MAZ is a synonym for serum amyloid A binding protein 1 (SAF-1), the function of which appears variable and incompletely understood. Activation and induction of MAZ by minimally modified LDL and lipopolysaccharides have been shown, leading to expression of serum amyloid A (SAA), a known inflammation-responsive protein within macrophages (6, 7). SAA has been long considered influential in atherosclerosis, with SAA mRNA having been identified in the endothelial and macrophage foam cells of atherosclerotic lesions of coronary and carotid arteries. Subsequently, MAZ has proven to be a transcription factor in active macrophages within atherosclerotic plaques triggering the formation of matrix metalloproteinases (MMPs) (8). Active macrophages are the hallmark of vulnerable plaques, with a high tendency to rupture. Originally identified in patients with coexisting CVD and autoimmune disease, MAZ-Ab has recently been evaluated in atherosclerosis using ¹⁸F-Fluorodeoxyglucose (¹⁸F-FDG) positron-emission and computer tomography (9). ¹⁸F-FDG uptake in vessel walls is known to correlate with macrophage activity in atherosclerotic plaques (10), with calculated plaque burden correlating with serum MAZ-Ab optical density (OD) (9).

Data regarding the further MAZ function have been emerging, with evidence supporting a role in certain cancers as regulator of oncogene transcription and angiogenesis (11–14). Furthermore, MAZ is known to regulate various inflammatory response genes (15), and associations to Alzheimer’s disease have been suggested in transcription factor analysis of 1,372-probe gene expression signatures (16).

The aim of this study was to examine the clinical relevance of MAZ-Ab in patients with a confirmed cardiovascular event, compared with healthy controls. Furthermore, the relationship of MAZ-Ab to traditional cardiovascular risk factors and its prognostic value were evaluated.

To assess the risk of premature ischemic heart disease, unselected adult patients below the age of 65 years admitted to Hannover Medical School between August 2007 and July 2011, and subsequently diagnosed with ACS were included. All patients provided both verbal and written informed consent before participation in the study, with the ethics committee of Hanover Medical School approving the study (Ethics Number: 2614). All patients completed a questionnaire assessing cardiac risk profile and provided a serum sample, which was subsequently used for MAZ-Ab testing. The questionnaire assessed age, gender, past-history of hypertension, hypercholesterolemia, statin treatment, diabetes, tobacco exposure, as well as previous vascular events including myocardial infarction and stroke. All patients underwent routine ACS diagnostics including laboratory investigations, electrocardiogram (ECG), and coronary angiography. Laboratory tests included N-terminal propeptide brain natriuretic peptide (NT-proBNP), cardiac TroponinT (cTnT), growth-differentiation factor-15 (GDF-15), total cholesterol, and low-density lipoproteins (LDL). Serum samples were stored at −70°C.

Acute coronary syndrome patients were divided into three groups: patients with ST-elevation myocardial infarction (STEMI), with non-ST elevation myocardial infarction (NSTEMI) and with unstable angina pectoris (AP).

Patients with AP required at least 1 angiographically documented stenosis ≥70% in a major coronary artery. Based on ECG changes and cTnT, using a decision threshold of 0.03 µg/L, STEMI or NSTEMI was diagnosed.

Serum samples of an apparently healthy control group, which has previously been described in detail (17), underwent MAZ-Ab testing. Control patients received cardiac magnetic resonance imaging with dobutamine or adenosine stress, 12-lead ECGs, and physical examination without any pathological findings. In addition, all patients exhibited normal serum creatinine, aspartate aminotransferase, alanine aminotransferase, thyroid-stimulating hormone, hemoglobin concentrations, leukocyte, platelet counts, oral glucose tolerance test, and N-terminal pro-B-type natriuretic peptide (NT-proBNP) levels. None were in receipt of current medications and had no conventional cardiovascular risk factors. All were aged ≥18 years at inclusion and previously provided written informed consent.

MAZ-Ab has been detected via protein array technique in patients with CVD as described in an earlier publication (9). A specially developed anti-MAZ-antibody ELISA kit (EUROIMMUN AG, Lübeck, Germany) based on our previously described ELISA protocol was used to test serum for MAZ-Ab (9).

MAZ-Ab optical densities were compared between ACS patients and controls, as well as within the ACS groups. Furthermore, MAZ-Ab values were correlated to traditional risk factors. To rule out an influence on MAZ-Ab by statins, LDL and cholesterol values of all patients with and without statins were analyzed additionally.

A well-defined group of 197 ACS patients, experiencing myocardial infarction, with up to 12 months subsequent follow-up were identified within the Platelet Inhibition and Patient Outcomes (PLATO) database (http://www.ClinicalTrials.gov NCT00391872) (18). An equal number (n = 199) of patients surviving 12 months follow-up without a subsequent spontaneous myocardial infarction were randomly drawn from the same cohort. Being matched for initial STEMI/NSTEMI, age, gender, and nationality the latter acted as controls. Serum taken at the time of initial admission was tested for MAZ-Ab in both groups.

Continuous variables were assumed non-parametric and tested using either the Mann–Whitney U test or Kruskal–Wallis test. Categorical variables were assessed using chi-square test or Fisher’s exact test. Correction for age and gender between groups was performed using Propensity Score matching, employing logistic regression with nearest neighbor matching and a 0.2 caliper. Prognostic relevance regarding reinfarction was calculated in conditional logistic regression. All p-values are two-tailed, with 95% cutoff being the declared level of significance. Statistical analysis was performed using IBM SPSS Statistics for Macintosh, Version 23 (IBM Corp. Armonk, USA), Prism 7 (GraphPad Software, La Jolla, CA, USA), and R Version 3.2.3 (R Foundation for Statistical Computing, Vienna, Austria).

Taking the entire cohort, the median age of the controls was significantly lower than the ACS group (41 vs. 53 years; p < 0.001). Similarly, a far greater proportion of the ACS group was male (87 vs. 47%; p = 0.001). To eliminate these potential confounders, a logistic regression analysis was performed using propensity score matching for age and gender between groups. In total, 206/270 patients (76%) were included in the sub-analysis, 168 from the ACS group and 38 control patients. The median ages were 49 [45–57] years among the controls and 53 [47–60] years in the ACS group (p = 0.46). While improved, a significant disparity in gender composition remained between groups, with a significantly greater male domination in the ACS group (83 vs. 68%; p = 0.003). Having eliminated age as a confounder, a significant difference in MAZ-Ab between groups remained with significantly higher ODs in the ACS group (0.47 vs. 0.24; p = 0.001 Figure 1).

Within the ACS group, 116/174 patients (67%) were diagnosed with a STEMI, 44 patients (25%) with NSTEMI, and 14 (8%) with AP. Within the different ACS phenotypes, no differences in MAZ-Ab OD were observed between groups (p = 0.682), all of which proved higher than that observed among controls (Figure 2), suggesting that MAZ-Ab was independent of the extent of myocardial damage occurring. Supplementing this further, a comparison of all the relevant blood markers (Table 1) revealed no significant correlations between these and MAZ-Ab OD.

Comprehensive data relating to the incidence of hypertension, diabetes mellitus, hyperlipidemia including treatment with statins as well as smoking habits and family history of CVD were available for all patients. No known risk factors were present among the entire control group. In the ACS group, the most prevalent risk factors were tobacco consumption (84%), hypertension (56%) and hyperlipidemia (49%). For all CVD risk factors assessed, MAZ-Ab optical densities were lower among control patients than in ACS patients without a given risk factor (Figures 3A–F). With the exceptions of evident hyperlipidemia at original admission and preexisting HMG-CoA reductase inhibitor use, no differences in MAZ-Ab OD were evident within the ACS group. Taken collectively, these findings would strongly support that MAZ-Ab is independent of hypertension, diabetes, dyslipidemia, smoking, and family history in terms of generating cardiovascular risk.

The findings regarding existing statin use are less clear (Figure 3D). In total, 35 patients in the ACS group were taking statins at admission. Considering the recommended European Society of Cardiology cutoff values for blood lipid concentrations (cholesterol <190 mg/dL, LDL <115 mg/dL), 6/35 (17%) still had elevated cholesterol at admission. 59/109 (54%) of those admitted without statin had an elevated cholesterol, showing no significant difference between the two groups (p = 0.384). Of those on statin 22/35 (63%) and 75/109 (69%) of those without had raised LDL levels at admission (p = 0.59).

Patients on statins did, however, have significantly higher MAZ-Ab ODs (p = 0.009). It is worth noting that while these patients were older 58 [52–60] vs. 52 [46–58] years (p = 0.004), no association between age and raised LDL (p = 0.145) or cholesterol (p = 0.553) was observed. Similarly, neither cholesterol >190 (p = 0.945) nor LDL >115 (p = 0.946) influenced MAZ-Ab OD. An alternative explanation, however, may exist that the findings are the result of a selection bias, in that only high MAZ-Ab patients on statins subsequently experienced an event. This, however, remains speculative and requires further evaluation in future studies.

Of the 197 PLATO patients experiencing a second cardiovascular event within one year of the initial ACS event, 42 died. No significant difference in MAZ-Ab OD between these patients and their 1997 matched controls was observed (OR 1.04 [95% CI: 0.94–1.16]; p = 0.436).