#523 Hotcakes: Left Atrial Appendage Closure vs AC for Afib, Apixaban vs Rivaroxaban for VTE, Intensive LDL Targeting, GLP1s and Substance Use Disorders, and more

April 27, 2026 | By Rahul Ganatra

Audio

Video

Transcript available via YouTube

Join us as we review recent practice-changing articles on left atrial appendage closure vs AC for AFib, apixaban vs rivaroxaban for VTE, intensive LDL targeting, GLP1s and substance use disorders, and more! Fill your brain hole with a delicious stack of hotcakes! Featuring Paul Williams (@PaulNWilliamz), Shani Herzig (@ShaniHerzig) Rahul Ganatra (@rbganatra), and Matt Watto (@doctorwatto).

Claim CME for this episode at curbsiders.vcuhealth.org!

Patreon | Episodes | Subscribe | Spotify | YouTube | Newsletter | Contact | Swag! | CME

Show Segments

• Intro, disclaimer
• Left atrial appendage closure vs anticoagulation in AF
• Apixaban vs rivaroxaban for VTE
• Intensive LDL targeting in ASCVD
• GLP1s and substance use disorders
• FDA approves Orforglipron
• E-cigarettes and cancer
• Early AM blood draws and sleep quality
• Outro

Hotcake #1: Left Atrial Appendage Closure vs Anticoagulation for Afib (Watto)

Doshi SK et al; CHAMPION-AF Investigators. Left Atrial Appendage Closure or Anticoagulation for Atrial Fibrillation. N Engl J Med. 2026 Mar 28. doi: 10.1056/NEJMoa2517213. Epub ahead of print. PMID: 41910347.

What was the research question? Is left atrial appendage closure or occlusion (LAAO) noninferior to NOAC therapy for prevention of cardiovascular death, stroke, or systemic embolism?

Why is this study important? Oral anticoagulation is the standard of care for stroke prevention in AF, with LAAO being reserved for patients at high bleeding risk. Whether LAAO can achieve comparable efficacy with less bleeding in patients who are otherwise anticoagulation-eligible is unknown.

How was the study done? This was an interim analysis of a randomized noninferiority trial with a margin of 4.8%. Patients were randomized 1:1 to LAAO with the Watchman FLX device or continued NOAC therapy, with drug choice left to treating clinicians. The primary efficacy outcome was a composite of cardiovascular death, stroke, or systemic embolism at 3 years, and the primary safety outcome (tested for superiority) was non–procedure-related bleeding.

Who were the patients? 3,000 adults (mean age: low 70s, mean CHA2DS2-VASc ≈3.5, and mean HAS-BLED ≈1.3) with nonvalvular AF were randomized. Patients with recent stroke, recent major bleeding, or clear contraindications to anticoagulation were excluded.  

Top-line results: This was a positive trial (for noninferiority): at 3 years, CV death, stroke, or systemic embolism occurred in 5.7% vs 4.8% (an absolute increase of 0.9% with LAAO, with 95% CI -0.8% to 2.6%, below the NI margin of 4.8%). Non-procedure related bleeding occurred in 10.9% vs 19.0% (HR 0.55, 95% CI 0.45 – 0.67). Device-related thrombus occurred in ~5% of patients in the LAAO group, requiring resumption of anticoagulation in 2% of those. Procedural complications included pericardial effusions requiring intervention in ~1% (consistent with contemporary practice).

Learning points & limitations: The top-line result of LAAO being noninferior to NOACs for MACE and superior for bleeding is misleading. First, the large noninferiority margin (4.8%) is itself a bias towards noninferiority and permits a clinically meaningful excess risk of stroke. Second, excluding periprocedural bleeding from the definition of the primary safety outcome is also a source of bias towards superiority in the safety outcome. Third, approximately 200 patients in the NOAC group crossed over to the LAAO group, which is more than the total number of primary outcome events and an additional source of bias towards noninferiority.   

Bottom line: I am worried that this headline-grabbing trial will cause harm by expanding LAAO from a narrow “contraindication-only” strategy to all patients with AF despite a signal for an increase in stroke with only a marginal reduction in bleeding. 

Hotcakes rating: 2.5 out of 5

Further reading:

• John Mandrola critiques CHAMPION-AF on the Sensible Medicine Substack https://www.sensible-med.com/p/champion-af-breaks-almost-every-rule accessed 5 April 2026

Hotcake #2: Apixaban vs Rivaroxaban for VTE (Rahul)

Castellucci LA, Chen VM, Kovacs MJ, et al. Bleeding risk with apixaban vs. Rivaroxaban in acute venous thromboembolism. N Engl J Med. 2026;394(11):1051-1060.

What was the research question? Does rivaroxaban carry a higher bleeding risk than apixaban when used for VTE treatment for 3 months?

Why is this study important? Prior studies comparing apixaban and rivaroxaban to VKAs for VTE treatment suggested a higher bleeding risk with rivaroxaban than apixaban, but no direct head-to-head comparison of the two has been available until now. Current professional society guidelines recommend DOACs for VTE with no preference among them.

How was the study done? This was a pragmatic trial using a PROBE design (prospective, randomized, open-label, blinded endpoint) carried out in Canada, Australia, and Ireland from 2017 – 2025. Patients with symptomatic VTE were randomized 1:1 to apixaban or rivaroxaban, and outcomes were adjudicated by a committee blinded to treatment group. The primary outcome, “clinically relevant bleeding,” was a composite of “major bleeding” and “clinically relevant nonmajor bleeding.” Secondary outcomes were death, VTE recurrence, and adherence. 

Who were the patients? 2,700 patients (mean age high 50s, half women, half-and-half DVT vs PE, 75% unprovoked, 15% recurrent) were enrolled, mostly in Canada. Patients with GFR <30 ml/min/1.73 m², obesity (body weight >120 kg), active cancer, and Child-Pugh B or C cirrhosis were excluded. Comorbidities were not reported. 

Top-line results: This was a positive trial: at 3 months, clinically relevant bleeding occurred in 3.3% of patients on apixaban vs. 7.1% of patients on rivaroxaban (ARR ~4%, NNT 25, RRR ~50%). VTE recurrence happened in ~1% of both groups, despite adherence being about 10% lower with apixaban (the BID medication).  

Learning points & limitations: These results are vulnerable to selection bias because the study population was somewhat narrow (only enrolled ~30% of eligible patients and excluded patients with active cancer, renal or liver disease, and weight >120 kg). However, I am reassured that there was no difference in recurrence despite adherence being about 10% lower in the apixaban group (65% vs 75%). Also, figure 2 makes it clear that the bleeding risk with rivaroxaban is highest during the initial 3-week high-intensity phase, and the accrual of bleeding events over time looks fairly similar to apixaban after that. This makes me wonder if bleeding risk might be more similar if we forgo the initial high-intensity phase (e.g. using enoxaparin instead, or skipping it altogether, as in AF).

Bottom line: This was a nice, clean trial – well-done and without sources of chance or bias to threaten the authors’ conclusions. These results might not generalize to AF – but fortunately the forthcoming COBRRA-AF trial should answer this. But for VTE, the answer is clear: I prefer apixaban.  

Hotcakes rating: 4 out of 5

Further Reading:

• Study Details | NCT04642430 | COmparison of Bleeding Risk Between Rivaroxaban and Apixaban in Patients With Atrial Fibrillation | ClinicalTrials.gov
• Oral Apixaban for the Treatment of Acute Venous Thromboembolism | New England Journal of Medicine 
• Oral Rivaroxaban for Symptomatic Venous Thromboembolism | New England Journal of Medicine 

Hotcake #3: Intensive LDL Cholesterol Targeting in Atherosclerotic Cardiovascular Disease (Paul)

Lee Y-J et al.  2026.  Intensive LDL cholesterol targeting in atherosclerotic cardiovascular disease.  NEJM 394(14): 1365-1375.

What was the research question? Is an LDL-C level <55 mg/dL superior to <70 mg/dL for preventing recurrent MACE in patients with established atherosclerotic cardiovascular disease (ASCVD)?

Why is this study important? ASCVD guidelines recommend an LDL-C target of <55 mg/dL for secondary prevention in patients at “very high risk,” defined as ≥2 major ASCVD events or 1 minor ASCVD event and ≥2 high risk conditions (age ≥65, CABG or PCI, current tobacco use, diabetes, CHF, HTN). However, evidence to support this target is lacking.

How was the study done? In this open-label superiority trial, patients with ASCVD were randomized 1:1 to intensive (<55 mg/dL) or conventional (<70 mg/dL) LDL-C targeting. Within each group, patients were further randomized to high-intensity statin monotherapy or a statin plus ezetimibe; patients in the statin-only group were further randomized to atorvastatin or rosuvastatin. The primary endpoint was a composite of cardiovascular death, MI, stroke, any revascularization, or hospitalization for unstable angina at 3 years. Many safety endpoints were also tested.

Who were the patients? 3,048 patients (mean age: mid 60s, mostly men) in South Korea with documented ASCVD (previous ACS, stable angina, coronary or other arterial revascularization, stroke or TIA, or PAD) were included. Patients with an LDL-C <70 mg/dL without medications were excluded.  

Top-line results: This was a positive trial: targeting an LDL-C <55 mg/dL resulted in a ~3% absolute risk reduction (~30% relative risk reduction) in cardiovascular events at 3 years compared with targeting <70 mg/dL. The intensive group achieved a median LDL-C of 56 mg/dL, compared with 66 mg/dL in the conventional group at 3 years. There were no concerning safety signals.

Learning points & limitations: There are not obvious sources of chance or bias to threaten the authors’ conclusions. Whether the impressive absolute and relative risk reductions seen here are really explained by a difference of just 10 mg/dL in LDL-C at 3 years is an open question. Because PCSK9 inhibitors were used very infrequently in this trial (<5% of patients), it suggests that the cardiovascular benefit comes from the LDL-C target, not the drugs used to achieve it. And, even though many in the intensive group didn’t quite reach the goal, this trial suggests that when it comes to LDL-C, if you shoot for the moon and miss, you still land among the stars.

Bottom line/Hotcakes rating: 4 out of 5 

Further reading:

• Paving the Road toward Targeted Lipid Lowering | New England Journal of Medicine 

Hotcake #4: GLP1a’s and SUD (Shani)

Cai M, Choi T, Xie Y, Al-Aly Z. Glucagon-like peptide-1 receptor agonists and risk of substance use disorders among US veterans with type 2 diabetes: cohort study. BMJ. 2026 Mar 4;392:e086886.

What was the research question? Is initiation of glucagon-like peptide-1 (GLP-1) receptor agonists associated with A) reduced risk of substance use disorders (SUDs) in people without a prior history, and B) reduced risk of SUD-related complications among people with a preexisting SUD?

Why is this study important? Animal studies suggest GLP-1s act on mesolimbic reward pathways to reduce drug reinforcement, and observational studies have linked GLP-1s use to lower risk of new and recurrent alcohol, tobacco, and cannabis use disorders. This is the first study to evaluate the effectiveness of GLP-1s for preventing a broad array of SUDs or improving hard clinical outcomes in people with established SUDs. 

How was the study done? This was a retrospective cohort study using the target trial emulation framework to compare US veterans with diabetes who initiated a GLP-1 receptor agonist versus a sodium-glucose cotransporter-2 (SGLT-2) inhibitor. In protocol 1, incident SUDs (alcohol, cannabis, nicotine, cocaine, opioids, other SUDs, and a composite) were compared between groups. In protocol 2, SUD-related complications (ED visits, hospitalizations, and mortality; drug overdose, and suicidal ideation or attempt) were compared between groups. 

Who were the patients? 600,000 US veterans with type 2 diabetes without contraindications to GLP1s or SGLT2s were included across the 12 main analyses performed. Included patients had a mean age in the mid 60s and were 90% male. In protocol 1, patients with a prior diagnosis of any SUD were excluded; in protocol 2, patients were required to have a prior diagnosis of SUD in the preceding year. Patients were followed for up to 3 years. 

Top-line results: This was a positive study: compared with SGLT-2 inhibitors, initiating a GLP-1 (mostly semaglutide) was associated with lower risk of all 7 SUDs tested (6-7 fewer new diagnoses per 1,000 users over 3 years). Among patients with preexisting SUD, initiating a GLP-1 was associated with lower risk of all 5 SUD-related complications tested (~12 fewer events per 1,000 new users over 3 years).    

Learning points & limitations: While any retrospective study will be vulnerable to residual confounding, several design features of this study strengthen causal inference: the active comparator / new user design, propensity score matching, steps to mitigate immortal time bias, prespecified falsification analysis, and multiple sensitivity analyses. The early separation of the cumulative incidence functions also has biologic plausibility, given GLP-1s’ rapid onset of action. One limitation is the potential for misclassification of SUD-related group assignment due to reliance on ICD-10 codes. Also, the cumulative adherence in both groups was only about 50% by the end of follow-up, which may bias towards underestimation of the true effect. 

Bottom line: This is about as good as it gets for an observational study of drug effects. That said, randomised trials are still needed, including people without diabetes and those receiving standard therapies for OUD and AUD, to test whether GLP-1 receptor agonists add benefit beyond standard addiction therapy. 

Hotcakes rating: 4 out of 5

Further reading:

• Metabolic medicines and addiction: what GLP-1 receptor agonists might add to substance use care | The BMJ 

Hot take #1: Orforglipron (Watto)

FDA Approves First New Molecular Entity Under National Priority Voucher Program

Summary: The US FDA approved orforglipron (a small-molecule, non-peptide oral GLP1 partial agonist) for weight loss. It is taken once daily with or without food, which differs from semaglutide tablets (taken once daily 30 minutes before food). It is available in 0.8 mg, 2.5 mg, 5.5 mg, 9 mg, 14.5 mg, or 17.2 mg oral tablets. Patients start at the lowest dose and progress monthly as needed for weight loss, with maintenance doses of 5.5 to 17.2 mg daily expected based on clinical trials (ATTAIN-1 and ATTAIN-2, linked below).

Bottom line: Oral orforglipron resulted in an average 9-11% body weight loss (vs 2% with placebo) at 72 weeks. It is not as potent as injectable GLP1 agonists (which lead to approximately 15-20% body weight loss), and we don’t yet have an active comparison to semaglutide tablets. Cardiovascular outcomes trials are ongoing, but data so far suggests orforglipron is associated with a 1-2% reduction in A1C, and I expect future approval for diabetes.

Further reading:

• FDA’s Commissioner’s National Priority Voucher Program https://www.fda.gov/industry/commissioners-national-priority-voucher-cnpv-pilot-program accessed 5 April 2026 
• Orforglipron package insert https://pi.lilly.com/us/foundayo-uspi.pdf accessed 5 April 2026
• Orforglipron, an Oral Small-Molecule GLP-1 Receptor Agonist for Obesity Treatment | New England Journal of Medicine 
• Orforglipron, an oral small-molecule GLP-1 receptor agonist, for the treatment of obesity in people with type 2 diabetes (ATTAIN-2): a phase 3, double-blind, randomised, multicentre, placebo-controlled trial – ScienceDirect 

Hot take #2: E-cigarettes and cancer (Paul)

Stewart BS et al.  The carcinogenicity of e-cigarettes: a qualitative risk assessment.  Carcinogenesis, 2026, 47, 1-14.

Summary: This is a narrative review of multiple types of evidence related to the potential carcinogenicity of electronic cigarettes. The review includes laboratory studies suggesting tissue injury from e-cigarettes that may predispose to malignant transformation, as well as biomarker studies, case reports, and animal studies. The authors conclude: “E-cigarettes delivering nicotine are likely to be carcinogenic to humans…The burden of disease cannot be quantified from data presently available.”

The problem is that this is not how the lay media reported on this. One headline reads: “Strongest evidence yet that vaping likely causes cancer,” and a similar headline is in The Guardian. It’s worth noting that a narrative review differs from a systematic review in that here, expert opinion guides evidence selection and interpretation, rather than a systematic or reproducible process. This is not to say that narrative reviews are without value, but in this case, assuming causality (as in the headlines above) that has not yet been demonstrated is both dangerous and unwarranted.

Bottom line: When you see a dramatic medical headline, make sure you go to the source and read the article(s) it is based on. This will allow you to be better informed and better able to discuss with your patients. Are electronic cigarettes unhealthy? Probably. But whether they are less carcinogenic than cigarettes remains an open question, and more research is needed to fully understand these risks.

Further reading:

• Does Vaping Cause Cancer? – Gideon M-K: Health Nerd 

Hot take #3: Early AM blood draws (Rahul)

Chakabva MS, Flanagan M, Drouin M, et al. Effect of timing of morning blood draws on sleep quantity and quality in hospitalized patients. Journal of Hospital Medicine. Published online February 4, 2026:jhm.70269 

Summary: In this quasi-experimental study done at a single hospital in Indiana, 130 medical inpatients admitted for >3 days to one of two similar medical units had AM blood draws done at either 4 AM or 6 AM, and self-reported sleep quality and quantity were compared between groups. Group assignment was dictated by bed availability at the time of admission (hence “quasi-random”). Patients in the 6 AM group reported higher subjective sleep quality on the Richards-Campbell sleep questionnaire (+11 points on a 0-100 scale) and an additional 1 hour of sleep (6 vs 7 hours). 

Bottom line: There are limitations: this was a small study with lots of unmeasured confounders, and the authors did not examine impact on throughput metrics (e.g. length of stay, discharges by noon, bed turnover, etc). That said, I am inclined to believe these unsurprising results, which suggest an easy way to improve sleep quality for inpatients: enough with the early morning labs! 

Further reading:

• Evaluation of the Order SMARTT: An Initiative to Reduce Phlebotomy and Improve Sleep‐Friendly Labs on General Medicine Services – Tapaskar – 2020 
• Timing of Blood Draws Among Patients Hospitalized in a Large Academic Medical Center 

Links

Links are included in the show notes above.

Goals

Listeners will review recent practice-changing articles and medical news.

Learning objectives

After listening to this episode listeners will…

1. Decide between left atrial appendage closure and anticoagulation for AF
2. Choose the safer DOAC between rivaroxaban and apixaban for VTE
3. Describe the benefits of intensive vs conventional LDL-C targets in ASCVD

Disclosures

The Curbsiders report no relevant financial disclosures. 

Citation

Ganatra RB, Herzig S, Williams PN, Watto MF. “#523 Hotcakes: Left Atrial Appendage Closure vs AC for Afib, Apixaban vs Rivaroxaban for VTE, Intensive LDL Targeting, GLP1s and Substance Use Disorders, and more”. The Curbsiders Internal Medicine Podcast. https://thecurbsiders.com/episode-list April 27, 2026.