💯
No FDA is better than this FDA.
At least return it to its pre Kefauver amendment days where its main focus was safety, not efficacy.
Everyone is not clear. And biotech world has created their own rules on what qualifies as working.
Best part is you get to pay for it.
Another must read on coronary CT from @anish_koka
The great debate on CT Plaque Volume Measurement - I took the opposing side in a live debate at an excellent local Indian Restaurant (Sura Bistro).
Thanks to @RogueRad for organizing.
My opponents side was unfortunately not recorded (I do summarize some of his points)
x.com/i/article/203009521099…
DOJ is targeting a doctor who treated over 100,000 patients during the pandemic with monoclonal antibodies.
@RonElfenbeinmd is now facing 50 years in prison over a bogus coding dispute.
@AGPamBondi should dismiss!
Another warrior for the truth sighted.
Ryan fin twit has some thoughts
Only looking out for greater good I’m sure.
Nah. I’ve seen enough.
Medicare should negotiate with drug companies the way they negotiate with doctors.
Just let them know what an appropriate price is.
Couldn’t sum it up better myself.
Yup. The public should wake up.
Medicare should have a global budget for pharma spend and set prices.. just like they do for physicians.
I see no other path given the current rigged system - the Ds and Rs that will push back —> all pharma funded.
lol. Why were they called? Who decided ? There is no way VP makes that decision given coverage by many of these outlets.
The origin story of this news cycle is uniqure CEO running to media after a rejection.
But you knew that. Again, all bad faith actors here.
What you’re seeing playing out by science media has little to do with actual science or how it’s supposed to be done.
These are friends of the Biden administration, political actors masquerading as journalists hell bent on proving how dysfunctional the FDA is with their former friends no longer in charge.
This campaign was set in motion the moment Marks/Fauci et. al. were forced out.
This is revenge.
I agree that randomized clinical trials, not comparisons to historical controls, needed for vast majority of approvals. I can only imagine how hard this could be for patients w/ terrible illnesses & their families, but truly helping such patients requires knowing what works.
Unless Marty colors within the lines the stat news reporters draw from now on, he’s out as well.
Prediction: Even if he does what they say and approves everything the Peter Marks FDA wanted, they wont stop until they claim his scalp as well.
My favorite illustration of the problems with propensity-score matching comes from Facebook
Their researchers had to use a massive set of 3,719 covariates to come close to matching the estimate produced by an experiment they ran
'Yeah, PSM works, just get 3,719 covariates, bro'
New Unreported Truths, about the FDA’s Catch-22…
Nothing to see here, folks! Move along. Thanks for posting, Anish. GKS
Who could have predicated that NPs who lobbied for independent practice without oversight to serve Rural and primary care shortages are mostly working in Med Spas, Addiction/Psych, Cardiology😳 with no supervision in FL.
Why CT-FFR Doesn't Work And Why Its Reference Standard Is Suspect Too
https://x.com/i/status/2029893569444524406 via @anish_koka
Finally someone entered boldly into this field.
MUST BE RED with open mind and will to challenge mainstream narrative.
Free , no paywall, so you know exactly who the FDA staffer is that’s speaking on condition of anonymity.
Sorry, not good faith.
Questionable tech, but hospital admins see a great marketing opportunity
Excellent! MUST READ read for anyone providing cardiac care to patients.
The visual is a Rube Goldberg machine - “an intentionally over-engineered contraption designed to perform a simple task in a comically complex, chain-reaction manner.”
R to @anish_koka: 14/14: Right question: Does acting on CT-FFR improve outcomes? No idea, and given issues discussed, doubtful a well done trial would be positive.
Until a positive trial on hard endpoints, it's a business model dressed up as diagnostic test.
R to @anish_koka: 12/14: Bottom line: CT-FFR is sophisticated but oversold. Rests on flawed assumptions, questionable reference standard, modest accuracy. No outcomes trial showing patient benefit. In stable CAD, revasc itself lacks mortality benefit (ORBITA/ISCHEMIA).
R to @anish_koka: 13/14: Good cardiologists use it as one data point, but push is to defer to the number. Reimbursement rewards more testing/procedures. Patients may demand fixes for "bad" numbers.
R to @anish_koka: 11/14: Incentives: Positive CT-FFR (often wrong) → cath → fees/stents. Belief cultivated via guidelines/pubs, but structure favors revenue over evidence.
R to @anish_koka: 9/14: PLATFORM trial: CT FFR reduced invasive angio by 61%, but standard of care arm was not offered standard of care non-invasive tests before cath. Regardless, "Fewer caths" ≠ better care if test is inaccurate ~50% time.
R to @anish_koka: 10/14: Business model drives adoption. FDA clearance on accuracy, not outcomes. Guideline inclusion (Class IIa) led to reimbursement: $1,017 for FFRCT + $950 for Plaque Analysis (2026). Centers earn >$2k/patient. Lobbying by ACC/SCCT—overlaps w/ HeartFlow KOLs.
R to @anish_koka: 8/14: What the Diagnostic Performance Data Actually Show : Independent reviews confirm major scatter, weak performance near decision points of clinical interest (FFR 0.6-0.8)
R to @anish_koka: 7/14: It's a black box: proprietary algorithm, no independent audit. Validation vs. invasive FFR: NXT trial sensitivity 86%, specificity 79%. But correlation r=0.73—explains only 53% variance. PPV ~49% in UK audit; worse (35%) for 50-69% stenoses. Poor agreement near 0.80 threshold.
R to @anish_koka: 6/14: CT-FFR builds on this shaky base. From CCTA, it reconstructs 3D model, applies CFD to simulate hyperemia, outputs FFRCT. Assumptions: fixed microvascular resistance drop (~75%), myocardial mass-based flow, no collaterals. Fails in microvascular disease—common in CAD patients.
R to @anish_koka: 5/14: Replication issues: Trials without original FFR developers (Pijls/De Bruyne) failed. FUTURE trial stopped early for mortality signal in FFR group. FLOWER-MI: no benefit in STEMI multivessel. FAME 3: FFR-PCI inferior to CABG. Data anomalies in FAME papers raise questions—unexplained clustering at FFR=0.5.
R to @anish_koka: 4/14: FFR's mechanistic reasoning is weak, predicting weak clinical evidence. Key trials: FAME (2009) showed FFR-guided PCI better than angio-guided, but comparator was overly aggressive stenting. Investigators had IP stakes. FAME 2 (2012) vs. med therapy: benefit driven by fewer revasc, not death/MI; high crossover.
R to @anish_koka: 2/14: To see why CT-FFR is doubtful, understand invasive FFR first. FFR measures the physiological significance of a coronary stenosis: max blood flow with stenosis divided by max flow if artery were normal. During cath, a pressure wire passes the lesion, adenosine induces hyperemia, FFR = Pd/Pa (distal/aortic pressure).
3/14: Normal FFR=1.0; threshold=0.80 for "significant." But it assumes uniform minimal microvascular resistance during hyperemia. This fails in real patients w/ diabetes, hypertension, LVH, etc.—microvascular dysfunction blunts response, making FFR unreliable. Healthy microvasculature can make mild lesions look severe, and vice versa.
CT-Derived Fractional Flow Reserve (HeartFlow FFRCT): A Critical Look 🧵
1/14: There's a familiar pattern in cardiology: take a test of uncertain validity, layer sophisticated tech on top, and present it as a breakthrough. CT-derived fractional flow reserve (HeartFlow FFRCT) fits this template. It's a proprietary, off-site analysis approximating invasive FFR—a number with shaky foundations.