Abstract
Purpose:
This study aimed to identify metabolic biomarkers and pathways that might be associated with total joint arthroplasty (TJA) early revision for heterogeneous failure modes using an individual data meta-analysis of metabolomics.
Methods:
Two independent osteoarthritis (OA) cohorts were included. Revision records of patients with primary knee and hip OA were extracted at an average of 11.1 and 7.8 years after primary TJA, respectively. Preoperative fasting plasma was metabolomically profiled. Concentrations of metabolites/metabolism indicators were natural log-transformed, and their associations with early revision for all reasons in each individual cohort were assessed using logistic regression; the summary statistics from each cohort were then subjected to random-effects meta-analysis modelling.
Results:
Five hundred seventy-two patients with primary OA in The Newfoundland Osteoarthritis Study and 368 in the Longitudinal Evaluation in the Arthritis Program: Osteoarthritis Study were included. The revision rates were 4% and 6%, and mean times to revision were 1.7 and 2.1 years, respectively. No metabolite reached the prespecified significance threshold for multiple testing correction. However, 119 metabolites including choline, tryptophan betaine, indole, ornithine, three acylcarnitines, four cholesteryl esters, two lysophosphatidylcholines, five long-chain diglycerides, and 101 unsaturated (very) long-chain triglycerides were nominally significant with P < 0.05, suggesting potential links between these metabolites and early revision. Among these, indole and one acylcarnitine were positively associated with revision (odds ratio ≥1.73), while all others were negatively associated (odds ratio ≤0.73).
Conclusion:
Overactivation of the tryptophan-indole metabolic pathway may be associated with early revision after primary TJA. However, the findings were suggestive and represented a composite signal from heterogeneous failure modes.
Hip and knee total joint arthroplasty (TJA) is one of the most performed surgical procedures to relieve joint pain and restore joint function for patients with end-stage osteoarthritis (OA). The annual volume of TJA exceeds 1.3 million in the United States alone. With the aging population and increased obesity rate, this number continues to grow and is projected to reach 5 million by 2060.1 TJA is largely successful, and the longevity of a TJA is expected to be 15 to 20 years, yet around 4% of patients require early revision for a variety of reasons.2 Compared with primary TJA, revision TJA is a more complex procedure with greater morbidity and mortality rates and a higher risk of subsequent revision, and it poses more notable economic challenges. Therefore, it is important to identify risk factors associated with TJA early revision to perform personalized risk assessment and help patients make informed decisions about their treatment plans.
Current studies on risk factors for early revision TJA mainly focus on patients' demographic factors, diagnosis and comorbidities, joint implant factors, medication use, and health services.3 To date, no study has been done to investigate biomarkers associated with early revision. Biomarkers are characteristics that can be objectively measured and evaluated to indicate pathogenic processes or responses to therapeutic interventions, and have been proven valuable in disease diagnosis, risk prediction, and treatment monitoring.4 This exploratory study aimed to identify metabolic biomarkers and pathways that might be associated with early revision TJA for heterogeneous failure modes using an individual data meta-analysis of metabolomics to generate hypotheses and raise interest in this field.
Methods
Study Participants
Two independent cohorts were included in the study. The Newfoundland OA Study (NFOAS) recruited patients undergoing knee or hip arthroplasty between 2011 and 2017 in St. John's, Newfoundland and Labrador, Canada.5 The Longitudinal Evaluation in the Arthritis Program: Osteoarthritis Study (LEAP-OA) recruited patients undergoing knee or hip arthroplasty between 2013 and 2018 in Toronto, Ontario (ON), Canada.6 Primary hip and knee OA diagnosis for the NFOAS cohort was based on the American College of Rheumatology (ACR) diagnostic criteria7,8 and confirmed by the attending orthopaedic surgeon and pathology reports of the removed articular cartilage after surgery, and that for the LEAP-OA cohort was based on ACR clinical criteria.7,8 Patients who were deceased/lost to follow-up within 2 years after the primary TJA were excluded from the analysis.
Data Collection
Revision records for both cohorts were extracted from electronic health databases. For the NFOAS, data on age and sex were collected using a questionnaire, and weight and height data were retrieved from the Eastern Health Meditech Health Care Information System. For the LEAP-OA, all demographic data were collected using a questionnaire. Body mass index was calculated as weight in kilograms divided by squared height in meters.
Metabolomics Profiling
Whole-blood samples were collected into Vacutainer EDTA tubes after at least 8 hours of fasting. Plasma was separated by centrifugation at 2000 revolutions per minute (RPM) for 10 minutes and stored at −80°C until analysis. Targeted metabolomic profiling was done using Biocrates MxP Quant 500 kit (BIOCRATES Life Sciences GmbH), which quantifies 630 metabolites and 234 metabolism indicators (Supplemental Digital Content Tables S1 and S2, http://links.lww.com/JG9/A502 and http://links.lww.com/JG9/A503). Plasma was profiled on an API5500 Qtrap tandem mass spectrometry instrument (Applied Biosystems/MDS Analytical Technologies) equipped with a liquid chromatography at The Metabolomics Innovation Centre (https://metabolomicscentre.ca). The Biocrates MetIQ software was used to control the entire assay workflow from sample registration, automated calculation of metabolite concentrations to the export of data into other data analysis programs. Metabolite concentrations were reported as μmol/L.
Statistical Analysis
As there is no clear timeframe definition for early revision TJA,9 based on patients' revision status in the two cohorts, early revision in this study was defined as having revision TJA within 6 years after the primary TJA to ensure the inclusion of all patients who underwent revision not due to the natural wear or loosening at the end of the life span of the prosthesis. Metabolomic data underwent quality control before analysis. Metabolites/metabolism indicators were removed if more than 25% of the samples had values below the limit of detection. Values < limit of detection for the remaining metabolites/metabolism indicators were imputed using a linear regression prediction model in which concentration of a given metabolite was regressed on age, sex, and body mass index. Principal component analysis was done on metabolite concentrations, and no batch effect was detected. Concentrations of metabolites/metabolism indicators were natural log-transformed, and their associations with early revision in each individual cohort were assessed using logistic regression; the summary statistics from each cohort were then subjected to random-effects meta-analysis modelling using inverse variance as weights. Concentrations of nominally significant metabolites of the same class were summed, and the associations between these sums and early revision were also assessed using the same method. Analyses were done for early revision for all reasons combined and early revision due to pain only. The significance level for metabolites/sums of nominally significant metabolites of the same class and metabolism indicators associated with revision was defined at ⍺ = 8.17 × 10−5 and 2.49 × 10−4, respectively, to control for multiple testing for the 612 metabolites and 201 metabolism indicators that passed quality control in both cohorts with the Bonferroni method, and that for all other variables was set at ⍺ = 0.05. All analyses were done in R Studio with R version 4.2.3. Visualizations of the results were created using the ggplot210 and forestplot11 R packages.
Results
Patient Characteristics
Five hundred seventy-two patients with primary OA in NFOAS and 368 in LEAP-OA were included in the study. Body mass index was significantly higher in NFOAS than in LEAP-OA (P < 0.001), but there was no difference in age or sex distribution between the two cohorts (Table 1). Revision data were extracted 11.1 and 7.9 years after the primary TJA for the two cohorts, respectively, and the date of data extraction was August 20, 2024, for NFOAS and May 3, 2023, for LEAP-OA. The total revision rates were 4% (n = 23) and 6% (n = 21), and mean times to revision were 1.7 and 2.1 years, respectively. In NFOAS, 14 patients had revision for total knee arthroplasty and nine for total hip arthroplasty. The top reason for revision was infection (n = 9) followed by pain (n = 6, Table 2). In LEAP-OA, 16 patients had revision for total knee arthroplasty and five for total hip arthroplasty. Two patients had revision due to infection and 12 due to pain (Table 2). For early revision due to pain specifically, the rates were 1% and 3% (P = 0.02), and times to revision were 1.9 and 3.3 years, respectively, for the two cohorts. No difference was observed in age, body mass index, or sex distribution between revision and nonrevision patients in either cohort (P ≥ 0.07, Table 1 and Supplemental Digital Content Table S3, http://links.lww.com/JG9/A504).
Table 1.
Characteristics of the Study Participants
| Factor or Variable | NFOAS (n = 572) | LEAP-OA (n = 368) | Between cohorts | ||||||
| Revision (n = 23) | Nonrevision (n = 549) | P a | All | Revision (n = 21) | Nonrevision (n = 347) | P a | All | P b | |
| Age (yrs) | 63.85 ± 7.53 | 66.02 ± 8.04 | 0.15 | 65.94 ± 8.02 | 62.71 ± 7.87 | 65.43 ± 9.12 | 0.21 | 65.28 ± 9.07 | 0.32 |
| BMI (kg/m2) | 35.42 ± 6.17 | 33.77 ± 6.85 | 0.19 | 33.84 ± 6.83 | 33.2 ± 7.86 | 30.24 ± 6.43 | 0.07 | 30.45 ± 6.52 | <0.001 |
| Female (%) | 52 | 55 | 0.80 | 55 | 43 | 61 | 0.11 | 60 | 0.15 |
| Revision rate (%) | — | — | — | 4.0 | — | — | — | 5.7 | 0.23 |
BMI = body mass index, LEAP-OA = the Longitudinal Evaluation in the Arthritis Program: Osteoarthritis Study, NFOAS = the Newfoundland Osteoarthritis Study
P value between revision and nonrevision groups within each cohort.
P value between NFOAS and LEAP-OA cohorts.
Values are mean ± standard deviation or percentage. P values were obtained by the Mann-Whitney U test or chi squared test wherever appropriate.
Table 2.
Revision Reasons in Study Cohorts
| Reason | NFOAS (n [%]) | LEAP-OA (n [%]) | P |
| 0.04 | |||
| Infection | 9 (39.1) | 2 (9.5) | |
| Pain | 6 (26.1) | 12 (57.1) | |
| Decreased range of motion | — | 1 (4.8) | |
| Displacement | 2 (8.7) | 3 (14.3) | |
| Instability | 1 (4.3) | — | |
| Loosening | 2 (8.7) | — | |
| Fracture | — | 2 (9.5) | |
| Ossification | 1 (4.3) | — | |
| Stiffness | 2 (8.7) | 1 (4.8) |
LEAP-OA = the Longitudinal Evaluation in the Arthritis Program: Osteoarthritis Study, NFOAS = the Newfoundland Osteoarthritis Study.
P value was obtained by the Fisher exact test.
Metabolites/Sums/Metabolism Indicators Associated With Early Revision Total Joint Arthroplasty
No metabolite was significantly associated with early revision after controlling for multiple testing (P ≥ 5.6 × 10−4); however, 119 metabolites including choline, tryptophan betaine, indole, ornithine, three acylcarnitines (ACs), four cholesteryl esters (CEs), two lysophosphatidylcholines, five unsaturated long-chain diglycerides (DGs), and 101 unsaturated (very) long-chain triglycerides (TGs) were nominally significant with P < 0.05 (Figure 1, Supplemental Digital Content Table S4, http://links.lww.com/JG9/A505), suggestive of their associations with early revision. Among these, indole and one acylcarnitine were positively associated with revision (odds ratio [OR] ≥ 1.73), while all others were negatively associated (OR ≤ 0.73, Figure 2, Supplemental Digital Content Table S4, http://links.lww.com/JG9/A505). Sums of the nominally significant metabolites of the same class were all negatively associated with revision except ACs with P < 0.05 (nominal significance, suggestive of their associations with early revision, OR ≤ 0.43, P ≤ 0.04, Figure 2, Supplemental Digital Content Table S5, http://links.lww.com/JG9/A506). Similar to individual metabolites, no metabolism indicator was significantly associated with revision after controlling for multiple testing (P ≥ 0.01), but 10 metabolism indicators were associated with P < 0.05 (nominal significance, suggestive of their associations with early revision, Supplemental Digital Content Table S6, http://links.lww.com/JG9/A507). Among these, sums of (unsaturated) TGs, very long-chain fatty acid (VLCFA) CEs, and polyunsaturated fatty acid (PUFA) ACs were negatively associated with revision, while the ratio of phosphatidylcholines (PCs) to choline was positively associated.
Figure 1.
Volcano plot for metabolites nominally significantly associated with total joint arthroplasty early revisionGreen dot indicates metabolites negatively associated with early revision; red dot indicates metabolites positively associated with early revision; black dot indicates metabolites not associated with early revision using alpha=0.05.
Figure 2.
Forest plot for nominally significant metabolites and sums of these metabolites of the same class associated with total joint arthroplasty early revisionCE: cholesteryl ester; CI: confidence interval; DG: diglyceride; LEAP-OA: Longitudinal Evaluation in the Arthritis Program, Osteoarthritis Study; LPC: lysophosphatidylcholine; NFOAS: Newfoundland Osteoarthritis Study; OR: odds ratio; TG: triglyceride.
In the subanalysis on early revision TJA due to joint pain, no metabolite was significantly associated with revision after controlling for multiple testing (P ≥ 0.001) and 65 metabolites including xanthine, glutamate, sarcosine, two ACs, three CEs, seven unsaturated long-chain DGs, and 50 unsaturated (very) long-chain TGs were nominally significant with P < 0.05 (Supplemental Digital Content Fig. S1, http://links.lww.com/JG9/A500, Table S7, http://links.lww.com/JG9/A508), suggestive of their associations with early revision. Among these, one DG was positively associated with revision (OR = 3.04), while all other metabolites were negatively associated (OR ≤ 0.57). Sums of the nominally significant metabolites of the same class were all negatively associated with revision (OR ≤ 0.20, P ≤ 0.03, Supplemental Digital Content Table S8, http://links.lww.com/JG9/A509, Fig. S2, http://links.lww.com/JG9/A501). No metabolism indicator was significantly associated with revision after controlling for multiple testing (P ≥ 0.007), but nine metabolism indicators were associated with revision with P < 0.05 (nominal significance, suggestive of their associations with early revision, Supplemental Digital Content Table S9, http://links.lww.com/JG9/A510), among which sums of (unsaturated) TGs, (unsaturated) DGs, VLCFA CEs, and PUFA ACs were negatively associated with revision.
Discussion
In this study, the early revision rate was 4% in the NFOAS, which was consistent with previous reports.2 The rate was slightly higher in the LEAP-OA cohort, but the difference was not statistically significant. Previous studies have shown that younger age and male sex are risk factors for early revision of TJA.12 We also observed similar trends in both cohorts in our study, but the differences between revision and nonrevision patients were not significant, which could be due to the small sample size and/or study populations.
To the best of our knowledge, this is the first exploratory study to investigate metabolic biomarkers associated with early revision TJA. We found that choline and tryptophan betaine were negatively associated with revision, while indole was positively associated with P < 0.05, suggesting likely associations between early revision TJA and these metabolites and/or metabolic pathways involving these metabolites. Choline is the precursor to form PC through the cytidine-5-diphosphocholine (CDP)-choline pathway. It has been reported that patients with arthritis have lower choline levels compared with control subjects,13 and we observed lower choline levels in early revision patients which could be due to the overactivation of CDP-choline pathway as evidenced by the negative association between the ratio of PCs to choline and revision. Choline can also be converted to acetylcholine (Ach), functioning as the primary neurotransmitter at the neuromuscular junction, facilitating the transmission of electrical signals from the motor neuron to the skeletal muscle fiber, ultimately triggering muscle contraction.14 Lowered choline and subsequent Ach levels could have led to weakened muscular strength, contributing to unfavorable surgery outcomes such as joint instability,15 which could eventually result in aseptic failure. Ach is also an important component of the cholinergic anti-inflammatory pathway (CAIP) in which Ach binds to its receptor to inhibit the production of proinflammatory cytokines.16 Compared with the diffusible anti-inflammatory network involving hormones and anti-inflammatory cytokines, this neural-based anti-inflammatory pathway is localized in injured tissues,17 making it more effective in fighting infections in the surgery joint. The recent observation of cholinergic nerves in subchondral bone in patients with OA and evidence on the influence of neural and nonneural cholinergic systems in various joint tissues18 suggest that the regulation of local inflammation through Ach release and downstream CAIP might be involved in OA development. It might be plausible that the decreased choline level in early revision TJA patients led to the reduced production of Ach and subsequent downregulation of CAIP, an important pathway in sepsis pathophysiology.19 Unfortunately, the metabolomic assay used in our study does not measure the level of Ach. This hypothesis requires additional investigation. In addition to exerting an anti-inflammatory effect through Ach, choline has also been reported to prevent inflammatory hyperalgesia without affecting the inflammatory response by a mechanism independent of CAIP, and such an effect was also observed in a postoperative pain model.20 This evidence further suggests the potential clinical use of choline supplementation in patients experiencing chronic pain after TJA. Conversely, in our subanalysis of early TJA revision due to pain, the significance of this metabolite diminished (P = 0.06), which was likely due to the smaller sample size.
In liver and kidney, choline indirectly serves as a methyl donor in the conversion of tryptophan to tryptophan betaine (Figure 3). Given that choline was downregulated in early revision patients, this pathway could also have been suppressed resulting in the negative association we observed between tryptophan betaine and early TJA revision. Tryptophan betaine exerts anti-inflammatory actions,21 making it a likely player in joint infection after TJA. It can also inhibit osteoclast differentiation and prevent inflammatory bone loss.22 The bone changes caused by the reduced tryptophan betaine level might be partially responsible for early revision TJA due to aseptic failure such as fracture, loosening, instability, or implant displacement. On the other hand, reduced tryptophan betaine synthesis would render more tryptophan for downstream metabolism. Tryptophan is an essential amino acid that can be metabolized through three pathways, kynurenine, serotonin, and indole pathways. In brain, gastrointestinal tract, and liver, tryptophan is degraded into kynurenine which is then further catabolized into kynurenic acid and quinolinic acid. In gastrointestinal tract and central nervous system, tryptophan is catalyzed into serotonin and then further converted into melatonin. In the gastrointestinal tract, approximately 5% of the tryptophan is converted into indole and indole derivatives such as indole-3-acetic acid (IAA), indole-3-aldehyde (IAld), indole-3-propionic acid (IPA), and tryptamine (Figure 3) by gut microbiota.23 We had data on four tryptophan derivatives: kynurenine, indole, IAA, and IPA. The only one that could be associated with early TJA revision (with P < 0.05) in our study was indole which was upregulated in revision patients. Seymour et al24 recently reported that bacterial dysbiosis associated with collagen-induced arthritis (CIA) can enhance the production of indole pathway metabolites, and adding indole to tryptophan-low diet promoted proinflammatory cytokine production, resulted in a 4.5-fold increase in CIA severity compared with feeding tryptophan-low diet alone. They also demonstrated that bacterial tryptophan catabolism was required to generate indole for the development of inflammatory arthritis in CIA and observed positive correlations between expanded bacteria taxa Firmicutes, Barnesiella, Rikenellaceae, and Lactobacillales and indole levels. Interestingly, in a recent study conducted by our group, Barnesiella was found to be positively associated with sustained postoperative joint pain,25 which was one of the main reasons for early revision in our cohorts. These findings suggest that intestinal dysbiosis during arthritis results in indole stimulation of arthritis development and joint damage and likely contributes to postoperative joint pain as well. The elevated indole level in early revision patients might indicate inflammation and/or reflect pain in the replaced joints, a signal for potential early revision for infection or pain. Blocking indole production or restoring a healthy gut microbiome might reduce the risk of early revision.
Figure 3.
Pathways involving metabolites associated with total joint arthroplasty early revisionAchR: acetylcholine receptor; CAIP: cholinergic anti-inflammatory pathway; ChAT: choline acetyltransferase; IAA: indole-3-acetic acid; IAld: indole-3-aldehyde; IPA: indole-3-propionic acid; ODC: ornithine decarboxylase; SAM: S-adenosylmethionine. Red arrow indicates negative association with early revision; blue arrow indicates positive association with early revision.
This exploratory study has some limitations. No metabolites or metabolic indicators survived multiple testing corrections which were used to control for the overall error rate across all 612 tests for metabolites and 201 for metabolic indicators, and our findings were based on metabolites/metabolic indicators with P < 0.05; hence, the associations identified were only suggestive. The number of patients who underwent revision was low, and the distribution of reasons for revision in the two cohorts was largely different, with the top reason being infection in NFOAS (39%) and pain in LEAP-OA (57%, P = 0.04, Table 2); thus, it was unfeasible to analyze revision for each reason individually. However, our findings might shed light on metabolic risk factors and pathways involved in TJA early revision because some metabolites could affect various aspects of TJA recovery and outcome. Although we did explore metabolomics for revision due to pain specifically, the sample size was further reduced, which might have limited our ability to discover more associated metabolites. In addition, the postoperative joint pain that led to early revision could also have been caused by various reasons such as component malrotation, aseptic loosening, and soft tissue impingement; thus, our findings represented a composite signal from heterogeneous failure modes. In addition, due to the low number of early revision patients, we were unable to perform analysis in knee and hip patients separately.
Conclusion
Findings of this study suggested that the overactivation of tryptophan-indole metabolic pathway was possibly associated with early revision for heterogeneous failure modes after primary TJA, highlighting the potential role of gut microbiome in TJA outcome and revision risk. However, our findings originated from a small number of patients who underwent early revision due to mixed reasons. Additional research with larger sample sizes and targeting early revision for more specific reasons is essential to identify metabolic risk factors associated with early revision TJA. This will enable us to better inform patients and develop personalized preoperative and postoperative care strategies for those patients undergoing TJA.
Acknowledgements
We thank all study participants who made this study possible.
Footnotes
None of the following authors or any immediate family member has received anything of value from or has stock or stock options held in a commercial company or institution related directly or indirectly to the subject of this article: Liu, Dr. Gill, Dr. Gandhi, Huang, Dr. Furey, Dr. Perruccio, Dr. Rampersaud, Dr. Mahomed, Dr. Rahman, Dr. Kapoor, and Dr. Zhai.
Dr. Zhai: conceptualization, supervision, funding acquisition, resources, data curation, methodology, formal analysis, investigation, writing—original draft preparation, writing—review and editing. Liu: data curation, formal analysis, investigation, writing—original draft preparation, visualization, writing—review and editing. Dr. Gandhi: data curation, formal analysis, investigation, writing—review and editing. Huang: formal analysis, investigation, writing—review and editing. Dr. Furey, Dr. Perruccio, Dr. Gandhi, Dr. Kapoor, Dr. Perruccio, Dr. Rampersaud, Dr. Mahomed: funding acquisition, resources, data curation, writing—review and editing. All authors have read and agreed to the published version of the manuscript.
The data sets used and/or analyzed during this study are available from the corresponding author on reasonable request.
Ethics approval and consent to participate: The NFOAS was approved by the Health Research Ethics Authority of Newfoundland and Labrador (HREB # 2011.311). The LEAP-OA study was approved by the University Health Network Research Ethics Board (REB # 16-5759). Informed written consent was obtained from all study participants.
Contributor Information
Ming Liu, Email: mingl@mun.ca.
Robert Gill, Email: rmg445@mun.ca.
Rajiv Gandhi, Email: rajiv.gandhi@uhn.ca.
Jingyi Huang, Email: jingyih@mun.ca.
Andrew Furey, Email: afurey99@gmail.com.
Anthony Perruccio, Email: anthony.perruccio@uhnresearch.ca.
Raja Rampersaud, Email: raja.rampersaud@uhn.ca.
Nizar N. Mahomed, Email: Nizar.Mahomed@uhn.ca.
Proton Rahman, Email: prahman@mun.ca.
Mohit Kapoor, Email: Mohit.Kapoor@uhnresearch.ca.
References
- 1.Shichman I, Roof M, Askew N, et al. : Projections and epidemiology of primary hip and knee arthroplasty in medicare patients to 2040-2060. JB JS Open Access 2023;8:e22.00112. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Dy CJ, Marx RG, Bozic KJ, Pan TJ, Padgett DE, Lyman S: Risk factors for revision within 10 years of total knee arthroplasty. Clin Orthop Relat Res 2014;472:1198-1207. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Jasper LL, Jones CA, Mollins J, Pohar SL, Beaupre LA: Risk factors for revision of total knee arthroplasty: A scoping review. BMC Musculoskelet Disord 2016;17:182. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Costello CA, Rockel JS, Liu M, et al. : Individual participant data meta-analysis of metabolomics on sustained knee pain in primary osteoarthritis patients. Rheumatology (Oxford) 2023;62:1964-1971. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Liu M, Huang Y, Zhang H, et al. : Restricting branched-chain amino acids within a high-fat diet prevents obesity. Metabolites 2022;12:334. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Power JD, Perruccio AV, Gandhi R, et al. : Neuropathic pain in end-stage hip and knee osteoarthritis: Differential associations with patient-reported pain at rest and pain on activity. Osteoarthritis Cartilage 2018;26:363-369. [DOI] [PubMed] [Google Scholar]
- 7.Altman R, Alarcon G, Appelrouth D, et al. : The American college of rheumatology criteria for the classification and reporting of osteoarthritis of the hip. Arthritis Rheum 1991;34:505-514. [DOI] [PubMed] [Google Scholar]
- 8.Altman R, Asch E, Bloch D, et al. : Development of criteria for the classification and reporting of osteoarthritis. Classification of osteoarthritis of the knee. Diagnostic and therapeutic criteria committee of the American Rheumatism Association. Arthritis Rheum 1986;29:1039-1049. [DOI] [PubMed] [Google Scholar]
- 9.Santarelli M, Wirth W, Engl M, Saccoccio M: Early revision in total joint arthroplasty surgery—A narrative review. Orthop Rev (Pavia) 2025;17:144006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Hadley Wickham DN, Thomas LP: ggplot2: Elegant graphics for data analysis. Springer-Verlag New York, 2016. [Google Scholar]
- 11.Max Gordon TL: Forestplot: Advanced forest plot using 'grid' graphics. in R Package Version 1.10.1, edn, 2020. [Google Scholar]
- 12.Santaguida PL, Hawker GA, Hudak PL, et al. : Patient characteristics affecting the prognosis of total hip and knee joint arthroplasty: A systematic review. Can J Surg 2008;51:428-436. [PMC free article] [PubMed] [Google Scholar]
- 13.Cedola FCR, Sanchez-Lopez E, Mateo L, et al. : 2019 ACR/ARP Annual Meeting: 2019, 2019.Choline metabolite is associated with inflammation in arthritis in the elderly [Google Scholar]
- 14.Kuo IY, Ehrlich BE: Signaling in muscle contraction. Cold Spring Harb Perspect Biol 2015;7:a006023. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Stambough JB, Edwards PK, Mannen EM, Barnes CL, Mears SC: Flexion instability after total knee arthroplasty. J Am Acad Orthop Surg 2019;27:642-651. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Borovikova LV, Ivanova S, Zhang M, et al. : Vagus nerve stimulation attenuates the systemic inflammatory response to endotoxin. Nature 2000;405:458-462. [DOI] [PubMed] [Google Scholar]
- 17.Tracey KJ: The inflammatory reflex. Nature 2002;420:853-859. [DOI] [PubMed] [Google Scholar]
- 18.Lauwers M, Courties A, Sellam J, Wen C: The cholinergic system in joint health and osteoarthritis: A narrative-review. Osteoarthritis Cartilage 2021;29:643-653. [DOI] [PubMed] [Google Scholar]
- 19.Kanashiro A, Sonego F, Ferreira RG, et al. : Therapeutic potential and limitations of cholinergic anti-inflammatory pathway in sepsis. Pharmacol Res 2017;117:1-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Kusuda R, Carreira EU, Ulloa L, Cunha FQ, Kanashiro A, Cunha TM: Choline attenuates inflammatory hyperalgesia activating nitric oxide/cGMP/ATP-sensitive potassium channels pathway. Brain Res 2020;1727:146567. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Sun H, Zhu X, Cai W, Qiu L: Hypaphorine attenuates lipopolysaccharide-induced endothelial inflammation via regulation of TLR4 and PPAR-Gamma dependent on PI3K/Akt/mTOR signal pathway. Int J Mol Sci 2017;18:844. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Chen H, Guo T, Wang D, Qin R: Vaccaria hypaphorine impairs RANKL-induced osteoclastogenesis by inhibition of ERK, p38, JNK and NF-kappaB pathway and prevents inflammatory bone loss in mice. Biomed Pharmacother 2018;97:1155-1163. [DOI] [PubMed] [Google Scholar]
- 23.Taleb S: Tryptophan dietary impacts gut barrier and metabolic diseases. Front Immunol 2019;10:2113. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Seymour BJ, Trent B, Allen BE, et al. : Microbiota-dependent indole production stimulates the development of collagen-induced arthritis in mice. J Clin Invest 2023;134:e167671. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Huang J, Liu M, Furey A, Rahman P, Zhai G: Gut microbiomics of sustained knee pain in patients with knee osteoarthritis. J Rheumatol 2024;51:1218-1225. [DOI] [PubMed] [Google Scholar]