No, you're not being unclear -- I understood what you were saying. However, it is funny that you linked a paper that explicitly states that time considerations should be taken into effect, and when they are, even with an alpha/beta that is 3 times as high as 1.33, the CHHIP regimen remains isoeffective. And the SBRT regimens even moreso.
Let me say first that CHHIP looked at 60/20 vs 74/37; the paper I linked to looked at 78/39 as the standard comparative regimen (just wanted to state that at the outset to remind myself as much as anything).
Ok well I somewhat regret linking to this paper now as it's become an idee fixe for us but now I see your concern. I was linking it just to show a wide range of alpha/betas for prostate CA, do a little back-of-the-envelope calc to show some caution in wholesale acceptance of 60/20 might be in order, and then move on from there. Per Bentzen, 60/20 would remain isoeffective (or be superior) to 78/39 but this is based on one key assumption at the outset (see Table 2): that the bNED for 78/39 (the comparative regimen) is 70% @5y.
CHHIP is showing us that the bNED for 74/37 at 5y is 88.3%. So Bentzen needs to go back the drawing board and plug that new data back in to all the equations which come after Equation 1. Bentzen in his Equation 1 arrived at what I calculated where e.g.
78*(1+2/x)=60*(1+3/x)
Dc = 78
dc= 2
De = 60
de= 3
x = alpha/beta ratio
So substituting in the variables you get: Dc*(1+dc/x)=De(1+de/x)
Solving for x:
x=(De*de-Dc*dc)/(Dc-De)
Which is exactly Bentzen's equation 1 (I had never seen it before so I'm fortunate that I came up with it independently lol). His remaining equations after that attempt of course to "finesse" based on reasons I won't go into here, but this is a good starting point for gedankenexperiments regarding doses and fractionation in prostate cancer. Bentzen will need to go back and substitute the new data from CHHIP, most importantly with a starting point of 5y bNED of 88.3% with 74/37 vs Bentzen's theoretical 5y bNED of 70% with 78/39; this will skew the outcome, but I don't know how or which way honestly.
Regarding the effect of treatment time, this should not factor into our discussions too much. The difference in local control at these alpha/betas at these treatment times (53 days for 78/39 vs 26 days for 60/20, again see Table 2 in Bentzen) will be minimal. Especially as the alpha/betas get very low. Of course, Bentzen says while there is some doubt, he seems to support the hypothesis "...that overall treatment time affects the outcome of prostate cancer radiation therapy." It affects it of course favorably for local control with shorter schedules. But again, the affect will be minimal (the fraction size becomes much more predominant especially as it goes up and up)--you can see this because, again looking at Table 2, he *does not* calculate the treatment time factor (
δprolif = 0) for a/b=0.47 as he knows (I assume) there would be no treatment time factor of import at an alpha/beta that low.
AND AGAIN... the point which is being glossed over even by Bentzen... is that we are focusing on 5y LC rates. This is a slow-growing tumor. We are using PSAs as the surrogate for LC. 5y LC rates tell us very little. 10y LC rates would tell us much more. Wanting this much followup for HNSCC or lung CA or even breast CA would be silly. Wanting it for prostate cancer seems reasonable in my opinion... based on the radiobiology!
What they means by all of the alpha-beta/Tpot mumbo jumbo
Just remember... well remember it or not but it's true... the alpha/beta and the Tpot are the complete basis of every normal tissue side effect and tumor effect you see in the clinic and in followup. It's not really mumbo jumbo in that regard.
The whole reason we are involved in hypofractionating prostate CA now is based on radiolobiology and the alpha/beta!
ADDITIONAL EDIT: The CHHIP might be speaking to us, even at 5y followup. Keep in mind 57/19 in the CHHIP (1077 of the 3216 men got 57 Gy) was not proven to be isoeffective. This shows the sensitivity to dosing/fractionation for a very low alpha/beta tumor and at least hints at the possibility (in my opinion) that 60/20 might be shown to not be isoeffective over time as well.
