Hypofrac for intermediate risk prostate - the new standard?

[napoleondynamite]

Curious on everyone's thoughts. Randomized trial. 6 years follow-up. Decent size (but I'm not a stats guy - is this trial large enough to demonstrate non-inferiority? Often these studies are criticized..)

http://ascopubs.org/doi/pdf/10.1200/JCO.2016.71.7397

---

Members don't see this ad.

 

[Gfunk6]

I hypofrac all prostate patients with Cleveland Clinic protocol (70 Gy/28 fx) if they have low or intermediate risk assuming I don't need to treat LNs.

I treat with conventional fx if I am treating LNs or if pretreatment AUA score > 20


Sent from my iPhone using SDN mobile

 

[napoleondynamite]

Grade 2 toxicity was higher for hypofrac - I'm sure that will be the main criticism?

 

[napoleondynamite]

I hypofrac all prostate patients with Cleveland Clinic protocol (70 Gy/28 fx) if they have low or intermediate risk assuming I don't need to treat LNs.

I treat with conventional fx if I am treating LNs or if pretreatment AUA score > 20


Sent from my iPhone using SDN mobile

Would you transition to 60/20 based on this data or do you think the Cleveland data is stronger than this randomized study?

Also, do you taylor this decision by extent of LUTS?

 

[Gfunk6]

Would you transition to 60/20 based on this data or do you think the Cleveland data is stronger than this randomized study?

Also, do you taylor this decision by extent of LUTS?

Don't think so. But that's mainly because our practice is about to open a dedicated SBRT/SRS center and I anticipate rapid transition to "extreme" hypofractionation.

 

[napoleondynamite]

Don't think so. But that's mainly because our practice is about to open a dedicated SBRT/SRS center and I anticipate rapid transition to "extreme" hypofractionation.

Whoa, really? Dude..

 

[John Rattan]

Whoa, really? Dude..

Norther California Kaiser Permanent in the Bay Area utilizes 5 fraction treatment frequently. Shows you what happens when you treat in a closed HMO system with salaries and no production incentives.

 

[seper]

Does Kaiser really have no incentive for their RadOnc's? I find it hard to believe. Probably some bonus related to patient satisfaction.

 

[napoleondynamite]

Disappointing to me. We need to retain ownership of fractionation decisions or insurance companies will dictate these decisions for us. We need solidarity on holding out for better data before embracing extreme hypofrac IMO.

 

[Palex80]

Not yet standard of care in my opinion. I would like to see 10 year toxicity data before switching.

I would be willing to switch, if hypofractionation showed singnificantly better BFS in comparison to standard fractionation. That has not materialized in any study yet.

 

[John Rattan]

Does Kaiser really have no incentive for their RadOnc's? I find it hard to believe. Probably some bonus related to patient satisfaction.

Yeah, not based on production as I said above. Other metrics.

 

[medgator]

Norther California Kaiser Permanent in the Bay Area utilizes 5 fraction treatment frequently. Shows you what happens when you treat in a closed HMO system with salaries and no production incentives.

Yup, Kaiser To Pay $2M In Transplant Mess

 

[mavrik13]

The difference in cost and time to the patient as well as society as a whole should not be undersold. Sure you can nitpick small differences in toxicity, but the number needed to harm is extraordinarily high for (relatively) minor toxicity. 6 year median follow-up should be sufficient to change practice. My own opinion.

 

[medgator]

The difference in cost and time to the patient as well as society as a whole should not be undersold. Sure you can nitpick small differences in toxicity, but the number needed to harm is extraordinarily high for (relatively) minor toxicity. 6 year median follow-up should be sufficient to change practice. My own opinion.

Not in a disease where men can live decades after tx. I fully embrace the hypofx data from astro last year in stage 3 lung in patients who aren't chemo candidates (given the expected OS there), as well as breast where the median fu is much longer than what we have in prostate and there actually appears to be superiority in some cases

 

[John Rattan]

Florida Fractionation Schema

1. <15 on treatment - bone mets get 40 Gy in 15 fx, prostates get 81/45

2. 16-25 on treatment - bone mets get 30 in 10 fx, prostates get 77.4/43

3. 26-35 on treatment - bone mets get 20/5, consider hypo fx for prostate

4. >35 on treatment - begin to follow standard of care, evidence based medicine, and consideration of patient convenience

 

[medgator]

Florida Fractionation Schema

1. <15 on treatment - bone mets get 40 Gy in 15 fx, prostates get 81/45

2. 16-25 on treatment - bone mets get 30 in 10 fx, prostates get 77.4/43

3. 26-35 on treatment - bone mets get 20/5, consider hypo fx for prostate

4. >35 on treatment - begin to follow standard of care, evidence based medicine, and consideration of patient convenience

Haters gonna hate . Btw, the old school docs down here treat bone mets 40/20, not 40/15

 

[RadOncDoc21]

Haters gonna hate . Btw, the old school docs down here treat bone mets 40/20, not 40/15

Sad, but true... I've seen it.

 

[MegaVoltagePhoton]

I personally await the 35-year followup data (median, not projected) from these hypofractionation studies, because I truly believe in following the evidence. Yes, I know there is no high-grade randomized evidence for 81 Gy, IGRT, IMRT over 3D, and yes I do these things routinely. I mean, I would have waited for 35 years of followup for that stuff too but the billing codes came first so what was I supposed to do?

Particularly, I'm worried about a minor risk of increased late toxicity because the 50-year followup data for dose escalation has shown me that this is a true risk, and I quote that data religiously to patients. I really spend the time to explain to them that we are talking about absolute risks of severe toxicities that are still below 10%. Specifically, for patients who are 75-80 years old, I think we really need the long-term followup data before I will even consider hypofractionation with a 28 fraction course.

Once this data is available in 2040 or so, I think we can start a phase III trial of 28 fractions versus 27 fractions, and I can posthumously recommend 27 fractions to patients when I am confident about the quantified risk of urethral strictures.

I'd post more detailed thoughts, but I have to work on my plan for a patient with 90 lung mets. I'm going to SBRT the **** out of one them. I've got my fingers crossed for the abscopal effect.

 

[scarbrtj]

I personally await the 35-year followup data (median, not projected) from these hypofractionation studies, because I truly believe in following the evidence. Yes, I know there is no high-grade randomized evidence for 81 Gy, IGRT, IMRT over 3D, and yes I do these things routinely. I mean, I would have waited for 35 years of followup for that stuff too but the billing codes came first so what was I supposed to do?

Particularly, I'm worried about a minor risk of increased late toxicity because the 50-year followup data for dose escalation has shown me that this is a true risk, and I quote that data religiously to patients. I really spend the time to explain to them that we are talking about absolute risks of severe toxicities that are still below 10%. Specifically, for patients who are 75-80 years old, I think we really need the long-term followup data before I will even consider hypofractionation with a 28 fraction course.

Once this data is available in 2040 or so, I think we can start a phase III trial of 28 fractions versus 27 fractions, and I can posthumously recommend 27 fractions to patients when I am confident about the quantified risk of urethral strictures.

Haha. Sarcasm. I like it. It's not a subtle point you're making.

Welllll... I face an internal conundrum. On one hand I had years and years of success and happy patients with 81/45. I have tried hypofx and switched to that several years ago for low-risk; it's not just recently that we've data showing hypofx as a viable tx option, after all. I see an appealing bandwagon, I 'll jump on it occasionally. Yet, *in my experience* there are more urinary side effects with hypofx. Such that it kind of fell out of favor with me. I might say the same thing about brachy; in my hands, brachy patients have more side effects vs IG-IMRT patients. I know people will want to argue these points with me, and point out the obvious pitfalls of going with my own experience and feelings versus what I read on the internet in The Best Journals, yet here I am: burdened with my own long-term experiences.

I'm sure very soon I will be forced by the insurance companies to jettison my feelings on this matter or they'll put me in the Reeducation Camp--so try not to pillory me too unkindly.

 

[scarbrtj]

If you wanna be a nerd, you can solve for the equivalent alpha/beta for 60/20 to be the BED cancer-killing equivalent of 78/39 and the number you arrive at is *drum roll* 1.33.

So if the alpha/beta of all prostate cancers is 1.33, or less, 60/20 will be equivalent (in terms of cNED rates) to 78/39. However, if the alpha/beta of prostate cancer is higher ("α/β values exceeding 4 Gy can still not be ruled out by the outcomes of the published studies"), unfortunately 60/20 will be inferior in terms of cure rates.

But I'm sure the smart people have thought that all through and this is a needless worry on my part.

 

[MegaVoltagePhoton]

If you wanna be a nerd, you can solve for the equivalent alpha/beta for 60/20 to be the BED cancer-killing equivalent of 78/39 and the number you arrive at is *drum roll* 1.33.

So if the alpha/beta of all prostate cancers is 1.33, or less, 60/20 will be equivalent (in terms of cNED rates) to 78/39. However, if the alpha/beta of prostate cancer is higher ("α/β values exceeding 4 Gy can still not be ruled out by the outcomes of the published studies"), unfortunately 60/20 will be inferior in terms of cure rates.

But I'm sure the smart people have thought that all through and this is a needless worry on my part.

4.14 (which "exceeds 4" technically, I suppose) is arrived using a "time factor" correction and is a worst-case scenario estimate. If you look in Table 2, where the actual EQD2 an bNED estimates are made, the 60/20 is actually isoeffective, as displayed in table 2, to 78/39. Further, the "extreme" hypofractionation schedules still have a higher EQD2 than 78/39 because of the condensed duration, despite the a/b of 4.

So yes, if all models are way off, then sure. Luckily, none of the modern hypofrac trials show that. Also, I'm sure we throw out alpha/beta assumptions when we bloviate over cord dose constraints in SBRT lung regimens etc (ah, that damn sarcasm again).

 

[RadOncDoc21]

How do you guys factor in RVU's though?

 

[John Rattan]

They use "patients on treatment" as a proxy for RVU to attain a target income. Has worked well for many years in the sunbelt.

 

[scarbrtj]

So yes, if all models are way off, then sure

The models need not be "way off" (and I don't know why you're harping on alpha/beta >= 4). 60/20 will only be isoeffective or superior to 78/39 for alpha/beta of 1.33 or less, and nearly all estimate exercises I've seen have quoted alpha/betas slightly, or much, higher than this. And if that is the case, which is a reasonable assumption to make, 60/20 would be BED inferior.

Luckily, none of the modern hypofrac trials show that.

Well that's not necessarily true per se, because 1) none of the trials are designed to be "alpha/beta finding," 2) we have to imply and/or guesstimate an alpha/beta from the local controls of trials with different fractionation schedules, and 3) with a tumor like prostate CA where the doubling rate can be very low, i.e. 6-18 months maybe as compared to 5-15 days for something like a squamous, local control time observations of 10 years or less--especially where local control is attached to an imprecise surrogate such as PSA--may be inadequate.

I am not saying it's wrong to hypofractionate. I am saying it's wrong to impugn the motives of those who have not yet chosen to do so.

 

[scarbrtj]

Also, I'm sure we throw out alpha/beta assumptions when we bloviate over cord dose constraints in SBRT lung regimens etc (ah, that damn sarcasm again)

Do you know of any SBRT regimen cord dose constraint that has a BED Gy3 higher than 50 Gy/25 fx? I do not know of any; nor would I ever surpass that. So, IMHO, I think we never throw out those assumptions. For me, I hold them tightly.

 

[medgator]

I am saying it's wrong to impugn the motives of those who have not yet chosen to do so.

It seems like we have a lot of angry/salaried/HMO doctors in this forum lately....

 

[RadOncDoc21]

It seems like we have a lot of angry/salaried/HMO doctors in this forum lately....

I'm salaried but against the urge to hypofractionate everyone when insurance companies are looking for reasons to pay less. If the data is there, I will offer it my patients, but I am not willing to give a more toxic treatment if the end results are equivalent (prostate). The breast data is there, but not for anything less than 15 fx despite recent data suggesting we can possibly be giving less fractions. I'm sure once these trials hit 5 yr follow up data, we will be forced to hypofractionate even more despite not having any long-term data.

On a separate but related topic. I think the current payer model needs to change. We do most of our work upfront with the consult and treatment planning and it takes as much time for me to plan an 8 Gy x's 1 bone met as it would a 30 Gy/10 fx.

I am all for hypofractionation if the data is there to support it, but don't believe it should be all or none. There are certainly some cases where a standard fractionation approach is appropriate. It's not all about money either, but I do believe that something definitely needs to change with the current reimbursement model.

 

[Phantom1]

I personally await the 35-year followup data (median, not projected) from these hypofractionation studies, because I truly believe in following the evidence...

Heh, couldn't stop laughing reading this.. awesome post man... (median, not projected)!

I think it is important to realize though that this is a problem in many fields... for example I have seen med onc's (in florida, surprise surprise..) give induction chemo (big revenue) to almost all head and neck cancer patients... this is clearly not recommended as there is no proven benefit, but worse, can cause significant long term toxicity and decreased survival, if they are unable to complete definitive treatment!!

So while an extra week or two of treatment maybe inconvienient and take up important time that could be used for other things... it doesn't risk significant long term toxicity... and may be slightly better tolerated for certain patients..

 

[RSAOaky]

So while an extra week or two of treatment maybe inconvienient and take up important time that could be used for other things... it doesn't risk significant long term toxicity... and may be slightly better tolerated for certain patients..

Furthermore, it is the standard of care and none of these trials do anything to change that standard other than saying that based on short term follow up the alternative may be acceptable. I'm all for making treatment easier on patients, but at the same time it is sad that all this money is being poured into these large non inferiority trials rather than doing things that will actually improve disease outcomes, and the study conclusions essentially amount to "well it's pretty much the same thing" while minimizing the fact that there actually is a toxicity difference (the RTOG trial specifically). Maybe the next trial will compare "moderate hypofractionation" to "extreme hypofractionation" and find that it's noninferior as well except for a bit of rectal toxicity that is unlikely to be relevant. Unfortunately you lower this threshold each time (i.e. if B is non-inferior to A and C is noninferior to B, that does not necessarily mean that C will be non-inferior to A).

At the end of the day we can get very philosophical about it all, but just ask yourself which treatment you would want if you had prostate cancer or what you would recommend if your family member had prostate cancer and to me, with the available data, the answer becomes pretty clear.

 

[John Rattan]

Do you think based on most physician treatment behavior (look at what it took for people to switch for breast and what isn't happening for bone mets), that they are actually doing what they would do for their family members? We have people talking about having one DCIS ever in their careers treated with HF. We have people that ignore pain outcomes data (not re-treat, read the papers) about bone mets.

Physician behavior in the US in a FFS will not reflect standard of care unless coerced or cajoled by outside forces. Unless said standard of care pays more. Then people will do it before it's a "standard".

Your pie in the sky statement that people will do what their family member had cancer is ludicrous. People are trying to pay the bills, man. Why is not exceeding 30/10 for bone mets a quality measure? That's like saying if someone has chest pain, you do a history and physical to meet a metric.

 

[medgator]

Do you think based on most physician treatment behavior (look at what it took for people to switch for breast and what isn't happening for bone mets), that they are actually doing what they would do for their family

Yup

We have people talking about having one DCIS ever in their careers treated with HF.

Trolling to get this thread shut down too?

Why is not exceeding 30/10 for bone mets a quality measure?

Because not everyone works in a capitated/salaried system like you do.

Why even give them 30/10? A morphine drip is cheaper...

If I had a single bone met/oligomet, I'd want 45/15 or sbrt... ever heard of that? (Both of those are done at well known academic places for oligomets). Does your CEO/resource administrator even let you do that?

Btw, enjoy being the very first person on my ignore list ever. Congrats!

 

[John Rattan]

So you don't read the papers
There is no difference in pain control between 30/10 and 8/1.

45/15 is stupid if you want to cure. Gosh. Wow. Golly. SBRT is more than reasonable. Do your weird mid fractionation if you want to. Sounds dumb A F.

My boss wants me to withhold treatment entirely. HMO4LYFE. Makes my life a living hell. The fact that I give radiation ever is an issue.

Don't be holier than thou about that. If I give 60/30 to a known stage 3 NSCLC, I'm having to discuss it. Do you have discuss with anyone when you give your 45/15 lunacy? My administrator, thank god, wouldn't allow that stupid amateur hour ****. You wouldn't last a day in a quality metric practice. 45/15. Eejit.

Yup



Trolling to get this thread shut down too?



Because not everyone works in a capitated/salaried system like you do.

Why even give them 30/10? A morphine drip is cheaper...

If I had a single bone met/oligomet, I'd want 45/15 or sbrt... ever heard of that? (Both of those are done at well known academic places for oligomets). Does your CEO/resource administrator even let you do that?

 

[Gfunk6]

All, please keep things civil.

 

[John Rattan]

All, please keep things civil.

My sincerest apologies

 

[RSAOaky]

Your pie in the sky statement that people will do what their family member had cancer is ludicrous. People are trying to pay the bills, man. Why is not exceeding 30/10 for bone mets a quality measure? That's like saying if someone has chest pain, you do a history and physical to meet a metric.

We're talking about a very specific topic here and that's hypofractionation in prostate cancer. I'm not talking about bone mets and I'm not talking about breast cancer, both of which have single fraction or hypofractionated regimens that are well established. I'm not making a pie in the sky statement about the virtuosity of radiation oncologists in a fee for service system. I'm saying that, based on the available evidence, I would not hypofractionate for prostate cancer save for very specific circumstances until the data has matured and we have a better idea of which subset of patients do worse from a toxicity standpoint. We have decades of data showing the safety and efficacy of the current standard of care, I think it would be a bit premature to change practice based on the available data. You may disagree with this approach and that doesn't offend me, but in this specific case I know what I would want if I were being treated and that's what I'd offer my patients. First do no harm.

 

[John Rattan]

Agree. Fine post!

HF for prostate is not settled science. If we are to be stewards of health finances, and if we are going to an era of greater patient "skin in the game", it's upon us to ask the patient "Hey, is a little more diarrhea okay for $5-10k less?" It's not okay to continue to ask people to pay for something that they don't see value in. I made the point about breast and bone because even when the data is clear, eejits don't do the right thing, because of economics, and because Florida. In this case, the data is less clear, so FFS doctors are never going to choose the lesser reimbursement option, unless they, too, have skin in the game.

We're talking about a very specific topic here and that's hypofractionation in prostate cancer. I'm not talking about bone mets and I'm not talking about breast cancer, both of which have single fraction or hypofractionated regimens that are well established. I'm not making a pie in the sky statement about the virtuosity of radiation oncologists in a fee for service system. I'm saying that, based on the available evidence, I would not hypofractionate for prostate cancer save for very specific circumstances until the data has matured and we have a better idea of which subset of patients do worse from a toxicity standpoint. We have decades of data showing the safety and efficacy of the current standard of care, I think it would be a bit premature to change practice based on the available data. You may disagree with this approach and that doesn't offend me, but in this specific case I know what I would want if I were being treated and that's what I'd offer my patients. First do no harm.

 

[radiaterMike]

Not in a disease where men can live decades after tx. I fully embrace the hypofx data from astro last year in stage 3 lung in patients who aren't chemo candidates (given the expected OS there), as well as breast where the median fu is much longer than what we have in prostate and there actually appears to be superiority in some cases

Does anyone know the esophageal dose constraints used in this study? I'd be cautious of 4 Gy x 10 and genuinely worried about 4 Gy x 15 partial esophagus exposure. I realize there was not a different in esophagitis but in a small study (where patients with low volume disease could be over represented) this doesn't mean that much to me to consider adopting this in my practice.


Sent from my iPhone using SDN mobile

 

[medgator]

Does anyone know the esophageal dose constraints used in this study? I'd be cautious of 4 Gy x 10 and genuinely worried about 4 Gy x 15 partial esophagus exposure. I realize there was not a different in esophagitis but in a small study (where patients with low volume disease could be over represented) this doesn't mean that much to me to consider adopting this in my practice.


Sent from my iPhone using SDN mobile

I haven't seen them, and to be fair I haven't had a lot of patients to hypofx yet, but it shouldn't be an issue in lower volume ipsi N2/N1 disease. The bigger hassle is getting imrt approval depending on the payor...

 

[Cancerdancer]

As above, this question is very analogous to the breast hypofrac discussion we had as a field 10 years ago.
At that time, people were saying "well, late effects happen late, and I'm worried about late effects, so I want to see 10 year data."
My question then is the same as it is now. Has there ever been a study where different fractionation schemes were compared (and meant to be isoeffective) head to head, and the late side effect curves tracked at 1 year, 3 years, 5 years, and then magically separated?
If the doses are meant to be isoeffective and the late side effect profile is the same at 3-5 years...it's gonna be the same at 10-20 years.

 

[napoleondynamite]

As above, this question is very analogous to the breast hypofrac discussion we had as a field 10 years ago.
At that time, people were saying "well, late effects happen late, and I'm worried about late effects, so I want to see 10 year data."
My question then is the same as it is now. Has there ever been a study where different fractionation schemes were compared (and meant to be isoeffective) head to head, and the late side effect curves tracked at 1 year, 3 years, 5 years, and then magically separated?
If the doses are meant to be isoeffective and the late side effect profile is the same at 3-5 years...it's gonna be the same at 10-20 years.

Boy, too bad we wasted 4 years learning Radbio with all those a/b ratios and other such nonsense.

Are you saying you don't believe in early vs late responding tissue?

Ever seen adults in the clinic that were radiated as children?

While these fractionation schedules seem normal to our generation of radoncs, still pretty new - and yes, I do actually believe we could see differences with 10 or 15 years of follow up as opposed to only 6.

 

[Gfunk6]

Boy, too bad we wasted 4 years learning Radbio with all those a/b ratios and other such nonsense.

Are you saying you don't believe in early vs late responding tissue?

Ever seen adults in the clinic that were radiated as children?

While these fractionation schedules seem normal to our generation of radoncs, still pretty new - and yes, I do actually believe we could see differences with 10 or 15 years of follow up as opposed to only 6.

You raise a good point. This is why I never hypofractionate my pediatric prostate or breast cancer patients.


Sent from my iPhone using SDN mobile

 

[napoleondynamite]

You raise a good point. This is why I never hypofractionate my pediatric prostate or breast cancer patients.

Sent from my iPhone using SDN mobile

Touche

 

[MegaVoltagePhoton]

The models need not be "way off" (and I don't know why you're harping on alpha/beta >= 4). 60/20 will only be isoeffective or superior to 78/39 for alpha/beta of 1.33 or less, and nearly all estimate exercises I've seen have quoted alpha/betas slightly, or much, higher than this. And if that is the case, which is a reasonable assumption to make, 60/20 would be BED inferior.

Well that's not necessarily true per se, because 1) none of the trials are designed to be "alpha/beta finding," 2) we have to imply and/or guesstimate an alpha/beta from the local controls of trials with different fractionation schedules, and 3) with a tumor like prostate CA where the doubling rate can be very low, i.e. 6-18 months maybe as compared to 5-15 days for something like a squamous, local control time observations of 10 years or less--especially where local control is attached to an imprecise surrogate such as PSA--may be inadequate.

I am not saying it's wrong to hypofractionate. I am saying it's wrong to impugn the motives of those who have not yet chosen to do so.

Um...I was "harping" on it because you posted a link to the article in support of your contention that the alpha/beta can exceed 4. Yet, while this is technically true as stated in the article (exceeds all the way to a whopping 4.14, which would still be a low alpha/beta), the very same article states that the treatment time must also be taken into consideration and clearly reports that the EQD2 and projected NED for 78/2 Gy and 60/3 Gy are equal. Napoleon just made an impassioned post about people forgetting radiobiology...did you all forget that the 4 Rs are a simplification, and treatment time is important?

All of this is besides the point, and is in reference to your remark about how these trials/this approach is radiobiologically questionable. Trials don't have to be designed to be "alpha/beta" finding for the results to provide clinical evidence of effectiveness. Were the rates of biochemical failure at long-term followup worse in CHHIP, HYPRO, PROFIT to an extent that would imply the alpha/beta estimates were wrong?

 

[scarbrtj]

Um...I was "harping" on it because you posted a link to the article in support of your contention that the alpha/beta can exceed 4.

No... my *real* contention is that 1) for NED 78/39 and 60/20 to be equal, the alpha/beta must be 1.33 or less, 2) I provided one meta-analysis that illustrated, historically, most of the time we've guessed that prostate alpha/beta is >1.33 (the alpha/beta of 4 is a red herring to this argument), and 3) since prostate has a known large Tpot, or slow doubling time, or is a slowly proliferating etc etc tumor, long followup times of 10-15+ years may be needed to see a difference in NED between 78/39 and 60/20.

I have furthermore not forgotten that treatment time is important. But as you will recall, from Hall, the treatment time correction factor for the BED only comes into play when the Tpot is small (like 30 days or less). This is why in his section dealing with correcting for treatment time, he mentions it only for the BED Gy-10 calculations (dealing with H&N Ca tumor control probabilities e.g.), not for the BED Gy-3 calculations. Up above BED Gy-3, perhaps BED Gy-4 or more, I can see how treatment time would be important. However, if we posit an alpha/beta of 1.33 (which we do for 78/39 vs 60/20), a BED Gy-1.33 time correction factor would be an equally moot point (ie there is no extra BED gained by shortening the treatment time from 39 fractions to 20 fractions).

 

[MegaVoltagePhoton]

The study you introduced to the argument (now conveniently a red herring, along with your reference to an a/b exceeding 4), with the senior author being a pre-eminent radiobiologist and an ASTRO Gold Medal winner, argues differently about CHHIP and particularly extreme hypofractionation. I'll probably go with his conclusions.

 

[scarbrtj]

The study you introduced to the argument (now conveniently a red herring, along with your reference to an a/b exceeding 4), with the senior author being a pre-eminent radiobiologist and an ASTRO Gold Medal winner, argues differently about CHHIP and particularly extreme hypofractionation. I'll probably go with his conclusions.

I know who Bentzen is yeesh. His conclusions are exactly what I'm concluding too. I'm trying to say that there's a rather inchoate alpha/beta value floating out there, but *specifically* appertaining to 78/39 vs 60/20... where

78*(1+2/x)=60*(1+3/x) ...

and solving for x (the alpha/beta) you get 1.33333.....

Dr Bentzen does not refute me anywhere (nor come to a different conclusion). You can plug in 4 for x, and then 60/20 is "weaker" than 78/39. You can plug in 1 for x, and then 60/20 is "stronger." Of course x=1.33, they're equal. And so on and so forth. I feel like I'm not being as clear as I should...

 

[Cancerdancer]

Boy, too bad we wasted 4 years learning Radbio with all those a/b ratios and other such nonsense.

Are you saying you don't believe in early vs late responding tissue?

Ever seen adults in the clinic that were radiated as children?

You mock my question, but don't actually refute/address my observation/question.
Of course I believe in early/late tissues, alpha/beta etc.
However, my question is...is there an example of a study where the observed rate of a late side effect (rectal bleeding, breast telangiectasia/fibrosis, whatever) is observed to be equal between two RT fractionation schedules at 5 years, and then thereafter separate?
I get that the rate of (rectal bleeding, whatever) will continue to slowly increase beyond 5 years, but I've never seen evidence that if two radiation schedules both result in say 6% rectal bleeding rate at 5 years, one of them will plateau and the other will continue to escalate, which is basically the argument you have to espouse if you insist on 10 year plus data.

 

[RadOncDoc21]

You mock my question, but don't actually refute/address my observation/question.
Of course I believe in early/late tissues, alpha/beta etc.
However, my question is...is there an example of a study where the observed rate of a late side effect (rectal bleeding, breast telangiectasia/fibrosis, whatever) is observed to be equal between two RT fractionation schedules at 5 years, and then thereafter separate?
I get that the rate of (rectal bleeding, whatever) will continue to slowly increase beyond 5 years, but I've never seen evidence that if two radiation schedules both result in say 6% rectal bleeding rate at 5 years, one of them will plateau and the other will continue to escalate, which is basically the argument you have to espouse if you insist on 10 year plus data.

Agree, especially in a disease where the majority of patients will live to experience long-term side effects. I rather get treated an extra 2 weeks then have rectal bleeding for the rest of my life.

I'm sure the outcomes will eventually be the same but right now the long-term data isn't there.

 

[John Rattan]

HA!

You raise a good point. This is why I never hypofractionate my pediatric prostate or breast cancer patients.


Sent from my iPhone using SDN mobile

You mock my question, but don't actually refute/address my observation/question.
Of course I believe in early/late tissues, alpha/beta etc.
However, my question is...is there an example of a study where the observed rate of a late side effect (rectal bleeding, breast telangiectasia/fibrosis, whatever) is observed to be equal between two RT fractionation schedules at 5 years, and then thereafter separate?
I get that the rate of (rectal bleeding, whatever) will continue to slowly increase beyond 5 years, but I've never seen evidence that if two radiation schedules both result in say 6% rectal bleeding rate at 5 years, one of them will plateau and the other will continue to escalate, which is basically the argument you have to espouse if you insist on 10 year plus data.

Of course there isn't. It's a "straw man" to delay implementation of lesser reimbursed treatments. Whelan/START at 5 years - "Not enough follow up!"; Whelan/START at 10 years "Not enough follow up!" If it's going to show up, it's going to show up at 5 years. Except for 2nd malignancy, probably.

 

[scarbrtj]

If it's going to show up, it's going to show up at 5 years. Except for 2nd malignancy, probably.

My "worry" as it were is not the late side effect profile past 5 years. I agree: for tissues which are "classically" late-responding and have alpha/betas in the ~3 range, 5 year Kaplan-Meier assessed late side effect rates should be adequate and will not appreciably diverge later in time. However for *very* late-responding cells such as prostate where the alpha/beta may be 2... or 1... 5 year followups will be, at least mathematically/statistically, inadequate. I would go so far as to say a 5-year local control comparison between fractionation schedules on an alpha/beta=1 cell line would be equivalent to assessing 6-month local controls for a rapidly proliferating tumor e.g. H&N sq c ca. (Now one must also come to the table and be honest and say there has been a bit of a trend toward higher chronic late side effects in these hypofractionated protocols; your assessment of the magnitude, severity, clinical impact, etc., of these effects is up to you.)

Pooh-pooh me if you wish. Could be a case of being too clever by half. Yet I say again... I have niggling reservations. Read too much Hall I reckon.