MAGIC trial vs. McDonald trial which is better for gastric cancer - we now know the answer

[Gfunk6]

Drum roll please . . . .

They are equivalent.

Link: http://www.esmo.org/Press-Office/Pr...tcomes-to-Post-Operative-Chemotherapy?hit=ehp

In this phase III study, 788 patients with stage Ib-IVa resectable gastric cancer were randomised upfront and were all given pre-operative chemotherapy consisting of three courses of epirubicin, a platinum compound (cisplatin or oxaliplatin) and capecitabine.

So the argument in tumor boards will continue to rage.

---

Members don't see this ad.

 

[BobbyHeenan]

My gut (and rad onc bias) feels like pre-op chemoXRT may be the way to go and would like this tested next. I'm not aware of a phase III in the U.S. that is looking at MAGIC vs. pre-op chemo XRT though.

The preop path CR on MAGIC was 0%, so there's room for better regimens in the preop setting.

 

[RadOncDoc21]

My gut (and rad onc bias) feels like pre-op chemoXRT may be the way to go and would like this tested next. I'm not aware of a phase III in the U.S. that is looking at MAGIC vs. pre-op chemo XRT though.

The preop path CR on MAGIC was 0%, so there's room for better regimens in the preop setting.

Seems like a reasonable approach to me, such as in almost all other major GI malignancies (esophagus, rectal). A neoadj chemoRT approach would at the least be less toxic (in my opinion).

 

[Neuronix]

Macdonald didn't have pre-op chemo... A big problem with MAGIC is that very few (43% IIRC) got through the entire course of chemotherapy.

And you see: "While the surgical quality was very high in the study, researchers noted that a significant number of patients did not start or complete the full course of either chemotherapy or chemoradiotherapy; 52% in the chemotherapy arm, and 47% in the chemoradiotherapy arm."

So it's kind of a flawed study off the bat. I'd like to see a per-protocol analysis, though it may lose all power.

Agree with neoadjuvant treatment. Problem with stomach of course is that it moves a lot and deforms a lot. Also it doesn't tolerate a whole lot of RT without toxicity. Could be an interesting question for modern image guidance...

 

[Palex80]

The main problem with the McDonald trial is the fact that patients didnt get a proper lymphadenectomy in the trial.
That's why radiation therapy added to local control and with it to overall survical.
With D2-lymphadenectomy the value of radiation therapy decreases, since the recurrence risk shifts to metastatic spread.

 

[BobbyHeenan]

The main problem with the McDonald trial is the fact that patients didnt get a proper lymphadenectomy in the trial.
That's why radiation therapy added to local control and with it to overall survical.
With D2-lymphadenectomy the value of radiation therapy decreases, since the recurrence risk shifts to metastatic spread.

That's definitely an issue, but there's still some applicability there, because though I don't see a lot of gastric cases (med onc usually gets a hold of them first and starts ECF), there are a lot of patients that don't get D2 dissections in real practice still for whatever reason. I saw this even at the academic center.

I'm hoping RTOG/NRG roll out a neoadjuvant study, would love to participate. I agree though that tolerance and image guidance will be tough, but I think it's doable.

High level anecdotal data here: I had a recent gastric MALT lymphoma patient that I treated first fraction in the morning (fasting) and she had an amazingly consistent stomach location.

 

[thesauce]

The main problem with the McDonald trial is the fact that patients didnt get a proper lymphadenectomy in the trial.
That's why radiation therapy added to local control and with it to overall survical.
With D2-lymphadenectomy the value of radiation therapy decreases, since the recurrence risk shifts to metastatic spread.

They've gone back and looked at the patients with D2 resection and found benefit even in those patients. And there's a more recent meta-analysis of 9 randomized trials similarly showing 5yr overall survival benefit with the addition of radiation. Also, 90% of stomach cancer patients that fail have some local component, so it's hard to argue against more aggressive local therapy.

One problem with Magic is it included lower esophageal patients who we already know benefit from N.A. therapy, so it may have skewed the results.

 

[BobbyHeenan]

They've gone back and looked at the patients with D2 resection and found benefit even in those patients. And there's a more recent meta-analysis of 9 randomized trials similarly showing 5yr overall survival benefit with the addition of radiation. Also, 90% of stomach cancer patients that fail have some local component, so it's hard to argue against more aggressive local therapy.

One problem with Magic is it included lower esophageal patients who we already know benefit from N.A. therapy, so it may have skewed the results.

And we have the Stahl POET trial as well that included lower esophagus and gastric cardia (neo adjv chemo vs. neoadjv chemo XRT) showing survival benefit for XRT and path CR 15% vs. 2% favoring XRT. As I recall though more post-op mortality and under powered study that didn't quite meet statistical significance for survival benefit but the curves seemed to diverge pretty well.

 

[Neuronix]

The main problem with the McDonald trial is the fact that patients didnt get a proper lymphadenectomy in the trial.
That's why radiation therapy added to local control and with it to overall survical.
With D2-lymphadenectomy the value of radiation therapy decreases, since the recurrence risk shifts to metastatic spread.

Absolutely true. Just want to point out that ARTIST still showed LRC benefit overall and DFS benefit for node positive gastric with adjuvant chemoradiotherapy vs chemotherapy after D2 dissection.

Again, no harsh ECF neoadjuvant therapy in that trial, so people actually got through it (mostly).

 

[radtothebone]

My gut (and rad onc bias) feels like pre-op chemoXRT may be the way to go and would like this tested next. I'm not aware of a phase III in the U.S. that is looking at MAGIC vs. pre-op chemo XRT though.

The preop path CR on MAGIC was 0%, so there's room for better regimens in the preop setting.

Not in the US, TOP GEAR will address neodjuvant chemo vs XRT (with some chemo). D2 not mandated, but recommended. Results expected many years down the road...

http://agitg.org.au/clinical-trials/trials-open-to-recruitment/8-top-gear/

 

[Haybrant]

We should change the title here as it's not quite magic vs McDonald. Brought it up at tumor board for an obstructed guy but was politely corrected

 

[Gfunk6]

We should change the title here as it's not quite magic vs McDonald. Brought it up at tumor board for an obstructed guy but was politely corrected

Interestingly, NCCN guidelines allows you to abort the MAGIC regimen in favor of chemoXRT if surgery shows R1/R2 resection.

To your point, I agree it is not a straight comparison since both arms received neoadjuvant ECF.

 

[Neuronix]

Interestingly, NCCN guidelines allows you to abort the MAGIC regimen in favor of chemoXRT if surgery shows R1/R2 resection.

That's based on zero evidence. This comes up every year at GI ASCO. What do I do adjuvantly if the neoadjuvant ECF didn't do anything? Then it's common sense: why continue something that isn't working?

You can make the same argument for poor pathological response at surgery to ECF or whatever neoadjuvant chemo. If it didn't work neoadjuvantly, maybe you should try something else adjuvantly like chemorads. You just need a very healthy person to tolerate neoadjuvant ECF and adjuvant RT+5-FU.

 

[evilbooyaa]

Surgeons here do a lot of surgeries off the bat (haven't seen a lot of pre-op chemo) D2 lymphadenectomies so anyone with positive LNs ends up getting treated per ARTIST's Sandwich method, usually with FOLFOX instead of XP, and 5-FU based CRT.

We've had two patients treated by POET. One guy absolutely did not tolerate treatment even to 30Gy and we stopped at 20Gy and sent the patient to surgery. Other lady went from T3N1 to T3N0 after completing POET and now has post-surgical issues.

 

[scarbrtj]

Thank you for posting. So many questions, so few good answers; but please let me hear from the "experts" at all the Meetings®