Soluble CD25 as a potential prognostic biomarker for pediatric septic shock and its comparison with hemophagocytic lymphohistiocytosis Authors: T B Garcia, V Knight, H F Scott Published: 2026 ABSTRACT AbstractObjectivesIdentifying children at highest risk for deterioration during evaluation for septic shock in emergency department (ED) settings remains challenging. Soluble CD25 (sCD25), a hemophagocytic lymphohistiocytosis (HLH) diagnostic criterion, has been associated with sepsis mortality in ICU settings. This study aimed to determine whether sCD25 is associated with adverse outcomes in children undergoing initial septic shock evaluation in a pediatric ED. We also sought to compare sCD25 in patients with septic shock, infection without sepsis, and HLH.ResultsChildren with HLH demonstrated significantly higher sCD25 levels [25,800 pg/mL (IQR: 1620–42,300)] compared with those with septic shock [5200 pg/mL (IQR: 2500–20,600)] or infection [5500 pg/mL (IQR: 2200–12,500)]. No statistical differences were observed between septic shock and infection groups; however, a subset of septic shock patients exhibited sCD25 elevations comparable to HLH patients. These patients experienced severe outcomes, including 100% PICU admission, 50% vasoactive support requirement, and prolonged hospitalization (16 days). Bloodstream infections were associated with the highest sCD25 levels. These findings suggest sCD25 elevation may be associated with adverse outcomes in children with septic shock. Larger studies are warranted to validate sCD25 as a prognostic marker for children with septic shock.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1186/s13104-026-07923-2.Keywords: Hemophagocytic lymphohistiocytosis, Pediatrics, Sepsis, Septic shock, Soluble CD25 Introduction Pediatric septic shock represents a significant cause of morbidity and mortality worldwide [1, 2]. Identifying patients at risk for adverse outcomes during early sepsis evaluation remains a challenge. Biomarkers that are detectable at initial emergency department (ED) evaluation and are associated with severe outcomes including fluid-refractory shock, requirement for positive pressure ventilation (PPV), and prolonged hospitalization are needed for more effective prognostication and identification of patients that may benefit from novel, targeted therapies. Patients with sepsis who develop concurrent hemophagocytic lymphohistiocytosis (HLH) have poor outcomes including increased mortality, suggesting markers of HLH may have prognostic utility in sepsis [3]. Up to one third of children who die of sepsis have evidence of concurrent HLH [4, 5]. Surrogate markers of macrophage activation in HLH including ferritin and soluble CD163 have been associated with mortality in children with sepsis. However, it is unclear whether markers associated with abnormal T cell activation, a key component of HLH pathogenesis, may be useful in early sepsis prognostication. Soluble CD25 (sCD25) is a surrogate marker of T cell activity and HLH diagnostic criterion [6]. sCD25 has high sensitivity and specificity for HLH diagnosis, and elevation is associated with disease severity, response to therapy, and mortality [6–8]. In adult sepsis, sCD25 has shown promise in distinguishing patients with bacterial infections and high risk of mortality [9–11]. Likewise, sCD25 elevation has been associated with increased mortality in children with sepsis admitted to an ICU [12]. However, the prognostic value of sCD25 in early sepsis evaluation in children remains unexplored. This study sought to describe sCD25 levels in pediatric patients with suspected septic shock and to determine whether higher levels were associated with illness severity. Additionally, we compared sCD25 between patients with suspected septic shock, those with infection without sepsis, and those with HLH. Methods Participants This was a prospective cohort study conducted at Children’s Hospital Colorado, an academic pediatric referral center that treats 75,000 patients yearly in the ED, over the period from March 2024 to April 2025. Children who presented to the ED were eligible for enrollment in the suspected septic shock cohort if they were ≥ 12 months to < 18 years old, had clinical concern for sepsis defined as use of a sepsis order set, had at least one hypotensive systolic blood pressure for age defined as less than 70 + 2(age in years), and required an IV for clinical care [13]. Patients less than 12 months old were excluded due to age-related variability in sCD25 levels in this group [14]. Patients were eligible for enrollment in the infectious control population if they were ≥ 12 months to < 18 years of age, had evidence of infection (defined as documented fever > 38.0 °C in the emergency department or within 24 h preceding the encounter), required intravenous access for clinical care, and had no clinical concern for sepsis. Age matching with the sepsis population was performed using the following brackets: 1–5 years, 6–10 years, and 11–17 years. Patients with immunodeficiencies, cancer not in remission, history of a solid organ transplant, or were receiving immunomodulatory medications including systemic corticosteroids within one week of ED presentation were excluded to minimize heterogeneity of the study population. This study was approved by the Colorado Multiple Institutions Review Board. To identify patients with a diagnosis of HLH, all patients throughout the hospital system (including inpatients and ED patients) who had a sCD25 level ordered for clinical care between January 1, 2022 and April 30, 2024 underwent retrospective chart review. Patients who met HLH-2004 diagnostic criteria as determined by the primary team, hematology, or rheumatology were included in this cohort. Patients with concerns for concurrent sepsis were excluded from this group. Outcome The primary outcome was to determine the association between sCD25 levels and a clinical diagnosis of sepsis in patients undergoing evaluation for septic shock in the ED. Secondary outcomes were comparison of sCD25 levels between patients with sepsis, HLH, and infection without sepsis and association of sCD25 with key outcomes (PICU admission, PPV requirement, vasoactive medication use, pediatric logistic organ dysfunction-2 (PELOD-2) score, infection source, and positive bacterial culture) in patients with sepsis. Data collection The study team identified potential subjects who had use of an ED sepsis order set (septic shock cohort) or had fever within the past 24 h without use of a sepsis order set (infection cohort). Treating physicians were then asked to complete a screening form to assess inclusion and exclusion criteria. Per US law and IRB regulations, consent from a parent or legal guardian is required for all children < 18 years old. Written informed consent was obtained from parents or guardians of all subjects, and written assent was obtained from children ≥ 7 years old. After consenting patients were enrolled, a single venous sample was collected for sCD25 measurement. sCD25 levels were measured by the Children’s Hospital Immunopathology and Hematopathology Laboratory using a commercial enzyme linked immunosorbent assay (ELISA) kit (Human sIL-2R Instant ELISA, Invitrogen). Institutional cutoffs were used to determine sCD25 elevation (Supplemental Table S1). Clinical data were collected for all patients including age, gender, chronic medical conditions, history of sepsis within the past year, vital signs at presentation and daily for the first 2 days of hospitalization, admitting service, discharge date, vasoactive medication requirement, respiratory support requirement, clinical diagnosis, and microbial culture results. Missing data was assumed to be normal. On days where multiple data points were available, the most abnormal value was recorded. For patients in the HLH cohort, day 0 was considered the day that the first sCD25 level of admission was obtained in order to compare data between patients with septic shock and HLH at the time the sCD25 was measured. Statistical analysis Demographic and clinical characteristics were described using counts and percentages for categorical variables, and medians with interquartile ranges for continuous variables with non-normal distributions. Normality was assessed with the Shapiro–Wilk test. For normally distributed data, t-tests or one-way analysis of variance (ANOVA) were used, as appropriate. Mann–Whitney U tests and Kruskal–Wallis tests were used for nonparametric data. We planned a subgroup analysis of all sepsis patients with a sCD25 level above the diagnostic criteria for HLH based on our institutional age-specific reference ranges (Supplemental Table S1). All analyses were conducted using GraphPad Prism 10 (GraphPad Software, La Jolla, CA). Results Study population Between March 2024 and April 2025, 7 children with suspected septic shock, 7 with infection without sepsis, and 28 with HLH were enrolled (Table 1). There was no statistical difference in age between the cohorts (p 0.40). PICU admission rates were the same for septic shock and HLH groups (43% each; p > 0.99) with no PICU admissions in the infection without sepsis group. HLH patients had longer hospitalizations (median 9 days [IQR: 4–50]) compared to septic shock patients (4 days [IQR: 0–20] p < 0.01). Although higher percentages of septic shock patients required vasoactive medications (29% vs 18%, p 0.40) and positive pressure ventilation (43% vs 29%, p 0.22) than HLH patients, these differences were not statistically significant. Regarding infectious etiology, patients with septic shock demonstrated higher rates of positive bacterial cultures compared to other groups, though this difference was not statistically significant (sepsis 29% vs. infection 0% vs. HLH 7%, p 0.24). Conversely, patients with HLH showed higher rates of viral infections, but this difference also did not reach statistical significance (HLH 43% vs. sepsis 0% vs. infection 14%, p = 0.06). The most common etiology of HLH was an underlying rheumatologic condition (46%) followed by unknown cause (21%), infection (11%), CAR T-cell therapy (11%), genetic (7%), and malignancy (4%). Table 1Study populationSepsis (n = 7)No Sepsis (n = 7)HLH (n = 28)pAge12 (8–14)11 (8–16)9 (2–15)0.40% Female6 (86)3 (43)14 (50)0.18Chronic medical condition6 (86)1 (14)21 (75) < 0.01Neuromuscular4 (57)1 (14)3 (11)0.04Pulmonary4 (57)05 (18)0.04Endocrine/metabolic001 (4) > 0.99GI/liver2 (29)01 (4)0.13Renal1 (14)01 (4)0.56Hematologic1 (14)05 (18)0.81Cardiac5 (71)01 (4) < 0.01Rheumatologic0013 (46) < 0.01History of sepsis within the past year3 (43)01 (4)0.03Clinical OutcomesPICU admission3 (42.9)012 (42.9) > 0.99Non-PICU admission2 (29)3 (42.9)16 (57.1)0.44Length of stay (d)4 (0–20)0 (0–1)9 (4–50) < 0.01Positive pressure ventilation3 (43)08 (29)0.22Vasoactive agent2 (29)05 (18)0.40Primary infection sourceBlood1 (14)08 (29)0.30Pneumonia2 (29)2 (29)5 (18)0.62Meningitis000Urinary tract1 (14)00Skin/soft tissue/musculoskeletal000Other2 (29)4 (57)2 (7) < 0.01Unknown2 (29)1 (14)1 (4)0.10Pathogen identifiedBacteria2 (29)02 (7)0.24Virus01 (14)12 (43)0.06Fungus000None5 (71)6 (86)15 (54)0.32aCategorical values are displayed as counts and percentagesContinuous variables are shown as median and interquartile (IQR) range values Study population aCategorical values are displayed as counts and percentages Continuous variables are shown as median and interquartile (IQR) range values Comparison of sCD25 levels between groups HLH patients demonstrated significantly higher sCD25 levels (25,800 pg/mL [IQR: 1,620–42,300]) compared to both septic shock (5,200 pg/mL [IQR: 2,500–20,600]) and infection (5,500 pg/mL [IQR: 2,200–12,500] p < 0.001) groups (Fig. 1). While no statistical differences were detected between septic shock and infection cohorts, a subset of septic shock patients exhibited elevated sCD25 levels comparable to those observed in HLH patients (Fig. 1a). When sCD25 levels were compared in patients evaluated for suspected septic shock based on a discharge diagnosis of sepsis or not, patients with clinically confirmed sepsis demonstrated a wider range of sCD25 [10,300 pg/mL (IQR: 3,600–18,875)] than those with an alternate diagnosis [4,200 pg/mL (IQR: 4,000–5,200)] (Fig. 1b). Notably, a subset of patients with a clinical diagnosis of sepsis had markedly elevated sCD25 levels while all patients with an alternative diagnosis had normal levels. Soluble CD25 levels across clinical groups. a Log transformed sCD25 values in patients with clinical concern for septic shock, infection without sepsis, and HLH. b Log transformed sCD25 values in patients evaluated for septic shock based on whether they received a clinical diagnosis of sepsis during admission. Data are displayed as mean ± standard deviation. *** < 0.001 Clinical outcomes We examined associations between sCD25 levels and clinical outcomes in sepsis patients. Those with elevated sCD25 demonstrated elevated PELOD-2 scores across all time points (Day 0: 5 vs 2, Day 1: 4 vs 2, Day 2: 3 vs 1), increased PICU admission rates (100% vs 20%), longer hospitalizations (16 days vs 4 days), and greater need for vasoactive support (50% vs 20%), though these differences did not reach statistical significance (Table 2). Table 2Clinical outcomes in suspected septic shock patients with normal vs elevated sCD25OutcomesCD25 normalsCD25 elevatedP valuePELOD-2 scoreDay 02.05.00.95Day 12.04.0 > 0.99Day 21.03.0 > 0.99PICU admission, n (%)1 (20)2 (100)0.14Length of stay, days4160.38Vasopressor use, n (%)1 (20)1 (50) > 0.99aCategorical values are displayed as counts and percentagesContinuous variables are shown as medians Clinical outcomes in suspected septic shock patients with normal vs elevated sCD25 aCategorical values are displayed as counts and percentages Continuous variables are shown as medians Analysis by infection site revealed that bloodstream infections were associated with the highest sCD25 levels (20,600 pg/mL, n = 1), followed by pneumonia (10,300 pg/mL, IQR: 6,900–13,700) (Fig. 2). sCD25 levels (pg/mL) displayed by primary source of infection in patients undergoing evaluation for suspected septic shock (n = 7). Data are presented as mean ± standard deviation for infection sources with more than one patient, or as individual values for infection sources represented by a single patient Patients requiring PICU admission, positive pressure ventilation, or vasoactive medications had elevations in sCD25 compared with those who did not, but these findings were not significant (Fig. 3a–c). Likewise, patients with positive bacterial cultures had nonsignificant elevations in sCD25 levels compared with culture-negative patients [13,750 pg/mL (IQR: 6,900–26,000) vs 4200 pg/mL (IQR: 3,250–9,450)] (Fig. 3d). sCD25 levels by clinical outcomes. sCD25 levels in patients undergoing evaluation for septic shock based on the requirement for a PICU admission, b vasoactive medication use, c PPV use, and d a positive bacterial culture. Data are displayed as median with 95% CI Discussion In this study, we demonstrate that a subset of patients undergoing ED evaluation for septic shock have sCD25 levels comparable to HLH patients. Patients with sepsis and sCD25 elevation had nonsignificant trends toward worse clinical outcomes, including higher disease severity scores, increased need for intensive care, greater requirement for advanced supportive therapies (i.e., vasoactive medications and positive pressure ventilation), and longer lengths of stay. Among patients undergoing sepsis evaluation, sCD25 levels appeared higher in those with a clinical diagnosis of sepsis than in those with alternative diagnoses. However, these preliminary findings did not reach statistical significance, likely due to the small sample size, and thus require validation in larger studies. The broad range in sCD25 levels in the suspected septic shock group is representative of the well-recognized clinical and pathophysiologic heterogeneity in this population. Our findings are consistent with prior reports that have interrogated HLH diagnostic criteria in the setting of sepsis. One-third of children who died of sepsis fulfilled 5/6 HLH diagnostic criteria [5]. Hyperferritinemia was present in up to 66% of patients admitted to a PICU with sepsis and organ failure [15, 16]. With the known variability of immune responses in sepsis, elevated sCD25 levels would be expected in only a subset of patients, as observed in our study. The severity of illness in the few sepsis patients who had an elevation in sCD25 levels suggests that this marker may be a helpful prognostic indicator in sepsis. Correlations between sCD25 and mortality have been observed in adults and children with sepsis in ICU settings [9, 15, 17, 18] Our study suggests sCD25 may be useful earlier in sepsis evaluation and associated with adverse outcomes beyond mortality. In conclusion, sCD25 may be a potential prognostic marker in children undergoing ED evaluation for septic shock. Our findings warrant validation in larger, multicenter studies but suggest sCD25 may be a useful tool for risk stratification and clinical management of pediatric patients with sepsis. Prospective studies are warranted to further investigate the clinical utility of sCD25 in pediatric sepsis evaluation. Limitations There are several limitations to this study. First, the limited sample size restricts our power and precludes firm conclusions about the potential role of sCD25 as a prognostic biomarker in pediatric sepsis. Our study is single-centered and may not be generalizable to other pediatric EDs with different patient populations or clinical practices. The patients in the suspected septic shock cohort were relatively mild in illness severity (some failed to meet clinical criteria for sepsis after initial diagnostic workup) as compared to studies in ICU settings where patients are generally more critically ill. In fact, none of our patients in suspected septic shock met the Phoenix sepsis criteria, which may limit generalizability of these results to more severely ill patients. We did not longitudinally evaluation sCD25 levels as our study was limited to the emergency department setting. Current evidence suggests that sCD25 elevation occurs early during infection. However, kinetics vary depending on underlying infection, severity of inflammatory response, and host factors. Further studies are necessary to establish the optimal timing of sCD25 measurement in pediatric sepsis. Finally, we excluded patients younger than 12 months old and those with malignancy or immune deficiencies. While this may have decreased sCD25 heterogeneity in this study, it limits generalizability to these high-risk populations. Supplementary Information Additional file 1. Acknowledgements Not applicable. Abbreviations Hemophagocytic lymphophistiocytosis Soluble CD25 Emergency department Positive pressure ventilation Pediatric intensive care unit Pediatric logistic organ dysfunction-2 Enzyme linked immunosorbent assay Author contributions TBG: conceptualization, data curation, data analysis, writing- original draft. VK: conceptualization, methodology, supervision, data- generation and analysis, writing- review and editing. HFS: conceptualization, methodology, supervision, data analysis, writing- review and editing. Funding This work was funded by the Elevate Grant Fund, Section of Emergency Medicine, Children’s Hospital Colorado 23–03 (Garcia). Data availability The data generated during the current study are available from the corresponding author on reasonable request. Declarations Ethics approval and consent to participate Ethics review and approval was given by the Colorado Multiple Institutions Review Board (23–0948) at the University of Colorado Denver. The study was conducted in accordance with the Declaration of Helsinki and its later amendments. Under US law and IRB regulations, patients < 18 years old are considered minors and cannot participate in research without parental consent and legally cannot consent themselves. Parental informed consent was obtained for all children < 18 years of age. Although no children < 18 years provided legal consent, all patients ≥ 7 years of age were present for the consent process and provided voluntary written informed assent. Consent for publication Written informed consent was obtained from the parents or guardians of all patients, and when appropriate, assent was obtained from the children themselves. Competing interests The authors declare no competing interests. Footnotes Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Additional file 1. Data Availability Statement The data generated during the current study are available from the corresponding author on reasonable request.