Read on Twitter
An extended synopsis and narrative of our published work @stream_research for 2020. 1/
Let me start by describing my lab's goals: to better understand the ethical, theoretical and scientific underpinnings of human research- in particular therapeutic development. 2/
The more I study this, the less I feel like I understand it, the more I doubt I have that I ever will understand it, and the more I doubt I have about those who claim they do understand it. A theme in this year's work. 3/
2020 began with a bunch of articles on cancer drug development. My former PhD student @bgcarlisle published 2 articles 4/
In this meta-analysis, we describe massive exploration of cancer combo therapies for recently approved drugs. Combo entailed >> toxicity, no survival benefit, & vanishingly small gains in terms of new Rx's 5/ https://journals.sagepub.com/doi/abs/10.1177/1740774519873883?journalCode=ctja#articleShareContainer
This meta-analysis is the monotherapy complement to the above: massive volume of repurposing trials. Almost no clinical gains. Conclusion of two papers? Much repurposing research in cancer has a bad 'burden to clinical impact' ratio 6/ @BMJ_Open https://bmjopen.bmj.com/content/10/2/e034306
See also the below (technically 2019). Imatinib: wonder drug of fast development for CML.
Ratio of patients in trials 2 clinical gains ~same as for crappy cancer drugs, when you include repurposing trials in the estimate. @bgcarlisle @OhLookATiger 7/ https://rdcu.be/ccPPo
Ratio of patients in trials 2 clinical gains ~same as for crappy cancer drugs, when you include repurposing trials in the estimate. @bgcarlisle @OhLookATiger 7/ https://rdcu.be/ccPPo
We thought, when we did that, most cancer drug trials were aimed at repurposing already approved drugs to new cancers. Squeezing one more drop from lemons. We were wrong, as this Eli Gumnit (undergrad) led paper revealed https://rdcu.be/ccPNV 8/
In 2021, stay tuned for @CharlotteOuimet (another undergrad) led work suggesting *some* albeit limited and uncertain value for post-approval label extension trials. 9/
Moving to pre-approval, my postdoc + RA @esoteric_esty and @MSKBiostats collaborator @AlexiaIasonos published this critical commentary on the ethics and validity of seamless P1 trials- a questionable trend in cancer drug dev. 10/ https://ascopubs.org/doi/10.1200/JCO.19.02456?url_ver=Z39.88-2003&rfr_id=ori%3Arid%3Acrossref.org&rfr_dat=cr_pub++0pubmed#.X-uEGl4gFss.twitter
I am esp. happy w/ this paper- led by Master's student Sean Zhang- showing odds of getting a drug in phase 1 cancer trials that will eventually be FDA approved for your cancer is ~1/80. 11/ https://academic.oup.com/jnci/article-abstract/112/9/886/5814931#.X-uEzbIx3No.twitter
Bottom line: despite a lot of dynamism in design- P1 cancer trials still not a good way to access curative therapy & some design changes (e.g. seamless) may be eroding basic ethical and statistical principles. 12/
Onto a second (complementary) line of research @stream_research: Prediction. Can we use wisdom of the crowd approaches to predict which Rx's will show efficacy? To prioritize drugs for testing? Do experts have realistic estimates of efficacy for Rx's they test in trials? 13/
This study, led by postdoc Danny Benjamin + @drmandel66 et al found cancer specialists wd generally be better off abstaining from predicting efficacy in RCTs than predicting it. : ( 14/ https://theoncologist.onlinelibrary.wiley.com/doi/full/10.1634/theoncologist.2020-0054#.X-uGlT6H-aM.twitter
Here, neurologist 'wisdom of crowd' anticipates (-) trial results, but no good at predicting evolution of disease in placebo or Rx group. : (
Led by @PavelDAtanasov + collaborators https://n.neurology.org/content/95/5/e488
15/
Led by @PavelDAtanasov + collaborators https://n.neurology.org/content/95/5/e488
15/
Bottom line (I think): be very very doubtful of expert opinion re: clinical promise of new drugs- esp experts who are confident! [There's a paradox here...] 16/
Maybe we can 'improve' expert predictions by mathematically processing them! My postdoc Patrick Kane published this series of predictions (w/ Dany Benjamin) on #parkinsons disease clinical development. https://pubmed.ncbi.nlm.nih.gov/32333550/ 17/
In a few years we'll see if coherence weighting improves forecasts. Meantime, we found that patients' expectations of drug development timelines pretty consistent w/ those of experts. Never condescend to PD patients! Collab w/ @MichaelJFoxOrg 18/ https://onlinelibrary.wiley.com/doi/10.1002/mds.28319#.X-uJgILsYYQ.twitter
Bottom line here: Bioethics (and drug development) depends on expert opinion re: future events (e.g. efficacy, recruitment, equipoise) at every turn. This is a huge liability that the field (and enterprise) has not reckoned with, IMO 19/
COVID. Am so sick of thinking/reading/writing about it. But @stream_research applied our 'brand' of research to various aspects of the pandemic. I'll mention 3: 20/
My former RA @amandamacp5 spearheaded a systematic investigation of timeline and success rates for viral vaccine development. Entire @stream_research lab deployed. Published here: 21/ https://www.acpjournals.org/doi/10.7326/M20-5350?url_ver=Z39.88-2003&rfr_id=ori%3Arid%3Acrossref.org&rfr_dat=cr_pub++0pubmed#.X-uLekYR_v4.twitter
My postdoc Patrick Kane spearheaded a 'wisdom of expert crowd' study for COVID vaccine development. Now dated (who'd have predicted it?) but published here: 22/ https://rdcu.be/ccPSt
[An aside, Kane published a gr8 article (w/ Scott Kim and I) on the concept of minimal risk in research ethics. Doubt anyone read it, sadly, but IMO this paper- in 1.5k words- shreds the view that head to head trials are automatically minimal risk.] https://pubmed.ncbi.nlm.nih.gov/31910134/ 23/
Finally- not @stream_research per se, but my long time collaborator @AlexJohnLondon & I wrote a widely circulated piece on why COVID trials must meet the same exacting standards we'd expect elsewhere. 24/ https://science.sciencemag.org/content/368/6490/476
Huge thanks to our '20 funders @MichaelJFoxOrg @GenomeCanada @GenomeQuebec @CIHR_IRSC @McGillMed @MI4Innovates + awesome team and many collaborators. And referees and journal editors. And @McGillBioethics administrative staff who hold everything together. Onto 2021. END/
