Resectability in stage IIIA NSCLC

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

Thaiger75

Senior Member
15+ Year Member
Joined
Nov 3, 2003
Messages
345
Reaction score
0
Anybody know of good guidelines as to what is considered resectable for N2 disease (IIIA) in NSCLC?

Members don't see this ad.
 
Resectability for IIIA is definitely a moving target. Our surgeons tend to use the following disease criteria to define as UN-resectable:

>1 involved N2 nodal station
Single nodal station >3cm
 
Also, based on the Intergroup study of CRT+surg vs. CRT alone, many surgeons are cautious if the patient will require a pneumonectomy, because of the mortality associated with this.
 
Members don't see this ad :)
There is also another randomized study out there:
EORTC 08941: No difference in PFS or OS for patients with Stage IIIA treated with induction Chemo + Surgery versus induction Chemo + RT.
Patients randomized had to have shown a response to chemotherapy.

Therefore if no consensus can be met on whether or not the patient is resectable, then an induction chemo may be a good idea, before deciding what to do with him.

Our chest surgeons see no difficulties in operating patients with multiple mediastinal positive lymph nodes. We usually have to treat the patients postoperatively then.


Also, based on the Intergroup study of CRT+surg vs. CRT alone, many surgeons are cautious if the patient will require a pneumonectomy, because of the mortality associated with this.
What I always failed to understand however is why SO many patients in the Intergroup study needed a pneumonectomy. Stage IIIA is not T4 (that would be Stage IIIB), therefore pneumonectomy should not be the standard procedure for such a tumour. Sleeve resections have also become popular in the past years. Therefore I expect that a correctly staged IIIA NSCLC should require a lobectomy or a sleeve resection as a standard procedure. Especially in the era of neoadjuvant chemo, we should be able to see less pneumonectomies in initially Stage IIIA NSCLC. This again raised the question if the Intergroup study results would not have been better, if not so many pneumonectomies were necessary (which substantially added to overall treatment letality in the surgery group).
 
Last edited:
There is also another randomized study out there:
EORTC 08941: No difference in PFS or OS for patients with Stage IIIA treated with induction Chemo + Surgery versus induction Chemo + RT.

just curious --- why did the EORTC proceed with that study given the RTOG/CALGB data that clearly showed the superiority of a combined approach over induction (for the RT arm)?
 
Therefore if no consensus can be met on whether or not the patient is resectable, then an induction chemo may be a good idea, before deciding what to do with him.

I don't agree with this approach. It essentially forces you to have, at best, induction chemo followed by RT (ala Dillman regimen) which has been shown to be inferior to concurrent chemoRT over and over again. (West Japan, Czech, French, and are-we-ever-going-to-see-the-manuscript-RTOG9410)

You can subgroup analyze until cows come home, but the bottomline for the intergroup trial was that it was negative. ChemoRT vs chemoRT->S for lobectomy candidates is a fine question to ask for a clinical trial but shouldn't be the standard of care in my opinion.

If there is truly really minimal mediastinal disease (for example, a single node after extensive CME), one could consider involving thoracic surgeons, but I would still push for concurrent chemoRT.

If someone is found to have occult IIIA disease after a surgery for a presumed N0-1 disease, then I would offer postop RT.

The real question is not whether the tumor is resectable, but whether surgery-based approach is what's best for the patient. I am not convinced.

*** See above for my answer at the Oral Board *** Below for the real life dilemma***

At our institution, we do push for chemoRT unless the patient is being considered for the protocol (now closed for poor accrual). We can do that because we have very strong thoracic rad onc (Dr. Choy) and thoracic med onc (Dr. Joan Schiller, ECOG thoracic committee chair). In real life where the thoracic surgeons see the patients first and send them to you for preop RT, I would have to be absolutely convinced to change the course of management. Surgery-based approach is not necessarily wrong, just not as strongly evidence-based. Thankfully, I don't have to worry about politics just yet (for 4 more months).
 
Alicia Keys sang it best ...

"No one, no one, no one/
Knows how to treat IIIA NSCLC"

I think those were the lyrics.

I don't think being too nuanced on oral boards is a great idea. Gotta go with concurrent chemoRT +/- consolidation chemo. Maybe not even get into the adjuvant treatment. Answering the question with data from a subset analysis from a negative trial sounds like a bad idea. The only IIIA treated with surgery is the "accidental" N2.

-S
 
In community setting any N2 disease, proven with either PET or med, is deemed unresectable.
 
Interesting to hear feedback on what other institutions do. Thanks for all the replies.

Our (UNC) group also heavily favors definitive chemorads for N2 patients and will rarely operate on any of them.

Just wasn't sure what happens in the "real" world outside our institution.

Are there certain mediastinal nodal stations that thoracic surgeons feel more excited about operating if they were to operate at your institution? Eg. R4 vs. subcarinal vs. AP window?

I would suspect so.....
 
In community setting any N2 disease, proven with either PET or med, is deemed unresectable.

This brings up an interesting question. If mediastinum is PET positive for a single node, does that make the patient N2, unresectable, and therefore chemoRT candidate? Or should you push for a mediastinal evaluation (CME/Chamberline/EBUS, etc)?

BTW, I wouldn't give the long-drawn answer at the oral board, either. I was just explaining my rationale for my answer which would be:
"IIIA should be treated with concurrent chemoRT with systemic dose chemo and no adjuvant chemo. Consolidation chemo should be considered if the patient did not receive full systemic dose during the chemoRT (e.g. weekly carbo/taxol)."
 
Last edited:
This brings up an interesting question. If mediastinum is PET positive for a single node, does that make the patient N2, unresectable, and therefore chemoRT candidate? Or should you push for a mediastinal evaluation (CME/Chamberline/EBUS, etc)?

At our institution, all PET positive med LNs need path confirmation through med evaluation.
 
Does anyone's institution do induction chemo followed by a med and then surgery if the med is negative?

Btw -- I thought the adjuvant chemo question was somewhat answered with the SWOG "swindle" (i.e. a great phase II SWOG study S9504 with adjuvant docetaxel after CRT, turned out to be a negative study in the phase III setting).

http://www.asco.org/ASCO/Abstracts+...w=abst_detail_view&confID=47&abstractID=33596

Induction chemo before CRT also didn't look too hot either:

http://jco.ascopubs.org/cgi/content/full/25/13/1698
 
Last edited:
The HOG trial was recently published. Hanna et al. JCO 26:5755, '08
 
Members don't see this ad :)
Didn't realize the manuscript was out. Thanks.




I guess at the end of the day, it's kinda like esophagus. Do CRT first, and +/- surgery in your fit pts.

I agree. Trimodality is likely to be the best treatment for appropriately selected patients with appropriate surgery (no pneumonectomies!!) ala esophagus.

Like the esophagus, it has been shown that those with pathCR after chemorads do better after surgery. But the question remains in esophagus and NSCLC, whether these pathCR do well with or without additional surgery. Question will remain unanswered unless we are able to cherry pick these patients noninvasively.
 
It is importatnt to remember in esophagus that the data supporting the equivalence of chemoRT and trimodality is only valid in SCC. Stahl et al. only enrolled SCC (they had another consecutively running trial in adenos comparing neoadjuvant chemo vs neoadjuvant chemoRT which was recently published in the JCO). If I remember correctly, in the FFCD trial (Bedenne et al) 88% were SCC. Also, both of these trials treated to high RT doses (64-66Gy). Therefore, you cannot say that definitive chemoRT is an equivalent approach to trimodality for adeno.

For SCC, you can say that definitive chemoRT using doses >60Gy is equivalent in terms of survival to trimodality therapy, but has never been tested against 50Gy. Depends on whether you truly believe that there is no dose response between 50.4Gy and 64.8Gy in esophageal SCC (as per the INT 0123 trial (Minsky et al.)). At our institution, we treat to 50.4Gy; but IMHO, I think that the negative intergroup dose escalation trial was a statistical abberation and, if repeated 10 times (like all breast cancer trials :)), the majority of the trials would be positive for dose escalation provided that appropriate lung constraints were met in the dose escalation arm. Problem with esophageal cancer is that many of the patients have poor PS or significant weight loss and are not medically fit for randomization, so we are stuck making treatment decisions based on small studies for which many of our patients would not have been eligible.

Moreover, both the FFCD and the Stahl trials demonstrated a significant improvement in LRC with trimodality therapy. The HR for local failure was 2.1-2.5 with definitive chemoRT versus trimodality. Similarly, in the intergroup 0139 trial in NSCLC, local control and progression-free survival (22% vs 11%) was significantly improved at 5 years in the trimodality arm. In both of these settings (stage IIIa NSCLC and esophageal SCC), the problem is an excess of perioperative deaths in the trimodality arm (9.8-12.3% in the esopheal trials; 7.8% in the stage IIIa NSCLC trial). This rate decreases with surgeon experience and at experienced institutions. It also strongly depends on patient selection. In the INT 0139 trial, peri-operative death was only 1% in lobectomy patients versus 27% in pneumonectomy patients. In addition, patients requiring a Rt sided pneumonectomy had worse outcomes compared to those having a Lt sided pneumonectomy. On subset analysis of patients requiring Lobectomy, the improvement in LRC and PFS translated into a doubling of 5yr OS (36% vs 18%). I agree that this is only a subset analysis and is not to be used for treatment justification on the oral boards; however, I also think that it is hypocritical for radiation oncologists to say that there is no role for a treatment that produces a LRC and PFS benefit, and may produce an overall survival benefit in appropriately selected patients. Remember, in many of the sites we treat adjuvantly, RT alone (without chemotherapy) does not produce a survival benefit!

All that being said, at our institution most of the patients with clinical N2 disease are treated with definitive chemoRT. I just think that we need to keep an open mind regarding trimodality therapy.
 
To clarify, I meant for oral boards or "evidence-based" practice. In real life, referring appropriate patients for surgery is not a bad idea after a response to ChemoRT for esophageal adenoCA or NSCLC.

"EBM" unfortunately leads to the situation where there is one answer for the boards and one answer for reality. That incongruity is baffling.

I wish they'd go through the case, and then follow-up with, "Now, what would you really do?"

-S
 
It is importatnt to remember in esophagus that the data supporting the equivalence of chemoRT and trimodality is only valid in SCC. .

I agree with most of what was said in this message. However, there IS data supporting trimodality for adenocarcinoma and GEJ in esophageal cancer: CALGB 9781 (Tepper JCO 2008). Keep in mind that there were many caveats to this study, mainly that they were only able to enroll 56 patients on the trial. Despite this, there was a statistically significant overall survival advantage in favor of trimodality over surgery alone.
 
Thaiger75 wrote: I agree with most of what was said in this message. However, there IS data supporting trimodality for adenocarcinoma and GEJ in esophageal cancer: CALGB 9781 (Tepper JCO 2008). Keep in mind that there were many caveats to this study, mainly that they were only able to enroll 56 patients on the trial. Despite this, there was a statistically significant overall survival advantage in favor of trimodality over surgery alone.

I agree with you Thaiger75, but I think that you missed what I was saying. I agree there is a lot of data for a survival benefit with trimaodality therapy in esophageal adeno. In addition to the referenced CALGB study (Tepper et al.), Walsh (NEJM 1996; 3yr OS 32% vs 6%; p=sig), Urba (JCO 2001; 3yr OS 30% vs 16%;p=trend ) and recent IPD meta-analysis (ASTRO 2008; 5yr OS improved by 7%) showed a survival advantage for trimodality therapy over surgery alone. As such, I think that trimodality therapy is the standard of care for esophageal adeno >T1N0. What I was saying is that there is no data to support definitive chemoRT in esopohageal adeno and I think that this would be a very dangerous answer on the board exam. The only data comparing surgical versus non-surgical approach for esopheal adenos that I am aware of is the Scandanavian trial (Carstens et al, ASCO 2008; not yet published) which compared chemoRT (64Gy) to transthoracic esophagectomy. This trial included about 100 patients and roughly 1/2 had adenocarcinoma. The two arms were equivalent in terms of DFS and OS. This does not, however, answer the question regarding the equivalence of neoadjuvant chemoRT plus esophagectomy versus definitive chemoRT. Also, in the famed RTOG 85-01 trial (Herskovic et al. NEJM) only 22 out of the 129 patients had adenocarcinoma histology and these patients had numerically lower (although not statistically significant due to small patient numbers; Cooper et al JAMA 1997) survival compared to the patients with SCC. Therefore, we are limited in applying the favorable survival data from this trial to patients with adenocarcinoma. As such, I think that the best answer boards and reality for esophageal adeno is trimodality.
 
Last edited:
Therefore, we are limited in applying the favorable survival data from this trial to patients with adenocarcinoma. As such, I think that the best answer boards and reality for esophageal adeno is trimodality.

very good point. I should have clarified that earlier. Most of the pts. in these trimodality vs. CRT trials were SCC pts, and over recent years, people have come to appreciate that you can't lump histologies together when it comes to esophagus.

As for dose escalation, I'm less than impressed, given the available data out there. People have tried everything, even brachy. which didn't turn out so great.
 
As for dose escalation, I'm less than impressed, given the available data out there. People have tried everything, even brachy. which didn't turn out so great.[/quote]

Good point about brachy. RTOG 92-07 did have limitations though (Brachytherapy catheter diameter of only 4-6mm), which undoubtedely contributed to the 14% incidence of fistulas and 10% incidence of treatment related death in patients receiving brachytherapy. Interestingly, the Stahl trial used brachytherapy using a 10-14mm applicator for patients with T3 tumors without high-grade stenosis and reported no late fistulas.

Regarding the intergroup study (INT 0123), we are all familiar with the excess of on-treatment deaths in the high dose arm (11 vs 2) and that 7/11 occured at doses less than 50.4Gy Even after an exploratory analysis excluding these patients, however, there was still no benefit to dose escalation. You can't argue with the results of this trial, but I think that it is important to remember that this was a trial of only ~230 patients and it is the only phase 3 trial of dose escalation with concurrent chemotherapy, that I am aware of, addressing this question.

I just think that it is interesting that esophageal SCC is the only disease that we treat to the same dose for neoadjuvant and definitive intent therapy. With the exception of anal canal cancer, there are no other squamous cell cancers in which 50Gy in conventional fractionation is a "curative" dose. Local failure and disease persistence still accounts for the majority of failures in patients treated with definitive chemoRT. As such, I think that there is still a potential role for dose escalation, particularly with more conformal treatment techniques. If dose escalation isn't the answer, I guess that we will have to hope for the best with concurrent Cetuximab...
 
We just had Dr. Willett visit us. His take on the dose escalation was that for a higher dose to make a difference, it may have to be really high, obviously difficult to do with esophagus. His conjecture was that both 50 Gy and 65 Gy were at the bottom of the sigmoid curve and that was why it made no difference in the Minsky trial.
 
Last edited:
Getting back to NSCLC, do you guys notice that your med oncs seem to love wussy carbo/taxol a lot as a typical backbone for unresectable IIIAs?

The SWOG regimen of Cis/etoposide seemed to be more efficacious, and based on the aggregate data, we probably should be using a cis-based regimen in our fittest unresectable patients. I think they do that in Europe a lot, although I'd love to hear from Palex regarding this?
 
We just had Dr. Willett visit us. His take on the dose escalation was that for higher dose to make a difference, it may have to be really high, obviously difficult to do with esophagus. His conjecture was that both 50 Gy and 65 Gy were at the bottom of the sigmoid curve and that was why it made no difference in the Minsky trial.

Makes sense to me. You can probably say that for pancreas/hepatobilliary too. It's probably a combo of not enough of dose, and not enough effective chemo for the systemic disease, given that these GI tumors like to go metastatic quite frequently.
 
Getting back to NSCLC, do you guys notice that your med oncs seem to love wussy carbo/taxol a lot as a typical backbone for unresectable IIIAs?

The SWOG regimen of Cis/etoposide seemed to be more efficacious, and based on the aggregate data, we probably should be using a cis-based regimen in our fittest unresectable patients. I think they do that in Europe a lot, although I'd love to hear from Palex regarding this?

Our med oncs use Cis/etoposide in almost all NSCLC chemorads patients (if they aren't healthy enough for cis/etoposide, then perhaps they aren't healthy enough for concurrent CRT). They feel that cisplatin is likely superior to carboplatin in upfront treatment of NSCLC.
 
Our med oncs use Cis/etoposide in almost all NSCLC chemorads patients (if they aren't healthy enough for cis/etoposide, then perhaps they aren't healthy enough for concurrent CRT). They feel that cisplatin is likely superior to carboplatin in upfront treatment of NSCLC.

I see many of ours get carbo/taxol
 
We just had Dr. Willett visit us. His take on the dose escalation was that for a higher dose to make a difference, it may have to be really high, obviously difficult to do with esophagus. His conjecture was that both 50 Gy and 65 Gy were at the bottom of the sigmoid curve and that was why it made no difference in the Minsky trial.

While I have a great deal of respect and admiration for Dr. Willet, I think that the argument that the doses evaluated in the Intergroup dose escalation trial "were not within sigmoid region of the dose response curve" is more convenient than plausible. Allow me to explain my reasoning. Even among presumed relatively radio-resistant tumors with high rates of local recurrence after radiotherapy or surgery and RT, such as stage III NSCLC or high grade glioma, there is a significant dose response which has been demonstrated between 40Gy and 60Gy (NSCLC: RTOG 73-01; HGG: MRC trial, Walker et al.). There IS, in fact, a dose response that has been suggested in unresectable pancreatic adenocarcinoma. In a GITSG trial comparing 40Gy and 60Gy with concurrent 5-FU chemotherapy, the 60Gy arm was associated with a non significant 3 month improvement in median survival (8.4mo vs 11.4mo) at the expense of increased GI toxicity (Moertel et al. Cancer 1981). Finally, in the notoriously radioresistant soft tissue sarcoma, a large retrospective series from MGH showed an improvement in local control with adjuvant RT doses >64Gy compared to doses ≤64Gy in patients with positive surgical margins (Delaney et al. Red J 2007).

Essentially, esophageal SCC is an upper aerodigestive malignancy with the same risk factors (smoking, ETOH) and, presumably, similar biologic behavior. Other upper aerodigestive maligancies (HNSCC, Lung SCC), exhibit dose response in the range of 50-65Gy. Moreover, there is biologic evidence of a dose response for esophageal cancer (both adeno and SCC) that can be derived from the pCR rates from trials of neoadjuvant chemoRT. In the Australian trial of neoadjuvant chemoRT (Burmeister et al. Lancet Oncol 2005) the neoadjuvant dose was 35Gy/15fx (2.33Gy/fx) and pCR was observed in 16/128 patients (12.5%) on the neoadjuvant arm. In the ECOG trial of RT versus chemoRT (Smith et al. Red J 1998), there was the option of having esophagectomy after 40Gy in 20 fractions. In the 23 patients randomized to chemoRT (5FU/MMC) who had surgery, 5 had pCR (21.7%). In the Urba study (JCO 2001) comparing neoadjuvant chemoRT (CDDP, 5FU, Vinblastine, 45Gy/30fx BID) to surgery alone, pCR was observed in 14 of 50 patients (28%) in the neoadjuvant arm. Finally, in the previously referenced CALGB study (Tepper et al.) comparing neoadjuvant chemoRT (5FU, CDDP, 50.4Gy) followed by esophagectomy to esophagectomy alone, 10 of 25 patients (40%) had pCR. While it is difficult to compare across studies, particularly in the setting of small patient numbers, these data suggest that there is a dose response in the range of 35-50.4Gy and, therefore, it is unlikely that the 50.4-64.8Gy dose range is below the steep region of the dose response curve.

I think that a more plausible explanation for the lack of dose response in esophageal cancer in the GI Intergroup study is a simple statistical aberration. The statistical calculations used for phase III clinical trials typically use a power of 0.8 or 80%, meaning that they have an 80% power to detect a difference if one truly exists. If you look back at any of the well-known individual patient data meta-analyses (chemoRT for head and neck, PCI for LS-SCLC, Thoracic RT for LS-SCLC, LACE meta-analysis of adjuvant chemo for resected NSCLC, any of the EBCTCG meta-analyses, etc), many of the individual trials are negative despite the overall benefit in the pooled analyses. One of the most striking examples of variable results in an individual trial is in the setting of chemoRT versus RT alone for cervix cancer. Five trials (GOG 85, GOG 120, GOG 123, GOG 109/Intergroup and RTOG 90-01) all showed significant improvements in PFS (~50%) and OS (~40-50%) with the addition of CDDP chemo to RT. The data was so compelling that the NCI issued a clinical practice announcement recommending the addition of CDDP to any patient requiring definitive or adjuvant RT for cervical cancer. Despite these drastic benefits, a trial performed by the NCIC (Pearcey et al. JCO 2002) showed no OS or DFS benefit with the addition of weekly CDDP to RT in 253 patients with large IB-IIA or IIB-IVA disease. The authors of this trial offered possible explanations, but in the end acknowledged that their results were most likely simply an outlier and they conceded that the preponderance of clinical evidence supported chemoRT. A negative result is known to occur in roughly 15-25% of trials, even if the experimental intervention is in fact beneficial. I think that this is the best explanation for the results of the Intergroup esophageal dose escalation study, but unfortunately it occurred in the first study rather than the sixth.

Obviously, we don't know the real reason that dose escalation hasn't panned out in esophageal SCC. In my practice, I will treat patients with SCC with chemoRT to 50.4Gy +/- esophagectomy beacuse I agree that this is the standard of care approach and higher doses are not supported by the currently available evidence. I'm just trying to provide some food for thought…
 
Last edited:
Getting back to NSCLC, do you guys notice that your med oncs seem to love wussy carbo/taxol a lot as a typical backbone for unresectable IIIAs?

The SWOG regimen of Cis/etoposide seemed to be more efficacious, and based on the aggregate data, we probably should be using a cis-based regimen in our fittest unresectable patients. I think they do that in Europe a lot, although I'd love to hear from Palex regarding this?

Regarding chemotherapy in unresectable stage IIIA NSCLC, I agree that weekly carbo/taxol is inadequate compared to systemic doses of CDDP/VP-16 (SWOG regimen). This is summed up nicely in the conclusion of the CALGB 39801 study (Vokes et al, JCO 2008). IMHO, that is the weakness of the RTOG 0617 study… you may see a benefit of 74Gy with weekly carbo/taxol or 74Gy with weekly carbo/taxol and C225, but it really doesn’t prove that it is superior to 63-66Gy with concurrent systemic doses of CDDP/VP-16. We use CDDP/VP-16 whenever possible...
 
just curious --- why did the EORTC proceed with that study given the RTOG/CALGB data that clearly showed the superiority of a combined approach over induction (for the RT arm)?
That is a very good question. I presume that the EORTC had some concerns when it came to delivering concurrent RCT after induction RT in terms of toxicity.

I don't agree with this approach. It essentially forces you to have, at best, induction chemo followed by RT (ala Dillman regimen) which has been shown to be inferior to concurrent chemoRT over and over again. (West Japan, Czech, French, and are-we-ever-going-to-see-the-manuscript-RTOG9410)[/quite]
I don't agree with your point.
What about delivering concurrect RCT after induction CT.
The EORTC did not give chemotherapy together with RT, but who says you can't do it. It probably wont be better than concurrent RCT upfront like recent CALGB evidence showed, but it may gie you the lucury of deciding which patient to operate and which not.
Surely there is a case for timing the RT in relation to the CT and the earlier you deliver CT, the better it is. However what you want to see here is if you are treating a cancer that responds to treatment and for this you do need some kind of induction treatment. If the tumour still does not shrink then you can either rethink of operating or going on with the concurrent RCT.

If someone is found to have occult IIIA disease after a surgery for a presumed N0-1 disease, then I would offer postop RT.
I would be very careful with generalizing this.
I would be very careful in offering RT in the case of 1 positive N2 node, when the surgeon extracted something like 20 lymph nodes and 19 of them were negative. Please bear in mind that the evidence for postOP RT is quite thin at the moment and mainly based on retrospective subgroup analysis of trials like ANITA.

In real life where the thoracic surgeons see the patients first and send them to you for preop RT, I would have to be absolutely convinced to change the course of management. In real life where the thoracic surgeons see the patients first and send them to you for preop RT, I would have to be absolutely convinced to change the course of management.
Based on what evidence do you do preoperative RT in Stage IIIA NSCLC patients? If I am not mistaken a recommendation for preoperative RT only exists in the case of Sulcus superior tumours. For a standard Stage IIIA NSCLC, the evidence for preoperative RT is very thin.

Just wasn't sure what happens in the "real" world outside our institution.
Well in Europe Stage IIIA NSCLC is regularly treated with surgery in our clinic and in other clinics I have been to. Most of the times, patients receive preoperative CT.
Primary RCT is reserved for those patients that either decline surgery or are medically unfit.

If mediastinum is PET positive for a single node, does that make the patient N2, unresectable, and therefore chemoRT candidate? Or should you push for a mediastinal evaluation (CME/Chamberline/EBUS, etc)?
PET positive nodes are explored in our clinin before decision of definitve treatment, via your mentioned methods. PET positive is not the evidence. And since the N2 +/- will decide what the therapy would be, we feel like having to prove it.

I agree. Trimodality is likely to be the best treatment for appropriately selected patients with appropriate surgery (no pneumonectomies!!) ala esophagus.
I would be careful with such a statement. Esophagus experience has not shown a clear overall survival benefit with such an approach and toxicity is often limiting.
 
I don't agree with this approach. It essentially forces you to have, at best, induction chemo followed by RT (ala Dillman regimen) which has been shown to be inferior to concurrent chemoRT over and over again. (West Japan, Czech, French, and are-we-ever-going-to-see-the-manuscript-RTOG9410)

I don't agree with your point.
What about delivering concurrect RCT after induction CT.
The EORTC did not give chemotherapy together with RT, but who says you can't do it. It probably wont be better than concurrent RCT upfront like recent CALGB evidence showed, but it may gie you the lucury of deciding which patient to operate and which not.
Surely there is a case for timing the RT in relation to the CT and the earlier you deliver CT, the better it is. However what you want to see here is if you are treating a cancer that responds to treatment and for this you do need some kind of induction treatment. If the tumour still does not shrink then you can either rethink of operating or going on with the concurrent RCT.

Induction chemo followed by concurrent chemoRT was also tested againt upfront concurrent (plus consolidation) in LAMP trial and was inferior. While I do see your point about assessing the tumor response for information on resectability and prognosis, I still think one should approach these cases with the most bang upfront.

One can give chemoRT then go with surgery (lobectomy only) or, if not a candidate for R0 lobectomy, complete with definitive dose chemoRT. If one really insists on info from induction, I think this is far more palatable approach then induction with chemo alone.

There is a role for induction chemo followed by RT; for example, I would give it to frail patients with a large mass who wouldn't be able to tolerate concurrent chemoRT.
 
Last edited:
Based on what evidence do you do preoperative RT in Stage IIIA NSCLC patients? If I am not mistaken a recommendation for preoperative RT only exists in the case of Sulcus superior tumours. For a standard Stage IIIA NSCLC, the evidence for preoperative RT is very thin.

Sorry, I meant preop RT as part of preop chemoRT. I do not advocate giving RT alone for induction. (I am in general against surgical approach with preop therapy, but I can be convinced to go with preop chemoRT) Thanks for the chance to clarify.
 
Sorry, I meant preop RT as part of preop chemoRT. I do not advocate giving RT alone for induction. (I am in general against surgical approach with preop therapy, but I can be convinced to go with preop chemoRT) Thanks for the chance to clarify.

Still even in the case of preoperative RCT, evidence is thin. No trial has proven a survival benefit through this approach. In fact, many would say that the patients who would do good with preop RCT should rather continue to full dose preop RCT, rather than being operated.
As I said, other than Sulcus superior tumors, preop RCT is not a general recommendation for NSCLC Stage IIIA.

One can give chemoRT then go with surgery (lobectomy only) or, if not a candidate for R0 lobectomy, complete with definitive dose chemoRT. If one really insists on info from induction, I think this is far more palatable approach then induction with chemo alone.
And far more toxic probably if one really does go for resection.


Our clinic's algorithm for Stage IIIA looks like this

Generally operable (even with pneumonectomy) and medically fit:
Preop CT then Resection. According to R and pN, then eventually consolidation RT. This is the general recommendation to the patient.
Concurrent RCT is an option and is offered to patients who don't want to get operated or want to get rid of their tumor as soon as possible (some patients do not like preoperative CT, they want to fight the tumor directly with resection of RT, rather than wait for a couple of months).


Inoperable and/or medically unfit:
Concurrent RCT. If tumor very big, bad lung function and chance to get a better DVH with smaller primary tumor, then induction CT followed by RCT.
 
Still even in the case of preoperative RCT, evidence is thin. No trial has proven a survival benefit through this approach. In fact, many would say that the patients who would do good with preop RCT should rather continue to full dose preop RCT, rather than being operated.

I agree the evidence for this approach is not solid. In the INT 0139, the survival rates appear to be separating with a long term follow up. The latest report (ASCO 05 --- anyone know of more recent update?) had them at 27 vs 20% at 5 years with p = 0.10 and PFS was outright improved (p=0.017). If you select out the patients (which consisted of about 1/3 of patients on the surgical arm) who had clearly unacceptable (26%) treatment related mortality after pneumonectomy, you could make a pretty strong argument for this approach for patients who can have R0 lobectomy.

Will I recommend this as my first choice in real life or on the board? No. But, I can be swayed if the right patient presented him/herself.


Our clinic's algorithm for Stage IIIA looks like this

Generally operable (even with pneumonectomy) and medically fit:
Preop CT then Resection. According to R and pN, then eventually consolidation RT. This is the general recommendation to the patient.
Concurrent RCT is an option and is offered to patients who don't want to get operated or want to get rid of their tumor as soon as possible (some patients do not like preoperative CT, they want to fight the tumor directly with resection of RT, rather than wait for a couple of months).


Inoperable and/or medically unfit:
Concurrent RCT. If tumor very big, bad lung function and chance to get a better DVH with smaller primary tumor, then induction CT followed by RCT.

It's very interesting to see how we practice differently in different continents. Thanks for sharing the algorithm.
 
Last edited:
Palex80: "I would be very careful with generalizing this.
I would be very careful in offering RT in the case of 1 positive N2 node, when the surgeon extracted something like 20 lymph nodes and 19 of them were negative. Please bear in mind that the evidence for postOP RT is quite thin at the moment and mainly based on retrospective subgroup analysis of trials like ANITA."


I think that this is one of the situations in which European and American Radiation Oncologists differ in their orientation. Of course the well-known PORT meta-analysis demonstrated a survival decrement with PORT in N0-1 patients and no survival advantage or decrement with PORT for N2 patients. The largest trial included in this meta-analysis was a French trial (Dautzenberg et al. Cancer 1999) which accounted for 728/2128 (34%) of the patients in the meta-analysis. This study demonstrated a significant survival decrement in patients treated with PORT (5yr OS 43% vs 30%) due to an excess of intercurrent deaths (which may be part of the reason why Europeans favor no PORT!), and this study strongly contributed to the 7% absolute decrease in survival found in the PORT meta-analysis. What people often forget, however, is that PORT was associated with a 24% reduction in the risk of local recurrence. An older study by the LCSG of patients with stage II-III SCC without involvement of the highest mediastinal nodal region sampled, demonstrated substantial reduction in ipsilateral lung and mediastinal recurrence with PORT compared to no PORT (3% vs 41%). Moreover, N2 patients had reduction in overall risk of recurrence. The PORT meta-analysis was also limited by antiquated RT techniques: 5/9 trials allowed RT doses up to 60Gy, 4/9 allowed fractional doses greater than or equal to 2.5Gy, 7/9 used Co60, 5/9 did not include the bronchial stump in the target volume (which is the most common site of recurrence after surgery in POF studies), and only 2/9 used CT planning.

In light of the improvement in local control with PORT and the reduction in overall disease recurrence among N2 patients in the LCSG study, I think that it is reasonable to revisit the issue of PORT for N2 disease in the era of adjuvant chemotherapy and modern 3D-RT/IMRT techniques. In addition to the previously referenced post-hoc analysis of PORT in the ANITA study (showing a benefit in patients with N2 disease, and N1 patients not receiving chemotherapy), A SEER study (Lally et al. JCO 2006) and a single institutional study from Mayo have suggested a survival benefit with PORT for N2 disease. Moreover, POF studies (Kelsey et al. Red J 2006), help to identify the high risk regions that should be included in the post-op target volume. Improved RT techniques (using CT simulation and 3D conformal or IMRT) and tailored treatment of more select target volumes, should decrease the incidence of intercurrent/treatment-related deaths.

Finally, recent data has demonstrated improvements in systemic disease control and overall survival with adjuvant CDDP-based chemotherapy (IALT, ANITA, JBR.10). None of the 9 trials included in the PORT meta-analysis included adjuvant chemo. As systemic disease control continues to improve with adjuvant chemotherapy, local control will likely become an increasingly important predictor of survival. Remember in breast cancer that a survival benefit with post-mastectomy radiation was only realized in trials using adjuvant systemic therapy (CMF chemo or Tamoxifen).

I think that it is reasonable to offer PORT to patients with N2 disease, provided that you discuss the absence of a proven survival benefit with this approach. However, I think the sufficient evidence exists to say that PORT improves intra-thoracic disease control, which might translate into improved survival in the setting of adjuvant chemotherapy. The ongoing French trial evaluating PORT for N2 disease will hopefully help resolve the controversy in these patients.
 
Getting more technical, how do you guys treat your patients? fields, dose, PET/CT fusion, IMRT?

Anyone care to comment on IFRT (involved field)? Obviously, in a desire to dose-escalate, IFRT started to come into favor over the last 1-2 decades. Does anyone use elective nodal RT for any reason?
 
Getting more technical, how do you guys treat your patients? fields, dose, PET/CT fusion, IMRT?

Anyone care to comment on IFRT (involved field)? Obviously, in a desire to dose-escalate, IFRT started to come into favor over the last 1-2 decades. Does anyone use elective nodal RT for any reason?

We still treat many of our patients with 3D Conformal, but over the past year or so have been treating more with IMRT. I think that IMRT has two potential pitfalls in lung treatment. First, it is associated with increased V5 and, in some cases, V10 volumes. This is readily apparent in the mesothelioma experience in which these low dose parameters are highly predictive of high grade pulmonary toxicity or pulmonary-related death (Allen et al. Red J 2006; Rice et al. Red J 2007; Miles et al. Red J 2008). Moreover, treatment planning systems tend to underestimate the dose in regions receiving <10Gy compared to Monte Carlo calculations (Jang et al. Red J 2008). These low dose parameters may be associated with increased rates of lung toxicity and radiation pneumonitis. Secondly, historically intra-fraction respiratory-related tumor motion has caused concern regarding the possibility of geographic miss with IMRT due to the steep dose gradient at the periphery of the target volume. This is less of a concern in patients planned with 4D techniques and internal target volumes. In patients planned using 4D CT, we have increased the proportion of patients treated with IMRT.

Most of our patients are treated using cone-beam CT image guidance, provided that the PTV can be encompassed within the 21cm x 16cm field size allowed on the Elekta Synergy S platform. Otherwise they are treated on standard liner accelerators. All of our patients planned using 4D CT technique are simulated and treated with respiratory inhibition using an abdominal compression plate, which reduces the amplitude of respiratory motion by 40-50% (Heinzerling et al UTSW Red J 2008). For CTV margins, I tailor the margin width based on histology. I like that MSKCC study (Giraud et al Red J 2000) which demonstrated that the necessary margin width to cover microscopic extension in 95% of patients was 8mm for adenoCa and 6mm for SCC (I generally treat large cell like adeno). ITV to PTV margin is a moving target, but we have mostly been using 5mm margins in patients treated using cone-beam IGRT.

Whenever possible, we obtain a PET/CT in the treatment position (in vacuum-type cradle). This is particularly helpful for patients with significant post-obstructive atelectasis. These are fused to the planning CT for GTV delineation. Unfortunately, PET-based planning and 4D-CT are mutually exclusive, unless you have the capability of 4D PET acquisition (which we do not). Therefore, you have to decide which is more important in an individual patient: PET-based planning or 4D planning. In patients with significant atelectasis which cannot be delineated from tumor on CT, I would argue that PET-based planning is more important. Conversely, in the patient with a lower lobe primary located near the diaphragm, with significant respiratory-associated tumor motion and no post-obstructive atelectasis, I would argue that 4D is the more important technology in treatment planning.

I would treat to 64-66Gy using 2Gy fractions with concurrent CDDP (50mg/m2 d 1,8) and VP-16 (50mg/m2 d1-5). The SWOG studies (SWOG 95-04 and SWOG 0023) and HOG 01-24 demonstrate median survival in the range of 19-26 months (for all enrolled patients from the start of treatment) using this approach. I also favor treating the involved-field only. Rosenzweig and colleagues recently updated the MSKCC experience treating with IFRT only (JCO 2007). In this study, actuarial elective nodal failure was observed in only 7.6% at 2yrs. Moreover, among patients treated using PET-defined IFRT, elective nodal failure was only 6% at 2 years. Similarly, a study from MDACC using PET-defined IFRT revealed isolated out-of-field failure in only 5.7% (Klopp et al Red J). A Chinese study randomized (Yuan et al. AJCO 2007) 200 patients with stage III NSCLC to IFRT (68-74Gy) versus elective nodal RT (60-64Gy) with concurrent CDDP-based chemo. This study reported significant improvements in local control (51% vs 36%) and 2 year OS (39% vs 26%) favoring the dose-escalated IFRT arm. Also, IFRT was associated with reduced incidence of RT pneumonitis (17% vs 29%). Given the low rate of elective nodal failure and increased toxicity associated with elective nodal treatment, I think that IFRT is a reasonable standard approach. That being said, we still treat the ipsilateral hilum in all stage III patients (whether or not it is involved). This is based on the post-hoc analysis of 1700 patients treated on RTOG studies (Emami et al) showing increased IF progression in patients with inadequate coverage of the ipsilateral hilum (11% vs 22%, p=0.01). We generally don't do elective mediastinal radiation for enyone with NSCLC.
 
Last edited:
I would treat to 64-66Gy using 2Gy fractions with concurrent CDDP (50mg/m2 d 1,8) and VP-16 (50mg/m2 d1-5). The SWOG studies (SWOG 95-04 and SWOG 0023) and HOG 01-24 demonstrate median survival in the range of 19-26 months (for all enrolled patients from the start of treatment) using this approach. I also favor treating the involved-field only. Rosenzweig and colleagues recently updated the MSKCC experience treating with IFRT only (JCO 2007). In this study, actuarial elective nodal failure was observed in only 7.6% at 2yrs. Moreover, among patients treated using PET-defined IFRT, elective nodal failure was only 6% at 2 years. Similarly, a study from MDACC using PET-defined IFRT revealed isolated out-of-field failure in only 5.7% (Klopp et al Red J). A Chinese study randomized (Yuan et al. AJCO 2007) 200 patients with stage III NSCLC to IFRT (68-74Gy) versus elective nodal RT (60-64Gy) with concurrent CDDP-based chemo. This study reported significant improvements in local control (51% vs 36%) and 2 year OS (39% vs 26%) favoring the dose-escalated IFRT arm. Also, IFRT was associated with reduced incidence of RT pneumonitis (17% vs 29%). Given the low rate of elective nodal failure and increased toxicity associated with elective nodal treatment, I think that IFRT is a reasonable standard approach. That being said, we still treat the ipsilateral hilum in all stage III patients (whether or not it is involved). This is based on the post-hoc analysis of 1700 patients treated on RTOG studies (Emami et al) showing increased IF progression in patients with inadequate coverage of the ipsilateral hilum (11% vs 22%, p=0.01). We generally don't do elective mediastinal radiation for enyone with NSCLC.

In light of the recent Red Journal editorial on ENI by Drs. Kelsey, Marks, and Glatstein, I wonder if anyone treats elective nodes. (Duke? UNC? UPenn?)

We don't do ENI, but we would take the dose higher than 64-66 Gy TarHeelRadOnc suggested. Off protocol, we routinely take the dose to 70-74 Gy to a 4D CT defined ITV + setup margin (around 5mm). We commonly use weekly carbo/taxol (I know, I know) followed by systemic dose consolidation chemo.
 
Last edited:
In light of the recent Red Journal editorial on ENI by Drs. Kelsey, Marks, and Glatstein, I wonder if anyone treats elective nodes. (Duke? UNC? UPenn?)

We don't do ENI, but we would take the dose higher than 64-66 Gy TarHeelRadOnc suggested. Off protocol, we routinely take the dose to 70-74 Gy to a 4D CT defined ITV + setup margin (around 5mm). We commonly use weekly carbo/taxol (I know, I know) followed by systemic dose consolidation chemo.

We do "limited" ENI up to 46-50 Gy, before shrinking the field for a total dose of 60-66 Gy to the primary tumour (according to DVH).
"Limited" ENI means we don't regularly treat both supraclavicular fossae and the entire mediastinum, but rather follow a more custom-tailored approach according to primary tumour location, extent of lymph node involvement, histology. For ttreatment planning we use FDG-PET-CT-fusion.

I firmly believe there is a role for "limited" ENI both in NSCLC and SCLC patients. However one must take care not to make the fields too big. Furthermore the more chemotherapy a patient is receiving, the smaller the fields can get and possibly should be to avoid sequlae.
 
There was a paper a year ago or so (off the top of my head dont have the refrence) that looked at patients in which the intent was to treat GTV + margin. Much of the neighboring nodal stations were in fact covered. This of course is impacted by patient, tumor and treatment factors (i.e. margin used, etc) factors.
 
There was a paper a year ago or so (off the top of my head dont have the refrence) that looked at patients in which the intent was to treat GTV + margin. Much of the neighboring nodal stations were in fact covered. This of course is impacted by patient, tumor and treatment factors (i.e. margin used, etc) factors.

I think that's the MSKCC paper on ENI that came out in the JCO a year ago.
 
At our institution it often comes down to what is reasonable. If we're covering most of the ipsilateral hilum to cover the GTV we draw the fields a little larger. If enlarging the fields to cover the nodes because "they're right there" results in an unacceptable V20 or MLD then we don't do it.

I'm curious how other groups interpreted the Chinese study comparing 74 Gy to involved field vs 60 Gy to ENI. One interpretation we've discussed is that if you don't give enough dose (i.e., 74 Gy) it doesn't matter how large your field is you're bound to fail.
 
At our institution it often comes down to what is reasonable. If we're covering most of the ipsilateral hilum to cover the GTV we draw the fields a little larger. If enlarging the fields to cover the nodes because "they're right there" results in an unacceptable V20 or MLD then we don't do it.

I'm curious how other groups interpreted the Chinese study comparing 74 Gy to involved field vs 60 Gy to ENI. One interpretation we've discussed is that if you don't give enough dose (i.e., 74 Gy) it doesn't matter how large your field is you're bound to fail.


Rhe Chinese comparison was not a standard randomised comparison where you change only one variable and treat both groups apart from that in the same way. They changed two variables, both field and dose.
Therefore the Chinese study was not designed to answer the question:
"Can we ommit ENI?"
The Chinese study was designed to answer the question:
"Is radiation therapy with higher dose on the tumour without ENI better than radiation therapy with less dose and ENI?"

Speculation:
Do ENI up to 46 Gy, covering tumour+affected lymph nodes+non-affected lymph nodes, then shrink the field up to 60 Gy for tumour+affected lymph nodes (using new CT), then reshrink the field and perform ESRT with something like 2x5Gy (60% tumour-enclosing isodose) only for the tumour.
How do you think this strategy would work out?
 
Last edited:
Top