Challenging NSCLC-case!

[Palex80]

Here is a challenging NSCLC case.

75 year old patient, good PS.

Small primary in the left lung, mediastinum looks mostly clear on PET, only a few very small nodes. MRI of the brain shows solitary <1cm lesion.
Histology come backs as adenocarcinoma in the lung, EGFR-mutated.

I just SRSed the brain lesion (EGFR-mutation was not known prior to the planning CT ).

The thoracic surgeons performed a mediastinoscopy this week to rule out any mediastinal involvement, they resected two nodes, which came back positive in the right mediastinum (with only minor involvement). So, it's a cT1c cN3 cM1 (BRA; solitary).


What to do?

Possible alternatives:

1. It's cN3 cM1. Give the patient Tagrisso and call it a day until progression occurs (hopefully in several years).
2. We just killed the met, so let's go for The Full Monty! Concurrent radiochemotherapy for the primary and mediastinum, add Tagrisso as adjuvant treatment (extrapolated from ADAURA).
3. Re-PET, if the mediastinum looks clear after the mediastinoscopy, then SBRT the primary and give Osimertinib.

---

Members don't see this ad.

 

[Ray D. Ayshun]

4. 5 gy x 5

In all seriousness, I'd let the patient choose between 1 and 2. Maybe could justify more of a port dose to the mediastinum to reduce toxicity a little I suppose.

 

[Lamount]

Here is a challenging NSCLC case.

75 year old patient, good PS.

Small primary in the left lung, mediastinum looks mostly clear on PET, only a few very small nodes. MRI of the brain shows solitary <1cm lesion.
Histology come backs as adenocarcinoma in the lung, EGFR-mutated.

I just SRSed the brain lesion (EGFR-mutation was not known prior to the planning CT ).

The thoracic surgeons performed a mediastinoscopy this week to rule out any mediastinal involvement, they resected two nodes, which came back positive in the right mediastinum (with only minor involvement). So, it's a cT1c cN3 cM1 (BRA; solitary).


What to do?

Possible alternatives:

1. It's cN3 cM1. Give the patient Tagrisso and call it a day until progression occurs (hopefully in several years).
2. We just killed the met, so let's go for The Full Monty! Concurrent radiochemotherapy for the primary and mediastinum, add Tagrisso as adjuvant treatment (extrapolated from ADAURA).
3. Re-PET, if the mediastinum looks clear after the mediastinoscopy, then SBRT the primary and give Osimertinib.

If he has a good PS, I would go full monte and do CRT. These EGFR mutant lung cancer can be aggressive, so observing the mediastinum is probably risky. You have localized disease (at this point) so you should start with a local modality (so no osi alone). As for the adjuvant Osi after CRT... my personal druthers would be too hold off. It's already metastatic and you can't really trust the PET to identify disease. With Osi, you get a deep initial response and about 1.5 years of disease control before the music stops. Either you are lucky and the brain + lung + mediastinum is the true extent of disease (in which case Osi is pointless), or you are just seeing the tip of the iceberg and you are using the Osi early. Would save Osi for progression.

 

[jondunn]

Save Osi for progression?

You have an incredibly active drug also with great CNS penetration in EGFR-mutated lung cancer which is a systemic disease, even when you happen to catch an early stage lung cancer.

Definitely give the Osi, there is zero question about it.

I would not give chemoradiation. if you want to add local therapy, do SBRT to the only gross disease left behind - the primary. Can do it now along with starting Osi, or wait 3-6 months and assess and do it then.


Do not hold Osi.

 

[Lamount]

Save Osi for progression?

You have an incredibly active drug also with great CNS penetration in EGFR-mutated lung cancer which is a systemic disease, even when you happen to catch an early stage lung cancer.

Definitely give the Osi, there is zero question about it.

I would not give chemoradiation. if you want to add local therapy, do SBRT to the only gross disease left behind - the primary. Can do it now along with starting Osi, or wait 3-6 months and assess and do it then.


Do not hold Osi.

Osi won’t cure this patient. If you give osi in this case, it will either do nothing (because it wasn’t needed) or will give you a PFS of 1.5 years


Edit. If you give Osi instead of local therapy, you will get a PFS of 1.5 years. Adjuvantly, it will either do nothing or give you a PFS if 1.5 years.

The response is so deep when you start osi, what advantage is there to adjuvant over salvage? I have yet to see a patient with EGFR mut metastatic nsclc cured with osi. The cancer always develops resistance. No reason to give it a head start.

 

[jondunn]

Osi won’t cure this patient. If you give osi in this case, it will either do nothing (because it wasn’t needed) or will give you a PFS of 1.5 years

lol.

Osi has a much higher chance of helping this patient than chemoradiation without Osi. this is a fundamental misunderstanding.

you talking about cure?

 

[Lamount]

lol.

Osi has a much higher chance of helping this patient than chemoradiation without Osi. this is a fundamental misunderstanding.

you talking about cure?

One of these curves has a tail, and the other one doesn't.

 

[jondunn]

You’re ignoring biology. And btw the Gomez trial was consolidation after initial systemic therapy

Like I said I don’t oppose local therapy. Doing it in a chemoradiajron style and delaying Osi or even holding Osi altogether makes little to no sense. I think that’s honestly potentially a legal case to not give a patient with EGFR mutated lung cancer a systemically active drug

 

[Lamount]

You’re ignoring biology. And btw the Gomez trial was consolidation after initial systemic therapy

Like I said I don’t oppose local therapy. Doing it in a chemoradiajron style and delaying Osi or even holding Osi altogether makes little to no sense. I think that’s honestly potentially a legal case to not give a patient with EGFR mutated lung cancer a systemically active drug

This question comes up in our tumor board quite often...

I don't think adjuvant osi is necessarily wrong... and it is certainly easy to justify from extrapolating from ADURA. However, you are more-or-less starting a 1-2 year clock before the cancer develops resistance. If there are isolated recurrences, you can play wack-a-mole, but if it is widespread, your patient is hosed.

In select patients with this circumstance, we have sometimes treated definitively and followed closely. If they develop a local extracranial recurrence, we use local therapy. If they develop a widespread or CNS recurrence, we start Osi. This way, we can add another 6 months+ to that clock. This is particularly appealing in elderly patients as Osi can cause a lot of GI issues and can lead to FTT.

As for the prospect of a lawsuit... (if I were the medical oncologist), I would offer both treatment options to the patient, presenting the pluses and minuses of each approach. I would share my opinion but would support them whatever they choose. It's hard to get in trouble doing that...

 

[Lamount]

You’re ignoring biology. And btw the Gomez trial was consolidation after initial systemic therapy

Like I said I don’t oppose local therapy. Doing it in a chemoradiajron style and delaying Osi or even holding Osi altogether makes little to no sense. I think that’s honestly potentially a legal case to not give a patient with EGFR mutated lung cancer a systemically active drug

P.S. I was hoping you wouldn't point that out about Gomez... I would have cited SABR-COMET but I suspect you would have mentioned that only 20% were lung

 

[Mandelin Rain]

1. SBRT primary (only known active disease site currently/low-risk procedure for small primary)
2. Osi
3. If mediastinal-only progression in future, treat with chemoradiation

 

[Ray D. Ayshun]

Why ignore the mediastinum if there's cancer there and you're giving systemic therapy that allows it to hang out for a couple years, hit the gym and get angry?

 

[Palex80]

Why ignore the mediastinum if there's cancer there and you're giving systemic therapy that allows it to hang out for a couple years, hit the gym and get angry?

I believe the argument is that all macroscopic disease has been resected and microscopic disease is probably all over the place anyway (since the patient had a brain met). So, the number of cancer cells in the patient's mediastinum right now may be the same as the number of cancer cells in the patient's liver.

 

[Mandelin Rain]

Why ignore the mediastinum if there's cancer there and you're giving systemic therapy that allows it to hang out for a couple years, hit the gym and get angry?

Is there cancer there? Probably, but no PET avid disease. Couple small, completely resected contralateral nodes taken out.

What's your target? Whole mediastinum?

I can name a couple reasons why I wouldn't want a broad mediastinal field concurrent with chemo if I was an active 75 year old with a different highly-effective option.

 

[Mandelin Rain]

I believe the argument is that all macroscopic disease has been resected and microscopic disease is probably all over the place anyway (since the patient had a brain met). So, the number of cancer cells in the patient's mediastinum right now may be the same as the number of cancer cells in the patient's liver.

Bingo.

 

[Palex80]

Bingo.

We think very much alike, all the time!

 

[Ray D. Ayshun]

So then, what's the typical post-op approach in resected n+ nsclc? Tki alone, or cytotoxic chemo/consideration of port? Give the effective tki alone, sure, but why sbrt if no symptoms.

 

[Neuronix]

SRS brain met is already done. Arguable whether that was necessary but certainly not wrong.

We typically give the osi for awhile and keep a close eye on them.

When the osi fails or patient can't tolerate, then consolidate with chemoRT usually 60/30 but could consider a hypofractionated course. I think something this extensive (N3) probably deserves a more protracted course.

I don't think there's a right answer here. I don't think there is a curative intent therapy given that this is metastatic disease and there is no curative regimen in metastatic NSCLC. I don't think that immunotherapy or targeted agents change the fact that this is incurable oligometastatic disease.

But if you wanted to start with definitive chemoRT and then give adjuvant osi, I think that would be reasonable. Our patients generally cling to osi as being less toxic than chemo and would be very hesitant to take a chemo upfront approach without med onc being on board.

 

[Ray D. Ayshun]

Is there cancer there? Probably, but no PET avid disease. Couple small, completely resected contralateral nodes taken out.

What's your target? Whole mediastinum?

I can name a couple reasons why I wouldn't want a broad mediastinal field concurrent with chemo if I was an active 75 year old with a different highly-effective option.

Why perform a med on this patient in the first place if it's going to change virtually nothing?

 

[jondunn]

The only wrong answer is to not give a patient with metastatic EGFR lung cancer Osi, with a known Brain met especially

 

[TheWallnerus]

mediastinum looks mostly clear on PET,

That's a little too gauzy for me to know what to tell you; like a mostly pregnant woman. If I felt like I could target some real disease in the mediastinum, I would treat primary and mediastinum to 60/30 with chemoRT (and adj immuno obv). If I felt I were blindly firing RT into the mediastinum 1980s style, no thank you. SBRT to primary only. The PET would be a decision point for me.

 

[Neuronix]

That's a little too gauzy for me to know what to tell you; like a mostly pregnant woman. If I felt like I could target some real disease in the mediastinum, I would treat primary and mediastinum to 60/30 with chemoRT (and adj immuno obv). If I felt I were blindly firing RT into the mediastinum 1980s style, no thank you. SBRT to primary only. The PET would be a decision point for me.

Good point. If you see nothing suspicious on CT/PET I'm not sure it's worth chasing all over the mediastinum. SBRT or a hypofrac regimen to gross disease only reasonable too. Not a simple case, this one!

 

[Palex80]

That's a little too gauzy for me to know what to tell you; like a mostly pregnant woman. If I felt like I could target some real disease in the mediastinum, I would treat primary and mediastinum to 60/30 with chemoRT (and adj immuno obv). If I felt I were blindly firing RT into the mediastinum 1980s style, no thank you. SBRT to primary only. The PET would be a decision point for me.

Absolutely fair point.

This is how it looked prior to the mediastinoscopy.
Small nodes, not PET-avid.
Yet, precisely at this area 2 nodes were positive with small metastatic burden in them (<5mm), which is probably why the PET missed them.

PET is now 7 weeks old, the patient has not started Osimertinib yet, so I was thinking of doing the PET again immediately prior to commencing any local treatment to make sure no other mets are progressing.

 

[communitydoc13]

I would SBRT lung unless clear gross disease in mediastinum. Preponderance of data that treating areas with likely metastatic disease in mediastinum is only going to impact regional control. Can treat mediastinum down the road if necessary.

Tagrisso not that easy to tolerate long term. But agree with adding now in the setting of a brain met (almost certainly microscopic disease in brain).

I have had multiple EGFR+ patients behave more like true oligometastatic patients over the years. Many of these patients go off of TKI therapy due to poor tolerance.

 

[TheWallnerus]

Absolutely fair point.

View attachment 344608


This is how it looked prior to the mediastinoscopy.
Small nodes, not PET-avid.
Yet, precisely at this area 2 nodes were positive with small metastatic burden in them (<5mm), which is probably why the PET missed them.

PET is now 7 weeks old, the patient has not started Osimertinib yet, so I was thinking of doing the PET again immediately prior to commencing any local treatment to make sure no other mets are progressing.

I will assume that what the CT surgeons are calling right mediastinum is more from the yellow area vs the green area (the med can probably only access nodes in the yellow and green area), so this is probably just a "barely N3" case. Nevertheless, pretty low volume disease and I assume the primary is smallish. I am a little concerned on the basis of the +med the suspicious looking paratracheal nodes in the blue area. I would cover the blue area and the yellow (+/- green) region and the primary and it probably wouldn't be a very bad V20 at all to 60 Gy/30 fx and minimal if any esophagitis.

(In a traditional med these are the regions usually sampled and what I'm saying is the anterior med nodes may be the "right sided nodes" in this case. Would ask the CT surgeon to look at my volumes if I were going to treat with chemoRT.)

 

[Mandelin Rain]

We think very much alike, all the time!

I'm so sorry.

I will assume that what the CT surgeons are calling right mediastinum is more from the yellow area vs the green area (the med can probably only access nodes in the yellow and green area), so this is probably just a "barely N3" case. Nevertheless, pretty low volume disease and I assume the primary is smallish. I am a little concerned on the basis of the +med the suspicious looking paratracheal nodes in the blue area. I would cover the blue area and the yellow region and the primary and it probably wouldn't be a very bad V20 at all to 60 Gy/30 fx and minimal if any esophagitis.

(In a traditional med these are the regions usually sampled and what I'm saying is the anterior med nodes may be the "right sided nodes" in this case. Would ask the CT surgeon to look at my volumes if I were going to treat with chemoRT.)

I imagine the positive nodes were in the blue zone i.e. right paratracheal (4R).

If rescan was done and overt gross disease remains in mediastinum, I could definitely be convinced to target it. Otherwise, you'll likely have a bunch of 4-6 mm shoddy nodes (maybe/probably reactive???) post mediastinoscopy to chase after.

 

[Palex80]

I'm so sorry.

Don't be!

I imagine the positive nodes were in the blue zone i.e. right paratracheal (4R).

Precisely!

If rescan was done and overt gross disease remains in mediastinum, I could definitely be convinced to target it. Otherwise, you'll likely have a bunch of 4-6 mm shoddy nodes (maybe/probably reactive???) post mediastinoscopy to chase after.

Precisely!

 

[TheWallnerus]

I imagine the positive nodes were in the blue zone i.e. right paratracheal (4R).

Precisely!

You can get there (blue) with a mediastinoscopy? (How do you get past the aorta in front of it.) I thought you'd need a Wang or something for that region. We need a CT surgeon to poke his head in here.*

*EDIT: yeah you can get 4R with a med... would cover that "area" if doing chemoRT. If you see shoddy slightly bright nodes above and below, what to do? If you can make a plan that looks reasonable with low esophagitis vibes, I would tx.

 

[Palex80]

The surgeons assessed the right mediastinum first (2R & 4R) and found one positive node in each area. Then, they stopped.
Their standard practice is to assess the contralateral mediastinum first and if it's clear on frozen section (rule out N3) then move on to the ipsilateral side.

This was a standard mediastinoscopy, not a TEMLA (transcervical extended procedure).
Our surgeons rarely perform TEMLA, they rather resort to VATS to assess stations 5&6.

From what I know, standard mediastinoscopy is able to assess mediastinal stations 1, 2R/L, 4R/L, 7.


The primary is a couple of cms higher:

 

[Mandelin Rain]

You can get there (blue) with a mediastinoscopy? (How do you get past the aorta in front of it.) I thought you'd need a Wang or something for that region. We need a CT surgeon to poke his head in here.

I assume a cervical and not a Chamberlain

Also, let's try to be more inclusive of our female CT surgeons.

 

[Ray D. Ayshun]

Fair enough, after reading up on details of ADAURA, if the argument is that the med was an oncologic resection and you sbrt the primary, and the use the ADAURA approach, considering these two aspects/findings:

Administration of standard postoperative adjuvant chemotherapy before randomization was allowed but not mandatory; decisions about whether patients would receive adjuvant chemotherapy were made by the physician and the patient and were made before trial enrollment.

The disease-free survival benefit with osimertinib was observed irrespective of whether patients received adjuvant chemotherapy or not. Of patients who received adjuvant chemotherapy, 89% who received osimertinib and 49% who received placebo were alive and disease-free at 24 months (overall hazard ratio for disease recurrence or death, 0.16); of patients who did not receive adjuvant chemotherapy, these percentages were 89% and 58%, respectively (overall hazard ratio, 0.23).

SBRT the primary to make this a complete oncologic resection and give osi, presuming negative PET/you don't just want to treat a sensical mediastinal volume to begin with, makes sense. At the same time, I still have trouble believing 24 month DFS is a good endpoint for a drug that works for 24 months.

 

[TheWallnerus]

The surgeons assessed the right mediastinum first (2R & 4R) and found one positive node in each area. Then, they stopped.

This was a standard mediastinoscopy, not a TEMLA (transcervical extended procedure). Our surgeons rarely perform TEMLA, they rather resort to VATS to assess stations 5&6.

The primary is a couple of cms higher:
View attachment 344613

Quite a discontiguousness between the primary and LN stations. Which I am generally OK with. This is a case on the cusp. You have already treated overt metastatic disease in the brain, now you have to decide whether to treat semi-overt disease in the mediastinum. I would, *if* the plan looks very good for V20 and esophagus. There is data of course that "aggressive" chemoRT to chest in solitary brain met or low-volume oligomet cases can be associated with improved outcomes; much of that though was before the current immunotherapy era.

 

[Palex80]

Quite a discontiguousness between the primary and LN stations. Which I am generally OK with. This is a case on the cusp. You have already treated overt metastatic disease in the brain, now you have to decide whether to treat semi-overt disease in the mediastinum. I would, *if* the plan looks very good for V20 and esophagus. There is data of course that "aggressive" chemoRT to chest in solitary brain met or low-volume oligomet cases can be associated with improved outcomes; much of that though was before the current immunotherapy era.

I am a bit concerned about giving large volume RT to the mediastinum because of the Osimertinib. There are reports of pneumonitis, when combining these.

 

[TheWallnerus]

I am a bit concerned about giving large volume RT to the mediastinum because of the Osimertinib. There are reports of pneumonitis, when combining these.

"I would, *if* the plan looks very good for V20 and esophagus."

I did put asterisks around the "if."

I have a sneaking suspicion that if you don't irradiate at least the suspicious area(s)--they at least look suspicious, and not negative--and especially if you SBRT the primary, the first site of progression will be the mediastinum. We can certainly argue whether this pattern of progression will affect his overall survival.

 

[metview]

What are thoughts on Gomez approach and starting with systemic therapy for few months and if there is a response/stable disease, consolidate both the primary and mediastinum (only involved levels) with chemoRT to 60 Gy? There is likely microscopic disease in the mediastinum not detected by initial PET and likely not removed by surgery.

 

[jondunn]

What are thoughts on Gomez approach and starting with systemic therapy for few months and if there is a response/stable disease, consolidate both the primary and mediastinum (only involved levels) with chemoRT to 60 Gy? There is likely microscopic disease in the mediastinum not detected by initial PET and likely not removed by surgery.

That’s fine.

My own personal preference if consolidating the mediastinum is to do 15 fraction RT alone.

 

[Mandelin Rain]

What are thoughts on Gomez approach and starting with systemic therapy for few months and if there is a response/stable disease, consolidate both the primary and mediastinum (only involved levels) with chemoRT to 60 Gy? There is likely microscopic disease in the mediastinum not detected by initial PET and likely not removed by surgery.

I completely understand this, but....

Is there not also likely untreated microscopic disease in the brain? Obviously, no one is going to advocate for WBRT here.

What if you treat the involved levels and they recur immediately above/below marginal to previous fields? Have you now burned that bridge for a rather dubious benefit?

 

[TheWallnerus]

I completely understand this, but....

Is there not also likely untreated microscopic disease in the brain? Obviously, no one is going to advocate for WBRT here.

What if you treat the involved levels and they recur immediately above/below marginal to previous fields? Have you now burned that bridge for a rather dubious benefit?

 

[TheWallnerus]

Don't want to derail the convo/case but at least somewhat tangentially related as we are talking about DFSs here in this case. If RT for NSCLC could give us a 0.66, p<0.004 hazard ratio after surgery for DFS we would be... big fans of PORT in NSCLC. Immunotherapies are, let's face it, pretty good for cancer patients. Lung cancer is just in no shape form or fashion today like it was 20 years ago.

 

[Lamount]

Don't want to derail the convo/case but at least somewhat tangentially related as we are talking about DFSs here in this case. If RT for NSCLC could give us a 0.66, p<0.004 hazard ratio after surgery for DFS we would be... big fans of PORT in NSCLC. Immunotherapies are, let's face it, pretty good for cancer patients. Lung cancer is just in no shape form or fashion today like it was 20 years ago.

Interested to see how the IIIa patients compare to similar patients on PACIFIC

 

[Lamount]

The only wrong answer is to not give a patient with metastatic EGFR lung cancer Osi, with a known Brain met especially

Are there any data showing that adjuvant Osi in this setting improves survival as compared to salvage... or are you just supposing?

In this particular case, there is a question as to whether one can reasonably target the involved areas of the mediastinum, so it may be reasonable to do SBRT -> Osi, but I I am not aware of data that would support your level of confidence regarding a treated solitary brain met in EGFR+ cancer.

Like I said above, upon initiation of Osi, these patients usually have a profound response... so do you really lose any ground by waiting?

If you have a reference support adjuvant Osi in this scenario, please share, because this actually does come up quite a bit in our patient population, and we always struggle with this question... and it would be nice to have some data to nudge us.

 

[jondunn]

Are there any data showing that adjuvant Osi in this setting improves survival as compared to salvage... or are you just supposing?

In this particular case, there is a question as to whether one can reasonably target the involved areas of the mediastinum, so it may be reasonable to do SBRT -> Osi, but I I am not aware of data that would support your level of confidence regarding a treated solitary brain met in EGFR+ cancer.

Like I said above, upon initiation of Osi, these patients usually have a profound response... so do you really lose any ground by waiting?

If you have a reference support adjuvant Osi in this scenario, please share, because this actually does come up quite a bit in our patient population, and we always struggle with this question... and it would be nice to have some data to nudge us.

Why are you calling this adjuvant?

The patient has metastatic cancer and a highly active drug. You give it.

 

[communitydoc13]

Why are you calling this adjuvant?

The patient has metastatic cancer and a highly active drug. You give it.

I agree. The only situation I'm aware of where we consider "definitive" treatment of oligometastatic disease with XRT while deferring highly effective systemic therapy is hormone sensitive pCA per something like the ORIOLE trial. (Not that I practice this way and I believe that the bulk of the data now would favor including not only ADT but also Abiraterone if the goal is longevity).

 

[Lamount]

Why are you calling this adjuvant?

The patient has metastatic cancer and a highly active drug. You give it.

Not to say that the NCCN guidelines always outline the best clinical choices, but it is worth you knowing that your approach isn't the only reasonable one...

When you can definitively treat all known sites of disease, pulling the trigger on systemic therapy isn't always the right choice. This is especially true with drugs like Osi that are highly effective at first (even with a high burden of disease) but only work for a finite window.

Again, if your argument is supported by some data, I am always eager to learns something new

 

[bubbachuck]

I agree with NCCN here. Small burden of disease so maximize therapy to give patient best chance of cure if not durable PFS/OS.
If you wanted to do 3m of osi first, it's not unreasonable, but I think most med oncs at academic institution would agree on primary treatment at some point. Whether to give adjuvant osi or IO is debatable.

I would not recommend (1) and I think most tumor boards would agree, if only because NCCN is pretty clear

 

[mavrik13]

Funny how opinionated some people are. It's a tough case without a clear, data driven answer. I'd be biased by the patient's age and favour SBRT to the primary -> Osi as a very tolerable treatment and eradicating all macroscopic disease with SBRT, and treating microscopic disease with Osi.

If it were a young, fit patient who really wanted to go for maximal treatment in the absence of data, chemoRT including involved mediastinal stations (+/- intervening stations) is quite reasonable. Have certainly done that for non EGFR mutated disease. If EGFR mutated, I'd probably favour Osi at progression.

 

[evilbooyaa]

Strange case. I agree with SRS upfront for a single brain met.
Not sure why patient had a mediastinoscopy. I would've likely done Osi x 3 months followed by SBRT to primary disease in the setting of negative mediastinal lymph nodes.
I think if you can localize where the + LNs were and they had some potential radiologic correlates (even if normally would've been equivocal) pre-mediastinoscopy, then can consider 60/30 with chemo.

I also would totally be on board with Osi x 3 months and consolidate all gross disease (likely SBRT to the primary assuming no mediastinal recurrence) at that time a la Gomez.