Septic peritonitis is the presence of bacteria or other infectious pathogens within the peritoneal space originating from a septic focus. In regard to septic foci, intestinal perforation is the most common cause seen in dogs; however, other causes can include penetrating trauma, urinary bladder rupture, and gallbladder rupture.1 Once within the peritoneal space, bacteria can transit into the bloodstream, a process known as bacteremia. Bacteremia then progresses to septicemia once an immune response is elicited. This immune response is often associated with uncontrolled systemic inflammation, which is the main driver of morbidity and mortality in septic patients, eventually leading to dysregulated coagulation and multiple organ dysfunction.2
Although there has been significant research into the diagnosis and treatment of sepsis, there has been little to no improvement in survival rates among dogs with sepsis over the last 30 years. One study3 comparing mortality rates in septic dogs from 1988 to 1993 and then from 1999 to 2003 reported mortality rates of 64% and 57%, respectively. Two additional studies4,5 from 2013 and 2016 reported mortality rates that ranged from 44% to 70%. There has been a similar lack of significant improvement in survival seen with sepsis in humans over the last several decades.6–9 This lack of progress around improving how we care for septic patients indicates a significant unmet need for improved diagnostic and treatment options for dogs, as well as humans, diagnosed with sepsis.
To date, there has been no diagnostic test available to specifically confirm the diagnosis of sepsis in veterinary patients other than through blood cultures. The culturing process can be complicated, requires a prolonged turnaround time, and necessitates the collection of large volumes of blood, which is not always feasible in critically ill septic patients. Additionally, achieving growth in these cultures is uncommon, with a 20% positive growth rate having previously been reported.10 This, in combination with the prolonged turnaround time for obtaining results, raises questions about the diagnostic value of blood cultures.10,11 Rather than the direct confirmation of sepsis, the value in these cultures more so lies with the susceptibility data that is reported. However, with the continued development of new diagnostic platforms, it is expected that novel tests will become available that will replace blood cultures as the gold standard, such as point-of-care PCR analysis.11,12
In addition to diagnosis confirmation and antimicrobial susceptibility profiling, another diagnostic need in sepsis is the ability to determine illness severity and to track a patient’s response to treatment. The acute patient physiologic and laboratory evaluation (APPLE) score and the measurement of procalcitonin are the most thoroughly demonstrated measures to estimate illness severity in septic dogs.13,14 Furthermore, procalcitonin has been demonstrated to have potential value in monitoring patient response to treatment.15 It is suspected that there are other blood-based targets that could be used in this fashion to both determine illness severity and provide useful information during treatment to track patient response, such as the need to change antimicrobials if the first-line agents are not proving effective.
With the primary driver of morbidity and mortality in sepsis being the systemic inflammatory response that develops, it seems reasonable that a therapeutic approach that aims to reduce this inflammation and its sequelae would be beneficial. However, previous studies16 that have investigated the use of immunosuppressive drugs such as corticosteroids in the setting of sepsis have demonstrated no discernable benefit or even potentially detrimental effects. Corticosteroids are known for their broad anti-inflammatory and immunosuppressive effects. They function by downregulating the expression of multiple inflammatory genes and upregulating anti-inflammatory genes, affecting a wide range of immune functions.17 In the context of sepsis, while the initial intention of using corticosteroids is to mitigate the overwhelming inflammatory response, their broad-spectrum activity could suppress other essential immune functions.
In contrast, targeting a specific molecule involved in the inflammatory process offers a more focused approach. This targeted approach might preserve the body’s ability to combat the infectious agent effectively while reducing some of the harmful inflammatory responses that contribute to septic pathophysiology. A targeted therapy might offer the advantage of mitigating specific deleterious processes in sepsis while avoiding the wide-ranging suppression of immune function that can complicate the use of corticosteroids.
This pilot study seeks to determine the feasibility of evaluating gene expression profiles on whole blood from dogs with and without septic peritonitis secondary to intestinal perforations and to identify potential molecular targets that could aid themselves in the future development of novel diagnostics and therapeutics for sepsis.
Methods
Eligibility criteria
Dogs with confirmed septic peritonitis of any cause were eligible for inclusion in this pilot study. Ethical approval was obtained on the study protocol through the ACI Biosciences Animal Care and Use Committee, and client-informed consent was obtained for each dog enrolled. Dogs were enrolled through 3 tertiary veterinary referral hospitals located across the United States: Massachusetts Veterinary Referral Hospital, Wheat Ridge Animal Hospital, and Veterinary Specialty Hospital of San Diego.
Confirmed septic peritonitis was defined as the positive identification of intracellular bacteria on cytologic examination of peritoneal effusion in combination with an identifiable septic focus. Dogs were excluded if they had received any transfusion products within the previous 7 days or any NSAIDs, glucocorticoids, or other immunosuppressive medications in the previous 14 days. Dogs were allowed to enroll if they had received antimicrobials given the importance of their timely administration in the setting of sepsis; however, dogs who had been exposed to antimicrobials for more than 8 hours before enrollment were excluded. Dogs with a platelet count less than 100 K/μL or with possible coagulopathy were also excluded. This exploratory study aimed to enroll 6 dogs with septic peritonitis and 6 healthy control dogs. The control dogs were confirmed as healthy based on physical examination, CBC, biochemistry, and a point-of-care ultrasound examination to confirm the absence of peritoneal effusion. This was a sample collection-only study, and all other care for these dogs was at the discretion of the attending veterinarian.
Sample collection
All dogs had a CBC and biochemistry performed. All of the septic peritonitis dogs also had lactate measured on a point-of-care device and blood oxygen saturation measured to be used to calculate the APPLE full score. Both the healthy and septic peritonitis dogs then had 2.5 mL of whole blood collected from a peripheral or jugular vein, which was immediately transferred into a 3-mL RNA stabilization tube (PreAnalytiX), which was gently inverted 10 times to ensure adequate mixing of the blood with the preservative. For the septic peritonitis dogs, the goal was for these samples to be collected before any significant treatments or interventions were initiated when possible in an attempt for the group to be standardized as much as possible. Each dog’s participation in the study was complete following the provision of this blood sample. All other patient care for these dogs was at the discretion of the attending veterinarian and the owner.
The tubes for all dogs were left to sit at room temperature for 2 hours before they were frozen at −20 °C. Within 48 hours of freezing, samples were then shipped overnight on ice packs to a centralized biobank (located within the Veterinary Specialty Hospital of San Diego) where they were stored at −80 °C. Once all samples were collected and batched at the central biobank, the samples were shipped overnight on dry ice to the Dartmouth Cancer Center Molecular Biology Core Facility for RNA extraction and transcript quantification. The samples were placed back into a −80 °C freezer pending their analysis. All samples were confirmed to have remained frozen without thawing on receipt to both the biobank and the lab.
RNA extraction and quantification
Before RNA extraction, the blood samples were thawed from −80 °C storage and equilibrated to room temperature for 2 hours. Total RNA was extracted using a commercial kit (PreAnalytiX) following the manufacturer’s protocol. The RNA was eluted by applying 40 μL elution buffer to the spin column provided in the kit, followed by centrifugation. Another 40 μL of elution buffer was then applied to the same column, and after a second centrifugation, the combined eluates were vortexed to ensure homogeneity. The samples were then treated with DNase to remove any potential contaminating genomic DNA, ensuring no interference with transcript quantification. RNA concentration was then quantified using another commercial kit (Thermo Fisher Scientific) to ensure sufficient yield and quality.18 Total RNA was used for the analysis without further purification or isolation of mRNA or small RNAs (ie, microRNAs).
Transcript quantification
For transcript quantification, 100 ng of total RNA was loaded into a commercial canine immuno-oncology assay consisting of 800 genes was used (NanoString Technologies; a full list of genes included is provided; Supplementary Table S1). The hybridization mixture was prepared by combining 3 μL of reporter code set, 5 μL of hybridization buffer, up to 5 μL of RNA (100 ng), and 2 μL of capture probe set. This standard input was used for all samples analyzed.
The hybridization mixture was added to labeled strip tubes, which were then centrifuged briefly. The hybridization was conducted overnight in a thermal cycler (Fisher Scientific) set to 62 °C with a lid temperature of 67 °C for a total of 15 hours.
Posthybridization processing
Posthybridization, samples underwent 2-step magnetic bead-based purification, which involved sequential binding to beads that are complementary to the capture and reporter probes. This process included several wash steps to remove nonspecifically bound nucleic acids and excess probes.
After purification, the target-probe complexes were transferred to a cartridge prearranged with 12 channels corresponding to the sample tubes. The cartridge was sealed and loaded into the analyzer (NanoString Technologies). Data collection was facilitated by epifluorescence microscopy, where each channel was imaged independently using a charge-coupled device camera. The resulting images were analyzed to count and tabulate target molecules, utilizing a reporter library file specific to the reporter code set.
Data analysis
Transcript data from the healthy and septic dogs were analyzed to identify differentially expressed genes and to characterize the molecular pathways involved in canine sepsis. The data were normalized using positive and negative control probes, with background thresholding set at the mean of the negative control probe counts plus 2 SD, as per the software’s recommendations (nSolver 4.0; NanoString Technologies). Genes with less than 20 counts across all samples were filtered out, and then counts were normalized to positive control spike-ins and housekeeping genes using the default settings. Unsupervised hierarchical clustering based on Euclidean distance was performed using normalized counts, and then a heat map was generated with genes z-score normalized.
All further analyses were performed using R Statistical Software (v4.1.1; R Core Team; The R Foundation). Fold changes between the septic samples and healthy samples were performed using the DESeq2 R package (Wald test).19 For both differential expression analyses, genes were considered significantly differentially expressed (DEGs) based on an adjusted P value threshold (using the Benjamin-Hochberg method) of less than .05 and an absolute log2-fold change of greater than .5. A principal component analysis (PCA) was performed using the DESeq2 package to assess the variability in gene expression across samples. Significantly upregulated and downregulated genes (septic peritonitis vs healthy controls) called using the DESeq2 package were visualized as a volcano plot using “ggplot2” in R.20 Heat map visualization of differentially expressed genes was performed of z-score normalized genes (scale = “row”) and plotted using “pheatmap” in R with default hierarchical clustering by rows (genes) and columns (samples) using Euclidean distance (cluster_rows = TRUE; cluster_cols = TRUE). Finally, DEGs were compared using “ggvenndiagram” in R.21
Results
The median age of the control group was 5.5 years (range, 2 to 11 years), and the median weight was 23.9 kg (range, 15.6 to 59.4 kg). There were 2 intact males, 2 castrated males, 1 intact female, and 1 spayed female. The breeds represented included 2 Welsh Springer Spaniels and 1 each of Bernese Mountain Dog, Boxer, Labrador Retriever, and Saint Bernard. The CBC and biochemistry results for this cohort are summarized (Supplementary Table S2).
The median age of the septic peritonitis group was 7 years (range, 3 to 9 years), and the median weight was 33.3 kg (range, 6.7 to 35.8 kg). There were 3 castrated males, 1 intact male, 1 spayed female, and 1 intact female. The breeds represented included 1 each of Collie, German Shepherd Dog, Greyhound, Labrador Retriever, Pomeranian, and Standard Poodle. The CBC and biochemistry results for this cohort are summarized (Supplementary Table S3). The median APPLE full score was 43 (range, 26 to 52). Intestinal perforation was the underlying etiology for all 6 of the dogs with septic peritonitis, with 4 of these being due to perforating intestinal obstructions and 2 due to incisional dehiscence of previous intestinal obstruction surgery sites.
The RNA isolation from the control dog samples had a median yield of 6,620 ng (range, 3,752 to 13,600 ng) and for the septic peritonitis group was 29,160 ng (range, 13,680 to 52,000 ng). Unsupervised hierarchical clustering of samples demonstrated broad transcriptional differences between the healthy and septic patient populations, with distinct differences in the gene expression signatures between the 2 groups (Figure 1). The full output of the analyses, including effect sizes (log2-fold change), raw P values, and false discovery rate–adjusted P values (q values), are summarized (Supplementary Table S4).
Heat map of gene expression of dogs with and without septic peritonitis that demonstrates broad transcriptional differences between the healthy and septic patient populations. Rows represent differentially expressed genes, and columns represent individual samples (B1 to B12). Hierarchical clustering of both genes and samples was performed using Euclidean distance. Gene expression values were z-score normalized (scale = “row”), with blue indicating increased expression and yellow indicating decreased expression relative to the mean expression level of each gene across all samples.
Citation: American Journal of Veterinary Research 2025; 10.2460/ajvr.24.06.0180
The PCA results further demonstrated a clear separation between the healthy and septic peritonitis blood samples along the first principal component (PC1), which accounted for 73% of the total variance (Figure 2). Additionally, there was noticeable interindividual variation within the septic peritonitis group, evident along the second principal component (PC2), which accounted for 7% of the variance.
First and second principal component (PC1 and PC2) analysis plot showing clear separation between the healthy and septic peritonitis samples, where PC1 accounts for 73% of the total variance, while the PC2 captures 7% variance, highlighting interindividual variation within the septic peritonitis group.
Citation: American Journal of Veterinary Research 2025; 10.2460/ajvr.24.06.0180
A total of 445 DEGs were identified using DESeq2 analysis. Examples of differential genes include matrix metallopeptidase 9, IL-1 receptor type 2, proliferating cell nuclear antigen, phosphatidylinositol 3-kinase (PI3K) catalytic-γ, and cluster of differentiation 55 (Figure 3).
Volcano plot of dogs with and without sepsis. Dots in red are individual genes increased in expression across septic samples. Dots in blue are genes with increased expression across healthy samples. Differentially expressed (diffexp) genes identified include activating transcription factor 1 (ATF1), complement component 3 (C3), cluster of differentiation 55 (CD55), integrin subunit-α4 (ITGA4), IL-1 receptor type 2 (IL1R2), lactotransferrin (LTF), lymphocyte cytosolic protein 1 (LCP1), matrix metallopeptidase 9 (MMP9), paired box 5 (PAX5), proliferating cell nuclear antigen (PCNA), phosphatidylinositol 3-kinase catalytic-γ (PIK3CG), S100 calcium-binding protein A8 and A9 (S100A8 and S100A9), and transporter associated with antigen-processing 1 gene (TAP1).
Citation: American Journal of Veterinary Research 2025; 10.2460/ajvr.24.06.0180
Discussion
This pilot study demonstrates that the evaluation of whole blood gene expression profiles from dogs is feasible and that gene expression differs between dogs with and without septic peritonitis. It is expected that the majority of the genes detected in this analysis are being expressed by mononuclear cells within the blood and less so from erythrocytes or platelets, which are anucleate and likely contribute less significantly to the whole blood gene expression profile.22 This is further supported by the understanding that the primary driver of morbidity and mortality in sepsis and septic peritonitis is the systemic inflammatory response that is elicited, increasing the expression of these gene pathways within the mononuclear cells.23 However, it is possible that some of the detected transcripts may originate from other tissues, particularly in the context of septic peritonitis, where organ dysfunction and cellular injury can release organ-specific gene transcripts into circulation.24,25
There was a more than 4-fold increase in the median RNA yield in the septic peritonitis dogs compared to healthy dogs with no overlap in the ranges observed. There were no differences in the preanalytical conditions for the samples between the groups, making potential RNA degradation less likely as the cause for the disparity. Likely explanations for the elevated RNA yield in the septic peritonitis dogs include increased gene expression, elevated levels of free circulating RNA from tissue injury and systemic inflammation, or increased levels of circulating mononuclear cells.
The PCA in this study demonstrated distinct clustering of gene expression profiles, clearly differentiating between the septic and healthy dogs (73% of variance along PC1). This distinct separation underscores significant alterations in gene expression associated with septic peritonitis, as would be expected.
Unsupervised hierarchical clustering of the samples using normalized counts revealed broad transcriptional distinctions between healthy and septic blood. Further evaluation of sepsis associated gene expression in a larger cohort is necessary to understand common genes and pathways; however, these data suggest that transcriptional biomarkers of sepsis can be discovered from whole blood.
The PCA findings revealed that the major source of variation in gene expression between the healthy and septic peritonitis dogs was captured by PC1, suggesting that the overall differences between these 2 groups were the primary driver of transcript variation. Additionally, the interindividual variability observed in the septic peritonitis group, captured by PC2, could be associated with clinical factors such as disease severity or treatment history. These findings highlight the potential for certain gene expression features that could serve as biomarkers for diagnosis or prognosis in septic peritonitis. Further analysis of the PCA coefficients, particularly those contributing to PC2, may reveal gene expression variations linked to clinical variables, offering deeper insight into sources of interindividual variation. The identification of a likely outlier in the septic peritonitis group along PC1 could be explained by unique clinical circumstances, although none were identified. Future studies could benefit from incorporating more comprehensive clinical annotation to explore these relationships in greater depth.
Differential expression analysis revealed candidate genes requiring further validation in a larger cohort. Genes of interest included matrix metallopeptidase 9, IL-1 receptor type 2, proliferating cell nuclear antigen, PI3K catalytic-γ, and cluster of differentiation 55. All of these genes could have potential relevance to the pathophysiology of sepsis and are commonly associated with key processes associated with a septic process such as inflammation, cell proliferation, and immune response.
Matrix metallopeptidase 9 is an enzyme involved in the breakdown of the extracellular matrix and is known to play a critical role in inflammatory processes, tissue remodeling, and healing.26 Interleukin-1 receptor type 2 is a cell surface receptor that acts as a decoy receptor for IL-1 (both IL-1α and IL-1β) and inhibits their activity.27 Proliferating cell nuclear antigen is a protein that is associated with DNA replication and repair, specifically acting as a sliding clamp for DNA polymerase during DNA replication.28 Phosphatidylinositol 3-kinase catalytic-γ is a subunit of PI3K, involved in the PI3K/Akt signaling pathway, and plays a fundamental role in the immune system, regulating the activation, differentiation, and survival of immune cells.29 Cluster of differentiation 55 is a glycoprotein attached to the cell surface and plays a crucial role in regulating the complement system, which is part of the innate immune response.30
There have been several previous studies evaluating the transcriptome in humans with sepsis.31–35 When used to examine peripheral blood mononuclear cells at the single-cell level, distinct immune cell profiles were documented in septic patients.31 Notably, nuclear enriched abundant transcript 1, a long noncoding RNA, was found to be upregulated in monocytes of septic patients and is associated with inflammatory signaling.31 In another study,32 genes identified as being differentially expressed in septic patients compared to healthy controls included integrin subunit-α M, cluster of differentiation 44, complement component 3a receptor 1, and IL-2 receptor subunit-γ. These genes are predominantly localized within macrophages, playing significant roles in immune regulation and the inflammatory response, further highlighting their potential as valuable biomarkers and therapeutic targets in sepsis.32 Additionally, transcriptomic studies31 have shown that human leukocyte antigen-DR isotype, an important marker for antigen presentation, is significantly downregulated in septic patients, emphasizing immune dysfunction as a key feature of sepsis progression.
This study has several limitations that should be considered. First, the sample size was small, including only 6 dogs with septic peritonitis and 6 healthy controls. The primary focus of this study was to assess the feasibility of evaluating gene expression profiles on canine whole blood and was not powered to identify potential transcriptomic targets with diagnostic or therapeutic value. Future studies with a larger cohort, including dogs with other causes of septic peritonitis as well as noninfectious causes of systemic inflammation, will be necessary to further identify and validate any potential targets. Additionally, the study analyzed whole blood samples, which are believed to have predominantly reflected mononuclear cell gene expression; however, discrete isolation of the mononuclear cells was not performed before transcript quantification. Future studies that would include mononuclear cell isolation before quantification may allow for better identification of potential targets that are associated with the inflammatory process that is associated with sepsis.
For this study, a commercial canine assay consisting of 800 genes specific to canine immuno-oncology was used. With there being an estimated 20,000 to 25,000 genes in the canine transcriptome, it is possible that genes that are relevant to septic peritonitis outside of the 800 included were not captured in the current analysis. Additionally, this study did not measure proteins or conduct any metabolic or immunologic assessments to determine whether the changes in gene expression were associated with changes in biologic functions, which could be a focus of future studies as potentially valuable transcriptomic targets are further identified and validated. However, this study has demonstrated that gene transcripts can be quantified from canine whole blood and thus could be used directly as potential biomarkers. Finally, it is possible that the higher RNA abundance in the septic peritonitis group could have made it more difficult to detect lower abundance transcripts. These input differences can artificially inflate differential expression analysis when not correcting for input.
With workflow feasibility now confirmed, future studies are proposed to further evaluate the gene expression profiles in dogs with septic peritonitis, including the upregulated genes documented in this report, to look for potentially valuable diagnostic and therapeutic targets. This would include evaluating expression profiles in a larger group of dogs with septic peritonitis and applying the APPLE full score as a stratifying factor to see if there are correlations between gene expression and illness severity. The results of these future efforts will aid in further defining genes as possible diagnostic and therapeutic targets in the setting of canine septic peritonitis.
Supplementary Materials
Supplementary materials are posted online at the journal website: avmajournals.avma.org.
Acknowledgments
None reported.
Disclosures
The study was conducted and funded by Ethos Discovery, a nonprofit veterinary scientific incubator. Eligible dogs were enrolled in this study through hospitals within the Ethos Veterinary Health network. All authors are employees of Ethos Veterinary Health.
No AI-assisted technologies were used in the generation of this manuscript.
Funding
This study was funded by Ethos Discovery.
ORCID
S. D. Stewart https://orcid.org/0000-0003-1148-2589
References
- 1.↑
Macintire DK, Drobatz KJ, Haskins SC, Saxon WD. Manual of Small Animal Emergency and Critical Care Medicine. John Wiley & Sons; 2012.
- 3.↑
Bentley AM, Otto CM, Shofer FS. Comparison of dogs with septic peritonitis: 1988–1993 versus 1999–2003. J Vet Emerg Crit Care. 2007;17(4):391–398.
- 4.↑
Abelson AL, Buckley GJ, Rozanski EA. Positive impact of an emergency department protocol on time to antimicrobial administration in dogs with septic peritonitis. J Vet Emerg Crit Care. 2013;23(5):551–556.
- 5.↑
Barfield DM, Tivers MS, Holahan M, Welch K, House A, Adamantos SE. Retrospective evaluation of recurrent secondary septic peritonitis in dogs (2000–2011): 41 cases. J Vet Emerg Crit Care. 2016;26(2):281–287.
- 6.↑
Luhr R, Cao Y, Soederquist B, Cajander S. Trends in sepsis mortality over time in randomised sepsis trials: a systematic literature review and meta-analysis of mortality in the control arm, 2002–2016. Crit Care. 2019;23(1):241.
- 7.
Bauer M, Gerlach H, Vogelmann T, Preissing F, Stiefel J, Adam D. Mortality in sepsis and septic shock in Europe, North America and Australia between 2009 and 2019—results from a systematic review and meta-analysis. Crit Care. 2020;24(1):239.
- 8.
Stevenson EK, Rubenstein AR, Radin GT, Wiener RS, Walkey AJ. Two decades of mortality trends among patients with severe sepsis: a comparative meta-analysis. Crit Care Med. 2014;42(3):625–631.
- 9.↑
Rhee C, Klompas M. Sepsis trends: increasing incidence and decreasing mortality, or changing denominator? J Thorac Dis. 2020;12(suppl 1):S89.
- 10.↑
Saarenkari HK, Sharp CR, Smart L. Retrospective evaluation of the utility of blood cultures in dogs (2009–2018): 45 cases. J Vet Emerg Crit Care. 2022;32(1):141–145.
- 11.↑
Silveira MM, Cândido SL, Santos KR, et al. Polymerase chain reaction and blood culture for diagnosis of canine sepsis. Ciênc Rural. 2018;48(6):e20170871. doi:10.1590/0103-8478cr20170871
- 12.↑
Heilmann RM, Xenoulis PG, Barr JW, et al. Comparison of PCR and conventional blood culture to analyze blood from dogs with suspected sepsis. Vet J. 2013;198(3):714–716. doi:10.1016/j.tvjl.2013.10.001
- 13.↑
Hayes G, Mathews K, Doig G, et al. The acute patient physiologic and laboratory evaluation (APPLE) score: a severity of illness stratification system for hospitalized dogs. J Vet Intern Med. 2010;24(5):1034–1047. doi:10.1111/j.1939-1676.2010.0552.x
- 14.↑
Goggs R, Milloway M, Troia R, Giunti M. Plasma procalcitonin concentrations are increased in dogs with sepsis. Vet Rec Open. 2018;5(1):e000255.
- 15.↑
Easley F, Holowaychuk MK, Lashnits EW, Nordone SK, Marr H, Birkenheuer AJ. Serum procalcitonin concentrations in dogs with induced endotoxemia. J Vet Intern Med. 2020;34(2):653–658.
- 16.↑
Annane D, Bellissant E, Bollaert P-E, et al. Corticosteroids in the treatment of severe sepsis and septic shock in adults: a systematic review. JAMA. 2009;301(22):2362–2375. doi:10.1001/jama.2009.815
- 17.↑
Coutinho AE, Chapman KE. The anti-inflammatory and immunosuppressive effects of glucocorticoids, recent developments and mechanistic insights. Mol Cell Endocrinol. 2011;335(1):2–13.
- 19.↑
Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014;15(12):550.
- 20.↑
Wickham H. Programming with ggplot2. In: ggplot2: Elegant Graphics for Data Analysis. Springer; 2016:241–253.
- 21.↑
Gao C. ggVennDiagram: a ‘ggplot2’ Implement of Venn Diagram. R package, version 1.2.2. 2022. Accessed June 15, 2024. https://gaospecial.github.io/ggVennDiagram/
- 22.↑
O’Connell GC, Wang J, Smothers C. Donor white blood cell differential is the single largest determinant of whole blood gene expression patterns. Genomics. 2023;115(6):110708.
- 23.↑
Lewis D, Chan D, Pinheiro D, Armitage-Chan E, Garden O. The immunopathology of sepsis: pathogen recognition, systemic inflammation, the compensatory anti-inflammatory response, and regulatory T cells. J Vet Intern Med. 2012;26(3):457–482.
- 24.↑
Yang JO, Zinter MS, Pellegrini M, et al. Whole blood transcriptomics identifies subclasses of pediatric septic shock. Crit Care. 2023;27(1):486. doi:10.1186/s13054-023-04689-y
- 25.↑
Petejova N, Martinek A, Zadrazil J, et al. Expression and 7-day time course of circulating microRNAs in septic patients treated with nephrotoxic antibiotic agents. BMC Nephrol. 2022;23(1):111. doi:10.1186/s12882-022-02726-6
- 26.↑
Van den Steen PE, Dubois B, Nelissen I, Rudd PM, Dwek RA, Opdenakker G. Biochemistry and molecular biology of gelatinase B or matrix metalloproteinase-9 (MMP-9). Crit Rev Biochem Mol Biol. 2002;37(6):375–536.
- 27.↑
Colotta F, Re F, Muzio M, et al. Interleukin-1 type II receptor: a decoy target for IL-1 that is regulated by IL-4. Science. 1993;261(5120):472–475. doi:10.1126/science.8332913
- 28.↑
Moldovan G-L, Pfander B, Jentsch S. PCNA, the maestro of the replication fork. Cell. 2007;129(4):665–679.
- 29.↑
Vanhaesebroeck B, Guillermet-Guibert J, Graupera M, Bilanges B. The emerging mechanisms of isoform-specific PI3K signalling. Nat Rev Mol Cell Biol. 2010;11(5):329–341.
- 30.↑
Nicholson-Weller A, Wang C. Structure and function of decay accelerating factor CD55. J Lab Clin Med. 1994;123(4):485–491.
- 31.↑
Wen M, Cai G, Ye J, Liu X, Ding H, Zeng H. Single-cell transcriptomics reveals the alteration of peripheral blood mononuclear cells driven by sepsis. Ann Transl Med. 2020;8(4):125. doi:10.21037/atm.2020.02.35
- 32.↑
Chen W, Guo W, Li Y, Chen M. Integrative analysis of metabolomics and transcriptomics to uncover biomarkers in sepsis. Sci Rep. 2024;14(1):9676.
- 33.
Tsakiroglou M, Evans A, Pirmohamed M. Leveraging transcriptomics for precision diagnosis: lessons learned from cancer and sepsis. Front Genet. 2023;14:1100352.
- 34.
Almansa R, Heredia-Rodríguez M, Gomez-Sanchez E, et al. Transcriptomic correlates of organ failure extent in sepsis. J Infect. 2015;70(5):445–456. doi:10.1016/j.jinf.2014.12.010
- 35.↑
Bauer M, Giamarellos-Bourboulis EJ, Kortgen A, et al. A transcriptomic biomarker to quantify systemic inflammation in sepsis—a prospective multicenter phase II diagnostic study. EBioMedicine. 2016;6:114–125. doi:10.1016/j.ebiom.2016.03.006
