Immune dysregulation score predicts mortality and identifies hydrocortisone responders in community-acquired pneumonia

This study presents a multicohort analysis and reanalysis of a randomized controlled trial aimed at deriving and validating an immune dysregulation score for patients with community-acquired pneumonia (CAP) across diverse care settings. The research involved 398 participants in the derivation phase and 1191 participants in the external validation phase, encompassing patients from emergency departments, general wards, and intensive care units with varying disease severities. The primary objective was to develop a categorical (DIP stages) and continuous (cDIP) immune dysregulation score independent of clinical presentation or outcome, using plasma biomarkers to quantify the host immune response.

The intervention was the application of a three-biomarker machine-learning framework measuring procalcitonin, soluble TREM-1, and IL-6 to predict the degree of immune dysregulation, which was originally derived from a panel of 35 biomarkers. The comparator was standard clinical severity assessment. The framework demonstrated high accuracy, with the categorical DIP stage classification achieving 91.2% accuracy and the continuous cDIP score showing a root mean square error of 0.056. For the trial reanalysis component, follow-up focused on 30-day mortality outcomes.

Regarding primary outcomes, the three-biomarker framework successfully predicted the immune dysregulation state. For key clinical associations, increased immune dysregulation was strongly linked to worse outcomes. The continuous cDIP score showed an odds ratio of 1.26 for mortality per 10% increase in dysregulation (95% CI 1.13-1.40, p<0.0001). For secondary infections, the cDIP score showed an odds ratio of 1.50 per 10% increase (95% CI 1.22-1.93, p=0.0005). These results indicate a gradual, dose-response relationship between the degree of immune dysregulation and adverse clinical events.

A critical secondary finding emerged from the reanalysis of the randomized trial data concerning hydrocortisone. The corticosteroid conferred a survival benefit exclusively in participants classified as severely dysregulated. For those in the DIP3 (severe dysregulation) category, the odds ratio for mortality with hydrocortisone was 0.25 (95% CI 0.05-0.85, p=0.042). Using the continuous cDIP score threshold of ≥0.63 for severe dysregulation, the odds ratio was 0.21 (95% CI 0.10-0.72, p=0.011). Furthermore, hydrocortisone treatment was accompanied by faster immune recovery over time, as indicated by a significant time × treatment interaction (p<0.0001).

Safety and tolerability data were not reported in the provided input. The study did not detail adverse event rates, serious adverse events, discontinuations, or general tolerability profiles associated with either the biomarker assessment or the hydrocortisone treatment in the reanalyzed cohort.

These results contribute to a growing body of evidence seeking to personalize immunomodulatory therapy in sepsis and pneumonia. Prior landmark studies, such as the ADRENAL and APROCCHSS trials, investigated hydrocortisone in broader septic shock populations with mixed results on mortality. This analysis suggests heterogeneity in treatment effect may be explained by underlying immune phenotype, aligning with the concept of precision medicine in critical care. It moves beyond clinical severity scores by attempting to biologically define a patient subgroup most likely to benefit from corticosteroid intervention.

Key methodological limitations must be acknowledged. The most significant is that the therapeutic finding is derived from a post-hoc analysis of a randomized trial. This introduces potential bias and means the observed treatment effect in the dysregulated subgroup is hypothesis-generating, not confirmatory. The specific dosing and protocol for hydrocortisone administration in the original trial were not detailed in the input. Furthermore, the absence of reported safety data for the biomarker-stratified group is a notable gap.

The clinical implication is that a simple, three-biomarker score (procalcitonin, sTREM-1, IL-6) shows promise for identifying CAP patients with severe immune dysregulation who have higher baseline risk and may derive mortality benefit from hydrocortisone. This suggests a potential future strategy for stratifying patients for corticosteroid therapy rather than applying it uniformly. However, this approach cannot be adopted into practice based on this evidence alone; it requires validation in a prospective, randomized trial designed a priori to test the biomarker-guided strategy.

Several important questions remain unanswered. The optimal timing for biomarker measurement and score calculation in clinical workflow is unclear. The cost-effectiveness and broad availability of the three-biomarker panel, particularly sTREM-1 and IL-6, are uncertain. The mechanism by which hydrocortisone accelerates immune recovery in the dysregulated phenotype needs further elucidation. Most critically, the efficacy and safety of biomarker-guided hydrocortisone administration must be prospectively tested against standard care in a randomized controlled trial setting before any clinical recommendations can be made.