Determination method for calprotectin and the use of calprotectin as a predictive marker for cardiovascular disease
10101325 ยท 2018-10-16
Assignee
Inventors
Cpc classification
G01N33/54393
PHYSICS
G01N33/54313
PHYSICS
G01N33/5306
PHYSICS
International classification
G01N33/543
PHYSICS
G01N33/53
PHYSICS
Abstract
An assay method for the detection of potential for CVD or propensity to CVD in a human or non-human animal subject, said method comprising assessing the concentration of calprotectin in a calprotectin-containing sample taken from said subject.
Claims
1. A direct turbidimetric assay method for the determination of calprotectin at concentrations of 0.5 to 50 mg/L in a calprotectin-containing body fluid, said method comprising the steps of: (a) obtaining a calprotectin-containing liquid sample of, or derived from, said body fluid, wherein said liquid sample comprises calprotectin at a concentration of 0.5 to 50 mg/L; (b) contacting said liquid sample of said body fluid with anti-calprotectin antibodies or antibody fragments, to bind said calprotectin, wherein said anti-calprotectin antibodies or antibody fragments are egg polyclonal antibodies or fragments thereof having a population of calprotectin binding sites and are immobilised by binding or coupling, either directly or indirectly, to nanoparticles to form antibody or antibody fragment coated nanoparticles; and (c) assessing the calprotectin content in said contacted liquid sample by turbidimetry, wherein the diameter of the antibody or antibody fragment coated nanoparticles is 65-140 nm, wherein increasing opacity correlates with increasing calprotectin concentration; and wherein said assay is capable of measuring calprotectin with precision in samples having calprotectin concentrations throughout the range of 0.5-50 mg/L.
2. The method of claim 1 wherein the diameter of the antibody or antibody fragment coated nanoparticles is 75-120 nm.
3. The method of claim 2, wherein the nanoparticles consist of polyethylene.
4. The method of claim 2, wherein the nanoparticles consist of gold.
5. The method of claim 2, wherein the nanoparticles consist of glass.
6. The method of claim 1 wherein said nanoparticles are substantially all of the same size.
7. The method of claim 1 wherein an opacity enhancer is added in between steps (b) and (c).
8. The method of claim 1 wherein said body fluid is selected from blood, serum, plasma, urine, cerebrospinal fluid, oral fluid, synovial fluid or empyema fluid.
9. The method of claim 1 performed as an automated assay.
10. The method of claim 1 wherein said nanoparticles are monodisperse.
11. The method of claim 1, wherein the body fluid is treated with calcium or ions or another alkaline earth metal prior to being used in the assay method.
12. The method of claim 11, wherein sufficient calcium or ions of another alkaline earth metal is added to the body fluid to saturate the calcium binding sites of calprotectin.
13. A direct turbidimetric assay method for the determination of calprotectin at concentrations of 0.5 to 50 mg/L in a calprotectin-containing body fluid, said method comprising the steps of: (a) obtaining a calprotectin-containing liquid sample of, or derived from, said body fluid, wherein said liquid sample comprises calprotectin at a concentration of 0.5 to 50 mg/L; (b) contacting said liquid sample of said body fluid with a composition consisting essentially of anti-calprotectin antibodies or antibody fragments immobilized by binding or coupling, either directly or indirectly, to nanoparticles to form antibody or antibody fragment coated microparticles; wherein said anti-calprotectin antibodies or antibody fragments are egg polyclonal antibodies or fragments thereof having a population of calprotectin binding sites; and (c) assessing the calprotectin content in said contacted liquid sample by turbidimetry, wherein the diameter of the antibody or antibody fragment coated nanoparticles is 65-140 nm, and wherein increasing opacity correlates with increasing calprotectin concentration; and wherein said is capable of measuring calprotectin with precision in samples having calprotectin concentrations throughout the range of 0.5-50 mg/L.
Description
BRIEF DESCRIPTION OF THE DRAWINGS
(1)
(2) TABLE-US-00001 Anderson-Darling Normality Test A.sup.2: 8.160 P-value: 0.000 Mean 0.403 Standard Deviation 0.238 Variance 0.057 Skewness 1.679 Kurtosis 3.302 N 199 Minimum 0.070 1st Quartile 0.240 Median 0.340 3rd Quartile 0.500 Maximum 1.370 95% confidence limit for Mu 0.370 0.436 95% confidence limit for Sigma 0.217 0.264 95% confidence limit for Median 0.310 0.370
(3)
(4) TABLE-US-00002 Anderson-Darling Normality Test A.sup.2: 25.037 P-value: 0.000 Mean 217.060 Standard Deviation 399.000 Variance 159201 Skewness 3.462 Kurtosis 15.341 N 200 Minimum 0.00 1st Quartile 0.00 Median 78.00 3rd Quartile 259.25 Maximum 2794.00 95% confidence limit for Mu 161.42 272.70 95% confidence limit for Sigma 363.35 442.46 95% confidence limit for Median 0.00 117.33
(5)
(6) TABLE-US-00003 Anderson-Darling Normality Test A.sup.2: 21.221 P-value: 0.000 Mean 2.331 Standard Deviation 2.970 Variance 8.819 Skewness 2.243 Kurtosis 5.103 N 197 Minimum 0.150 1st Quartile 0.540 Median 1.150 3rd Quartile 2.580 Maximum 16.30 95% confidence limit for Mu 1.913 2.748 95% confidence limit for Sigma 2.703 3.296 95% confidence limit for Median 0.933 1.375
(7)
(8) TABLE-US-00004 Anderson-Darling Normality Test A.sup.2: 2.535 P-value: 0.000 Mean 8.391 Standard Deviation 1.966 Variance 3.867 Skewness 0.875 Kurtosis 0.730 N 199 Minimum 5.100 1st Quartile 7.100 Median 8.100 3rd Quartile 9.600 Maximum 15.300 95% confidence limit for Mu 8.117 8.666 95% confidence limit for Sigma 1.790 2.181 95% confidence limit for Median 7.620 8.400
(9)
(10)
EXAMPLE 1
Anti-Calprotectin Antibody
(11) (a) Isolation of Calprotectin
(12) Calprotectin may be isolated according to the methods described in Examples 1 and 2 of U.S. Pat. No. 4,833,074 (Fagerhol).
(13) Calprotectin may alternatively be purified from human buffy coats. A cell-suspension in 2.5 mM EDTA is made by the addition of EDTA (50 mM, pH 7) to cells. The cells are then washed in 160 mM ammonium chloride/10 mM sodium hydrogen carbonate for 3 minutes and centrifuged (160g) for 10 minutes at 4 C. The resulting pellet is washed in EDTA (2.5 mM)/NaCl (150 mM) and centrifuged (55g) for a further 10 minutes at 4 C. The pellet is then resuspended in 0.625 mM EDTA/18.75 mM Diemal, pH 7.4 and frozen at 70 C. for at least 24 hours.
(14) Following thawing, the resulting material is centrifuged (at 3700g) for 30 minutes, then the supernatant is removed and filtered (with a 0.45 m filter available from Millipore), then loaded onto a DEAE (diethylaminoethyl) Sepharose ion-exchange column (available from Pharmacia), pre-prepared using a binding buffer (e.g. 0.63 mM EDTA/18.75 mM Diemal, pH 7.4). Any non-binding material passes through the column and is eluted. Once all of the non-binding material is eluted from the column, pure calprotectin is eluted using a calcium-containing elution buffer (e.g. 75 mM Diemal buffer/10 mM CaCl.sub.2). About 25 mg calprotectin is obtained per buffy coat.
(15) (b) Preparation of Anti-Calprotectin Antibody
(16) Anti-calprotectin antibodies may be prepared according to the method described in Example 3 of U.S. Pat. No. 4,833,074 (Fagerhol)
(17) Chicken egg polyclonals may alternatively be prepared. A solution comprising calprotectin (0.5 mg/ml) and Freund's adjuvant is injected into chickens every 14 days four times (or for two months), and then once every 1 month. After 12 weeks, the eggs of the calprotectin-injected chicken may be collected and their yolks removed (without the film). Following dilution in HCl (5 mM), the yolk is centrifuged and the supernatant is collected. The supernatant is then filtered and treated with saturated ammonium sulfate to a final concentration of 3.8 M. The mixture is centrifuged and the precipitate produced is collected and dissolved in buffer (0.11 M sodium acetate, 0.15 M NaCl, pH 7.4). The resulting solution is finally dialysed with a membrane having a pore size of 10,000 kD and then purified by affinity chromatography.
(18) The column typically used for affinity chromatography comprises an activated matrix of succinimide-activated sepharose (HiTrap NHS activated available from Amersham-Pharmacia) which is suitable for the immobilisation of calprotectin. More specifically, the activated resin reacts spontaneously, at pH 7-8, with free amines in the calprotectin. For chromatography the dialysis solution is usually diluted to a concentration of about 3 mg/ml in PBS prior to its application to the column. The anti-calprotectin antibodies are subsequently eluted using 6 M urea in ice cold PBS or 0.1 M sodium citrate solution, pH 3.0. Preferably, 0.1 M sodium citrate solution is used. Following elution, the anti-calprotectin antibody containing fractions are immediately diluted and dialysed in PBS.
EXAMPLE 2
Turbidimetric Assay for Calprotectin
(19) (a) Preparation of Avidin-coated Nanoparticles
(20) 600 m of 4.2% w/v chloromethyl activated nanoparticles (diameter 44 nm) available from Interfacial Dynamic Corporation, US are dialysed against water with a membrane having a pore size of 10,000 kD. 0.5 ml of a borate (10 mM) and sodium chloride (15 mM) solution at pH 9.0 is added and mixed. 10 mg avidin, dissolved in 0.5 ml of a 10 mM borate and 15 mM NaCl solution at pH 9 (available from Pierce Chemical Company) is added and the mixture is agitated at room temperature for 24 hours. 40 l of glycine solution (2M, pH 9.0) is then added and the mixture is agitated for a further 4 hours at room temperature.
(21) The particles are then diluted to a volume of 100 ml and diafiltrated, firstly in 500 ml of a 10 mM borate and 15 mM sodium chloride solution at pH 9.0 and secondly in a 25 mM Tris, 150 mM sodium chloride and 0.01% Tween 20 solution at pH 7.4 (available from Sigma US) using a Pellicon XL Filter (cut off 300,000) and a labscale TTF System (available from Millipore) in accordance with the instructions supplied from the instruments suppliers. The desired concentration of avidin-coated nanoparticles is finally obtained by centrifugation and re-suspension of the particles in a 25 mM TRIS, 150 mM sodium chloride and 0.01% Tween 20 solution. Any aggregates formed during this preparation procedure may be removed by slow centrifugation.
(22) (b) Assay for Calprotectin using Avidin-coated Nanoparticles
(23) A suspension having a concentration of about 0.30 mg particles of the above-described avidin-coated nanoparticles per ml is prepared by centrifugation and re-suspension of the above-described preparation in a 25 mM TRIS, 150 mM NaCl, 0.1% Tween 20 and 2% PEG 6000 solution at pH 7.4 (available from Sigma). 500 l of this particle suspension is mixed with a plasma sample (about 20 l), taken from a subject being tested for propensity to CVD, in a reading quartz cuvette of a recording spectrophotometer (e.g. a Shimadzu UV-160). The absorption of 340 nm monochromatic light is recorded and after 60 s, 75 g of anti-calprotectin antibody labelled with 0.15 nmol biotin (e.g. biotin labelled affinity purified egg polyclonal purchased from Norwegian Antibodies AS, Norway), diluted in 50 l of a 25 mM TRIS, 150 mM NaCl and 0.1% Tween 20 solution at pH 7.4 is added to the quartz cuvette and mixed. The absorption of 340 nm monochromatic light is immediately recorded using a reference cuvette containing a solution of 25 mM TRIS, 150 mM NaCl and 0. 1% Tween 20 at pH 7.4, and again at regular intervals (e.g. every 2 minutes) until about 15 minutes has elapsed. The increase in absorption at each time point is calculated in accordance with standard turbidimetric reading in kinetic mode or end-point readings. That is, the increase in light absorption at each time-point is calculated relative to the reading made prior to the addition of antibody-coated nanoparticles and/or at the end of the recording.
(24) A calibration curve is constructed by carrying out an identical procedure with standards having a known concentration of calprotectin. The concentration of calprotectin in the sample can then be calculated from the calibration curve.
EXAMPLE 3
Alternative Turbidimetric Assay for Calprotectin
(25) (a) Preparation of Anti-Calprotectin Antibody Coated Nanoparticles
(26) 1 ml of 4.2% w/v chloromethyl activated nanoparticles (diameter 44 nm) available from Interfacial Dynamic Corporation, US are dialysed against water with a membrane having a pore size of 10,000 kD. 0.5 ml of a 10 mM borate and 15 mM sodium chloride solution at pH 9.0 is then added. 27 mg of purified anti-calprotectin antibodies (e.g. affinity purified egg polyclonal antibodies available from Norwegian Antibodies AS, Norway) are dialysed against a 10 mM borate and 15 mM sodium chloride solution at pH 9.0.
(27) Following addition of the nanoparticles to the purified anti-calprotectin antibodies the mixture is agitated for 24 hours at room temperature. 40 l of a glycine solution (2 M at pH 9.0) is then added and the mixture is agitated for a further 4 hours at room temperature.
(28) The particles are then diluted to total volume of 100 ml and diafiltrated against 1000 ml of a 10 mM borate and 15 mM sodium chloride solution at pH 9.0 to which 0.1% Tween 20 and 3 mg/ml egg albumin is added using a Pellicon XL filter (cut of 300,000) and a labscale TFF system (available from Millipore) in accordance with the instructions supplied from the instruments suppliers. The desired concentration of anti-calprotectin antibody-coated nanoparticles is finally obtained by centrifugation and re-suspension of the particles in solution. Any aggregates formed during this preparation procedure may be removed by slow centrifugation.
(29) (b) Assay for Calprotectin using Anti-calprotectin Antibody-coated Nanoparticles
(30) A suspension comprising 400 g of the above-described antibody-coated nanoparticles in 50 l of a 10 mM borate, 15 mM NaCl, 0.1% Tween 20, 3 g/l egg albumin solution at pH 9.0 is prepared.
(31) Simultaneously, 20 l of plasma, taken from the subject being tested for potential for CVD, in 500 l assay buffer (25 mM TRIS, 150 mM NaCl, 0.1% Tween 20 and 2% PEG 6000 at pH 7.4 (available from Sigma) is put in a reading quartz cuvette of a recording spectrophotometer (e.g. Shimadzu UV-160) and the light absorption of 340 nm monochromatic light is measured. After 60 s, the above-mentioned suspension comprising 400 g of antibody-coated nanoparticles is added, and mixed in the cuvette. The light absorption immediately after adding the antibody-coated nanoparticles is recorded, and again at regular intervals (e.g. every 2 minutes) until about 15 minutes has elapsed. The increase in light absorption at each time-point is calculated relative to the reading made prior to the addition of antibody-coated nanoparticles and/or at the end of the recording. In other words, turbidimetric readings in kinetic mode or end-point readings are made.
(32) A calibration curve is also constructed by carrying out an identical procedure with standards having a known concentration of calprotectin. The concentration of calprotectin in the sample can then be calculated from the curve.
EXAMPLE 4
Turbidimetric Assay for Calprotectin
(33) (a) Preparation of Streptavidin-coated Nanoparticles
(34) 600 m of 4.2% w/v chloromethyl activated nanoparticles (diameter 67 nm) available from Interfacial Dynamic Corporation, US are dialysed against water with a membrane having a pore size of 10,000 kD. 0.5 ml of a phosphate (10 mM) and sodium chloride (150 mM) buffer solution at pH 7.4 is added together with 10 mg streptavidin, dissolved in 0.5 ml of a 10 mM phosphate and 150 mM NaCl buffer solution at pH 7.4 (available from Pierce Chemical Company) and the mixture is agitated at room temperature for 24 hours. 40 l of glycine solution (2M, pH 9.0) is then added and the mixture is agitated for a further 4 hours at room temperature.
(35) The particles are then diluted to a volume of 100 ml and diafiltrated, firstly in 500 ml of a 10 mM borate and 15 mM sodium chloride solution at pH 9.0 and secondly in a 25 mM Tris, 150 mM sodium chloride and 0.01% Tween 20 solution at pH 7.4 (available from Sigma US) using a Pellicon XL Filter (cut off 300,000) and a labscale TTF System (available from Millipore) in accordance with the instructions supplied from the instruments suppliers. The desired concentration of avidin-coated nanoparticles is finally obtained by centrifugation and re-suspension of the particles in a 25 mM TRIS, 150 mM sodium chloride and 0.0% Tween 20 solution. Any aggregates formed during this preparation procedure may be removed by slow centrifugation.
(36) The mean particle size of the streptavidin coated nanoparticles was measured to be 82 nm by Sinteff AS, Norway.
(37) (b) Assay for Calprotectin using Streptavidin-coated Nanoparticles
(38) A suspension having a concentration of about 0.60 mg particles of the above-described avidin-coated nanoparticles per ml is prepared by centrifugation and re-suspension of the above-described preparation in a 25 mM TRIS, 150 mM NaCl, 0.1% Tween 20 and 1% PEG 6000 solution at pH 7.4 (available from Sigma). 500 l of this particle suspension is mixed with a plasma sample (about 5 l), taken from a subject being tested for propensity to CVD, in a reading quartz cuvette of a recording spectrophotometer (e.g. a Shimadzu UV-160). The absorption of 560 nm monochromatic light is recorded and after 60 s, 75 g of anti-calprotectin antibody labelled with 0.15 nmol biotin (e.g. biotin-labelled affinity purified egg polyclonal purchased from Norwegian Antibodies AS, Norway), diluted in 50 l of a 25 mM TRIS, 150 mM NaCl and 0.1% Tween 20 solution at pH 7.4 is added to the quartz cuvette and mixed. The absorption of 340 nm monochromatic light is immediately recorded using a reference cuvette containing a solution of 25 mM TRIS, 150 mM NaCl and 0.1% Tween 20 at pH 7.4, and again at regular intervals (e.g. every 2 minutes) until about 15 minutes has elapsed. The increase in absorption at each time point is calculated in accordance with standard turbidimetric reading in kinetic mode or end-point readings. That is, the increase in light absorption at each time-point is calculated relative to the reading made prior to the addition of antibody-coated nanoparticles and/or at the end of the recording.
(39) A calibration curve is constructed by carrying out an identical procedure with standards having a known concentration of calprotectin. The concentration of calprotectin in the sample can then be calculated from the calibration curve.
EXAMPLE 5
(40) (a) Preparation of Anti-Calprotectin Antibody Coated Nanoparticles
(41) 1 ml of 4.2% w/v chloromethyl activated nanoparticles (mean diameter 67 nm) available from Interfacial Dynamic Corporation, US, are dialysed against water with a membrane having a pore size of 10.000 kD, and then diluted to 10 ml with water. 27 mg of purified egg polyclonal antibodies ( e.g. affinity purified egg polyclonal antibodies available from Norwegian Antibodies AS, Norway) are dialysed against a 10 mM borate and 15 mM sodium chloride buffer solution and finally diluted to 6 ml in the same 10 mM borate and 15 mM sodium chloride solution at pH=9.0.
(42) Under agitation, the particles are mixed with the antibodies, and agitation is continued at room temperature for 24 hours. 40 l of a glycine solution (2 M at pH 9.0) is then added and the mixture is agitated for a further 4 hours at room temperature.
(43) The particles are then diluted to a total volume of 100 ml in 10 mM borate, 15 mM sodium chloride buffer to which 0.1% Tween 20 and 3 mg/ml egg albumin is added and diafiltrated against 1000 ml of said borate/sodium chloride buffer to which 0.1% Tween 20 and 3 mg/ml egg albumin is added, using a Pellicon XL filter (cut off 300.000 D) and a labscale TFF system (available from Millipore) in accordance with the instructions supplied from the instrument suppliers, and in the end the particles are concentrated to a volume of 40 to 100 ml.
(44) The mean diameter of the particles obtained was measured to be 81 nm by Sinteff AS, Norway.
(45) (b) Assay for Calprotectin using Anti-calprotectin Antibody-coated Nanoparticles
(46) A suspension of 0.7 mg/ml of the above described anti-calprotectin antibody coated particles is made in 0.25 mM TRIS, 0.15 M NaCl and 0.1% Tween at pH 8.0.
(47) 5 l of a plasma sample, taken from the subject being tested for CVD, is dissolved in assay buffer (460 l, 25 mM TRIS, 150 mM NaCl, 0.1% Tween, 1.0% polyethyleneglycol 6000, pH=7.4) in a reading quartz cuvette in a recording spectrophotometer (e.g. Shimadzu UV-160) and the light absorption of 560 nm monochromatic light is measured. After 60 s, 100 l of the above mentioned suspension comprising 0.7 mg/ml of antibody coated nanoparticles is added, and mixed in the cuvette. The light absorption before and immediately after adding the antibody-coated nanoparticles is recorded, and again at regular intervals (eg. every 20 s) until 15 minutes has elapsed. The increase in light absorption at each time point is calculated relative to the reading made prior to the addition of antibody coated nanoparticles and at the end of the recording. In other words, turbidimetric readings in kinetic mode and/or end point readings are made.
(48) A calibration curve is also constructed by carrying out an identical procedure with standards having a known concentrations of calprotectin. The concentration of calprotectin ib the sample can then be calculated from the curve.
EXAMPLE 6
Statistical Analysis
(49) Comparison of Calprotectin and Other Markers for Detection of Potential for CVD or Propensity to CVD
(50) Coronary calcification has been shown to be strongly associated with the occurrence of CVD and has also been demonstrated to be a useful method for predicting potential for CVD (e.g. mycocardial infarction or stroke).
(51) The extent of coronary calcification is quantitatively measured used electron-beam computed tomography (EBCT) and is represented by a calcium score (CS). A high calcium score represents a high level of calcification and a high risk of developing CVD.
(52) In the following study the CS of 200 subjects (100 controls having a CS<100 and 100 cases having a CS>100) aged 45 or greater was-tested as well as their calprotectin, CRP and homocysteine plasma or serum levels. The subjects were either self- or physician-referred asymptomatic individuals and had had an EBCT scan within the previous 2 years (usually within the previous 6 months) to testing of their plasma or serum for the concentration of calprotectin, CRP and homocysteine.
(53) Methods
(54) Calprotectin
(55) Calprotectin was measured by the Calprest test (distributed by Eurospital, Italy) and data were summarised for the patient and the control group using standard deviation, median, minimum and maximum. 95% confidence intervals for median were calculated. The Anderson-Darling calculation was used as a test for normality.
(56) Calprotectin was also summarised by gender using the same summary statistics.
(57) Calcium Score (CS)
(58) Calcium score was determined by Electron-Beam Computed Tomography (EBCT) scanning. Data for case and control groups were summarised.
(59) High Sensitive C-Reactive Protein (hsCRP)
(60) hsCRP was determined by Dade Behring's N High Sensitivity CRP assay (Roberts et al., Clinical Chemistry, 2000, 46:4, p 461-468). Data were summarised for the patient and the control groups using mean, standard deviation, median, minimum and maximum.
(61) Homocysteine (Hyc)
(62) Plasma Homocysteine (Hcy) was determined by the Abbott IMx method (Shipchandler, M. T. and Moore E. G., Clinical Chemistry, 1995, 41:7, p. 991-994). Data were summarised for the patient and the control groups using standard deviation, median, minimum and maximum.
(63) Comparison between CS and Calprotectin, hsCRP, Hcy
(64) (i) Chi-squared
(65) The Chi-squared test was used to test for covariance between pairs of markers.
(66) (ii) Odds Ratio
(67) The odds ratio used with the 22 cross-tabulation (i.e the chi-squared table) is the ratio of the odds of two tests co-varying to the odds of two tests disagreeing. Therefore, the odds ratio may be interpreted as a measure of the magnitude of association between the two tests.
(68) Odds ratio was calculated for CS versus calprotectin, hsCRP and Hcy respectively.
(69) Comparison of Median Values of the Markers
(70) (i) Mann-Whitney Test
(71) This is a non-parametric (the data does not need to be normally distributed) which was used to test if the median values of two markers are significantly different. Minitab ranks all the data from both sets of data in order, assigning 1 to the lowest up to 200 for the highest. The software then adds up the rank-score for the two groups to be compared and reports the P value which indicates the chance that random sampling would result in medians as far apart as that observed in the experiment.
(72) Calprotectin, hsCRP and Hcy median values were compared to CS split into case and control groups and the Mann-Whitney test was used to test for significance.
(73) (ii) ROC-Curves for Calprotectin and Calcium Score
(74) The ability of a test to discriminate diseased cases from normal cases can be evaluated using Receiver Operating Characteristic (ROC) curve analysis. ROC curve analysis was used for both calprotectin and hsCRP and as a comparison of the two markers.
(75) The ROC curves were produced using 11 cut-off limits for calculation of the sensitivity (y-axis) and the 1-specificity (x-axis). The data from the control and the case groups were pooled together and ranked from lowest to highest value. Analysis was performed on the total population and on the male population. The areas under the ROC curves were calculated using the trapezoidal rule.
(76) Results
(77) Summary Statistics for Calprotectin
(78)
(79) Table 1 below shows the summary statistics for calprotectin in the control and the case groups. The median calprotectin concentration in the control group is 0.31 mg/L compared to 0.38 mg/L in the case group. The median calprotectin concentration is 0.31 mg/L for the males and 0.30 mg/L for the females in the control group. In the case group the median calprotectin concentration for males is 0.39 mg/L compared to 0.31 mg/L for females.
(80) TABLE-US-00005 TABLE 1 Summary statistics for Calprotectin Calprotectin Controls Cases All Females Males All Females Males N 100 59 41 100 15 85 SD 0.228 0.232 0.225 0.503 0.202 0.537 95% CI for median 0.28-0.34 0.28-0.33 0.27-0.39 0.34-0.43 0.19-0.47 0.34-0.45 Median (mg/L) 0.31 0.30 0.31 0.38 0.31 0.39 Min (mg/L) 0.09 0.09 0.11 0.07 0.08 0.07 Max (mg/L) 1.37 1.37 1.36 4.85 0.73 4.85
Summary Statistics for Calcium Score
(81)
(82) Table 2 below shows the summary statistics for EBCT calcium score in the control and case groups. The median calcium score in the control group is 0 compared to 259 in the case group. The median calcium score is 0 for both females and males in the control group and 315 and 256, respectively, in the case group.
(83) TABLE-US-00006 TABLE 2 Summary statistics for Calcium score Calcium score Controls Cases All Females Males All Females Males N 100 59 41 100 15 85 SD 5.6 0.7 8.7 474.7 365.1 493.3 95% CI for median 0-0 0-0 0-0 313.5-215.0 174.8-538.0 300.4-207.5 Median 0 0 0 259 315 256 Min 0 0 0 100 100 100 Max 56 5 56 2794 1507 2794
Summary Statistics for hsCRP
(84)
(85) Table 3 below shows the summary statistics for hsCRP in the control and case groups. The median hsCRP in both the control group and the case group is 1.2 mg/L. The median hsCRP is 1.6 mg/L for females and 0.7 mg/L for males in the control group and 1.3 mg/L and 1.1 mg/L, respectively, in the case group.
(86) TABLE-US-00007 TABLE 3 Summary statistics for hsCRP hsCRP Controls Cases All Females Males All Females Males N 100 59 41 100 15 85 SD 6.63 8.36 1.53 12.70 3.16 13.71 95% CI for median 0.79-1.47 1.36-2.74 0.56-0.97 0.94-1.56 0.65-2.53 0.89-1.59 Median (mg/L) 1.17 1.59 0.68 1.23 1.28 1.11 Min (mg/L) 0.15 0.15 0.15 0.17 0.34 0.17 Max (mg/L) 62.1 62.1 9.29 120.0 11.9 120.0
Summary Statistics for Hcy
(87)
(88) Table 4 below shows the summary statistics for Hcy in the control and case groups. The median Hcy concentration in the control group is 7.5 mol/L compared to 8.5 mol/L in the case group. The median Hcy concentration is 7.15 mol/L for females and 8.55 mol/L for males in the control group and 7.35 mol/L and 8.55 mol/L respectively in the case group.
(89) TABLE-US-00008 TABLE 4 Summary statistics for Hcy Hcy Controls Cases All Females Males All Females Males N 100 59 41 100 15 85 SD 1.94 1.58 2.18 6.03 2.18 6.46 95% CI for median 7.7-8.5 7.1-7.9 8.1-9.5 8.1-10.5 6.8-9.2 8.1-10.9 Median (mol/L) 7.5 7.1 8.5 8.5 7.3 8.5 Min (mol/L) 5.3 5.3 5.3 5.1 5.1 5.2 Max (mol/L) 14.7 11.8 14.7 65.8 12.4 65.8
Comparison between Calcium Score and Calprotectin, hsCRP, Hcy
(i) Chi-squared Test
(90) Using Minitab, chi-squared tables were performed (see table 5 below). This test compares expected distributions between sets of data (assuming a random distribution) and those observed. The significance value is a measure of the degree to which the data is not randomly distributed. For example, considering the data for calprotectin positive it would be expected that there would be an even spilt between calcium positive and negative (i.e. 30.5 expected in both columns). The observed split, however, is 22 negative and 39 positive indicating a tendency for the calprotectin positives to co-vary with the calcium positives.
(91) Minitab provides an overall measure of the significance of these differences (both agreement and disagreement) and in this case P is 0.009. Thus there is a significant co-variation of calcium with calprotectin.
(92) TABLE-US-00009 TABLE 5 Chi-Square comparison of Calprotectin versus Calcium score Calcium (co 100) Calprotectin Negative Positive All (co 0.45 mg/L) (N) (N) (N) Negative Obs 78 61 139 Exp 69.5 69.5 139 Positive Obs 22 39 61 Exp 30.5 30.5 61 All Obs 100 100 200 Exp 100 100 200
(93) An analogous comparison was made between calprotectin and hsCRP (see Table 6 below).
(94) TABLE-US-00010 TABLE 6 Chi-Square comparison of Calprotectin versus hsCRP (P = 0) hsCRP (co 1.69 mg/L) Calprotectin Negative Positive All (co 0.45 mg/L) (N) (N) (N) Negative Obs 100 39 139 Exp 88.96 50.04 139 Positive Obs 28 33 61 Exp 39.04 21.96 61 All Obs 128 72 200 Exp 128 72 200
(ii) Odds Ratio
(95) Odds ratios can be derived from tables 5 and 6 and give a measure of how far the values observed deviate from the expected. If the expected agreement figures are multiplied together and divided by the product of the disagreement expected figures a value of 1 should be obtained.
(96) Taking the data in Table 6, for instance, the odds ratio on the expected values is 1: (88.9621.96)/(50.0439.04)=1953.64/1953.6=1.
(97) The odds ratio observed in this table is: (10033)/(3928)=3300/1092=3.02.
(98) The further the odds ratio from 1 the more pronounced the co-variation; an odds ratio of 3 is significant.
(99) The results of an odds ratio analysis carried out on the co-variance data obtained from the study are shown in Table 7.
(100) TABLE-US-00011 TABLE 7 Summary of Odds-Ratio findings Risk markers Odds ratio (N1#pos, N2#neg) Calcium Calprotectin CRP Calcium Score (co 100) N1 = 100, N2 = 100 Calprotectin (co 0.45 mg/L) 2.27 N1 = 61, N2 = 139 CRP (co 1.69 mg/L) 1.00 3.02 N1 = 72, N2 = 128 Hcy (co 12 mol/L) 1.21 1.32 1.52 N1 = 11, N2 = 189 (co 10 mol/L) 1.63 1.55 1.12 N1 = 42, N2 = 158 co = cut off
(101) A high odds ratio indicates a high degree of co-variation between the tests. It can be seen from Table 7 that the test for calprotectin gives the highest odds ratio to calcium score and therefore it can be deduced that calprotectin has the highest degree of covariance with calcium score out of calprotectin, CRP and homocysteine.
(102) Additionally, calprotectin gives a high odds ratio with CRP, another marker for CVD.
(103) Chi-Squared Analysis using Minitab
(104) A chi-squared test on the above data, using the same cut-offs as the odds-ratio test, showed significant (P 0.009) agreement between the calcium score and calprotectin results. In contrast, neither the test for CRP nor the test for homocysteine showed any significant co-variance with calcium score.
(105) Significant agreement (P 0.000) was also found between calprotectin and CRP.
(106) Comparing Median Values
(107) (i) Mann-Whitney Test
(108)
(109) The median values for the two groups are 0.310 mg/L for the negative calcium and 0.375 mg/L for the positive calcium. Mann-Whitney provides that the sum of the ranks for the negative calcium is 912 and 1108 for calcium negative and yields a P value of 0.0113.
(110) A Mann-Whitney analysis to test for significant increases in the medians of each of calprotectin, CRP, and homocysteine concentration in the high (CS>100) and low (CS<100) calcium score groups was also carried out on the above data. The results showed that the median values for calprotectin and homocysteine concentration were significantly raised in the high calcium group (P=0.0112 and 0.0037 respectively) whereas CRP did not show any significant difference in either group (see Table 8 below).
(111) TABLE-US-00012 TABLE 8 Comparison of median values of Calprotectin and Hcy Median value in Median value in low CS group high CS group P value Calprotectin 0.3 mg/L 0.4 mg/L 0.0112 Homocysteine 7.5 mol/L 8.5 mol/L 0.0037
(ii) ROC Analyses and Curves
(112)
(113) TABLE-US-00013 TABLE 9 Area under the ROC curves Group Calprotectin hsCRP All subjects 0.604 0.524 Male subjects 0.622 0.502
Agreement Rate Against Calcium Score
(114) The accuracy of each of the calprotectin, CRP and homocysteine (Hcy) tests at the cut-off levels (Hcy co 12 mol) in Table 7 as a test for potential to CVD was also assessed assuming that a calcium score >100 is reflective of high risk to CVD.
(115) The results of the analysis are shown in Table 10.
(116) TABLE-US-00014 TABLE 10 Calprotectin & Calprotectin Calprotectin Calprotectin & CRP correct & CRP incorrect CRP incorrect incorrect & CRP correct correct 58 42 25 75 29% 21% 12.5% 37.5% Calprotectin & Calprotectin Calprotectin Calprotectin & Hcy correct & Hcy incorrect Hcy incorrect incorrect & Hcy correct correct 59 40 24 77 29.5% 20% 12% 38.5% Agreement rate against calcium for calprotectin = 58.5% (42 + 75/200) Agreement rate against calcium for CRP = 50% (25 + 75/200) Agreement rate against calcium for homocysteine = 50.5% (24 + 77/200)