Immunosuppression is a critical risk factor for infection and sepsis, and as the global burden of acquired immunosuppression continues to grow, optimizing care for this vulnerable population represents an urgent public health priority. The prevalence of immunosuppression within the general population is increasing, with up to one-third of all critically ill patients being immunocompromised at baseline. Three converging trends necessitate renewed focus. First, survival among patients with immunocompromising conditions has improved dramatically. Contemporary data demonstrate that intensive care unit (ICU) survival in selected patients with cancer approaches that of immunocompetent patients when adjusted for organ failure severity. Yet triage thresholds remain inconsistent, and delayed ICU transfer is associated with excess mortality. ^, Second, the therapeutic landscape has fundamentally altered immune function by introducing unique infectious risks and inflammatory syndromes that may mimic sepsis. Third, immunocompromised patients are underrepresented in sepsis trials. The Surviving Sepsis Campaign recommendations are derived largely from immunocompetent populations. Whether fluid strategies, vasopressor thresholds, steroid responsiveness, and antimicrobial duration should differ remains uncertain. Importantly, sepsis in the immunocompromised host is not simply sepsis plus comorbidity.

Understanding the relationship between immunity and infection risk is crucial in developing personalized approaches to patient care and improving clinical outcomes. However, with prolonged survival among those with varying immunosuppressive conditions, compounded by an aging and increasingly comorbid population, this relationship is more complex than ever before. The recent seismic expansion in immunomodulating therapies, including biologics and T-cell-engaging therapies, has required physicians to recognize a new repertoire of immune toxicities, including novel infections and immune syndromes masquerading as sepsis. With clear evidence supporting an association between immune dysfunction and adverse clinical outcomes among critically ill patients, early sepsis identification, awareness of sepsis-mimickers, and tailored management strategies are of paramount importance for optimizing patient care and long-term outcomes.

This article aims to summarize the current understanding of sepsis in the immunocompromised host, highlighting diagnostic challenges and management controversies while raising awareness of important parallel immune syndromes and future research opportunities. Immunosuppression may result from primary immunodeficiencies that is, caused by genetic or congenital abnormalities, or secondary immunodeficiencies, which are acquired throughout the lifespan owing to underlying disease, comorbidities, and immunosuppressive agents. This article focuses on the latter; however, the general approaches and considerations outlined here may be applicable to the broader population. Importantly, sepsis-induced immunosuppression in critical illness is an increasingly recognized entity, underscoring the importance of viewing the immune system as a dynamic rather than static state, the kinetics of which evolve over the course of a hospital admission. Integrating temporal patterns of immune dysfunction into the care of critically ill patients represents an important area of ongoing research that, while beyond the scope of this article, has been described elsewhere.

One of the greatest challenges facing the field is the variability in how immunocompromised patients are defined, described, and stratified. Without consistent definitions, clinical and trial populations will remain detrimentally heterogeneous, limiting the application and generalizability of findings. Unfortunately, there is no consensus definition that currently exists. Traditional disease-based categories include: hematologic malignancy, solid tumor receiving active therapy, hematopoietic stem cell transplantation (HSCT), solid organ transplantation (SOT), autoimmune or vasculitic disease, advanced human immunodeficiency virus (HIV) infection, and chronic corticosteroid exposure.

While a useful starting point, this approach is limited by a lack of granularity and obscures heterogeneity. With the rising complexity of the modern immunocompromised host, many patients carry multiple diagnoses with overlapping or competing immune defects, making it difficult to identify a single dominant immune pathology resulting in intervention uncertainty. The number of diseases and exposures that may result in an immunocompromised state are also very broad and therefore creating a comprehensive list of possible conditions may be cumbersome. For example, patients with HIV, sick cell disease or history of splenectomy, may not fit into any of the above-mentioned categories. Furthermore, these conditions are not static and may evolve and differ within the same disease state over time (eg, a patient’s immunocompromised state at the time of HIV diagnosis verses at the time of follow-up after years of appropriate antiviral therapy). Importantly, immunosuppression may affect the immune system in terms of both quantity—as seen in neutropenia or lymphopenia—and quality—as seen in functional asplenia, chronic steroid use, or uremia, where cell counts may appear normal despite profound functional impairment. These important considerations remain agnostic to the disease-based classification system.

Immune Defect Classification

An alternative framework stratifies patients according to predicted baseline immune defects, categorized as barrier disruption, innate immune dysfunction, adaptive immune suppression, and immune reprogramming ( Fig. 1 , Table 1 ). This approach offers a more precise and flexible alternative to disease-based classification, compelling clinicians to consider multiple simultaneous forms of immune compromise and their associated infection risks.

Classification of immunocompromised state based upon immune deficiencies. Allo-HSCT, Allogeneic hematopoietic stem cell transplant; AML, Acute myeloid leukemia; CLL, Chronic lymphocytic leukemia; CMV, Cytomegalovirus; CVID, Common variable immune deficiency; EBV, Epstein-barr virus; ESBL, Extended spectrum beta-lactamase; HPV, Human papilloma virus; HSV, Herpes simplex virus; JAK, Janus kinase; JC, John Cunningham virus (human polyomavirus 2); MDS, Myelodysplastic syndrome; MRSA- Methicillin-resistant staphylococcus aureus; PJP, Pneumocystis jirovecii pneumonia; SOT, Solid organ transplant; TB, Tuberculosis; TNF, Tumor necrosis factor; VZV, Varicella zoster virus.

( Created in MS Power Point and pictograms were taken from SERVIER MEDICAL ART. Servier Medical Art is licensed under CC BY 4.0.)

Table 1

Immune defect driven risk stratification

Immune Defect	Common Causes	Infection Risk	Sepsis Phenotype	Management Implications
Barrier disruption	Chemotherapy, mucositis, lines	Gram-positive, gut translocation	Rapid bacteremia	Broad early antibiotics
Innate dysfunction	Neutropenia, steroids	Bacterial + fungal (Candida, Aspergillus)	Blunted inflammation	Consider antifungals early
Adaptive (T-cell)	HIV, calcineurin inhibitors	Viral, intracellular (TB, PJP)	Atypical presentations	Broaden to viral/opportunistic
Adaptive (B-cell)	Rituximab, myeloma	Encapsulated organisms	Recurrent infections	IVIG ± targeted antibiotics
Immune reprogramming	CAR-T, checkpoint inhibitors	Mixed (CRS vs infection)	Hyperinflammatory mimic	Immunomodulation vs antibiotics
Mixed defects	Most ICU patients	Broad spectrum	Heterogeneous	Individualized approach

Abbreviation: IVIG, intravenous immunoglobulin.

Barrier disruption

The outermost layer of the host’s defense system is the anatomic barrier that is, the patient’s physical envelope , which provides both mechanical and chemical protection against pathogen entry. Barrier defense encompasses skin integrity, mucosal membrane function, mucociliary clearance, and the absence of foreign bodies breaching sterile compartments. Barrier disruption is particularly prevalent in patients with hematological malignancy or those undergoing cytotoxic chemotherapy, in whom mucositis, chemotherapy-induced colitis, and indwelling central venous catheters are common. Another important patient population that cannot be overlooked in this category are burn patients, whose mortality due to nosocomial infection are significant. These disruptions facilitate translocation of cutaneous and gastrointestinal flora into the bloodstream, where even organisms of low intrinsic virulence, such as coagulase-negative staphylococci, can cause life-threatening sepsis.

Innate immune dysfunction

The innate immune system constitutes the first cellular line of defense, providing rapid, nonspecific responses to invading pathogens. Its principal functions include pathogen recognition, opsonization via complement activation, and phagocytic clearance, with neutrophils and monocytes serving as the primary effector cells. Neutrophils eliminate pathogens through phagocytosis, degranulation, and oxidative killing mechanisms. Monocytes differentiate into macrophages and dendritic cells, which not only perform phagocytosis but also serve as critical antigen-presenting cells (APCs), bridging innate and adaptive immunity by initiating downstream lymphocyte responses. Derangements may be seen in the early phase following chemotherapy, allogeneic hematopoietic stem cell transplantation, acute myeloid leukemia and aplastic anemia. Impairment of innate immune function, whether through quantitative depletion or qualitative dysfunction, increases susceptibility to both gram-positive and gram-negative bacterial infections, fungal infections including candidemia, and molds including invasive aspergillosis.

Adaptive immune suppression

Adaptive immunity comprises 2 principal arms: (1) cellular immunity, mediated by antigen-specific T lymphocytes activated through interactions with APCs; and (2) humoral immunity, mediated by B lymphocytes that differentiate into plasma cells or memory B cells and produce antigen-specific antibodies. T lymphocyte subsets, principally CD8 ⁺ cytotoxic T cells and CD4 ⁺ helper T cells, are activated following major histocompatibility complex-dependent antigen presentation by dendritic cells and macrophages. These subsequently proliferate to mount pathogen-specific effector responses. Cytotoxic T cells induce apoptosis of cells displaying foreign peptides on their surface, while helper T cells augment this response through macrophage activation and cytokine secretion including interferon-γ. Collectively, cellular immunity is directed primarily against intracellular pathogens-including mycobacteria, Listeria, Leishmania, and Histoplasma—and against virus-infected cells that cannot be targeted by circulating antibodies alone. The clinical consequences of cellular immune deficiency are well illustrated by HIV infection, in which progressive CD4 ⁺ T cell depletion produces a predictable hierarchy of opportunistic infections that forms the basis of well-established prophylactic guidelines. In non-HIV immunocompromised patients however, CD4 ⁺ dysfunction induced by corticosteroids, calcineurin inhibitors, and purine analogs are responsible for the mechanisms of cellular immunosuppression.

Unlike T cells, B cells recognize antigens directly via surface-expressed immunoglobulins. Upon activation, B cells differentiate into plasma cells that secrete antigen-specific antibodies, which mediate pathogen elimination through complement activation, promotion of opsonization and phagocytosis, and recruitment of innate immune effectors. While a detailed review of immunoglobulin class structure and function is beyond the scope of this article, 2 clinically important causes of functional verses depletion hypogammaglobulinemia warrant specific consideration in the context of the immunocompromised host. In multiple myeloma, overproduction of a nonfunctional monoclonal paraprotein paradoxically suppresses normal antibody production, rendering patients functionally hypogammaglobulinemic despite elevated total protein levels. The use of a common biologic, Rituximab, an anti-CD20 monoclonal antibody, depletes mature B cells while sparing CD20-negative plasma cells; hypogammaglobulinemia therefore develops in a delayed fashion as the existing plasma cell pool turns over in the periphery and cannot be replenished from the depleted B cell compartment. Patients with humoral deficiency from either qualitative or quantitative mechanism are at heightened risk of infection with encapsulated bacteria, including Streptococcus pneumoniae , Haemophilus influenzae, and Neisseria meningitidis, which exploit their polysaccharide capsule to resist phagocytosis in the absence of the targeted opsonizing antibodies. Recognition of this specific vulnerability should inform prophylactic vaccination strategies and empiric antimicrobial selection in this population.

Immune reprogramming

The advent of immune effector cell therapy has created a unique category of immune dysregulation that may include cytokine excess, immune exhaustion via PD-1 axis activation and a mixed hyper and hypoinflammatory state. Of note, chimeric antigen receptor (CAR)-T cell therapy exploits the same cytotoxic mechanisms described earlier to eliminate tumor cells; however, T cell activation may simultaneously trigger pathologic cytokine release, producing sepsis-mimicking syndromes including cytokine release syndrome (CRS) and immune effector cell-associated neurotoxicity syndrome (ICANS), which add to the diagnostic complexity and ambiguity at the bedside.

The epidemiology of sepsis is evolving in parallel with transformative advances in cancer therapeutics, transplantation, and immune-modifying agents. Current estimates indicate that 6.6% of adults in the United States were immunocompromised in 2021 based on reported health conditions, prescription medications, and medical treatments, representing a substantial increase from 2.7% in 2013. Up to 15% of ICU patients in European registries carry a cancer diagnosis. ^, This expanding demographic is driven, in part, by the increasing use of immunosuppressive medications.

Notably, cancer-specific mortality has declined substantially, with a 33% reduction since 1991 attributable to advances in preventive strategies, alongside the development of novel therapeutics. Despite these gains, patients with cancer face a nearly ten-fold greater risk of developing sepsis compared with the general population, constituting a vulnerable and growing subpopulation with substantial health care resource needs. A large epidemiologic study by Yang and colleagues, examining emergency department utilization in the United States from 2006 to 2012, identified approximately 4 million annual adult emergency visits involving patients with active cancer diagnoses, of whom 60% were admitted primarily for severe infections and associated complications.

Resource utilization and clinical outcomes are not uniform across malignancy types however, with considerable heterogeneity even within the hematological malignancy population. Among patients with solid organ tumors, sepsis-related mortality accounts for approximately 9% of all deaths, with the highest mortality rates (5%–13%) observed in the context of severe infections complicating postchemotherapy neutropenia. In contrast, 12% to 27% of patients with hematological malignancies develop sepsis within the first year of diagnosis, with an associated in-hospital mortality rate of 14%. ^, The incidence of severe sepsis is approximately 5 times higher among recipients of hematopoietic stem cell transplantation compared with hospitalized patients without a history of hematopoietic stem cell or solid organ transplantation. Collectively, these data underscore the growing public health significance of immunosuppression.

The pathophysiology of sepsis is characterized by a dysregulated immune response to infection, involving the synchronized activation and cross-talk between the innate and adaptive immune systems. The rapid, nonspecific effector functions of innate immunity followed by the sophisticated and targeting adaptive immune response, are both amplified by proinflammatory cascades mediated by cytokines, coagulation factors, and markers of endothelial and epithelial injury. This inflammatory cascade underlies the immune dysregulation hypothesis central to sepsis pathobiology and has been implicated in driving deleterious vasodilation, microvascular leak, microthrombotic formation, and acute organ injury. A concurrent anti-inflammatory response also exists, which may, when persistently dysregulated, contribute to subacute adaptive immune exhaustion and heightened susceptibility to secondary infections.

The determinants of the host immune response to infection are numerous and likely encompass a broad range of endogenous and exogenous factors, including genetic composition, age, biological sex, comorbidities, metabolic state, prior microbial exposure, and concurrent medications. The interplay of these variables gives rise to profound interindividual heterogeneity in immune response. Over the past decade, significant progress has been made in characterizing this heterogeneity through the discovery of molecular phenotypes of sepsis. Approaches leveraging plasma protein biomarkers, transcriptomics, and metabolomics combined with unsupervised clustering techniques have identified biologically distinct patient subgroups with differential risks of adverse outcomes and heterogeneity in treatment effects with supportive and targeted therapies. Collaborative efforts, such as the Subtyping in Sepsis and Critical Illness Consortium, aim to harmonize findings across independent research groups to define generalizable patient cohorts characterized by distinct molecular pathways and immune cell-specific dysregulation. These initiatives hold promise for a biology-driven classification framework that may inform future prospective targeted trials to test novel and supportive interventions. However, phenotypic frameworks have been derived in immunocompetent patient cohorts shaped by carefully curated inclusion criteria and therefore may not be generalizable immunocompromised patients.

Efforts to address this gap are nonetheless emerging. Notably, studies by Ronner and colleagues and Weingart and colleagues have investigated how baseline perturbations in host immune function may influence molecular phenotype assignment and its generalizability. ^, Both studies examined the performance of the inflammatory subphenotypes of sepsis, developed by Calfee and colleagues. Weingart and colleagues examined the relationship between immunocompromised status and subphenotype assignment, demonstrating that patients with hematological malignancies were disproportionately classified as hyperinflammatory and that this phenotype was significantly associated with impaired survival. Interestingly, elevated cytokine levels of interleukin (IL)-8, a powerful neutrophil chemoattractant, emerged as a key determinant of phenotype assignment in this population, highlighting retained mechanistic relevance even in patients who are often plagued by neutropenia. Ronner and colleagues found that unlike in immunocompetent cohorts, parsimonious models incorporating IL-6 and IL-8 inconsistently predicted hospital mortality and discordantly classified patients by inflammatory phenotype, particularly among those with leukemia. Collectively, these studies suggest that immunocompromised patients may not easily mirror existing sepsis phenotypes. Simply put, the assumption that immunocompromised patients are uniformly hypoinflammatory is overly simplistic. In reality, what may exist is a phenotypic spectrum that spans various categories including hyperinflammatory, hypoinflammatory, mixed hyper–hypo and immunoparalysis. Temporal evolution likely occurs, but longitudinal data in immunocompromised cohorts are sparse.