/SRS can be curative. For stage 3, cat's out of the bag, and surgery/high dose rt isn't curative. Yet, as I said above, I think 70 Gy with or without chemotherapy is not enough. Chemotherapy adds 8 weeks to median survival for NSCLC, not the significant leap as seen with cervical or head and neck cancer. Again - improvement in LC in those sites leads to an overall improval in outcome, like in early stage lung. Not so much in lung, or pancreas, or other diseases that are micrometastatic at presentation.


Absolutey untrue however. If local therapy does not affect overall outcome you need to explain innumerable studies for various primary cancers dating back decades showing that local therapy (surgery or radiation) improves survival outcomes.

Also, microscopic mets at diagnosis may or may not be true- just because you can isolate cancer cell in blood or marrow does not mean that they will seed in distant parenchymal sites.
 
  • Like
Reactions: 1 user
If local therapy does not affect overall outcome you need to explain innumerable studies for various primary cancers dating back decades showing that local therapy (surgery or radiation) improves survival outcomes.

Also, microscopic mets at diagnosis may or may not be true- just because you can isolate cancer cell in blood or marrow does not mean that they will seed in distant parenchymal sites.

The onus is on you. We aren't talking about 'various primary cancers' we're talking about inoperable lung cancer. To my knowledge there is no RCT showing that the presence of XRT (versus no XRT) provides a survival advantage, or that dose escalation beyond 40 Gy split-course provides a statistically significant survival advantage. Am I missing something? :)
 
Sounds right to me. 73-01 is still the standard (but not positive for survival as CancerDancer said), though treating based on 94-10 makes sense too.

For advanced stage lung cancer, the biology of the disease is still calling the shots. The promising results of the phase II trials for dose escalation can be explained by patient selection. This is a reason to do phase III trials in the first place.

I would also caution against using BED comparisons for extreme hypofractionation. Not only is the biology of stage I-II lung cancer possibly (slightly) different, but comparing stereotactic regimens to fractionated regimens using the alpha-beta model is not necessarily correct.
 
The onus is on you. We aren't talking about 'various primary cancers' we're talking about inoperable lung cancer. To my knowledge there is no RCT showing that the presence of XRT (versus no XRT) provides a survival advantage, or that dose escalation beyond 40 Gy split-course provides a statistically significant survival advantage. Am I missing something? :)

to be clear, you are referring to inoperable stage III correct?
 
to be clear, you are referring to inoperable stage III correct?

No, I'd be more inclusive, not less, actually. I believe my statement holds true for all NSCLC, all stages, operable or not...no randomized data showing OS benefit to XRT, or dose escalation. In fact, we get excited in some situations where XRT (vs not) just doesn't show a survival decrement (PORT).

My point isn't that we can't (or shouldn't) do stuff because a primary endpoint of a RCT isn't met, or hasn't been conducted. We all do that all the time, just way too many of us don't realize it (i.e. all NSCLC thoracic RT, or hormones w/ dose-escalated prostate RT). These practices aren't 'wrong' just b/c they aren't supported by randomized data.

It's funny how people are saying that we can't justify going above 60 Gy b/c of a failure of one RCT, but they don't think about how/why they are justified in giving that 60 Gy in the first place.

If I have a T3N0 lung SCC, and you can give me 70 Gy w/ a V30 of 15%, you're damn right I want those extra 10 Gy...
 
No, I'd be more inclusive, not less, actually. I believe my statement holds true for all NSCLC, all stages, operable or not...no randomized data showing OS benefit to XRT, or dose escalation. In fact, we get excited in some situations where XRT (vs not) just doesn't show a survival decrement (PORT).

While that may be the case, modern prospective data using SBRT or dose-escalated 3DCRT does show compelling OS rates as compared to historical cohorts of medically-inoperable patients who received standard-dose XRT in early-stage NSCLC. There are SEER studies that document a survival difference in patients treated with XRT compared to no treatment at all.

The RCTs may not exist, but that doesn't mean that the effect of dose escalation isn't there.

My point isn't that we can't (or shouldn't) do stuff because a primary endpoint of a RCT isn't met, or hasn't been conducted. We all do that all the time, just way too many of us don't realize it (i.e. all NSCLC thoracic RT, or hormones w/ dose-escalated prostate RT). These practices aren't 'wrong' just b/c they aren't supported by randomized data.

Yeah. Pretty much the situation above.

It'll be interesting to see how the RTOG studies in medically-operable early-stage patients turns out. In Japan, SBRT has been a SOC for patients who are candidates for lobectomy.
 
Last edited:
My comment was in response to the 'Fisher hypothesis' and the assertion that local therapy does not impact overall survival in general. Fisher's hypothesis was actually based upon breast cancer. True- you do need to look at each disease state (primary site, stage, histology, and probably gene expression profiles) independently, which is an even more compelling reason to dismiss the Fisher hypothesis as a general oncology tenet.


The onus is on you. We aren't talking about 'various primary cancers' we're talking about inoperable lung cancer. To my knowledge there is no RCT showing that the presence of XRT (versus no XRT) provides a survival advantage, or that dose escalation beyond 40 Gy split-course provides a statistically significant survival advantage. Am I missing something? :)
 
True. diseases are too different to make such a generalization. However, locally advanced NSCLC is an excellent illustration of what Fisher meant.



My comment was in response to the 'Fisher hypothesis' and the assertion that local therapy does not impact overall survival in general. Fisher's hypothesis was actually based upon breast cancer. True- you do need to look at each disease state (primary site, stage, histology, and probably gene expression profiles) independently, which is an even more compelling reason to dismiss the Fisher hypothesis as a general oncology tenet.
 
There is a trial showing benefit of RT vs no RT for inoperable NSCLC from either the late 60s or early-mid 1970s. I actually was able to obtain a copy when I did a presentation as a student. Forgot author, but I know it is mentioned in the UpToDate article on inoperable NSCLC. Shows small but significant benefit.

It works (a little bit).
S
 
There is a trial showing benefit of RT vs no RT for inoperable NSCLC from either the late 60s or early-mid 1970s. I actually was able to obtain a copy when I did a presentation as a student. Forgot author, but I know it is mentioned in the UpToDate article on inoperable NSCLC. Shows small but significant benefit.

It works (a little bit).
S

You probably mean this trial (it's mentioned in the RadOnc wikibook):
http://www.ncbi.nlm.nih.gov/pubmed/4170866?dopt=Abstract
"The survival of patients with inoperable lung cancer: a large-scale randomized study of radiation therapy versus placebo."
 
James Cox writes an editorial discussing 0617 in the latest IJROBP

http://download.journals.elsevierhealth.com/pdfs/journals/0360-3016/PIIS0360301611037199.pdf


On June 17, 2011, two of the four arms in the Radiation Therapy Oncology Group (RTOG) 0617 protocol were closed to accrual when a planned interim analysis showed that the higher radiation dose being tested, 74 Gy, could not produce an overall survival benefit compared with the lower, standard dose of 60 Gy (1). This four-arm study was designed to test radiation with carboplatin and paclitaxel, with or without cetuximab, for Stage III non–small-cell lung cancer (NSCLC). Survival was compared between the two 74-Gy groups and the two 60-Gy groups, and overall survival was found to be better in the lower dose groups (1-year overall survival rate, 70.4% in the 74-Gy groups vs. 81% in the 60-Gy groups—both respectable rates for this stage of disease). However, even though 17 patients died in the 74-Gy arms compared with 7 in the 60-Gy arms, the toxicity rates were no different between the two dose groups. These findings are not only counterintuitive, they run counter to a large body of evidence showing that higher radiation doses lead to better tumor control at numerous sites. Deaths related to the effects on the normal lungs and perhaps the heart from high-dose three dimensional conformal radiotherapy (3D-CRT) and intensitymodulated RT (IMRT) are the most likely explanation of the findings from the RTOG 0617. Considerable evidence, summarized below, supports the hypothesis that the pulmonary or cardiopulmonary effects of thoracic RT can contribute to death
 
How does an investigator become part of the RTOG? Are only attendings allowed to become part of the RTOG? Can non-RTOG members get access to RTOG data? Thanks a lot!
 
Hitting ASCO shortly.... Cetuximab data should be available in 2014.

Not only was OS worse in the 74 Gy arm, so was LC :confused:

http://www.medpagetoday.com/MeetingCoverage/ASCO/39177
http://www.medscape.com/viewarticle/804282
http://www.onclive.com/conference-c...dvanced-NSCLC-More-Radiotherapy-Is-Not-Better

The final results of the large phase 3 RTOG 0617 trial will be presented at the 2013 Annual Meeting of the American Society of Clinical Oncology

Boosting the dose of radiation therapy in stage III non-small cell lung cancer (NSCLC) actually compromises outcomes and cuts survival, a randomized trial showed.

Overall survival was 56% higher with standard 60 Gy radiation than with a 74 Gy dose (median 28.7 versus 19.5 months, P=0.0007), Jeffrey D. Bradley, MD, of the Washington University School of Medicine in St. Louis, Mo., and colleagues found.

Local control, which usually refers to local recurrence, favored the conventional arm as well, they reported at a telephone press briefing in advance of their presentation here at the American Society of Clinical Oncology meeting, similar to an interim analysis of the trial in 2011.

After a decade of research, it's time to close the book on high-dose radiation in lung cancer therapy, ASCO President Sandra M. Swain, MD, of Georgetown University in Washington, D.C., commented at the briefing.

While the trial results were surprising, they "really should put an end to higher-dose treatment, given the better outcomes in the standard dose arm," she said, although noting that the findings wouldn't generalize broadly to early stage cancer.

The conventional thinking was that more radiation would more effectively kill the tumor and improve outcomes, as suggested in prior phase II studies, Bradley explained.

"That's why you do a phase III trial," he said.

However, the trial didn't yield clues as to why less was more in NSCLC.

Physician-reported side effects recorded in the trial didn't appear to account for the difference, nor did protocol deviations, Bradley noted.

The only significant difference in adverse events between arms was esophagitis, with rates of 21% versus 7% in the high- and standard-dose groups, respectively. There were numerically more deaths from toxicity in the high-dose group, although the difference didn't rise to statistical significance.

"Possible explanations are increased heart dose, extended therapy duration, unreported toxicities, possible too-tight margins in the high-dose arms, or a combination of these factors," Bradley suggested.

Unmeasured side effects, particularly to the heart, are top candidates, he told reporters, noting that left ventricular ejection fraction as an indicator of cardiac damage wasn't routinely monitored.

The 18-month rate of survival was 67% with standard radiation versus 54% with the high dose.

Local failure likewise came in at 34% versus 25%, respectively, for a 37% relative difference that was significant at P=0.0319.

The results were independent of cetuximab, although results from the cetuximab versus none comparison in the trial are expected to come out in 2014, Bradley noted.
 
Not only was OS worse in the 74 Gy arm, so was LC :confused:

Thanks for sharing this info. I have to hand it to RTOG leadership, 0617 was a well-designed trial and it gave a surprising answer to a very pertinent question.

To paraphrase Chris Berman . . .

"Everyone thought that 74 Gy would be superior to 60 Gy. It made logical sense, it was obvious! But you know what? That's why they run Phase III trials."
 
Thanks for sharing this info. I have to hand it to RTOG leadership, 0617 was a well-designed trial and it gave a surprising answer to a very pertinent question.

To paraphrase Chris Berman . . .

"Everyone thought that 74 Gy would be superior to 60 Gy. It made logical sense, it was obvious! But you know what? That's why they run Phase III trials."

I am looking forward to the manuscript. I am hoping they put some detailed DVH data from the the 60 Gy and 74 Gy groups.

The main message from this study is that 60 Gy remains the standard dose for radiation administered concurrent with chemotherapy for stage III NSCLC, Dr. Bradley concluded.

How are you guys going to modify your practice, if at all? Any changes for Stage III patients who are chemo ineligible?
 
Last edited:
How are you guys going to modify your practice, if at all? Any changes for Stage III patients who are chemo ineligible?

As soon as RTOG 0617 reported that high-dose arms were closing, I defaulted to 2 Gy x 30 fractions for all my Stage III patients.

Without chemo? "Game over man, game over . . ."

Per the results of German randomized trial between 60Gy/30 fractions and 32 Gy/16 fractions delivered b.i.d., I perform the latter. Equivalent results, more convenient for patient, less late toxicity.
 
I treat stage III to 66/33 fx if volume is limited or to 60/30 if mediastinal disease is widespread.

As for reasons for inefrior LC in high dose arm, I also suspect margins. As a resident, I helped to put a few patients on the protocol. 4D sim and image guidance were not mandated and CTV-PTV expansion could be as small as 5 mm as I recall.
 
I treat stage III to 66/33 fx if volume is limited or to 60/30 if mediastinal disease is widespread.

As for reasons for inefrior LC in high dose arm, I also suspect margins. As a resident, I helped to put a few patients on the protocol. 4D sim and image guidance were not mandated and CTV-PTV expansion could be as small as 5 mm as I recall.

Correct. I think that was mentioned in some of the commentary in the links above. Also, patients were enrolled as assigned even if V20 > 37%. I am really curious how the DVH parameters differed between the groups.

The Actual ASCO abstract can be found here: http://abstracts.asco.org/AbstView_132_110125.html

It's definitely practice changing. I think my new dose is going to be 60, with 66 reserved for the excellent PS patient with small-volume disease.
 
Curious why people would still give 66 Gy if 60 Gy has 9 months better median survival and local control than 74 Gy. It would be like giving 66 Gy if 74 Gy was better than 60 Gy. 9 months is incredible in lung cancer. I can't think of one intervention in stage III lung cancer that has led to a 9 month improvement in survival. Dose de-escalation is tremendous, and we should start trying it in other sites. Maybe if we go down to 50 Gy, can achieve a median survival of 3 years ... But, but seriously...

I think that certain trials have to be negative, because of statistics. Doesn't mean that higher doses don't work. It's just that this trial was negative. I don't think 9 months can be explained away by smaller margins or cardiotoxicity (Come on! This is lung cancer - if they live long enough for me to give them a heart attack, I think I've done my job). This is a statistical anomaly. It happens. I think I said in an older post, I think 74 Gy may not be high enough. We are giving 60 Gy in 3 fractions for stage I disease. How the heck is 74 Gy in 37 Fx going to cut it? There is a dose-response in lung cancer and we know this. Maybe it's time to explore SRS (definitive or as boost) in stage III? I don't know the answer. I know if we decide to stay with lower doses, we're not going to make a dent in overall survival for the next 20 years, as chemotherapy is not getting any better.

That being said, I've been going with 60 Gy/30 Fx for pretty much every stage III, even though I'm certain it's not enough. No reason to subject yourself to any scrutiny for a patient that dies after 66 Gy, when we have definitive "evidence" that dose escalation is doing more harm than good.

And... The median survival of 28.5 months in the conventional arm and 19 months in the high dose arm? Both of those are much higher than historical controls (17 months MS in best arm of RTOG 9410, 16 months in LAMP). I wonder why that's the case. (EDIT: Just learned PET-CT staging accounts for this)

Anyway, good trial. Wish it was just two arms, but even though it had a confusing outcome, seemed to be well executed.
 
Last edited by a moderator:
I have a question about the margins everyone assumes accounts for the local control difference. To meet V20, the CTV was allowed to be moved to the minimum, which would be from 1cm to 5mm according to the protocol. I know volumetrically that can make a difference for large tumors. My question really is, how quickly does the dose fall off in these plans? I have never done dose-escalated treatment. You would think the 60 Gy wash would be fairly big for a 74 Gy PTV coverage, perhaps big enough to cover the maximal PTV for the non-escalated plans. In other words, would the 60 Gy line (even if penumbra dose for 74 Gy) cover the same volume for each plan and thus negate the "margin" cause for difference in local control?

It would be interesting to see if 3D and IMRT are different for local control in the high dose arm as it was stratified. The 3D plans may not have as sharp a dose falloff and could have better LC if margins are the true cause. I am not convinced yet that IMRT provides better treatment in lung as yet due to the motion of tumors and difficulty of contouring compared to other sites. I have seen some people just do a 7 or 8 field fan like a prostate and spill dose everywhere
 
So the ASCO media releases are out, and already the ASCO president is quoted as saying that we "should close the book" on dose escalation in lung cancer.

The reason for this is that the straightforward interpretation seems to be the following: Patients treated with high-dose radiation died quicker and more often, therefore high dose radiation is bad.

What I am surprised about is how little light is being cast on the completely counter-intuitive, dare-I-say nonsensical result that high dose XRT led to more local failures.

Here is my Occam's Razor interpretation of 0617: Patients treated with high-dose radiation had higher local failures (P<0.05), and then those patients died from CANCER. If you have recurrent cancer, it tends to be a terminal event, and it's even likely that as a group, those with higher risk of early recurrences would die about 9 months sooner on average.

Am I missing something here? Is there a reason that the local failure interpretation is being downplayed by commentators and the press, and the much more suspicious, hand-wavy explanation of increased toxicity due to, perhaps, coronary and myocardium dose is being endorsed instead. After all, the local failure interpretation is a statistically significant result, while the toxicity interpretation is complete, chin-rubbing conjecture.

The corollary question, of course, is why should higher dose patients have increased local failures. The answer is seems obvious: marginal misses due to tight margins. It seems that Bradley and colleagues did not put enough emphasis on quality assurance, and as a field, we should be vocal in pointing this out. Otherwise, the "close the book" chorus will be the dominant voice. Yes, Bradley's feelings and ego will be hurt if we aggressively point this out, but that is the price to be paid for running a bad trial (and anyone who disagrees should point out how a "good trial" results in more failures with higher radiation dose despite a century of basic science and clinical trial results to the contrary).

I hope that the Europeans, Canadians, or a large single-institution will run a trial of 60 v. 66 with 4D CT and daily IGRT with quality assurance which includes careful attention paid to the margins. Until then, the question of dose-escalation is unresolved by 0617.
 
I think the coverage is irresponsible.

After a decade of research, it's time to close the book on high-dose radiation in lung cancer therapy, ASCO President Sandra M. Swain, MD, of Georgetown University in Washington, D.C., commented at the briefing.

There has been one abstract release at ASTRO, now this abstract. This particular abstract has not even been presented yet. There are at least 5 different hypotheses as to why the results may not be due to the high-dose radiation or could be due to some other technical aspect. So why is this any new news from ASTRO 2011? What is the point of generating this media buzz and this buzz within our own field for data that is incomplete, and that is inexplicable with what we currently know about lung cancer.

It's an interesting abstract. In my opinion, this should not be practice changing until it's published and we've really had a chance to analyze these interesting results. It's time to close the book on high-dose radiation? :rolleyes:



My take: the published ASCO abstract states: "The effect of the anti-EGFR antibody (cetuximab) awaits further follow up." Has everyone forgotten about the SWOG adjuvant gefitinib trial where placebo had a one year median survival advantage? http://www.ncbi.nlm.nih.gov/pubmed/18378568. Everyone assumes that these receptor targeted agents are more benign than the traditional chemotherapy, but it's clearly not the case in some circumstances.
 
There does appear to be a local control advantage of 60 Gy > 74 Gy. I agree that this does not make logical sense on the face of it. There are alternative explanations aplenty:

1. Perhaps 74 Gy arms used more IMRT = tighter margins = more beams with greater tissue-air dose build up issues
2, More toxicity as alluded above. Perhaps due to high V5 lung.
3. Longer treatment planning process to meet dose constraints in 74 Gy arm and/or longer breaks due to toxicity breaks

In the end I agree that the final publication and not the abstract would be most enlightening.

QA is a thorny subject for trials of this size and magnitude. Ideally, it should be rigorous but keep in mind there is already stringent criteria just to enroll in RTOG, let alone perform IMRT in RTOG trials. Rapid review is best suited for Phase II single arm trials that are testing a novel concept (e.g. hippocampal sparing IMRT).

How you choose to interpret the data is up to you. However, unless you are a thoracic RO specialist you probably have no business escalating dose beyond 60 Gy under normal circumstances.
 
Last edited:
Top