Destined to fail, page-35

Thanks for the replies so far. I'd like to explore my assertion A.

'A) If the 60% of liver only mCRC had a PFS of 6 months more than the standard care arm, then that would have put the whole population at at least 3 months PFS, as a mean (?) and probably achieved the Overall PFS result we had hoped for.'

What do we know:
Almost 600 mCRC suffers were recruited into the trial. (for the sake of the maths lets call it 600)
A statistically significant result is ~3 months more than the control arm. (source Van Huesen et al)
60% of the trail participants had predominately liver mCRC (call these the liver only group, although they may have had other organs where cancer either spread once on the trail, or minor lesions grew to the point of being counted)


Assumptions:
A 1 to 1 randomization to the control arm (There may be data but I cant find it right now, so lets run with this and adjust later if required)
300 had SIRT and 300 in the control arm.
From the 300 on the SIRT arm we are told that there was a 3 month, or greater, PFS in the liver.
180 had mainly liver only (at recruitment), and 120 had mCRC measurable in other organs.

Assertion:
If the PFS in the liver for the liver only sub group was more than 6 months over the control arm, then it would have lifted the SIRT arm of 300, to more than 3 months PFS over the control arm of 300.


My logic may be in error. However if the facts and assumptions are reasonable then the trial could have shown overall PFS > 3 month advantage over the control arm.

If the assertion has any validity then it points to the Liver PFS being in the range of 3 to 6 months, which I would have thought is a very good result for those folks that have liver only mCRC. A large number of cases each year within SRXs countries of operation.

Please tell me if I'm totally nuts, and have overlooked the blindingly obvious.