METHOD OF DIAGNOSIS OR PROGNOSIS OF A NON-HEALING OR CHRONIC WOUND

Abstract

The present invention relates to a method of diagnosis or prognosis of a non-healing or chronic wound comprising the step of determining the ratio of carnitine to ceramide and/or ceramide derivative in a sample, or the level of at least one metabolite in a sample.

Claims

1. A method of diagnosis or prognosis of a non-healing or chronic wound comprising the step of determining the ratio of carnitine to ceramide and/or ceramide derivate in a sample from a mammalian, wherein said ratio is decreased when compared with a reference value.

2. The method of diagnosis or prognosis of a non-healing or chronic wound according to claim 1, comprising the following steps: a) determining the ratio of carnitine to ceramide and/or ceramide derivate in a sample from a mammalian, b) comparing the ratio to a reference value, c) providing a favorable diagnosis or prognosis of healing when the ratio is higher than said reference value, or providing a poor diagnosis or prognosis of healing when the ratio is lower than said reference value.

3. The method of diagnosis or prognosis of a non-healing or chronic wound according to claim 1, wherein said reference value is 40.

4. The method of diagnosis or prognosis of a non-healing or chronic wound according to claim 1, wherein said reference value has been determined in people who healed.

5. The method of diagnosis or prognosis of a non-healing or chronic wound according to claim 1, wherein the said ceramide is Cer d18:1/24:0, Cer d18:1/24:1, Cer d18:1/23:0, Cer d18:2/23:0, Cer d18:1/22:0, and/or Cer d18:2/22:0.

6. The method of diagnosis or prognosis of a non-healing or chronic wound according to claim 1, wherein the ceramide derivative is a ceramide-1-phosphate, or a sphingomyelin.

7. The method of diagnosis or prognosis of a non-healing or chronic wound according to claim 6, wherein the ceramide-1-phosphate is CerP d18:1/18:0.

8. The method of diagnosis or prognosis of a non-healing or chronic wound according to claim 6, wherein the sphingomyelin is SM d18:1/23:0.

9. The method of diagnosis or prognosis of a non-healing or chronic wound according to claim 1, wherein the wound is an ulcer.

10. The method of diagnosis or prognosis of a non-healing or chronic wound according to claim 1, wherein the sample is chosen among: blood, serum, urine and ulcer fluids.

11. A method of diagnosis or prognosis of a non-healing or chronic wound comprising the step of determining the level of at least one metabolite in a sample from a mammalian, wherein: at least carnitine is lower than a reference value, and/or at least one ceramide and/or one ceramide derivative is higher than a reference value, and/or at least one phosphatidylethanolamine is higher than a reference value.

12. The method of diagnosis or prognosis of a non-healing or chronic wound according to claim 11, wherein: the ceramide is Cer d18:1/24:0, Cer d18:1/24:1, Cer d18:1/23:0, Cer d18:2/23:0, Cer d18:1/22:0, and/or Cer d18:2/22:0, and/or the ceramide derivative is: a ceramide-1-phosphate, preferably CerP d18:1/18:0, and/or a sphingomyelin, preferably SM d18:1/23:0, and/or the phosphatidylethanolamine is PE 18:4/22:6.

Description

FIGURE LEGENDS

[0113] FIG. 1 represents multivariate OPLS-DA cross-validated scores for metabolic profiling of serum, urine and ulcer fluid samples of non-healed (red) and Healed (blue) vein leg ulcer patients. (A) positive mode (ESI+) HILIC-MS and (B) negative mode (ESI−) HILIC-MS profiling analysis of serum; (C) negative mode (ESI−) of lipid RPLC-MS profiling analysis of serum; (D) positive mode (ESI+) RPLC-MS and (E) negative mode (ESI−) of RPLC-MS profiling analysis of urine; (F) negative mode (ESI−) of lipid RPLC-MS profiling analysis of ulcer fluid. The R.sup.2 and Q.sup.2 values for OPLS-DA scores plots are shown in Table 2.

[0114] FIG. 2 represents carnitine to ceramide ratio as a predictive biomarker of ulcer healing. Significance was measured via Student's unpaired T-test (p-value<0.0001)

EXAMPLES

A—Methods

1. Obtention of the Sample

[0115] Samples consisting of serum, urine and ulcer fluids were obtained from 28 patients suffering from a venous leg ulcer. After 20 weeks follow up, 15 patients had healed venous leg ulcers and 13 were unhealed.

2. Experimental Analysis

Sample Preparation and UPLC-MS Analysis

[0116] Serum and urine samples were analysed by Reversed phase (RP) and hydrophilic interaction liquid chromatography (HILIC) UPLC-MS profiling methods. For ulcer fluid profiling, organic and aqueous metabolites were extracted, for example as described by Anwar et al. 2015. Lipid profiling of serum and ulcer fluid organic metabolite extracts was performed using RP-UPLC-MS as described by Isaac et al. 2011. HILIC-UPLC-MS metabolic profiling of serum and urine was conducted, for example as described by Spagou et al. 2011. The separated samples were analysed with electrospray ionisation (ESI)+ and ESI− modes.

Sample Preparation and NMR Analysis

[0117] For NMR analysis of serum and the ulcer fluid, aqueous extract buffer solution of 0.075 M Na.sub.2HPO.sub.4×7H.sub.2O, 4% sodium azide (NaN.sub.3) in H.sub.2O, 20% deuterium oxide (D.sub.2O), and 0.08% 3-(trimethyl-silyl) propionic acid-d4 (TSP) was added to serum and ulcer fluid aqueous extract samples. For NMR urine analysis, a 0.075M Na.sub.2HPO.sub.4×7H.sub.2O, 17% deuterium oxide (D.sub.2O) containing 0.1% TSP was added to the urine samples. .sup.1H-NMR data acquisition was performed as previously described by Dona et al. 2014.

Data Processing and Statistical Analysis

[0118] Selected characteristics of the study populations were described and compared according to healing status. Differences between the healing groups were tested using the Kruskal-Wallis one-way analysis of variance.

[0119] The raw MS data were collected in centroid mode and converted to netCDF format using the DataBridge tool implemented in MassLynx™ software (Waters Corporation, Milford, USA). The data were pre-processed using the freely available package in R programming software, XCMS (Smith et al. 2006, Tautenhahn et al. 2008) and an output table was obtained comprising pairs of mass/charge ratio (m/z), retention time (RT) and intensity values of the detected metabolite features in each sample. The resulting data underwent total area normalization. Then, univariate statistical analysis was applied to each discriminatory metabolite followed by a two tailed Student t-test with a Benjamini-Hochberg multiple tests (adjusted p-value). An adjusted p-value of less than 0.05 was assigned as significant.

Metabolite Assignment/Metabolic Profiling Analysis

[0120] Metabolite identification by MS was conducted by matching accurate m/z measurements of detected chromatographic peaks and molecular fragments.

[0121] Five experiments were performed for urine and serum analysis: NMR, reversed-phase UPLC-MS in ESI+ and ESI− modes, HILIC-UPLC-MS in ESI+ and ESI− modes. Three experiments were performed for each ulcer fluid sample: NMR, lipid profiling and reversed-phase UPLC-MS in ESI+ and ESI− modes.

B—Results

Patient Demographics

[0122] Demographic and clinical details of patients that healed and failed to heal are presented in Table 1.

TABLE-US-00001 TABLE 1 Patients demographics Healed Non-Healed (N = 15) (N = 13) P value Age, years, (SD) 73 (12) 71 (16) 0.7191 Male, n (%) 9 (60) 9 (69) 0.5024 Ulcer Area, cm.sup.3, (SD) 27 (38) 65 (50) 0.0333 Ulcer Perimeter, cm, (SD) 13.6 (15.7) 27.86 (16.48) 0.0281 Ulcer Age, months, (SD) 27 (39) 65 (50) 0.0333

[0123] Table 1 shows that there is no statistically significant differences in age distribution or gender ratio between healed and non-healed groups analysed. However, there are significant differences in ulcer area, perimeter and age between healed and non-healed groups.

Reversed Phase UPLC-MS Analysis of Serum, Urine and Ulcer Fluid of the Venous Leg Ulcer Patients

[0124] Serum, urine and ulcer fluid samples were analysed using reversed phase UPLC-MS. The serum analysis revealed 8,296 metabolic features in the positive ionisation mode (ESI+), whilst 2,929 features were detected in the negative mode (ESI−). The urine examined showed 5,263 metabolic features in ESI+, whereas 7,026 features in ESI−. The ulcer fluid highlighted 16,457 metabolic features in ESI+ and 7,204 features in ESI−. Integrals of metabolic features with coefficient of variation (CV) more than 30% in the quality control (QC) samples were removed. Moreover, contaminant peaks in either QCs and blank controls were also excluded. The retained metabolic features were used for the final stages of analyses. Initial differences between healed and non-healed groups were highlighted when OPLS-DA models were performed.

[0125] The significance and robustness of the corresponding OPLS-DA of the serum (ESI−, p=0.0005 for R.sup.2Y=90% and Q.sup.2Y=57%), urine (ESI+, p=0.046 for R.sup.2Y=64% and Q.sup.2Y=25% and ESI−, p=0.024 for R.sup.2Y=87% and Q.sup.2Y=31%) and ulcer fluid (ESI−, p=0.047 for R.sup.2Y=58% and Q.sup.2Y=20%) were confirmed by cross-validated ANOVA (CV-ANOVA) testing; see Table 2 and FIG. 1. The s-plot in SIMCA was used to visualise metabolic features that contributed to the differences in the significant OPLS-DA models. The integrals of the metabolic features responsible for the class discrimination (i.e. healed vs non-healed) were further subjected to statistical testing and the lipid profiling analysis of the serum has leaded the inventors to identify eight unique discriminant between the healed versus the non-healed group. Features included six ceramides (Cer d18:1/24:0, d18:1/24:1, d18:1/23:0, d18:2/23:0, d18:1/22:0, and d18:2/22:0), one ceramide-1-phosphate (CerP d18:1/18:0) and one sphingomyelin (SM d18:1/23:0). Overall, six different ceramide and ceramide-1-phosphate (CerP d18:1/18:0) intensities were significantly shown to be reduced in the healed patients versus the non-healed. SM d18:1/23:0 also showed significantly reduced intensities in the healed versus non healed group. Results are shown in Table 2.

TABLE-US-00002 TABLE 2 Summary of model characteristics from significant OPLS-DA multivariate statistical analysis of all analysed data. Sample CV type Analyses Components R2X R2Y Q2 ANOVA Serum HILIC-MS ESI+ 1 + 2 47.20% 86.00% 43.90% 6.66E−03 HILIC-MS ESI− 1 + 2 24.70% 96.60% 34.20% 3.76E−02 RPLC-MS ESI− 1 + 1 35.50% 90.40% 57.10% 5.25E−04 Urine RPLC-MS ESI+ 1 + 1 24.20% 64.40% 24.90% 4.55E−02 RPLC-MS ESI− 1 + 2 27.30% 87.40% 31.90% 2.78E−02 Ulcer fluid RPLC-MS ESI− 1 + 1 30.30% 58.10% 19.70% 4.72E−02

HILIC-UPLC-MS Analysis of Serum, Urine and Ulcer Fluid of the Venous Leg Ulcer Patients

[0126] Serum and urine samples were further examined using HILIC-UPLC-MS. Overall, 4,302 metabolic features were identified in the ESI+, whilst 3,623 features were observed in the ESI− of the run serum samples. The urine samples showed 15,637 metabolic features in the ESI+ and 5,491 metabolic features in ESI−. Similar to the RP-UPLC-MS analysis, metabolic peaks with coefficient of variance (CV)>30% across the QCs plus the contaminants' peaks in either the QCs or blanks were removed. This resulted in a remainder of 3,410 features for ESI+ and 2,442 features for the ESI− of the serum samples. In the urine samples, the ESI+ had 12,394 features and the ESI− had 3,701 features.

[0127] OPLS-DA modes were performed, and results are shown in FIG. 1. Differences were shown in the profiles derived from the serum samples when the corresponding OPLS-DA modes were performed. CV-ANOVA testing confirmed the robustness of the OPLS-DA models across the profiles generated from the serum (ESI+, p=0.006 for R.sup.2Y=86% and Q.sup.2Y=44% and ESI−, p=0.034 for R.sup.2Y=97% and Q.sup.2Y=34%). The SIMCA derived s-plot visualised the metabolic features highlighted by the significant OPLS-DA models generated from the serum HILIC-UPLC-MS profiles. Further statistical testing was performed on the integrals of the metabolic features that differentiate between classes (i.e. healed vs non-healed).

[0128] These analyses have shown that significant metabolic features included carnitine and phosphatidylethanolamine (PE 18:4/22:6). In total, the integral intensity of carnitine was significantly elevated in the healed group compared to the intensity of the non-healed. Furthermore, PE 18:4/22:6 intensity was significantly reduced in the healed patient serum samples compared to the non-healed samples.

UPLC-MS Metabolic Profiling of Serum The metabolic profiles of serum were analysed using UPLC-MS. Results are shown in Table 3.

TABLE-US-00003 TABLE 3 Discriminatory metabolites observed in serum via UPLC-MS Significance measured via Student's T- test with Benjamini-Hochberg multiple test correction adjusted Metabolite Fold Change(H/NH) P value P value Carnitine 1.840284485 8.76E−04 3.75E−02 Cer(d18:1/24:0) 0.790378672 3.18E−04 2.80E−02 CerP(d18:1/18:0) 0.782541073 9.10E−04 4.17E−02 Cer(d18:2/23:0) 0.781021223 3.35E−04 2.80E−02 Cer(d18:1/22:0) 0.744297674 2.45E−04 2.68E−02 Cer(d18:1/24:1) 0.738650889 1.23E−03 4.88E−02 Cer(d18:1/23:0) 0.722764682 1.97E−04 2.68E−02 Cer(d18:2/22:0) 0.656515972 2.32E−04 2.68E−02 SM(d18:1/23:0) 0.654286489 9.28E−04 4.17E−02 PE (18:4/22:6) 0.349720924 9.68E−04 3.75E−02 Abbreviations: Cer, ceramide; CerP, ceramide-1-phosphate; SM, Sphingomyelin; PE, phosphatidylethanolamine; H, healers; NH, non-healers.

Ratio of Carnitine to Ceramide as a Serum Biomarker Predictive of Ulcer Healing

[0129] Serum-based carnitine to ceramide ratio has also been determined in patients. The results are shown in FIG. 2 and highlight that healed leg ulcer patients have a significantly (p-value<0.0001) elevated carnitine to ceramide ratio compared to non-healing patients.

REFERENCES

[0130] Anwar, M. A. et al., 2015. Optimization of metabolite extraction of human vein tissue for ultra performance liquid chromatography-mass spectrometry and nuclear magnetic resonance-based untargeted metabolic profiling. Analyst, 140(22), pp. 7586-7597. [0131] Broszczak, D. A. et al., 2017. Molecular Aspects of Wound Healing and the Rise of Venous Leg Ulceration: Omics Approaches to Enhance Knowledge and Aid Diagnostic Discovery. The Clinical Biochemist Reviews, 38(1), p. 35. [0132] Callam, M. J. et al., 1985. Chronic ulceration of the leg: extent of the problem and provision of care. Br Med J (Clin Res Ed), 290(6485), pp. 1855-1856. [0133] Dona, A. C. et al., 2014. Precision high-throughput proton NMR spectroscopy of human urine, serum, and plasma for large-scale metabolic phenotyping. Analytical chemistry, 86(19), pp. 9887-9894. [0134] Ennis, W. J., Driscoll, D. M. & Meneses, P., 1994. 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, 6(5), pp. 166-173. [0135] Gohel, M. S. et al., 2005. Randomized clinical trial of compression plus surgery versus compression alone in chronic venous ulceration (ESCHAR study)—haemodynamic and anatomical changes. British Journal of Surgery: Incorporating European Journal of Surgery and Swiss Surgery, 92(3), pp. 291-297. [0136] Gohel, M. S. et al., 2018. A randomized trial of early endovenous ablation in venous ulceration. New England Journal of Medicine. [0137] Gottrup, F. et al., 2001. A new concept of a multidisciplinary wound healing center and a national expert function of wound healing. Archives of Surgery, 136(7), pp. 765-772. [0138] Hampton, S., 2016. The difficulty and the solution of compression therapy in a healed venous leg ulcer. British journal of community nursing, 21(Sup9), pp. S34-S38. [0139] Isaac, G., McDonald, S. & Astarita, G., 2011. Lipid separation using UPLC with charged surface hybrid technology. Waters App note. [0140] James, T. J. et al., 2003. Evidence of oxidative stress in chronic venous ulcers. Wound Repair and Regeneration, 11(3), pp. 172-176. [0141] Junka, A. et al., 2017. Metabolic profiles of exudates from chronic leg ulcerations. Journal of pharmaceutical and biomedical analysis, 137, pp. 13-22. [0142] Kalkhof, S. et al., 2014. Proteomics and metabolomics for in situ monitoring of wound healing. BioMed research international, 2014. [0143] Mannello, F. et al., 2014. Omics profiles in chronic venous ulcer wound fluid: innovative applications for translational medicine. Expert review of molecular diagnostics, 14(6), pp. 737-762. [0144] Onida, S. & Davies, A. H., 2016. Predicted burden of venous disease. Phlebology, 31(1_suppl), pp. 74-79. [0145] Smith, C. A. et al., 2006. XCMS: processing mass spectrometry data for metabolite profiling using nonlinear peak alignment, matching, and identification. Analytical chemistry, 78(3), pp. 779-787. [0146] Spagou, K. et al., 2010. HILIC-UPLC-MS for exploratory urinary metabolic profiling in toxicological studies. Analytical chemistry, 83(1), pp. 382-390. [0147] Spencer et al., Stat Med. 2016 Nov. 30; 35(27): 4909-4923. [0148] Tautenhahn, R., Boettcher, C. & Neumann, S., 2008. Highly sensitive feature detection for high resolution LC/MS. BMC bioinformatics, 9(1), p. 504. [0149] Trengove, N. J., Langton, S. R. & Stacey, M. C., 1996. Biochemical analysis of wound fluid from nonhealing and healing chronic leg ulcers. Wound Repair and Regeneration, 4(2), pp. 234-239.