NEJM This Week — June 25, 2026

This week's high-yield read is the BEAT-TB trial — a 6-month all-oral, delamanid-based regimen was noninferior to 9-month standard care for rifampicin-resistant TB and, crucially, can be used in children and pregnant patients (below). Three other trials are summarized for awareness but flagged lower-yield for general IM/ID: AVANT GUARD (first-line pulsed field ablation beats drugs for persistent AF — an EP procedural decision), TRACTION (hospital-policy tranexamic acid cuts surgical transfusion without raising clots — a perioperative decision), and a 10-year CAR T-cell follow-up (durable remissions, but a single-arm oncology study). A peanut-allergy review and a high-yield diabetic-amyotrophy teaching case round out the issue.

Infectious DiseaseBEAT-TB · RCTHigh yield · ID

A Pragmatic Trial of a 6-Month Strategy for Rifampicin-Resistant Tuberculosis

In rifampicin-resistant TB — including fluoroquinolone-resistant disease, children, and pregnancy — is a 6-month all-oral regimen noninferior to 9-month standard care?

Population: 403 South Africans ≥6 yr with pulmonary rifampicin-resistant TB; 21% fluoroquinolone-resistant, 51% HIV-positive, 9 pregnant, 30 children 8–17 yr.
Intervention: 6 months of bedaquiline + delamanid + linezolid + levofloxacin or clofazimine (or both).
Comparison: 9-month standard-of-care RR-TB regimen current in South Africa.
Outcome: Successful outcome 86.1% vs 86.0% (adjusted difference −0.2 pp, 95% CI −6.9 to 6.5; P=0.001 for noninferiority, 10-pp margin). Grade ≥3 adverse events 31.2% vs 37.0%; linezolid-related anemia the most common SAE; 10 deaths in each group.

Clinical takeaway. A delamanid-based 6-month regimen matches standard care for cure with a similar safety profile. Its decisive advantage over pretomanid-based BPaL/BPaLM is usability in the populations where pretomanid is contraindicated — children and pregnant/breastfeeding patients. Watch linezolid toxicity (anemia, peripheral neuropathy, optic neuritis) and emerging bedaquiline resistance, which clustered in fluoroquinolone-resistant isolates. What this changes: extends the short-regimen paradigm (Nix-TB, ZeNix, TB-PRACTECAL) to priority populations — already in the June 2024 WHO RR-TB update.

Cardiology / EPAVANT GUARD · RCTLower yield · IM/ID

Pulsed Field Ablation as Initial Therapy for Persistent Atrial Fibrillation

In previously untreated persistent AF, does first-line pulsed field ablation vs antiarrhythmic drugs improve 12-month treatment success?

Population: 310 adults with previously untreated symptomatic persistent AF (continuous 7–365 days); +100 PFA-assigned for the safety analysis.
Intervention: First-line PFA (PV isolation + posterior-wall ablation, pentaspline FARAWAVE), 2:1.
Comparison: Guideline-directed antiarrhythmic-drug therapy (amiodarone excluded).
Outcome: 12-month treatment success 56% vs 30% (HR for failure 0.46; 95% CI 0.33–0.65; P<0.001). Six procedure-related strokes (all NIHSS <5), none after protocol changes (LA thrombus screening, anticoagulation, excluding CHA₂DS₂-VASc ≥4).

Clinical takeaway. Extends the ablation-over-drugs advantage from paroxysmal to persistent AF. Per the editorial (Sapp), the gain was driven mostly by reduced asymptomatic recurrence, the mandated posterior-wall isolation is of uncertain added benefit, and real-world procedural stroke safety is not yet settled. What this changes: extends first-line ablation to persistent AF, but it is a procedural electrophysiology decision — not practice-mandating for the general internist pending real-world safety.

Perioperative / HematologyTRACTION · RCTLower yield · IM/ID

Hospital Policy of Tranexamic Acid to Reduce Transfusion in Major Noncardiac Surgery

In major noncardiac surgery at high transfusion risk, does a hospital-wide tranexamic acid policy reduce transfusion without increasing venous thromboembolism?

Population: 8,273 patients at 10 Canadian hospitals undergoing major noncardiac surgery at high transfusion risk (60.5% oncologic surgery).
Intervention: Hospital-wide policy of IV tranexamic acid (1 g at incision + 1 g before closure; 2 g if >100 kg).
Comparison: Matching placebo (cluster-crossover, 4-week blocks).
Outcome: Red-cell transfusion 7.4% vs 9.8% (RR 0.73; 95% CI 0.61–0.86; NNT 37). 90-day VTE 2.1% vs 2.1% (RR 0.96; 95% CI 0.65–1.38) — met noninferiority, including in cancer surgery (2.4% vs 2.6%).

Clinical takeaway. A single perioperative TXA dose reliably cuts transfusion ~25% without raising thrombosis; per the editorial (Murphy & Roberts) the remaining barrier is implementation, not evidence. The actionable decision sits with anesthesia/surgery rather than the admitting internist or ID fellow. What this changes: confirms/extends POISE-3 and the antifibrinolytic literature (CRASH-2, WOMAN) to a diverse high-risk noncardiac — including oncologic — surgical population.

Allergy & ImmunologyReview · Lower yield · IM

Prevention and Treatment of Peanut Allergy

When should peanut be introduced to prevent allergy, and how is established peanut allergy treated?

•Early dietary introduction of peanut (4–6 months in high-risk infants, ~6 months otherwise) prevents ~80% of peanut allergy; the window is narrow and every month of delay raises risk, especially with severe atopic dermatitis or non-White race. •Once established, oral / epicutaneous / sublingual immunotherapy and omalizumab are more effective and safer the younger the child, but none is curative — protection wanes within ~1 year of stopping. •Adult-IM-adjacent point: omalizumab is now FDA-approved for IgE-mediated food allergy in persons ≥1 yr.

Teaching Case · Case Records of the MGH

Case 18-2026: A 53-Year-Old Man with Leg Weakness, Pain, and Weight Loss · Neurology

High yield · IM

Key learning point. Final diagnosis: diabetic lumbosacral radiculoplexus neuropathy (diabetic amyotrophy). Painful, asymmetric, proximal leg weakness with profound weight loss appearing at or shortly after a new diabetes diagnosis — with normal/unrevealing imaging and inflammatory markers — should raise it; EMG/NCS confirm a severe, subacute, asymmetric lumbosacral radiculoplexus neuropathy. It is a microvasculitic, self-limited/monophasic disorder: management is supportive (early PT, neuropathic-pain agents), glycemic control is not disease-modifying (and aggressive control may even provoke it), and most patients recover only partially (often residual foot drop).

Also in this issue (not randomized trials, so not summarized as such): Ten-Year Outcomes after CAR T-Cell Therapy for B-Cell Lymphomas — single-arm 10-yr follow-up of tisagenlecleucel (no relapse beyond 5.4 yr; 10-yr lymphoma-free survival 32% large B-cell / 47% follicular; but second primary cancer 21% and B-cell aplasia 44%; lower yield — narrow oncology, observational); Phase 3 Bepirovirsen for Chronic Hepatitis B — RCT reprinted from the May 28 issue, already summarized/archived; Images in Clinical Medicine — Kwashiorkor; Unheralded Syncope from Ventricular Arrhythmia; plus Perspectives and Correspondence (GLP-1 access for Medicare; air pollution / EPA rulemaking; Andes hantavirus cruise-ship outbreak; childhood undervaccination & measles clusters in South Carolina; stopping beta-blockers after MI).

Summaries are original paraphrases prepared for educational use; figures are drawn from the published full text. Read the full articles at NEJM.org (subscription required). Trial registrations: BEAT-TB (NCT04062201), AVANT GUARD (NCT06096337), TRACTION (NCT04803747). Citations: N Engl J Med 2026; Vol. 394, No. 24. DOIs: 10.1056/NEJMoa2503687, 10.1056/NEJMoa2600929 (editorial 10.1056/NEJMe2606012), 10.1056/NEJMoa2515820 (editorial 10.1056/NEJMe2603105), 10.1056/NEJMcp2314424, 10.1056/NEJMcpc2517891; also 10.1056/NEJMoa2518035, 10.1056/NEJMoa2515131.