Breast IMRT Choosing Wisely... We Knew Ye

[RickyScott]

I think I'm like many folks who will open tangents superiorly to cover first 3 intercostal interspaces for IM nodes with trimming inferiorly. Will also sim flat and push ISO as superiorly as reasonable to reduce lung dose from Sclav. Coverage to 45 or even 40 Gy often accepted.

Matched electron was one of those things that I learned in residency and then I had some real skin toxicity in the real world that I never wanted to cause again. Matched photon will blow up heart dose, which we emphasize more now than 15 years ago.

I have used IMRT for left sided breast/CW with RNI. The coverage recs per these protocols aren't much more than what I described above. A lot of allowing 90% of volume to get prescription dose. Sometimes necessary for heart/lung dose but rarely and always a compromise IMO regarding contralateral breast dose and integrated dose.

If I am treating the imn (pt w really bad disease)contralateral breast is my lowest concern.

---

Members don't see this ad.

 

[RadRadRad]

This may have already been mentioned but I think ma20 allowed imn to be covered by 80% idl
Definitely helps keep the lung and heart lower to not push the optimizer to full coverage

 

[communitydoc13]

If I am treating the imn (pt w really bad disease)contralateral breast is my lowest concern.

Agree. We often don't meet protocol on contralateral breast with IMRT and I don't worry.

More interesting is heart dose. Is the mean dose thing real or just real and reflective of toxicity with tangents? IMRT will often have high mean heart dose due to low dose bath. But is this equivalent to low mean heart dose and the 60% IDL going through anterior left ventricle and in proximity to LAD?

 

[taserlaser]

Tangential anecdote time:

Few years back I treated a man’s whole heart 30 Gy with chemo for locally progressive pericardial angiosarcoma. Sudden death maybe a year later, I attended the autopsy. Cause of death was a spontaneous bleeding met to his abdomen, not his heart. Heart musculature and coronaries were grossly pristine. Pericardium was fibrosed for sure though.

One more that comes to mind - I also SBRT’d an LV myocardial met picked up on whole body MR surveillance a few years back for a 30 year old. Something like her 7th course of RT for oligomet/progression. Heart still functionally great and NED to date. Still need to finish writing this one up when I have time.

 

[TheWallnerus]

Ok so nobody is answering the question.

This may have already been mentioned but I think ma20 allowed imn to be covered by 80% idl
Definitely helps keep the lung and heart lower to not push the optimizer to full coverage

good point

I think if one signs on to the idea that four fifths of the dose cures IMNs of breast cancer cells, backing down to two thirds of Rx dose to save heart seems like a reasonable fallback position when necessary?

 

[TheWallnerus]

Tangential anecdote time:

Few years back I treated a man’s whole heart 30 Gy with chemo for locally progressive pericardial angiosarcoma. Sudden death maybe a year later, I attended the autopsy. Cause of death was a spontaneous bleeding met to his abdomen, not his heart. Heart musculature and coronaries were grossly pristine. Pericardium was fibrosed for sure though.

One more that comes to mind - I also SBRT’d an LV myocardial met picked up on whole body MR surveillance a few years back for a 30 year old. Something like her 7th course of RT for oligomet/progression. Heart still functionally great and NED to date. Still need to finish writing this one up when I have time.

Those are wild stories all around! You were probably the only rad onc in North America who went to an autopsy that year.

 

[taserlaser]

Those are wild stories all around! You were probably the only rad onc in North America who went to an autopsy that year.

Thanks, I can't claim credit for that one. The med onc got the consent and arranged it - his office is right next to mine, and we have a great working relationship. It absolutely was worthwhile, and I learned a lot.

 

[Ray D. Ayshun]

I use b51 to inform constraints. Acceptable variation is 90% covered by 80% for imns. I push based on location etc, but accept this without regret in left sided uoq er/pr+ rni and push harder if inner triple negative. Blah blah blah

 

[OTN]

Tangential anecdote time:

Few years back I treated a man’s whole heart 30 Gy with chemo for locally progressive pericardial angiosarcoma. Sudden death maybe a year later, I attended the autopsy. Cause of death was a spontaneous bleeding met to his abdomen, not his heart. Heart musculature and coronaries were grossly pristine. Pericardium was fibrosed for sure though.

One more that comes to mind - I also SBRT’d an LV myocardial met picked up on whole body MR surveillance a few years back for a 30 year old. Something like her 7th course of RT for oligomet/progression. Heart still functionally great and NED to date. Still need to finish writing this one up when I have time.

Let's not forget that SBRT for VT goes to 25 Gy x 1, and the volume of the heart irradiated is not small.

 

[Mandelin Rain]

Let's not forget that SBRT for VT goes to 25 Gy x 1, and the volume of the heart irradiated is not small.

As certain as death and taxes.... breast is the worst.

 

[TheWallnerus]

I use b51 to inform constraints. Acceptable variation is 90% covered by 80% for imns. I push based on location etc, but accept this without regret in left sided uoq er/pr+ rni and push harder if inner triple negative. Blah blah blah

Some guy(s) in some place just came up with those “acceptable variations.” They’re wholly invalidated and untested. They just “make sense” I suppose. What if someone came ‘round and said “I have allowed 60 to 70% IDL to 90% volume for decades and never saw a single IMN failure in thousand patients.” Wouldn’t that be compelling/edifying too.

Yes. I am saying that if someone writes “this is an acceptable variation” in a protocol, and people latch on to that, think of why that acceptable variation seems to have worked out so nicely here for these IMNs. People almost never fail in an isolated fashion in the IMNs that’s why. That rate can be as low as 1 in 500 to 1000 in some analyses. Coverage of 50% of the IMNs with the 50% IDL would’ve been shown to be an acceptable variation.

 

[Ray D. Ayshun]

Some guy(s) in some place just came up with those “acceptable variations.” They’re wholly invalidated and untested. They just “make sense” I suppose. What if someone came ‘round and said “I have allowed 60 to 70% IDL to 90% volume for decades and never saw a single IMN failure in thousand patients.” Wouldn’t that be compelling/edifying too.

Yes. I am saying that if someone writes “this is an acceptable variation” in a protocol, and people latch on to that, think of why that acceptable variation seems to have worked out so nicely here for these IMNs. People almost never fail in an isolated fashion in the IMNs that’s why. That rate can be as low as 1 in 500 to 1000 in some analyses. Coverage of 50% of the IMNs with the 50% IDL would’ve been shown to be an acceptable variation.

Don't hate the player. If this trial was designed by a bunch of breast radoncs who literally have nothing better to do, I'm gonna roll with. I think its symbolic of everything you just said. It's a mathematical way of saying, "we hate breast."

 

[RickyScott]

Ok so nobody is answering the question.

good point

I think if one signs on to the idea that four fifths of the dose cures IMNs of breast cancer cells, backing down to two thirds of Rx dose to save heart seems like a reasonable fallback position when necessary?

I am reaching way back but I thought in flethecers orgiginalnstudies 30gy/2 gy fraction would kill microscopic disease 50-70.% of time but 45-50 was needed for 95% kill. And this was withou highly active systemic agents we have today in breast. Therefore it is very unlikely to matter if imn gets 80, 90 or 100 percent of dose.

 

[taserlaser]

I am reaching way back but I thought in flethecers orgiginalnstudies 30gy/2 gy fraction would kill microscopic disease 50-70.% of time but 45-50 was needed for 95% kill. And this was withou highly active systemic agents we have today in breast. Therefore it is very unlikely to matter if imn gets 80, 90 or 100 percent of dose.

I’ll have to look that up, interesting

 

[Lamount]

Tangential anecdote time:

Few years back I treated a man’s whole heart 30 Gy with chemo for locally progressive pericardial angiosarcoma. Sudden death maybe a year later, I attended the autopsy. Cause of death was a spontaneous bleeding met to his abdomen, not his heart. Heart musculature and coronaries were grossly pristine. Pericardium was fibrosed for sure though.

One more that comes to mind - I also SBRT’d an LV myocardial met picked up on whole body MR surveillance a few years back for a 30 year old. Something like her 7th course of RT for oligomet/progression. Heart still functionally great and NED to date. Still need to finish writing this one up when I have time.

I’ve SBRTed quite a few hearts at this point for VT… no cardiac tox that we have see. and some have actually had their EF improve

 

[communitydoc13]

I know everyone knows this, but when we talk about heart dose in breast it is not in regard to acute cardiac toxicity, which is extremely rare at any of the doses we give. We are just talking about potentiating symptomatic atherosclerotic disease over time. This is obviously not relevant in the angiosarcoma patient or refractory VT patient.

There is reason to get pretty conservative regarding heart dose in a 60 y/o ER+ breast cancer patient.

 

[Lamount]

I know everyone knows this, but when we talk about heart dose in breast it is not in regard to acute cardiac toxicity, which is extremely rare at any of the doses we give. We are just talking about potentiating symptomatic atherosclerotic disease over time. This is obviously not relevant in the angiosarcoma patient or refractory VT patient.

There is reason to get pretty conservative regarding heart dose in a 60 y/o ER+ breast cancer patient.

Agree.

I nonetheless think the whole cardiac dose paradigm is strange. We know that in many disease sites (breast, lung, esophagus etc...) when heart isn't the target, less dose is better. However, the mechanisms of morbidity are nebulous.

Is it the coronary arteries, the mean heart dose, or some other DVH metric we should worry about? Is the RT causing CAD, arrhythmias, CHF... is it something else entirely (like immunosuppression), or all of the above? It's just strange how little we know

 

[medgator]

I know everyone knows this, but when we talk about heart dose in breast it is not in regard to acute cardiac toxicity, which is extremely rare at any of the doses we give. We are just talking about potentiating symptomatic atherosclerotic disease over time. This is obviously not relevant in the angiosarcoma patient or refractory VT patient.

There is reason to get pretty conservative regarding heart dose in a 60 y/o ER+ breast cancer patient.

Stage 3 lung probably somewhere in the middle on that spectrum

 

[evilbooyaa]

Agree.

I nonetheless think the whole cardiac dose paradigm is strange. We know that in many disease sites (breast, lung, esophagus etc...) when heart isn't the target, less dose is better. However, the mechanisms of morbidity are nebulous.

Is it the coronary arteries, the mean heart dose, or some other DVH metric we should worry about? Is the RT causing CAD, arrhythmias, CHF... is it something else entirely (like immunosuppression), or all of the above? It's just strange how little we know

It is strange, but until somebody comes up with a better marker for predicting late cardiac toxicity in cancer patients (think about how we also use MHD in Lung and Esophageal caner), it will continue to be used. Pontificating that "MHD doesn't matter, how does that even make sense" and then blowing through lax guidelines is not a recipe for success.

B51 allowed 90% of PTV to receive 40Gy (80% of Rx dose)

 

[TheWallnerus]

I am reaching way back but I thought in flethecers orgiginalnstudies 30gy/2 gy fraction would kill microscopic disease 50-70.% of time but 45-50 was needed for 95% kill. And this was withou highly active systemic agents we have today in breast. Therefore it is very unlikely to matter if imn gets 80, 90 or 100 percent of dose.

True.

Actually for “subclinical” (he coined that term) which was one level below microscopic (needs 60 Gy) which was one level below gross (needs 70 Gy).

 

[dieABRdie]

Agree.

I nonetheless think the whole cardiac dose paradigm is strange. We know that in many disease sites (breast, lung, esophagus etc...) when heart isn't the target, less dose is better. However, the mechanisms of morbidity are nebulous.

Is it the coronary arteries, the mean heart dose, or some other DVH metric we should worry about? Is the RT causing CAD, arrhythmias, CHF... is it something else entirely (like immunosuppression), or all of the above? It's just strange how little we know

I found this helpful when not able to meet the mean heart dose; at least try to meet LAD.

DEFINE_ME

 

[taserlaser]

I know everyone knows this, but when we talk about heart dose in breast it is not in regard to acute cardiac toxicity, which is extremely rare at any of the doses we give. We are just talking about potentiating symptomatic atherosclerotic disease over time. This is obviously not relevant in the angiosarcoma patient or refractory VT patient.

There is reason to get pretty conservative regarding heart dose in a 60 y/o ER+ breast cancer patient.

Oh absolutely. No disagreement - the specific clinical concerns and timelines are completely separate.

 

[RickyScott]

I’ve SBRTed quite a few hearts at this point for VT… no cardiac tox that we have see. and some have actually had their EF improve

Does this get reimbursed?

 

[OTN]

Does this get reimbursed?

not yet

 

[HolySmokes]

We are adding flash to our vmat plans. Heats up the superficial tissues a smidge but as I understand it is more robust.

I should also mention our default is static field imrt 5 fraction whole breast +- sequential boost

How are people doing this in practice? We are on Pinnacle and despite multiple conversations with Phillips, they have yet to give us a good way of forcing flash on VMAT for breast.

 

[RickyScott]

How are people doing this in practice? We are on Pinnacle and despite multiple conversations with Phillips, they have yet to give us a good way of forcing flash on VMAT for breast.

I used to plan on pinnacle years ago and would paint in bolus.

 

[evilbooyaa]

How are people doing this in practice? We are on Pinnacle and despite multiple conversations with Phillips, they have yet to give us a good way of forcing flash on VMAT for breast.

I've only used Eclipse for TPS but here is my work flow for VMAT Flash - This actually applies to vulva cases but can be translated to anything requiring VMAT Flash.

Put a virtual bolus of whatever amount of flash you want on skin. Let's say 3cm of Flash, so 3cm virtual bolus to the area overlying the PTV.

Take your previous PTV and expand 3cm extra from your eval structure. Your PTV may be outside the body, but all within the virtual bolus (not outside of the virtual bolus)
With the bolus present, create an IMRT plan planned to the PTV + 3cm structure you created. The optimizer will attempt to deposit dose (successfully) into the bolus region. Optimize as desired for constraints and such.

Copy the plan, remove the virtual bolus and re-calculate the plan without re-optimizing. Verify that MLCs are spraying dose into air anterior and lateral (in this case) to the structure.

Your PTV_Eval and OAR doses may look different on the re-calculated plan, but usually end up within a few percentage points.