This tweetorial explains a Dec 2019 preprint I posted on OSF. See also this short video.
I’ve been mulling over this Dec 2017 Palmer & Sorger @CellCellPress article for a long time now 2/https://t.co/Z35ivL0SHw
— David C. Norris, MD moved to Mastodon 🦣 (@davidcnorrismd) December 5, 2019
Even a few weeks ago, I was still (yes, it’s true) exploring a category-theoretic formulation of the intuition! But as happens so often in this #DTAT research programme, simplicity has won the day. 4/
— David C. Norris, MD moved to Mastodon 🦣 (@davidcnorrismd) December 5, 2019
I open the discussion by presenting the Palmer-Sorger notion of ‘independent action’ in a scenario (Fig. 1) that excludes mechanistic synergy by construction. Rescaling the axes (needs animation!) then shows concave population-level isoboles despite zero mechanistic synergy. 6/ pic.twitter.com/efQmKD9ZFC
— David C. Norris, MD moved to Mastodon 🦣 (@davidcnorrismd) December 5, 2019
Of course, to discuss a scenario of actual synergism, I must depart somehow from the zero-synergy Palmer-Sorger world of ‘independent action’. How exactly to make this departure is the question I’ve struggled with for months. I found my answer in a 1985 paper by Emil Frei III 8/ pic.twitter.com/4qoUqSTxg9
— David C. Norris, MD moved to Mastodon 🦣 (@davidcnorrismd) December 5, 2019
Much of the nutritional content of this paper is to be found its figure captions, but let’s zoom in on the Figures themselves …
— David C. Norris, MD moved to Mastodon 🦣 (@davidcnorrismd) December 5, 2019
Fig.3 shows a shared ‘background’ of therapeutic isoboles common to the whole population, i.e. absent any heterogeneity of treatment effect (HTE) 10/ pic.twitter.com/4AX1oiuWpm
Monotherapy doses in this figure lie on the horizontal (A) and vertical (B) axes. Thus, patient ‘x’ could tolerate drug A at a dose that confers a 10% chance of response, and can tolerate drug B at a dose that would confer nearly a 20% chance of responding. 12/
— David C. Norris, MD moved to Mastodon 🦣 (@davidcnorrismd) December 5, 2019
What happens when we overlay the methods of 1-size-fits-all #dosefinding upon such therapeutic synergism? The key ‘contribution’ of 1-size-fits-all #dosefinding is the (atrocious) #TargetToxicityRate concept demolished in the tweetorial linked below 14/https://t.co/I1QMPPP3eg
— David C. Norris, MD moved to Mastodon 🦣 (@davidcnorrismd) December 5, 2019
The point of tangency of this toxrate isobole with a therapeutic isobole (red dot) is the 1-size-fits-all #dosefinding methodologist’s idea of ’the’ recommended combination dose. Who benefits from these methodologists’ ideas, and by how much? 16/
— David C. Norris, MD moved to Mastodon 🦣 (@davidcnorrismd) December 5, 2019
These 3 doses put you (respectively) on the green, blue & gray isoboles. In this crazy, 1-size-fits-all world, a few patients (≈2%) in the hatched area couldn’t tolerate either monotherapy dose, but CAN take the recommended combo dose. They get a big benefit (black arrow) 18/
— David C. Norris, MD moved to Mastodon 🦣 (@davidcnorrismd) December 5, 2019
Everyone else (white areas) gets NO benefit from synergy, since they cannot tolerate ’the’ recommended combo dose. This of course recalls the partitioning of a heterogeneous population as explored below for 1-size-fits-all dosing of monotherapy: 20/https://t.co/6jPnV1rpFn
— David C. Norris, MD moved to Mastodon 🦣 (@davidcnorrismd) December 5, 2019
Stepping back for broader view, I suggest viewing this working paper as complementary to my Aug 2018 “Costing ’the’ MTD … in 2-D” preprint. Whereas that earlier work stressed concreteness, here I stress genericity. 22/https://t.co/cAbSczt8s9
— David C. Norris, MD moved to Mastodon 🦣 (@davidcnorrismd) December 5, 2019
Here BTW is the poster, stressing genericity as essential in all useful clinical heuristics:
— David C. Norris, MD moved to Mastodon 🦣 (@davidcnorrismd) December 5, 2019
“Amid the perplexity and uncertainties of clinical practice, clinicians place special value on robust heuristics of broad applicability [such as] Osler’s aphoristic distillations” 24/ pic.twitter.com/n5UOGk5YAe
YES, I do! In fact, I no longer think it credible that a lack of knowledge is the actual barrier to wider adoption of the dose-titration designs I advocate. A more investigative-journalistic approach now seems required. “Follow the money,” as it were… 26/https://t.co/reMOxpfQQg
— David C. Norris, MD moved to Mastodon 🦣 (@davidcnorrismd) December 5, 2019
But #DTAT becomes now less scholarship, more politics. I look forward to reaching out to phase 1 trialists 1-by-1—appropriately for a precision-medicine endeavor!—while uncovering whatever political barriers I can identify to the adoption of #DoseIndividualization designs. 28/fin
— David C. Norris, MD moved to Mastodon 🦣 (@davidcnorrismd) December 6, 2019