Misinformation reaches 1 billion people in seconds. Debunking it takes work. At #ACC26, former FDA Commissioner and DCRI founder, Rob Califf encouraged clinicians to engage in new media — or risk losing the battle for public trust without ever competing. #HealthPolicy #CardioSky
True, but what's the cost?
Normal CMR = 3.5% 10-yr MACE.
CMR+ MI = 47% MACE (Konst 2023, n=252).
If we're missing small MIs, they're clinically benign. www.ahajournals.org/doi/10.1161/...
Bottom line:
Maybe “vulnerable plaque” features (rupture, erosion, intraplaque hemorrhage) are just common findings in CAD patients and don’t necessarily identify THE culprit that caused THIS event?
Are we over-interpreting OCT? Under-trusting CMR?
Thoughts?
7/7
ISSUE #4: Ischemic CMR Without OCT Culprit
31% had definite infarction on CMR but NO identifiable culprit on OCT.
Possible explanations:
a) OCT missed it (sensitivity issue)
b) The culprit already healed/resolved
c) Vasospasm
d) Embolism
e) It wasn’t atherothrombotic
6/7
ISSUE #3: The Discordant Cases
17 patients had OCT culprit + NON-ISCHEMIC CMR (myocarditis pattern, Takotsubo, etc.)
Final diagnosis:
• 7 → MINOCA mimic (CMR as arbiter)
• 10 → MI (Despite non ischemic CMR findings)
How is this “complementary”?
5/7
ISSUE #2: OCT Culprit + Normal CMR
35% (43/124) had OCT “culprit” lesions but COMPLETELY NORMAL CMR - no infarction, no injury, nothing.
If there’s a “culprit” that caused MI, where’s the infarct?
Are these just bystander plaques?
4/7
ISSUE #1: Territory Mismatch
When BOTH OCT culprit AND ischemic CMR were present, they matched only 81% of the time.
19% had the OCT “culprit” in a DIFFERENT coronary territory than the actual infarct.
How is the OCT lesion the “culprit” if it’s not in the infarcted territory?
3/7
First, the "good news": 79% had an abnormality on one or both tests, vs 44% with OCT alone or 63% with CMR alone.
Sounds complementary, right? But let's look closer...
2/7
🧵 Are OCT and CMR really "complementary" in MINOCA? Or are they telling us conflicting stories?
Reynolds et al (Circulation 2026) combined both modalities in 284 MINOCA patients. The results raise more questions than answers.
📄 www.ahajournals.org/doi/abs/10.1...
1/7
The paradox: if baseline screening had used CMR and truly excluded ALL pre-existing LVT, I'd expect the 1-month incident rate to be lower than 16.6%, not higher. The high rate despite TTE exclusion is itself indirect evidence that prevalent CMR-occult thrombi entered the trial.
Your point about early thrombus formation actually reinforces this concern. If thrombi form within days of MI — before they're TTE-visible but already CMR-detectable — then the 1-month CMR is capturing a mix of truly incident and pre-existing but occult thrombi.
Great point — but APERITIF actually did enroll high-risk patients. The control arm LVT rate was 16.6% on CMR at 1 month — well above the ~6% all-comer STEMI and ~12% anterior STEMI background rates from the Bulluck 2018 meta-analysis.
Although the use of an ultrasound-enhancing agent improves the sensitivity of echocardiography, it is still substantially lower compared with LGE CMR.
www.ahajournals.org/doi/10.1161/...
Takeaway: Before concluding low-dose rivaroxaban doesn't prevent LVT post-anterior STEMI, we need a trial gating enrollment on baseline LGE CMR — not TTE. Using a ~30% sensitivity screen to test a CMR-detected endpoint is a fundamental design mismatch. The question remains open.
The trial was already underpowered (lower-than-expected event rates, wide CIs: −8.9% to +3.2%). Add in LGE CMR-occult baseline LVT contaminating both arms, and the null result becomes even harder to interpret. A negative trial ≠ no effect.
If patients with pre-existing LVT were inadvertently enrolled, you've essentially randomized established thrombus to treatment vs. no treatment — a fundamentally different question than prevention. This dilutes the treatment effect and biases toward the null.
Eligibility required no LVT on baseline TTE. The authors themselves cite TTE sensitivity for LVT detection at just ~29% vs LGE CMR. That means up to 70% of pre-existing thrombi could have been missed at enrollment — and those patients randomized anyway.
APERITIF tested low-dose rivaroxaban+ DAPT vs DAPT alone to prevent LV thrombus after anterior STEMI. Primary endpoint: LVT on contrast-enhanced CMR at 1 month. Result: no significant difference (13.7% vs 16.6%, p=.34).But there's a screening problem worth discussing. jamanetwork.com/journals/jam...
academic.oup.com/eurheartj/ad...
@escardio.bsky.social
#Cardiosky #WhyCMR #Epeeps #Medsky
Please DM me if you would like a full-text PDF of the paper!
Thank you, @toddneale.bsky.social, for your excellent reporting on our EHJ paper!
@tctmd.bsky.social @escardio.bsky.social @paragbawaskar.bsky.social
Query sarcoid, normal #WhyCMR, so then should we do FDG-PET? Not much benefit.
Read our accompanying editorial for our thoughts on the important paper –
academic.oup.com/ehjcimaging/...
@paragbawaskar.bsky.social #CardioSky #MedSky #Sarcoidosis
In patients with suspected cardiac sarcoidosis, cardiac FDG-PET is recommended after a normal CMR if there is a high clinical suspicion. What are the data supporting this recommendation? #CardioSky #MedSky #Sarcoidosis
This paper reports the coexistence of cardiac sarcoidosis and arrhythmogenic cardiomyopathy in 5 patients.
heart.bmj.com/content/earl...
#CardioSky #MedSky
In honor of Dr. Braunwald’s 95th birthday. The number of areas he has touched in cardiology is staggering, and number of cardiovascular professionals is unquantifiable. @harvardmed.bsky.social
Amazing achievement by rising star and post-doc in my lab @paragbawaskar.bsky.social, winner of the prestigious 2024 Melvin Judkins Early Career Investigator Award at #AHA24!! Congratulations!!!
@ahascience.bsky.social
It’s great that journals are coming over here, but what I have really missed after my departure from Twitter 2 years ago is this, the authors themselves presenting their works and discussing it with peers. #cardiosky #medsky #emimcc
Figure 2 from the paper. Incidence of embolism in left ventricular (LV) thrombus patients compared with matched non-LV thrombus patients. Kaplan-Meier curves demonstrate the cumulative incidence of the composite embolic end point in the LV thrombus (in red) and in the matched non-LV thrombus (in blue) groups. Note the significant difference in the cumulative incidence of embolic events between the 2 groups.
In 2019, we wrote a paper showing that patients with left ventricular thrombus have a long-term risk of embolism, extending to at least 8 years.
We were puzzled by the finding…
#CardioSky
www.ahajournals.org/doi/10.1161/...
Finally, many thanks to @cshenoy.bsky.social for his mentorship, our outstanding team for their contributions, and the editors and reviewers at Circulation for all their efforts to make our paper better!
