The recruitment of leukocytes initiates the atherogenic process. This occurs through expression on the surface of the endothelial cell, particularly when it encounters pro-inflammatory triggers of adhesion molecules, which increase the usually very transient or weak interaction with leukocytes and the endothelial surface.

Vascular cell adhesion molecule one, or VCAM-1, binds the leukocytes found in the early atherosclerotic lesion. Although VCAM-1 is not expressed on the normal endothelium or the unstimulated endothelial cell in culture, it is readily induced by cytokines, pro-inflammatory mediators, in the atherosclerotic plaque.

Libby and colleagues, in collaboration with other investigators, showed the kinetics of VCAM-1 expression in the cholesterol-fed rabbit. At about week one, a very early lesion of atherosclerosis is seen. At week three, VCAM, a molecular mediator of the adhesion of this leukocyte, is present in the endothelial cells. Ordinarily, the aortic endothelium of the rabbit does not express VCAM-1.

Transmigration of the monocyte into the intima occurs next. Because this requires directed migration due to a chemoattractant gradient, the effect of specialized cytokines that stimulate mononuclear cell chemoattraction has been of interest. MCP-1 is produced by endothelial smooth muscle cells (SMC) present when atherosclerosis begins and was localized in human and experimental atherosclerosis, and thus was an early candidate.

Descriptive biology has transitioned to molecular causality. Genetically modified animals are used to test in a very rigorous way the causal role of various inflammatory mediators in atherosclerosis.

A genetically modified mouse susceptible to atherosclerosis, for example, by deleting the low density lipoprotein (LDL) receptor gene so it develops fatty lesions when it consumes a cholesterol-enriched diet, was used to study this molecular causality. Further altering this model by inactivating MCP-1, thus introducing a genetically engineered mutation, markedly reduces lesion formation. In their work, animals that were atherosclerosis-susceptible and lacked MCP-1 had attenuation of lesion formation, compared to atherosclerosos-susceptible animals with wild-type MCP-1.

Chemokines and the types of leukocytes found in plaque are multiple. MCP-1 is not the only chemokine important in atherosclerosis. Work by other investigators has shown that IL-8 may be important by binding to CXCR2, and that a trio of gamma interferon inducible chemokine interacting with CXCR3 as a chemoattractant for the T cells are very important in the adaptive immune response in atherogenesis, and overexpression of eotaxin, another chemokine, in human atherosclerotic plaque. Eotaxin may be recruiting mast cells by binding to CCR3. Other work importantly linked inflammation and thrombosis, the ultimate complication of atherosclerosis, by describing stromal derived factor one, SDF-1, which interacts with CXCR4 as an activator of platelets. Platelet aggregation is a very uncommon property of the chemokines, but SDF-1 is a very potent platelet aggregator.

After chemoattraction is complete and the monocyte is resident in the intima, the phenotype changes and it expresses scavenger receptors that engulf modified lipoprotein particles. It also divides as frequently as SMC in the artery wall. The monocyte may become a macrophage foam cell. Macrophage colony stimulating factor (M-CSF), known to induce scavenger receptor, is a co-mitogen for mononuclear phagocytes, produced by cells present in the artery wall, may cause the change to a foam cell.

Work by Libby’s laboratory and other investigators showed a causal role of M-CSF in atherogenesis. With the inactivation of one or both alleles that encode M-CSF, there is a stepwise reduction in fatty lesion formation in the aortic root. There is a gene dosage dependent relationship between M-CSF and atherogenesis; marked diminution with inactivation of both alleles and lesser diminution with only one inactivated. Interestingly, this attenuation of atherogenesis occurs despite increased cholesterol levels.

Modern molecular genetics has led to a fundamental understanding of the initiation of atherosclerosis. It has implicated specific mediators in various steps of this process: VCAM-1 as one of the adhesion molecules, MCP-1 as one of the chemoattractants, and MCSF as one of the activators of the macrophage and stimuli to maturation of the monocyte.

More recent work implicates interleukin (IL)-18 in atherogenesis. IL-18 may have a very proximal role in a cytokine cascade, as it is an inducer of gamma interferon, a molecule implicated in atherosclerosis and known to be capable of producing many pro-atherogenic effects and stimulating both the innate and the adaptive immune response.

Libby’s group showed overexpression of IL-18, mostly by the macrophages in the plaque rather than the SMC or the endothelial cells.  However, the receptor of the alpha chain and the beta chain is expressed promiscuously by many cells in the plaque. They demonstrated a role for IL-18 as a novel step proximal to interferon gamma in the pro-inflammatory pathways of atherosclerosis. A surprising role of the SMC as a source of gamma interferon in the human atherosclerotic plaque was identified as a byproduct of this research. A sub-population of patients will have SMC that can be induced by a combination of IL-12 and IL-18 to make gamma interferon.

Compound mutant analysis of the role of IL-18 in atherosclerosis in vivo, by Libby’s laboratory in collaboration with another lab, showed an early attenuation of atherogenesis in the compound mutant mouse with inactivated IL-18. But, in the cholesterol-fed mice lesion formation continued. IL-18 is clearly important in the early phases of atherosclerosis.

Another signaling system implicated in atherogenesis provides a transition to clinical application. Work in Libby’s lab showed overexpression of CD40, the receptor, and CD-40 ligand in the human atherosclerotic plaque. CD40 is a TNF-receptor-like molecule; it binds CD40 ligand, CD154. Other work has also shown that CD154, the CD40 ligand, may have many pro-atherogenic effects.

Among the soluble cytokines, IL-1 and TNF poorly induced tissue factor, but CD40 ligation readily induces tissue factor expression in macrophages, activates caspase-1, which is important in processing IL-1 beta and IL-18 precursor to their active form, and can augment the expression of a special matrix metalloproteinase, MMP-11. These are properties not shared by the usual soluble cytokines.

In vivo testing of the role of CD40 signaling in the initiation and progression of atherosclerosis by Libby’s group showed that neutralization of CD40 signaling can inhibit atherogenesis in the experimental setting, along with an improvement in the inflammatory status of the plaque. Hence, not only was the lesion size decreased, but the lesion biology altered in an anti-inflammatory sense, as shown by the decreased expression of VCAM-1 when the CD40 ligand was neutralized. Further, this group showed that inhibiting the CD40 ligand arrests atherosclerosis progression.

Together with work by other investigators, it has been shown that CD40 signaling is important in the formation of atherosclerotic lesions in the evolution of established lesions. The character of these lesions was changed; the atherosclerotic plaque was stabilized, by an increase in the collagenous character of the plaque.

The ability to harness inflammation biology and apply it to patients has been a recent, exciting development. Soluble CD40 ligand is shed into the blood where it can be measured. Libby and colleagues, in collaboration with Ridker, tested whether or not plasma concentrations of soluble CD40 ligand might have diagnostic or prognostic value.

Using data from the Women’s Health Study and matched controls within the population, they found that increased levels of CD40 ligand is associated with an increase in stroke and acute coronary syndromes. Hence, initial basic science observations led to the concept of soluble CD40 as a marker of disease in humans.

Despite the established risk factors of cardiovascular disease (CVD), inflammatory markers as predictors of cardiovascular risk are needed. For example, the lipid profile does not identify the entire at-risk population; many patients have atherosclerotic complications, despite having below median cholesterol levels.

A schema of an inflammatory pathway in atherogenesis, based on the integrated findings from Libby’s laboratory and Ridker’s clinical trial findings was developed. Pro-inflammatory triggers or established risk factors elicit a first wave of cytokines, such as IL-1 or TNF-alpha, that increase the expression of IL-6 in an amplification loop: IL-6 travels to the liver and tells it to make acute phase reactants such as C-reactive protein, serum amyloid A, and many others, which can be sampled in venous blood. These inflammatory cytokines can increase the expression on the surface of the endothelial cell and promote the shedding of various adhesion molecules such as ICAM-1, which can be shed and tested in the blood. 

C-reactive protein (CRP) as an inflammatory marker of disease has been the recent focus. CRP is very stable chemically, and thus does not require special care to preserve the samples, and the assay is reliable, internationally standardized, convenient, and inexpensive. CRP has a long half-life, no diurnal variation, and little fluctuation in its level in well persons.

Can inflammatory markers guide therapy?  Data from the Physicians Health Study show that the presence of dyslipidemia and inflammation results in greater risk than the presence of either alone. So, CRP adds to the predictive value of the total cholesterol-to-HDL ratio to determine the risk of a first myocardial infarction. The Women’s Healthy Study provided a robust data set to show that CRP was a better predictor of CV events and coronary heart disease than the gold-standard LDL level. By using both of these predictors, a greater number of at-risk patients can be identified.

CRP is a non-specific marker of inflammation, with both vascular and non-vascular stimuli for CRP. It is a global integrator of the inflammatory burden in the body. Some vascular sources of CRP are extra-coronary atheroma and the aorta. Obesity is an important non-vascular stimulus of inflammation and CRP. Adipose tissue is a source of pro-inflammatory cytokines. As body mass index increases, CRP increases.

The National Cholesterol Education Panel Adult Treatment Panel III guidelines stipulated that the presence of 3 of 5 different factors diagnosed the metabolic syndrome. In the Women's Health Study there was a stepwise increase in CRP as the number of these factors increased.

The onset of diabetes is preceded by inflammation, as shown by the Women's Health Study. Women in the higher quartiles of CRP at the study outset developed diabetes more often than their cohorts without elevated CRP, even after adjustment for body mass index. Hypertension is a pro-inflammatory stimulus. IL-6 and soluble ICAM-1, two markers of inflammation, increase in a stepwise fashion with increases in systolic blood pressure or pulse pressure.

Other potential non-vascular stimuli may apply to some patients.  For example, low grade systemic inflammation or infection, such as periodontal disease with gingivitis, a chronic bronchitis, or chronic prostatitis.

Recent guidelines from the American Heart Association (AHA) and the US Centers for Disease Control and Prevention (CDC) defined three strata of risk based on high-sensitivity CRP (hsCRP). CRP screening is likely to be useful in the population with a 10% to 20% ten-year risk of developing coronary heart disease. In this group, the results of CRP may help guide therapy; e.g., to initiate drug therapy, or motivate patients to adopt a healthy lifestyle. Screening the entire adult population for CRP is not cost-effective and the information gained would not change clinical practice, based on current knowledge. CRP should not be tested in persons with established coronary disease or in persons at low risk (<10% ten-year risk).

How should CRP screening be used in clinical practice? Measure CRP only in healthy people. The risk prediction does not work if the patient is ill.  Measure it twice, two weeks apart, and average the results. A CRP level greater than 10 mg/dL indicates the patient may have had the flu or may have an inflammatory condition, like lupus erythematosus or rheumatoid arthritis, which renders the test totally irrelevant in terms of cardiac risk prediction. Currently, there is no current evidence that the degree of CRP lowering predicts the outcome of treatment in a patient. Hence, serial testing of CRP or measuring CRP to determine the efficacy of treatment is not warranted.

An analysis by Ridker of the AFCAPS/TexCAPS study divided the study population into 4 groups: above or below median HDL:total cholesterol ratio, and above or below median CRP. In the group with both inflammation and dyslipidemia, treatment was very effective. The number needed to treat (NNT) to prevent 1 event was less than 100. Therapy was very cost effective. In the group with dyslipidemia but no inflammation, the NNT was 1000 patients, and treatment was not cost effective, as the exposure to risk was small.
In the group with inflammation and no dyslipidemia, the statin reduced the number of events and the NNT was similar to the group with dyslipidemia and inflammation.
This hypothesis-generating data set spurred the initiation of a clinical study (JUPITER) to provide evidence for whether or not inflammatory status can be used to guide therapy and reduce events.