NEJM This Week — June 18, 2026

This week's headline read is the SNAP trial — cefazolin was noninferior to antistaphylococcal penicillins for MSSA bacteremia and caused less kidney injury (high-yield, below). Two concordant phase 3 trials (SWiFT and TOWAR) found prehospital whole blood no better than components, and C-TRACT showed iliac-vein stenting modestly helps post-thrombotic syndrome — all summarized for awareness but flagged lower-yield for general IM. A practical thyroid-cancer review and a parvovirus B19 teaching case round out the high-yield items.

Infectious DiseaseSNAP · RCTHigh yield · IM/ID

Cefazolin for Methicillin-Susceptible Staphylococcus aureus Bacteremia

In adults with MSSA bacteremia, is cefazolin noninferior to flucloxacillin/cloxacillin for 90-day mortality — and how do the harms compare?

Population: 1,287 adults (median age 66; most common focus osteoarticular) with penicillin-resistant, methicillin-susceptible S. aureus bacteremia, enrolled ≤72 h after first positive culture in an international Bayesian adaptive platform; >99% had ID consultation. Excluded: penicillin/cefazolin allergy, maintenance dialysis, polymicrobial bacteremia.
Intervention: Cefazolin 2 g IV q8h (q6h if critical illness/endocarditis).
Comparison: Flucloxacillin 2 g IV q6h or cloxacillin 2 g IV q4h.
Outcome: 90-day all-cause mortality 15.0% vs 17.0% (adjusted OR 0.81; 95% CrI 0.59–1.12; probability of noninferiority 99.2%, superiority 89.8%). Acute kidney injury 13.9% vs 19.6% (aOR 0.67; 95% CrI 0.50–0.89); fewer serious adverse reactions (aOR 0.41) and far fewer discontinuations (aOR 0.21). Endocarditis subgroup consistent.

Clinical takeaway. In MSSA bacteremia, cefazolin is at least as effective for survival and clearly better tolerated — notably less acute kidney injury. The narrow noninferiority margin and a consistent effect in the endocarditis subgroup substantially allay long-standing inoculum-effect concerns (microbiologic subgroup report pending). Findings likely extend to nafcillin/oxacillin. What this changes: elevates observational data (pooled OR ~0.73) to randomized evidence — supports cefazolin as a first-line option for MSSA bacteremia, including endocarditis, not a reluctant alternative.

Critical Care / TraumaSWiFT · RCTLower yield · IM/ID

Prehospital Whole Blood in Traumatic Hemorrhage — A Randomized Controlled Trial

In major traumatic hemorrhage, does prehospital whole blood reduce death or massive transfusion vs. components?

Population: 616 analyzed patients with major traumatic hemorrhage attended by 10 English air ambulance services (median Injury Severity Score 33).
Intervention: Up to 2 units leukocyte-depleted, platelet-rich group O whole blood.
Comparison: Up to 2 units each of red cells and plasma.
Outcome: Death or massive transfusion at 24 h: 48.7% vs 47.7% (adjusted RR 1.02; 95% CI 0.80–1.31; P=0.84) — not superior. Safety similar.

Clinical takeaway. Up to 2 units of cold-stored, leukoreduced whole blood was no better than red cells + plasma. Per the shared editorial, this is not "whole blood fails": baseline mortality is far lower than in older trials, neither trial was an equivalence trial, the 2-unit cap may be too small, and most deaths were from brain injury, not hemorrhage. What this changes: with TOWAR, choose whole blood vs. components by availability/logistics/cost — no practice mandate.

Critical Care / TraumaTOWAR · RCTLower yield · IM/ID

Prehospital Resuscitation with Type O Whole Blood for Trauma and Hemorrhage

In hemorrhagic shock, does prehospital group O whole blood lower 30-day mortality vs. components — and does storage age matter?

Population: 993 analyzed injured patients with hemorrhagic shock; cluster-randomized across 44 U.S./Canadian air medical bases (2:1).
Intervention: Up to 2 units low-titer group O whole blood (shelf life ≤21 days).
Comparison: As-indicated blood components (plasma, red cells, or both).
Outcome: 30-day mortality 25.9% vs 20.5% (adjusted OR 1.24; 95% CI 0.87–1.76; P=0.24) — no benefit. Storage age 15–21 vs 1–14 days: 27.1% vs 26.4% (aOR 0.99) — no difference.

Clinical takeaway. Concordant with SWiFT: no survival advantage for prehospital whole blood at ≤2 units, and both products are safe. Substantial product crossover and low prehospital transfusion volume bias toward the null. What this changes: reinforces SWiFT; storage age through 21 days appears not to matter for outcomes.

Cardiology / VascularC-TRACT · RCTLower yield · IM/ID

Endovascular Therapy for Post-Thrombotic Syndrome — A Randomized Trial

In moderate/severe post-thrombotic syndrome with iliac-vein obstruction, does stenting reduce symptom severity vs. standard care?

Population: 225 patients with moderate/severe post-thrombotic syndrome (VCSS ≥8, Villalta ≥10, or open ulcer) and imaging-confirmed iliac-vein obstruction, ≥3 months after DVT.
Intervention: Iliac-vein stenting + enhanced antithrombotic therapy, plus standard care.
Comparison: Standard post-thrombotic syndrome care alone (compression, anticoagulation, wound care).
Outcome: 6-month VCSS 8.1 vs 10.0 (adjusted difference −2.0; 95% CI −3.2 to −0.8; P=0.001); venous-specific QoL +14.5 and SF-36 physical +6.1 (both P<0.001). Bleeding higher (11.6% vs 3.6%; mostly nonmajor).

Clinical takeaway. Stenting modestly improved severity and meaningfully improved quality of life in carefully selected patients, at the cost of more bleeding. Per the editorial: the bundled intensified antithrombotic therapy can't be disentangled from stenting, and only 6-month durability is known — not yet a routine guideline recommendation. What this changes: strongest evidence to date, but not practice-mandating pending durability/cost data. Reasonable via shared decision making for severe refractory disease with acceptable bleeding risk.

EndocrinologyReview · High yield · IM

Management of Differentiated Thyroid Cancer

When is active surveillance — rather than immediate surgery — appropriate for papillary thyroid cancer?

•Active surveillance is a guideline-endorsed alternative to immediate surgery for properly selected papillary thyroid cancer nodules ≤1 cm (microcarcinoma): a meta-analysis of 5,685 low-risk patients showed 100% disease-specific survival, progression in 14.5%, distant metastasis in only 0.03%. •Surveillance = serial neck/thyroid ultrasound every 6–12 months for 1–2 years, then less often while stable. •BRAF V600E (MAPK-pathway activation) drives loss of radioactive-iodine sensitivity; RAI-refractory disease is treated with multikinase inhibitors — lenvatinib (SELECT) and sorafenib (DECISION) improved progression-free survival.

Teaching Case · Case Records of the MGH

Case 17-2026: A 74-Year-Old Man with Pancytopenia · Hematology / Infectious Disease

High yield · IM/ID

Key learning point. Final diagnosis: parvovirus B19 infection. Parvovirus B19 is the most common infectious cause of pure red-cell aplasia — it infects erythroid progenitors via the P-antigen receptor, producing pathognomonic giant proerythroblasts in the marrow. In immunocompetent hosts the suppression is transient, but in immunocompromised hosts (here, 3 months post–heart transplant) it can persist as chronic anemia/pancytopenia. No direct antiviral exists: treat with IVIG, reduction of immunosuppression, and supportive transfusions; recurrence occurs in ~25% of transplant recipients.

Also in this issue (not randomized trials, so not summarized as such): NEJM Outbreaks Update — Andes Hantavirus (outbreak update); Images in Clinical Medicine — immunosuppression-associated peripheral T-cell lymphoma; chromoblastomycosis; Clinical Implications of Basic Research — stemness and checkpoint durability (mechanistic); plus Perspectives and Correspondence (incl. Sudan virus disease in Uganda; IV artesunate in artemisinin-resistant malaria).

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: SNAP (NCT05137119), SWiFT (ISRCTN23657907), TOWAR (NCT04684719), C-TRACT (NCT03250247). Citations: N Engl J Med 2026; Vol. 394, No. 23. DOIs: 10.1056/NEJMoa2506905, 10.1056/NEJMoa2516043, 10.1056/NEJMoa2602167, 10.1056/NEJMoa2519001, 10.1056/NEJMra2416814, 10.1056/NEJMcpc2517867 (shared whole-blood editorial 10.1056/NEJMe2605735).