cT1N0 breast cancer. RNI?

[scarbrtj]

Is the final conclusion that anyone with N+ at any given time (pre/post-chemo, surgery), we should just go ahead and give RNI (including IM’s)?

Did we simplify breast?

Rad onc needs breast fellows.

---

Members don't see this ad.

 

[RadOncDoc21]

Rad onc needs breast fellows.

You may have found your true calling!

 

[scarbrtj]

You may have found your true calling!

 

[RadOncDoc21]

Is the final conclusion that anyone with N+ at any given time (pre/post-chemo, surgery), we should just go ahead and give RNI (including IM’s)?

Did we simplify breast?

Leave it to breast for it to be the first time I ever officially replied to my own post: but just think this through with me.

If she was 65, presented at Tumor Board as a clinical T1N0M0, luminal A, we would all likely say it’s ok for her to omit RT if she is to get hormones.

If pathology confirms she has one node positive now we are saying it’s ok to treat. Now the volumes, techniques, dose, +\- a boost, RNI +\- IM’s. We have a 30 page discussion about it all but technically everything is ok to do.

If this lady never had her axilla addressed by the surgeon, she likely would have had a positive lymph node we were all ok with leaving completely alone because we didn’t know that she had disease in there and we would all be doing the patient a disservice by treating an elderly patient with stage I disease as discussed and preached by anyone who ever heard an ASTRO lecture. If she is to get hormonal therapy then the node may be of no issue but look at how one patient can make so many different pathways and decisions. Again from no treatment to treat it all with anything from a local regional recurrence is bad to lymphedema is bad and likely no OS benefit!

We went from being able to sleep peacefully at night not caring anything about the radiation we just blew off initially and now the same radiation we discussed in significant detail on how important it is to go after the nodes but yet unsure if there is truly a significant benefit or not in this patient. All because the surgeon discovered a positive lymph node in someone we were all willing to leave alone in the first place.

This is why breast is the worst!

 

[gmsquid]

To remind us all re: MA.20....

"The primary outcome was overall survival (82.3% vs 80.7%, p=0.06).
Secondary outcomes were disease-free survival (82.0% vs 77.0%, p=0.01),
isolated locoregional disease-free survival (95.2% vs 92.2%, p=0.009),
and distant disease-free survival (86.3% vs 82.4%, p=0.03)."

And...
"Patients in the nodal-irradiation group had higher rates of grade 2 or greater acute pneumonitis (1.2% vs. 0.2%, P=0.01),
and lymphedema (8.4% vs. 4.5%, P=0.001)."

And from discussion...
"Although subgroup analyses were prespecified, they were generally not adequately powered to assess the benefit of treatment in different subgroups. Furthermore, the P values of the subgroup analyses were not adjusted for multiple testing.15 Patients with ER-negative or PR-negative tumors appeared to benefit more from regional nodal irradiation than those with ER-positive or PR-positive tumors. Although this effect was not observed in previous trials of postmastectomy radiotherapy,16 it supports the hypothesis that further research on the molecular characterization of the primary tumor may identify patients who are more likely to benefit from regional nodal irradiation.17,18 Since the number of node-negative patients in our trial was relatively small, the application of our results to node-negative patients is unclear. In addition, at the time of our study, the size of nodal metastasis was not routinely measured, so it is difficult to generalize our findings to patients with micrometastases."

Unknown: proportion of patients Her2+. Study was 2000-2007; from 2005 onward, patients could get herceptin.

As a counterpoint to this, on nsabp b-28 the addition of paclitaxel to AC did not improve 5 year OS 85% both arms) and resulted in a small improvement in 5 year DFS (72 -> 76%). Calgb 9344 showed a similar improvement in DFS with paclitaxel (65->70%) with an ultimate small improvement (3%) in OS (77-> 80%). Looking at the DFS, likely a worse patient population in CALGB. Both of those trials had 3000 patients and were all in node positive women.

adding regional nodal rt improves DFS from 77-82% (I.e. a similar absolute improvement in dfs but a greater relative improvement than paclitaxel on calgb) but in a more favorable population on ma20. In the eortc study this was 69-> 72% and there was a trend towards a survival benefit. Ma.20 is a Much smaller trial than calgb (1800 vs 3000 patients) of a way more favorable population (as evidenced by the dfs in the non rt group). Eortc was a large trial (4000 patients) but also of a somewhat more favorable population than calgb and also included a substantial percentage of node negative patients (44%) yet almost showed a survival benefit.

Think about all the women who get peripheral neuropathy from paclitaxel compared to the small percentage of women who get late toxicity from rt. Paclitaxel is routinely given for a marginal OS benefit. This benefit is even questionable as one trial had it and one did not.

If you select more unfavorable patients node positive patients as done on calgb you’re likely to find a survival benefit to rt. And even the very favorable population on ma.20 with fewer patient numbers you have a substantial dfs benefit both absolutely and relatively greater than paclitaxel.

 

[drewdog1973]

As a counterpoint to this, on nsabp b-28 the addition of paclitaxel to AC did not improve 5 year OS 85% both arms) and resulted in a small improvement in 5 year DFS (72 -> 76%). Calgb 9344 showed a similar improvement in DFS with paclitaxel (65->70%) with an ultimate small improvement (3%) in OS (77-> 80%). Looking at the DFS, likely a worse patient population in CALGB. Both of those trials had 3000 patients and were all in node positive women.

adding regional nodal rt improves DFS from 77-82% (I.e. a similar absolute improvement in dfs but a greater relative improvement than paclitaxel on calgb) but in a more favorable population on ma20. In the eortc study this was 69-> 72% and there was a trend towards a survival benefit. Ma.20 is a Much smaller trial than calgb (1800 vs 3000 patients) of a way more favorable population (as evidenced by the dfs in the non rt group). Eortc was a large trial (4000 patients) but also of a somewhat more favorable population than calgb and also included a substantial percentage of node negative patients (44%) yet almost showed a survival benefit.

Think about all the women who get peripheral neuropathy from paclitaxel compared to the small percentage of women who get late toxicity from rt. Paclitaxel is routinely given for a marginal OS benefit. This benefit is even questionable as one trial had it and one did not.

If you select more unfavorable patients node positive patients as done on calgb you’re likely to find a survival benefit to rt. And even the very favorable population on ma.20 with fewer patient numbers you have a substantial dfs benefit both absolutely and relatively greater than paclitaxel.

What he said.

 

[scarbrtj]

Eortc was a large trial (4000 patients) but also of a somewhat more favorable population than calgb and also included a substantial percentage of node negative patients (44%) yet almost showed a survival benefit.

So I ask again: why no RNI in node-negative patients then? Would it not be well supported by the Janusian NEJM RNI studies of 2015. (Don't let the recent EORTC update cast a pall.) But I'm not being very honest. I know why. It was so hammered into the historical rad onc's head that PMRT was only good in 4 or more LNs positive, it just seems "weird" to do RNI in zero LN positive cases. And, yet, there's data (for some +/- DFS benefits, and other nouns I can't recall).

As a counterpoint to this, on nsabp b-28 the addition of paclitaxel to AC did not improve 5 year OS 85% both arms) and resulted in a small improvement in 5 year DFS (72 -> 76%). Calgb 9344 showed a similar improvement in DFS with paclitaxel (65->70%) with an ultimate small improvement (3%) in OS (77-> 80%). Looking at the DFS, likely a worse patient population in CALGB. Both of those trials had 3000 patients and were all in node positive women.

Time has so thoroughly marched onward since those studies; do they even apply nowadays? More "cool story bro?" I am not med-onc-y enough to explain it, but breast patients just do REALLY well nowadays. So well I think we rely on MA.20 and EORTC 22922 a little too much, too; two studies that started accruing around 20-25 years ago. That's almost pre-Chicxulub, br CA-wise. Her2 therapies have been substantial game changers as well. ("Six years ago, a medical oncologist told me he had never seen a patient with HER2-positive breast cancer develop local recurrence. I set out to find patients who had recurrence in this setting. Approximately 2,000 patients later, I have not seen one, either.") But, really, as I said, it's all changing. Survival is constantly improving in breast cancer. Just look at this:

I think... I can confidently say rad oncs are not contributing to this survival improvement trend. At least not in the last decade. Of course I know XRT can/does improve br CA survival. But how (ie tx volume debates) is up for debate. (We have made our survival improvements just by applying RT vs not applying RT at all.)

If you select more unfavorable patients node positive patients as done on calgb you’re likely to find a survival benefit to rt.

Well, IDK if a 65 yo lady with cT1N0 and a good response to ("weak") NAC and a 2 mm/4 mm primary/nodal residual is the one in which we're going to find this illusive survival benefit. If it's there, it's well hidden. Show me the money. "Follow me and I will make you Fishers of men." Not Matthew fishers: Bernie Fishers.

 

[gmsquid]

Why no rni in node negative patients? Risk stratification. The predominant potential benefit of rni in node negative patients is likely in triple negative breast cancer. in node positive patients, that benefit is higher cause you’re more likely to address the micronodal disease that results in locoregional and distant recurrence. Why not give chemo to a 0.3 mm breast cancer that is node negative? The risk is low. That even holds for triple negative disease.

The eortc analysis update is not a difficult thing to argue. The more time passes the more deaths you have and the more events you have in regards to dmfs. Breast cancer events are still lower with rt in a largely node negative population

Of course mortality has improved in breast cancer substantially. It comes down to many reasons. Improved screening, better systemic therapy, better delivery of systemic therapy, better radiation, better delivery of radiation. One quarter of breast cancers are her2 positive. Before herceptin these women all metted out. At the same time, there are women who still recur despite herceptin and perjeta. Just because an oncologist says he’s never seen a local recurrence of her2 breast cancer doesn’t mean it doesn’t exist. Also, why does rni reduce distant metastasis. It likely addresses that nodal disease that seeds distant nets. And honestly, we underdetect locoregional recurrence in breast cancer. We don’t routinely look at the axilla or imns with routine imaging and how many patients do you actually go back and assess those areas? And how many breast cancer patients have you treated for Palliation of axillary nodes that already had distant mets? Once somebody is called distant the locoregional is forgotten.

radiation is a cost effective treatment that reduces the risk of local, regional and distant recurrence of breast cancer by percentages similar or better than most systemic therapies. It is also minimally toxic. As much as you harp on academics diminishing the role of rt, you really want to remove it from breast cancer.

 

[gmsquid]

Well, IDK if a 65 yo lady with cT1N0 and a good response to ("weak") NAC and a 2 mm/4 mm primary/nodal residual is the one in which we're going to find this illusive survival benefit. If it's there, it's well hidden. Show me the money. "Follow me and I will make you Fishers of men." Not Matthew fishers: Bernie Fishers.

Also, the assumption is that the 4 mm node post chemo was too small to be found on axillary us pre chemo? Either it was 4 mm before chemo and had no response to systemic therapy or was much larger prechemo and was just missed. Either suggests more aggressive disease than initially thought.

 

[gmsquid]

Leave it to breast for it to be the first time I ever officially replied to my own post: but just think this through with me.

If she was 65, presented at Tumor Board as a clinical T1N0M0, luminal A, we would all likely say it’s ok for her to omit RT if she is to get hormones.

If pathology confirms she has one node positive now we are saying it’s ok to treat. Now the volumes, techniques, dose, +\- a boost, RNI +\- IM’s. We have a 30 page discussion about it all but technically everything is ok to do.

If this lady never had her axilla addressed by the surgeon, she likely would have had a positive lymph node we were all ok with leaving completely alone because we didn’t know that she had disease in there and we would all be doing the patient a disservice by treating an elderly patient with stage I disease as discussed and preached by anyone who ever heard an ASTRO lecture. If she is to get hormonal therapy then the node may be of no issue but look at how one patient can make so many different pathways and decisions. Again from no treatment to treat it all with anything from a local regional recurrence is bad to lymphedema is bad and likely no OS benefit!

If she was 65 and pN0 yes, could consider avoiding rt (though I wouldn’t) You would omit nodal staging in a 65 year old woman with a 10 year life expectancy? If she is sentinel lymph node positive at surgery, one could argue for z11. Z11 was a trial of axillary dissection not of rt. Half the patients on the sln arm had micromets, people who could be observed based on the ibcsg study. Many patients did not receive breast only rt. It’s both a good and a bad trial. Part of the confusion with nodal management in breast cancer is over reading z11.

As for if the axilla had not been assessed, this argument is a stage migration argument. We treat people based on what we know. If we don’t look we will never know so we may have observed her. She would make stage I look worse when she’s actually a stage II.

 

[RadOncDoc21]

If she was 65 and pN0 yes, could consider avoiding rt (though I wouldn’t) You would omit nodal staging in a 65 year old woman with a 10 year life expectancy? If she is sentinel lymph node positive at surgery, one could argue for z11. Z11 was a trial of axillary dissection not of rt. Half the patients on the sln arm had micromets, people who could be observed based on the ibcsg study. Many patients did not receive breast only rt. It’s both a good and a bad trial. Part of the confusion with nodal management in breast cancer is over reading z11.

As for if the axilla had not been assessed, this argument is a stage migration argument. We treat people based on what we know. If we don’t look we will never know so we may have observed her. She would make stage I look worse when she’s actually a stage II.

I personally would have the discussion and would go over the Hughes data, Prime II etc that basically say it’s ok to hold RT for “elderly” women (65 was Prime II).

I would actually prefer to give RT to these patients but we are preached down upon from the Ivory tower to not treat these women and to instead offer them hormone therapy. My argument (rant) is to bring up the issues with how we go about determining in our minds what we “should” do. How do we go from one extreme (no RT) to full RNI when both are acceptable depending on what study you like the most.

 

[gmsquid]

Yes but those women were pN0 not cN0. And the counseling should include the competing risks of hormonal therapy which is often poorly complied with.

 

[RadOncDoc21]

Yes but those women were pN0 not cN0. And the counseling should include the competing risks of hormonal therapy which is often poorly complied with.

Not all patients on those trials underwent a SLNB.

 

[medgator]

If she was 65 and pN0 yes, could consider avoiding rt (though I wouldn’t) You would omit nodal staging in a 65 year old woman with a 10 year life expectancy? If she is sentinel lymph node positive at surgery, one could argue for z11. Z11 was a trial of axillary dissection not of rt. Half the patients on the sln arm had micromets, people who could be observed based on the ibcsg study. Many patients did not receive breast only rt. It’s both a good and a bad trial. Part of the confusion with nodal management in breast cancer is over reading z11.

As for if the axilla had not been assessed, this argument is a stage migration argument. We treat people based on what we know. If we don’t look we will never know so we may have observed her. She would make stage I look worse when she’s actually a stage II.

More than a few women prefer rt over ai and end up coming back after initial consult to get it after they permanently dc the ai

 

[drewdog1973]

Yes but those women were pN0 not cN0. And the counseling should include the competing risks of hormonal therapy which is often poorly complied with.

Not on CALGB. The surgeons were told not to do SLN Bx and I think close to 1/2 did nothing. Choosing Wisely suggests not to check the axilla in these patients.

 

[communitydoc13]

I would actually prefer to give RT to these patients but we are preached down upon from the Ivory tower to not treat these women and to instead offer them hormone therapy.

Hormone Therapy Linked to Cognitive Decline

Findings from a subgroup of the TAILORx trial on early breast cancer reveal that endocrine therapy can contribute to extended cognitive decline, whether or not women receive initial chemotherapy. The results suggest that women taking hormone therapy should be periodically assessed for worsening...

cancerdiscovery.aacrjournals.org

It is amazing how as a community, we underreport and under-educate patients regarding toxicity. (This goes for us as well. As a resident, I did not appreciate the typical amount of breast shrinkage and firmness experienced by many patients long term and thought risk of breast RT induced low grade pneumonitis was close to zero (still low but definitely not close to zero with RNI and taxanes). That being said, for the vast majority of women, RT is much better tolerated in terms of meaningful QOL impact than endocrine therapy. This analysis from TailorX was a rare confirmation of a clinical suspicion. Namely that endocrine therapy has a significant impact on the cognitive function of (particularly younger) women. I will have more sympathy for that high powered, 40 year old professional, who is lamenting that "it's just harder to do her job" a year after treatment. I used to chalk a lot of this up to psychosocial issues. Bad doctoring on my part.

It was so hammered into the historical rad onc's head that PMRT was only good in 4 or more LNs positive, it just seems "weird" to do RNI in zero LN positive cases.

This drives me crazy and I must credit my training program with not espousing this nonsense even years ago. Those original post-mastectomy XRT trials have been looked at every which way for low burden nodal patients, and in the context of those patient populations and those treatment options, the XRT benefit held down to a single node positive. I remember back in the day, the argument (a ridiculous one in retrospect) was that the mean number of nodes taken was low (~8) on these trials and was suboptimal therapeutically. Good luck even getting a breast surgeon to take 15+ nodes routinely now-a-days.

 

[scarbrtj]

First of all, thanks for responding so cogently. I appreciate it. But I have a few points...

We don’t routinely look at the axilla or imns with routine imaging

As much as you harp on academics diminishing the role of rt, you really want to remove it from breast cancer.

I have not used the verb "omit" once. I am simply talking about the style of RT, not the omission of it. Every time you want to irradiate, I do too. We'd just maybe do it differently. I know you keep coming back to the probability that her nodal disease was bigger pre-chemo, and I have no reason to doubt it. OTOH, she was cT1N0. There is an orgy of data that "routine axillary and IMN imaging" in cT1N0 average risk disease is fruitless and bootless and many other bad connotation adjectives. So in most ways that count, it doesn't matter if the nodal disease was bigger pre-chemo.

Since 2015 and MA.20 and EORTC 22922, there's been a great "swooshing" sound of rad oncs suddenly doing RNI even in cases like the OP's; a case where she's just two millimeters away from being pathologic stage I. Were we in a medically conservative, "p<0.005" world, here'd be the headlines from MA.20 and EORTC 22922: "RNI has no effect on any oncologic outcomes but is proven to increase lymphedema."

Also, why does rni reduce distant metastasis. It likely addresses that nodal disease that seeds distant mets.

As for if the axilla had not been assessed, this argument is a stage migration argument. We treat people based on what we know. If we don’t look we will never know so we may have observed her. She would make stage I look worse when she’s actually a stage II.

"Stage migration" arguments were supposed to have gone away in breast cancer for the most part. Z0011 could put one in a BIG stage migration argument. And yet it showed "never knowing" is just fine, just great.

More importantly though, your statement "nodal disease that seeds distant mets" is the biggie here. This is the crux, right? This type of thinking is 100-plus-years-old. It's Halstedian. It's disproven. Were it not disproven, we might still be advocating internal mammary nodal dissection (along with axillary dissection! gasp). Not to be too strident but it's the type of thinking that "...does a disservice to women with breast cancer and, moreover, repudiates science."

And yet, we in rad onc are still perseverating on controlling micromets in the IMNs in every woman with a 1cm breast primary and 0.5cm of cancer in a single sentinel node. Yes, I think that's a bit crazy if not delusions of grandeur.

Of course mortality has improved in breast cancer substantially. It comes down to many reasons.
Improved screening, WEAK DATA
better systemic therapy, DATA
better delivery of systemic therapy, DATA
better radiation, NO DATA
better delivery of radiation. NO DATA

 

[evilbooyaa]

Is the final conclusion that anyone with N+ at any given time (pre/post-chemo, surgery), we should just go ahead and give RNI (including IM’s)?

Did we simplify breast?

While this is very simplified, this is probably the most data driven way to practice breast RT in the current era, and I would not bat an eye if somebody did this for every single patient who met this clinical situation, as long as dose constraints could be met.

 

[scarbrtj]

"Is the final conclusion that anyone with N+ at any given time (pre/post-chemo, surgery), we should just go ahead and give RNI (including IM’s)?"
"While this is very simplified, this is probably the most data driven way to practice breast RT in the current era."

Why is it the "most" data driven way. The conclusion from EORTC 22922 was "regional nodal irradiation was beneficial to women with early-stage breast cancer"; the nodal status un-nuanced. And MA.20 concluded RNI was beneficial in "high-risk node-negative breast cancer." So focusing on N+.... is that the "most" data driven way? Because RNI is not really a firm SOC without those two studies.

 

[evilbooyaa]

Leave it to breast for it to be the first time I ever officially replied to my own post: but just think this through with me.

If she was 65, presented at Tumor Board as a clinical T1N0M0, luminal A, we would all likely say it’s ok for her to omit RT if she is to get hormones.

If pathology confirms she has one node positive now we are saying it’s ok to treat. Now the volumes, techniques, dose, +\- a boost, RNI +\- IM’s. We have a 30 page discussion about it all but technically everything is ok to do.

If this lady never had her axilla addressed by the surgeon, she likely would have had a positive lymph node we were all ok with leaving completely alone because we didn’t know that she had disease in there and we would all be doing the patient a disservice by treating an elderly patient with stage I disease as discussed and preached by anyone who ever heard an ASTRO lecture. If she is to get hormonal therapy then the node may be of no issue but look at how one patient can make so many different pathways and decisions. Again from no treatment to treat it all with anything from a local regional recurrence is bad to lymphedema is bad and likely no OS benefit!

We went from being able to sleep peacefully at night not caring anything about the radiation we just blew off initially and now the same radiation we discussed in significant detail on how important it is to go after the nodes but yet unsure if there is truly a significant benefit or not in this patient. All because the surgeon discovered a positive lymph node in someone we were all willing to leave alone in the first place.

This is why breast is the worst!

No. Multiple points below.

Per OP:
1) She is 65, so meets age criteria for PRIME-II
2) She is ER/PR- per OP, so PRIME-II/CALGB is immediately thrown in the trash.
3) She is Her2+ - this is more of a soft call but I do not routinely recommend omission of RT in Her2+ patients given unknown numbers of Her2+ pts in at least CALGB.

Per your post:
4) Even if she was as you describe, Prime-II mandated surgical axillary evaluation. So yes, she would have been discovered to be N+ on SLNB. I hate skipping SLNB in general, but I do not think it is even OK to do so in somebody < 70 years old.



In my practice, N1mi gets high tangents. Any N+ gets 3-field. 100% of the time that is my primary recommendation. People poo-poo 1-3+ LNs in terms of SCV but they are practicing OUTDATED medicine given the previous obsession with 4+ LNs based on the EBCTCG meta-analyses.


I have not seen ANY data suggesting that an elderly population that is identified as LN positive does not benefit from 3-field RT short of underpowered subset analyses. This, IMO should be a clinical trial question (enrolling >= 65yo with ER+ disease and is in part being investigated by MA.39 (TAILOR-RT). I'm sure they'll do pre-specified subset analyses and we can see if there's any interaction effect of age on the results (assuming that there is a benefit to SCV field in the entire MA.39 patient cohort).

 

[evilbooyaa]

"Is the final conclusion that anyone with N+ at any given time (pre/post-chemo, surgery), we should just go ahead and give RNI (including IM’s)?"
"While this is very simplified, this is probably the most data driven way to practice breast RT in the current era."

Why is it the "most" data driven way. The conclusion from EORTC 22922 was "regional nodal irradiation was beneficial to women with early-stage breast cancer"; the nodal status un-nuanced. And MA.20 concluded RNI was beneficial in "high-risk node-negative breast cancer." So focusing on N+.... is that the "most" data driven way? Because RNI is not really a firm SOC without those two studies.

EBCTCG meta-analyses.

While EORTC had 44% pN0, MA.20 had a whopping 9% of patients be pN0. I treat in America, so I pay more attention to MA.20 than EORTC.

 

[FrostyHammer]

No. Multiple points below.

Per OP:
1) She is 65, so meets age criteria for PRIME-II
2) She is ER/PR- per OP, so PRIME-II/CALGB is immediately thrown in the trash.
3) She is Her2+ - this is more of a soft call but I do not routinely recommend omission of RT in Her2+ patients given unknown numbers of Her2+ pts in at least CALGB.

Per your post:
4) Even if she was as you describe, Prime-II mandated surgical axillary evaluation. So yes, she would have been discovered to be N+ on SLNB. I hate skipping SLNB in general, but I do not think it is even OK to do so in somebody < 70 years old.



In my practice, N1mi gets high tangents. Any N+ gets 3-field. 100% of the time that is my primary recommendation. People poo-poo 1-3+ LNs in terms of SCV but they are practicing OUTDATED medicine given the previous obsession with 4+ LNs based on the EBCTCG meta-analyses.


I have not seen ANY data suggesting that an elderly population that is identified as LN positive does not benefit from 3-field RT short of underpowered subset analyses. This, IMO should be a clinical trial question (enrolling >= 65yo with ER+ disease and is in part being investigated by MA.39 (TAILOR-RT). I'm sure they'll do pre-specified subset analyses and we can see if there's any interaction effect of age on the results (assuming that there is a benefit to SCV field in the entire MA.39 patient cohort).

Isn't this concept what the SUPREMO trial is looking at?

 

[RadOncDoc21]

I know this isn’t going to win me any praises and may get me banned but for a disease site that is so simple (yes folks, breast cancer is simple), we sure as hell know how to complicate it! There I said it!

 

[scarbrtj]

EBCTCG meta-analyses.

While EORTC had 44% pN0, MA.20 had a whopping 9% of patients be pN0. I treat in America, so I pay more attention to MA.20 than EORTC.

And yet with that "whopping" 9%, that was (the Canadian!) MA.20's main conclusion: RNI is effective in "high risk" node negative. ("I don't do breast hypofractionation; that's Canadian data" - rad onc from 15 years ago.) We practice in America, but that EBCTCG you mention... is it American *cough*. I think I'll give 36/20 to the IMNs with orthovoltage. (I like the EBCTCG, don't get me wrong.)

 

[metallica81788]

 

[gmsquid]

First of all, thanks for responding so cogently. I appreciate it. But I have a few points...


I have not used the verb "omit" once. I am simply talking about the style of RT, not the omission of it. Every time you want to irradiate, I do too. We'd just maybe do it differently. I know you keep coming back to the probability that her nodal disease was bigger pre-chemo, and I have no reason to doubt it. OTOH, she was cT1N0. There is an orgy of data that "routine axillary and IMN imaging" in cT1N0 average risk disease is fruitless and bootless and many other bad connotation adjectives. So in most ways that count, it doesn't matter if the nodal disease was bigger pre-chemo.

Since 2015 and MA.20 and EORTC 22922, there's been a great "swooshing" sound of rad oncs suddenly doing RNI even in cases like the OP's; a case where she's just two millimeters away from being pathologic stage I. Were we in a medically conservative, "p<0.005" world, here'd be the headlines from MA.20 and EORTC 22922: "RNI has no effect on any oncologic outcomes but is proven to increase lymphedema."


"Stage migration" arguments were supposed to have gone away in breast cancer for the most part. Z0011 could put one in a BIG stage migration argument. And yet it showed "never knowing" is just fine, just great.

More importantly though, your statement "nodal disease that seeds distant mets" is the biggie here. This is the crux, right? This type of thinking is 100-plus-years-old. It's Halstedian. It's disproven. Were it not disproven, we might still be advocating internal mammary nodal dissection (along with axillary dissection! gasp). Not to be too strident but it's the type of thinking that "...does a disservice to women with breast cancer and, moreover, repudiates science."

And yet, we in rad onc are still perseverating on controlling micromets in the IMNs in every woman with a 1cm breast primary and 0.5cm of cancer in a single sentinel node. Yes, I think that's a bit crazy if not delusions of grandeur.

regarding stage migration and z11, Only 13% of patients were upstaged to n2 with alnd. It only speaks to surgery being not beneficial. There’s quite a bit of incidental axillary rt with tangents let alone with high tangents. Half the patients assessed in the sln arm had incidental axillary rt with high tangents. Jagsi JCO 2014. It’s a tough trial to evaluate when it’s possible 50% of the patients got axillary treatment. Thus, a substantial portion of these patients probably got there high axilla treated which would probably make up for the fact that they didn’t get nodal dissections.

That’s a false statement to say treatment of microscopic nodal disease with surgery equates to rt. Halsteds surgery was morbid. RT is not. Additionally, that study was done in an era of no systemic therapy. The likelihood of distant metastasis was extraordinary high. Additionally, when you do a dissection you don’t get every single node. You have to spare plexus and vessels. So there’s likely microscopic disease left behind in a substantial number of patients. Similar to HN.

Multiple (not just one) randomized trials of regional nodal radiation in the systemic therapy era have shown improvement in distant metastatic free survival. How does that happen if not for disease left in the nodes?

 

[gmsquid]

I know this isn’t going to win me any praises and may get me banned but for a disease site that is so simple (yes folks, breast cancer is simple), we sure as hell know how to complicate it! There I said it!

It’s because so many people have breast cancer risk stratification can exist. It’s a key example of personalized medicine

 

[RadOncDoc21]

It’s because so many people have breast cancer risk stratification can exist. It’s a key example of personalized medicine

Is this why every woman in the world has their very own personalized stage in the new staging system?

 

[scarbrtj]

regarding stage migration and z11, Only 13% of patients were upstaged to n2 with alnd. It only speaks to surgery being not beneficial. There’s quite a bit of incidental axillary rt with tangents let alone with high tangents. Half the patients assessed in the sln arm had incidental axillary rt with high tangents. Jagsi JCO 2014. It’s a tough trial to evaluate when it’s possible 50% of the patients got axillary treatment. Thus, a substantial portion of these patients probably got there high axilla treated which would probably make up for the fact that they didn’t get nodal dissections.

I have read this Jagsi paper about, IDK, 50 times in my life. The "high tangent rate" was balanced between arms. But it (their analysis of a kind of small proportion of all Z0011 patients) is not the most enlightening thing in breast cancer. That said (and speaking of omitting RT), here's a quote from Jagsi et al:

Finally, it is critical to recognize that our observations should not be taken to suggest that the nodal radiation administered to patients in Z0011 was necessary or beneficial. Although we found that a nontrivial minority of patients received more extensive RT than prescribed by the protocol, it is important to note that a subgroup received no RT at all. This is noteworthy not because these protocol violations offset each other but rather because these findings together suggest that future studies are needed to determine whether certain patients might safely avoid RT in these circumstances as well as whether others might benefit from more extensive treatment.

So to "There’s quite a bit of incidental axillary rt with tangents let alone with high tangents [in Z0011]," I say there's quite a bit of patients in Z0011 who got no RT at all. Restated: quite a bit of N+ patients who got no RT at all. And there were no differences in LRR or OS.

That’s a false statement to say treatment of microscopic nodal disease with surgery equates to rt. Halsteds surgery was morbid. RT is not. Additionally, that study was done in an era of no systemic therapy. The likelihood of distant metastasis was extraordinary high. Additionally, when you do a dissection you don’t get every single node. You have to spare plexus and vessels. So there’s likely microscopic disease left behind in a substantial number of patients. Similar to HN.

Is it false? Do we see a lot of desire to irradiate "fully dissected" positive axillae? Isn't there a "industry standard" of not irradiating dissected axillae? Isn't surgery/RT equivalence for microscopic disease a conclusion of AMAROS? There are about 30 LNs in the average woman's armpit. The average dissection thus obviously never gets all the LNs out of there (thankfully). If there's microscopic disease left behind after axillary surgery that somehow RT-sans-surgery mops up in a more superior fashion, there's zero data I'm aware of that supports that theory. HN is as far from breast biology as I am from the North Pole.

 

[gmsquid]

I have read this Jagsi paper about, IDK, 50 times in my life. The "high tangent rate" was balanced between arms. But it (their analysis of a kind of small proportion of all Z0011 patients) is not the most enlightening thing in breast cancer. That said (and speaking of omitting RT), here's a quote from Jagsi et al:

Finally, it is critical to recognize that our observations should not be taken to suggest that the nodal radiation administered to patients in Z0011 was necessary or beneficial. Although we found that a nontrivial minority of patients received more extensive RT than prescribed by the protocol, it is important to note that a subgroup received no RT at all. This is noteworthy not because these protocol violations offset each other but rather because these findings together suggest that future studies are needed to determine whether certain patients might safely avoid RT in these circumstances as well as whether others might benefit from more extensive treatment.

That's a fair point to quote this.. So what she's saying here is that this trial said nothing about whether there was a benefit of radiation to the axilla. You are using a trial of surgery in a low-risk population (50% micrometastasis powered for an overall survival benefit (by the way based on a 5 year OS assumption of 80% for ALND versus non-inferiority at 75% for SLN) of the addition of an axillary nodal dissection at 5 years to avoid giving radiation. This population was so low risk the survival analysis power calculation underestimated the actual survival on the trial by 10%. Additionally, this trial had no standard radiation across arms and clear protocol violations in 50% of the patients. So this trial says nothing about radiation in this population. Unlike MA.20 and EORTC which are both direct randomized studies of the benefits of RNI.

Is it false? Do we see a lot of desire to irradiate "fully dissected" positive axillae? Isn't there a "industry standard" of not irradiating dissected axillae? Isn't surgery/RT equivalence for microscopic disease a conclusion of AMAROS? There are about 30 LNs in the average woman's armpit. The average dissection thus obviously never gets all the LNs out of there (thankfully). If there's microscopic disease left behind after axillary surgery that somehow RT-sans-surgery mops up in a more superior fashion, there's zero data I'm aware of that supports that theory. HN is as far from breast biology as I am from the North Pole.

Again, AMAROS was a trial of what? Non-inferiority of RT compared to ALND for the endpoint of 5-year axillary recurrence in clinically node negative sentinel lymph node positive breast cancer. Isolated axillary recurrence in both arms was 1% or less. (Similar to the risk of axillary recurrence on Z11 where no dissection was done) 75% of patients had 1 positive SLN. A low risk group. RT was also allowed in the ALND arm if 4 positive lymph nodes were found, so the higher risk patients who were found due to ALND (at least 10% of the population got both surgery and RT) got both surgery and RT, which were never found in the RT alone group. Finally, the highest risk of local recurrence in post mastectomy studies is in the chest wall > SCV > axilla/IMNs. Isolated axillary recurrence is low throughout. Is that due to underdiagnosis of axillary recurrence or changing drainage patterns after surgery.

So neither of these studies has anything to do with the reduction in distant metastasis with RNI: the first was a study of surgery, and the second was a study of axillary recurrence comparing RT and surgery in regards to that risk factor.

You have still not addressed the fact that both MA.20 and EORTC showed a DMFS benefit to the addition of RNI in breast cancer. Two large independent randomized trials. It's okay if you don't have a mechanism, nobody knows. But, its reasonable to presume that if RT's benefit is locoregional and it reduces distant metastasis, it's because reducing microscopic nodal disease reduces distant recurrence. Or you just have statistical fluke.

Finally, the benefits in MA.20 in 10-year DFS by number of positive LNs (10% of patients had 0 LNs positive, ~50% had 1, ~25% had 2 and ~11% had 3, 5% had > 3):
10-year DFS
0 nodes: 83.7% vs 72.4%
1 node: 83.5% vs 80.9%
2 nodes: 74.8% vs 67.6%
3 nodes: 69.8% vs 60.3%

The stem and leafs all cross 1 but the studies weren't powered for subgroup analysses. So maybe you don't consider the 3.5% DFS benefit with a true single lymph node positive (96% of patients had an ALND on MA.20). But a potential 7% benefit in patients with 2-3 nodes, and a near 10% benefit in patients with more than 3 nodes? Pretty impressive. And MA.20 was largely a study of relatively low-risk patients in the anthracycline era, so you can't argue that systemic therapy was poor.

Now how do you look at that in patients with a SLN positive. If the SLN is small what's the probability of a second undetected node? What if the SLN is large? How do you tell? How does that drive your decision to do RNI? Probably the only way you can do, is to use nomograms which only exist in the non-neoadjuvant setting. (i.e. Breast Cancer Nomogram to Predict Additional Positive Non-SLN, without Neoadjuvant Chemotherapy and Breast Cancer Nomogram: Breast Additional Non SLN Metastases | Memorial Sloan Kettering Cancer Center)

So if there's a 30% probabilty of additional positive nodes do you recommend an ALND and then radiate if positive? Or do you radiate without ALND?

Ultimately after neoadjuvant chemotherapy, all is jumbled. So in this patient who at least has a 0.4 cm positive node after chemotherapy, what would be the probability that she had additional SLNs if it was 0.4 cm before neoadjuvant chemotherapy with a 1.3 cm primary? (according to MDACC nomogram 27%, according to MSKCC 34% with routine H&E). We don't have studies after neoadjuvant chemotherapy. So I think there's a high likelihood of additional nodes because this node was likely bigger than 0.4 cm before. Since there's no difference between neoadjuvant and adjuvant systemic therapy, and there's a modestly high likelihood of additional nodes in this woman, and there's 7% DFS benefit to RNI in 2 positive axillary nodes according to MA.20 for RNI, I would radiate her regional nodes. Let alone the fact that she didn't pCR.

 

[scarbrtj]

This [Z0011] population was so low risk

Only recently could we call a population with 100% LN positivity "so low risk"

Finally, the highest risk of local recurrence in post mastectomy studies is in the chest wall > SCV > axilla/IMNs

a great argument to consider s'clav RT alone when doing "RNI"
EDIT:
To expound, risks of recurrences in MA.20 (not PMRT obv) e.g.:
distant>any other cancer occurrence>local recurrence>"ipsi/contra breast second cancer">s'clav>axillary>IMN ... and re: IMNs, 0.8% (16/2002) recurrence without IMN RT and 0.2% (4/2002) recurrence with IMN RT. Which means you need to treat roughly 170 IMN stations in 170 patients to prevent one recurrence there.

Isolated axillary recurrence is low throughout. Is that due to underdiagnosis of axillary recurrence or changing drainage patterns after surgery.

how about option #3... the axilla is almost entirely prognostic, "treating" it is not therapeutic, and positive axillae just rarely manifest breast cancer recurrences in the axillae

You have still not addressed the fact that both MA.20 and EORTC showed a DMFS benefit to the addition of RNI in breast cancer.

I addressed it. DMFS and a dollar will get you a cup of coffee, OS-wise. You haven't addressed the fact that the DMFS benefit in EORTC was significant in 2015 (p=0.02), but with five more years of follow-up it lost significance (p=0.18). Puts a dent in the nodes-seed-metastases theory if weren't dented already. And there's not been an update of MA.20's outcomes since 2015. This harkens back to my point about conservatism, reproducibility, etc.

 

[drewdog1973]

Only recently could we call a population with 100% LN positivity "so low risk"


a great argument to consider s'clav RT alone when doing "RNI"
EDIT:
To expound, risks of recurrences in MA.20 (not PMRT obv) e.g.:
distant>any other cancer occurrence>local recurrence>"ipsi/contra breast second cancer">s'clav>axillary>IMN ... and re: IMNs, 0.8% (16/2002) recurrence without IMN RT and 0.2% (4/2002) recurrence with IMN RT. Which means you need to treat roughly 170 IMN stations in 170 patients to prevent one recurrence there.

how about option #3... the axilla is almost entirely prognostic, "treating" it is not therapeutic, and positive axillae just rarely manifest breast cancer recurrences in the axillae

I addressed it. DMFS and a dollar will get you a cup of coffee, OS-wise. You haven't addressed the fact that the DMFS benefit in EORTC was significant in 2015 (p=0.02), but with five more years of follow-up it lost significance (p=0.18). Puts a dent in the nodes-seed-metastases theory if weren't dented already. And there's not been an update of MA.20's outcomes since 2015. This harkens back to my point about conservatism, reproducibility, etc.

Well written
PM me about that $1 coffee. What a steal!

 

[gmsquid]

Only recently could we call a population with 100% LN positivity "so low risk"


a great argument to consider s'clav RT alone when doing "RNI"

how about option #3... the axilla is almost entirely prognostic, "treating" it is not therapeutic, and positive axillae just rarely manifest breast cancer recurrences in the axillae

I addressed it. DMFS and a dollar will get you a cup of coffee, OS-wise. You haven't addressed the fact that the DMFS benefit in EORTC was significant in 2015 (p=0.02), but with five more years of follow-up it lost significance (p=0.18). Puts a dent in the nodes-seed-metastases theory if weren't dented already. And there's not been an update of MA.20's outcomes since 2015. This harkens back to my point about conservatism, reproducibility, etc.

The only trials that have shown the DFS benefit are trials of RNI. You may or may not include the axilla, but the incidental axillary dose with all these trials (even EORTC) is substantial. You can remove the axillary nodal RT if you would like, though it snearly impossible. Of course the axilla is prognostic. It is also the single best predictor of SCV and IMN involvement (Breast cancer subpopulation with high risk of internal mammary lymph nodes metastasis: analysis of 2,269 Chinese breast cancer patients treated with extended radical mastectomy - PubMed), which are addressed with RNI.

I did address the DMFS aspect. DMFS includes distant metastasis and survival. An increasing number of patients dying with 5 years of follow-up from non breast-cancer related causes. Everyone dies eventually. Survival curves always meet at the end of time. Breast cancer related mortality was reduced by 4% on EORTC which is basically the DFS benefit of the earlier study. Further, EORTC also included a largely node negative population, who presumably are less likely to benefit from RNI.

 

[scarbrtj]

Well written
PM me about that $1 coffee. What a steal!

at a HS football game. maybe

 

[scarbrtj]

The only trials that have shown the DFS benefit are trials of RNI.

EORTC also included a largely node negative population, who presumably are less likely to benefit from RNI.

The DFS benefit was also lost with the recent EORTC update.

EORTC 2015: we love thee
EORTC 2020: MA.20 who dis

 

[gmsquid]

The DFS benefit was also lost with the recent EORTC update.

EORTC 2015: we love thee
EORTC 2020: MA.20 who dis

Again dfs includes survival events. Death from other causes

 

[gmsquid]

Scarbrtj before this thread: nodal rt doesnt matter
Scarbrtj after this thread: touché salesman

 

[RO5231]

Again dfs includes survival events. Death from other causes

I think that partly explains the loss of DFS benefit at 15 years in EORTC. In EORTC, DFS was defined as time to local recurrence, regional recurrence, distant metastasis, second breast cancer, or death.

In MA.20, DFS was defined as time to first recurrence in the ipsilateral breast or in nodal or distant sites, a contralateral breast cancer, or death from breast cancer.

 

[gmsquid]

I think that partly explains the loss of DFS benefit at 15 years in EORTC. In EORTC, DFS was defined as time to local recurrence, regional recurrence, distant metastasis, second breast cancer, or death.

In MA.20, DFS was defined as time to first recurrence in the ipsilateral breast or in nodal or distant sites, a contralateral breast cancer, or death from breast cancer.

And that’s why analyses are done at a particular time point. They are powered for that time point.

 

[medgator]

Well written
PM me about that $1 coffee. What a steal!

That's pretty much the price at mickey Ds these days

 

[scarbrtj]

Scarbrtj before this thread: nodal rt doesnt matter
Scarbrtj after this thread: touché salesman

Ha I had to google touche salesman. But RNI doesn't matter (depending on how RNI is defined). How doesn't it matter? It doesn't matter for survival or affecting the overall natural history of breast cancer. How does it matter? It has effects (NNTs>=~25) on things like DMFS, DFS; although the data on these metrics exhibit non-reproducibility between trials. The number needed to harm (NNH) is equivalent to the aforementioned NNTs. The NNH RNI effects don't show mortal harms as long as we ignore the higher risk of second cancers seen in one nodal RT trial and make the assumption that the nodal RT was not associated with the second cancer risks. If the higher second cancer risks are ignored, the NNH RNI effects aren't life-threatening—but neither is the RNI itself life-saving.

If what I've stated in all the above is true about RNI treatment, it's been the usual course of events in onco-history that such a treatment would be on the controversial side vs the straight-ahead standard of care side. This has not been the case in breast RNI. I don't know if I'm getting the meme right (maybe it's a trap), but this irony as it were over breast RNI probably oft leads me to being touche salesman guy.

 

[drewdog1973]

Your premise, Scarb, is that the harm is as high as past data suggests. I think with a 5% DFS, with modern techniques, there is going to be an improvement in survival. That’s speculative, but the Danish study gives us an idea, as the right sided tumors appeared to benefit. It’s all iterative - a few percent here, a few percent there, and suddenly we show some benefits. I think a lot of this has to do with technique. We “know” that there is local/regional control benefit, and this isn’t translating into survival. So, if the technique improves and cardio toxicity is mitigated, then why wouldn’t it benefit patients?

 

[scarbrtj]

Your premise, Scarb, is that the harm is as high as past data suggests. I think with a 5% DFS, with modern techniques, there is going to be an improvement in survival. That’s speculative, but the Danish study gives us an idea, as the right sided tumors appeared to benefit. It’s all iterative - a few percent here, a few percent there, and suddenly we show some benefits. I think a lot of this has to do with technique. We “know” that there is local/regional control benefit, and this isn’t translating into survival. So, if the technique improves and cardio toxicity is mitigated, then why wouldn’t it benefit patients?

Because the systemics are outpacing the "few percent here, few percent there" the RNI RT might bring to the table. And I haven't seen (surprisingly) a significant cardiotoxicity signal from MA.20 or EORTC 22922. And because the Fisherian hypothesis prevents minor variations in LR treatments from affecting overall survival... because, for example, the OP's patient has a ~30-40% chance of microscopic marrow positivity (were this to be checked).

And, I will believe everything you say re: technique... once there's good data to show it.

 

[gmsquid]

Because the systemics are outpacing the "few percent here, few percent there" the RNI RT might bring to the table. And I haven't seen (surprisingly) a significant cardiotoxicity signal from MA.20 or EORTC 22922. And because the Fisherian hypothesis prevents minor variations in LR treatments from affecting overall survival... because, for example, the OP's patient has a ~30-40% chance of microscopic marrow positivity (were this to be checked).

And, I will believe everything you say re: technique... once there's good data to show it.

What systemics in the adjuvant setting? Other than HER-2+ disease (truly game changed by transtuzumab and pertuzumab), standard of care remains AC +/- T for breast cancer, what it was on MA.20 and EORTC.

Let's consider the metastatic setting:

Hormone positive disease/Luminal A disease:

Standard of care is palbociclib + AI which improved OS by 6 months over AI alone. This disease has a long natural history. Most women don't develop metastatic disease until at least 3-5 years after adjuvant therapy if not longer. The disease is usually hormone responsive at the time of relapse and patients do okay for a few years and there are no good second line options.

Triple negative disease:

No advances in the recurrent setting. Limited options other than systemic chemo, though immuno may have some promise for a small percentage of good responders (as it has in other diseases). Also, the HR for benefit of RNI in the ER- population on MA.20 was about .55 (15% improvement in 10 year DFS)

HER-2 disease:

Trastuzumab-pertuzumab-chemo was definitely an impressive improvement.

Still patients with metastatic disease are incurable. No matter the number of systemic therapies you give most solid tumors you can't clear them forever. Even if they have a path CR tumors aren't curable. They will recur.

Only 20-25% of patients developed recurrent breast cancer in the population of the EORTC study at 15 years (which btw was 45% node negative). Its hard to show survival benefits in the adjuvant setting unless you are treating patients with more aggressive disease (node positive). Additionally, most patients on MA.20 had luminal A disease and had 5 yeras of hormones. Those cancers come back latera. Like I said earlier the addition of paclitaxel only showed an OS benefit in a completely node positive patient population with baseline high risk in a trial of 3000 patients (as evidenced by the low DFS of both arms in the CALGB trial). MA.20 and EORTC were half that size.

These were favorable populations on EORTC and MA.20 and there was a significant DMFS benefit. 60% on MA.20 had 0-1 positive nodes after ALND. See my diagram earlier. So like I said, maybe in the single node positive population after ALND with luminal A disease a 3% DFS benefit may not be worth it. How do you deal with that in patients who don't have ALNDs and have 1/1 positive SLNs? Or clincially positive nodes for that matter.

 

[scarbrtj]

Still patients with metastatic disease are incurable.

Since ~30% of cT1N0 patients have detectable disease in their marrow, why is the cure rate in Stage I not 70%... or less.

What systemics in the adjuvant setting? Other than HER-2+ disease (truly game changed by transtuzumab and pertuzumab), standard of care remains AC +/- T for breast cancer, what it was on MA.20 and EORTC.

I don't know; as I said I'm no med onc. All I know is survival has improved significantly in breast cancer in the last 2 decades. I would have a hard time arguing that rad onc (or surgery) has had an impact on this improvement. RNI hasn't.

Most women don't develop metastatic disease until at least 3-5 years after adjuvant therapy if not longer

Technically/on paper, this is not prevailingly true in breast cancer. Forty years ago Fisher wrote "Breast cancer is a systemic disease at diagnosis. The metastatic phenotype is either present or absent but is not acquired over time." I realize this notion doesn't "sit well" with rad oncs. Were it really to have obtained purchase in rad oncs' noggins, trials like MA.20 and EORTC 22922, and Jagsi et al's analysis of Z0011, wouldn't have happened.

These were favorable populations on EORTC and MA.20 and there was a significant DMFS benefit. 60% on MA.20 had 0-1 positive nodes after ALND.

Also, the HR for benefit of RNI in the ER- population on MA.20 was about .55 (15% improvement in 10 year DFS)

There's not a significant DMFS benefit in EORTC 22922. And in MA.20, the smallest HR was for N0 patients (0.55), not ER- patients (0.56); N0 patients got a strikingly (graphically) better benefit from RNI than N1 patients. However, the CI crossed over 1.0 in all nodal subsets (N0, N1, N2, and N3). But for ER-, the CI was 0.39-0.81.

If we were to discuss the need for RNI on the basis of molecular features, versus "Is she node positive or not," I might be inclined to buy into that line of reasoning a little more. As such, we have one randomized trial in all of rad onc to show there's a DMFS benefit to RNI and that's MA.20. And in that trial, the benefit of RNI was irrespective of nodal status (p=0.65). So why does everyone and their brother keep making RNI decisions essentially purely on the basis of was the axilla positive or not when it's not really well data-supported. (The EBCTCG included trials where the only systemic therapies were anti-estrogens, or ovarian radiotherapy... no taxanes... not really applicable in 2021.) Happy New Year btw.

 

[gmsquid]

Since ~30% of cT1N0 patients have detectable disease in their marrow, why is the cure rate in Stage I not 70%... or less.



I don't know; as I said I'm no med onc. All I know is survival has improved significantly in breast cancer in the last 2 decades. I would have a hard time arguing that rad onc (or surgery) has had an impact on this improvement. RNI hasn't.

Technically/on paper, this is not prevailingly true in breast cancer. Forty years ago Fisher wrote "Breast cancer is a systemic disease at diagnosis. The metastatic phenotype is either present or absent but is not acquired over time." I realize this notion doesn't "sit well" with rad oncs. Were it really to have obtained purchase in rad oncs' noggins, trials like MA.20 and EORTC 22922, and Jagsi et al's analysis of Z0011, wouldn't have happened.


There's not a significant DMFS benefit in EORTC 22922. And in MA.20, the smallest HR was for N0 patients (0.55), not ER- patients (0.56); N0 patients got a strikingly (graphically) better benefit from RNI than N1 patients. However, the CI crossed over 1.0 in all nodal subsets (N0, N1, N2, and N3). But for ER-, the CI was 0.39-0.81.

If we were to discuss the need for RNI on the basis of molecular features, versus "Is she node positive or not," I might be inclined to buy into that line of reasoning a little more. As such, we have one randomized trial in all of rad onc to show there's a DMFS benefit to RNI and that's MA.20. And in that trial, the benefit of RNI was irrespective of nodal status (p=0.65). So why does everyone and their brother keep making RNI decisions essentially purely on the basis of was the axilla positive or not when it's not really well data-supported. (The EBCTCG included trials where the only systemic therapies were anti-estrogens, or ovarian radiotherapy... no taxanes... not really applicable in 2021.) Happy New Year btw.

I have to say you are thoroughly misquoting data. That’s a 30% micromet rate in a pooled analysis of Stage I-Iii patients. Not cT1N0. And yes there is a subset of pT1N0 patients who develop distant Mets without nodal Mets first. About 22%. Doesn’t mean that all patients do that without nodal. In fact that happens more frequently in triple negative and her2 positive breast cancer. Which is why even pT1N0 patients get systemic therapy unless their disease is sufficiently small. Additionally, a subset of those pT1N0 patients also have sub clinical imn disease not assessed with an axillary surgery.


quoting Bernie fisher from 40 years ago when most breast cancer was diagnosed clinically without presurgical nodal staging or screening mammography is disingenuous. Expert opinion is a low form of evidence. And our understanding of cancer has changed since then. It’s also disingenuous to discredit the survival benefit of post mastectomy rt on three studies in the immediate post fisher era. Danish 82b, 82c and Ontario.


also, even in the modern era of other malignancies we frequently are treating patients with oligometastatic disease with long term survival using sbrt. So even though these patients had metastatic disease, they may survive a long time.

finally, I’ve addressed the lack of a dmfs benefit on long term follow up of eortc multiple times on this thread yet you continue to bring it up. More non breast cancer events. That’s why breast cancer mortality was still lower in the rni group on long term follow up. The study was powered at the initial publication date. It wasn’t powered to see a difference beyond that time point. It’s important to power a study at a particular time point because that drives the number of patients you enroll. Beyond that time point you don’t take into account the event frequency and the important competing risks!

 

[gmsquid]

Also, with better detection there is both lead time bias and stage migration. More mammographicaly detected cancers that are smaller causes stage migration and lead time bias. Thus, you tend to find more stage I tumors before they met. Secondly lymph node dissection is better. SLNs detect micromets and small nodes all the time. That also leads to stage migration.

chemo hasn’t changed, but imaging has gotten better tomosynthesis as well as sln technique.

You can’t argue that lead time bias and stage migration can’t have that much of an impact on survival of stage I patients .

What Cancer Screening Statistics Really Tell Us

Cancer screening studies have shown that more screening does not necessarily translate into fewer cancer deaths. This article explains how to interpret the statistics used to describe the results of screening studies.

www.cancer.gov

 

[scarbrtj]

I have to say you are thoroughly misquoting data. That’s a 30% micromet rate in a pooled analysis of Stage I-Iii patients. Not cT1N0.

I said ~30% for cT1N0, which the OP patient was. More precisely, the OP patient was cT1 (25.2% marrow positivity rate), cN0, ER- (34.5% marrow positivity rate), and pN1 (30.0% marrow positivity rate). And she was Her2+ which also increases the marrow positivity rate. "Thoroughly misquoting"? I could easily be more thorough. Or less thorough. IDK at this point. Would you give RNI to a patient w/ a positive marrow?

Additionally, a subset of those pT1N0 patients also have sub clinical imn disease not assessed with an axillary surgery.

So why no IMN RT in pT1N0, as per this logic it would improve DMFS and DFS.

quoting Bernie fisher from 40 years ago when most breast cancer was diagnosed clinically without presurgical nodal staging or screening mammography is disingenuous. Expert opinion is a low form of evidence.

Well, I mean, the Fisherian hypothesis (aka the Alternative Hypothesis) is the current working hypothesis in breast cancer. Saying Fisher is "expert opinion" in breast cancer is like saying Newton is expert opinion in physics. (Don't touche salesman me on that.) In all discussions on breast cancer you have to start with the hypothesis "nodal disease DOES NOT seed distant mets." To say that nodal disease seeds distant mets is Halstedian and is quoting ideas from more than 100 years ago... empiric ideas that Fisher and the NSABP helped disprove.

finally, I’ve addressed the lack of a dmfs benefit on long term follow up of eortc multiple times on this thread yet you continue to bring it up. More non breast cancer events. That’s why breast cancer mortality was still lower in the rni group on long term follow up. The study was powered at the initial publication date. It wasn’t powered to see a difference beyond that time point. It’s important to power a study at a particular time point because that drives the number of patients you enroll. Beyond that time point you don’t take into account the event frequency and the important competing risks!

I think I understand what you're saying here. See if I do... if a study looking at failure event data, with censoring, is positive at 10-year followup, but becomes negative with 15-year followup, we should still believe the 10-year data and call the study positive because "studies are powered at the initial publication date," (who knew power was linked with time of publication) and power doesn't stretch beyond that time point. What if a study is positive at 10 years and still positive at 15 years? Do we still call that negative (or the 15y publication superfluous) because it's past its initial publication date? Are 10 and 15 and 20 year updates of studies, which we see all the time, a total waste of time? Should we even read these studies? Alert the biostatisticians.

 

[gmsquid]

I said ~30% for cT1N0, which the OP patient was. More precisely, the OP patient was cT1 (25.2% marrow positivity rate), cN0, ER- (34.5% marrow positivity rate), and pN1 (30.0% marrow positivity rate). And she was Her2+ which also increases the marrow positivity rate. "Thoroughly misquoting"? I could easily be more thorough. Or less thorough. IDK at this point. Would you give RNI to a patient w/ a positive marrow?

First of all, cT1N0 is not pN0. She may have a higher marrow positivity rate, but she also has a higher additional LN rate. Probably on the order of 30+% according to the MSKCC nomograms. Now suppose she has a 50% chance of having distant mets at the time of diagnosis. So in 15% of cases RNI would not benefit her but in the other 15% it may reduce her risk of locoregional recurrence by lets say 1/2-3/4th. And if we use metanalysis data, she may derive a 5% improvement in OS with RNI. It's risk stratification. Now if you wanted to confirm she would benefit you could send her for a bone marrow biopsy, but that's your call.

So why no IMN RT in pT1N0, as per this logic it would improve DMFS and DFS.

It depends on the risk. Of course RT will reduce the risk of locoregional recurrence and distant metastasis. But in somebody who is pT1N0, that risk is small so to detect an improvement in DMFS or DFS it would take ten thousand or more patients probably. (That's why you power a study). It also means your NNT would be 100+. And as you alluded to earlier the NNH would probably outweigh. ITs why not everybody gets chemo up front either. This is all about pre-test and post-test probabilities.

Well, I mean, the Fisherian hypothesis (aka the Alternative Hypothesis) is the current working hypothesis in breast cancer. Saying Fisher is "expert opinion" in breast cancer is like saying Newton is expert opinion in physics. (Don't touche salesman me on that.) In all discussions on breast cancer you have to start with the hypothesis "nodal disease DOES NOT seed distant mets." To say that nodal disease seeds distant mets is Halstedian and is quoting ideas from more than 100 years ago... empiric ideas that Fisher and the NSABP helped disprove.

In a substantial number of cases nodal disease does not seed distant metastasis. Absolutely correct. But 5 trials ahve shown that PMRT or RNI imrpoves breast cancer OS and/or DMFS. How many breast cancer patietns die of locoregional disease. So that means RNI is doing something. 5 trials have shown it is on the order of > 5% absolute benefit at 10 years for DMFS in a low-risk patient population.

I think I understand what you're saying here. See if I do... if a study looking at failure event data, with censoring, is positive at 10-year followup, but becomes negative with 15-year followup, we should still believe the 10-year data and call the study positive because "studies are powered at the initial publication date," (who knew power was linked with time of publication) and power doesn't stretch beyond that time point. What if a study is positive at 10 years and still positive at 15 years? Do we still call that negative (or the 15y publication superfluous) because it's past its initial publication date? Are 10 and 15 and 20 year updates of studies, which we see all the time, a total waste of time? Should we even read these studies? Alert the biostatisticians.

Almost exactly what I'm saying. Suppose the following situation. Let's assume you powered a study to show a 10% risk reduction for a particular outcome at 5 years. You assumed the risk of that otucome without intervention was 50%. So you want to show that it improves to 45%. And for ease lets say that's a 200 patient study (each arm). Now you conduct th4e study and at 5 years you find that in the control arm the risk of the outcome was 80% without intervention and 82 with intervention. No way is a 2 patient differnece going to be statistically significant but the 10% risk reduction is there. This is probably why CALGB was postiive for paclitaxel and NSABP was not. This is also why if you enroll a low risk population on study or survival improves for other reasons (such as lead time bias), an intervention may be successful at the rate you expect, but the study will still be negative.

Suppose the second situation. Let's take a patient population of 75 year old Females with some disease. We want to see if an intervention reduces a particular outcome X or death at 5 years. Let's take the same numbers 50% risk of death or X and 100 patients each arm. The 5 year study shows that there is a benefit. However in 10 years 80% of the 75 year olds are deadin both arms. There will be no benefit at 10 years.

Powering a study has to be taken into understanding the outcomes of a trial.

 

[gmsquid]

And also, 15% of patients with pCR to her2 directed therapy have died or recurred at 5 years albeit in a higher risk population than this patient (had to be at least T2). It’s still bad disease relative to hormone positive disease: 5-year analysis of neoadjuvant pertuzumab and trastuzumab in patients with locally advanced, inflammatory, or early-stage HER2-positive breast cancer (NeoSphere): a multicentre, open-label, phase 2 randomised trial - PubMed