F1000Research F1000Research 2046-1402 F1000 Research Limited London, UK 10.12688/f1000research.163800.1 Study Protocol Articles Assessing the effects of fluids and antibiotics in an acute murine model of sepsis: study protocol for the National Preclinical Sepsis Platform-01 (NPSP-01) Study

[version 1; peer review: 1 approved, 1 approved with reservations]

 Jahandideh Forough Conceptualization Methodology Project Administration Writing – Original Draft Preparation Writing – Review & Editing 1 Mendelson Asher A. Conceptualization Funding Acquisition Writing – Review & Editing 2 Liaw Patricia C. Writing – Review & Editing 3 Gill Sean E. Writing – Review & Editing 4 5 6 Bourque Stephane Writing – Review & Editing 7 Fox-Robichaud Alison E. Writing – Review & Editing https://orcid.org/0000-0001-9912-3606 3 Cepinskas Gediminas Writing – Review & Editing 6 8 9 Macala Kimberly F. Writing – Review & Editing 7 10 Sunohara- Neilson Janet Writing – Review & Editing 11 Engelberts Doreen Methodology Writing – Review & Editing 1 Sontag David Methodology Writing – Review & Editing 2 Dwivedi Dhruva J. Methodology Writing – Review & Editing 3 Yu Ian-Ling Methodology Writing – Review & Editing 12 Batnyam Onon Methodology Writing – Review & Editing 6 Mbuta Kashimbi Methodology Writing – Review & Editing 7 Fergusson Dean A. Conceptualization Resources Writing – Review & Editing 1 13 14 Taljaard Monica Writing – Review & Editing https://orcid.org/0000-0002-3978-8961 13 McDonald Braedon Conceptualization Funding Acquisition Writing – Review & Editing 12 Lalu Manoj M. Conceptualization Funding Acquisition Resources Writing – Original Draft Preparation Writing – Review & Editing https://orcid.org/0000-0002-0322-382X a 1 14 15 16 Blueprint Translational Research Group, Acute Care Research and Methods and Implementation Research Programs, Ottawa Hospital Research Institute, Ottawa, ON, Canada Department of Internal Medicine, Section of Critical Care, University of Manitoba, Winnipeg, Manitoba, Canada Department of Medicine, McMaster University, Hamilton, Ontario, Canada Department of Physiology and Pharmacology, Western University Schulich School of Medicine & Dentistry, London, Ontario, Canada Division of Respirology, Western University Schulich School of Medicine & Dentistry, London, Ontario, Canada Centre for Critical Illness Research, London Health Sciences Centre Research Institute, London, ON, Canada Department of Anesthesiology and Pain Medicine, University of Alberta, Edmonton, Alberta, Canada Department of Anatomy & Cell Biology, Western University Schulich School of Medicine & Dentistry, London, Ontario, Canada Department of Medical Biophysics, Western University Schulich School of Medicine & Dentistry, London, Ontario, Canada Department of Critical Care Medicine, Royal Alexandra Hospital, University of Alberta, Edmonton, AB, Canada Animal care services, Office of Research, University of Guelph, Guelph, ON, Canada Department of Critical Care Medicine, University of Calgary Cumming School of Medicine, Calgary, Alberta, Canada Methods and Implementation Research Program, Ottawa Hospital Research Institute, Ottawa, Ontario, Canada Faculty of Medicine, University of Ottawa, Ottawa, Ontario, Canada Department of Cellular and Molecular Medicine, University of Ottawa, Ottawa, Ontario, Canada Department of Anesthesiology and Pain Medicine, The Ottawa Hospital, University of Ottawa, Ottawa, Ontario, Canada mlalu@toh.ca

No competing interests were disclosed.

 1 5 2025 2025 14 472 17 4 2025 Copyright: © 2025 Jahandideh F et al. 2025 This is an open access article distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Background

Sepsis remains a leading cause of mortality in critical care. Despite extensive preclinical research on sepsis pathophysiology, the development of effective therapies has been largely unsuccessful. Key obstacles include limited construct validity of animal models, insufficient methodological rigor and the lack of collaborative frameworks akin to clinical trials. These issues plague not only sepsis research, but preclinical research in general. The National Preclinical Sepsis Platform (NPSP), an interdisciplinary network under Sepsis Canada, addresses these challenges in sepsis research through multilaboratory, randomized, controlled preclinical studies. NPSP-01 will establish baseline conditions for future investigations using an acute fecal-induced peritonitis model of sepsis.

 Methods

This randomized, controlled study will evaluate the effect of standard sepsis therapy in a mouse model of sepsis across six centres. Interlaboratory variability and the interaction of biological sex on outcomes will also be examined. C57BL/6 mice of both sexes will be randomized into sham (healthy control) + treatment, sepsis, or sepsis + treatment groups. Sepsis will be induced via intraperitoneal injection of fecal slurry, while sham mice will receive vehicle control. Antibiotics and fluids will be administered to treatment groups at 4 hours post-induction, and mice with be euthanized at 8 hours post-induction. The primary outcome is plasma interleukin-6 levels. Secondary outcomes include biological (blood gas and chemistry, white blood cell count, bacterial load), clinical (body weight, core temperature, sepsis score, mortality as measured by surrogate humane endpoints), and feasibility measures.

 Conclusions

NPSP-01 will be the first multilaboratory study of sepsis and represents a shift in preclinical critical illness research, mirroring the rigor of clinical multicenter trials. By addressing procedural standardization, interlaboratory variability, and sex-based differences, this study aims to enhance the reliability and translational relevance of preclinical findings. The outcomes of NPSP-01 will establish foundational data for future investigations and provide a roadmap for rigorous collaborative preclinical studies to accelerate the evaluation of novel sepsis therapies.

 Registration

PreclinicalTrials.eu PCTE0000552

Protocol Version 1.0, October 21, 2024

 Acute Sepsis Antibiotic therapy Feasibility Fluid resuscitation Inter-laboratory variability Multicenter study National Preclinical Sepsis Platform Sepsis Canada New Frontiers in Research Fund The National Preclinical Sepsis Platform is funded by Sepsis Canada (a network funded by the Canadian Institutes of Health Research) as well as an Exploration Grant from the New Frontiers in Research Fund. M.M.L. is supported by the Ottawa Hospital Anesthesia Alternate Funds Association, the Canadian Anesthesiologists’ Society Career Investigator Award and holds a University of Ottawa Junior Research Chair in Innovative Translational Research. AAM is supported by the Manitoba Medical Services Foundation Dr. F. W. DuVal and John Henson Clinical Research Professorship. PCL holds the Jack Hirsh and Clive Kearon Endowed Chair in Thrombosis Research. AFR is the Scientific Director of Sepsis Canada and holds the Hamilton Health Sciences Chair in Sepsis Research. The funding agencies had no role in the design of the study, analysis, or writing of the manuscript. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Introduction

Preclinical research has significantly advanced our understanding of sepsis pathophysiology, shedding light on key mediators such as inflammatory cytokine production (e.g., interleukin [IL]6, tumor necrosis factor [TNF]α) and mechanisms of organ dysfunction (e.g., mitochondrial failure, endothelial permeability, immunothrombosis). Despite these insights, septic shock mortality rates persist at 25 – 30%, 1 and decades of research have failed to translate findings from preclinical models into effective therapies. This translational challenge is not unique to sepsis: across biomedical fields, fewer than 5% of high-impact preclinical findings are translated from “bench-to-bedside”. 2, 3

Several modifiable factors could improve translational success in sepsis research. First, construct/translational validity 1 of animal models can be augmented to mimic some of the complexity and heterogeneity of human sepsis. For instance, researchers could incorporate standard clinical interventions, such as fluid resuscitation and antibiotics, use true bacterial or viral pathogens instead of surrogates like lipopolysaccharide, and study both male and female animals within a wider range of age and genetic diversity. 2– 4 In addition, methodological rigour could be improved to reduce risk of bias (e.g., through best practices like randomization and blinding) and ensure adequate statistical power (e.g., through sample size calculations). Other factors such as variability in laboratory conditions and protocols, lack of transparency (e.g., insufficient methods details provided), and different baseline conditions, contribute to variability in findings between laboratories. 5 Addressing these challenges requires a coordinated and collaborative approach.

A promising strategy to increase the translational potential of preclinical findings is the use of multilaboratory collaborative studies. While multicenter studies are a gold standard approach in clinical research, they remain underutilized in preclinical settings. 6 To address this gap in sepsis research, we established the National Preclinical Sepsis Platform (NPSP), an interdisciplinary network of scientists, veterinarians, and interest holders under Sepsis Canada. 7 Our objective is to conduct randomized, controlled, multilaboratory studies in sepsis, employing harmonized protocols across several laboratories. This will establish a network capable of conducting rigorous studies that can accelerate testing of promising therapies for sepsis.

For our initial study, NPSP-01, we are implementing procedural standardization and rigorous practices to establish baseline conditions for future investigations. Specifically, we will compare outcomes in six independent laboratories between untreated septic mice and those treated with clinical standard of care antibiotics and fluid resuscitation. We hypothesize that administering antibiotics and fluids will reduce systemic inflammation as assessed by plasma IL6, a clinically relevant pro-inflammatory cytokine and biomarker for sepsis. 8 We will also assess feasibility, interlaboratory variability, and the interaction of biological sex on study outcomes. To our knowledge, this will be the first multilaboratory preclinical study of critical illness in the world.

Objectives

To use a multilaboratory approach to compare outcomes between untreated septic mice and septic mice treated with usual therapy (i.e., antibiotic and fluids).

To evaluate inter-laboratory variability in estimated treatment effects.

To evaluate biological sex as a potential effect modifier on study outcomes. 7

To assess the feasibility of conducting a multicenter preclinical sepsis study across six laboratories in Canada.

Methods and design

Animal experiments will be conducted in accordance with the Canadian Council on Animal Care Guide to Care and Use of Experimental Animals. We have received approval from the institutional animal care and use committees at the University of Ottawa (Protocol OHRIe3562), McMaster University (Protocol #21-07-19), University of Western Ontario (Protocol #2022-023), University of Manitoba (Protocol #21-019), University of Alberta (Protocol #3381), and University of Calgary (Protocol AC23-0116). Elements of the PREPARE guidelines 2 and the Minimum Quality Threshold in Pre-Clinical Sepsis Studies (MQTiPSS) guidelines are considered in the design of this protocol. 9 The SPIRIT guidelines are followed, where applicable, in the reporting of this protocol, to optimally match the reporting practices use for human clinical trial protocols. 10 The SPIRIT checklist can be found on our Open Science Framework National Preclinical Sepsis Platform-01 (NPSP-01) project homepage, DOI 10.17605/OSF.IO/R5G7Y. ARRIVE guidelines will be used when reporting the final results of our study. Our study protocol has been registered in preclinicaltrials.eu (PCTE0000552). All protocol amendments will be reflected in updates to the registered protocol.

Study design and setting

This is a randomized controlled multilaboratory preclinical study with three arms: sham + treatment (i.e., healthy animals), sepsis, and sepsis + treatment. The allocation ratio will be 1:2:2. The inclusion of a sham group will allow us to distinguish the effects of sepsis and its treatment from baseline physiological responses, providing essential control data for interpretation of study outcomes. Research staff will be blinded for disease induction, treatment administration, and outcome evaluation. All experiments will be carried out in animal facilities of six academic or research institutions across Canada. Table 1 lists the participating centers in the study.

Participating centers and roles.
Center Role
Ottawa Hospital Research Institute Coordinating center; Conducting experiments
McMaster University Conducting experiments
University of Western Ontario Conducting experiments
University of Manitoba Conducting experiments
University of Alberta Conducting experiments
University of Calgary Conducting experiments
Eligibility criteria

Inclusion criteria: C57BL/6 mice at 10-12 weeks of age will be included in the study. Mice will be specifically bred for this study by the Charles River facility in Saint-Constant, Quebec, Canada where one male is paired with two females for breeding. Pups will remain with their parents until shipment to the designated centers.

Exclusion criteria: Prior to randomization, mice that do not pass routine health assessments by local veterinary staff will be excluded. Mice <16 g or >30 g will be excluded. No exclusions will occur after mice have been allocated to an experimental group.

 Experimental model: fecal-induced peritonitis (FIP)

The FIP model was selected as the first model to be tested by the NPSP. The FIP model only requires an intraperitoneal injection of fecal slurry for sepsis induction, which is less technically demanding than surgical models (e.g., cecal ligation and puncture). This results in less interoperator variability and can facilitate standardization of baseline conditions across multiple laboratories. To refine and optimize the model for this study, 24 pilot experiments were conducted, with some of the early results published by our group. 11 Although initial pilot studies employed a 72-hour experimental timepoint, an 8-hour endpoint model was believed to be more feasible for implementation by all participating laboratories given available personnel, intensity of monitoring, and funding, while still inducing a severity of illness reflective of acute sepsis. Additionally, all mice in the FIP pilot studies that received 0.75 mg/g of fecal slurry survived beyond 8-hours after sepsis induction, suggesting that the 8-hour endpoint model may ensure optimal survival and tissue collection in the mice.

Preparation of fecal slurry: To minimize batch to batch variability, the rat fecal slurry required for disease induction was prepared in bulk (100mg/mL in dextrose-glycerol) by one center (McMaster University), aliquoted, and stored at -80°C until use. The microbial composition of this batch of slurry was previously characterized. 11 Fecal slurry aliquots will be shipped on dry ice to all participating centers, where their frozen state will be verified upon receipt prior to immediate storage at -80°C. For sham animals, vehicle aliquots containing 5% dextrose (Fisher Chemical, Cat#: D16-3) in 10% glycerol (Fisher Bioreagents, Cat #: BP229-1) alone will be autoclaved and stored at -80°C in each center for use in the experiments.

Mouse acquisition, housing and husbandry: All mice for the NPSP-01 will be shipped to individual institutions directly from Charles-River at six weeks of age in 3-4 batches. Upon receipt, mice will be housed in groups of 2-3 per HEPA-filtered cages (segregated by sex). Housing will be equipped with corncob bedding (1/4-inch or 1/8-inch, irradiated or autoclaved), nesting material (Nestlet and crinkle paper), and one structure in the form of cardboard hut or dome. Each cage will be provided with chow (Teklad Irradiated Global 18% Protein Rodent Diet 2919) and water (autoclaved reverse osmosis). To minimize in-fighting/barbering, male mice in each cage will be from the same litter. Mice will be kept in the animal facility for a 4-6 week acclimation period prior to the start of the experiments.

 Interventions

All mice will receive a subcutaneous injection of buprenorphine (0.05 mg/kg body weight, Ceva, DIN: 02342510) for pain management at 4 hours post-induction, to meet current MQTiPSS recommendations 2 for animal welfare. Additionally, at 4 hours post-induction, mice in the treatment groups will receive imipenem-cilastatin (25 mg/kg body weight, Sandoz, DIN: 02358344) and Ringer's lactate (15 mL/kg body weight, Baxter, DIN: 00061085) in the same injection subcutaneously.

 Experimental groups

A total of 192 mice will be randomly allocated to one of the following three groups:

Sham + treatment (dextrose-glycerol injection at t = 0, treated with imipenem-cilastatin and Ringer’s Lactate at t = 4), both sexes; n = 36 mice.

Sepsis alone (fecal slurry injection at t = 0, no treatment), both sexes; n = 78 mice.

Sepsis + treatment (fecal slurry injection at t = 0, treated with imipenem-cilastatin and Ringer’s Lactate at t = 4), both sexes; n = 78 mice.

Timeline

The schedule of enrollment, interventions, and assessments for NPSP-01, designed in accordance with SPIRIT guidelines, is outlined in Figure 1. Additional details on the study flow and the timeline of different procedures are illustrated in Table 2.

Experimental flow diagram for NPSP-C01 depicting pre-study, day of study, and post-study protocols for FIP sepsis induction in mice. Schedule of enrolment, induction, interventions, assessments, and biobanking for NPSP-01.
Study period
Enrolment Allocation Post-Allocation Close-out
Timepoint -t 4 weeks -t 1 day t 0 t 4 t 6 t 8 >t 8 t 4 months
Enrollment
Charles River shipment of mice to individual centers X
Receipt by animal facilities X
Housing and husbandry requirements X
Group caging (2-3 mice per cage) X
Central randomization X
Allocation X
Intention-to-treat form submission X
Induction
Induction syringe administration X
Interventions
Buprenorphine X
Imipenem-cilastatin and Ringer’s lactate X
Assessments
Baseline variables (wellness check, body weight, and core temperature) X
Outcome variables (wellness check ± body weight and core temperature) X X X
Other data variables (blood chemistry analysis, white blood cell counts, blood and peritoneal lavage fluid bacterial counts) X Plasma multiplex analysis
Biobanking
Tissue and sample harvest X Processing & storage X
Data transfer to online forms X
Randomization and blinding

Mice will be randomized in a 1:2:2 ratio to sham + treatment, sepsis alone, and sepsis + treatment arms. The allocation sequence will be computer-generated by a statistician at the central coordinating center using a stratified permuted block design with randomly varying lengths. Allocations will be stratified by laboratory and biological sex of animals to ensure balanced distribution across groups. Mice will be assigned a study identification number on arrival to the facility and the project coordinator will receive an email with the study ID number and treatment allocation. The randomization scheme will ensure allocation concealment from study personnel who are conducting the experiments in each laboratory.

Each center will designate one unblinded individual to be solely responsible for preparing induction and treatment syringes. This person will not participate in other study procedures. The project coordinator will send the allocation list to the designated unblinded individual at each center one day prior to the experiment. On the day of the experiment, unblinded individuals at each center will prepare the syringes and label with mouse numbers. The allocation list will be securely discarded post-syringe preparation. A separate, blinded individual will administer the syringes by matching to mouse numbers and thus remain blinded to the experimental groups.

Assessments: Except for the individual responsible for preparing the syringes at each center, all other personnel will remain blinded to group allocations. Consequently, all experimental protocols, animal monitoring, tissue collection, and point-of-care analyses will be conducted in a blinded manner. Data analysis will be performed without knowledge of center or group allocations, with unblinding occurring only after the final analysis is completed.

 Experimental procedures

Mice aged 10–12 weeks will be injected intraperitoneally with either the fecal slurry (0.75 mg/g body weight) or an equivalent volume of vehicle (dextrose-glycerol) under isoflurane anesthesia, targeting the right or left lower abdominal quadrant using a syringe with a 26-gauge needle. After injection, the abdomen will be massaged for 10 seconds, and the mice will be returned to their cages to recover. Heating blankets set at 37-38°C will be placed under half of each cage to support the mice in regulating their body temperature.

Overall health will be assessed with the murine sepsis score 12 modified for use in NPSP-01. Briefly, mice will be evaluated for locomotor activity, orbital tightening (indicative of pain or distress from clinical condition), fur appearance (reflective of grooming and hydration), and posture before handling for body weight or temperature measurements. This will be completed at baseline, 4, 6, and 8 hours post-induction. Each mouse will be evaluated at each timepoint by at least two blinded personnel and the calculated mean sepsis score will be used at each timepoint. Body weight and core temperature (as assessed by a rodent rectal probe thermometer) will be measured at baseline, 4 and 8 hours post-induction.

At 4 hours, mice will receive a subcutaneous injection of either buprenorphine alone or a combination of buprenorphine, imipenem-cilastatin, and Ringer’s lactate. At the study endpoint (8 hours post-induction) mice will be anesthetized with isoflurane and carotid blood will be collected into EDTA-coated tubes. A portion of the blood will be used for analysis of blood gas and chemistry, while the remaining blood will be used to measure white blood cell counts and bacterial load, and for plasma isolation. Animals will be euthanized under general anesthesia and then exsanguinated. Comprehensive biobanking will be performed with up to 20 samples collected per mouse (2-3 samples per tissue). These include plasma, peritoneal lavage fluid, tissues (brain, heart, lung, kidney, spleen, and muscle), and cecal content. Plasma will be prepared by centrifugation of blood at 5000 x g for 10 minutes and will be stored in aliquots of 20 to 100 μL at -80°C. Collected samples will either be stored at -80°C or fixed in formalin at 4°C for a minimum of 24 hours for future histology assessments. Formalin-fixed tissues will subsequently be washed in 1x PBS (Fisher BioReagents, Cat#: BP24384) and stored in either 0.1% sodium azide (VWR Avantor, Cat #: RC7144.8-16) with 30% sucrose (Sigma Aldrich, Cat#: S7903-1KG) (brain samples) or 70% ethanol (Greenfield Global, Cat#: P016EAAN) (all other tissues) for long-term storage.

 Protocol harmonization

To assess reproducibility and inter-laboratory variability in the NPSP-01 study, we will implement harmonized experimental conditions, including consistent mouse sourcing (as described above), animal care (e.g., cage conditions, husbandry, food and water), consensus based standard operating procedures, centralized biobanking protocols, and standardized supplies and devices across all participating labs. Multiple sources of heterogeneity and variation remain, to avoid the ‘standardization fallacy’, where excessive standardization can obscure real world variability and confound external validity. 13 The use of male and female mice, differences in facility environments as well as the number and experience of personnel in each lab will be uncontrolled variables. 14

Training materials: Each center will be responsible for ensuring that its research personnel are adequately trained in experimental techniques including injection protocols, animal handling, monitoring, and tissue collection. To facilitate this training, we have provided training videos and resources. Detailed standard operating procedures were created and shared with all members via a SharePoint (Microsoft, USA) repository to ensure consistent implementation across centers.

Supplies and devices: To reduce confounding due to equipment variability, all necessary supplies (e.g. tubes, vials, EDTA-coated tubes for blood collection, imipenem-cilastatin antibiotic (Sandoz Canada)) as well as devices including thermometers (INTELLIBIO, A-2205-00389), hemocytometers (Hausser Scientific Hemacytometer, #0267151B), Epoc analyzers (Siemens, Model PD470SH-B), and cartridges have been provided by the coordinating center to all participating centers.

 Outcomes

Primary outcome

Plasma IL6 will be assessed at the end of experiment (8 h) or at surrogate humane-endpoint of death, measured using a multiplex analysis (Eve Technologies, Calgary, Canada). IL6 is a well-established biomarker of systemic inflammation in sepsis and plays a critical role in the pathophysiology of the disease. Elevated plasma IL6 levels are strongly associated with disease severity, organ dysfunction, and mortality in both preclinical models and human sepsis patients; IL6 levels also have demonstrated sensitivity to therapeutic interventions. 8, 15– 17 Its consistent elevation in septic conditions and rapid decline with effective treatment make it a suitable primary outcome for evaluating inflammatory responses, as well as a useful biomarker to help guide translational efforts.

Secondary outcomes

Clinical outcomes: Murine sepsis score, body weight, and core temperature will be measured as described above. Mortality, evaluated by surrogate humane endpoints will also be assessed. Mice meeting any of the predefined humane endpoint criteria (body weight drop exceeding 20% from baseline, a core temperature drop exceeding 20% from baseline, inability to right when placed on their side, labored breathing, an average modified murine sepsis score greater than 2.5 for any two criteria, or a total average murine sepsis score greater than 9.4) will be humanely euthanized via isoflurane anesthesia with exsanguination, and their tissues will be collected for analysis. Those mice euthanized before study end will be included in an intention-to-treat analysis, using their most recent data for the analysis.

Biomarker outcomes: Bacterial load will be quantified in blood and peritoneal lavage fluid by plating samples on blood agar and assessing colony-forming units after incubation. White blood cell count will be measured manually by a hemocytometer as an indicator of immune response. Arterial blood gas and chemistry will be assessed with a point of care device (Epoc Blood Analysis System, Siemens Healthineers). Plasma inflammatory markers (granulocyte-macrophage colony-stimulating factor [GM-CSF], interferon [IFN] γ, IL1β, IL2, IL4, IL10, IL12p70, monocyte chemotactic protein [MCP]1, TNFα) and cardiovascular markers (pro-matrix metalloproteinase [MMP]9, plasminogen activator inhibitor [PAI]1, platelet endothelial cell adhesion molecule [PECAM]1, soluble platelet [sP]-Selectin, soluble endothelial [sE]-Selectin, soluble intercellular adhesion molecule [sICAM]1, thrombomodulin) will be measured by the multiplex analysis.

Feasibility outcomes: Several measures of feasibility, including study completion, technical success, biobanking completion, and protocol adherence will be assessed as follows:

Study completion: Defined as adherence to the allocation and successful plasma collection for IL6 analysis in ≥80% of mice.

Technical success: Defined by the occurrence of any of the following six technical issues in ≤20% of mice: Subcutaneous injection at T0, incomplete injection at T0, incomplete injection at T4, accidental organ punctures, arterial blood collection failures, and compromised blood quality (e.g., clotting). If multiple technical issues are observed in a single mouse, they will be recorded as a single event for the purposes of outcome assessment.

Biobanking completion: Defined as biobanking of ≥ 90% tissues in ≥ 80% of mice reaching the study endpoint (8 hours).

Protocol Adherence (Yes/No): Defined as meeting at least 90% of 15 specified criteria for each mouse in ≥80% of mice. Criteria include husbandry practices (such as the use of appropriate bedding, nestlets, crinkle paper, cardboard huts/domes, food, water, and ventilated cages), group caging, use of the correct fecal slurry batch, strict adherence to SOPs, maintenance of blinding procedures, appropriate time staggering, use of external heating, assessment of mice wellness by multiple individuals, and proper sample storage.

 Sample size

For this feasibility study, our sample size calculation is based on detecting differences in plasma IL-6 between septic and septic + treatment groups at each of the six participating centers (two-arm comparison with equal allocation) at the study endpoint. Sample sizes of 12 mice in each arm (total 24 mice per center) will achieve 80% power to detect a mean difference of 6,000 in fluorescence intensity using a two-sided t-test with equal variance at the 5% level of significance. We assumed a standard deviation of 5,000 based on pilot multiplex analysis data. An additional two animals per group will be included at each center to account for potential attrition. Thus, our total sample size allocated to the two treatment arms across 6 centers is 156 mice. An additional 36 mice will be allocated to the sham+treatment arm for a grand total of 192 mice across 6 centers.

 Statistical analysis plan

Baseline description of the groups

Descriptive analyses will be used to compare baseline characteristics of mice allocated to the different study arms. Measures of central tendency (e.g., mean or median) and dispersion (e.g., standard deviation or inter-quartile ranges) will be calculated for all continuous variables (e.g., body weight, core temperature, murine sepsis score) and counts and proportions for categorical variables (e.g., sex). As differences in treatment effects across centers is of specific interest, we will also explore differences in baseline characteristics of mice across centers using descriptive statistics.

Analysis of differences between non-treated and treated septic mice

Descriptive statistics will be used to describe outcomes in each arm (mean and standard deviation or median and inter-quartile range as appropriate) or frequency and proportion. Histograms and normal probability plots will be used to investigate skewness in continuous outcome variables and normalizing transformations will be applied as necessary. Additional descriptive analyses will characterize outcomes in each arm by center.

Primary analysis

Analysis of our primary outcome, plasma IL6, will be conducted using a linear regression analysis (after normalizing transformation if necessary). To obtain correct inferences, the model will include the stratification factors: sex and center as fixed covariates. Adjusted least square mean differences with 95% confidence intervals will be obtained from the model and used to compare mean plasma IL-6 levels at 8 h (or when humane endpoint is reached) between Sepsis and Sepsis + treatment groups (primary comparison).

Inter-laboratory variability analysis

To formally assess differences in treatment effects on mean plasma IL6 levels between laboratories, we will extend the primary analysis described above by including the interaction between center and treatment as a fixed effect into the statistical model. The statistical significance of the treatment by center interaction will be used to test the hypothesis that treatment effects differ by center. Least square mean differences from the model with 95% confidence intervals will be used to estimate the treatment effects at each center. We will use forest plots to visualize mean differences and 95% confidence intervals across the six laboratories and contrast them with the overall pooled finding to visually assess heterogeneity. Additional analyses will allow for extra variability in the pooled treatment effect estimates due to differences between centers by modeling center as a random effect and including center and center by treatment interaction as random terms in the model, provided the center by treatment variance component is positive. The random effects analyses will be used to quantify variability in outcomes between centers using intracluster correlation coefficients.

Secondary analyses

For all secondary outcomes, the effect of the treatment will be evaluated as for our primary analysis for continuous variables (i.e., mean differences and 95% confidence intervals). For dichotomous secondary variables (e.g., mortality), we will conduct logistic regression analyses adjusted for the stratification factors and obtain odds ratios and risk differences with 95% confidence intervals.

Feasibility analysis

Descriptive analyses will be carried out using classification in categorical variables and using means and SD in numerical variables.

Additional subgroup analyses

Subgroup analyses by male and female animals will be carried out for primary and secondary outcomes. These analyses will be carried out by adding interaction terms between sex and treatment to the statistical models and using least square mean differences from the model to estimate sex-specific treatment effects.

Population analysis and missing data

Primary analyses: Primary analyses will follow an intention-to-treat (ITT) approach, including all randomized animals to ensure a comprehensive evaluation of treatment effects. The presence of missing data will be limited through rigorous training but missingness can arise (e.g., in plasma IL-6 measurements due to insufficient blood collection). We will explore the potential impact of missing data on inferences by comparing characteristics of animals with missing outcome data and those with complete data. Multivariable logistic regression analyses will be used to explore factors associated with missingness under a Missing At Random assumption, i.e., that missingness depends only on observed characteristics. If any characteristics associated with missingness are identified, additional sensitivity analyses will be carried out by adjusting for these characteristics in the linear and logistic regression analyses.

We will also conduct a per-protocol analysis for animals that received injections at T0 intraperitoneally and all interventions (e.g. fluids and antibioitics) as detailed.

Secondary analyses: Several secondary analyses are being planned for NPSP-01, including assessment of laboratory personnel characteristics on the study outcomes, histological analyses of biobanked tissue, microbiome analysis of cecal contents, and assessment of the modified murine sepsis score. Prior to initiating these studies, protocols for each will be developed and posted.

 Study management

Coordinating center

The trial will be centrally coordinated by a project coordinator (FJ) based at the Ottawa Hospital Research Institute, under the supervision of the NPSP-01’s primary investigator (MML) and a senior scientist from the methods center (DAF). The study structure also includes principal investigators as well as personnel at each participating center. The project coordinator will maintain regular contact with participating laboratories prior to study initiation to ensure that all study infrastructure is in place and that personnel at each center have received adequate training. After each experimental day, the project coordinator will perform online monitoring of collected data to ensure validity and proper study execution, including intervention implementation, outcome measurement, biobanking, and sample storage. Statistical consulting, randomization, and analyses will also be conducted at the Methods Center of the Ottawa Hospital Research Institute.

 Data management

Study data will be collected and managed through a user-friendly interface built with React, with data stored in a PostgreSQL database hosted on Heroku. The backend, built with express, ensures reliable and secure data processing. Backup of data will be scheduled daily with Heroku PGBackups.

All data will be recorded on preformatted ‘mouse report forms’, which include designated fields for dates and IDs of responsible personnel. Each center’s principal investigator will review the forms for accuracy and completeness. Data forms will then be scanned and uploaded to center-specific folders in the study's SharePoint within 48 hours of collection. Data will also be uploaded to a web-based platform with automated completeness checks. A designated individual will transfer data to the online database, with a second individual verifying accuracy to minimize input errors. Once submitted, forms will lock to prevent any further modifications.

 Data monitoring

Quality assessments will be conducted by the coordinating center through remote monitoring. The project coordinator will verify the completeness of uploaded forms in SharePoint and randomly audit a selection of forms for consistency between the manual and electronic submissions for each center on the experimental dates. In the event of any inconsistencies, the project coordinator will inform the principal investigator of the specific center about discrepancies identified and work collaboratively to resolve issues.

 Access to data

During the course of the study, access to the data will be granted only for input and verification.

Upon completion of NPSP-01, a complete cleaned dataset will be prepared, together with its related dictionaries. This dataset will be made available on the publication of the scientific papers using an online data repository.

 Discussion

The NPSP-01 study will unite diverse expertise across various laboratories to address longstanding barriers in preclinical sepsis research and advance the field. Our evaluation of interlaboratory and intralaboratory variability will provide critical insights into variability thresholds, which are essential for designing future multilaboratory sepsis studies. By incorporating biological sex as a variable into experimental design, the NPSP-01 will also directly address the need for preclinical models that mirror heterogeneity of human sepsis. In addition, assessing how standard clinical interventions for sepsis (antibiotics and fluid resuscitation) affect this model will help establish a baseline to compare novel interventions in future studies. Furthermore, the rigorous implementation of harmonized protocols, randomization, blinding, and transparent reporting sets a high standard that supports the generation of high-quality and reliable data.

Beyond its immediate scope, NPSP-01 also establishes a scalable and versatile platform that can be readily adapted to test other models of sepsis. This platform could also be used to address pandemic response efforts and other emergent public health challenges requiring rapid and rigorous evaluation of therapies. By mitigating the risks associated with non-reproducible research, this platform has the potential to maximize the impact of scientific investments and streamline the translation of promising treatments to early-phase clinical trials.

Disclosures Ethics approval and consent to participate

The study is approved by the University of Ottawa, Approval # OHRIe3562.

 Consent for publication

Not applicable.

 Data availability statement

No data are associated with this article.

Reporting guidelines

SPIRIT Guideline checklists can be publicly accessed at National Preclinical Sepsis Platform-01 (NPSP-01) on Open Science Framework, DOI 10.17605/OSF.IO/R5G7Y. 18 Given the lack of reporting guidelines for animal protocols, we felt this was the closest relevant guideline.

 References Drude NI Gamboa LM Danziger M : Science Forum: Improving preclinical studies through replications. elife. 10. 33432925 10.7554/eLife.62101 PMC7817176 Osuchowski MF Ayala A Bahrami S : Minimum Quality Threshold in Pre-Clinical Sepsis Studies (MQTiPSS): An International Expert Consensus Initiative for Improvement of Animal Modeling in Sepsis. Shock Augusta Ga. 2018;50(4):377–380. 30106875 10.1097/SHK.0000000000001212 PMC6133201 Nandi M Jackson SK Macrae D : Rethinking Animal Models of Sepsis - Working towards Improved Clinical Translation Whilst Integrating the 3Rs. Clin. Sci. Lond. Engl. 1979. 2020;134(13):1715–1734. 32648582 10.1042/CS20200679 PMC7352061 Cavaillon J-M Singer M Skirecki T : Sepsis Therapies: Learning from 30 Years of Failure of Translational Research to Propose New Leads. EMBO Mol. Med. 2020;12(4):e10128. 32176432 10.15252/emmm.201810128 PMC7136965 França TF Monserrat JM : Reproducibility Crisis in Science or Unrealistic Expectations?. EMBO Rep. 2018;19(6):e46008. 29695587 10.15252/embr.201846008 PMC5989743 Hunniford VT Grudniewicz A Fergusson DA : A Systematic Assessment of Preclinical Multilaboratory Studies and a Comparison to Single Laboratory Studies. elife. 2023;12:e76300. 36892457 10.7554/eLife.76300 PMC10168693 Mendelson AA Lansdell C Fox-Robichaud AE : Sepsis Canada’s National Preclinical Sepsis Platform. National Preclinical Sepsis Platform: Developing a Framework for Accelerating Innovation in Canadian Sepsis Research. Intensive Care Med. Exp. 2021;9(1):14. 33738642 10.1186/s40635-020-00366-4 PMC7973346 Hamilton FW Thomas M Arnold D : Therapeutic Potential of IL6R Blockade for the Treatment of Sepsis and Sepsis-Related Death: A Mendelian Randomisation Study. PLoS Med. 2023;20(1):e1004174. 36716318 10.1371/journal.pmed.1004174 PMC9925069 Smith AJ Clutton RE Lilley E : PREPARE: guidelines for planning animal research and testing. 2018;52:135–141. (accessed 2024-11-20). 28771074 10.1177/0023677217724823 PMC5862319 SPIRIT 2013 explanation and elaboration: guidance for protocols of clinical trials. The BMJ. (accessed 2024-12-03). Reference Source Sharma N Chwastek D Dwivedi DJ : Development and Characterization of a Fecal-Induced Peritonitis Model of Murine Sepsis: Results from a Multi-Laboratory Study and Iterative Modification of Experimental Conditions. Intensive Care Med. Exp. 2023;11(1):45. 37460911 10.1186/s40635-023-00533-3 PMC10352196 Shrum B Anantha RV Xu SX : A Robust Scoring System to Evaluate Sepsis Severity in an Animal Model. BMC. Res. Notes. 2014;7(1):233. 24725742 10.1186/1756-0500-7-233 PMC4022086 Voelkl B Würbel H Krzywinski M : The Standardization Fallacy. Nat. Methods. 2021;18(1):5–7. 10.1038/s41592-020-01036-9 Kortzfleisch VT von Ambrée O Karp NA : Do Multiple Experimenters Improve the Reproducibility of Animal Studies? PLoS Biol. 2022;20(5):e3001564. 35511779 10.1371/journal.pbio.3001564 PMC9070896 Bozza FA Salluh JI Japiassu AM : Cytokine Profiles as Markers of Disease Severity in Sepsis: A Multiplex Analysis. Crit. Care. 2007;11(2):R49. 17448250 10.1186/cc5783 PMC2206478 Song J Park DW Moon S : Diagnostic and Prognostic Value of Interleukin-6, Pentraxin 3, and Procalcitonin Levels among Sepsis and Septic Shock Patients: A Prospective Controlled Study According to the Sepsis-3 Definitions. BMC Infect. Dis. 2019;19(1):968. 31718563 10.1186/s12879-019-4618-7 PMC6852730 Wang J Yang X Li Y : Specific Cytokines in the Inflammatory Cytokine Storm of Patients with COVID-19-Associated Acute Respiratory Distress Syndrome and Extrapulmonary Multiple-Organ Dysfunction. Virol. J. 2021;18(1):117. 34088317 10.1186/s12985-021-01588-y PMC8177255 Lalu MM : National Preclinical Sepsis Platform-01 (NPSP-01). 2025, April 10. 10.17605/OSF.IO/R5G7Y 10.5256/f1000research.180201.r437806 Reviewer response for version 1 Mansart Arnaud 1 Referee https://orcid.org/0000-0003-1442-8924 Université Paris-Saclay, Montigny le bretonneux, France

Competing interests: No competing interests were disclosed.

 31 12 2025 Copyright: © 2025 Mansart A 2025 This is an open access peer review report distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. recommendation approve

Overall synthesis

This study protocol describes a well-conceived and ambitious multicenter preclinical study that directly addresses several long-standing limitations in translational sepsis research. By adopting clinical-trial standards, the NPSP-01 study represents a meaningful methodological advance. Beyond its biological objectives, the study’s main value lies in establishing a scalable and reproducible framework for future preclinical trials.

Major strengths

A key strength of the protocol is its six-center randomized controlled design, which remains rare in animal research. This approach directly tackles issues of reproducibility and inter-laboratory variability, two major contributors to translational failure in sepsis. Modeling the study after multicenter clinical trials is therefore a notable and commendable advance.

The methodological rigor is high. Centralized randomization stratified by center and sex, blinding at multiple stages, and clearly predefined primary and secondary outcomes all contribute to strong internal validity.

The inclusion of clinically relevant standard-of-care interventions, particularly antibiotics and fluid resuscitation, substantially enhances construct and translational validity. Many experimental sepsis models omit these elements, limiting clinical relevance. Establishing a treated baseline is especially valuable for future evaluation of novel therapeutic strategies.

The explicit inclusion of both male and female mice aligns well with current best practices and strengthens the biological relevance of the findings.

Another major strength is the extensive feasibility and process evaluation. Metrics such as technical success rates, protocol adherence, and biobanking completion will generate highly informative data for the design of future multicenter preclinical studies, not only in sepsis but also in other disease areas.

Finally, the authors strike a thoughtful balance between harmonization and real-world variability, avoiding excessive standardization. This reflects a sophisticated understanding of external validity and enhances the generalizability of the findings.

Areas for improvement and clarification

The relatively short experimental time frame, with an 8-hour endpoint, limits insight into the dynamic kinetics of sepsis and the progression of organ dysfunction. While pragmatic, this constraint should be clearly acknowledged when interpreting results and extrapolating to human sepsis.

Important dimensions such as age-related dysfunction are not addressed. Although sex diversity is intentionally introduced, age (a major determinant of sepsis susceptibility, immune response, and outcome) is not considered.

The protocol would also benefit from a clearer discussion of downstream decision rules. It is not specified what degree of inter-laboratory heterogeneity would be considered acceptable, or how such variability will inform future study design, sample size calculations, or protocol refinement.

Additional methodological considerations

Several sources of variability occurring before and after the experimental protocol deserve further attention. The sanitary status of the animal facilities prior to study initiation is critical, as undetected infections or contaminations will significantly influence outcomes. Environmental factors such as room temperature are not monitored, despite their impact on hypothermia and long-term survival in septic mice. These aspects could be documented.

Internal variability within individual laboratories is not explicitly assessed. This could be evaluated using simple, pragmatic parameters all of which can serve as indicators of technical reproducibility.

Finally, there is a potential procedural bias related to fluid administration. The sepsis group does not receive a dextrose–glycerol injection at T4, unlike the other groups, which may induce unnecessary differences in long-term hypovolemia. Does the buprenorphine injection adequately compensate for this difference?

Conclusion

Overall, this protocol represents a landmark step forward in preclinical sepsis research. Its greatest contribution may lie less in its biological findings and more in its rigorous, transparent, and reproducible trial-like framework. With additional analyses that explicitly account for, rather than avoid, sources of variability, the study has strong potential to shape the design of future multicenter preclinical investigations and improve translational success in sepsis research.

Is the study design appropriate for the research question?

Yes

Is the rationale for, and objectives of, the study clearly described?

Yes

Are sufficient details of the methods provided to allow replication by others?

Yes

Are the datasets clearly presented in a useable and accessible format?

Yes

Reviewer Expertise:

preclinical sepsis researches

I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard.

Lalu Manoj Ottawa Health Research Institute, Canada

Competing interests: N/A

 6 2 2026

Arnaud Mansart, Université Paris-Saclay, Montigny le bretonneux, France 

  Approved

info_outline

1. Overall synthesis

This study protocol describes a well-conceived and ambitious multicenter preclinical study that directly addresses several long-standing limitations in translational sepsis research. By adopting clinical-trial standards, the NPSP-01 study represents a meaningful methodological advance. Beyond its biological objectives, the study’s main value lies in establishing a scalable and reproducible framework for future preclinical trials.

Major strengths

A key strength of the protocol is its six-center randomized controlled design, which remains rare in animal research. This approach directly tackles issues of reproducibility and inter-laboratory variability, two major contributors to translational failure in sepsis. Modeling the study after multicenter clinical trials is therefore a notable and commendable advance.

The methodological rigor is high. Centralized randomization stratified by center and sex, blinding at multiple stages, and clearly predefined primary and secondary outcomes all contribute to strong internal validity.

The inclusion of clinically relevant standard-of-care interventions, particularly antibiotics and fluid resuscitation, substantially enhances construct and translational validity. Many experimental sepsis models omit these elements, limiting clinical relevance. Establishing a treated baseline is especially valuable for future evaluation of novel therapeutic strategies.

The explicit inclusion of both male and female mice aligns well with current best practices and strengthens the biological relevance of the findings.

Another major strength is the extensive feasibility and process evaluation. Metrics such as technical success rates, protocol adherence, and biobanking completion will generate highly informative data for the design of future multicenter preclinical studies, not only in sepsis but also in other disease areas.

Finally, the authors strike a thoughtful balance between harmonization and real-world variability, avoiding excessive standardization. This reflects a sophisticated understanding of external validity and enhances the generalizability of the findings.

We thank the reviewer for their careful review and insightful comments.  We have addressed their concerns below and in the manuscript. We believe the edits made should address outstanding concerns.

2. Areas for improvement and clarification

The relatively short experimental time frame, with an 8-hour endpoint, limits insight into the dynamic kinetics of sepsis and the progression of organ dysfunction. While pragmatic, this constraint should be clearly acknowledged when interpreting results and extrapolating to human sepsis.

Response: We thank the reviewer for this insightful comment. We acknowledge that the 8h experimental timeframe limits assessment of sepsis progression and organ dysfunction over longer periods. We have added this to the limitation section in the discussion.

“ There are several aspects that should be considered when applying the model described in this protocol. 1) The 8h experimental timeframe limits assessment of longer-term sepsis progression and organ dysfunction, and modifications would be required for studies addressing these aspects….”

4. Important dimensions such as age-related dysfunction are not addressed. Although sex diversity is intentionally introduced, age (a major determinant of sepsis susceptibility, immune response, and outcome) is not considered.

Response: We agree, however as first step we want to ensure feasibility in a younger, healthier cohort prior to considering more resource intensive experiments in comorbid and aged animals. We plan to extend the model in future studies to incorporate additional complexity. This has now been addressed as a limitation in the discussion.

“… 2) The use of young animals does not capture age-related physiological changes or disease susceptibility; future adaptations incorporating aged animals, neonatal models, or relevant comorbidities may improve construct validity…”

5. The protocol would also benefit from a clearer discussion of downstream decision rules. It is not specified what degree of inter-laboratory heterogeneity would be considered acceptable, or how such variability will inform future study design, sample size calculations, or protocol refinement.

Response: We agree that clearer downstream decision rules strengthen the protocol. There are no universally accepted thresholds defining “acceptable” versus “high” inter-laboratory variability in multicentre clinical trials (i.e., where these issues have been studied extensively). Rather, interpretation is context-dependent and informed by empirical distributions of site-level variance.

We have now clarified how inter-laboratory heterogeneity will be interpreted using intraclass correlation coefficients (ICCs), and how these estimates will inform protocol refinement, sample size assumptions, and design choices for future multilaboratory studies.

Statistical analysis plan:

“The random effects analyses will be used to quantify variability in the IL6 outcome between centers using intracluster correlation coefficients (ICCs). Intracluster correlation coefficient estimates the proportion of total outcome variance attributable to between-laboratory differences. Consistent with practice in multicentre clinical trials, smaller ICC values will be interpreted as indicating limited between-laboratory heterogeneity, whereas larger values will prompt closer examination of protocol implementation and contextual sources of variability rather than being treated as a priori unacceptable.”

Discussion:

“ Estimates of interlaboratory variance from NPSP-01 may help inform the design of subsequent multilaboratory studies by guiding assumptions for sample size calculations, determining whether center should be modeled as a random effect, and identifying domains where additional harmonization or training may be required. Where substantial heterogeneity is observed, future protocols will need to consider incorporating targeted refinements, or modified designs to improve interpretability.”

6. Additional methodological considerations

Several sources of variability occurring before and after the experimental protocol deserve further attention. The sanitary status of the animal facilities prior to study initiation is critical, as undetected infections or contaminations will significantly influence outcomes. Environmental factors such as room temperature are not monitored, despite their impact on hypothermia and long-term survival in septic mice. These aspects could be documented.

Response: We thank the reviewer for this comment. Procedures for temperature and humidity monitoring, as well as the cage change schedule, are defined in our protocols. This information has now been added to the manuscript:

“ Temperature and humidity will be documented within each animal facility. To maintain sanitary status, cage changes will be performed every 10-14 days, or earlier if cage soiling exceeds each centre’s standard thresholds, in line with local animal care procedures.”

7. Internal variability within individual laboratories is not explicitly assessed. This could be evaluated using simple, pragmatic parameters all of which can serve as indicators of technical reproducibility.

Response: Intra-centre variability will be reported using 95% confidence intervals around the primary outcome (IL6) estimate. The approach was described in section “ Inter-laboratory variability analysis” in the manuscript.

8. Finally, there is a potential procedural bias related to fluid administration. The sepsis group does not receive a dextrose–glycerol injection at T4, unlike the other groups, which may induce unnecessary differences in long-term hypovolemia. Does the buprenorphine injection adequately compensate for this difference?

Response: None of the groups receive dextrose-glycerol at T4; only the treatment groups receive fluids and antibiotics, while the sepsis group does not. We acknowledge this as a potential limitation, but fluid resuscitation and antibiotics represent the primary intervention under study. Therefore, we could not compensate for administered volume in the sepsis group.

8. Conclusion

Overall, this protocol represents a landmark step forward in preclinical sepsis research. Its greatest contribution may lie less in its biological findings and more in its rigorous, transparent, and reproducible trial-like framework. With additional analyses that explicitly account for, rather than avoid, sources of variability, the study has strong potential to shape the design of future multicenter preclinical investigations and improve translational success in sepsis research.

We again thank the reviewer for their positive comments and recognizing the value of our planned approach.

 10.5256/f1000research.180201.r437807 Reviewer response for version 1 Amenyogbe Nelly 1 Referee https://orcid.org/0000-0002-7817-1534 Dalhousie University, Dalhousie, Canada

Competing interests: No competing interests were disclosed.

 31 12 2025 Copyright: © 2025 Amenyogbe N 2025 This is an open access peer review report distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. recommendation approve-with-reservations

The manuscript describes a rigorous multi-centre preclinical protocol designed to assess the impact of iv fluids and broad-spectrum antibiotics on plasma IL-6 levels in an adult mouse model of polymicrobial sepsis.  The manuscript describes a thorough governance structure intended to meet the standards of clinical trials involving humans. This level of standardization both improves animal welfare and aims to improve the applicability of findings within preclinical models to human translation. Below are minor comments to improve the work.

Introduction:  The introduction can be written more clearly to describe the objectives of the work.  Currently, the introduction states that “For our initial study, NPSP-01, we are implementing procedural standardization and rigorous practices to establish baseline conditions for future investigations.” However, eventually in the methods section (where the challenge dose was justified) the authors reference their prior work (doi: 10.1186/s40635-023-00533-3) whereby multi-centre dose titrations were performed and the administration of fluids and antibiotics evaluated at 4 and 12 hours post sepsis challenge. In this manuscript, the early administration of Abics + fluids at 4h and at this dose rescued 100% of the mice. Those details are important to put the current work into context. For example, based on prior work, do the authors intend to later implement the sepsis model with 4h fluids and antibiotics with an 8 hour readout as a standard way to evaluate sepsis interventions in a clinical setting or is the current study simply evaluating the impact of the intervention on IL-6 levels?

Exclusion criteria:  Please elaborate on the welfare criteria put in place to exclude animals that arrive to the facility and indicate whether these are the same for all participating institutions.

Endpoint selection:  can the authors elaborate on the pilot experiments that informed the design of this pilot?  All mice survived to the eight hour timepoint.  However, When were the endpoints reached in these models?  How variable were the clinical scores of animals across different groups at this time?  What was the experiment LD for these pilot studies at the 0.75 mg/g dose?

Mouse acquisition:  At what age are the mice shipped to the respective institutions?  The authors wrote above that mice will remain with their dams (i.e. will not be weaned) until shipment to the laboratories.  Does the shipment time also represent the weaning time for these mice?

Experimental groups:  I would suggest to clarify here and in the flow chart sample size at each respective institution.  While eventually the authors clarify that the study size of 192 mice is the pooled sample size across all centres, it is not clear at this point nor is the sample size per site given in the Figure.

Randomization and blinding:  how are individual mice distinguished (e.g. ear notch? Another method?)

Experimental procedures:  What is the volume of injections given to each animal? Do the study centres also aim to synchronize the time of challenge? For euthanasia, the authors describe general anaesthesia and exsanguination. Can the authors comment specifically on the type, dose and duration of anaesthetic and whether this will also be constant across study centres?  Is there a secondary method of euthanasia following anaesthesia (e.g. cervical dislocation)?

Technical success:  Why would multiple events with the same mouse be counted as one event?  It would be more informative to analyze the frequency of each specific event separately across the study.

Figure:  A flow chart that illustrates the sample numbers by treatment group for each study centre would be more informative, and can still contain the same information as was already placed within the overview figure.

Discussion:  The discussion is very brief.  The authors can describe some of the potential challenges to deploying a standardized early treatment model to evaluate sepsis interventions.  For example, if the plasma IL-6 levels in the fluid and antibiotic treated groups is dramatically reduced, then would this readout allow for the detection of benefit from other interventions?  How would this reflect the timing of IV fluid and antibiotic support used in the human setting, if known?  Given that the team had previously identified that IV fluids and antibiotics given 12 hours post challenge are not effective, interventions that can effectively treat or prevent sepsis in cases where IV fluids or broad spectrum antibiotics are not given early enough in the challenge course would need to be evaluated in a different context.

Can the authors comment on whether plasma IL-6 as an indicator of sepsis severity varies across age groups? For example, in neonates the specificity of IL-6 for sepsis detection is generally lower than for adults (see doi: 10.1007/s00431-025-06409-w). The authors could suggest future work to evaluate the model in other physiological states (e.g. during pregnancy, ageing, neonates).

Other comments:

Can the authors comment on why rat cecal slury was selected instead of mouse cecal slurry?  How might the species difference impact the overall validity of the model?

How was the 4-hour timepoint for treatment administration selected?

Is the study design appropriate for the research question?

Yes

Is the rationale for, and objectives of, the study clearly described?

Partly

Are sufficient details of the methods provided to allow replication by others?

Partly

Are the datasets clearly presented in a useable and accessible format?

Not applicable

Reviewer Expertise:

innate immunity, neonatal sepsis, vaccines

I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.

Lalu Manoj Ottawa Health Research Institute, Canada

Competing interests: N/A

 6 2 2026

Nelly Amenyogbe, Dalhousie University, Dalhousie, Canada 

Approved With Reservations

info_outline

1. The manuscript describes a rigorous multi-centre preclinical protocol designed to assess the impact of iv fluids and broad-spectrum antibiotics on plasma IL-6 levels in an adult mouse model of polymicrobial sepsis.  The manuscript describes a thorough governance structure intended to meet the standards of clinical trials involving humans. This level of standardization both improves animal welfare and aims to improve the applicability of findings within preclinical models to human translation. Below are minor comments to improve the work.

Response: We thank the reviewer for their careful review of our protocol and providing thoughtful comments. We have addressed concerns below and in the manuscript. We believe the manuscript has been significantly strengthened as a result.

2. Introduction:  The introduction can be written more clearly to describe the objectives of the work.  Currently, the introduction states that “For our initial study, NPSP-01, we are implementing procedural standardization and rigorous practices to establish baseline conditions for future investigations.” However, eventually in the methods section (where the challenge dose was justified) the authors reference their prior work (doi: 10.1186/s40635-023-00533-3) whereby multi-centre dose titrations were performed and the administration of fluids and antibiotics evaluated at 4 and 12 hours post sepsis challenge. In this manuscript, the early administration of Abics + fluids at 4h and at this dose rescued 100% of the mice. Those details are important to put the current work into context. For example, based on prior work, do the authors intend to later implement the sepsis model with 4h fluids and antibiotics with an 8 hour readout as a standard way to evaluate sepsis interventions in a clinical setting or is the current study simply evaluating the impact of the intervention on IL-6 levels?

Response: Our previously published dose-titration study was not conducted as a fully implemented multicenter investigation, but rather served as an initial evaluation of fluid and antibiotic administration at 4 and 12 hours following sepsis induction to inform the development of NPSP-01. During subsequent feasibility assessments, we identified logistical and technical constraints that limited consistent implementation of this 72h model. To ensure standardization and feasibility across sites, the consortium therefore adopted an 8-hour experimental model for our multicenter investigations. We have now added this to the main manuscript for more clarity:

“As an initial step, we published a 72 hour dose-titration study evaluating fluid and antibiotic administration at 4 and 12 hours following sepsis induction to inform the development of NPSP-01 8 . During subsequent feasibility assessments, we identified logistical and technical constraints that limited consistent implementation of this model across laboratories. To ensure standardization and feasibility, we have adopted an 8 hour experimental model for our multicenter investigations. At this time point mice develop a septic phenotype, however, this precedes the 60-90% mortality we observed in our initial study. Thus, the 8 hour model also mitigates survivorship bias and helps ensure data collection from a full cohort.”

3. Exclusion criteria:  Please elaborate on the welfare criteria put in place to exclude animals that arrive to the facility and indicate whether these are the same for all participating institutions.

Response: We do not have a priori criteria for this purpose, and institutes will determine general mice wellness based on their institutional animal care standards and veterinarian recommendations. This has been now added to the manuscript:

“ At six weeks of age, trained Charles-River personnel will perform dual verification prior to packaging and transport to participating centers. Upon receipt, each institute will assess general mouse wellness based on its routine animal care standards and veterinary recommendations.”

4. Endpoint selection:  can the authors elaborate on the pilot experiments that informed the design of this pilot?  All mice survived to the eight hour timepoint.  However, When were the endpoints reached in these models?  How variable were the clinical scores of animals across different groups at this time?  What was the experiment LD for these pilot studies at the 0.75 mg/g dose?

Response: Fecal slurry (from 2 batches prepared in one center) at 0.75 mg/g resulted in heterogeneous mortality between the two centers. Mortality ranged from 10–60% at 16 h post-FIP, increased to 30–90% at 24 h, and reached 60–90% by 72 h post-FIP. Corresponding changes in MSS over the 72-h period for the 0.75 mg/g dose at each center are shown below.

Img1:   https://f1000research-files.f1000.com/linked/781635.163800_Img1.jpeg

(From: Sharma et al. Intensive Care Medicine Experimental (2023) 11:45; https://doi.org/10.1186/s40635-023-00533-3)

This has now been added to the introduction for context. 

“ At this time point mice develop a septic phenotype, however, this precedes the 60-90% mortality we observed in our initial study. Thus, the 8 hour model also mitigates survivorship bias as killing all animals at this timepoint will help ensure data collection from a full cohort.”

5. Mouse acquisition:  At what age are the mice shipped to the respective institutions?  The authors wrote above that mice will remain with their dams (i.e. will not be weaned) until shipment to the laboratories.  Does the shipment time also represent the weaning time for these mice?

Response: Mice will be weaned at 3-4 weeks of age and shipped to the study centers at 6 weeks of age. These details have now been clarified:

“Mice will be specifically bred for this study by the Charles River facility in Saint-Constant, Quebec, Canada (Charles River barrier facilities K61 and K92 (specific pathogen–free). Animals will originate from breeding units (one male, two females) and wean according to standard procedures.”

“ …At six weeks of age, trained Charles-River personnel will perform dual verification prior to packaging and transport to participating centers.”

“… Mice will be kept in the animal facility for a 4-6 week acclimation period prior to the start of the experiments to minimize stress and cage aggression. Experiments will begin when animals are 10-12 weeks old.”

6. Experimental groups:  I would suggest to clarify here and in the flow chart sample size at each respective institution.  While eventually the authors clarify that the study size of 192 mice is the pooled sample size across all centres, it is not clear at this point nor is the sample size per site given in the Figure.

Response: Thank you for the suggestion. The sample size (n) per center will be between 28 and 36. The exact numbers will be determined by feasibility and human resources at each center. This is similar to clinical trials, where a goal sample per centre is declared, but final numbers are determined by feasibility and resources.  Figure 2 has been revised to reflect this information.

Img2:  https://f1000research-files.f1000.com/linked/781637.163800_Img2.jpg

Revised Figure 2.

7. Randomization and blinding:  how are individual mice distinguished (e.g. ear notch? Another method?)

Response: Tail marking was used to distinguish individual mice. This has now been clarified in the manuscript.

“Mice will be assigned to each center randomly by Charles River on the day of shipment. Mouse ID will be assigned using tail markings on the day of the experiment. Identification will be independently verified by an additional HQP before the start of the experiment, during outcome assessments, and biobanking to ensure accuracy.”

8. Experimental procedures:  What is the volume of injections given to each animal? Do the study centres also aim to synchronize the time of challenge? For euthanasia, the authors describe general anaesthesia and exsanguination. Can the authors comment specifically on the type, dose and duration of anaesthetic and whether this will also be constant across study centres?  Is there a secondary method of euthanasia following anaesthesia (e.g. cervical dislocation)?

Response: The following information has been added to the manuscript:

“All centres will initiate experiments in the morning and stagger induction time so that animals reach their endpoints within a narrow time window (8 ± 10 minutes) to minimise potential timing-related confounders.”

“ Injection volumes will be adjusted according to body weight and experimental group.”

“Euthanasia procedures are standardised across centres. Terminal anaesthesia will be induced with isoflurane (5% in an induction chamber followed by 1–2% via nose cone in oxygen at 0.5–1 L/min). The duration of anaesthesia will be recorded in standardised ‘ mouse report forms’ . Cervical dislocation will be used as a secondary method of euthanasia following anaesthesia.”

9. Technical success:  Why would multiple events with the same mouse be counted as one event?  It would be more informative to analyze the frequency of each specific event separately across the study.

Response: Feasibility was assessed at the level of the experimental unit (mouse) and was a process-level question. Multiple deviations could occur across different, unrelated SOP categories, but any single event in a mouse was sufficient to classify it as a deviation. This is sensitive – a single deviation would count as a process failure.

10. Figure:  A flow chart that illustrates the sample numbers by treatment group for each study centre would be more informative, and can still contain the same information as was already placed within the overview figure.

Response: Thanks for the suggestion. As noted above, the total number of animals per center will range between 28 and 36, with exact numbers determined by feasibility and capacity at each center. Consequently, the exact number of mice per treatment group at each center cannot be specified at this stage. These details will be provided once the study is completed.

11. Discussion:  The discussion is very brief.  The authors can describe some of the potential challenges to deploying a standardized early treatment model to evaluate sepsis interventions.  For example, if the plasma IL-6 levels in the fluid and antibiotic treated groups is dramatically reduced, then would this readout allow for the detection of benefit from other interventions?  How would this reflect the timing of IV fluid and antibiotic support used in the human setting, if known?  Given that the team had previously identified that IV fluids and antibiotics given 12 hours post challenge are not effective, interventions that can effectively treat or prevent sepsis in cases where IV fluids or broad spectrum antibiotics are not given early enough in the challenge course would need to be evaluated in a different context.

Response: We thank the reviewer for these important comments. As this manuscript describes the protocol, we have intentionally kept the discussion brief. In order to avoid any bias toward the study outcomes, we plan to address potential challenges, translational considerations, and limitations in detail in the main paper once the data are collected, analyzed, and interpreted.

Nonetheless, based on this comment and a similar suggestion by the second reviewer, we have now added a paragraph on the limitations.

“There are several aspects that should be considered when applying the model described in this protocol. 1) The 8h experimental timeframe limits assessment of longer-term sepsis progression and organ dysfunction, and modifications would be required for studies addressing these aspects; 2) The use of young animals does not capture age-related physiological changes or disease susceptibility; future adaptations incorporating aged animals, neonatal models, or relevant comorbidities may improve construct validity;  3) additional readouts may be needed when evaluating other therapeutics, as outcomes depend on the characteristics of the animal model and the effects of standard of care treatments, such as fluids and antibiotics.”

12. Can the authors comment on whether plasma IL-6 as an indicator of sepsis severity varies across age groups? For example, in neonates the specificity of IL-6 for sepsis detection is generally lower than for adults (see doi: 10.1007/s00431-025-06409-w). The authors could suggest future work to evaluate the model in other physiological states (e.g. during pregnancy, ageing, neonates).

Response: We thank the reviewer for this suggestion. This aspect has now been added to the limitation paragraph in the protocol. Please see the response to Question 11.

Other comments:

13. Can the authors comment on why rat cecal slury was selected instead of mouse cecal slurry?  How might the species difference impact the overall validity of the model?

Response: Rat cecal slurry was selected by our group to provide a single, standardized batch that could be used across all centres. Rats were also selected to minimize animal use - if we were to use mouse cecal slurry a very large number of mice would need to be euthanized to produce an adequate volume. We acknowledge that species differences between the slurry donor (rat) and experimental subjects (mouse) could potentially affect some aspects of the model. We have now added this information to the manuscript:

“ Rats were selected to minimise animal use, as producing an adequate volume of fecal slurry from mice would require euthanizing a substantially larger number of animals.”

14. How was the 4-hour timepoint for treatment administration selected?

Response: The 4h timepoint for treatment administration was selected based on pilot data and the aims of the study. Administering at 4 hours post FIP allows interventions to be given neither too early avoiding masking the disease phenotype, nor too late when their efficacy would be minimal. This timing also reduces handling stress by aligning treatment with the scheduled 4h body weight and core temperature assessments and a single analgesic dose sufficient for pain control for the rest of the experiment, thereby avoiding additional injections in the 8h model and supporting animal welfare.