February 3rd, 2024
I have seen (with great pleasure!) the rising recent interest in the survival curves as they have progressed to tentative interpretability. The story is basically unchanged this month (see below), with relative performance reflecting good additivity with the exception of two of the male three-intervention groups.
I'm SLIGHTLY disappointed that the average all-control survival (across both sexes) is only 40% at this point, because we're right on the 31-month age that is the historical median (i.e. 50%) survival age, but that's not a shortfall that I'll lose too much sleep over.
Also, I've tried to respond to feedback that the graphs are rather hard to read. The ten groups are now distinguished not only by colour but by style of line: the all-treatment and all-control groups are thick lines (red and blue respectively), the single-treatment groups are dashed lines and the three-treatment groups are solid lines. I hope this helps!
As for further details, well, people pointed out last time that rapamycin is doing the best out of the single-intervention groups, and that's still true except that telomerase has caught up with rapamycin in females. A few people went further and tried to claim that rapa on its own is doing virtually as well as all four interventions combined, but I think it's even clearer now that that's not the case, especially in males.
There's a bit of a story emerging, however, concerning the relative benefit of the interventions. I think we're starting to be able to say that rapa and mTERT are more beneficial, in both sexes, both individually and in combination, than HSCs and Gal-Nav. In males the anomalously poor-performing three-intervention groups are those that lack rapa and mTERT, and rapa-only is the best single intervention. In females, similarly, Gal-Nav and HSCs are the single-intervention groups that are performing the worst, not really distinguishable from all-controls, and reciprocally the no-HSC and no-Gal-Nav groups are doing about as well as the all-four group while the other two three-intervention groups are doing slightly worse. It's far too early to describe this as anything more than a hint of a trend, but I'll be watching it to see if it persists.
But for now, this gives me the opportunity to highlight one important aspect of our study design. Insofar as the HSC and Gal-Nav treatments are underperforming, that could be because they are just not benefiting the mice, or it could be that the benefits are being counteracted by negative effects of the delivery process. The latter is quite plausible because multiple injections were required.
So, for the first time, I've done a comparison of the two types of controls, mock and naive. (If you don't remember what these are, please go back to the description of the study at the LEVF website.) For each intervention, there are five treatment groups (so, 250 animals of each sex) that get it and five that don't. Of the ones that don't, therefore, there are 125 animals of each sex - 25 from each relevant treatment group - that got the mock control and 125 that got the naive control. So, here are the curves!
And you'll immediately see that the stories for HSCs and for NavGal are significantly different. In the case of HSCs, as of now it looks rather strongly as though the delivery process explains most of the poor performance in males, but little or none of the poor performance of females. In the case of GalNav, on the other hand, delivery doesn't seem to explain much. But another quite conspicuous aspect of these curves is how they compare early on. For both HSCs and GalNav, for quite a long time (up to age 800 days or so), the MOCK females did quite considerably better than the NAIVE ones! I'm definitely at a loss to explain that.
The biggest questions are still a few months away from being answered, though. I'm talking, of course, about the impact on maximum lifespan (which, as is conventional, we will define as the point when only 10% of the cohort is alive).
