The Critical Therapeutic Window Lies Before Intubation: HFNC and Mortality Across Ventilatory Escalation Tiers in Pediatric HSCT.

BACKGROUND: Pediatric allogeneic hematopoietic stem cell transplant (HSCT) recipients face high morbidity from respiratory complications, with mortality exceeding 20% once intensive care is required, rising above 45% when invasive mechanical ventilation is needed. While early identification of patients at risk for respiratory deterioration is widely recognized as a clinical priority, the role of noninvasive respiratory support before intubation remains unresolved. Recent evidence has raised concern that preintubation high-flow nasal cannula (HFNC) or noninvasive positive-pressure ventilation may worsen post-intubation outcomes - yet most analyses have been restricted to patients who ultimately required mechanical ventilation, excluding the substantial group who respond to noninvasive support and avoid intubation altogether. OBJECTIVE: To examine HFNC use and outcomes across the full ventilatory escalation pathway in pediatric allogeneic HSCT recipients, with particular attention to identifying which patients benefit from noninvasive support and which progress to invasive ventilation. STUDY DESIGN: Single-center retrospective cohort study of 530 pediatric allogeneic HSCT recipients (<21 years) transplanted between 2011 and 2025, of whom 67 (12.6%) were admitted non-electively to the pediatric intensive care unit (PICU) within 548 days of transplantation. Patients were stratified by highest ventilatory escalation tier reached. Multivariable logistic regression identified independent predictors of PICU mortality. RESULTS: Overall PICU mortality was 67.2%. Half of HFNC-treated patients (13/26, 50%) avoided intubation altogether, with a mortality of only 23.1%. HFNC use was associated with improved overall survival in univariate analysis (odds ratio 0.21; 95% CI 0.07-0.62; p = 0.005). Year of admission (OR 0.81) and organ support score (OR 2.89) were independent multivariable predictors. Prior noninvasive support did not increase post-intubation mortality compared with direct intubation (84.4% vs. 76.9%; p = 0.672); patients intubated after HFNC failure showed lower mortality than those intubated without prior HFNC (69.2% vs. 87.5%). Mortality rose sharply from 35.3% with noninvasive support only to 80-88% across all post-intubation tiers, with no consistent further increase upon escalation to high-frequency oscillatory ventilation or extracorporeal membrane oxygenation. In an exploratory subgroup analysis, higher Organ Support Score at admission (median 3 vs. 1; p = 0.001) and respiratory failure as the primary admission reason (92.3% vs. 38.5%; p = 0.011) distinguished HFNC failures from responders. CONCLUSIONS: In pediatric allogeneic HSCT recipients, HFNC use was associated with improved overall survival, and prior noninvasive support was not associated with worse post-intubation outcomes, although this comparison was underpowered. The critical mortality inflection occurred at the transition to invasive ventilation, not with subsequent escalation. These findings reframe the central clinical question: not whether to offer HFNC, but how to identify the patients most likely to respond - a research priority of immediate relevance to the pediatric transplant community, requiring prospective validation of clinical and biological predictors that allow timely escalation in non-responders while sparing responders from invasive ventilation.