Jing Liang 🇺🇦

36.1K posts

Jing Liang 🇺🇦 banner
Jing Liang 🇺🇦

Jing Liang 🇺🇦

@AppleHelix

Entrepreneur. Drug hunter. Anti-ideologues, Anti-medical nihilists, Optimist My NGO: https://t.co/yDWRnOYAyY https://t.co/SOf8c5OZm6

Katılım Aralık 2009
4.9K Takip Edilen10.9K Takipçiler
Jing Liang 🇺🇦
Jing Liang 🇺🇦@AppleHelix·
The future is here
Daniel V. Araujo@DVAraujoMD

PROTRACT (#ASCO26, Abstr 5017) — ctDNA-guided treatment selection in mCRPC after abiraterone. 🔹 The idea Use ctDNA fraction (ctDNA%) to pick the next therapy. Rationale: high ctDNA% (≥2%) predicts a poor response to a second ARPI but preserved sensitivity to chemo. So the biomarker arm sent ≥2% → docetaxel and <2% → enzalutamide, vs clinician's choice. 🔹 The trial Small randomized phase II, stopped early for slow accrual — only 42 pts (17 biomarker-directed vs 25 clinician's choice). 🔹 Results Both PFS (5.6 vs 2.5 mo, HR 0.4) and OS (46.3 vs 15.3 mo, HR 0.4) favored the ctDNA-guided arm. PSA50 numerically better but not significant. 🔹 My take: Nice study! Thought-provoking, but very small numbers. Most pts on clinician's choice got enzalutamide (second ARPI), which tends to be the weaker option after abiraterone (à la CARD trial). So it's unclear whether the benefit is really from better biomarker selection or just from the guided arm using docetaxel more often (11/17 vs 4/25). What I'd really want to see is how the pts whose biomarker favored ENZA actually did on ENZA — IMO that's the real test of the biomarker. Looking forward to seeing that. 🔗 asco.org/abstracts-pres… #pcsm #ctDNA #ASCO26

English
1
0
3
3.4K
Jing Liang 🇺🇦
Jing Liang 🇺🇦@AppleHelix·
@Papa_Heme @Eddie_Cliff This is where I disagree philosophically. Destined to have bad cancer vs not. This is not black and white. If will be a Gaussian distribution meaning there will be people whom even the best test classify is “not bad” will get “bad cancers”
English
1
0
0
80
Jing Liang 🇺🇦
Jing Liang 🇺🇦@AppleHelix·
@Papa_Heme @Eddie_Cliff I do think the future of cancer treatment is to treat it early. We may disagree philosophically. Hopefully they will be therapies with mild AEs. Some cancer vaccines are promising in this regard. Very safe and single injection.
English
1
0
1
44
Jing Liang 🇺🇦 retweetledi
MTS
MTS@MTSlive·
SITUATION DETECTED: Google DeepMind’s AI agent autonomously solved 9 of 353 open Erdos problems in mathematics, at a cost of a few hundred dollars per problem.
English
151
379
5.8K
2.3M
Jing Liang 🇺🇦
Jing Liang 🇺🇦@AppleHelix·
@Eddie_Cliff @Papa_Heme The debate always comes down to crossover timing trials. While I understand that may be ideal, it will take a much larger and longer trial
English
1
0
1
98
Eddie Cliff
Eddie Cliff@Eddie_Cliff·
@Papa_Heme @AppleHelix The OS is very difficult to interpret given the lack of dara for patients who had disease progression
English
1
0
3
130
Papa Heme
Papa Heme@Papa_Heme·
@AppleHelix I agree but control arm AQUILA had worse OS than recent myeloma trials (PERSEUS) How can it be people with smoldering are living shorter time than people with myeloma OS results do not apply US where pts get better imaging surveillance,access to better Rx IF develop myeloma
English
6
3
35
2.7K
Jing Liang 🇺🇦 retweetledi
Alasdair Munro
Alasdair Munro@apsmunro·
If you’ve seen this chart, it’s important to note this is NOT the current growth rate of the Ebola outbreak This is the speed we are uncovering the outbreak *that has already occurred* It’s already enormous, and we’re only just unearthing it
Alasdair Munro tweet media
English
3
25
86
27.3K
Jing Liang 🇺🇦 retweetledi
Annals of Oncology
Annals of Oncology@Annals_Oncology·
🆕 Article in press: Dato-DXd combo with durvalumab as 1L treatment for unresectable locally advanced or metastatic triple-negative breast cancer: results from the phase Ib/II BEGONIA study annalsofoncology.org/article/S0923-…
Annals of Oncology tweet media
English
0
25
55
14.6K
Jing Liang 🇺🇦 retweetledi
Rob Shaffer
Rob Shaffer@ShafferBiotech·
Just added bsAbs, mRNA, and oligo estimators (ASO + siRNA) to biologics.tools 2 surprises I did not know: 1) ds siRNA actually slightly less $/g than ss ASO due to cheaper "typical" 2' chemistry. 2) LNP formulation is roughly same cost as making the mRNA DS Still free!
Rob Shaffer@ShafferBiotech

Built a free biologics COGS estimator covering mAbs, recombinant proteins, ADCs, and C&GT, with GMP/Tox/RG options. For founders, investors, and BD folks who need a defensible cost figure for pitching/evaluating. Feedback welcome; still iterating! biologics.tools

English
0
4
25
4K
Rahul Banerjee, MD, FACP
Rahul Banerjee, MD, FACP@RahulBanerjeeMD·
Grateful to my @fredhutch @UWMedicine mentors for the privilege to work and learn here, and to my family for putting up with my Epic logons & manuscript revisions at all hours 🙏🏽 Now I have no excuses for getting lost in building basements or stairwells anymore...
Rahul Banerjee, MD, FACP tweet media
English
51
5
176
9.7K
Jing Liang 🇺🇦 retweetledi
gilberto lopes
gilberto lopes@GlopesMd·
Remarkable is an understatement! #ASCO26 7-yr update (Abstract 8502, CROWN): 1L lorlatinib in advanced ALK+ NSCLC — median PFS STILL not reached at 7 years, the longest ever in advanced NSCLC. 7-yr PFS 55% vs 3% for crizotinib. Clear 24 mo and you have a 79% chance of being PFS at yr 7. No new brain progression after 30 mo. #LCSM @OncoAlert @OncBrothers @StephenVLiu @Jani_Chinmay @asco @myESMO @glopesmd @SylvesterCancer @latinamd @iaslc @COlazagasti
gilberto lopes tweet media
English
2
24
68
17.5K
Jing Liang 🇺🇦
Jing Liang 🇺🇦@AppleHelix·
@SimonDBarnett Simon, If I read correctly, you just implicitly said AI x Bio needs to be in the drug development business, not service business
English
1
0
4
528
Simon Barnett
Simon Barnett@SimonDBarnett·
I don't understand the original chart. Well, I do, it's an attempt to dunk on Ron. It's just that I'm very, deeply tired of this discussion around AI x Bio companies. AI is not a separate technology, it is another tool in the armamentarium. As such, there aren't AI x Bio companies. There are Bio companies who increasingly route their operating expenses through compute tokens. That's it. I could post 100 charts of near-dead trad-bio companies for every 1 "AI x Bio" chart. Neither would be useful to have a discussion. There will be (are) amazing biotechs who lean increasingly into ML to develop their assets. There will also continue to be great companies who don't care to do this at all. So the business model is the same for right now. Just make good drugs, and if you think you have a structural edge by pushing more of your spend into ML, then please do so because I think it's the right move. Yes, there are platform companies that don't want to make drugs, but that's a very different debate than the one I've made here. There are no AI drugs. There are just drugs.
English
3
1
36
1.1K