To the Editor: Trimethoprim–sulfamethoxazole (TMP-SMX) is among the most frequently prescribed antimicrobials in pediatric practice and remains the cornerstone of treatment and prophylaxis for Pneumocystis jirovecii pneumonia (PJP) in immunocompromised children. In addition to PJP, TMP-SMX is routinely used for urinary tract infections and skin and soft-tissue infections, resulting in widespread exposure among pediatric populations. TMP-SMX is generally well tolerated, with delayed hypersensitivity reactions, cytopenias and renal toxicity being the most recognized adverse effects.1 However, rare but fulminant, life-threatening pulmonary toxicity has emerged as a distinct and underrecognized adverse drug reaction, particularly in children and young adults,1–3 with accumulating evidence implicating specific human leukocyte antigen (HLA) susceptibility alleles.4 We report a pediatric case of TMP-SMX–associated acute lung injury occurring after initial clinical response to microbiologically confirmed PJP, highlighting a critical diagnostic pitfall for pediatric infectious diseases clinicians: distinguishing refractory infection from catastrophic drug-induced lung injury. A 12-year-old female with steroid-resistant minimal change disease complicated by nephrotic syndrome, receiving chronic prednisone and tacrolimus, presented with fever, hypoxemia and progressive respiratory distress requiring pediatric intensive care unit admission. Chest radiography and computed tomography demonstrated bilateral pulmonary opacities and interlobular septal thickening. Bronchoalveolar lavage revealed P. jirovecii polymerase chain reaction detection, confirming PJP pneumonia. Serum β-D-glucan was markedly elevated, while galactomannan and other fungal biomarker testing were negative. She was treated with intravenous TMP-SMX and adjunctive corticosteroids, resulting in clinical stabilization and improvement in oxygen requirements. She was transitioned later to oral TMP-SMX and discharged home to complete the course for 21 days, followed by TMP-SMX prophylaxis. On day 19 of TMP-SMX therapy, she developed recurrent fever and rapidly progressive hypoxemic respiratory failure. Repeat imaging revealed diffuse bilateral ground-glass opacities and consolidations, substantially worse compared with prior studies. Despite broad antimicrobial escalation, she deteriorated rapidly, requiring endotracheal intubation, mechanical ventilation and ultimately venovenous extracorporeal oxygenation (ECMO) for refractory respiratory failure. Extensive infectious reevaluation was performed, including repeat bronchoalveolar lavage with comprehensive bacterial, viral and fungal testing, all of which were negative, including repeat P. jirovecii polymerase chain reaction. The tempo and severity of deterioration were disproportionate to the expected course of treated PJP, raising concern for a noninfectious etiology. Given the temporal association with prolonged TMP-SMX exposure and the absence of an alternative diagnosis, TMP-SMX–associated lung injury was strongly suspected. Pharmacogenomic testing demonstrated positivity for HLA-B07:02 and HLA-C07:02, alleles previously associated with TMP-SMX–related fulminant respiratory failure.2,5 Owing to hemodynamic instability and ECMO dependence, lung biopsy was initially deferred; however, persistent respiratory failure and steroid dependence prompted surgical lung biopsy once clinically feasible. Histopathologic examination revealed diffuse alveolar injury with delayed epithelialization, characterized by extensive alveolar epithelial denudation, relative paucity of hyaline membranes and prominent macrophage lining of denuded alveolar walls (Fig. 1). No organisms were identified on special stains, and there was no evidence of vasculitis or organizing pneumonia. These findings, together with the clinical course and HLA profile, provided compelling evidence for TMP-SMX–induced acute lung injury rather than refractory PJP.FIGURE 1.: Lung biopsy demonstrating diffuse alveolar injury with delayed epithelialization (DAIDE) across multiple regions of the specimen. (A) Alveolar epithelial denudation with loss of epithelial lining. (B) Prominent macrophage accumulation along denuded alveolar surfaces. (C) Expanded interstitium with inflammatory infiltrates and relative absence of hyaline membrane formation. (D) Patchy areas of delayed epithelial regeneration, highlighting the diffuse nature of injury. Hematoxylin and eosin stain; original magnification ×200.TMP-SMX prophylaxis was discontinued, and high-dose systemic corticosteroids were initiated, leading to gradual improvement in oxygenation and successful liberation from ECMO and mechanical ventilation. Attempts at rapid corticosteroid tapering or discontinuation were associated with clinical worsening, consistent with a steroid-responsive, immune-mediated drug toxicity. This case illustrates several critical teaching points. First, TMP-SMX–associated lung injury can closely mimic progressive or refractory PJP, particularly in immunocompromised hosts, leading to diagnostic delay and prolonged exposure to the offending agent. Second, initial clinical improvement on TMP-SMX does not exclude subsequent catastrophic drug toxicity, a feature repeatedly described in pediatric and young adult cohorts. Third, lung biopsy may be essential in establishing the diagnosis when respiratory failure remains unexplained despite exhaustive infectious evaluation.6 The identification of diffuse alveolar injury with delayed epithelialization is particularly compelling, as this histopathologic pattern has been described almost exclusively in association with TMP-SMX exposure and is distinct from infection-related acute respiratory distress syndrome, ECMO-associated lung injury and other drug-induced pneumonitis.7,8 Equally important is the role of pharmacogenomics. The association of TMP-SMX–induced lung injury with HLA-B07:02 and HLA-C07:02 places this syndrome within the broader framework of HLA-restricted immune-mediated adverse drug reactions, analogous to abacavir (HLA-B57:01) and carbamazepine (HLA-B15:02).4,9 While routine pre-prescription HLA screening is not currently recommended, postevent pharmacogenomic testing provides diagnostic clarity and may inform future antimicrobial avoidance.4 Emerging data also suggest a dose- and duration-dependent risk, with severe pulmonary toxicity occurring predominantly after high-dose, prolonged TMP-SMX exposure, as used for PJP treatment or other serious infections, whereas low-dose prophylaxis appears less frequently implicated. This distinction carries immediate clinical relevance, given the widespread use of TMP-SMX for common infections and prophylaxis.3 Reported cases of TMP-SMX–associated acute lung injury typically occur after a latency period of approximately 10–21 days of continuous exposure with therapeutic dosing, a timeline mirrored in our patient.2,3,10 Typically, TMP-SMX is not routinely used for PJP prophylaxis in children with nephrotic syndrome, as most patients are not exposed to prolonged immunosuppression and receive relatively short, tapering courses of corticosteroids. However, in patients requiring long-term or intensified steroid therapy, such as in this case, the risk of PJP increases, making prophylaxis a reasonable and commonly adopted strategy.11 Despite its well-established efficacy, clinicians must remain vigilant for rare but severe complications, including TMP-SMX–associated lung injury, where early recognition and prompt drug discontinuation may be lifesaving.
Muayad Alali (Fri,) studied this question.