Abstract
An analysis of high-sensitive factor I and C-reactive proteins as biomarkers for coronary artery disease has been performed from 19 anticipated cohort studies that included 21,567 participants having no information about coronary artery disease. Besides, the clinical implications of statin therapy initiated due to assessment of factor I and C-reactive proteins have also been modeled during studies. The measure of risk discrimination (C-index) was increased (by 0.0101) as per the prognostic model for coronary artery disease with respect to sex, smoking status, age, blood pressure, total cholesterol level along with diabetic history characteristic parameters. The C-index was further raised by 0.0045 and 0.0053 when factor I and C-reactive proteins based information were added, respectively which finally predicted 10-year risk categories as: high (> 20%), medium (10% to < 20%), and low (< 10%) risks. We found 2,254 persons (among 15,000 adults (age ≥ 45 years)) would initially be classified as being at medium risk for coronary artery disease when only conventional risk factors were used as calculated risk. Besides, persons with a predicted risk of more than 20% as well as for persons suffering from other risk factors (i.e. diabetes), statin therapy was initiated (irrespective of their decade old predicted risk). We conclude that under current treatment guidelines assessment of factor I and C-reactive proteins levels (as biomarker) in people at medium risk for coronary artery disease could prevent one additional coronary artery disease risk over a period a decade for every 390-500 people screened.
Keywords: Factor I protein, C-reactive protein, biomarkers, coronary artery disease
Introduction
The C-reactive protein is described as the first acute-phase protein and known it is capable to precipitate the somatic C-polysaccharide of Streptococcus pneumonia. Besides, it is presently known as an exquisitely sensitive systemic inflammation and tissue damage based biomarker and to date, more than 30 epidemiological studies have shown its strong association with cardiovascular heart disease [1,2]. This molecule has specific characteristics as inflammatory stimuli as it binds to phosphocholine on microbes and extrinsic ligands bearing phosphocholine residues like native and modified plasma lipoproteins, phospholipids and related compounds, apoptotic cells and other damaged cell membranes [3].
We are living in the era of multivariable statistical derived models as “global risk assessment” scores to check who is cardiovascular risk at but these models have some limitations which have prompted us to the search of novel biomarkers for coronary artery disease. To achieve this goal C-reactive protein and factor I biomarkers have been explored. C-reactive protein has been chosen due to its high-sensitive assays which had received widespread interest and an accumulated large database as its potential role as a predictor of cardiovascular risk by several associations, Canadian Cardiovascular Society [4], Centers for Disease Control and Prevention and American Heart Association [5], European Society of Cardiology [6], National Academy of Clinical Biochemistry Laboratory Medicine Practice Guidelines [7], as well as American College of Cardiology Foundation-American Heart Association Task Force [8], recently. While usefulness of factor I as biomarker is still inconclusive and hence inspired us to explore it as biomarker of inflammation for coronary artery disease.
On account of usage of factor I and C-reactive protein and need for biomarker of inflammation for coronary artery disease we analyzed records of 21,567 people without a background of coronary artery disease at baseline from 19 anticipated cohort studies. Thus, this approach was used to quantify the improvement in the prediction of a first coronary artery disease risk event by adding the assessment of circulating biomarkers of inflammation into the risk factors’ assessment to explore the usage of factor I and C-reactive proteins as biomarkers for coronary artery disease.
Materials and methods
Study cohorts
For anticipated cohort studies, participants were chosen (during follow-up), specifically who did not have a background of baseline coronary artery disease were chosen (during follow-up). Participants were followed up for more than a year to record the data. When we obtained baseline level data of factor I, C-reactive protein or both only then we included them for studies. Additionally, complete information on sex, age, smoking status, systolic blood pressure, diabetic history, as well as the total and high-density lipoprotein cholesterol’s levels had also been collected. Besides these, some other data related to albumin-levels, leukocyte-counts, socioeconomic factors, family history of coronary artery disease and low-density lipoprotein cholesterol levels were also available, but only as a subset of study participants. Definition of coronary artery disease was used in the entire studies on the basis of World Health Organization criteria. The study was designed by us and then approved by the ethics review committee of China-Japan Union Hospital of Jilin University, Changchun, Jilin, China.
Statistical analysis
The main outcome of the studies was a first coronary artery disease event. The loge-transformed values of leukocyte count and C-reactive protein were used during entire analyses. The risk prediction models were based on a Cox proportional hazards model as per reported literature [9] according to study, sex, and (where applicable) randomized trial arm, but log hazard ratios (common coefficients) were estimated across studies. Briefly, for each stratum k = 1, 2, •••, K (for K distinct combinations of study, sex, and trial arm), with i = 1, 2, •••, nk individuals in stratum K and the ith individual having baseline covariate values Xi = (xi1, xi2, ••• xip), the probability of surviving beyond t years after baseline without a coronary artery disease event was modeled as
S(t | Xi, k) = S0k(t)exp(βXi’)
However, the probability of a coronary artery disease event within t years after baseline was simply one minus the survival probability, such as
Pr(T ≤ t | Xi, k) = 1 - S(t | Xi, k) = 1 - S0k(t)exp(βXi’)
We censored the outcomes for patients who died from non-coronary artery disease event causes, so the above probability refers to the risk of a coronary artery disease event in the absence of deaths from other causes. In both the equations the evolution of risk over time was modeled independently for each stratum (i.e. study, sex, and trial arm), which is represented by the non-parametric baseline survival S0k(t).
Prognostic models were compared with the assessment of discrimination using the C-index [10] and the D measure [11] as per reported work to quantify the extent of the model up to which it can predict the disease events [12]. On the other hand for an arbitrary selection of participants, the C-index is the probability that, the individual with the shorter survival time has the higher value of the linear predictor βXi’. The overall measure is calculated as
C = (Co + 0.5u)/(Co + d + u)
where Co, u, and d are the number of concordant, undecided and discordant pairs respectively. In present case we allowed a constrained pair-wise comparisons i.e. only pairing of participants within the same strata. In other words we can say that the concordance/discordance counts did not include comparison of males to females nor participants from different studies.
However, the D measure was by first transforming each participant’s linear predictor βXi’ from the fitted Cox proportional hazards model to give standard normal order rank statistics (using Blom’s approximation). During present analyses the standard normal rank order statistics were formed within studies to avoid potential influences from between study differences in covariate distributions. Division of the rank statistics with a factor of √π/8 gave zi and then a second stratified Cox proportional hazards model is then fitted to these values. Thus, D is the regression coefficient of z from this second model. Thus we used two-stage method to examine the between-study heterogeneity through C-index and D measure calculations. As per number of coronary artery disease outcomes’ contribution our studies were weighted. The I2 statistic approach [13] was used to quantify the heterogeneity lied in the risk-discrimination and in this way it has measured all the changes. However, the well known chi-square tests were used to determine the discrimination measurements (across subgroups).
Prognostic models were also compared with the use of risk reclassification assessment [14]. Reclassification tables were constructed as per data from studies containing both the fatal and the nonfatal coronary artery disease outcomes which had examined the participants among three predicted decades coronary artery disease categories (“high” > 20%; “medium” 10% to < 20%; and “low” < 10%) when factor I and C-reactive proteins were included to the model of conventional risk factors. A reduction of coronary artery disease by 20% with statin treatment was assumed for clinical modeling. All P values were two-sided. The latest version of Stata software was utilized for the entire analyses.
Results
Characteristics of the study participants
We started the analysis with 19 studies involving 21,567 participants. However, the mean (± SD) age at baseline was 62 ± 8 years. Overall, there were 61% men out of the all the participants. Information on factor I levels was explored for total a 15,781 number of participants from 14 cohorts studies (Table 1), among whom there were first 737 coronary artery disease events. Information on C-reactive protein levels was available for 14,565 participants from 12 anticipated cohorts (Table 1), among whom there were first 810 coronary artery disease events. Information on both factor I and C-reactive proteins levels was available for 8,410 participants. Table 1 is shown here to express the baseline characteristics of participants in the factor I and C-reactive proteins analyses, along with the most important factor viz. the adjusted hazard ratios of first coronary artery disease events. Detailed data related to anticipated cohort studies are not shown to avoid mere over-compilation of data.
Table 1.
Baseline characteristics and hazard ratios for first coronary artery disease
| Characteristics | Assessment of factor I protein | Assessment of C-Reactive protein | ||
|---|---|---|---|---|
|
| ||||
| Participants/total* (N = 737/15,781) | Hazard ratio (95% CI)! | Participants/total* (N = 810/14,565) | Hazard ratio (95% CI)! | |
| Sex (male; no.) | 8363 | Not Available# | 6991 | Not Available# |
| Age (mean ± SD; yr) | 60.9 ± 8.3 | 1.80 (1.75-1.84) | 61.3 ± 8.9 | 1.89 (1.87-1.93) |
| Smoking Status (Current; no.) | 3787 | 1.72 (1.65-1.79) | 3204 | 1.65 (1.61-1.69) |
| Systolic B.P. (mean ± SD; mm Hg) | 138 ± 17 | 1.29 (1.26-1.31) | 137 ± 18 | 1.27 (1.25-1.29) |
| Diabetic History (no.) | 947 | 1.90 (1.81-2.02) | 873 | 1.73 (1.61-1.84) |
| Total Cholesterol (mean ± SD; mmol/L) | 5.78 ± 1.11 | 1.21 (1.18-1.23) | 5.82 ± 1.09 | 1.19 (1.16-1.22) |
| HDL Cholesterol (mean ± SD; mmol/L) | 1.34 ± 0.34 | 0.84 (0.81-0.86) | 1.34 ± 0.39 | 0.87 (0.85-0.90) |
| Factor I (g/L) | 3.11 ± 0.73 | 1.16 (1.14-1.18) | - | - |
| Loge(C-Reactive Protein) (mg/L) | - | - | 0.60 ± 1.07 | 1.22 (1.17-1.23) |
| Anticipated Cohort Studies$ | BRUN; BWHHS; COPEN; EAS; FINRISK97; HISAYAMA; IKNS; LEADER; MESA; NHANESIII; NSHS; PROCAM; SHHEC; SHS | AFTCAPS; ARIC; BRUN; BWHHS; EAS; FINE-FIN; FINE-IT; FINRISK97; KIHD; NHANESIII; TARFS; WOSCOPS | ||
Data on factor I are from 14 studies in which 737 participants had a first-ever coronary artery disease outcome during follow-up and data on C-reactive protein are from 12 studies in which 810 participants had a first-ever coronary artery disease outcome during follow-up.
The hazard ratios were calculated per 1-SD increment in the measured level or as compared with the relevant reference category.
Hazard ratios were adjusted for sex, age, smoking status, systolic blood pressure, diabetic history, and levels of total and high-density lipoprotein (HDL) cholesterol and factor I or C-reactive protein.
Because these models were stratified by sex.
SD: Standard Deviation; yr: Year;
AFTCAPS: Air Force/Texas Coronary Atherosclerosis Prevention Study [15]; ARIC: Atherosclerosis Risk in Communities Study [16]; BRUN: Bruneck Study [17]; BWHHS: British Women’s Heart and Health Study [18]; COPEN: Copenhagen City Heart Study [19]; EAS: Edinburgh Artery Study [20]; FINE-FIN: Finland, Italy and Netherlands Elderly Study-Finland cohort [21]; FINE-IT: Finland, Italy and Netherlands Elderly Study-Italian cohort [21]; FINRISK97: Finrisk Cohort 1997 [22]; HISAYAMA: Hisayama Study [23]; IKNS: Ikawa, Kyowa, and Noichi Study [24]; LEADER: Lower Extremity Arterial Disease Event Reduction Trial [25]; MESA: Multi-Ethnic Study of Atherosclerosi [26]; NHANESIII: Third National Health and Nutrition Examination Survey [27]; NSHS: Nova Scotia Health Survey [28]; PROCAM: Prospective Cardiovascular Munster Study [29]; SHHEC: Scottish Heart Health Extended Cohort [30]; SHS: Strong Heart Study [31]; TARFS: Turkish Adult Risk Factor Study [32]; WOSCOPS: West of Scotland Coronary Prevention Study [33].
Factor I and C-reactive proteins as biomarkers: Incremental values in C-index
To the prognostic model for first coronary artery disease events included data on sex, age, smoking status, systolic blood pressure, diabetic history, along with total and high-density lipoprotein cholesterol’s level, the addition of HDL cholesterol led to increase in C-index by 0.0101 (Figure 1; Table 2). And, addition of factor I and C-reactive proteins biomarkers of inflammation to this model enhanced the C-index by 0.0045 for factor I and by 0.0053 for C-reactive protein (Figure 1; Table 3). Thus factor I and C-reactive proteins biomarkers yielded 0.91% and 1.61% respective net reclassification improvement for the predicted decade risk categories of high (> 20%), medium (10% to < 20%), and low (< 10%) (P < 0.02; data not shown). The corresponding values for the integrated discrimination index were 0.0029 and 0.0038 (data not shown). We can easily observe in Figure 1 that the combined predictive value of total and HDL cholesterol was higher than that of either factor I or C-reactive proteins biomarkers.
Figure 1.
Modifications in C-index for coronary artery disease risk prediction after addition of lipids markers and biomarkers (factor I protein or C-reactive protein) to a non-lipid based model. The non-lipid based model denotes age, smoking status, systolic blood pressure, diabetic history (stratified as per sex). CI denotes confidence interval and HDL denotes high-density lipoprotein.
Table 2.
Modifications in C-index and net reclassification improvement for coronary artery disease risk prediction from addition of lipids markers and factor I protein (or C-reactive protein) to a non-lipid based model
| C-index (95% CI) | C-index modifications (95% CI) | P value for C-index modification differences | Net reclassification improvement (95% CI) | |
|---|---|---|---|---|
| Factor I protein | 14 studies, 15,781 participants, 737 coronary artery disease cases | 9 studies, 7,051 participants, 424 coronary artery disease cases | ||
| Non-lipid based model | 0.6022 (0.5981, 0.6067) | Reference model | Reference model | |
| Plus total cholesterol | 0.6127 (0.6081, 0.6172) | 0.0107 (0.0084, 0.0130) | Reference | 1.45 (0.72, 2.17) |
| Plus total and HDL cholesterol | 0.6239 (0.6195, 0.6285) | 0.0217 (0.0198, 0.0236) | < 0.0001 | 2.64 (1.73, 3.51) |
| Plus total and HDL cholesterol and Factor I protein | 0.6284 (0.6239, 0.6330) | 0.0262 (0.0243, 0.0281) | < 0.0001 | 3.40 (2.44, 4.36) |
| C-reactive protein | 12 studies, 14,565 participants, 810 coronary artery disease cases | 6 studies, 6,344 participants, 408 coronary artery disease cases | ||
| Non-lipid based model | 0.6009 (0.5965, 0.6055) | Reference model | Reference model | |
| Plus total cholesterol | 0.6121 (0.5976, 0.6166) | 0.0112 (0.0093, 0.0131) | Reference | 1.22 (0.53, 1.93) |
| Plus total and HDL cholesterol | 0.6221 (0.6177, 0.6268) | 0.0213 (0.0194, 0.0232) | < 0.0001 | 1.89 (0.91, 2.78) |
| Plus total and HDL cholesterol and C-reactive protein | 0.6275 (0.6230, 0.6319) | 0.0266 (0.0247, 0.0285) | < 0.0001 | 3.41 (2.42, 4.42) |
The non-lipid based model denotes age, smoking status, systolic blood pressure, diabetic history (stratified as per sex). CI denotes confidence interval and HDL denotes high-density lipoprotein.
Table 3.
C-index for coronary artery disease after addition of biomarkers (factor I or C-reactive proteins) to conventional risk factors
| Reference model | Modifications after addition of biomarkers | |||||
|---|---|---|---|---|---|---|
|
|
|
|||||
| C-index (95% CI) | I2 (95% CI) | C-index modifications (95% CI) | Z value | p value | I2 (95% CI) | |
| Factor I protein (14 studies, 15,781 participants, 737 coronary artery disease cases) | ||||||
| Age | 0.5530 (0.5485, 0.5576) | 96 (95, 97) | 0.0163 (0.0141, 0.0181) | 13.9 | < 0.0001 | 79 (69, 88) |
| Above plus smoking status | 0.5705 (0.5661, 0.5751) | 95 (94, 96) | 0.0131 (0.0106, 0.0154) | 10.5 | < 0.0001 | 65 (44, 84) |
| Above plus SBP* | 0.5931 (0.5885, 0.5976) | 95 (94, 96) | 0.0092 (0.0076, 0.0109) | 9.1 | < 0.0001 | 48 (27, 69) |
| Above plus diabetic history | 0.6022 (0.5981, 0.6067) | 95 (94, 96) | 0.0065 (0.0045, 0.0086) | 8.3 | < 0.0001 | 40 (19, 61) |
| Above plus total cholesterol | 0.6127 (0.6081, 0.6172) | 95 (94, 96) | 0.0059 (0.0046, 0.0074) | 7.3 | < 0.0001 | 39 (8, 68) |
| Above plus HDL cholesterol | 0.6239 (0.6195, 0.6285) | 95 (94, 96) | 0.0045 (0.0036, 0.0054) | 5.1 | < 0.0001 | 18 (0, 39) |
| C-reactive protein (12 studies, 14,565 participants, 810 coronary artery disease cases) | ||||||
| Age | 0.5115 (0.5076, 0.5159) | 95 (94, 96) | 0.0161 (0.0144, 0.0176) | 13.2 | < 0.0001 | 68 (47, 89) |
| Above plus smoking status | 0.5236 (0.5190, 0.5280) | 95 (94, 96) | 0.0146 (0.0131, 0.0162) | 11.3 | < 0.0001 | 47 (26, 68) |
| Above plus SBP* | 0.5948 (0.5902, 0.5992) | 95 (94, 96) | 0.0104 (0.0090, 0.0119) | 9.2 | < 0.0001 | 32 (11, 51) |
| Above plus diabetic history | 0.6009 (0.5965, 0.6055) | 94 (93, 95) | 0.0081 (0.0064, 0.0096) | 8.4 | < 0.0001 | 25 (5, 44) |
| Above plus total cholesterol | 0.6121 (0.5976, 0.6166) | 94 (93, 95) | 0.0063 (0.0051, 0.0075) | 8.1 | < 0.0001 | 17 (0, 37) |
| Above plus HDL cholesterol | 0.6221 (0.6177, 0.6268) | 94 (93, 95) | 0.0053 (0.0044, 0.0062) | 6.5 | < 0.0001 | 4 (0, 34) |
SBP, systolic blood pressure;
CI denotes confidence interval; and HDL denotes high-density lipoprotein.
In analyses that included measures of family history of coronary artery disease, body-mass index, or both risk factors (Table 4) both the biomarkers have shown similar effect, especially, in those analyses which had excluded extreme factor I and C-reactive protein values and also in those analyses which had excluded participants known to be taking blood pressure or lipid levels lowering medications at study entry (data not shown).
Table 4.
Risk determination (in terms of C-index) for coronary artery disease after addition of biomarkers (factor I or C-reactive proteins) to conventional risk factors plus body mass index and family history of the disease
| Reference model | Modifications after addition of biomarkers | |||||
|---|---|---|---|---|---|---|
|
|
|
|||||
| C-index (95% CI) | I2 (95% CI) | C-index modifications (95% CI) | Z value | p value | I2 (95% CI) | |
| Factor I protein (7 studies, 7,927 participants, 433 coronary artery disease cases) | ||||||
| Conventional risk factors# | 0.6372 (0.6322, 0.6423) | 95 (93, 96) | 0.016 (0.011, 0.021) | 3.9 | < 0.0001 | 39 (29, 47) |
| Above plus body mass index | 0.6371 (0.6322, 0.6423) | 95 (93, 96) | 0.0020 (0.0013, 0.0027) | 4.2 | < 0.0001 | 35 (14, 53) |
| Above plus family history of coronary artery disease | 0.6387 (0.6339, 0.6441) | 95 (93, 96) | 0.0020 (0.0013, 0.0027) | 4.2 | < 0.0001 | 28 (0, 48) |
| C-reactive protein (9 studies, 8,669 participants, 529 coronary artery disease cases) | ||||||
| Conventional risk factors# | 0.6333 (0.6282, 0.6381) | 95 (93, 96) | 0.021 (0.014, 0.027) | 3.5 | < 0.0001 | 28 (0, 52) |
| Above plus body mass index | 0.6332 (0.6281, 0.6380) | 95 (93, 96) | 0.0023 (0.0014, 0.0065) | 4.2 | < 0.0001 | 21 (0, 46) |
| Above plus family history of coronary artery disease | 0.6345 (0.6295, 0.6394) | 95 (93, 96) | 0.0023 (0.0014, 0.0065) | 4.2 | < 0.0001 | 14 (0, 39) |
Conventional risk factors include age, smoking status, systolic blood pressure, Diabetic history, total cholesterol and HDL cholesterol (stratified by sex);
CI denotes confidence interval.
Interestingly, in those analyses which had excluded one risk factor at a time from conventional risk factors plus measures of factor I and C-reactive protein values based model and thus we observed broadly a similar pattern when compared to those analyses which had excluded the values of factor I and C-reactive protein levels (Figure 2).
Figure 2.
Modifications in C-index for coronary artery disease risk prediction after exclusion of individual risk factors from a full model containing conventional risk factors and biomarkers (factor I protein or C-reactive protein). CI denotes confidence interval and HDL denotes high-density lipoprotein. *14 studies, 737 coronary artery disease cases; #12 studies, 810 coronary artery disease cases.
Factor I and C-reactive proteins as biomarkers: Assessment in clinically relevant subgroups
We observed that the factor I and C-reactive proteins have shown improved coronary artery disease risk discrimination only in men (P < 0.001 for the interaction) (Figure 3). Besides, C-reactive protein as biomarker was efficient in current smokers (P < 0.001 for the interaction) (Figure 3). However, there were no significant differences in coronary artery disease risk discrimination observed in other clinically relevant subgroups (Figure 3). Nevertheless, in further analyses of stratified participants as per their exclusion as smokers (or to their smoking status) we analyzed the persistence of the differences in the improvement of coronary artery disease risk discrimination with change in sex with the use of factor I and C-reactive proteins levels. The similar trend was observed (although information on such treatments was incomplete) in analyses in which hormonal treatment receiving women at baseline were not incorporated.
Figure 3.
Modifications in C-index for coronary artery disease risk prediction with addition of information on biomarkers ((A) factor I protein or (B) C-reactive protein) to a model with conventional risk factors (based on the subgroups). CI denotes confidence interval. Only Studies with information on all subgroup levels were used to fulfill the comparison purpose and hence, subgroup total numbers do not always equal overall total numbers. *For heterogeneity.
Reclassification from 8 studies that involved only both the genders as 4,510 men with 456 first coronary artery disease events and 6,795 women with 367 first coronary artery disease events with at least 10 years of follow-up to further explore sex-specific findings was assessed. In these studies, we observed, the net reclassification improvement along with C-reactive protein’s measurement among men was 1.31% (95% confidence interval [CI], -0.17 to 2.82; P = 0.08). On the other hand the same characteristic value among the opposite gender was 0.41% (95% CI, -0.34 to 1.57; P = 0.51). According to sex, the effect of modification was not observed with measures of either systolic blood pressure or HDL cholesterol (data not shown). In the similar way, when analyses of factor I and C-reactive proteins levels were conducted with the data that included participants of three age groups (45- < 55 years, 55- < 65 years, and ≥ 65 years; P = 0.59) we observed no significant effect of modification according to the change in age groups.
Estimation of factor I and C-reactive proteins as biomarkers’ potential for coronary artery disease prevention
To investigate the potential of factor I and C-reactive proteins as targeted biomarkers in people whose risk of a coronary artery disease risk event was 10% to < 20% over the subsequent 10 years we modeled this work. In a population of 15,000 adults (of age ≥ 45 years) 2,254 persons were classified as a predicted risk feature of coronary artery disease of 10% to < 20% over a period of 10 years (when only conventional risk factors were used for risk calculation). According to the Adult Treatment Panel III guidelines [34] allocation to statin treatment was conducted and found that 1,979 participants at “medium” risk were ineligible for statin treatment at that time. In addition to this, the assessment of factor I in these participants (1,979) reclassified 95 participants (4.8%) to a predicted risk of ≥ 20%. Interestingly, approximately 20 out of 95 participants were expected to have a coronary artery disease risk within 10 years; correspondingly, additional assessment of C-reactive protein in the same participants had reclassified 113 participants (5.7%) to a predicted risk of ≥ 20%. Besides, approximately 25 out of these 113 participants were expected to have a coronary artery disease risk within a decade. When we assumed that a predicted risk of ≥ 20% reclassified person would begin statin therapy then as per Adult Treatment Panel III guidelines, such targeted factor I we found that about 4 (i.e. 0.20 × 20) additional coronary artery disease risk over a 10-year period could be prevented; the corresponding assessment of C-reactive protein led us to a result where about 5 (i.e. 0.20 × 25) additional coronary artery disease risk over a period of 10 years could be prevented.
We may say that targeted assessment of factor I in people at “medium” risk for a coronary artery disease risk could prevent an additional event over a decade for every 494 (1,979 ÷ 4) screened participants which has also resulted in about 23 additional participants began with statin therapy (i.e. 4.8% × 494). Similarly the corresponding would be screened number with targeted C-reactive protein assessment were about 396 (1,979 ÷ 5).
Discussion
In the present studies the effects of addition of factor I or C-reactive protein to the standard risk factors were evaluated to predict the risk of a first coronary artery disease risk in the analyses comprising of individual participant data from 19 anticipated cohort studies. Thus to establish the utility of factor I or C-reactive proteins as biomarkers of inflammation in coronary artery disease we then modeled a scenario in which factor I or C-reactive protein was assessed in participants having “medium” risk (a predicted risk of 10% to < 20% over a decade) after initial screening by using mere factors of conventional risk. As per the recommendations of the Adult Treatment Panel-III criteria [34], when measurement of a biomarker of inflammation coupled with initiation of statin therapy, we achieved a result data which suggest that one extra coronary artery disease risk would be prevented in the decade for approximately every 495 people were being assessed with factor I protein levels or approximately every 396 people in which C-reactive protein levels were assessed, on account of the initiation of statin therapy in about 23 additional participants. Our studies also suggested that analyses in which both factor I and C-reactive proteins levels were included had similar outcome in comparison to those in which either of the protein biomarker was used.
We assumed in our main model that the assessment of factor I or C-reactive protein would provide similar predictions of the coronary artery disease risk across population subgroups. Surprisingly, we found from an exploratory analysis that these biomarkers improve risk discrimination characteristic only in one gender (men but no significant improvement was observed in women). This is the limitation of our work that we could not find a proper explanation of this kind of result (sex-differentiated outcome of biomarkers). This area should be explored further to find out a good explanation for the same. Our results also confirm that we have developed such risk prediction scores which have tendency to combine the results of fatal and nonfatal coronary artery disease.
This study has its strengths and potential limitations. We involved analyses of population of several countries along with China and found harmonious outcomes among most of the measures based on risk discrimination and reclassification. We would like to acknowledge that even we used a standard decade-time frame along with conventional clinical risk categories, but these types of reclassification analyses are very fragile and highly dependent on how we chose the risk categories and landmark time.
We also observed that the differences due to studies in levels of factor I or C-reactive protein biomarkers of inflammation contributed relatively very little to the heterogenic characteristic feature. The reason behind this would be mostly to differing age ranges across cohorts. We assumed that all eligible people would receive statin and hence it might be possible that we overestimated the potential benefits of statin in our models. Besides this, we had incomplete information on statin use, and hence estimation of individual risk factors’ or risk models’ effect on results might be influenced.
At this point we would like to issue some points that measurement of factor I or C-reactive protein biomarkers of inflammation (used in present studies) must be checked at the comparative level using other emerging biomarkers (i.e. imaging biomarkers) with respect to their practicability, clinical benefit and cost-effectiveness.
In conclusion, after analyzing individual records of 19 cohort studies in the due process of investigation of the value of addition of factor I or C-reactive protein levels to conventional models to predict coronary artery disease risk among people of no background of coronary artery disease. We also estimated that initial screening having conventional risk factors, under current treatment guidelines, has led the additional assessment of factor I or C-reactive protein levels in people at “medium” risk for a coronary artery disease risk to prevent an additional event over a period of decade for every 390-500 screened participants and hence established the utility of factor I and C-reactive proteins as biomarker of inflammation for coronary artery disease.
Acknowledgements
The authors thank members of China-Japan Union Hospital of Jilin University, Changchun, Jilin, China for their help in subject recruitment throughout the period of the study.
Disclosure of conflict of interest
None.
References
- 1.Calabro P, Golia E, Yeh ET. CRP and risk of atherosclerotic events. Semin Immunopathol. 2009;31:79–94. doi: 10.1007/s00281-009-0149-4. [DOI] [PubMed] [Google Scholar]
- 2.Jialal I, Devaraj S, Venugopal SK. C-reactive protein: risk marker or mediator in atherothrombosis? Hypertension. 2004;44:6–11. doi: 10.1161/01.HYP.0000130484.20501.df. [DOI] [PubMed] [Google Scholar]
- 3.Deodhar SD. C-reactive protein: the best laboratory indicator available for monitoring disease activity. Cleve Clin J Med. 1989;56:126–130. doi: 10.3949/ccjm.56.2.126. [DOI] [PubMed] [Google Scholar]
- 4.Genest J, McPherson R, Frohlich J, Anderson T, Campbell N, Carpentier A, Couture P, Dufour R, Fodor G, Francis GA, Grover S, Gupta M, Hegele RA, Lau DC, Leiter L, Lewis GF, Lonn E, Mancini GB, Ng D, Pearson GJ, Sniderman A, Stone JA, Ur E. 2009 Canadian Cardiovascular Society/Canadian guidelines for the diagnosis and treatment of dyslipidemia and prevention of cardiovascular disease in the adult-2009 recommendations. Can J Cardiol. 2009;25:567–579. doi: 10.1016/s0828-282x(09)70715-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Pearson TA, Mensah GA, Alexander RW, Anderson JL, Cannon RO 3rd, Criqui M, Fadl YY, Fortmann SP, Hong Y, Myers GL, Rifai N, Smith SC Jr, Taubert K, Tracy RP, Vinicor F Centers for Disease Control and Prevention and American Heart Association. Markers of inflammation and cardiovascular disease: application to clinical and public health practice: a statement for healthcare professionals from the Centers for Disease Control and Prevention and the American Heart Association. Circulation. 2003;107:499–511. doi: 10.1161/01.cir.0000052939.59093.45. [DOI] [PubMed] [Google Scholar]
- 6.Graham I, Atar D, Borch-Johnsen K, Boysen G, Burell G, Cifkova R, Dallongeville J, De Backer G, Ebrahim S, Gjelsvik B, Herrmann-Lingen C, Hoes A, Humphries S, Knapton M, Perk J, Priori SG, Pyorala K, Reiner Z, Ruilope L, Sans-Menendez S, Scholte op Reimer W, Weissberg P, Wood D, Yarnell J, Zamorano JL, Walma E, Fitzgerald T, Cooney MT, Dudina A European Society of Cardiology (ESC) Committee for Practice Guidelines (CPG) European guidelines on cardiovascular disease prevention in clinical practice: executive summary. Eur Heart J. 2007;28:2375–2414. doi: 10.1093/eurheartj/ehm316. [DOI] [PubMed] [Google Scholar]
- 7.NACB LMPG Committee Members. Myers GL, Christenson RH, Cushman M, Ballantyne CM, Cooper GR, Pfeiffer CM, Grundy SM, Labarthe DR, Levy D, Rifai N, Wilson PW. National Academy of Clinical Biochemistry Laboratory Medicine Practice Guidelines: emerging biomarkers for primary prevention of cardiovascular disease. Clin Chem. 2009;55:378–384. doi: 10.1373/clinchem.2008.115899. [DOI] [PubMed] [Google Scholar]
- 8.Greenland P, Alpert JS, Beller GA, Benjamin EJ, Budoff MJ, Fayad ZA, Foster E, Hlatky MA, Hodgson JM, Kushner FG, Lauer MS, Shaw LJ, Smith SC Jr, Taylor AJ, Weintraub WS, Wenger NK, Jacobs AK, Smith SC Jr, Anderson JL, Albert N, Buller CE, Creager MA, Ettinger SM, Guyton RA, Halperin JL, Hochman JS, Kushner FG, Nishimura R, Ohman EM, Page RL, Stevenson WG, Tarkington LG, Yancy CW American College of Cardiology Foundation; American Heart Association. 2010 ACCF/AHA guideline for assessment of cardiovascular risk in asymptomatic adults: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. J Am Coll Cardiol. 2010;56:e50–e103. doi: 10.1016/j.jacc.2010.09.001. [DOI] [PubMed] [Google Scholar]
- 9.Cox DR. Regression models and life tables. J Roy Stat Soc B. 1972;74:187–220. [Google Scholar]
- 10.Harrell FE Jr, Califf RM, Pryor DB, Lee KL, Rosati RA. Evaluating the yield of medical tests. JAMA. 1982;247:2543–2546. [PubMed] [Google Scholar]
- 11.Royston P, Sauerbrei W. A new measure of prognostic separation in survival data. Stat Med. 2004;23:723–748. doi: 10.1002/sim.1621. [DOI] [PubMed] [Google Scholar]
- 12.Fibrinogen Studies Collaboration. Measures to assess the prognostic ability of the stratified Cox proportional hazards model. Stat Med. 2009;28:389–411. doi: 10.1002/sim.3378. [DOI] [PubMed] [Google Scholar]
- 13.Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med. 2002;21:1539–1558. doi: 10.1002/sim.1186. [DOI] [PubMed] [Google Scholar]
- 14.Pencina MJ, D’Agostino RB Sr, D’Agostino RB Jr, Vasan RS. Evaluating the added predictive ability of a new marker: from area under the ROC curve to reclassification and beyond. Stat Med. 2008;27:157–172. doi: 10.1002/sim.2929. [DOI] [PubMed] [Google Scholar]
- 15.Downs JR, Beere PA, Whitney E, Clearfield M, Weis S, Rochen J, Stein EA, Shapiro DR, Langendorfer A, Gotto AM Jr. Design & rationale of the Air Force/Texas Coronary Atherosclerosis Prevention Study (AFCAPS/TexCAPS) Am J Cardiol. 1997;80:287–293. doi: 10.1016/s0002-9149(97)00347-0. [DOI] [PubMed] [Google Scholar]
- 16.The ARIC Investigators. The Atherosclerosis Risk in Communities (ARIC) Study: design and objectives. Am J Epidemiol. 1989;129:687–702. [PubMed] [Google Scholar]
- 17.Willeit J, Kiechl S. Prevalence and risk factors of asymptomatic extracranial carotid artery atherosclerosis. A population-based study. Arterioscler Thromb. 1993;13:661–668. doi: 10.1161/01.atv.13.5.661. [DOI] [PubMed] [Google Scholar]
- 18.Lawlor DA, Ebrahim S, Davey Smith G. The association between components of adult height and Type II diabetes and insulin resistance: British Women’s Heart and Health Study. Diabetologia. 2002;45:1097–1106. doi: 10.1007/s00125-002-0887-5. [DOI] [PubMed] [Google Scholar]
- 19.Schnohr P, Jensen G, Nyboe J, Eybjaerg HA. The Copenhagen City Heart Study. A prospective cardiovascular population study of 20,000 men and women. Ugeskr Laeger. 1977;139:1921–1923. [PubMed] [Google Scholar]
- 20.Fowkes FG, Housley E, Cawood EH, Macintyre CC, Ruckley CV, Prescott RJ. Edinburgh Artery Study: prevalence of asymptomatic and symptomatic peripheral arterial disease in the general population. Int J Epidemiol. 1991;20:384–392. doi: 10.1093/ije/20.2.384. [DOI] [PubMed] [Google Scholar]
- 21.Menotti A, Mulder I, Nissinen A, Feskens E, Giampaoli S, Tervahauta M, Kromhout D. Cardiovascular risk factors and 10-year all-cause mortality in elderly European male populations; the FINE study. Finland, Italy, Netherlands, Elderly. Eur Heart J. 2001;22:573–579. doi: 10.1053/euhj.2000.2402. [DOI] [PubMed] [Google Scholar]
- 22.Laatikainen T, Critchley J, Vartiainen E, Salomaa V, Ketonen M, Capewell S. Explaining the decline in coronary heart disease mortality in Finland between 1982 and 1997. Am J Epidemiol. 2005;162:764–773. doi: 10.1093/aje/kwi274. [DOI] [PubMed] [Google Scholar]
- 23.Arima H, Kubo M, Yonemoto K, Doi Y, Ninomiya T, Tanizaki Y, Hata J, Matsumura K, Iida M, Kiyohara Y. High-Sensitivity C-Reactive Protein and Coronary Heart Disease in a General Population of Japanese: The Hisayama Study. Arterioscler Thromb Vasc Biol. 2008;28:1385–1391. doi: 10.1161/ATVBAHA.107.157164. [DOI] [PubMed] [Google Scholar]
- 24.Iso H, Naito Y, Sato S, Kitamura A, Okamura T, Sankai T, Shimamoto T, Iida M, Komachi Y. Serum triglycerides and risk of coronary heart disease among Japanese men and women. Am J Epidemiol. 2001;153:490–499. doi: 10.1093/aje/153.5.490. [DOI] [PubMed] [Google Scholar]
- 25.Meade TW. Design and intermediate results of the Lower Extremity Arterial Disease Event Reduction (LEADER) trial of bezafibrate in men with lower extremity arterial disease. Curr Control Trials Cardiovasc Med. 2001;2:195–204. doi: 10.1186/cvm-2-4-195. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Bild DE, Bluemke DA, Burke GL, Detrano R, Diez Roux AV, Folsom AR, Greenland P, Jacob DR Jr, Kronmal R, Liu K, Nelson JC, O’Leary D, Saad MF, Shea S, Szklo M, Tracy RP. Multi-Ethnic Study of Atherosclerosis: Objectives and Design. Am J Epidemiol. 2002;156:871–881. doi: 10.1093/aje/kwf113. [DOI] [PubMed] [Google Scholar]
- 27.Gardner CD, Winkleby MA, Fortmann SP. Population frequency distribution of non-high-density lipoprotein cholesterol (Third National Health and Nutrition Examination Survey [NHANES III] , 1988-1994) Am J Cardiol. 2000;86:299–304. doi: 10.1016/s0002-9149(00)00918-8. [DOI] [PubMed] [Google Scholar]
- 28.Davidson KW, Schwartz JE, Kirkland SA, Mostofsky E, Fink D, Guernsey D, Shimbo D. Relation of inflammation to depression and incident coronary heart disease (from the Canadian Nova Scotia Health Survey [NSHS95] Prospective Population Study) Am J Cardiol. 2009;103:755–761. doi: 10.1016/j.amjcard.2008.11.035. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Heinrich J, Balleisen L, Schulte H, Assmann G, van de LJ. Fibrinogen and factor VII in the prediction of coronary risk. Results from the PROCAM study in healthy men. Arterioscler Thromb. 1994;14:54–59. doi: 10.1161/01.atv.14.1.54. [DOI] [PubMed] [Google Scholar]
- 30.Smith WC, Crombie IK, Tavendale R, Irving JM, Kenicer MB, Tunstall PH. The Scottish Heart Health Study: objectives and development of methods. Health Bull (Edinb) 1987;45:211–217. [PubMed] [Google Scholar]
- 31.Lee ET, Welty TK, Fabsitz R, Cowan LD, Le NA, Oopik AJ, Cucchiara AJ, Savage PJ, Howard BV. The Strong Heart Study. A study of cardiovascular disease in American Indians: design and methods. Am J Epidemiol. 1990;132:1141–1155. doi: 10.1093/oxfordjournals.aje.a115757. [DOI] [PubMed] [Google Scholar]
- 32.Onat A, Sansoy V, Yildirim B, Keles I, Uysal O, Hergenc G. C-reactive protein and coronary heart disease in western Turkey. Am J Cardiol. 2001;88:601–607. doi: 10.1016/s0002-9149(01)01799-4. [DOI] [PubMed] [Google Scholar]
- 33.A coronary primary prevention study of Scottish men aged 45-64 years: trial design. The West of Scotland Coronary Prevention Study Group. J Clin Epidemiol. 1992;45:849–860. doi: 10.1016/0895-4356(92)90068-x. [DOI] [PubMed] [Google Scholar]
- 34.Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults. Executive summary of the third report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III) JAMA. 2001;285:2486–2497. doi: 10.1001/jama.285.19.2486. [DOI] [PubMed] [Google Scholar]