Corticosteroids in immunocompromised ICU patients with pneumonia: risks, evidence, and clinical practice.

Adjunctive corticosteroids are increasingly used in Severe Community-Acquired Pneumonia (sCAP) requiring admission to the intensive care unit (ICU), based on evidence derived largely from immunocompetent populations. Immunocompromised hosts, however, represent a distinct and growing subgroup among critically ill patients with sCAP, characterised by baseline immune defects, broader pathogen spectra, and a heightened susceptibility to treatment-related harm. Whether corticosteroid strategies validated in the general population can be safely and effectively extrapolated to immunocompromised ICU patients with sCAP remains uncertain, despite their growing use in contemporary ICU practice. Immunocompromised patients admitted to the ICU with sCAP experience compounded alterations in host defence, driven by pre-existing immune dysfunction and superimposed critical illness-associated immune dysregulation. In this setting, corticosteroids may attenuate inflammation-mediated lung injury, but may also deepen immune suppression, distort clinical evolution, and increase susceptibility to opportunistic superinfections and pathogen reactivation. Randomised trials and meta-analyses supporting corticosteroid use in sCAP have systematically excluded immunocompromised patients, treating immunosuppression as an exclusion criterion rather than a stratification variable, and commonly rely on short-term endpoints that fail to capture delayed infectious complications relevant to this population. Available observational data and pathogen-specific evidence suggest substantial heterogeneity of treatment effect (HTE) across immunocompromised subgroups, influenced by immune substrate, pathogen context, and cumulative immunosuppressive burden. Among the pathogen-specific contexts in which corticosteroids have the strongest evidence, Pneumocystis jirovecii pneumonia (PJP) stands out as the principal indication: adjunctive corticosteroids have demonstrated clear mortality benefit in AIDS-associated severe PJP, though this benefit does not extend uniformly to HIV-negative immunocompromised patients with PJP. Consequently, standard corticosteroid treatment in all CAP subgroups is unlikely to be appropriate. In immunocompromised ICU patients with sCAP, corticosteroid therapy should neither be routinely applied nor categorically avoided. Instead, decisions should be individualised, guided by a clearly defined therapeutic target, careful assessment of immune status, and consideration of the risk for opportunistic infections. When used, corticosteroids should be administered at the lowest effective dose and for the shortest feasible duration. Dedicated studies incorporating immune stratification, pathogen-informed approaches, and outcomes relevant to delayed infections are urgently needed. Until such evidence is available, careful risk-balance assessments to decide steroids in these patients.