FormalPara Key Summary Points

Chronic pain assessment and post-treatment evaluation continues to be challenging for medical providers due to a lack of validated, objective tools to measure patient outcomes.

Validation of mechanistic pain biomarkers would allow clinicians to objectively identify abnormal biochemistry contributing to painful symptoms in patients.

The Foundation Pain Index (FPI) is a multi-biomarker assay derived from algorithmic analysis of abnormal urinary metabolites observed in a cross-sectional observational study.

FPI scores strongly segregate biomarker profiles of healthy control subjects and chronic pain patients and correlate with worsening clinical assessments of chronic pain.

This test provides novel, objective data that evaluates the role of biochemistry in chronic pain, which may pave the way for targeted, non-opioid therapeutic strategies.

Introduction

Pain is a subjective sensation with objective, biochemical underpinnings that continues to challenge healthcare providers and regulatory bodies worldwide. This is mainly due to the risks associated with opioid analgesics commonly used in chronic pain treatment. The relatively high efficacy of opioids for the management of acute pain has been a major driver in the evolution of the current opioid crisis, due mainly to well-documented risks such as tolerance, dependence, and addiction [1]. Unfortunately, a simple solution to the opioid crisis does not exist, as simply removing these addictive analgesics from the clinic has led to a host of additional challenging issues for the healthcare community. The lack of strong evidence for alternative medications to replace opioids in chronic pain adds fuel to this dilemma. Instead, a greater understanding of the biochemical mechanisms that underlie and drive the subjective sensation of chronic pain would allow for novel, personalized, non-opioid therapies that directly modulate pain pathways. Mechanistic pain biomarkers revealing biochemical derangement driving the development or worsening of pain may represent one potential solution to the opioid crisis. Such biomarkers would not only provide physicians with novel, objective data pertaining to the biochemical origin of pain, but could also directly indicate safe, non-opioid therapies aimed at correcting the biochemical dysfunction to alleviate pain and improve function. Biomarker monitoring would also allow for longitudinal evaluation of the efficacy of modulating therapies. This will also provide patients and healthcare providers with objective updates on the biochemical status in a non-invasive manner, which would certainly aid with treatment compliance.

In previous work, we established a high prevalence of abnormal biochemical function in a population of chronic pain patients. We proposed that identifying and correcting such abnormalities would provide a safe and novel treatment modality for the management of pain [2]. As a follow-up to this study, we introduce a novel, multi-biomarker assay, known as the Foundation Pain Index (FPI), which evaluates diverse biochemical pathways involved in the pathogenesis of chronic pain to provide clinicians with objective and reproducible measures of the biochemical, metabolic, and nutritional drivers of chronic pain. This assay with algorithmic analysis is comprised of 11 well-characterized urinary biomarkers associated with nerve health, chronic inflammation, oxidative stress, and neurotransmitter turnover [3,4,5,6,7,8,9,10]. Clinical interpretations of abnormal biomarker findings in the context of chronic pain are detailed in Table 1.

Table 1 Component biomarkers of the FPI and clinical interpretation in the context of chronic pain
Full size table

In the present study, we sought to validate the FPI as an indicator of abnormal biochemical function in a chronic pain population. We also sought to determine the discriminant validity by comparing FPI scores of chronic pain subjects to age- and sex-matched pain-free controls. We established the validity criterion of the FPI by demonstrating a significant correlation between FPI scores and standard clinical assessments of chronic pain (i.e., SF-36 survey). We also discuss in detail the face validity of this innovative test offering. This study provides valuable insight into the validity and utility of the FPI in a clinical setting.

Methods

Study Subjects

A total of 153 chronic pain patients were enrolled in the ERD2019-02 prospective, multi-center clinical study. This study (ERD2019-02) was reviewed and approved by the Quorum Institutional Review Board (QIRB) and was performed in accordance with the Helsinki Declaration of 1964 and its later amendments. All participants provided informed consent prior to their inclusion of the study. All subjects completed the SF-36 short-form survey and the Hospital Anxiety and Depression Scale (HADS) at the time of sample collection. A single urine sample was obtained from each subject following enrolment and completion of all study questionnaires. Samples were packaged and shipped to Ethos R&D (Newport, KY, USA) where they were accessioned, prepared, and analyzed according to standard operating procedures. The number of samples provided from pain clinics satisfied criteria of adequate power for data analysis (80%, P value < 0.05).

FPI Scoring Structure

Quantitative results for each of the 11 biomarkers (methylmalonic acid [MMA], homocysteine [HCYS], xanthurenic acid [XAN], 3-hydroxypropylmercapturic acid [3-HPMA], quinolinic acid [QA], kynurenic acid [KA], pyroglutamic acid [PGA], ethylmalonate [EMA], hydroxymethylglutarate [HMG], vanilmandelate [VMA], and 5-hydroxyindoleacetic acid [5-HIAA]) are first normalized using creatinine concentrations and then analyzed via our proprietary algorithm to generate a single numerical score (FPI score) on a scale of 0–100, which describes the severity of abnormal findings. In the pre-validation model, FPI scores were subclassified into tiers that further describe the severity of the abnormal findings based on the severity and number of detected abnormalities (Table 2).

Table 2 Tiers used to categorize FPI scores
Full size table

Our working hypothesis is that higher FPI scores, which are represented by a greater number of biochemical abnormalities, are more likely to be contributing to painful symptoms. Patients who exhibit higher FPI scores are therefore more likely to benefit from correction therapy aimed at lowering their baseline FPI score. Conversely, low FPI scores indicate that biochemical function is unremarkable and less likely to be contributing to painful symptomologies. Patients who exhibit low FPI scores at baseline are unlikely to benefit from correction therapy. Low FPI scores indicate that abnormal biochemical function is less likely to be a causative factor in a patient’s pain but in no way invalidates a patient’s complaint of pain. Such a test could be used in combination with standard clinical assessments and could prove particularly useful in complex cases involving pain of unknown etiology. The assay may also be helpful in cohorts of patients with comorbidities (anxiety, depression, T2D), unresponsive to traditional treatments and poor candidates for opioid therapy.

Inclusion/Exclusion Criteria

Table 3 describes the inclusion/exclusion criteria used for subject enrolment during the ERD2019-02 clinical study.

Table 3 Inclusion/exclusion criteria for the ERD2019-02 study
Full size table

Pain-Free Control Samples

Control samples from individuals with no history of chronic pain or opioid use, were collected prospectively from Lee Biosolutions (Maryland Heights, MO, USA) and their associated clinical collection sites across the United States. All pain-free control samples were shipped frozen to Ethos R&D (Newport, KY, USA) where they were prepared and analyzed according to standard operating procedures.

Analytical Methodologies

Samples were accessioned into the Laboratory Information System (LIS) and frozen at − 20 °C on the date of receipt and remained frozen until the time they were thawed at room temperature and prepared for analysis. Each sample underwent only one freeze/thaw cycle. Samples were prepared on 96-well plates with four-point calibrators, two quality controls, and one negative sample per plate. The urinary biomarkers and creatinine were each analyzed in separate liquid chromatography tandem mass spectrometry (LC–MS/MS) assays. All analytes were quantified using isotopic dilution and analyzed with MassHunter software. Biomarker concentrations were corrected for urinary dilution using creatinine and normalized values reported as µg/mg creatinine. Samples from healthy and pain subjects were prepared and analyzed in tandem to eliminate any instrument or user bias during LC–MS/MS analysis.

FPI Algorithm and Score Derivation

With the intent to distinguish between chronic pain and pain-free populations, the FPI was developed to represent an algorithm based on binary logistic regression. Individual biomarker results that deviate from previously established normal concentrations collectively contribute to a decision-making process that sorts samples as either normal (pain free) or abnormal (pain). Various types of supervised machine learning techniques, including principal component analysis, clustering, and linear discriminant analysis (LDA) were compared among randomly selected pain-free and chronic pain samples. The LDA proved to provide the most robust segregation between groups (P value < 0.001). This model of multiple analysis of variance (MANOVA) with LDA generated weighted coefficients for each biomarker according to the function below:

$$Y\, = \,\beta_{0} \, + \,\beta_{1} *E\, + \,\beta_{2} \times F\, + \,\beta_{3} \times H\, + \,\beta_{4} \times I\, + \,\beta_{5} \times J\, + \,\beta_{6} \times K \ldots$$

Y refers to the model output, E–K… refer to metabolite concentrations as normalized to creatinine levels (µg/mg) in urine, β0 is the y-intercept, and β1–β6… are coefficients generated by analysis that most distinguishes control (pain free) and experimental (chronic pain) groups for each metabolite.

Biomarkers used in the development of the FPI model were transformed for normal distribution by Box-cox transformations (xλ, where λ = lambda, R 3.5.1 programming) and verified by normality analysis (D’Agostino and Pearson test, GraphPad Prism 8.3.0).

Statistical Methods

To compare univariate patient characteristics between groups, non-parametric t tests were used to compare means of variables. Multivariate analysis (least squares) were used to determine distinguishable power of the multi-biomarker score. To achieve normal distributions of data prior to MANOVA and correlation analyses, box-cox transformations (λ, lambda) and outlier analyses (determined by > 2.5 × standard deviation) were applied (Table 4). Biomarkers are expressed as normalized to creatinine concentrations (µg/mg) for each urine sample. Data was analyzed with the level of significance at α = 0.05.

Table 4 Lamba (λ) conversion table for biomarkers (x)
Full size table

Receiver operating characteristics (ROC) and its area under the curve (AUROC) were performed to determine sensitivity and specificity of FPI of pain-free vs. pain subjects and low vs. high FPI among pain patient SF-36 scores.

To account for non-normal distribution of clinical assessments among chronic pain patients, Spearman rank’s coefficient (r) was used. Correlation analysis was performed for group characteristics (age, sex, creatinine levels, biomarker concentrations) between pain-free and chronic pain subjects, as well as among chronic pain patients (biomarker levels, FPI, BMI, daily MME, current VAS, SF-36 score, HADS Anxiety and Depression scores). Components within the SF-36 used for analysis include assessments of physical functioning, limits due to physical health and emotional problems, energy/fatigue, emotional well-being, social functioning, pain, general health, and health change.

Results

Cohort Characteristics

FPI scores were calculated using creatinine-normalized biomarker results from 153 chronic pain patients meeting all inclusion criteria and 334 age- and sex-matched pain-free controls with no history of chronic pain or opioid use. The mean age of chronic pain subjects was 55.5 (55.5 ± 11.4) with a 52% female population. Mean BMI in the chronic pain cohort was 30.9 (30.9 ± 7.57) and 24.8% reported smoking cigarettes. Primary pain diagnoses across the pain cohort included, but were not limited to, low back pain, lumbar radiculopathy, chronic pain syndrome, and cervicalgia. The mean daily dose of opioid medication was 65.3 morphine milligram equivalents (MME) (65.3 ± 42.8). At the time of sample collection, the mean self-reported VAS score was 5.92 (5.92 ± 1.9), while the mean SF-36 score across the pain cohort was 1674 (1674 ± 667). Hospital anxiety and depression surveys revealed a mean anxiety score of 7.12 (7.12 ± 4.11) and a mean depression score of 7.43 (7.43 ± 12.05). Mean component scores from the SF-36 assessment are detailed further in Table 5. Age, sex, and creatinine levels did not correlate with biomarker concentrations or FPI scores among pain-free and pain subjects. Prescription medications, daily MME, current VAS, and HADS Anxiety and Depression scores did not correlate with biomarker concentrations or FPI scores in subjects with chronic pain.

Table 5 Characteristics of patients suffering from chronic pain
Full size table

FPI Performance and Validation

Abnormal biomarker results were detected in a large majority of the chronic pain cohort with significant deviation from normal ranges observed across most individual biomarkers (Table 6).

Table 6 Biomarker validation across patient and healthy cohorts
Full size table

The majority of pain subjects (86%) exhibited at least one abnormal biomarker (outside of the double-sided, 95% confidence interval reference range). Mean comparisons of individual biomarkers using non-parametric t testing revealed extremely strong discriminatory power (P < 0.0001) of five of the 11 biomarkers (methylmalonic acid, xanthurenic acid, pyroglutamate, kynurenic acid, and hydroxymethylglutarate) while an additional three biomarkers (homocysteine, quinolinic acid, and 3-HPMA) exhibited statistically significant discriminatory power (P < 0.05) between the pain and pain-free cohorts. Multi-biomarker statistical models (FPI) outperformed individual biomarkers when discriminating between pain-free and chronic pain cohorts (AUROC: 0.7490, P < 0.0001) (Fig. 1). Mean FPI scores were evaluated across pain-free and chronic pain cohorts and confirm a greater degree of pain-relevant biochemical perturbations in chronic pain subjects (Fig. 1). Mean FPI scores (0-100 scale) among the chronic pain cohort was 43.9 (43.9 ± 28.0) while the pain-free cohort exhibited a mean FPI score of 20.4 (20.4 ± 20.6) (P < 0.0001) (Fig. 2).

Fig. 1
figure 1

ROC curve of FPI. Healthy (n = 334) and pain (n = 153) subjects were matched for sex (female: 50%) and age (avg per group: 55 years old). AUROC area under the receiver operating characteristic

Full size image
Fig. 2
figure 2

Comparison of means (non-parametric t test) of FPI. Healthy (n = 334) and pain (n = 153) subjects were matched for sex (female: 50%) and age (avg per group: 55 years old). ***P value < 0.0001

Full size image

Association of FPI Scores with Clinical Assessments of Pain

Criterion validation analysis evaluated the significance of any correlations between FPI scores and validated clinical assessments for chronic pain. FPI scores were significantly associated with overall SF-36 scores (P = 0.0141) (Fig. 2), general health (P = 0.0457) and even more significantly, with emotional well-being (P = 0.0044), and limitations due to emotional problems (P = 0.0011) (Table 7).

Table 7 Relationship between the FPI score and clinical assessments of chronic pain
Full size table

In the pre-validation model of FPI, scores (0–100 scale) were further categorized into tiers that represent the likelihood of detected abnormalities being pain determinants (Table 2). Validation of scoring tiers was carried out by examining associations between SF-36 scores and FPI tiers (low, moderate, moderately high, and high). We hypothesized that subjects with severe biochemical disarrangement evident by higher FPI scores would be more impacted, both physically and emotionally by their pain. The SF-36 was selected as the most appropriate, validation clinical assessment to evaluate both the physical and emotional impact of pain on subjects. Validation analysis was conducted by first comparing SF-36 scores in pain subjects who exhibited LOW FPI scores (< 20) to those who exhibited moderately high and high FPI scores (> 75). Comparison of SF-36 means (non-parametric t test) for the LOW FPI cohort and the high PI cohort reveal a strong discrimination between SF-36 scores of low and high FPI scores (AUROC: 0.7715; P = 0.0088) (Figs. 3 and 4).

Fig. 3
figure 3

Comparison of means (non-parametric t test) of FPI severity and SF-36 scores. Clinical evaluations were compared between randomly selected pain patients with moderately high and high FPI severity scores (> 75 FPI; n = 20) and low FPI severity scores (< 20 FPI; n = 20). **P value < 0.001

Full size image
Fig. 4
figure 4

ROC curve of FPI severity and SF-36 scores. ROC analysis was performed between randomly selected pain patients with moderately high and high FPI severity scores (> 75 FPI; n = 20) and low FPI severity scores (< 20 FPI; n = 20). AUROC area under the receiver operating characteristic curve

Full size image

Further validation analysis was conducted to evaluate the association between overall SF-36 scores and FPI as a continuous score. Figure 5 illustrates the continuous, significant association (P = 0.0177) between FPI scores and overall SF-36 scores across the four tiers of FPI scores (low, moderate, moderately high, and high) (Fig. 5).

Fig. 5
figure 5

Association between SF-36 scores and FPI severity among chronic pain patients. Data is represented as mean ± SEM and was analyzed by linear trend analysis of one-way ANOVA. MOD moderate, M.HIGH moderately high

Full size image

Final validation of the scoring tiers was conducted by evaluating both the biomarker and clinical (SF-36) characteristics of subjects within each tier (Table 8). Criteria for validation required a significant (P value < 0.05) difference between the number of abnormal biomarker findings and SF-36 scores across all four categories with a linear trend confirming that worsening clinical assessments correlate with worsening FPI scores.

Table 8 Biomarker and clinical characteristics of subjects across all four FPI scoring tiers
Full size table

Discussion

These results demonstrate the clinical validity and associated correlations between FPI scores and validated clinical assessments of chronic pain (limitations due to emotional problems, emotional well-being, general health, and overall SF-36 scores) in a heterogenous cohort of chronic pain patients across multiple sites. The FPI score describes the degree and severity of underlying metabolic derangement that may be driving painful symptomology, and therefore, provides novel objective information that will compliment current subjective assessments. In addition to providing mechanistic insight into underlying biochemical derangement in chronic pain, the FPI also directly underscores the potential for novel non-opioid therapies that have been shown to modulate component biomarkers. Novel, biochemical intelligence combined with safe, personalized, non-opioid therapy options will increase the likelihood of successful and prolonged pain control, while reducing healthcare costs and reliance on opioid medications.

During criterion validation analysis, the FPI was strongly associated with limitations due to emotional problems (P = 0.0011) and emotional well-being (P = 0.0044), both components of the SF-36 survey. The validity and importance of these findings is further supported by examining the component biomarkers and the diverse set of biochemical and metabolic pathways represented by these surrogate markers. Component biomarkers exhibiting the strongest distinguishing power between pain-free and pain cohorts included methylmalonic acid, xanthurenic acid, pyroglutamic acid and the kynurenine pathway metabolites, quinolinic acid, and kynurenic acid. While each of these biomarkers lies along pathways that can directly impact the development, worsening and/or perception of pain, they are also intimately linked to mental health and emotional well-being.

Methylmalonic acid is a sensitive and specific marker of intracellular vitamin B12 status. Elevated levels of this urinary metabolite indicate an increased demand for this critical micronutrient. Vitamin B12 deficiencies are commonly detected in the chronic pain population and can drive painful and neuropsychiatric symptomologies [2, 3, 40,41,42,43]. Xanthurenic acid is a metabolomic biomarker of vitamin B6 status. Elevated levels of this kynurenine pathway metabolite indicate an increased demand for vitamin B6. Vitamin B6 deficiencies have long been recognized as causes of painful peripheral neuropathies, migraine, chronic pain, depression, and other neuropsychiatric diseases [13, 44, 45]. Pyroglutamic acid is a well-characterized metabolomic biomarker of glutathione depletion or a reduced glutathione response capacity. Elevated levels of this urinary marker, indicating a need for glutathione support, is a common finding in chronic pain patients taking daily doses of acetaminophen (APAP). APAP is capable of directly depleting glutathione stores [2, 26]. A reduced glutathione response capacity renders nerve cells susceptible to oxidative damage, which can cause or worsen peripheral neuropathies and drive the development of neuropsychiatric disease [28, 46, 47]. Quinolinic and kynurenine acid are neuroactive metabolites of the kynurenine pathway (KP), which can impact the development of pain hypersensitivity and depression through their direct action on NMDA receptors and relationship to serotonin synthesis. The KP is the major pathway responsible for the catabolic degradation of dietary tryptophan. Chronic activation of the KP occurs under conditions of systemic inflammation due to the ability of pro-inflammatory cytokines to directly upregulate this critical pathway. Under normal conditions, a small percentage of dietary tryptophan is utilized for serotonin synthesis but in the presence of a persistent inflammatory response, tryptophan is preferentially shunted down the KP, at the cost of serotonin synthesis. This phenomenon leads to the accumulation of quinolinic acid (leading to NDMA-mediated excitotoxicity and depression) and decreased serotonin synthesis, which further amplifies heightened pain and depressive symptoms [35, 48,49,50].

Considering the high prevalence of mental health disorders including anxiety, depression, bipolar disorder, and post-traumatic stress disorder in chronic pain populations, it is timely and significant that we observe and report prevalent perturbations in biochemical pathways capable of afflicting both physical and mental health. Our results provide evidence of criterion validity for the FPI by demonstrating a significant association with the SF-36 and multiple sub-components of this validated assessment for chronic pain. The clinical significance of specific objective markers of pain, which provide mechanistic insight into subjective reports, cannot be overstated, and it has long been recognized that such tools, capable of encompassing self-reported pain scores, would represent a significant advancement for the field of pain management [51]. Figure 5 and Table 8 illustrate the potential clinical significance of the FPI and its associated treatments. To the best of our knowledge, this is the first report of a validated pain biomarker panel that shows association with validated clinical assessments of pain. Strong association between higher FPI scores and worsening SF-36 responses provides mechanistic insight into the potential role of deranged biochemistry in the worsening and emotional impact of chronic pain. Potentially more exciting than the prospect of objective measures encompassing self-report is the fact that perturbed biochemistry can be corrected with appropriate low-cost therapies and supplementation. If metabolic correction proves capable of improving SF-36 scores, while simultaneously improving FPI scores, it will represent the first laboratory tool of its kind by allowing providers to truly modulate the course of disease.

Face validity of the FPI is supported by the direct relationships between component biomarkers and biochemical pathways known to be involved in the pathogenesis of chronic pain. Many of the metabolic, biochemical, and nutritional abnormalities indicated by the FPI test have been described in the literature as being the underlying causes of pain [13, 22, 25, 27, 52,53,54,55]. FPI biomarkers represent and evaluate cytokine-mediated chronic inflammation (kynurenine pathway metabolites), oxidative stress (pyroglutamic acid, ethylmalonate, and hydroxymethylglutarate), micronutrient deficiencies (methylmalonic acid, xanthurenic acid, homocysteine), and neurotransmitter turnover (5-hydroxyindoleacetic acid and vanilmandelic acid), all of which have direct links to the development, worsening, or heightened perception of pain. In addition, many of the non-opioid, biomarker-modulating compounds that would be directly indicated by the FPI test have exhibited significant pain-relieving effects in numerous randomized controlled clinical trials [28, 56,57,58,59,60,61,62]. FPI testing will facilitate and simplify patient selection for these important compounds by affording providers the ability to objectively identify patients who exhibit abnormalities and hence require metabolic correction therapies.

Mean comparison testing of FPI scores in pain and pain-free cohorts also provides evidence of discriminant validity by demonstrating significant differences in the mean FPI scores from pain-free subjects, when compared to the chronic pain cohort (P value < 0.0001). Despite strong discriminatory power, this assay is not intended to diagnose chronic pain. Rather, this innovative assay has been designed and validated to identify patients whose pain complaints may be due, at least in part, to biochemical, metabolic, and nutritional abnormalities. The importance of identifying such patients with a non-invasive, cost-effective test cannot be overstated, as these abnormalities can be treated and corrected with safe, widely available compounds. In addition, patients whose pain is due, at least in part, to underlying biochemical derangement will likely experience no long-term benefit from opioid therapy, as this class of medication is in no way addressing the underlying pathology driving the painful symptoms. Only with objective identification of abnormalities and targeted metabolic correction will these patients experience prolonged pain relief. Therefore, identification of such patients prior to the initiation of opioid therapy represents an immediately available and cost-effective strategy to reduce the opioid burden.

Validation study subjects were recruited from diverse geographic locations and exhibited a wide variety of primary pain complaints including, but not limited to, low back pain, lumbar radiculopathy, chronic pain syndrome, and cervicalgia (43%, 31%, 22%, and 14% of pain cohort, respectively). Study subjects were also prescribed a variety of opioid analgesics for pain control. Because of these diverse characteristics, which reflect real-world demographics, the FPI test can be considered a valid objective assessment for chronic pain patients in the clinic. Considering the complexity and biopsychosocial nature of chronic pain, it is not surprising that composite, multi-biomarker statistical models outperformed individual biomarkers during discriminant validation.

Limitations

Chronic pain subjects exhibited abnormal biochemical function across various relevant pathways indicating systemically perturbed biochemistry. While these abnormalities may be responsible for the onset of pain in some patients, other detected abnormalities may reflect the long-term metabolic cost of chronic pain. Regardless of whether detected abnormalities drove the acute onset of pain or its subsequent chronification, the evidence suggests biologically comprehensive test panels, evaluating multiple pathways are better equipped to characterize pain biochemistry than single marker assays. Furthermore, as biomarkers and FPI scores were assessed by single-point, cross-sectional analysis, longitudinal monitoring through repeat FPI testing in future studies will provide data on the efficacy of modulating therapies, as well as provide patients with objective updates on their biochemical status over time. Such tools will likely improve compliance and motivate patients to adhere to the metabolic correction protocols.

Conclusions

Objective identification of patients exhibiting high FPI scores will enable providers to initiate novel conversations and implement innovative treatment plans to correct underlying abnormalities and modulate the course of the disease. Ongoing research efforts will seek to determine the impact of FPI testing and subsequent metabolic correction on patients’ outcomes in both new and long-term chronic pain patients. Studies will also investigate the impact of FPI testing and metabolic correction following acute injury in order to determine whether the transition from acute to chronic pain can be slowed, or even prevented, by optimizing biochemical function to ensure an appropriate inflammatory response.