Next Article in Journal
Knockout of the Amino Acid Transporter SLC6A19 and Autoimmune Diabetes Incidence in Female Non-Obese Diabetic (NOD) Mice
Next Article in Special Issue
Circulating Metabolites as Biomarkers of Disease in Patients with Mesial Temporal Lobe Epilepsy
Previous Article in Journal
Serum Metabolomic and Lipoprotein Profiling of Pancreatic Ductal Adenocarcinoma Patients of African Ancestry
Previous Article in Special Issue
Differential Glycosylation Levels in Saliva from Patients with Lung or Breast Cancer: A Preliminary Assessment for Early Diagnostic Purposes
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Metabolites
Volume 11
Issue 10
10.3390/metabo11100664
metabolites-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Metabolomics and Inflammatory Mediator Profiling for the Differentiation of Life-Threatening and Non-Severe Appendicitis in the Pediatric Population

by
Nusrat S. Shommu
1,
Jaime Blackwood
2,
Craig N. Jenne
3,4,
Ari R. Joffe
5,
Dori-Ann Martin
2,
Beata Mickiewicz
2,
Mary Brindle
6,7,
Robin Eccles
6,
Hans J. Vogel
1,
Graham C. Thompson
2,8,*,† and
on behalf of the Alberta Sepsis Network
†
1
Bio-NMR Centre, Department of Biological Sciences, University of Calgary, Calgary, AB T2N 1N4, Canada
2
Department of Pediatrics, Cumming School of Medicine, University of Calgary, Calgary, AB T2N 1N4, Canada
3
Department of Critical Care Medicine, Cumming School of Medicine, University of Calgary, Calgary, AB T2N 1N4, Canada
4
Department of Microbiology, Immunology and Infectious Disease, University of Calgary, Calgary, AB T2N 1N4, Canada
5
Department of Pediatrics, University of Alberta, Edmonton, AB T6G 2R3, Canada
6
Department of Surgery, Cumming School of Medicine, University of Calgary, Calgary, AB T2N 1N4, Canada
7
Department of Community Health Sciences, Cumming School of Medicine, University of Calgary, Calgary, AB T2N 1N4, Canada
8
Department of Emergency Medicine, Cumming School of Medicine, University of Calgary, Calgary, AB T2N 1N4, Canada
*
Author to whom correspondence should be addressed.
†
Membership of the Alberta Sepsis Network.
Metabolites 2021, 11(10), 664; https://doi.org/10.3390/metabo11100664
Submission received: 30 July 2021 / Revised: 18 September 2021 / Accepted: 24 September 2021 / Published: 28 September 2021
(This article belongs to the Special Issue Omics Technologies in Evaluating New Insights into Molecular Aspects of Human Health)
Browse Figures
Versions Notes

Abstract

:
While children with appendicitis often have excellent clinical outcomes, some develop life-threatening complications including sepsis and organ dysfunction requiring pediatric intensive care unit (PICU) support. Our study applied a metabolomics and inflammatory protein mediator (IPM) profiling approach to determine the bio-profiles of children who developed severe appendicitis compared with those that did not. We performed a prospective case-control study of children aged 0–17 years with a diagnosis of appendicitis. Cases had severe disease resulting in PICU admission. Primary controls had moderate appendicitis (perforation without PICU); secondary controls had mild appendicitis (non-perforated). Serum samples were analyzed using Proton Nuclear Magnetic Resonance (1H NMR) Spectroscopy and Gas Chromatography-Mass Spectrometry (GC-MS); IPM analysis was performed using plasma bead-based multiplex profiling. Comparisons were made using multivariate data statistical analysis. Fifty-three children were included (15 severe, 38 non-severe). Separation between severe and moderate appendicitis demonstrated excellent sensitivity and specificity (100%, 88%; 14 compounds), separation between severe and mild appendicitis also showed excellent sensitivity and specificity (91%, 90%; 16 compounds). Biomarker patterns derived from metabolomics and IPM profiling are capable of distinguishing children with severe appendicitis from those with less severe disease. These findings provide an important first step towards developing non-invasive diagnostic tools for clinicians in early identification of children who are at a high risk of developing severe appendicitis.

1. Introduction

Appendicitis is the most common non-traumatic surgical emergency in children [1]. It has been estimated that appendicitis related health care cost surpasses $680 million per year in the United States [2]. Moreover, appendicitis has the second highest litigation rate in pediatric emergency medicine [3]. Pediatric patients who develop severe disease need substantial critical care resources and prolonged hospital stays in order to manage serious consequences, such as intra-abdominal sepsis, organ dysfunction and, at times, death. Although severe appendicitis can be life-threatening for children, only a limited number of studies are reported in this area. This small body of literature demonstrates a case fatality rate that is significantly higher in children less than 10 years of age compared to all other age groups [4,5]. Factors associated with appendicitis-related morbidity and mortality include sepsis, anesthesia- or surgically-related complications, pre-existing comorbidities, age and male sex [6].
Appendicitis in the pediatric population often has an atypical presentation, particularly in young children who may exhibit several extra-intestinal non-specific symptoms, such as crying, lethargy, and anorexia and a paucity of the characteristic signs [1,7,8]. Moreover, this age group often lacks the developmental capacity to properly describe their symptoms and pain. These challenges result in delayed and/or missed diagnosis, which can increase perforation rate by 50% [1]. While there are many studies related to pediatric appendicitis diagnostic strategies including clinical scoring systems (e.g., Alvarado Score, Pediatric Appendicitis Score) [9,10,11], traditional laboratory tests of infection and inflammation (i.e., white blood cell count, erythrocyte sedimentation rate, C-reactive protein) [12,13,14,15] and imaging studies [16,17,18,19], there are obvious knowledge gaps related to our ability to accurately diagnose appendicitis and predict disease progression. What remains unknown is why some children rapidly progress to perforation and subsequent severe conditions such as sepsis and organ dysfunction while others remain physiologically stable throughout the course of the illness.
A recent focus on bio-profiling through Nuclear Magnetic Resonance (NMR) spectroscopy and/or Mass Spectrometry (MS) [20,21] and cytokine analyses [22,23] has identified biomarker panels that accurately separate specific populations and point to potential target pathways allowing for innovation in therapeutic options. NMR and MS platforms are used to detect various metabolites, the low molecular weight compounds which characterize the intermediate and final products of the metabolic pathways within an organism. The NMR technique is quantitative and highly reproducible [24]; however, NMR has the drawback of low sensitivity. By using NMR in combination with more sensitive methodologies such as MS, it is possible to quantitatively detect a wide range of metabolites. Similarly, using multiplexed cytokine/chemokine analysis, it is possible to simultaneously profile numerous inflammatory protein mediators and obtain a better understanding of the host immune response under differing physiological conditions [22,23]. These analytical approaches have also empowered the comparison of metabolic and protein mediator profiles, thus identifying potential biomarkers for disease diagnosis and prognosis. In addition, requirement of a small sample volume makes this approach quite convenient for disease diagnosis. In previous studies, we have combined metabolomics and protein mediator profiling data for differentiating pediatric appendicitis patients from pediatric controls (patients with no abdominal pain), pediatric appendicitis patients from non-appendicitis patients presenting with similar abdominal pain, and children with simple appendicitis from those with perforated appendicitis [25,26,27]. In the present study, we have applied a similar integrated profiling approach for identifying differences in the bio-profiles of children with appendicitis who require critical care management compared to those who do not. We anticipate that this approach will (1) allow clinicians to accurately predict disease severity and provide early escalation of care to those at risk of severe disease, and (2) provide a better understanding of the underlying pathophysiology.

2. Results

2.1. Sample Description

This study evaluated samples from 53 pediatric patients (15 ‘severe appendicitis’, 9 ‘moderate appendicitis’ and 29 ‘mild appendicitis’). Table 1 provides demographic and clinical characteristics of the patient cohorts. A total of 86 known metabolites by 1H NMR and GC-MS, and 57 protein mediators by Luminex techniques could be consistently detected and analyzed for each study subject.

2.2. Metabolic Profile

The common metabolites that were detected from 1H NMR and GC-MS platforms illustrated similar trends in different patient groups. Three principal components (PC1, PC2 and PC3) were calculated to build the unsupervised PCA model for the metabolic dataset (1H NMR and GC-MS) contributing 16.9%, 10.1% and 6.6% of variation, respectively (Figure 1a). Three outliers (two mild and one severe appendicitis) were identified and excluded from further downstream analysis. Supervised OPLS-DA models consisted of only two classes. The severe appendicitis group was separately compared with moderate and mild appendicitis groups (Figure 2a and Figure 3a). The score scatterplots for each of the OPLS-DA analyses illustrate substantial separation of the two groups with strong variance and predictive ability (Figure 2a and Figure 3a, Table 2). The R2Y:Q2 ratio for the model comparing severe and moderate groups (0.90:0.70) is higher compared to severe vs. mild groups (0.60:0.48); both models show excellent significance in statistical parameters including sensitivity and specificity (severe vs. moderate—1.00:0.86, severe vs. mild—0.71:0.89, respectively), CV-ANOVA p-values, and AUROC values (Table 2).

2.3. Inflammatory Protein Mediator Profile

The unsupervised PCA model for the inflammatory protein mediator dataset was built based on three principal components (PC1, PC2 and PC3), contributing to 18.8%, 13.1% and 10.2% of the variation, respectively (Figure 1b). Two outliers were identified (one mild and one severe appendicitis) and excluded from subsequent analysis. Supervised OPLS-DA methodology was applied to compare the variances between each pair of case-control groups: severe vs. moderate (Figure 2b) and severe vs. mild (Figure 3b). Score scatterplots for the OPLS-DA analysis demonstrate separation of each of the two classes having considerably strong variance and predictive values (severe vs. moderate: R2Y = 0.66, Q2 = 0.56; severe vs. mild: R2Y = 0.63, Q2 = 0.55, Figure 2b and Figure 3b and Table 2). The strength of the models is supported by the sensitivity and specificity (severe vs. moderate—0.92:0.80, severe vs. mild—0.92:0.90, respectively), AUROC and CV-ANOVA p-values (Table 2).

2.4. Integrated Metabolic and Protein Mediator Profile

A total of 16 metabolites and 15 inflammatory protein mediators provided the most significant contribution for the separation between severe and control patient groups in the corresponding OPLS-DA models. These 31 compounds were combined to develop an integrated metabolic and inflammatory protein mediator dataset. The unsupervised PCA model for the integrated dataset was based on three principal components: PC1 = 24.6%, PC2 = 13.7% and PC3 = 9.6% (Figure 1c). Two outliers from the ‘severe’ group were identified and excluded from the supervised analysis. Supervised OPLS-DA models were built to illustrate the differences between case and control patient classes. Score scatterplots demonstrated excellent separation of each of the two pairs of groups with strong R2Y and Q2 values (severe vs. moderate—0.75:0.65; severe vs. mild—0.65:0.57, Figure 2c and Figure 3c and Table 2). The combined dataset analysis also revealed high sensitivity and specificity (severe vs. moderate—1.00:0.88, severe vs. mild—0.91:0.90, respectively), and AUROC values for the models (Table 2). A bio-pattern comprised 5 serum metabolites and 9 inflammatory protein mediators that contributed significantly (p < 0.05) for the separation between pediatric patients with severe and moderate appendicitis (Figure 4a). Similarly, a bio-pattern constituting of 7 significant metabolites and 9 inflammatory protein mediators was detected for the distinction between the severe and mild appendicitis groups (Figure 4b).
To confirm that the exclusion of outliers from the supervised analyses did not bias our results, we re-developed the OPLS-DA models including all outliers, for all three datasets. The exclusion had negligible influence on the discriminative and predictive ability of the models (Table 3).

3. Discussion

This study has provided us with a novel opportunity to explore biochemical processes underlying the most severe presentations of appendicitis. We believe our study is unique in its use of an expansive, multi-modal biomarker profiling approach. Analyses were performed through advanced quantitative techniques capable of analyzing hundreds of different metabolites and inflammatory protein mediators with high sensitivity. From this arsenal of biological compounds, we identified biomarker fingerprints that distinguished severe appendicitis from mild and moderate appendicitis for the children included in our study. Our analyses demonstrated promising results separating the cohorts of different severity with strong predictive values, sensitivity, specificity and AUROC. The excellent performance of our models suggests that the identified biomarker patterns distinguishing severe appendicitis from moderate and mild presentations of the disease may have clinical significance.
Several significant metabolites in the bio-profiles were identified as key in separating the children with severe appendicitis from those with mild and moderate appendicitis. This finding implies that significant disruption of multiple metabolic processes underlies the development of severe disease pediatric patients with appendicitis. The depleted level of 2-hydroxybutanoate and increased o-acetylcarnitine indicate higher fatty acid degradation during severe appendicitis [28,29,30,31] compared to the moderate condition (Figure 4). Similarly, comparison with the mild disease group revealed increased levels of isopropanol and o-acetylcarnitine along with reduced levels of octadecadienoic acid in the severe group indicating elevated fatty acid degradation. Our findings suggest that fatty acid breakdown is induced during appendicitis, which elevates with the severity of the disease. However, whether this fatty acid manifestation is the cause or effect of appendicitis and its severity, is not clear. Another pathway, glycine, serine and threonine metabolism, is markedly disrupted during severe appendicitis as implied by the presence of a number of metabolites from this pathway in the detected biopatterns. For instance, the metabolites threonine, glyceric acid, glycine and serine detected in the bio-profile of severe vs. mild, and the metabolite serine found in severe vs. moderate analysis are essential parts of the glycine, serine and threonine metabolism pathway [28,29,30,31]. Depletion of glutamine in both bio-profiles indicates that one or more of the pathways, such as arginine biosynthesis, purine metabolism, pyrimidine metabolism, and alanine, aspartate and glutamate metabolism may be affected because glutamine is an essential component of all these pathways [28].
Appendicitis is characterized by an excessive inflammatory response [32], which is also reflected by the identification of a notable number of inflammatory mediators identified in our biomarker patterns. The levels of inflammatory markers procalcitonin, ferritin, interleukin-10 (IL-10), monocyte chemoattractant protein-1 (MCP-1), IL-8 and hepatocyte growth factor (HGF) were higher whereas the concentrations of signaling lymphocyte activation molecule (SLAM)-associated protein (SAP) and tissue plasminogen activator (tPA) were reduced in the children with severe appendicitis compared to children with moderate and mild appendicitis (Figure 4).
Many of these inflammatory biomarkers are consistent with previous studies. For instance, as an essential strategy of the host immune response, ferritin is almost always elevated during infection to chelate iron and deprives the invading pathogen from this essential nutrient [33,34]. We have also identified elevated ferritin in our previous studies that compared appendicitis with non-appendicitis, and perforated with simple appendicitis, which suggests that with increasing severity of appendicitis inflammation, the concentration of ferritin also increases in the host blood. We have identified elevated levels of IL-10 in the severe group compared to both control groups. In addition to be an important inflammatory biomarker for acute appendicitis [35,36], IL-10 has also been found in other studies to be increased during severe appendicitis [37,38]. The concentration of IL-8 has been elevated during acute appendicitis [32,39]; detection of this interleukin in our study implies that it’s level also increases with the increasing complexity of the disease. Similar to our finding, previous studies reported that procalcitonin is an indicator of the severity of acute appendicitis [4]. However, in contrast to our finding, the cytokine IL-4 has been found to be elevated in mild appendicitis when compared with the more severe form [38].
Overall, the metabolites and protein mediators detected from pediatric patients provide distinctive biomarker patterns to distinguish severe and non-severe appendicitis. Although various studies in the literature focused on selected inflammatory biomarkers of appendicitis, gaps in timely and accurate diagnosis and prediction of disease progression continue to exist. There is very little literature that reports on why some children develop life-threatening appendicitis, and there are still knowledge gaps about why some children with perforated appendicitis occasionally develop serious complications while most of the others respond well to ED and surgical management. Our study is an important initial step towards answering this question.
The presence of a small number of compounds in the detected bio-patterns should make it possible to develop a time-sensitive point of care aid for the emergency pediatric physicians to identify children with risk of developing severe appendicitis. A similar type of biomarker phenotyping approach using integrated metabolomics and protein mediator profiling has shown promising outcomes in early diagnosis and prognosis of sepsis [40,41]. Hence, it is rational to suggest that the promising findings from our study have also the potential to improve the prediction of disease progression in children with appendicitis.
Our results demonstrated that the separation between severe and moderate conditions were somewhat better than that between severe and mild conditions. It is possible that children with mild appendicitis have yet to establish a common set of metabolic derangements that converge on a common profile which would be identified as different from severe disease. In addition, the differences in sample sizes of the cohorts might explain this outcome [42,43]. In an attempt to minimize statistical bias [42,43], we maintained a ratio close to 1:2 during our analyses. Another limitation is that while blood samples were collected from the majority of the children before antibiotic treatment, some children had their samples drawn in PICU (albeit at the earliest time point achievable) due to severity of the condition. These children would have undergone additional resuscitation efforts including antibiotics, fluids, inotropes etc. Finally, due to the small sample size of this investigation, follow-up studies will be needed to confirm our findings across ages.

4. Materials and Methods

4.1. Study Design, Setting and Population

We performed a post hoc sub-group analysis of children aged 0–17 years who had been prospectively enrolled in the Alberta Sepsis Network (ASN) study and who had pathologically proven appendicitis. The ASN study was investigating the metabolic and inflammatory processes in children managed in the Emergency Department (ED) or the Pediatric Intensive Care Unit (PICU) for an infectious illness [44]. Children were eligible for enrolment into the ASN study if they met criteria for systemic inflammatory response syndrome, had a blood culture performed and antibiotics ordered. Those not expected to survive ≥24 h, refusing intubation or vasoactive infusions (e.g., palliative care), or already having had severe sepsis for ≥48 h (defined as sepsis with cardiovascular dysfunction, acute respiratory distress syndrome, or two other organ dysfunctions) were excluded. Enrolment occurred at the two pediatric tertiary hospitals in the province of Alberta, Canada.
In this current analysis, cases (severe cohort) were defined as children requiring PICU admission due to the severity of appendicitis, following consultation between the most responsible clinician (Emergency, Surgery) and the PICU attending clinician. Factors associated with a decision for PICU admission include, but are not limited to: requirement for significant fluid resuscitation, requirement of vasoactive support, initiation of mechanical ventilation (including both invasive and non-invasive). The primary control group (moderate cohort) included those with appendiceal perforation but without PICU admission. The secondary control group (mild cohort) included children with positive pathology for appendicitis but neither appendiceal perforation nor PICU admission. The diagnosis of appendicitis was defined by either (a) pathology report indicating any evidence of acute inflammation of the appendix, or (b) surgical placement of a percutaneous drain (for abscess collection related to the appendix) with interval appendectomy. Perforation of the appendix was defined by (a) pathology report indicating any level of perforation/rupture/disruption of the wall of the appendix, or (b) placement of a percutaneous drain for an abscess collection related to the appendix.
Approval for this study was obtained from the Health Research Ethics Board of the University of Alberta and the Conjoint Health Research Ethics Board of the University of Calgary. Children with severe life-threatening disease requiring active resuscitation and PICU admission were enrolled using a deferred consent model; for children with moderate and mild disease, informed consent/assent to participate in the study was obtained from the child and/or their parent/caregiver.

4.2. Sample Collection

Blood samples were obtained from children at the earliest possible time following enrolment and within 24 h after admission, whether that was in the PICU or ED. The volume of blood obtained was as follows: 2 mL from infants (0–<1 years), 4 mL from preschool and school-aged (1–12 years), and 8 mL from teenagers (13–17 years). For metabolic profiling, whole blood was collected into serum (red top) tubes. Immediately after collection, blood samples were gently inverted several times and allowed to clot at room temperature for an hour. The samples were then placed on ice for 2 h. The samples were centrifuged at 1200× g for 10 min at 4 °C and serum was then collected in a 4 mL cryovial and stored at −70 °C until distribution for metabolomics analysis. For inflammatory mediator profiling, whole blood was collected into a sodium-heparin plasma tube. The tubes were then gently inverted several times, and immediately placed on ice. The samples were centrifuged at 1200× g for 10 min at 4 °C in a swinging bucket centrifuge and the separated plasma was collected in a 4 mL cryovial. The plasma samples were stored at −70 °C until distribution for multiplex analysis.

4.3. Proton Nuclear Magnetic Resonance (1H NMR) Spectroscopy

Samples were prepared for the acquisition of 1H NMR spectra following a procedure described in earlier studies [44,45]. Briefly, serum samples were filtered using 3 kDa NanoSep microcentrifuge filters to remove large molecules. Sodium azide and D2O were added to the filtrate and the final pH of each sample was adjusted to 7.00 ± 0.04 by using phosphate buffer. The prepared samples were analyzed on a 600 MHz Bruker Ultrashield Plus NMR spectrometer (Bruker BioSpin Ltd., Milton, ON, Canada) to obtain 1H NMR spectra using the standard Bruker 1D spectroscopy pulse program ‘noesypr1d’ [46,47]. Each spectrum was manually processed for phasing, baseline correction, and referencing to the DSS peak at 0.0 ppm using the Chenomx NMR Suite 7.5 software (Chenomx Inc., Edmonton, Canada). The same software was used for profiling of each sample spectrum, which implicates identification of different metabolites by their distinctive spectral signature using an external metabolite reference database [46,47].

4.4. Gas Chromatography-Mass Spectrometry (GC-MS)

A previously described protocol was followed for sample extraction and derivatization for GC-MS21. In brief, metabolites from serum were extracted using a two-phase chloroform and methanol mixture. The aqueous layer was then separated and vacuum-dried by SpeedVac (Eppendorf, Germany). The metabolites were subsequently derivatized by methoxyamine hydrochloride in pyridine solution and N-methyl-N-(trimethylsilyl) trifluoroacetamide. An internal standard (phenylalanine D5) was used for acquiring the relative concentrations of the analytes. Mass spectra were obtained using Agilent chromatograph 7890A (Agilent Technologies Canada, Inc, Mississauga, ON, Canada) coupled with a Time-of-Flight Mass Spectrometer (Waters Corp., Milford, MA, USA); the MS range used for scanning was 50–800 m/z. The generated GC-MS spectra were then processed using the Metabolite Detector software (version 2.06, Technische Universität Carolo-Wilhelmina zu Braunschweig, Braunschweig, Germany) followed by identification of metabolites based on the GOLM metabolome database [48].

4.5. Inflammatory Mediator Profiling

The inflammatory mediators in the plasma samples were analyzed using three human cytokine and chemokine assay kits (Bio-Plex Pro Human Cytokine 21-Plex Assay, Bio-Plex Pro Human Cytokine 27-Plex Assay and Bio-Plex Pro Human Acute Phase Protein 5- + 4- Plex Assay) manufactured by Bio-Rad Laboratories, Inc. (Hercules, CA, USA). A Luminex 200 apparatus (Applied Cytometry Systems, Sheffield, UK) was used for reading the assay plates by following the manufacturer guidelines. Detection and analysis of the inflammatory protein mediators were done using the software Bio-Plex Manager 6.0 (Bio-Rad Laboratories, Inc.). Samples were run in duplicate and the data for a replicate were considered as a missing value if two replicates had a coefficient of variance greater than 20%.

4.6. Statistical Analysis

The SIMCA-P+ software (v12.0.1; Umetrics, Umeå, Sweden) was used for performing multivariate data statistical analysis of the metabolite and protein mediator profiles [42,43,49,50,51,52]. The common metabolites that were detected by both 1H NMR and GC-MS techniques were averaged and all the metabolites identified from the two platforms were combined to generate the metabolomics profile. The metabolomics and protein mediator datasets were first analyzed separately, then combined and re-analyzed. Metabolites and mediators having >20% missing values were excluded from the statistical analysis followed by data preprocessing (median fold change normalization, logarithmic transformation, centering, and unit variance scaling), for separate and combined datasets [52].
An initial overview of the multivariate datasets was obtained by unsupervised principal component analysis (PCA), which also identified outliers situated beyond of the 95% confidence interval [42]. The outliers were excluded from further analysis to prevent bias in the supervised models. Supervised orthogonal partial least squares discriminant analysis (OPLS-DA) was performed. Potentially relevant metabolites were selected using the variable importance to projection (VIP); those variables with VIP >1 were included in developing the OPLS-DA models. R2Y (the percentage of variation explained by the model), Q2 (the predictive ability of the model) and CV-ANOVA (Cross-Validated Analysis of Variance) p-value were calculated for the OPLS-DA models using a sevenfold cross-validation method. Regression coefficients for the OPLS-DA models were calculated from the combined dataset to detect the most critical compounds contributing to the class separations [51]. Those metabolites and inflammatory mediators which demonstrated significant differences in concentration (p < 0.05) between the classes were considered for biomarker fingerprint development. Clinically important parameters including sensitivity (Sn), specificity (Sp), and area under the receiver operating characteristic curve (AUROC) [53] were calculated for each OPLS-DA model from sample class prediction during sevenfold cross-validation (YpredCV) in the SIMCA-P+ software. Metz ROC (The University of Chicago, IL, USA) was used to calculate the AUROC values. MetaboAnalyst 3.0 was used to identify the metabolic pathways perturbed during severe appendicitis [28].

5. Conclusions

We have explored the metabolic and inflammatory processes underlying life-threatening presentations of appendicitis in children on a provincial level and obtained bio-profiles that separate the children with severe appendicitis from those with mild and moderate presentations of the disease. We have also identified key metabolic pathways that are disrupted during severe appendicitis. The results from this project have the potential to improve the diagnosis, management and outcomes of children presenting with appendicitis, in particular those with life-threatening illness. In addition, identifying pathways responsible for the progression of disease holds the potential for future innovation in therapeutics aimed at mitigating systemic inflammatory response accompanying severe disease, thus leading to improved patient outcomes. The promising pilot data from this study have been used by our team to inform a large multi-center follow-up study focused on validating the bio-profiles developed by metabolomics and protein mediator analyses. Additional future studies could ultimately result in the development of a highly-accurate, accessible, cost-effective, rapid (<1 h), minimally invasive (capillary finger-prick) point-of-care diagnostic test.

Author Contributions

Conceptualization, C.N.J., H.J.V., G.C.T.; methodology, N.S.S., C.N.J., H.J.V., G.C.T.; formal analysis, N.S.S., C.N.J., B.M., H.J.V.; investigation, J.B., A.R.J., D.-A.M., M.B., R.E., G.C.T.; resources, J.B., C.N.J., A.R.J., H.J.V., G.C.T.; data curation, J.B., A.R.J., D.-A.M., M.B., R.E., G.C.T.; writing—original draft preparation, N.S.S., H.J.V., C.N.J.; writing—review and editing, J.B., C.N.J., A.R.J., D.-A.M., B.M., M.B., R.E., H.J.V., G.C.T.; supervision, C.N.J., H.J.V., G.C.T.; project administration, D.-A.M.; funding acquisition, J.B., C.N.J., A.R.J., G.C.T. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by the Alberta Health Services Critical Care Strategic Clinical Network Seed Grant and Alberta Innovates—Health Solutions Team Grant awarded to the Alberta Sepsis Network.

Institutional Review Board Statement

The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Health Research Ethics Board of the University of Alberta and the Conjoint Health Research Ethics Board of the University of Calgary.

Informed Consent Statement

Children with severe life-threatening disease requiring active resuscitation and PICU admission were enrolled using a deferred consent model; for children with moderate and mild disease, informed consent/assent to participate in the study was obtained from the child and/or their parent/caregiver.

Data Availability Statement

The data presented in this study are available on request from the corresponding author. Data is not publicly available because we do not have ethics approval to publish our data set.

Acknowledgments

The authors wish to thank the research and clinical teams in the Pediatric Emergency Departments and Pediatric Intensive Care Units at the Alberta Children’s Hospital, Calgary AB, and Stollery Children’s Hospital, Edmonton AB.

Conflicts of Interest

The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

References

  1. Williams, N.; Kapila, L. Acute appendicitis in the under-5 year old. J. R. Cool Surg. Edinb. 1994, 39, 168–170. [Google Scholar]
  2. Guthery, S.L.; Hutchings, C.; Dean, J.M.; Hoff, C. National estimates of hospital utilization by children with gastrointestinal disorders: Analysis of the 1997 kids’ inpatient database. J. Pediatr. 2004, 144, 589–594. [Google Scholar] [CrossRef]
  3. Selbst, S.M.; Friedman, M.J.; Singh, S.B. Epidemiology and etiology of malpractice lawsuits involving children in US emergency departments and urgent care centers. Pediatr. Emerg. Care 2005, 21, 165–169. [Google Scholar]
  4. Andersson, R.E. Short and long-term mortality after appendectomy in Sweden 1987 to 2006. Influence of appendectomy diagnosis, sex, age, co-morbidity, surgical method, hospital volume, and time period. A national population-based cohort study. World J. Surg. 2013, 37, 974–981. [Google Scholar] [CrossRef]
  5. Blomqvist, P.G.; Andersson, R.E.; Granath, F.; Lambe, M.P.; Ekbom, A.R. Mortality after appendectomy in Sweden, 1987–1996. Ann. Surg. 2001, 233, 455–460. [Google Scholar] [CrossRef]
  6. Andersson, M.N.; Andersson, R.E. Causes of short-term mortality after appendectomy: A population-based case-controlled study. Ann. Surg. 2011, 254, 103–107. [Google Scholar] [CrossRef]
  7. Bundy, D.G.; Byerley, J.S.; Liles, E.A.; Perrin, E.M.; Katznelson, J.; Rice, H.E. Does this child have appendicitis? JAMA 2007, 298, 438–451. [Google Scholar] [CrossRef] [Green Version]
  8. Hennelly, K.E.; Bachur, R. Appendicitis update. Curr. Opin. Pediatr. 2011, 23, 281–285. [Google Scholar] [CrossRef] [PubMed]
  9. Rothrock, S.G.; Pagane, J. Acute appendicitis in children: Emergency department diagnosis and management. Ann. Emerg. Med. 2000, 36, 39–51. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  10. Vissers, R.J.; Lennarz, W.B. Pitfalls in appendicitis. Emerg. Med. Clin. North. Am. 2010, 28, 103–118. [Google Scholar] [CrossRef] [PubMed]
  11. Zouari, M.; Jallouli, M.; Louati, H.; Kchaou, R.; Chtourou, R.; Kotti, A.; Dhaou, M.B.; Zitouni, H.; Mhiri, R. Predictive value of C-reactive protein, ultrasound and Alvarado score in acute appendicitis: A prospective pediatric cohort. Am. J. Emerg. Med. 2016, 34, 189–192. [Google Scholar] [CrossRef]
  12. Abbas, M.H.; Choudhry, M.N.; Hamza, N.; Ali, B.; Amin, A.A.; Ammori, B.J. Admission levels of serum amyloid a and procalcitonin are more predictive of the diagnosis of acute appendicitis compared with C-reactive protein. Surg. Laparosc. Endosc. Percutaneous Tech. 2014, 24, 488–494. [Google Scholar] [CrossRef]
  13. Soomro, A.G.; Siddiqui, F.G.; Abro, A.H.; Abro, S.; Shaikh, N.A.; Memon, A.S. Diagnostic accuracy of Alvarado scoring system in acute appendicitis. J. Liaquat. Univ. Med. Health Sci. 2008, 7, 93–96. [Google Scholar]
  14. Yegane, R.P.; Hajinasrollah, E.; Ahmadei, M. Evaluation of modified Alvarado score in acute appendicitis among Iranian patients. Acta Med. Iran. 2008, 46, 501–506. [Google Scholar]
  15. Curran, T.J.; Muenchow, S.K. The treatment of complicated appendicitis in children using peritoneal drainage: Results from a public hospital. J. Pediatr. Surg. 1993, 28, 204–208. [Google Scholar] [CrossRef]
  16. Acosta, R.; Crain, E.F.; Goldman, H.S. CT can reduce hospitalization for observation in children with suspected appendicitis. Pediatr. Radiol. 2005, 35, 495–500. [Google Scholar] [CrossRef]
  17. Andersson, R.E. Meta-analysis of the clinical and laboratory diagnosis of appendicitis. Br. J. Surg. 2004, 91, 28–37. [Google Scholar] [CrossRef] [PubMed]
  18. Nance, M.L.; Adamson, W.T.; Hedrick, H.L. Appendicitis in the young child: A continuing diagnostic challenge. Pediatr. Emerg. Care 2000, 16, 160–162. [Google Scholar] [CrossRef] [PubMed]
  19. Rothrock, S.G.; Skeoch, G.; Rush, J.J.; Johnson, N.E. Clinical features of misdiagnosed appendicitis in children. Ann. Emerg. Med. 1991, 20, 45–50. [Google Scholar] [CrossRef]
  20. Mickiewicz, B.; Duggan, G.E.; Winston, B.W.; Doig, C.; Kubes, P.; Vogel, H.J. Metabolic profiling of serum samples by 1H nuclear magnetic resonance spectroscopy as a potential diagnostic approach for septic shock. Crit. Care Med. 2014, 42, 1140–1149. [Google Scholar] [CrossRef] [PubMed]
  21. Mickiewicz, B.; Kelly, J.J.; Ludwig, T.E.; Weljie, A.M.; Wiley, J.P.; Schmidt, T.A.; Vogel, H.J. Metabolic analysis of knee synovial fluid as a potential diagnostic approach for osteoarthritis. J. Orthop. Res. 2015, 33, 1631–1638. [Google Scholar] [CrossRef] [PubMed]
  22. Roberts, D.J.; Jenne, C.N.; Léger, C.; Kramer, A.H.; Gallagher, C.N.; Todd, S.; Parney, I.F.; Doig, C.J.; Yong, V.W.; Kubes, P.; et al. A prospective evaluation of the temporal matrix metalloproteinase response after severe traumatic brain injury in humans. J. Neurotrauma 2013, 30, 1717–1726. [Google Scholar] [CrossRef] [PubMed]
  23. Roberts, D.J.; Jenne, C.N.; Léger, C.; Kramer, A.H.; Gallagher, C.N.; Todd, S.; Parney, I.F.; Doig, C.J.; Yong, V.W.; Kubes, P.; et al. Association between the cerebral inflammatory and matrix metalloproteinase responses after severe traumatic brain injury in humans. J. Neurotrauma 2013, 30, 1727–1736. [Google Scholar] [CrossRef]
  24. Soininen, P.; Kangas, A.J.; Würtz, P.; Tukiainen, T.; Tynkkynen, T.; Laatikainen, R.; Järvelin, M.R.; Kähönen, M.; Lehtimäki, T.; Viikari, J.; et al. High-throughput serum NMR metabonomics for cost-effective holistic studies on systemic metabolism. Analyst 2009, 134, 1781–1785. [Google Scholar] [CrossRef]
  25. Shommu, N.S.; Jenne, C.N.; Blackwood, J.; Joffe, A.R.; Martin, D.A.; Thompson, G.C.; Vogel, H.J. Metabolomic and inflammatory mediator based biomarker profiling as a potential novel method to aid pediatric appendicitis identification. PLoS ONE 2018, 13, e0193563. [Google Scholar] [CrossRef]
  26. Shommu, N.S.; Jenne, C.N.; Blackwood, J.; Martin, D.A.; Joffe, A.R.; Eccles, R.; Brindle, M.; Khanafer, I.; Vogel, H.J.; Thompson, G.C. The use of metabolomic and inflammatory mediator profiling provides a novel approach to identifying pediatric appendicitis in the emergency department. Sci. Rep. 2018, 8, 4083. [Google Scholar] [CrossRef]
  27. Naqvi, S.A.; Thompson, G.C.; Joffe, A.R.; Blackwood, J.; Martin, D.A.; Brindle, M.; Barkema, H.W.; Jenne, C.N. Cytokines and chemokines in pediatric appendicitis: A multiplex analysis of inflammatory protein mediators. Mediat. Inflamm. 2019, 2019. [Google Scholar] [CrossRef] [PubMed]
  28. Xia, J.; Sinelnikov, I.V.; Han, B.; Wishart, D.S. MetaboAnalyst 3.0—making metabolomics more meaningful. Nucleic Acids Res. 2015, 43, W251–W257. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  29. Kanehisa, M.; Goto, S. KEGG: Kyoto encyclopedia of genes and genomes. Nucleic Acids Res. 2000, 28, 27–30. [Google Scholar] [CrossRef] [PubMed]
  30. Kanehisa, M.; Goto, S.; Sato, Y.; Kawashima, M.; Furumichi, M.; Tanabe, M. Data, information, knowledge and principle: Back to metabolism in KEGG. Nucleic Acids Res. 2014, 42, D199–D205. [Google Scholar] [CrossRef] [Green Version]
  31. Wishart, D.S.; Knox, C.; Guo, A.C.; Eisner, R.; Young, N.; Gautam, B.; Hau, D.D.; Psychogios, N.; Dong, E.; Bouatra, S.; et al. HMDB: A knowledgebase for the human metabolome. Nucleic Acids Res. 2009, 37, D603–D610. [Google Scholar] [CrossRef]
  32. Arlt, A.; Bharti, R.; Ilves, I.; Häsler, R.; Miettinen, P.; Paajanen, H.; Brunke, G.; Ellrichmann, M.; Rehman, A.; Hauser, C.; et al. Characteristic changes in microbial community composition and expression of innate immune genes in acute appendicitis. Innate Immun. 2015, 21, 30–41. [Google Scholar] [CrossRef]
  33. Ong, D.S.T.; Wang, L.; Zhu, Y.; Ho, B.; Ding, J.L. The response of ferritin to LPS and acute phase of Pseudomonas infection. J. Endotoxin Res. 2005, 11, 267–280. [Google Scholar] [CrossRef] [Green Version]
  34. Orino, K.; Watanabe, K. Molecular, physiological and clinical aspects of the iron storage protein ferritin. Vet. J. 2008, 178, 191–201. [Google Scholar] [CrossRef] [PubMed]
  35. Hachim, M.Y.; Ahmed, A.H. The role of the cytokines and cell-adhesion molecules on the immunopathology of acute appendicitis. Saudi Med. J. 2006, 27, 1815–1821. [Google Scholar] [PubMed]
  36. Yildirim, O.; Solak, C.; Koçer, B.; Ünal, B.; Karabeyoğlu, M.; Bozkurt, B.; Aksaray, S.; Cengiz, Ö. The role of serum inflammatory markers in acute appendicitis and their success in preventing negative laparotomy. J. Investig. Surg. 2006, 19, 345–352. [Google Scholar] [CrossRef] [PubMed]
  37. Ruber, M.; Berg, A.; Ekerfelt, C.; Olaison, G.; Andersson, R.E. Different cytokine profiles in patients with a history of gangrenous or phlegmonous appendicitis. Clin. Exp. Immunol. 2006, 143, 117–124. [Google Scholar] [CrossRef] [PubMed]
  38. Rivera-Chavez, F.A.; Wheeler, H.; Lindberg, G.; Munford, R.S.; O’Keefe, G.E. Regional and systemic cytokine responses to acute inflammation of the vermiform appendix. Ann. Surg. 2003, 237, 408–416. [Google Scholar] [CrossRef] [PubMed]
  39. de Oliveira Machado, S.L.; Bagatini, M.D.; da Costa, P.; Baldissarelli, J.; Reichert, K.P.; de Oliveira, L.S.; Lemos, J.G.; Duarte, T.; Chitolina Schetinger, M.R.; Morsch, V.M. Evaluation of mediators of oxidative stress and inflammation in patients with acute appendicitis. Biomarkers 2016, 21, 530–537. [Google Scholar] [CrossRef]
  40. Mickiewicz, B.; Holinski, P.; Vogel, H.J.; Winston, B.W.; Wong, J.; Thompson, G.; Blackwood, J.; Joffe, A.R. Metabolomic profiling of serum samples using NMR spectroscopy to diagnose pediatric sepsis in the emergency department and intensive care unit. Pediatric Crit. Care Med. 2014, 15, 146. [Google Scholar] [CrossRef]
  41. Mickiewicz, B.; Thompson, G.C.; Blackwood, J.; Jenne, C.N.; Winston, B.W.; Vogel, H.J.; Joffe, A.R. Biomarker Phenotype for Early Diagnosis and Triage of Sepsis to the Pediatric Intensive Care Unit. Sci. Rep. 2018, 8, 16606. [Google Scholar] [CrossRef]
  42. Eriksson, L.; Byrne, T.; Johansson, E.; Trygg, J.; Vikström, C. Multi- and megavariate data analysis. In Part I: Basic Principles and Applications; Umetrics Academy: Umea, Sweden, 2006; Volume 425. [Google Scholar]
  43. Eriksson, L. Multi- and megavariate data analysis. In Part II: Advanced Applications and Method Extensions; Umetrics Academy: Umea, Sweden, 2006; Volume 307. [Google Scholar]
  44. Mickiewicz, B.; Thompson, G.C.; Blackwood, J.; Jenne, C.N.; Winston, B.W.; Vogel, H.J.; Joffe, A.R. Development of metabolic and inflammatory mediator biomarker phenotyping for early diagnosis and triage of pediatric sepsis. Crit. Care 2015, 19, 320. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  45. Mickiewicz, B.; Tam, P.; Jenne, C.N.; Leger, C.; Wong, J.; Winston, B.W.; Doig, C.; Kubes, P.; Vogel, H.J. Integration of metabolic and inflammatory mediator profiles as a potential prognostic approach for septic shock in the intensive care unit. Crit. Care 2015, 19, 11. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  46. Nicholson, J.K.; Foxall, P.J.; Spraul, M.; Farrant, R.D.; Lindon, J.C. 750 MHz 1H and 1H-13C NMR spectroscopy of human blood plasma. Anal. Chem. 1995, 67, 793–811. [Google Scholar] [CrossRef]
  47. Weljie, A.M.; Newton, J.; Mercier, P.; Carlson, E.; Slupsky, C.M. Targeted profiling: Quantitative analysis of 1H NMR metabolomics data. Anal. Chem. 2006, 78, 4430–4442. [Google Scholar] [CrossRef]
  48. Hummel, J.S.; Walther, D.; Kopka, J. The Golm Metabolome Database: A database for GC-MS based metabolite profiling. In Metabolomics; Nielsen, J., Jewett, M.C., Eds.; Springer: Berlin/Heidelberg, Germany, 2007; Volume 18, pp. 75–95. [Google Scholar]
  49. Madsen, R.; Lundstedt, T.; Trygg, J. Chemometrics in metabolomics--a review in human disease diagnosis. Anal. Chim. Acta 2010, 659, 23–33. [Google Scholar] [CrossRef]
  50. Nicholson, J.K.; Lindon, J.C.; Holmes, E. Metabonomics: Understanding the metabolic responses of living systems to pathophysiological stimuli via multivariate statistical analysis of biological NMR spectroscopic data. Xenobiotica 1999, 29, 1181–1189. [Google Scholar] [CrossRef]
  51. Trygg, J.; Holmes, E.; Lundstedt, T. Chemometrics in metabonomics. J. Proteome Res. 2007, 6, 469–479. [Google Scholar] [CrossRef]
  52. van den Berg, R.A.; Hoefsloot, H.C.; Westerhuis, J.A.; Smilde, A.K.; van der Werf, M.J. Centering, scaling, and transformations: Improving the biological information content of metabolomics data. BMC Genom. 2006, 7, 142. [Google Scholar] [CrossRef] [Green Version]
  53. Metz, C.E. Basic principles of ROC analysis. Semin. Nucl. Med. 1978, 8, 283–298. [Google Scholar] [CrossRef]
Metabolites 11 00664 g001 550
Figure 1. Three-dimensional PCA-X score scatter plots of pediatric patients with appendicitis using (a) metabolic profile, (b) inflammatory mediator profile, (c) combined metabolic and inflammatory mediator profile. Each plot is based on three principal components - PC1, PC2 and PC3. Red diamonds represent distribution of pediatric patients with severe appendicitis, blue diamonds represent patients with moderate appendicitis and green diamonds represent patients with mild appendicitis.
Figure 1. Three-dimensional PCA-X score scatter plots of pediatric patients with appendicitis using (a) metabolic profile, (b) inflammatory mediator profile, (c) combined metabolic and inflammatory mediator profile. Each plot is based on three principal components - PC1, PC2 and PC3. Red diamonds represent distribution of pediatric patients with severe appendicitis, blue diamonds represent patients with moderate appendicitis and green diamonds represent patients with mild appendicitis.
Metabolites 11 00664 g001
Metabolites 11 00664 g002 550
Figure 2. OPLS-DA score scatter plots distinguishing pediatric patients with severe appendicitis and moderate appendicitis using (a) metabolic profile (R2Y = 0.90, Q2 = 0.70), (b) inflammatory mediator profile (R2Y = 0.66, Q2 = 0.56) and (c) combined metabolic and inflammatory mediator profile (R2Y = 0.78, Q2 = 0.65). Red diamonds represent distribution of pediatric patients with severe appendicitis, and blue diamonds represent patients with moderate appendicitis. OPLS—Orthogonal Partial Least Squares; PLS—Partial Least Squares.
Figure 2. OPLS-DA score scatter plots distinguishing pediatric patients with severe appendicitis and moderate appendicitis using (a) metabolic profile (R2Y = 0.90, Q2 = 0.70), (b) inflammatory mediator profile (R2Y = 0.66, Q2 = 0.56) and (c) combined metabolic and inflammatory mediator profile (R2Y = 0.78, Q2 = 0.65). Red diamonds represent distribution of pediatric patients with severe appendicitis, and blue diamonds represent patients with moderate appendicitis. OPLS—Orthogonal Partial Least Squares; PLS—Partial Least Squares.
Metabolites 11 00664 g002
Metabolites 11 00664 g003 550
Figure 3. OPLS-DA score scatter plots distinguishing pediatric patients with severe and mild appendicitis using (a) metabolic profile (R2Y = 0.60, Q2 = 0.48), (b) inflammatory mediator profile (R2Y = 0.63, Q2 = 0.55), and (c) combined metabolic and inflammatory mediator profile (R2Y = 0.65, Q2 = 0.57). Red diamonds represent distribution of pediatric patients with severe appendicitis, and green diamonds represent patients with mild appendicitis. OPLS—Orthogonal Partial Least Squares; PLS—Partial Least Squares.
Figure 3. OPLS-DA score scatter plots distinguishing pediatric patients with severe and mild appendicitis using (a) metabolic profile (R2Y = 0.60, Q2 = 0.48), (b) inflammatory mediator profile (R2Y = 0.63, Q2 = 0.55), and (c) combined metabolic and inflammatory mediator profile (R2Y = 0.65, Q2 = 0.57). Red diamonds represent distribution of pediatric patients with severe appendicitis, and green diamonds represent patients with mild appendicitis. OPLS—Orthogonal Partial Least Squares; PLS—Partial Least Squares.
Metabolites 11 00664 g003
Metabolites 11 00664 g004 550
Figure 4. Regression coefficient plots for the statistically significant (p < 0.05) metabolites and inflammatory mediators distinguishing children with (a) severe vs. moderate appendicitis and (b) severe vs. mild appendicitis. Red bars with positive coefficient values represent increased concentrations in the severe appendicitis children, blue bars and green bars with negative values represent reduced concentration in the severe appendicitis patients, compared with moderate and mild appendicitis. IL—Interleukin, HGF—Hepatocyte growth factor, PCT—procalcitonin, MIF—macrophage migration inhibitory factor, MCP-1—Monocyte chemoattractant protein-1, SAP—Signaling lymphocyte activation molecule (SLAM)-associated protein and tPA—tissue plasminogen activator.
Figure 4. Regression coefficient plots for the statistically significant (p < 0.05) metabolites and inflammatory mediators distinguishing children with (a) severe vs. moderate appendicitis and (b) severe vs. mild appendicitis. Red bars with positive coefficient values represent increased concentrations in the severe appendicitis children, blue bars and green bars with negative values represent reduced concentration in the severe appendicitis patients, compared with moderate and mild appendicitis. IL—Interleukin, HGF—Hepatocyte growth factor, PCT—procalcitonin, MIF—macrophage migration inhibitory factor, MCP-1—Monocyte chemoattractant protein-1, SAP—Signaling lymphocyte activation molecule (SLAM)-associated protein and tPA—tissue plasminogen activator.
Metabolites 11 00664 g004
Table
Table 1. Demographics of children presenting with severe, moderate and mild appendicitis.
Table 1. Demographics of children presenting with severe, moderate and mild appendicitis.
Patient DemographicSevere AppendicitisModerate AppendicitisMild Appendicitis
(n = 15)(n = 9)(n = 29)
Age, median (IQR)10 (6–14)9 (8–14)11 (9–13)
Sex—Female; n (%)6 (40.0)6 (40.0)13 (44.8)
Documented RLQ Pain, n (%)12 (80.0)7 (77.8)23 (79.3)
Documented Nausea, n (%)6 (40.0)5 (55.6)19 (65.5)
Documented Vomiting, n (%)10 (66.7)5 (55.6)19 (65.5)
Documented Anorexia, n (%)12 (80.0)6 (66.7)17 (58.6)
Documented Fever in the ED, n (%)11 (73.3)5 (55.6)16 (55.2)
WBC in the ED; mean (SD)15.9 (2.1)19.8 (1.5)17.1 (1.2)
Neutrophils in the ED, mean (SD)13.0 (2.0)17.0 (1.5)14.2 (1.1)
US completed in the ED, n (%)10 (66.7)8 (88.9)29 (100.0)
CT completed in the ED, n (%)2 (13.3)1 (11.1)1 (3.5)
ED Disposition, n (%)
Surgery Ward0 (0.0)0 (0.0)5 (17.2)
OR10 (66.7)9 (100.0)24 (82.8)
PICU5 (33.3)0 (0.0)0 (0.0)
PRISM 3, median (IQR)3 (0–4)--
PELOD (Day 1), median (IQR)2 (1–11)--
Ventilator Days, median (IQR)0 (0–2)--
Inotrope Days, median (IQR)1 (0–1)--
PICU Length of Stay (days), median (IQR)3 (3–4)--
CT—Computed Tomography; ED—Emergency Department; IQR—Interquartile Range; OR—Operating Room; PELOD—Pediatric Logistic Organ Dysfunction Score; PICU—Pediatric Intensive Care Unit; PRISM—Pediatric Risk of Mortality Score; RLQ—Right Lower Quadrant; SD—standard deviation; US—Ultrasound; WBC—White Blood Cell.
Table
Table 2. Summary statistics from OPLS-DA models differentiating each pair of pediatric patient groups.
Table 2. Summary statistics from OPLS-DA models differentiating each pair of pediatric patient groups.
Analysis Sn:SpAUROCCV-ANOVA
p-Value		R2Y:Q2
Severe vs. moderate
appendicitis		Metabolic profile1.00:0.861.00 ± 0.001.3 × 10−40.90:0.70
Inflammatory mediator profile0.92:0.800.92 ± 0.089.9 × 10−40.66:0.56
Combined profile1.00:0.881.00 ± 0.002.4 × 10−40.78:0.65
Severe vs. mild
appendicitis		Metabolic profile0.71:0.890.93 ± 0.044.8 × 10−60.60:0.48
Inflammatory mediator profile0.92:0.900.96 ± 0.039.3 × 10−60.63:0.55
Combined profile0.91:0.900.97 ± 0.035.1 × 10−60.65:0.57
Sn—Sensitivity, Sp—Specificity, AUROC—Area Under the Receiver Operating Characteristic curve, CV-ANOVA—Cross Validated Analysis of Variance, SD—Standard Deviation.
Table
Table 3. Comparison of R2Y:Q2 values calculated for the OPLS-DA models with and without outliers.
Table 3. Comparison of R2Y:Q2 values calculated for the OPLS-DA models with and without outliers.
Analysis R2Y:Q2
Outliers Excluded		R2Y:Q2
Outliers included
Metabolic profile0.90:0.700.83:0.70
Severe vs. moderate appendicitisInflammatory mediator profile0.66:0.560.61:0.53
Combined profile0.78:0.650.73:0.59
Severe vs. mild appendicitisMetabolic profile Inflammatory mediator profile
Combined profile		0.60:0.480.63:0.40
0.63:0.550.62:0.55
0.65:0.570.63:0.56
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Share and Cite

MDPI and ACS Style

Shommu, N.S.; Blackwood, J.; Jenne, C.N.; Joffe, A.R.; Martin, D.-A.; Mickiewicz, B.; Brindle, M.; Eccles, R.; Vogel, H.J.; Thompson, G.C.; et al. Metabolomics and Inflammatory Mediator Profiling for the Differentiation of Life-Threatening and Non-Severe Appendicitis in the Pediatric Population. Metabolites 2021, 11, 664. https://doi.org/10.3390/metabo11100664

AMA Style

Shommu NS, Blackwood J, Jenne CN, Joffe AR, Martin D-A, Mickiewicz B, Brindle M, Eccles R, Vogel HJ, Thompson GC, et al. Metabolomics and Inflammatory Mediator Profiling for the Differentiation of Life-Threatening and Non-Severe Appendicitis in the Pediatric Population. Metabolites. 2021; 11(10):664. https://doi.org/10.3390/metabo11100664

Chicago/Turabian Style

Shommu, Nusrat S., Jaime Blackwood, Craig N. Jenne, Ari R. Joffe, Dori-Ann Martin, Beata Mickiewicz, Mary Brindle, Robin Eccles, Hans J. Vogel, Graham C. Thompson, and et al. 2021. "Metabolomics and Inflammatory Mediator Profiling for the Differentiation of Life-Threatening and Non-Severe Appendicitis in the Pediatric Population" Metabolites 11, no. 10: 664. https://doi.org/10.3390/metabo11100664

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop