PEDIATRIC SEPSIS: EVOLVING SUBPHENOTYPES, OVERLAP SYNDROMES, AND THE PATH TO PRECISION THERAPIES

I. Introduction

Sepsis remains a leading cause of in-hospital mortality and critical illness worldwide. Data from 2017 estimate that 49 million people, including 25 million children, experience sepsis annually.^1–3 Despite the high burden of pediatric sepsis, standardized criteria with high specificity and positive predictive value for pediatric sepsis-related mortality were lacking until a Society of Critical Care Medicine (SCCM) task force released the 2024 International Consensus Criteria for Pediatric Sepsis and Septic Shock (“Phoenix criteria”). The Phoenix criteria introduced the Phoenix Sepsis Score (PSS) for identification of pediatric sepsis using clinically available characteristics.^4,5

Prior to 2024, the most recent pediatric-specific sepsis and septic shock criteria were established by the International Pediatric Sepsis Consensus Conference (IPSCC) in 2005 which defined pediatric sepsis as ≥2 out of 4 systemic inflammatory response syndrome (SIRS) criteria in the presence of an infection.^6,7 The IPSCC criteria adopted the structure of the adult consensus conference statements in 1991 (Sepsis-1)⁸ and 2001 (Sepsis-2)⁹ that sepsis is the consequence of immunoinflammatory damage from infection-induced SIRS, and provided age-specific vital sign (temperature, heart rate, respiratory rate) and laboratory (leukocyte count) parameters for diagnosis.^6,7 However, insights from the past two decades – including failed trials of targeted anti-inflammatory therapies^10–15 – suggest sepsis is not driven solely by hyperinflammation.^16,17 Newer data suggests that sepsis is better characterized as immune dysregulation resulting from an imbalance between pro- (SIRS) and anti-inflammatory (compensatory anti-inflammatory response syndrome [CARS]) processes, which leads to organ dysfunction.^16–18

This framework was reflected in the 2016 Third International Consensus Conference for Sepsis and Septic Shock (Sepsis-3) revised definition of adult sepsis as life-threatening organ dysfunction caused by a dysregulated host response to infection.¹⁹ This consensus statement also recommended use of the Sequential Organ Failure Assessment (SOFA) Score for the clinical identification of adults with sepsis.¹⁹ This paradigm shift was supported in pediatric literature demonstrating that SIRS-based criteria, when compared to pediatric organ dysfunction scores like PEdiatric Logistic Organ Dysfunction 2 (PELOD-2) Score, lacked the specificity to identify children with increased risk of mortality in the setting of an infection.^20–23 However, since SOFA was neither developed with pediatric data nor validated for pediatric patients, there was no consensus on a standardized, organ dysfunction-based scoring system for diagnosing sepsis in patients <18 years of age until the development of the PSS.^4,5 Table 1 highlights the differences between current diagnostic criteria for adult and pediatric sepsis.

Table 1:

Comparison of Adult-Based vs. Pediatric-Specific Sepsis Criteria

Organ System	Phoenix Sepsis Score29	SOFA Score30
Respiratory	P/F ratio, S/F ratio	P/F ratio
Cardiovascular	Hypotension, Vasopressors, Lactate	Hypotension, Vasopressors
Coagulation	Platelets, INR, D-dimer, Fibrinogen	Platelets
Neurologic	GCS, Pupillary response	GCS
Hepatic	Not included *	Bilirubin
Renal	Not included *	Creatinine, Urine output

^*Hepatic dysfunction (total bilirubin ≥4 mg/dL and/or ALT >102 IU/L) and Renal dysfunction (creatinine above age-based thresholds) are included in an expanded scoring system known as “Phoenix-8” which also includes benchmarks for endocrine and immunologic dysfunction. However, since PSS had comparable performance to Phoenix-8 during validation, the SCCM task force voted to promote PSS given its parsimony and generalizability to low resource settings.⁴

While development of the Phoenix criteria was a crucial first step in improving the identification of children with sepsis with high risk of mortality, these criteria fail to capture the biologic complexity of sepsis. Furthermore, unlike Sepsis-3, the Phoenix criteria do not explicitly define sepsis as a state of immune dysregulation. There is emerging evidence from large-scale electronic health record²⁴ and multi-omic profiling^25–27 studies suggesting that the heterogeneity of pediatric sepsis requires more granular definitions. Additionally, when compared to adult sepsis, developmental immunology and age-specific physiology influence the pediatric host response to infection, necessitating unique diagnostic and therapeutic considerations for children.²⁸ This review synthesizes current knowledge of pediatric sepsis as a distinct clinical entity with emphasis on developmental immunology, immunologic subtypes, overlap syndromes (like hemophagocytic lymphohistiocytosis [HLH] and thrombotic microangiopathy [TMA]), and opportunities for precision-guided treatment. By integrating the latest evidence on subphenotypes, biomarkers, and targeted immunomodulation, this article aims to provide a practical and forward-looking framework for clinicians and researchers seeking to improve recognition, treatment, and outcomes for children with sepsis.

II. Discussion

A. Epidemiology and Evolving Risk Profiles

The epidemiology of pediatric sepsis varies by geographic region, healthcare access, and socioeconomic status. While pediatric sepsis remains a significant cause of morbidity and mortality globally, its true burden of disease is unknown. In 2017, there were an estimated 3 million sepsis-related deaths in patients younger than 5 years old and 500,000 sepsis-related deaths in patients aged 5 to 19 years old worldwide.³ Recognizing the global impact of sepsis, the World Health Organization (WHO) designated the prevention, diagnosis, and management of sepsis as a global health priority in 2017.³¹ This initiative is aligned with the WHO’s Sustainable Development Goal of minimizing preventable deaths in neonates and children under 5 years of age by 2030.³²

Even before this call, the need to improve the management of patients with sepsis and septic shock was recognized by SCCM and the European Society of Intensive Care Medicine (ESICM), who together launched the Surviving Sepsis Campaign (SSC) in 2001.^33,34 The SSC released its initial guidelines in 2004, which have been reviewed and revised every four years since its inception.³³ In 2020, SCCM and ESICM released their “Surviving Sepsis Campaign International Guidelines for the Management of Septic Shock and Sepsis-associated Organ Dysfunction in Children” (pediatric SSC guidelines).³⁴ This was the first set of comprehensive, international guidelines for the management of children with sepsis and septic shock.

A pediatric-specific systemic review and meta-analysis found that pediatric sepsis-related case-fatality rates (CFR) have declined when comparing studies from 1991-2000 to those from 2011-2016.³⁵ This secular trend likely reflects improvements in public health, advancements in pediatric critical care medicine, and the impact of the SSC.^35,36 However, developing countries (31.7% CFR) continue to bear a disproportionate burden of pediatric-sepsis related mortalities when compared to developed countries (19.3% CFR), with increased mortality rates associated with younger age and the presence of shock.³⁵ This disparity is mirrored by findings in the PSS derivation dataset, where children with sepsis in the first 24 hours of hospitalization had 28.5% mortality in lower-resource settings versus 7.1% mortality in higher-resources settings.⁴ A retrospective study externally validating the PSS in a cohort of children with suspected community-acquired sepsis in Australia and New Zealand found a similar in-hospital mortality rate of 10.8%.³⁷

As pediatric sepsis mortality rates decline, there has been increased attention on the morbidities associated with pediatric sepsis survivorship. The Life After Pediatric Sepsis Evaluation (LAPSE) investigation was a prospective study of long-term functional status and health-related quality of life (HRQL) in sepsis survivors. This study found that 51% of survivors had significant chronic comorbid conditions, and after one year 35% of survivors had not regained their baseline HRQL.³⁸ Advancements in pediatric resuscitation and critical care medicine have improved patient survival, but this trend has also overlapped with a rise in children with medical complexity (CMC) and children with technology dependence.^39,40 Overall, CMC are increasing in prevalence and represent a vulnerable population at risk for higher mortality rates and worse outcomes from sepsis.^41–43 The rising prevalence of CMC also reflects increased survival of infants born prematurely, children with congenital anomalies, and those with other chronic conditions.⁴¹ A separate multicenter PICU cohort study found that immunocompromised diagnoses were present in 28% of children admitted for sepsis and 39% of patients dying with sepsis.⁴⁴ It is important to recognize the intersection of acute and chronic illness and the rising prevalence of at-risk subpopulations, such as CMC and immunocompromised children, as pediatric sepsis diagnosis and management evolves.

B. Age-Related Immunology and Pathophysiology

Pediatric patients with sepsis exhibit significant differences in innate and adaptive immunity across various age groups from neonates to adolescents, and these responses differ significantly from those observed in adults. Neonates rely heavily on their innate immune response due to the immaturity of their immune system. Sepsis pathobiology in neonates is primarily characterized by dysfunctional innate immune cells, which are less effective in pathogen recognition and response.⁴⁵ Neonatal monocytes and dendritic cells produce less IL-12p70 and IFN-α, which are crucial for Th1-type responses, but more IL-6 and IL-23, supporting Th17-type immunity.⁴⁶ This skewed cytokine profile results in a reduced ability to combat intracellular pathogens effectively. Neonatal septic shock is also marked by a predominantly downregulated transcriptome, with reduced expression of genes involved in both innate and adaptive immunity.⁴⁷ Additionally, neonatal CD4⁺ T cells are developmentally programmed to be anti-inflammatory and tolerogenic, biased toward regulatory rather than effector phenotypes.⁴⁸

Infants and toddlers show a rapid maturation of both innate and adaptive immune responses. Whole-blood transcriptomic analyses indicate that septic infants cluster immunologically closer to older children than to neonates, and infants and toddlers demonstrate a much broader upregulation of genes in sepsis compared to neonates.⁴⁹ By preschool age, innate immunity becomes more developed, and CD4⁺ and CD8⁺ T cells begin to mature.⁵⁰ However, infants still demonstrate an altered transcriptomic response to sepsis compared to older children, indicating a less regulated immune response.^49,51 School-age children and adolescents develop more robust adaptive immunity and can mount very high circulating levels of cytokines like IL-6, IL-8, and IL-17 during septic shock, indicating potent activation of neutrophils, monocytes, and T cells.⁴⁹ They also exhibit higher levels of regulatory and memory T cells, which are positively associated with age in sepsis.⁵² Both pro- and anti-inflammatory cytokines surge in children with sepsis during the acute phase, signifying a robust inflammatory response.^53,54 Accordingly, pediatric sepsis morbidity and mortality is less pronounced in older children,^55,56 though high-risks subgroups exist across the age spectrum.⁴⁴

Within this developmental context, poor outcomes in pediatric sepsis are typically associated with markedly dysregulated immune responses to infection. The result is a dynamic inflammatory course, with an initial cytokine storm that causes widespread tissue damage, and a subsequent compensatory anti-inflammatory phase, which may lead to an immunosuppressed state, increasing vulnerability to secondary infections.⁵⁷ Hyperinflammation in pediatric sepsis often results in multiple organ dysfunction syndrome (MODS) from unchecked cytokine storm.⁵⁸ In severe cases, pediatric patients with sepsis develop secondary HLH/macrophage activation syndrome (MAS) – an extreme hyperinflammatory sepsis overlap syndrome characterized by hyperferritinemia, liver dysfunction, coagulopathy, and excessive T cell activation.^59,60 Conversely, immunoparalysis in pediatric sepsis occurs in the setting of a dysregulated compensatory anti-inflammatory response and leads to diminished capacity to fight infection, and increased risk of secondary or nosocomial infections.⁶¹ Hallmarks of immunoparalysis in children include loss of HLA-DR expression on circulating monocytes^53,62–64 and impaired ability of their immune cells to respond to antigen stimulation.^65–67 Due to these immune dysregulated syndromes, pediatric sepsis mortality follows a bimodal distribution across the time course of illness, with about half of deaths occurring early (<24 hours) due to overwhelming inflammatory response, and the other half occurring late (>7 days) due to inability to restore immune homeostasis leading to opportunistic bacterial, viral, and fungal disease.⁶⁸

C. Pediatric Sepsis Subphenotypes and Overlap Syndromes

Pediatric sepsis is increasingly recognized as a heterogeneous syndrome driven by distinct biologic endotypes. Investigators have leveraged clinical data and omics technologies to define subphenotypes that go beyond the classic “one-size-fits-all” view. Hector Wong and colleagues were pioneers in this area, identifying multiple transcriptomic subgroups in pediatric septic shock⁶⁹ and later deriving parsimonious human and murine models of these sepsis subgroups.⁷⁰ Building on this, a recent multicenter cohort study identified two reproducible sepsis subgroups by integrating clinical and gene-expression data.⁷¹ One subgroup was characterized by a hyperinflammatory state, with neutrophil-dominated innate immune activation and a cytokine surge causing endothelial and microvascular injury. In contrast, the second phenotype exhibited a more tempered inflammatory profile. Other investigators have used vital sign trajectories⁷² patterns of organ dysfunction,²⁴ and biologic phenotypes⁷³ to identify reproducible sepsis subgroups with prognostic significance and evidence of heterogeneity of treatment effect. These contributions illustrate that pediatric sepsis encompasses shock variants ranging from hyperinflammatory to immunosuppressive, each with distinct prognostic signatures and clinical trajectories.

Complicating the sepsis landscape are overlap syndromes such as HLH and TMA, which can be precipitated by severe infection and sepsis. HLH/MAS is a severe hyperinflammatory state of immune dysregulation that can masquerade as protracted sepsis. Clinically, HLH is distinguished by marked hyperferritinemia, typically with ferritin >500-1000 ng/mL, and cytopenias, driven by immune-mediated hemophagocytosis.⁷⁴ Additional hallmark features of HLH include persistent fevers, hepatosplenomegaly, and coagulopathy. In septic children, the emergence of very high ferritin levels and new-onset pancytopenia should prompt evaluation for an HLH overlap syndrome.⁶⁰ While both pediatric and adult sepsis patients can develop HLH/MAS, the condition is more frequently recognized and diagnosed in children.^75–77 It is theorized that neonatal and pediatric patients may present with more profound pro-inflammatory responses during the acute phase of sepsis, which may lead to a higher clinical suspicion and diagnosis of HLH/MAS.⁵⁴ This concept of “hyperferritinemic sepsis” defines a particularly high-risk subset: these patients exhibit extreme cytokinemia (e.g. elevated IL-6, IL-18, sCD25), evidence of immune exhaustion (low ex vivo monocyte TNF-α response), often meet MAS criteria, and have significantly increased mortality risk.⁷⁵

Another pediatric sepsis overlap syndrome is TMA. Sepsis-induced secondary TMA, also referred to as thrombocytopenia-associated multiple organ failure (TAMOF), is characterized by microangiopathic hemolysis with the presence of schistocytes on peripheral smear, rapid platelet consumption, and endothelial injury leading to disseminated microthrombi.⁷⁸ A hallmark laboratory finding is severely reduced ADAMTS13 activity, reflecting failure to cleave ultra large von Willebrand multimers and a pathophysiology akin to thrombotic thrombocytopenic purpura (TTP).⁷⁹ However, a key mechanistic distinction between sepsis-induced secondary TMA and TTP is the lack of ADAMTS13 autoantibodies in TMA.⁷⁹ Prompt plasma exchange to replenish ADAMTS13 and remove pro-thrombotic mediators has been studied as a life-saving intervention in sepsis-associated TMA, however, evidence supporting its use is limited.^80,81

A successful precision medicine approach to the immunologic heterogeneity of sepsis requires an understanding of the molecular mechanisms underlying subphenotypes.⁸² As inferred from past failed randomized clinical trials in sepsis, uniformly administering immunomodulatory therapies without patient stratification is unlikely to be successful. However, as suggested by a reanalysis of the negative PROWESS-SHOCK (Prospective Recombinant Human Activated Protein C Worldwide Evaluation in Severe Sepsis and Septic Shock) trial demonstrating varied response to treatment with Drotrecogin alfa (recombinant human activated protein C) across biomarker-defined sepsis subgroups,⁸³ identification of sepsis subphenotypes with deep immune phenotyping may allow for future trials to overcome the immunologic heterogeneity that led to the demise of prior precision medicine efforts. Additionally, HLH/MAS and TMA represent critical overlap syndromes in septic children: their hallmark labs (ferritin, cytopenias for HLH; schistocytes, low ADAMTS13 for TMA) serve as red flags, and the institution of targeted therapies can be decisive for patient outcomes.

D. Biomarkers and Targeted Immunomodulation for a Precision Medicine Approach to Pediatric Sepsis

With the identification of pediatric sepsis subphenotypes, critical overlap syndromes, and advancements in omics technologies, a successful precision medicine approach to pediatric sepsis is within our grasp. However, this approach has not yet been translated to the bedside. For risk stratification, clinicians still rely on nonspecific biomarkers such as lactate, a byproduct of anaerobic metabolism used to determine the degree of tissue hypoperfusion in shock.⁸⁴ However, no cutoffs have been established to define clinically meaningful pediatric sepsis-associated hyperlactatemia, thus lactate levels are used as part of the comprehensive assessment of a child’s perfusion.³⁴ Other commonly used biomarkers for infection related inflammation include C-reactive protein (CRP) and procalcitonin, but they lack specificity for immunologic phenotypes and may be best interpreted along dynamic immune profiling.^85–90 Additionally, cytokine panels and immune cell markers have been increasingly proposed as a means of assessing the degree of a patient’s immune dysregulation, with hyperactivation of specific arms of the immune system theoretically allowing for targeted immunomodulation.^86,90–93

Similarly, current guidelines for the management of pediatric sepsis, as outlined by the pediatric SSC guidelines, have not adopted precision medicine approaches. The pediatric SSC guidelines emphasize the importance of recognizing pediatric sepsis early and then administering broad-spectrum antibiotics and hemodynamic support promptly.³⁴ There are additional weak recommendations for nonspecific immunomodulatory interventions such as corticoids, intravenous immunoglobulin (IVIG), and extracorporeal therapies.³⁴

Emerging insights into sepsis subphenotypes has also allowed for revamped clinical trial efforts based on stratification with rapid, real-time immunophenotyping. Trials like PROVIDE (a Personalized Randomized trial Of Validation and restoration of Immune Dysfunction in severE infections and Sepsis) trial⁹⁴ and its pediatric analog, the PRECISE (Personalized Immunomodulation in Pediatric Sepsis-induced MODS) trial,⁹⁵ exemplify efforts to stratify sepsis patients by subphenotypes to guide targeted therapies. PROVIDE and PRECISE call for biomarker-based immune classification, with hyperinflammatory macrophage activation-like syndrome (MALS) characterized by elevated ferritin and immunoparalysis characterized by low monocyte HLA-DR expression (PROVIDE) or whole blood ex vivo TNF-α production capacity (PRECISE). Early results from PROVIDE suggest that anakinra, an IL-1 receptor antagonist improves 7-day outcomes in adult patients with sepsis and MALS.⁹⁴ Additionally, while MALS and immunoparalysis represent two extremes of sepsis-related immune dysregulation, there are also states of indeterminate immune dysregulation between these two states characterized by emerging endotypes, like IFN-γ driven sepsis (IDS)⁹⁶, which is being targeted with emapalumab in the recently opened EMBRACE (Emapalumab Treatment for Anticipated Clinical Benefit in Sepsis Driven by the Interferon-Gamma Endotype) clinical trial.

Understanding sepsis subclasses and overlap syndromes is essential for therapeutic stratification in pediatric sepsis: an immunoparalysis-dominant patient might benefit from immune stimulants (e.g. GM-CSF or IFN-γ), whereas a hyperinflammatory HLH-like patient might require aggressive immune suppression (e.g. steroids, anakinra) or blood purification techniques (e.g. therapeutic plasma exchange, high-volume hemofiltration). Similarly, a child with TMA features may benefit from plasma exchange that would not be indicated in other sepsis phenotypes. Embracing this complexity through refined subclassification and targeted treatments holds promise to enhance outcomes in pediatric patients with sepsis, turning a long-recognized diagnostic challenge into an opportunity for precision therapy and improved patient care. Furthermore, despite the increasing prevalence of children with medical complexity and immunocompromised children in pediatric critical care, the lower burden of comorbidities in children makes pediatric sepsis an optimal cohort, with potentially cleaner immunobiologic signatures, for studying sepsis subphenotypes and their underlying mechanisms.

III. Summary

Pediatric sepsis is a biologically distinct syndrome that spans a dynamic immune continuum shaped by developmental immunology, age-specific physiology, and exposure to overlapping hyperinflammatory disorders such as HLH and TMA. Unlike in adults, where organ-dysfunction scores are derived from mature physiology, children mount age-dependent transcriptomic and immunometabolic responses that give rise to complex subphenotypes which we are just beginning to understand. Large electronic-health-record studies and multi-omic profiling have confirmed that these endotypes carry disparate risks of organ failure and mortality and, critically, exhibit heterogeneity of treatment effect with existing interventions. Accordingly, recently published Phoenix Sepsis Score criteria and emerging rapid immune-profiling tools together provide a framework for earlier, more granular recognition that can guide precision resuscitation and immunomodulation in real time.

The field now stands at an inflection point: future progress hinges on rigorous, pediatric-specific investigations that deliberately stratify patients by biologic endotype and overlap syndrome. Future prospective trials must embed adaptive designs that test targeted agents (e.g., IL-1/IL-6 blockade for hyperferritinemic states, immune adjuvants for immunoparalysis, plasma exchange for sepsis-associated TMA) within clearly defined subgroups, while simultaneously validating bedside biomarker panels and parsimonious gene-expression classifiers for rapid diagnostics. Multinational consortia, harmonized data definitions, and interoperable omics repositories will be pivotal to power these studies and translate bench insights into bedside precision. Recognizing and operationalizing pediatric sepsis subphenotypes will transform this historically monolithic diagnosis into a set of actionable, biologically coherent entities, laying the groundwork for the next generation of precision therapeutic approaches.

Clinics Care Points:

• Maintain a High Index of Suspicion

• Pediatric sepsis often presents with subtle or nonspecific signs (e.g., poor feeding, lethargy) rather than classic adult features like hypotension. Early screening tools should be tailored to age-specific physiology and developmental status.

• Consider Overlap Syndromes Early

• Children with persistent fever, elevated ferritin, cytopenias, or unexplained organ dysfunction may have HLH or TMA overlapping with sepsis. Prompt recognition and specialized treatment can significantly impact outcomes.

• Tailor Immunomodulatory Therapies

• Not all pediatric septic patients benefit from the same adjunctive interventions. Multidisciplinary collaboration is necessary to guide the use of immunomodulation (e.g. corticosteroids, IVIG, and cytokine inhibitors) in children with sepsis and hyperinflammatory overlap syndromes.

Key Points:

• Pediatric sepsis is highly heterogeneous, with age-dependent immune responses and distinct subphenotypes not typically seen in adults.

• Recent data suggest many pediatric patients experience a coexisting hyperinflammatory and immunoparalyzed state, requiring real-time immune monitoring.

• Overlap syndromes (e.g., HLH, TMA) are increasingly recognized and call for multidisciplinary evaluation and management.

• Precision-based approaches leveraging omics technologies may improve early identification and guide immunomodulatory treatments.

Synopsis:

Pediatric sepsis is a unique syndrome shaped by developmental considerations, age-specific physiology, immunologic heterogeneity, and hyperinflammatory overlap syndromes. Recent evidence underscores a dynamic interplay between hyperinflammation and immunoparalysis in pediatric patients with sepsis, emphasizing the need for rapid, personalized diagnostic tools. While revised pediatric scoring systems and targeted immunotherapies have been developed, significant knowledge gaps remain regarding optimal timing and patient selection for immunomodulatory interventions. This review consolidates current understanding of pediatric sepsis pathobiology, explores emerging pediatric subphenotypes, and examines early-stage clinical trials investigating precision medicine approaches in pediatric sepsis.