Template for Electronic Submission to ACS Journals



TITLE: Metabolic phenotyping in venous disease: the need for standardisation.Sarah Onida1?*, Matthew K. H. Tan1?, Marina Kafeza1?, Richmond T. Bergner2?, Joseph Shalhoub1?, Elaine Holmes2,3?? , Alun H. Davies1? 1 ? Academic Section of Vascular Surgery, Department of Surgery and Cancer, Imperial College London, Floor 4 East, Charing Cross Hospital, Fulham Palace Road, London W6 8RF, UK2 ? Section of Computational and Systems Medicine, Department of Surgery and Cancer, Imperial College London Sir Alexander Fleming Building, Prince Consort Road, Kensington, London SW7 2BB, UK3 ? Health Futures Institute, Murdoch University, Discovery Way, Perth, W.A., Australia* Corresponding author – s.onida@imperial.ac.ukABSTRACTVenous thromboembolism (VTE), chronic venous disease (CVD) and venous leg ulceration (VLU) are clinical manifestations of a poorly functioning venous system. Though common, much is unknown of the pathophysiology and progression of these conditions. Metabolic phenotyping (metabonomics) has been employed to explore mechanistic pathways involved in venous disease.A systematic literature review was performed: full text, primary research articles on the applications of nuclear magnetic resonance spectroscopy (NMR) and mass spectrometry (MS) in human participants and animals were included for qualitative synthesis.Seventeen studies applying metabolic phenotyping to venous disease were identified: six on CVD, two on VLU and nine on VTE; both animal (n=6) and human (n=10) experimental designs were reported, with one study including both. NMR, MS and MS imaging were employed to characterize serum, plasma, urine, wound fluid and tissue.Metabolites found to be upregulated in CVD included lipids, branched chain amino acids (BCAA), glutamate, taurine, lactate and myo-inositol identified in vein tissue.?Upregulated metabolites in VLU included lactate, BCAA, lysine, 3-hydroxybutyrate and glutamate identified in wound fluid and ulcer biopsies. VTE cases were associated with reduced carnitine levels, upregulated aromatic amino acids, 3-hydroxybutyrate, BCAA and lipids in plasma, serum, thrombus and vein wall; kynurenine and tricarboxylic acid pathway dysfunction were reported. Future research should focus on targeted studies with internal and external validation.KEYWORDS (Word Style “BG_Keywords”). Metabonomics; Metabolomics; Metabolic phenotyping; Systems biology; Omics; Biomarker; Chronic Venous Disease; Venous Thromboembolism; Venous Leg Ulceration.IntroductionClinical presentations of venous disease may reflect abnormalities in the superficial and/or deep venous systems. Chronic venous disease (CVD) encompasses common presentations, including aching legs, swelling and restlessness, with signs ranging from spider veins (telangiectasia), to varicose veins, skin changes and, ultimately, venous leg ulceration (VLU). CVD is common, with telangiectasia affecting up to 80% and varicose veins up to 40% of the populationADDIN BEC{Evans et al., 1999, #13665}1. VLU affects 1-2 % of the population, with the prevalence increasing to up to 4% in those over 65 years of ageADDIN BEC{Lal, 2015, #55754}2. Both conditions are expensive, together responsible for up to 2% of the annual healthcare budget expenditure of Western societiesADDIN BEC{Rabe and Pannier, 2010, #13764}3.Venous thromboembolic disease (VTE) describes venous thrombosis (DVT) and pulmonary embolism (PE). Although relatively uncommon, with a prevalence of 1:1000ADDIN BEC{Heit, 2015, #26566}4, VTE is the single, most common cause of hospital acquired mortalityADDIN BEC{Hunt, 2009, #43526}5 and a serious public health concern due to its complications: PE (concurrent with DVT in 50% of patients)ADDIN BEC{Paik et al., 2017, #95677}6 and post thrombotic syndrome (PTS, following up to 50% of DVTs)ADDIN BEC{Kahn, 2006, #96504}7. PE is a life-threatening sequela of DVT, that can lead to cardiac arrest and death; PTS is not life threatening but can lead to longstanding symptoms of venous insufficiency, with swelling, skin changes and ulceration. This impacts heavily on function, resulting in unemployment, social isolation and reduced quality of lifeADDIN BEC{Ghanima et al., 2018, #88227}8.Management of VTE has historically been conservative with anticoagulation administered for at least three months depending on thrombus aetiologyADDIN BEC{Kearon et al., 2016, #9170}9. Recently, with the advent of new pharmaceutical agents and technology, thrombolysis is increasingly being employed, particularly in patients with extensive DVTs diagnosed within the first two weeks. Thrombolysis aims to reduce the clot burden and the risk of developing PTS, although further studies are required in this areaADDIN BEC{Vedantham et al., 2017, #55451}{Comerota et al., 2019, #70708}10, 11.Research challenges in venous diseaseAlthough established pathways exist for the management of patients with CVD, VLU and DVT, there are a number of areas where further research should translate into improved care.Treatment of CVD yields clinical and quality of life improvements, whilst being cost effective, and is advocated by numerous clinical practice guidelinesADDIN BEC{National Institute of Health and Care Excellence (NICE), 2013, #83411}{Wittens et al., 2015, #12317}12, 13. Left untreated, estimated annual progression rates across stages of disease are in the region of 4%ADDIN BEC{Pannier and Rabe, 2011, #48130}14, with a small but important proportion of patients progressing from varicose veins, to skin changes and venous ulceration. What determines why certain individuals progress over others is not known; identification of these patients is important in the era of personalized medicineADDIN BEC{Nicholson et al., 2012, #34514}15 to ensure management is targeted to the needs of the individual.Despite the advantages of treating CVD, recurrent disease (defined as new veins and symptoms) requiring repeated intervention is an issue affecting approximately 7% of those undergoing treatmentADDIN BEC{Bush et al., 2014, #36589}16. Some patients exhibit more aggressive recurrence patterns than others; what influences this is currently unknown. Better characterization of these patients, with the ability to prognosticate recurrent disease, can help both develop targeted patient care and, potentially tailored, cost effective follow-up pathways.Venous leg ulceration is most commonly managed with compression therapy; this is cumbersome and costly for both patients and the healthcare serviceADDIN BEC{Rabe and Pannier, 2010, #19423}17. Despite best treatment, recurrence rates are in the region of 20% per annum and 40% at 5 yearsADDIN BEC{Gohel et al., 2005, #71803}18. Identifying patients who are likely not to heal or experience recurrence is important to develop targeted management strategies for patients.Unlike CVD and VLU, DVTs are a diagnostic challenge, with patients presenting with non-specific symptoms and symptom onset that is difficult to ascertain; they may miss the thrombolysis window, resulting in potentially less effective treatment. Furthermore, the existing diagnostic test for DVT, D-dimer, is only moderately sensitive and poorly specific. The development of a new biomarker for venous thrombosis could improve the diagnostic pathway for these patients. In terms of prognostication, 50% of those with a DVT will develop PTSADDIN BEC{Kahn, 2006, #96504}7. Being able to identify who will develop PTS would promote development of targeted strategies to help improve the care of these patients.Addressing the aforementioned challenges has been explored in various studies relating to omic disciplines, including proteomics, transcriptomics and genomics, with limited translational impactADDIN BEC{Mannello et al., 2014, #49764}{Jensen et al., 2018, #19322}19, 20. Metabonomics provides the means to explore these conditions in greater detail than that which was previously possible. Using high resolution spectroscopic platforms, molecules less than 1 kDa in size can be identified as candidate biomarkers to help explore the pathophysiology, progression and response to treatment in different conditions. This is of particular use in venous disease, where both innate and acquired factors are important. Metabolic profiling has been increasingly applied to vascular diseaseADDIN BEC{Vorkas et al., 2016, #35398}21 with developing biostatistical analytical techniques that now enable researchers to control for confounding factors and variablesADDIN BEC{Blaise et al., 2016, #77417}{Vorkas et al., 2015, #17159}{Vorkas et al., 2015, #82555}22–24. Both untargeted analyses, testing for the presence of all metabolites in a given sample, and targeted assays, identifying specific metabolite features of interest and their concentrations, can be performed. Proton (1H) nuclear magnetic resonance spectroscopy (NMR) and mass spectrometry (MS) are common analytical platforms employed in metabolic phenotyping studies. NMR is non-destructive and, historically, has been described as a more robust and reproducible assay compared to MS. MS, though destructive in nature, is highly sensitive but more time consuming, requiring different chromatographic techniques to assess for distinct metabolite classes (e.g. polar vs non polar). Furthermore, its destructive nature means that repeat experiments cannot be performed on a single sample. Together, both assays provide a comprehensive analysis of the metabolic phenotype of a given sampleADDIN BEC{Emwas, 2015, #53221}25.The aim of this review is to summarise the existing evidence regarding the use of metabolic profiling in venous disease. In addition, areas of translational research priority will be highlighted and considerations made with respect to study and experimental design.Materials and MethodsA systematic literature search was performed in accordance to the Preferred Reporting Items for Systematic Review and Meta – Analyses (PRISMA) guidelines. This review was registered with PROSPERO (Ref 107647). EMBASE and Medline databases were searched using the Ovid and Pubmed engines on the 7th June 2019 with no restriction on publication date. The search algorithm is reported in the supplemental information (Supplement 1). The search was performed by two authors (SO, MT) and any disagreements discussed with the senior author (AHD). The PRISMA diagram summarising the flow of analysis for the review can be viewed in Figure 1. Due to the heterogeneity in study methodology and experimental design, a narrative review was performed.Inclusion and exclusion criteriaAll full text, English language articles employing metabolic phenotyping technology (NMR and MS) in the context of CVD, VLU and VTE were included. Both human and animal studies were analysed due to the paucity of published data in this field. Non-English language articles and abstracts of non-published studies were excluded.ResultsThe search yielded 1,063 citations. Following removal of duplicates and application of exclusion criteria, 17 full text articles fulfilled the inclusion criteria: six relating to CVD, two to VLU and nine to VTE.Metabolic profiling in chronic venous disease Of the six studies pertaining to CVD, four were in human participantsADDIN BEC{Anwar et al., 2012, #113, Anwar et al., 2017, #35; Tanaka et al., 2010, #137; Tanaka et al., 2012, #122}26–28 and two in animal models, both in male Sprague Dawley ratsADDIN BEC{Anwar et al., 2016, #50; Tanaka et al., 2016, #43}29, 30. Studies employed 1H-NMR, MS and imaging MS (IMS) technology applied to vein tissue only; no studies analysed circulating or urinary biomarkers.Human studies explored differences between varicose vein and control vein by analysing tissue either from the refluxing truncal vein (great saphenous vein, GSV)ADDIN BEC{Anwar et al., 2012, #113}{Tanaka et al., 2010, #137}{Tanaka et al., 2012, #122}26–28 or from varicosities retrieved at phlebectomyADDIN BEC{Anwar et al., 2017, #35}31 and comparing them to control GSV. Methodology differed, with either analysis of aqueous extracts or tissue IMS employed. CVD stage of disease severity, as described by the clinical, etiology, anatomy, pathophysiology (CEAP) scaleADDIN BEC{Ekl?f et al., 2004, #76583}32, also differed in the human studies, with CEAP class ranging from C2 (varicose veins) to C5 (healed ulceration), as detailed in Table 1.Studies analysing aqueous extracts from human superficial vein identified a number of metabolites differentially expressed between cases and controls (listed in Table 1). A pilot study employing NMR spectroscopyADDIN BEC{Anwar et al., 2012, #113}26 observed creatine, myo-inositol and lactate as increasingly present within the varicose veins tissue, whilst triglycerides were more abundant in the control tissues. Veins were macroscopically described as thick walled, explaining the presence of creatine, a metabolite stored in skeletal and smooth muscleADDIN BEC{Wallimann et al., 1992, #31861}33. Myo-inositol is involved in cell signaling pathways; its presence was related to the known upregulation of growth factors and imbalance of matrix metalloproteinase (MMP)/ tissue inhibitor of matrix metalloproteinase (TIMP) activity known to occur in varicose veinsADDIN BEC{Chen et al., 2017, #48388}34. Lactate was described in association with hypoxia and anaerobic metabolism. Hypoxia inducible pathway activation has been explored in the pathophysiology of varicose veinsADDIN BEC{Lim et al., 2012, #3346}{Lim et al., 2013, #49093}35, 36, lending support to this theory. The validation study, employing NMR and ultra-performance liquid chromatography (UPLC)-MS, additionally performed microRNA expressionADDIN BEC{Anwar et al., 2017, #35}31 ADDIN BEC{Anwar, 2013, #32642}37. In the larger sample size (115 vs 13) a greater proportion of metabolites were differentially identified (Table 1). These included the metabolites identified in the pilot study with, additionally, glutamate, taurine and inosine detected via NMR and phosphatidylcholine and phosphatidylethanolamine via MS. Inflammation, regulation of cellular proliferation, apoptosis and hypoxia were again identified as pathways of relevance, providing further evidence for their role in varicose vein pathophysiology.A separate group performed imaging MS exploring lipid distribution across the vein wall in the valvular and perivalvular area of GSV tissue from patients with different disease stages (n = 10) versus controls (n = 6), identifying differential pro-inflammatory lipid expression at the vein valve in the disease groupADDIN BEC{Tanaka et al., 2010, #137}27 (Table 1). Such lipid species have been associated with the expression of cell adhesion molecules and inflammatory pathways, further supporting the role of inflammation in the development of CVDADDIN BEC{MacColl and Khalil, 2015, #75431}38. No statistically significant difference was identified between disease stages, likely due to the small sample size. A further study assessed segments of GSV from patients with CVD and controls in a slightly larger population (38 vs 10)ADDIN BEC{Tanaka et al., 2012, #122}28. The GSVs from the disease group were statistically significantly thicker and demonstrated higher lipid and triglyceride concentrations. Specifically, the vessel wall intima, media and adventitia had differential expressions of lipid molecules, particularly in advanced stages of CVD, which are associated with skin changes and chronic inflammation. Interestingly, compared to controls, lymphatic vessel presence was reduced in the CVD group, highlighting evidence of a possible association with lymphoedemaADDIN BEC{Tanaka et al., 2012, #122}28. The association between venous and lymphatic dysfunction is well described in phlebolymphoedemaADDIN BEC{Farrow, 2010, #79625}39, though poorly characterised mechanistically. These findings were further explored in one of the two animal studies identified in this review; lymphatic vessel ligation in a murine model resulted in increased accumulation of lipid moieties and thickening of the vein wall, suggesting a mechanistic pathway for the association between venous and lymphatic insufficiencyADDIN BEC{Tanaka et al., 2016, #43}30.An ex vivo animal study explored the response to mechanical stretch, aiming to replicate mechanisms of prolonged venous hypertension seen in CVDADDIN BEC{Anwar et al., 2016, #50}29. Prolonged stretching resulted in metabolic changes in the vein tissue (Table 1) associated with muscle breakdown and inflammation.These studies highlight the applicability of NMR, MS and matrix assisted laser desorption ionization (MALDI)-IMS technology in venous disease and, in many ways, complement existing information in the literature derived by genomic and proteomic studiesADDIN BEC{Chen et al., 2017, #48388}{Serralheiro et al., 2017, #67372}34, 40. However, none have assessed the concentration of metabolites across different stages of disease nor have they assessed participants longitudinally in a prospective manner to identify trends over time relating to progression. Furthermore, the studies are of a small sample size (ranging from 5 – 115) with heterogeneous control populations that are not matched with respect to demographics, comorbidities and medications, all confounding factors that may influence the statistical analysis. The use of alternative experimental platforms to provide validation data is also uncommon, resulting in conclusions of limited translational value. From a data analysis perspective, only two studies reported correcting for false discovery rate (FDR)ADDIN BEC{Anwar et al., 2016, #50; Anwar et al., 2017, #35}29, 31, three studies reported using Tukey’s post hoc testADDIN BEC{Tanaka et al., 2010, #137}27 ADDIN BEC{Tanaka et al., 2012, #122}28 ADDIN BEC{Tanaka et al., 2016, #43}30 and one study Scheffe’s post hoc testADDIN BEC{Ennis et al., 1994, #84035}41. These are important approaches in studies where multiple statistical tests and comparisons are performed on a single dataset, which can result in an increased rate of false positive, or Type I, errors; adjusted p values are generated to control for this occurrenceADDIN BEC{Noble, 2009, #2032}42. Future studies should be of a larger sample size, targeting metabolites associated with the mechanistic pathways highlighted to be involved. Additionally, more accessible samples, such as serum and urine, should be analysed to permit the assessment and development of less invasive assays.Metabolic profiling in venous leg ulcerationOnly two studies explored venous leg ulceration via 1H-NMR spectroscopy in small sample sizes. An early study employing 31P-NMR, comparing the results with histological analysis, assessed punch biopsies from venous ulcers, ischaemic and non-ischaemic legs. Venous leg ulcers had lower energy profiles, characterized by reduced levels of phosphocreatine and nucleoside triphosphate, suggestive of a low prevalence of metabolites associated with energy pathways. This may represent an association with poor wound healing, which is a common challenge in these ulcersADDIN BEC{Ennis et al., 1994, #84035}41.More recently, wound exudate analysis was performed via 1H NMR, with associated microbiology assessmentADDIN BEC{Junka et al., 2017, #41}43. The metabolites identified were associated with hypoxia, immune response activation and with wound colonisation, for example the presence of acetate in wounds inhabited by P. aeruginosa (Table 1). This highlights the effect of the local wound microbiome on NMR profiling. The importance of the microbiome in multiple pathologies has been well characterized, both in woundADDIN BEC{Misic et al., 2014, #46750}44, gutADDIN BEC{Li et al., 2008, #29503}45 and vaginalADDIN BEC{Kindinger et al., 2017, #58674}46 microbiota. Furthermore, the microbiome is known to have profound effects on the inflammatory response and the ability of tissues to healADDIN BEC{Johnson et al., 2018, #95297}47.Despite preliminary use of metabolic phenotyping in the context of venous leg ulceration, studies have been untargeted in nature with no validation via alternative technologies; furthermore, statistical analysis has not taken into account formal FDR correction for either study. Further research exploring the translational applications of this technology are still awaited. Importantly, these should be longitudinal in nature to permit serial sampling and assessment of any association with healing status and take into account the wound microenvironment.Metabolic profiling in venous thromboembolic diseaseNine studies relating to venous thromboembolism (VTE) fulfilled the inclusion criteria; of these four were human studies, two murine studies, one a porcine study, one a rabbit study and one a hybrid human / murine study.Of the human studies, only one performed a targeted analysis with metabolite quantificationADDIN BEC{Deguchi et al., 2015, #70}48. Serum samples from patients with an idiopathic VTE and age and sex matched controls underwent liquid chromatography (LC)-MS analysis and assessment of prothrombin activation. Except for warfarin use (controlled for in the statistical analysis), a family history of VTE, and a higher body mass index (BMI), the two populations were well matched. The untargeted (global metabolite discovery) and targeted (specific metabolite identification and quantification) analyses revealed a lower concentration of acylcarnitines in the VTE group, while the mechanistic assays highlighted the anticoagulant activity of acylcarnitines as inhibitors of factor Xa, resulting in an inability to convert prothrombin to thrombin and therefore generate thrombus. A reduction in acylcarnitine levels conferred a prothrombotic status and an increased VTE risk.VTE was explored in the context of trauma in critically ill patients, with untargeted MS plasma analysis performed in 20 patients who developed a VTE in the first 28 days of hospitalisation compared to 20 controls (non-VTE trauma patients)ADDIN BEC{Voils et al., 2018, #22}49. A number of metabolites were differentially expressed, particularly two kynurenine metabolites (Table 2). Pathway analysis revealed that tryptophan metabolism was implicated, with kynurenine metabolism associated with inflammation, oxidative stress, endothelial dysfunction and immune activationADDIN BEC{Karu et al., 2016, #83837}50. Tryptophan and kynurenine metabolisms are also associated with the gut microbiomeADDIN BEC{Agus et al., 2018, #62309}51, highlighting a connection that requires further elucidation. Though acylcarnitines were not statistically significant between the experimental groups, the design of these two human VTE studies was very different.Pulmonary embolism with differential risk stratification (46 low risk and 46 intermediate / high risk) was assessed with MS analysis performed on plasma samples via an untargeted approach, and comparisons made between low vs intermediate/high risk and intermediate vs high risk cases. A number of metabolites were identified belonging to the tricarboxylic acid (TCA) cycle, fatty acid, purine and amino acid metabolism, amongst others (Table 2). Specifically, amino acid, nucleotide and energy metabolisms appeared to be downregulated in intermediate and high-risk PE patientsADDIN BEC{Zeleznik et al., 2018, #26}52. Interestingly, short and medium chain acylcarnitines were found to be reduced in the intermediate/high-risk group compared to the low-risk group.In the hybrid human / murine study, untargeted NMR analysis assessing serum samples obtained from patients with VTE and controls, and from Sprague Dawley rats with experimentally induced deep venous thrombosis (inferior vena cava (IVC) ligation), healthy controls and sham controls identified differential metabolite expression across the different groups. Interestingly, there was a degree of overlap between the murine and human results (Table 2). Overall, pathways of relevance included ketone body, pyruvate, butanoate, phenylalanine, amino acid metabolism and glycine, serine and threonine metabolismADDIN BEC{Cao et al., 2018, #16}53.Of the four animal studies, two employed murine models. One study performed NMR analysis in a DVT electrolytic model in male C57BL/6 mice, with concomitant assessment of P- and E-selectin concentrations in the vein wall, which are cell adhesion molecules and proinflammatory proteinsADDIN BEC{Obi et al., 2016, #61}54. Both young and old mice with DVT and controls were assessed, although no sham control mice were included. Serum, thrombus and vein wall tissue samples were assayed at day 2 post thrombus generation, identifying that older mice with DVT had a larger clot burden than younger ones and a higher abundance of amino acids (Table 2). Vein P-selectin levels were higher in the wall of DVT mice compared to controls, particularly in the older group, highlighting an increased inflammatory burden. Furthermore, increasing P-selectin protein concentration was seen to correlate with increasing thrombus weight. Conversely, E-selectin levels were higher in both older and younger controls. This study highlighted the association between age-related metabolites, the vein wall and vein wall P-selectin, and identified age-related oxidative stress as an important pathway in experimental DVT.In a second animal VTE study, a murine inferior vena cava (IVC) ligation model was employed to investigate the metabolic phenotypes of DVT in disease and sham control mice. Serum and vein wall samples were assayed via NMR and MSADDIN BEC{Sung et al., 2018, #23}55. Ten mice in each group were tested, with samples harvested at day 2 post thrombus generation. A large number of metabolites were identified as differentially expressed, including lipids and acylcarnitines (Table 2). The pathways identified in this study as relevant to experimental DVT were energy metabolism, sphingolipid and adenosine metabolism.Pre- and post- thrombosis serum and thrombus samples from a jugular vein rabbit model were analysed via LC-MS in five rabbits that were sacrificed four hours following DVT generation via endothelial denudation. This animal VTE study identified that differentially abundant metabolites associated to glycolysis, purine, tryptophan and redox metabolism were associated with the presence of new DVT (Table 2); lactate was particularly abundant, possibly reflecting the role of active glycolysis of thrombus component cellsADDIN BEC{Maekawa et al., 2019, #81683}56.A further animal VTE model is the injection of polydextran microspheres through the femoral vein; sixteen pigs underwent baseline blood sampling, followed by further sampling following the development of PEADDIN BEC{Bujak et al., 2014, #97}57. Post-PE samples were compared to baseline samples and revealed an increase in metabolites associated with hypoxia, lipid metabolism, energy metabolism, cell signalling and mitochondrial function (Table 2).Many of the aforementioned human studies have limited sample sizes (Table 2). Recently, the largest VTE metabolic profiling assay was performed, exploring the metabolic phenotype of samples from 240 incident VTE cases and 6963 controls in the context of a prospective, longitudinal nested case control study performed in healthcare professionals. This study was comprehensive, with VTE rates recorded prospectively in individuals free of cardiovascular disease at the time of sample collection, with regression analysis and FDR testing performed to account for confounding variables and to ensure findings were truly statistically significant. Metabolites related to incident VTE were explored, with 60 identified as being of interest. Following correction for multiple testing, only carnitine C5 was identified as discriminant, with diacylglycerols found to be enriched in both VTE and PE. Following adjustment for BMI this did not remain significant, supporting the important role of high BMI as a major risk factor for VTEADDIN BEC{Jiang et al., 2018, #74014}58. Though a very large study in itself, the authors suggest that larger studies controlling for BMI are required to assess the baseline metabolic phenotype of incident VTE in further detail.DiscussionThe studies identified in this review have provided evidence for the role of metabolic phenotyping in the assessment of venous disease and important information on pathways of relevance with respect to the pathophysiology of disease. Nonetheless, it is clear that the current literature has a number of limitations. Studies are varied, including both animal and human work. Both experimental designs are valuable; the ultimate aim of these studies is to identify mechanistic pathways or biomarkers that can be explored for patient benefit. Animal studies are extremely valuable due to the controlled experimental conditions that help reduce the effect of confounding factors, particularly when exploring mechanistic pathways; nonetheless, their direct translational relevance, and ability to predict what may happen in humans with certainty, is limitedADDIN BEC{van der Worp et al., 2010, #80006}59. In this review, animal studies were identified in both CVD and VTE research; however, the interpretation of their results is difficult due to the wide variation in models used, particularly in VTE, with rodent (mouse, rat, rabbit) and porcine experiments performedADDIN BEC{Bujak et al., 2014, #97; Obi et al., 2016, #61; Sung et al., 2018, #23}{Maekawa et al., 2019, #81683}54–57. For animal research to be translational, it should be standardized in methodology and reporting criteria. A collaborative, society-driven consensus advising on this may be a suitable approach; this has recently been performed for mouse models in VTEADDIN BEC{Diaz et al., 2019, #78905}60, but should be applicable to all animal models in the future.Human studies are more representative of the general patient population, though more prone to influence by confounding factors that are difficult to control, even by case matching, such as genetics, diet and stressADDIN BEC{Bujak et al., 2014, #97; Obi et al., 2016, #61; Sung et al., 2018, #23}{Maekawa et al., 2019, #81683}54–57. Furthermore, the studies included in this review explored different research questions, populations and experimental designs, with samples collected via different standard operating procedures and at different time points (Tables 1, 2). These differences make any results from metabolic profiling platforms difficult to interpret. Furthermore, the vast majority of studies are untargeted in nature, identifying the presence of metabolites but providing limited insight on their concentration. Additionally, there is a lack of robust validation via alternative platforms. This study design can provide important mechanistic information that can be developed in future workADDIN BEC{Deguchi et al., 2015, #70}48 ADDIN BEC{Suhre et al., 2011, #81133}61. However, this review has found that only a handful of studies take this approach (Table 1, 2). Independent cohort validation following pilot data was performed in two of the CVD studiesADDIN BEC{Anwar et al., 2012, #113}26 ADDIN BEC{Anwar et al., 2017, #35}31 ADDIN BEC{Tanaka et al., 2010, #137}27 ADDIN BEC{Tanaka et al., 2012, #122}28, though not in venous leg ulceration and VTE.Sample size is an important limiting factor in these studies; large datasets are required to achieve appropriate statistical power and enable assessment of possible confounding variablesADDIN BEC{Maekawa et al., 2019, #81683}56, particularly in human studies, where experimental conditions are not tightly controlled. This can provide findings that are more robust; however, the importance of confounders cannot be underestimated, as these have a significant impact on the results of the analysis. An example is gender; metabolism differences exist between males and females, likely due to hormonal factorsADDIN BEC{Kochhar et al., 2006, #4860}62. Of the studies included in this review, one CVD publication assessed the role of gender, concluding that it had no influence on the metabolic phenotype of cases and controlsADDIN BEC{Anwar et al., 2017, #35}31. None of the VLU publications controlled their analyses for gender; of the VTE human studies, two recruited both male and female participants, adjusting their statistical analyses for genderADDIN BEC{Zeleznik et al., 2018, #26}{Jiang et al., 2018, #74014}52, 58. The influence of gender on the metabolic phenotype was not assessed. Case and control matching are important but can be challenging to perform during recruitment and may not necessarily be representative of real-world data. Sample size calculation methodologies for metabolic phenotyping studies have been developed in recent yearsADDIN BEC{Blaise et al., 2016, #77417}{Nyamundanda et al., 2013, #66665}22, 63 and are tools that can be used to estimate sample size in future studies.Despite this, the aforementioned studies have demonstrated the applicability of metabolic phenotyping to tissue and biofluids, with identification of metabolites and relevant pathways. Metabolites such as acylcarnitines have been shown to play an important role in VTEADDIN BEC{Deguchi et al., 2015, #70; Obi et al., 2016, #61}48, 54, although differential study design has resulted in somewhat inconsistent findings. The limited CVD data has identified energy metabolism, inflammation, hypoxia, amino acid metabolism and oxidative stress as important pathways; these are also of relevance in VTE and VLU. Figure 2 highlights the overlap in metabolic pathways between different venous pathologies. It is important to take this into account when planning further research; clearly, it is not enough to perform untargeted analyses providing metabolite identification when exploring mechanistic processes involved in these conditions. Furthermore, many of the aforementioned metabolites have been identified as candidate biomarkers in other pathological processes, such as atherosclerosis, diabetes, obesity and cancerADDIN BEC{Newgard, 2017, #29232}64 ADDIN BEC{Trivedi et al., 2017, #37225}65. Further studies should demonstrate the involvement of specific pathways by performing enzyme assays, as well as exploring the relative concentrations of metabolites via quantification. It is likely that only a finite number of metabolic pathways are involved in physiological and pathological processes, and it is possible that these are affected to a different extent in different conditions. This would potentially enable the generation of different metabolite cut-off levels, or ratios, for specific disease processes. Finally, multi-omic validation, including proteomic, transcriptomic and genomic assays would be valuable to explore the mechanistic pathways in further detail.Ultimately, study design and experimental methodology need to be more consistent across the literature to enable appropriate data interpretation. Studies should be translational in nature, aiming to address areas where data are lacking, such as CVD progression and recurrence; ulcer healing potential and factors associated with this; VTE diagnosis, prognostication with respect to PTS and response to treatment, whether by conservative or interventional means.ConclusionsMetabolic phenotyping has several clear potential applications to venous disease. Studies to date are largely untargeted, lacking validation via alternative experimental methodologies and include heterogeneous populations and assays. It is important that future studies are designed in a hypothesis-testing fashion with adequate sample sizing, consideration given to a multi-omic approach and embarked upon with clear translational potential.AcknowledgementsInfrastructure support for the Academic Section of Vascular Surgery is provided by the NIHR Imperial Biomedical Research Centre (BRC).Conflict of interest disclosureThe authors declare no competing financial interest.FIGURESFigure 1PRISMA diagram delineating search results.Figure 2 - Representation of metabolites and pathways involved as reported in the included studies. Circle colour relates to the disease process (chronic venous disease / venous leg ulceration / deep venous disease) – text colour relates to the metabolic pathway involved. Evident overlap between pathologies and pathways involved.3-HB – 3-hydroxybutyrate; 3-HIB – 3-hydroxyisobutyrate; AC – acylcarnitine; Alpha KG – alpha ketoglutarate; AMP – adenosine monophosphate; ATP – adenosine triphosphate; GMP – guanine monophosphate; Glucose 6-P – glucose 6-phosphate; GPC – glycerophosphocholine; LPC – lysophosphatidylcholine; NADP+ - nicotinamide adenine dinucleotide phosphate+; N-N DG – dimethylglycine; PC – phosphatidylcholine; PE – phosphatidylethanolamine; PI – phosphatidylinositol; PS – phosphatidylserine; SM – sphingomyelin; TG – triglyceride; VMA – vanlillylmandelateTABLESTable 1 – CVD and VLU Results – Table presenting the list of articles identified by the systematic review and relevant data extracted.Tentative / preliminary metabolite allocations in italics.TABLE 1H /APathology / ModelComparatorAssayT /UNSubstrateStatistical analysisMultiple testing correctionp-value thresholdControl for variablesUpregulated metabolites in casesUpregulated metabolites in controlsDownregulated metabolites in casesComparative testingTanaka 2010 (27)HCVD C2 - n = 4C3 - n = 1C4 - n = 4C5 - n = 1Total n = 10Controls with PAD n = 6MALDI-IMSU16Tissue – GSVT test One-way ANOVATukey’s test< 0.05Lysophosphatidylcholine (LPC); phosphatidylcholine (PC); sphingomyelin (SM) around valvesLPC uniformly distributed; PC ubiquitous-Tanaka 2012(28)HCVD C2 - n = 16C3 - n = 11C4 - n – 21C5 - n = 2Total n = 50 (38 patients)Controls with PADn = 10MALDI-IMSU48 Tissue – GSVOne-way ANOVATukey’s test< 0.05LPC in the intima and media; PC in the intima and media; PC ubiquitous; Triglyceride (TG) in adventitiaPC ubiquitousHistology (HE staining); lipid staining; biochemical quantitation; immunostaining(Anti vWF and anti D2-40) Anwar 2012(26)HCVDC2 – n = 8Non varicose GSV tissuen = 8 MAS NMRU16Tissue – GSVMultivariate (PCA, OPLS)ROC curveCreatine; Myo-inositol; Lactate; GlutamateTG-Tanaka2016(30)AMale Sprague Dawley rats (8 weeks) lymphatic ligation Contralateral limbMALDI-IMSUN/ATissue – Femoral vein wall DescriptiveTukey Kramer test< 0.05LPC; PC; TG in vein wall PC ubiquitousPC ubiquitousImmunohistoche-mistry; Immuno-fluorescent staining; Adipo-cyte detection; TNF alpha and Caspase 3 expressionAnwar 2016(29)AMale Sprague Dawley rats (12 weeks) Stretch IVC – 2 g stretch for 4 hours; 2 g stretch for 18 hoursn = 50.5 g and 2g stretch applied for either 4 or 18 hours (4 groups)NMRMSU5Tissue - IVCMultivariate One-way ANOVASpearman correlationBenjamini Yekutieli FDR correction< 0.05Valine; Choline; TG Increased in veins stretched for 18h compared to non-stretched 18h-Anwar 2017(31, 37)HCVD C2 - n = 4C3 - n = 66C4 - n = 9C5 - n = 1Total n = 80Non – CVD controls n = 35NMRMSU115Tissue – GSV or tributaryMultivariate (PCA, OPLS-DA)Benjamini Yekutieli FDR correction< 0.0001Age, sex past medical history; medications included in multivariate analysisPhosphatidylserines (PS); PC; SM; Phosphatidylethanolamines (PhE); Phosphatidylinositol (PI); Myo-inositol; Inosine; Taurine; Uridine; Creatine; Alanine; Guanosine; Glutamine; GlutamateCeramides Lysophosphocholine (lysoPC)TG Messenger RNA (mRNA) expressionEnnis1994(41)HVLUn = 4Ischaemic skin samples / Controlsn = 13NMRU17Ulcer biopsyOne way ANOVATest of Scheffe < 0.05Phospho-creatineAdenosine triphosphate (ATP)-Junka2017(43)HVLUn = 20-NMRU20Wound fluidMultivariate (PLS -DA)Lactate; Lipids (L1, L2, L3); Glycerophospho-rylcholine (GPC); Lysine; Acetate associated with pseudomonas aeruginosa; Valine; Leucine; Isoleucine?; 3-hydroxybutyrate(3-HB)?; Propylene glycol; Alanine; Ethanol; Aspartate; UreaHistamine; N-N-dimethylglycine; Phenylalanine; Tyrosine; Glutamate; Pyruvate; SuccinateMicrobiologyTable 2 – VTE Results – Table presenting the list of articles identified by the systematic review and relevant data extracted.Tentative / preliminary metabolite allocations in italics.TABLE 2 H /APathology / ModelComparatorAssayT /UNSubstrateStatistical analysisMultiple testing correctionp-value thresholdControl for variablesUpregulated metabolites in casesUpregulated metabolites in controlsDownregulated metabolites in casesComparative testingDeguchi 2015(48)HUnprovoked DVT n = 40Non DVT controls n = 40MS T + U80PlasmaMann Whitney; Spearman; Fisher’s exact probability test< 0.05Acylcarnitines (AC) Palmitoleoyl-carnitine Decenoyl-carnitineThrombosis assaysBujak 2015(57)APig (2-3 month old castrated male) PE(polydextran microspheres 100? 300 μm in diameter injected in femoral vein; increase of mPAP >40mmHg for at least 20 minutes)n = 16Baseline bloodLC – GC – QTOF MSU16PlasmaUnivariate Multivariate (PCA, OPLS DA) Welch’s paired t test Mann Whitney U test < 0.05Pyruvate; Lactate; Glycerol; Palmitic acid; Oleic acid; 3- hydroxybutyric acid; 2-ketoisocaproic acid; Galacturonic acidPC; Sphingosine; Alpha-tocopherol; Leukotriene C4Docasotetraenoic acid; Docosapentaenoic acid; Hydroxyoxohexadecanoic acid; Dihydroxyoctadecadienoic acid; Oxoheptadecatrienoic acid; Dodecadienoic acid; Methyltridecanedioic acid; Ceramide; SM; PS; PA; PI; LPE; LPI; LPA; Creatine; Arginine; Didesmethyl tocotrienol; Desmosine-Obi 2016(54)AMouse (youngand old C57BL/6) DVT (electrolytic IVC model)Young n = 15Old n = 16ControlsNMRU31SerumANOVAT testHolm-?ídák?testp < 0.05Linear regression for association between thrombosis parameters, P and E selectin levels and metabolite concentrationGlutamine; Proline; PhenylalanineVein wall P- and E- selectin levelsCao 2018(53)AHMale Sprague – Dawley ratsDVT (IVC ligation model)n = 10Unprovoked distal DVTn = 61Controls (rats) n = 10Sham controls (rats) n = 10Controls (human) n = 61NMRU30122SerumCV ANOVA Multivariate analysis (PCA, OPLS DA)Student’s t-testFisher’s exact test< 0.05RAT DVTLipids?; Leucine?; Valine?; N-acetylglycoproteins (NAC)?; O-acetylglycoproteins (OAC)?; Acetoacetate?; PyruvateHUMAN DVTLipids; Valine; 3-HB; Lactate; Lysine; Acetate; Glutamine?; Acetoacetate; Pyruvate; Creatine; GPC; Glycine; Tyrosine; Phenylalanine; Formate-RAT DVTLactateAlanineGlucoseMethanolHUMAN DVTNAGAcetoneGlutamateGlucoseMethanolSung 2018(55)AMouse (Male Balb/C)DVT (IVC ligation model) n = 10Sham control n = 10MS and NMRU20SerumIVC wallStudent’s t testMultivariate (PCA, OPLS DA) FDR correction< 0.05SERUMAlphahydroxyisobutyrate; L carnitine; LPC; PC; SM VEIN WALLAcetylcarnitine; Adenosine; Ceramide; GPC; PC -SERUMAlanine; Allantoin; Citrate; Creatine; Fumarate; Glycine; Succinate; Taurine; Tyrosine; Valine; Acetylcarnitine; Adenine; Adenosine; Creatine; Indoxysulfuric acid; L-tyrosine; PC; Taurine; TG VEIN WALLCholine; PA; TG Voils 2018(49)HCritically ill patients with VTE (provoked, blunt trauma); DVT n = 10 and PE n = 10Critically ill patients without VTE n = 20UPLC MSU40PlasmaChi square Fisher’s exact testT testWilcoxon rank sum testsFDR < 0.05N-formylkynurenine; 5-hydroxy N-formylkynurenineIn silico gene expression analysisZeleznik 2018(52)HPE with risk stratification (low n = 46; intermediate n = 28; high n = 18) -UPLC MSU92PlasmaMann Whitney U test Fisher’s exact testBenjamini Hochberg FDR < 0.02Multivariable linear regression to differenti-ate between PE riskLow vs intermediate risk Xanthosine; Alpha-ketoglutarate; Arachidoylcarnitine; Picolinate; Vanillylmandelate (VMA)Intermediate vs high riskPhenylalanine; Lactose; Tyrosine; Caffeine; 3-HIB; Citrate Jiang 2018(58)HIncident VTE DVT n = 240, PE n = 125Controls n = 6963LC-MSU7,203SerumMultivariate (PCA, OPLS DA)Sensitivity analysis, Pearson correlationBonferroni correction< 0.05Logistic regression for VTE association.Adjustment for BMINone significant Maekawa 2019(56)ARabbit (male Japanese white rabbits)DVT (jugular venous thrombus; jugular endothelial denudation and vessel ligation)n = 5-CE-TOFMSU5PlasmaThrombusUnpaired t testOne way ANOVASpearman’s testBonferroni correction< 0.05PLASMACitrate; Glucose-6-phosphate (glucose 6-P); Nicotinamide adenine dinucleotide phosphate+ (NADP+); Tryptophan; Fructose-6-phosphateTHROMBUSLactate; Adenosine monophosphate (AMP); Choline; Hypoxanthine; Guanine monophosphate (GMP); Guanine; GlutathioneThromboelastometryCellular contentsAnd Glut 1 expression in human aspirated DVTHistologyTable 1 and 2 abbreviations: 3-HB – 3-hydroxybutyrate; 3-HIB – 3-hydroxyisobutyrate; A – animal; AC – acylcarnitine; Alpha KG – alpha ketoglutarate; AMP – adenosine monophosphate; ANOVA – analysis of variance; ATP – adenosine triphosphate; CE-TOF MS – capillary electrophoresis time of flight; CV-ANOVA – cross-validated residuals analysis of variance; CVD – chronic venous disease; C2/3/4/5 – CEAP stages; GC – gas chromatography; GMP – guanine monophosphate; Glucose 6-P – glucose 6-phosphate; GPC – glycerophosphocholine; H – human; HE – haematoxylin-eosin; FDR – false discovery rate; IVC – inferior vena cava; LPA – lysophosphatidic acid; LPC – lysophosphatidylcholine; LPE – lysophosphatidylethanolamine; LPI – lysophosphatidylinositol; lysoPC – lysophosphocholine; MAS – magic angle spinning; mPAP - mean pulmonary arterial pressure; NAC – N-acetylglycoproteins; NADP+ - nicotinamide adenine dinucleotide phosphate+; NAG – N-acetylglutamate; OAC - O-acetylglycoproteins; OPLS – DA – orthogonal partial least square regression discriminant analysis; PA – phosphatidic acid; PAD – peripheral arterial disease; PC – phosphatidylcholine; PCA – principal component analysis; PE – pulmonary embolism; PhE – phosphatidylethanolamine; PI – phosphatidylinositol; PS – phosphatidylserine; QTOF – quad time of flight; ROC – receiver operating characteristic curve analysis; SM – sphingomyelin; T – targeted; TG – triglyceride; TMAO – trimethylamine N-oxide; TNF – tumour necrosis factor; U – untargeted; VLU – venous leg ulceration; VMA – vanlillylmandelate; VTE – venous thromboembolism; vWF – von Willebrand FactorFull tables in Supplement 2ASSOCIATED CONTENTSupporting Information. The following files are available free of charge.Supplement 1; Search terms for systematic review (PDF).AUTHOR INFORMATIONCorresponding AuthorMiss Sarah OnidaNIHR Clinical Lecturer in Vascular Surgery, Academic Section of Vascular SurgeryDepartment of Surgery and Cancer, Imperial College London s.onida@imperial.ac.ukAuthor ContributionsThe manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. Funding SourcesNo funding sources to report for this publication.ABBREVIATIONS3-HB – 3-hydroxybutyrate; 3-HIB – 3-hydroxyisobutyrate; A – animal; AC – acylcarnitine; Alpha KG – alpha ketoglutarate; AMP – adenosine monophosphate; ANOVA – analysis of variance; ATP – adenosine triphosphate; BCAA, branched chain amino acid; BMI, body mass index; CE-TOF MS – capillary electrophoresis time of flight; CV-ANOVA – cross-validated residuals analysis of variance; CVD – chronic venous disease; C2/3/4/5 – CEAP stages; FDR – false discovery rate; GC – gas chromatography; GMP – guanine monophosphate; Glucose 6-P – glucose 6-phosphate; GPC – glycerophosphocholine; GSV, great saphenous vein; H – human; HE – haematoxylin-eosin; IMS, imaging mass spectrometry; IVC – inferior vena cava; LPA – lysophosphatidic acid; LPC – lysophosphatidylcholine; LPE – lysophosphatidylethanolamine; LPI – lysophosphatidylinositol; lysoPC – lysophosphocholine; MALDI, matrix assisted laser desorption ionisation; MAS – magic angle spinning; MMP, matrix metalloproteinase; mPAP - mean pulmonary arterial pressure; MS, mass spectrometry; NAC – N-acetylglycoproteins; NADP+ - nicotinamide adenine dinucleotide phosphate+; NAG – N-acetylglutamate; N-NDG, N-N dimethylglycine; NMR, nuclear magnetic resonance spectroscopy; OAC - O-acetylglycoproteins; OPLS – DA – orthogonal partial least square regression discriminant analysis; PA – phosphatidic acid; PAD – peripheral arterial disease; PC – phosphatidylcholine; PCA – principal component analysis; PE – pulmonary embolism; PhE – phosphatidylethanolamine; PI – phosphatidylinositol; PRISMA, preferred reporting items for systematic reviews and meta-analyses; PS – phosphatidylserine; PTS, post thrombotic syndrome; QTOF – quad time of flight; RNA, ribonucleic acid; ROC – receiver operating characteristic curve analysis; SM – sphingomyelin; T – targeted; TCA, tricarboxylic acid; TG – triglyceride; TIMP, tissue inhibitor of matrix metalloproteinase; TMAO – trimethylamine N-oxide; TNF – tumour necrosis factor; U – untargeted; UPLC, ultra-performance liquid chromatography; VLU – venous leg ulceration; VMA – vanlillylmandelate; VTE – venous thromboembolism; vWF – von Willebrand FactorReferencesADDIN BB1.Evans CJ, Fowkes FG, Ruckley CV, Lee AJ. Prevalence of varicose veins and chronic venous insufficiency in men and women in the general population: Edinburgh Vein Study. J Epidemiol Community Health. 1999;53:149-153.2.Lal BK. Venous ulcers of the lower extremity: Definition, epidemiology, and economic and social burdens. Semin Vasc Surg. 2015;28:3-5.3.Rabe E, Pannier F. Societal costs of chronic venous disease in CEAP C4, C5, C6 disease. Phlebology. 2010;25 Suppl 1:64-67.4.Heit JA. Epidemiology of venous thromboembolism. Nat Rev Cardiol. 2015;12:464-474.5.Hunt BJ. The prevention of hospital-acquired venous thromboembolism in the United Kingdom. Br J Haematol. 2009;144:642-652.6.Paik B, Joh JH, Park HC. Anatomic and clinical risk factors for pulmonary embolism in patients with deep venous thrombosis of the lower extremity. Ann Surg Treat Res. 2017;92:365-369.7.Kahn SR. Frequency and determinants of the postthrombotic syndrome after venous thromboembolism. Curr Opin Pulm Med. 2006;12:299-303.8.Ghanima W, Wik HS, Tavoly M, Enden T, Jelsness-J?rgensen LP. Late consequences of venous thromboembolism: Measuring quality of life after deep vein thrombosis and pulmonary embolism. Thromb Res. 2018;164:170-176.9.Kearon C, Akl EA, Ornelas J et al. Antithrombotic Therapy for VTE Disease: CHEST Guideline and Expert Panel Report. Chest. 2016;149:315-352.10.Vedantham S, Goldhaber SZ, Julian JA et al. Pharmacomechanical Catheter-Directed Thrombolysis for Deep-Vein Thrombosis. N Engl J Med. 2017;377:2240-2252.erota AJ, Kearon C, Gu CS et al. Endovascular Thrombus Removal for Acute Iliofemoral Deep Vein Thrombosis. Circulation. 2019;139:1162-1173.12.(NICE) NIOHACE. Varicose veins: diagnosis and management [CG168]. 201313.Wittens CDAH, Davies AH, B?kgaard N et al. Editor’s choice-management of chronic venous disease: clinical practice guidelines of the European Society for Vascular Surgery (ESVS). European Journal of Vascular and Endovascular Surgery. 2015;49:678-737.14.Pannier F, Rabe E. Progression of chronic venous disorders: results from the Bonn Vein Study. Journal of Vascular Surgery. 2011;53:254-255.15.Nicholson JK, Holmes E, Kinross JM, Darzi AW, Takats Z, Lindon JC. Metabolic phenotyping in clinical and surgical environments. Nature. 2012;491:384.16.Bush RG, Bush P, Flanagan J et al. Factors associated with recurrence of varicose veins after thermal ablation: results of the recurrent veins after thermal ablation study. The Scientific World Journal. 2014;201417.Rabe E, Pannier F. Societal costs of chronic venous disease in CEAP C4, C5, C6 disease. Phlebology. 2010;25:64-67.18.Gohel MS, Taylor M, Earnshaw JJ, Heather BP, Poskitt KR, Whyman MR. Risk factors for delayed healing and recurrence of chronic venous leg ulcers--an analysis of 1324 legs. Eur J Vasc Endovasc Surg. 2005;29:74-77.19.Mannello F, Ligi D, Canale M, Raffetto JD. Omics profiles in chronic venous ulcer wound fluid: innovative applications for translational medicine. Expert review of molecular diagnostics. 2014;14:737-762.20.Jensen SB, Hindberg K, Solomon T et al. Discovery of novel plasma biomarkers for future incident venous thromboembolism by untargeted synchronous precursor selection mass spectrometry proteomics. Journal of Thrombosis and Haemostasis. 2018;16:1763-1774.21.Vorkas PA, Shalhoub J, Lewis MR et al. Metabolic Phenotypes of Carotid Atherosclerotic Plaques Relate to Stroke Risk: An Exploratory Study. Eur J Vasc Endovasc Surg. 2016;52:5-10.22.Blaise BJ, Correia G, Tin A et al. Power Analysis and Sample Size Determination in Metabolic Phenotyping. Anal Chem. 2016;88:5179-5188.23.Vorkas PA, Shalhoub J, Isaac G et al. Metabolic phenotyping of atherosclerotic plaques reveals latent associations between free cholesterol and ceramide metabolism in atherogenesis. J Proteome Res. 2015;14:1389-1399.24.Vorkas PA, Isaac G, Anwar MA et al. Untargeted UPLC-MS profiling pipeline to expand tissue metabolome coverage: application to cardiovascular disease. Anal Chem. 2015;87:4184-4193.25.Emwas AH. The strengths and weaknesses of NMR spectroscopy and mass spectrometry with particular focus on metabolomics research. Methods Mol Biol. 2015;1277:161-193.26.Anwar MA, Shalhoub J, Vorkas PA et al. In-vitro identification of distinctive metabolic signatures of intact varicose vein tissue via magic angle spinning nuclear magnetic resonance spectroscopy. Eur J Vasc Endovasc Surg. 2012;44:442-450.27.Tanaka H, Zaima N, Yamamoto N et al. Imaging mass spectrometry reveals unique lipid distribution in primary varicose veins. Eur J Vasc Endovasc Surg. 2010;40:657-663.28.Tanaka H, Zaima N, Sasaki T et al. Loss of lymphatic vessels and regional lipid accumulation is associated with great saphenous vein incompetence. J Vasc Surg. 2012;55:1440-1448.29.Anwar MA, Vorkas PA, Li J et al. Prolonged Mechanical Circumferential Stretch Induces Metabolic Changes in Rat Inferior Vena Cava. Eur J Vasc Endovasc Surg. 2016;52:544-552.30.Tanaka H, Yamamoto N, Suzuki M et al. Insufficient Lymph Drainage Causes Abnormal Lipid Accumulation and Vein Wall Degeneration. Ann Vasc Dis. 2016;9:277-284.31.Anwar MA, Adesina-Georgiadis KN, Spagou K et al. A comprehensive characterisation of the metabolic profile of varicose veins; implications in elaborating plausible cellular pathways for disease pathogenesis. Sci Rep. 2017;7:2989.32.Ekl?f B, Rutherford RB, Bergan JJ et al. Revision of the CEAP classification for chronic venous disorders: consensus statement. J Vasc Surg. 2004;40:1248-1252.33.Wallimann T, Wyss M, Brdiczka D, Nicolay K, Eppenberger HM. Intracellular compartmentation, structure and function of creatine kinase isoenzymes in tissues with high and fluctuating energy demands: the’phosphocreatine circuit’for cellular energy homeostasis. Biochemical Journal. 1992;281:21.34.Chen Y, Peng W, Raffetto JD, Khalil RA. Matrix metalloproteinases in remodeling of lower extremity veins and chronic venous disease. Progress in molecular biology and translational science 147. Elsevier; 2017. p. 267-299.35.Lim CS, Kiriakidis S, Paleolog EM, Davies AH. Increased activation of the hypoxia-inducible factor pathway in varicose veins. Journal of vascular surgery. 2012;55:1427-1439. e1.36.Lim CS, Kiriakidis S, Sandison A, Paleolog EM, Davies AH. Hypoxia-inducible factor pathway and diseases of the vascular wall. Journal of vascular surgery. 2013;58:219-230.37.Anwar M, Davies AH, Holmes E, Want E. Characterisation of biological factors in the pathogenesis of varicose veins [dissertation]. Imperial College London; 2013.38.MacColl E, Khalil RA. Matrix metalloproteinases as regulators of vein structure and function: implications in chronic venous disease. Journal of Pharmacology and Experimental Therapeutics. 2015;355:410-428.39.Farrow W. Phlebolymphedema–a common underdiagnosed and undertreated problem in the wound care clinic. The Journal of the American College of Certified Wound Specialists. 2010;2:14-23.40.Serralheiro P, Novais A, Cairr?o E, Maia C, Costa Almeida C, Verde I. Variability of MMP/TIMP and TGF-β1 receptors throughout the clinical progression of chronic venous disease. International journal of molecular sciences. 2017;19:6.41.Ennis WJ, Driscoll DM, Meneses P. A PRELIMINARY-STUDY ON P-31 NMR-SPECTROSCOPY-A POWERFUL TOOL FOR WOUND ANALYSIS USING HIGH-ENERGY PHOSPHATES. WOUNDS-A COMPENDIUM OF CLINICAL RESEARCH AND PRACTICE. 1994;6:166-173.42.Noble WS. How does multiple testing correction work. Nat Biotechnol. 2009;27:1135-1137.43.Junka A, Wojtowicz W, Zabek A et al. Metabolic profiles of exudates from chronic leg ulcerations. J Pharm Biomed Anal. 2017;137:13-22.44.Misic AM, Gardner SE, Grice EA. The wound microbiome: modern approaches to examining the role of microorganisms in impaired chronic wound healing. Advances in wound care. 2014;3:502-510.45.Li M, Wang B, Zhang M et al. Symbiotic gut microbes modulate human metabolic phenotypes. Proceedings of the National Academy of Sciences. 2008;105:2117-2122.46.Kindinger LM, Bennett PR, Lee YS et al. The interaction between vaginal microbiota, cervical length, and vaginal progesterone treatment for preterm birth risk. Microbiome. 2017;5:6.47.Johnson T, Gómez B, McIntyre M et al. The cutaneous microbiome and wounds: New molecular targets to promote wound healing. International journal of molecular sciences. 2018;19:2699.48.Deguchi H, Banerjee Y, Trauger S et al. Acylcarnitines are anticoagulants that inhibit factor Xa and are reduced in venous thrombosis, based on metabolomics data. Blood. 2015;126:1595-1600.49.Voils SA, Shahin MH, Garrett TJ, Frye RF. Metabolomic association between venous thromboembolism in critically ill trauma patients and kynurenine pathway of tryptophan metabolism. Thromb Res. 2018;165:6-13.50.Karu N, McKercher C, Nichols DS et al. Tryptophan metabolism, its relation to inflammation and stress markers and association with psychological and cognitive functioning: Tasmanian Chronic Kidney Disease pilot study. BMC nephrology. 2016;17:171.51.Agus A, Planchais J, Sokol H. Gut Microbiota Regulation of Tryptophan Metabolism in Health and Disease. Cell Host Microbe. 2018;23:716-724.52.Zeleznik OA, Poole EM, Lindstrom S et al. Metabolomic analysis of 92 pulmonary embolism patients from a nested case-control study identifies metabolites associated with adverse clinical outcomes. J Thromb Haemost. 2018;16:500-507.53.Cao J, Jin QQ, Wang GM et al. Comparison of the serum metabolic signatures based on (1)H NMR between patients and a rat model of deep vein thrombosis. Sci Rep. 2018;8:7837.54.Obi AT, Stringer KA, Diaz JA et al. 1D-(1)H-nuclear magnetic resonance metabolomics reveals age-related changes in metabolites associated with experimental venous thrombosis. J Vasc Surg Venous Lymphat Disord. 2016;4:221-230.55.Sung Y, Spagou K, Kafeza M et al. Deep Vein Thrombosis Exhibits Characteristic Serum and Vein Wall Metabolic Phenotypes in the Inferior Vena Cava Ligation Mouse Model. Eur J Vasc Endovasc Surg. 2018;55:703-713.56.Maekawa K, Sugita C, Yamashita A et al. Higher lactate and purine metabolite levels in erythrocyte-rich fresh venous thrombus: Potential markers for early deep vein thrombosis. Thromb Res. 2019;177:136-144.57.Bujak R, Garcia-Alvarez A, Ruperez FJ et al. Metabolomics reveals metabolite changes in acute pulmonary embolism. J Proteome Res. 2014;13:805-816.58.Jiang X, Zeleznik OA, Lindstr?m S et al. Metabolites Associated With the Risk of Incident Venous Thromboembolism: A Metabolomic Analysis. J Am Heart Assoc. 2018;7:e010317.59.van der Worp HB, Howells DW, Sena ES et al. Can animal models of disease reliably inform human studies. PLoS Med. 2010;7:e1000245.60.Diaz JA, Saha P, Cooley B et al. Choosing a Mouse Model of Venous Thrombosis. Arterioscler Thromb Vasc Biol. 2019;39:311-318.61.Suhre K, Shin SY, Petersen AK et al. Human metabolic individuality in biomedical and pharmaceutical research. Nature. 2011;477:54-60.62.Kochhar S, Jacobs DM, Ramadan Z, Berruex F, Fuerholz A, Fay LB. Probing gender-specific metabolism differences in humans by nuclear magnetic resonance-based metabonomics. Anal Biochem. 2006;352:274-281.63.Nyamundanda G, Gormley IC, Fan Y, Gallagher WM, Brennan L. MetSizeR: selecting the optimal sample size for metabolomic studies using an analysis based approach. BMC Bioinformatics. 2013;14:338.64.Newgard CB. Metabolomics and Metabolic Diseases: Where Do We Stand. Cell Metab. 2017;25:43-56.65.Trivedi DK, Hollywood KA, Goodacre R. Metabolomics for the masses: The future of metabolomics in a personalized world. New Horiz Transl Med. 2017;3:294-305.Table of Contents (TOC) GraphicFor TOC Only3 x 2 inches TOC imageFor TOC Only ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download