Breast micromets

[Neuronix]

I've been going back and forth with the breast surgeons and attendings in my own department about cases like this that have been coming up a lot lately. Here's a hypothetical case:

56 year old woman with early breast cancer elects for mastectomy w/ sentinel lymph node procedure. Final pathology demonstrates a single area of G2 IDC 2.1cm in the UOQ, ER+, PR+, HER2-, widely negative margins, no LVI, no EIC. Immediate reconstruction performed and expander placed.

SLNBx final path shows 1/2 axillary LNs with micromet 1.0mm in size. No ALND performed.


Would you radiate this patient? Surgery is pushing not to radiate due to the risk of capsular contraction and impaired cosmesis. They do not want to do ALND, citing Z11. If no radiation, assuming the patient still has T1-2N1mi(sn), are there risk factors that would tip you over to radiating?

I have done my own literature search and come to my own conclusions, but I'm curious what others think before I give my (resident level) thoughts.

---

Members don't see this ad.

 

[Palex80]

I probably wouldn't. If the patient is very keen to have maximum treatment, you can do it.

I would do it, if it was a macrometastasis though.

 

[deleted4401]

So, can't really cite Z11 for one obvious reason - all those patients got radiation. I hate it when they don't do a dissection based on this. Think we've discussed these cases in the past on SDN.

Anyway, I do MSKCC nomogram to predict risk of any more positive lymph nodes and Katz to predict 4 or more. Katz says <0.5% risk of having 4 or more, in this case. Can't plug this case into MSKCC b/c not sure if frozen performed or whether it was routine H&E or serial. Using this info, I try to make up a logical game plane. If the risk of additional lymph nodes are very high, like >30%, then I'd recommend treatment, if 10-20%, I'd say go ahead and have a discussion about pros and cons of treatment, leaning towards treating. If it's lower, I think I can sleep at night without treating. You didn't mention laterality, I think I'd be a little more wary with left side, but wouldn't sway me a huge amount.

My threshold is getting lower and lower, especially in healthy patients with MA.20 and in ER+ patients, the benefit of additional LR control may lead to a much better improvement in survival (More like 2:1, rather than 4:1, based on JCO analysis of Danish results).

What other approach are people using?

 

[Gfunk6]

I assume that she is going to get chemotherapy, is that the case?

 

[deleted4401]

NCCN would recommend OncoType for these folks.

 

[Neuronix]

NCCN would recommend OncoType for these folks.

That's what we would do, yes. Does whether the patient receives chemotherapy or oncotype score change your decision?

Let's say for sake of argument that this was detected on routine H&E, frozen was performed. Nomogram risk on this hypothetical patient is 13%. In their defense some of the surgeons will cite Z11/B04, some will cite the MSKCC nomogram risk, though we have no good guidelines as to what percentage risk is acceptable to us.

 

[medgator]

So, can't really cite Z11 for one obvious reason - all those patients got radiation.

Bingo

 

[medgator]

My threshold is getting lower and lower, especially in healthy patients with MA.20 and in ER+ patients, the benefit of additional LR control may lead to a much better improvement in survival (More like 2:1, rather than 4:1, based on JCO analysis of Danish results).

What other approach are people using?

The combined danish analysis is compelling especially when we're talking about an OS benefit here.

Plus there is this study out in Cancer that's been all over the news lately (I know it's retrospective but it definitely is thought provoking). Plus we have these small studies suggesting a benefit for mastectomy in T1/2 triple-neg pts (not this scenario, but showing that simply using traditional T/N cutoffs just isn't enough when considering PMRT). I also look at the data from some of the retrospective series out of MGH/Canada which suggests that G3, T2 primaries with LVI may do bad enough to warrant local therapy.

I am with you in that my threshhold has been going lower.

 

[seper]

Regarding micromets, I've heard an expert opinion that they can be ignored in figuring out PMRT.

 

[ShirleyT]

Where I work at, I would recommend that she get the axillary nodal dissection because Z11 does not apply since all of these patients got radiation. If remaining nodes negative, no XRT but if additional positive, I would recommend XRT. May be hard to talk the patient and/or surgeon into the nodal dissection though.

 

[TarHeelRadOnc]

Where I work at, I would recommend that she get the axillary nodal dissection because Z11 does not apply since all of these patients got radiation. If remaining nodes negative, no XRT but if additional positive, I would recommend XRT. May be hard to talk the patient and/or surgeon into the nodal dissection though.

I would not treat.

I think that you have to consider both the LRR risk for the chest wall and the axilla. We perseverate about nodal status and number of positive nodes when considering PMRT, but the most common site of failure is still the skin/chest wall. Nodal status aside, this patient has only one RF for LRR (T2... And very small T2 at that). Not gr 3, no LVSI, favorable subtype (ER/PR+, Her2 -), widely negative margins, and post menopausal. If SLN bx was negative, patient would not be considered in the group of high risk node negative (from Harvard and BCCA studies) in whom we might consider PMRT.

Regarding nodal status, a patient with 1mm N1mic would have been categorized as N0, and not included in Danish or BC studies. Risk of additional nodes in this patient with micromet in a single sentinel node is rather low based on MDACC and MSKCC nomograms, and the risk of 4 or more nodes is exceedingly low. Hard to accurately estimate this pts risk of locoregional recurrence with mastectomy alone (b/c large series with long-term data do not include patients with N0i+ or N1mic), but I would estimate that risk to be in the range of 4-7%. Hard to make a compelling argument for PMRT for recurrence risk in this range.

Agree that Z11 study doesn't apply, as all of those patients has breast conserving surgery + XRT. MA.20 also doesn't apply, as all of those patients had breast conserving surgery. I think that we (and the surgeons alike) should be careful applying data specific to the breast conservation setting to the post mastectomy setting. Patterns of failure are different in each setting and, as such, I don't think that we should lump them together for locoregional therapy decisions.

 

[Neuronix]

I would estimate that risk to be in the range of 4-7%. Hard to make a compelling argument for PMRT for recurrence risk in this range.

I think that's a good estimate. An ASCO abstract from 2009 shows the risk of axillary failure to be 6.2% at 5 years vs. 1% with ALND or XRT.

http://www.asco.org/ASCOv2/Meetings/Abstracts?&vmview=abst_detail_view&confID=65&abstractID=32586

 

[RadRadRad]

the Dutch Mirror numbers are only the risk of axillary failure with and without axillary treatment. presumably there is some risk (albeit relatively small) of chest wall/skin/supraclav failure on top of this such that the overall failure rate could be higher than those numbers.

That said, agree with Tarheel that overall this patient doesn't have many risk factors and that we often are too caught up on nodal status without looking at the whole picture. This is a tough case and don't think there is a definite right answer.

 

[hot sauce]

No ALND and no XRT with a positive node even if it is only 1 mm makes me feel nervous, particularly in a younger patient. If 5 year risk is in the 6% range what is the 10 or 15 year risk? She doesn't have other risk factors but the fact that it has spread to a node does suggest cancer's biology may be more aggressive.

I would probably offer treatment and at the least I think it warrants a discussion with the patient.

 

[deleted4401]

So... if we can't extrapolate results from ACOSOG/Z11 to post-mastectomy setting (i.e. omitting ALND in patients with 1 or 2 positive SLNs) because the patients all received XRT in Z11, then wouldn't the logic be that if a patient isn't getting an ALND after a mastectomy but has evidence of nodal disease, then you would give XRT to clear the axilla and may as well just hit the chest wall with tangents?

It just gets very muddy with conflicting data and surgeons omitting ALND for mastectomy patients. One that we work with is doing it, but he does it with the presumption that the patient will get RT.

For that MIRROR study, 6% risk of recurrence in axilla - but what was the chest wall/breast recurrence rate? Typically axilla is like 1-5% of recurrences, but undissected might make up, I don't know - 20-40% of recurrences, so estimated CW recurrence rate of 20-30% without treatment?

It's just rampant speculation. I'm a treater. One of the few sites in RT where we see a survival benefit. Weigh the risks and benefits, and block out the damn heart completely

S

 

[TarHeelRadOnc]

So... if we can't extrapolate results from ACOSOG/Z11 to post-mastectomy setting (i.e. omitting ALND in patients with 1 or 2 positive SLNs) because the patients all received XRT in Z11, then wouldn't the logic be that if a patient isn't getting an ALND after a mastectomy but has evidence of nodal disease, then you would give XRT to clear the axilla and may as well just hit the chest wall with tangents?

It just gets very muddy with conflicting data and surgeons omitting ALND for mastectomy patients. One that we work with is doing it, but he does it with the presumption that the patient will get RT.

For that MIRROR study, 6% risk of recurrence in axilla - but what was the chest wall/breast recurrence rate? Typically axilla is like 1-5% of recurrences, but undissected might make up, I don't know - 20-40% of recurrences, so estimated CW recurrence rate of 20-30% without treatment?

It's just rampant speculation. I'm a treater. One of the few sites in RT where we see a survival benefit. Weigh the risks and benefits, and block out the damn heart completely

S

Agree that it gets very muddy when surgeons apply data from BCS setting to patients treated with mastectomy. I think that Shirley T is right to ask for an axillary dx (which is the evidence based recommendation in this setting), but think that most surgeons would balk at that request in this setting.

Regarding LRR risk, I think that an estimate of 20-30% is VERY high. To start, the rate of axillary recurrence in the MIRROR study is comparably high relative to other studies looking at this issue. A recent MSKCC publication (Cancer 2012) suggests that the rate of axillary recurrence in patients with N1mic on SLN treated w/o completion ALN dx or Axillary Rt (not even high tangents) is in the range of ~1%. Rate from ACOSOG Z11 ~2%. We speculate that some of the Z11 patients received axillary RT in form of high tangents; but a post hoc analysis of this study looking at RT field design, to my knowledge, has not been performed. This is not the first time that LRR rates from a high profile European breast cancer study have been considerably higher than the corresponding rates from North American studies (seen LRR rate for N1 pts from Danish PMRT studies vs rates from ECOG/NSABP/MDACC patterns of failure analyses).

Assuming for a second that the axillary recurrence rate is actually 6%, that would mean that the chest wall/SCV recurrence rate would have to be 14-24% to account for your cumulative LRR estimate. Seems pretty high for a postmenopausal patient with 2.1cm grade 2, luminal A IDC with no LVSI, widely negative margins, and who will at a minimum receive an adjuvant aromatase inhibitor (+/- chemo)?!? In ECOG study (Recht 1999), risk of isolated LRR at 10yrs after mastectomy (granted w axillary dx) for patient with 2.1cm primary and 1 positive node is 4.3%. In that era, this patient (w N1mic) would have actually been scored as N0, so risk likely even lower.

We don't often think this way, but axillary recurrences in an undissected and unirradiated axilla are actually quite salvageable. In NSABP B04, 18% of patients in the no axillary tx arm had axillary recurrence which was primarily managed with axillary dx. These patients had no difference in distant DFS or OS at 25yrs. If you think that the true axillary recurrence risk in this patient is closer to 2-3% (as I do), then the option of surveillance with axillary surgery and PMRT as salvage is a reasonable option.

You could always use the oncotype as a tie breaker. There is pretty good evidence that both oncotype and mammoprint are predictive of locoregional recurrence risk. If high risk oncotype, may be worthwhile. Otherwise, PMRT seems like over treatment.

Good topic!

 

[Mandelin Rain]

Any thoughts on 59 yo T2N1mic s/p MRM with immediate reconstruction yielding a 2.5 cm, Grade 2, ER+, PR+, HER2-, Ki67 30%, +LVSI, 1/23 nodes with micromet only. MammaPrint = High Risk so she is getting full AC+T.

Combining LVSI and Size >2 cm alone may bring you to about a 15% LRR rate, independent of the N1mi, based on retrospective data. While I don't know that the N1mi adds much risk, it certainly can't lower it. That said, it was huge ALND and low ratio and potential reduced cosmesis with reconstruction.

What say you?

 

[OTN]

Any thoughts on 59 yo T2N1mic s/p MRM with immediate reconstruction yielding a 2.5 cm, Grade 2, ER+, PR+, HER2-, Ki67 30%, +LVSI, 1/23 nodes with micromet only. MammaPrint = High Risk so she is getting full AC+T.

Combining LVSI and Size >2 cm alone may bring you to about a 15% LRR rate, independent of the N1mi, based on retrospective data. While I don't know that the N1mi adds much risk, it certainly can't lower it. That said, it was huge ALND and low ratio and potential reduced cosmesis with reconstruction.

What say you?

I would treat. 2+ cm and LVSI would be enough for me- with 80+% of recurrence risk in the chest wall, though, would you need to irradiate the axilla/SCV? With 23 nodes resected one could argue to treat chest wall only to spare lymphedema risk.

 

[Palex80]

I woul

Any thoughts on 59 yo T2N1mic s/p MRM with immediate reconstruction yielding a 2.5 cm, Grade 2, ER+, PR+, HER2-, Ki67 30%, +LVSI, 1/23 nodes with micromet only. MammaPrint = High Risk so she is getting full AC+T.

Combining LVSI and Size >2 cm alone may bring you to about a 15% LRR rate, independent of the N1mi, based on retrospective data. While I don't know that the N1mi adds much risk, it certainly can't lower it. That said, it was huge ALND and low ratio and potential reduced cosmesis with reconstruction.

What say you?

I would irradiate her too.

Had she had only BCS, she would have been eligible for MA20 actually too, meaning that she would potentially benefit from an irradiation of the lymphatics too. However if you look at the subgroup analysis she would have been in a group, which would benefit only very little.

Thus I would probably only irradiate the chest wall without the lymphatics. Bearing in mind the big axillary dissection she had, I think it's a good trade off between an increased risk for severe lymphedema and only a marginal benefit in isolated nodal recurrence.

 

[evilbooyaa]

That's a good ALND to get 23 nodes. Data says to treat N1mi like N1 (versus isolated tumor cells being treated like N0), so it depends on what you'd do if she was N1. Depending on the trial you look at, you can have evidence to support either position. NCCN would tell you treat including regional nodal.

I don't think 2+cm and LVSI alone would buy you treatment at my institution. Assuming margins are negative.

 

[seper]

Is that how experts interpret MA.20 results? Or is it the the other way around (no benefit)?

I woul


I would irradiate her too.

Had she had only BCS, she would have been eligible for MA20 actually too, meaning that she would potentially benefit from an irradiation of the lymphatics too. However if you look at the subgroup analysis she would have been in a group, which would benefit only very little.

Thus I would probably only irradiate the chest wall without the lymphatics. Bearing in mind the big axillary dissection she had, I think it's a good trade off between an increased risk for severe lymphedema and only a marginal benefit in isolated nodal recurrence.

 

[Palex80]

Is that how experts interpret MA.20 results? Or is it the the other way around (no benefit)?

It depends n what you degine as benefit. "OS benefit" is little, the trial was negative. But PFS and DFS were better.

 

[seper]

I read the study too. I'm no expert so I was wondering what the message from higher ups is.