So I was reading mednet at Kavanaugh gives a discussion regarding SBRT of liver cancers that are near bowel which I think is germane and useful to read. In hindsight I think considering 10 x 4 (prescribed to as low a isodose as possible) with a 4cc < 40 Gy constraint would have been a reasonable thought based on his discussion
>>>>
Dr. Crane's paper (
JCO 2016 34:3, 219-226) is a very good report on IHC, and given the fact that few other institutions can accumulate such a sizeable experience, I think it is hard to argue with their nuanced approach in that setting. Their discussion acknowledges all of the caveats inherent in any such retrspective analysis, but this is an uncommon disease where prospective trials are impractical to conduct, and so their study is very helpful.
The vast majority of liver tumors I personally treat are metastases of substantially smaller volume than those treated in the MDAH IHC experience, and our approach is usually much simpler. I am not convinced that biliary stenosis is a frequent enough occurrence outside of the case of an IHC or other lesion causing or threatening stenosis by its intrinsic disruption/invasion of a major duct, and so we generally don't apply a dose limit on that structure--but if anyone knows of compelling data that suggest otherwise, I would appreciate being directed to it. Chest wall dose considerations are always in the background, and we try not to exceed a V30 (absolute) >30cc limit in 3-5 fractions, but the tumor dose is considered a higher priority, so the chest wall limit is soft in that sense. We do pay attention to a liver critical volume (at least 700 cc <15 Gy) and also mean dose per QUANTEC limits, but this issue is not commonly a dose-limiting factor, either.
In nearly every case where a normal tissue structure gives us pause, it is a segment of the GI tract that is receiving an excessive dose. If the maximum point dose to any region is above 9 Gy or so, I usually think about going from 3 up to 5 fractions, and maybe take the total PTV prescription dose down to 45/3 or 50/5, but I am not really so sure that that small difference in fraction number makes a big difference, because I get uncomfortable when the max point dose is above 34-35 Gy, since in the larger pancreas SBRT prospective studies, a typical limit would be on the order of D1cc >33 Gy (eg
Herman, J. M.,et al. Cancer, 121: 1128–1137), which is often a max dose of around 35 or so. Ultimately, when we are in that situation, my own default is a 10 fraction regimen using a "4-40" rule, ie <4cc of GI tract above 40 Gy. This local guideline was derived with all the usual hand-waving and guesstimation and blind reckoning and application of models that might have no basis in reality whatsoever, and also our local preference to have simple rules with easy mnemonics, but I think we have done ok with it though haven't done a structured restrospective review of our entire experience. In those cases, the nominal prescription dose is labeled 40 Gy in 10 fractions, but here's the rub: we always use hot hotspots to steepen the dose gradient away from the target and intensify the dose inside the ITV, and so the true prescription dose is essentially as much as we can safely give it, and the biological impact is not necessarily well represented by the nominal prescription dose. There was a time long ago when I thought EUD would be a good metric (
Medical physics, 30(3), 321-324, 2003), but that is a relatively unfamiliar notion, and it didn't catch on at all. In any case, while this all might seem like we do things in a far less rigorous manner than Chris and his team (and I wouldn't dispute that he has more self-disciline than I do
🙂), I think there is a fundamental similarity insofar as I agree that the intent should be to give as aggressive a dose as possible if local control is the objective...and inter-institutional stylistic differences about how to get there abound.
