The debate between yourself and Cafa is terrific, so thanks to...

The debate between yourself and Cafa is terrific, so thanks to both of you.

Question: can we boil it down this way ....

If we take an Overall Survival as a proxy for PFS (which I know will not be announced until the conference, and will not be finalised until the FOXFIRE trial reports) then:

1. If you have inoperable mCRC in the liver only, the use of SIRT as a first-line therapy will lengthen OS (by at least by 3 months). Crudely, 100% of these patients die of liver cancer and if the trial sample was representative, then 60% of all mCRC patients fall in this category.

2. If you have inoperable mCRC in the liver and elsewhere AND you were going to die of liver cancer, then its a reasonable assumption that the use of SIRT as a first-line therapy would lengthen OS. Its not proved yet, but its possible that OS may be lengthened by the same amount of time as group 1 (even though at the time of death, tumors in the other parts of the body would be significantly larger). Further, if the trial sample was representative, then 75% of these patients die of liver cancer (since 90% of all mCRC patients die of liver cancer) - ie, this group represents 30% of people with mCRC.

3. If you have inoperable mCRC in the liver and elsewhere AND you were going to die of a cancer other than liver cancer, then the use of SIRT as a first-line therapy would have no impact on OS. If the trial sample was representative, then this group represents 10% of people with mCRC.

Based on my basic understanding of statistics, this 3rd group would have been enough for the trial to fail its primary end-point. Nevertheless, if the assumptions / inferences about the 2nd group are correct, then I would still expect doctors to prescribe SIRT as a first line of treatment for anyone with mCRC in the liver (regardless of whether its in the liver only).

Thoughts appreciated.