NEJM This Week — September 3, 2026

This week features a practical Clinical Practice review on pulmonary nodules (the essential IM read below) and the PARADIGM trial, which extends the azacitidine–venetoclax paradigm from induction-ineligible to induction-eligible AML — both high-yield. Two pediatric RCTs (achondroplasia, balanced fluid vs saline) and a teaching case on hypophosphatasia round out the issue.

Pulmonary & Critical CareReview · High yield · IM

Pulmonary Nodules

The week’s most broadly relevant read — pulmonary nodules are among the most common incidental findings an internist manages. Key points:

•Stability = benign? Solid nodules stable for 2 years (or 1 year with volumetric CT) can be considered benign. Subsolid nodules need ≥4 years of stability — they grow more slowly but have a higher malignancy rate. •Risk stratification: Use validated models (Brock, Mayo). Low risk (<10%) → CT surveillance. Intermediate (10–70%) → PET-CT or biopsy. High (>70%) → surgical resection, even without preoperative tissue. •Biopsy comparison (from an RCT): Navigational bronchoscopy and CT-guided transthoracic needle biopsy have similar diagnostic accuracy (~74–79%), but pneumothorax is dramatically different — 3% vs 28% (chest tube: 1% vs 12%). •Sublobar resection is now acceptable for stage I NSCLC ≤2 cm (noninferior to lobectomy for disease-free and overall survival per a recent RCT). •PET-CT: Sensitivity 89%, specificity 75% for intermediate-risk nodules ≥8 mm. NOT indicated for characterizing growing nodules (may be FDG-negative yet malignant).

Hematology & OncologyPARADIGM · RCTHigh yield · IM/ID

Azacitidine–Venetoclax vs Induction Chemotherapy in Induction-Eligible AML

Population: 172 adults with newly diagnosed AML eligible for induction chemotherapy; median age 64; 72% adverse-risk per ELN 2022. Excluded: FLT3 mutations (VAF ≥5%), core binding factor fusions, NPM1 mutations if <60 years.
Intervention: Azacitidine 75 mg/m² IV days 1–7 + venetoclax 400 mg PO days 1–28, repeating 28-day cycles.
Comparison: Induction chemotherapy — 7+3 (cytarabine + idarubicin/daunorubicin) or CPX-351, investigator’s preselected choice.
Outcome: Median EFS 14.5 mo vs 6.2 mo (HR 0.57; 95% CI 0.39–0.84; P=0.002). Composite CR 78% vs 53%. HCT rate 60% vs 40%. Grade ≥3 infection 28% vs 41%; grade ≥3 hemorrhage 2% vs 12%. 30-day mortality 0% vs 3%. Mean inpatient days (first 30 d) 12.5 vs 27.3. OS not significantly different (21.5 vs 18.0 mo) but trial not powered for OS.

Clinical takeaway. Azacitidine–venetoclax — already the standard for induction-ineligible AML (VIALE-A) — showed superior EFS with a better safety profile and far fewer hospital days in induction-eligible patients. The editorial (Schiffer) cautions that the newer approach is “unlikely to be curative in most patients with higher-risk disease” even with HCT, and notes that at least five larger randomized trials are ongoing. Phase 2 (n=172); phase 3 confirmation awaited. Does NOT replace targeted therapy for FLT3-mutated or core-binding-factor AML.

EndocrinologyPROPEL 3 · RCTLower yield · IM/ID

Oral Infigratinib in Children with Achondroplasia

Population: 114 children (age 3–17) with genetically confirmed achondroplasia, 2:1 randomized. Pediatric rare bone disease.
Intervention: Oral infigratinib (FGFR1-3 TKI) 0.25 mg/kg daily for 52 weeks.
Comparison: Placebo.
Outcome: LS mean change in annualized height velocity: +1.74 cm/yr vs placebo (95% CI 1.31–2.17; P<0.001). Height z-score improvement: +0.32 (P<0.001). No FGFR1/2-related safety signals at this dose.

Clinical takeaway. Infigratinib is the first oral therapy for achondroplasia (existing options — vosoritide, navepegritide — are subcutaneous injections). The editorial (Salusky/Jüppner) highlights the advantage of oral administration for pediatric patients. Does not change IM practice; included for awareness of the therapeutic advance.

Pulmonary & Critical CarePRoMPT BOLUS · RCTLower yield · IM/ID

Balanced Fluid vs 0.9% Saline in Pediatric Septic Shock

Population: 8,482 children (2 months to <18 years) with suspected septic shock and abnormal perfusion, across 47 EDs in 5 countries.
Intervention: Balanced crystalloid (lactated Ringer’s or Plasma-Lyte) for bolus and maintenance, up to 48 hours.
Comparison: 0.9% saline for bolus and maintenance, up to 48 hours.
Outcome: MAKE30 (death, new RRT, or persistent kidney dysfunction at 30 d): 3.4% vs 3.0% (RR 1.10; 95% CI 0.88–1.40; P=0.85) — no significant difference. Hyperchloremia lower with balanced fluid (31.4% vs 49.0%) but no clinical outcome benefit.

Clinical takeaway. This large pragmatic trial found no benefit of balanced fluid over normal saline for pediatric septic shock, paralleling the adult BaSICS and PLUS trials. The editorial notes that the low overall event rate (3%) limited power for small differences, and point estimates favored balanced fluid in the highest-volume and most-acidotic subgroups. A definitive answer in the sickest patients remains elusive; for now, no evidence to prefer one crystalloid over the other at the bedside.

Teaching Case · Clinical Problem-Solving

Break a Leg · Endocrinology / Metabolic Bone Disease

High yield · IM · board-style pearl

Key learning point. A 46-year-old woman with recurrent fragility fractures (metatarsal → femoral → atypical femoral fracture and pseudofractures) treated with bisphosphonates was ultimately diagnosed with hypophosphatasia. The cardinal clue: low alkaline phosphatase. Bisphosphonates are structural pyrophosphate analogues and are contraindicated in hypophosphatasia — they worsen the mineralization defect. Denosumab is also contraindicated. Confirm with elevated vitamin B6 (pyridoxal 5′-phosphate), elevated urinary phosphoethanolamine, and ALPL gene sequencing. Measure ALP before starting any antiresorptive therapy. Treatment: asfotase alfa (enzyme replacement for childhood-onset disease) or teriparatide.

Also in this issue (not randomized, so not summarized as a trial): Daraxonrasib for previously treated RAS-mutant NSCLC — phase 1–2, open-label, nonrandomized dose-escalation/expansion study of a first-in-class oral RAS(ON) multiselective tri-complex inhibitor (RMC-6236). Response rate ~31–37%; median PFS 8.3 mo. Phase 3 (RASolve 301, daraxonrasib vs docetaxel) is ongoing.

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: PARADIGM (NCT04801797), PROPEL 3 (NCT06164951), PRoMPT BOLUS (NCT04102371). Citations: N Engl J Med 2026; Vol. 395, No. 9. DOIs: 10.1056/NEJMoa2602804, 10.1056/NEJMoa2604565, 10.1056/NEJMoa2601969, 10.1056/NEJMoa2504059, 10.1056/NEJMcp2515063, 10.1056/NEJMcps2602868.