Introduction

By sealing dissection planes and enlarging luminal dimensions, implantation of coronary artery stents reduces the risk of early and late recurrent ischemia and reocclusion of the infarct-related artery compared to balloon angioplasty alone in patients undergoing primary percutaneous coronary intervention (PCI) for ST-segment elevation myocardial infarction (STEMI) [1]. This reduction in restenosis has resulted in a decrease in the need for repeat coronary intervention, but did not reduce mortality or reinfarction. Compared with bare-metal stents, drug-eluting stents (DES) are associated with a further reduction in in-stent late loss and repeat revascularization with similar rates of death, reinfarction, and stent thrombosis [2]. Nonetheless, repeat revascularization is still performed in ~5–7 % of patients treated with primary PCI and DES for STEMI, and angiographic restenosis occurs in ~15–20 % of lesions [3].

Restenosis after stent implantation is caused by a multifactorial process including neointima formation, smooth muscle proliferation, and lipid oxidation [4, 5]. Thrombus may also participate in the restenotic process [6]. The use of inflammatory and thrombotic cardiac biomarkers may help to identify patients at high risk for clinical and angiographic restenosis. The identification of a subgroup of patients at high risk to develop restenosis can be important to target specific pharmacologic or alternative therapies to reduce the risk of restenosis in these patients. We therefore performed an exploratory study to determine the relationship between 26 established and novel biomarkers and restenosis in STEMI patients treated with primary PCI and DES from a pre-specified substudy of the Harmonizing Outcomes With Revascularization and Stents in Acute Myocardial Infarction (HORIZONS-AMI) Trial.

Methods

Study population and design

The design and results of the HORIZONS-AMI trial have been previously published [7–10]. Briefly, 3,602 patients with STEMI were randomized in an open-label fashion and in a 1:1 ratio to treatment with bivalirudin alone (1,800 patients) or with unfractionated heparin plus a GPI (1,02 patients). Consecutive patients ≥18 years of age who presented within 12 h after the onset of symptoms and who had ST-segment elevation of 1 mm or more in two or more contiguous leads, new left bundle-branch block, or true posterior myocardial infarction were considered for enrollment. Emergency coronary angiography with left ventriculography was performed after randomization, with subsequent triage to treatment with PCI, coronary artery bypass grafting (CABG), or medical management at physician discretion. After patency was restored in the infarct-related vessel, eligible patients were randomly assigned again, in a 3:1 ratio, to either paclitaxel-eluting stents (PES, TAXUS Express, Boston Scientific) or an uncoated, but otherwise identical BMS (Express, Boston Scientific).

Clinical and angiographic follow-up

Clinical follow-up was planned at 30 days, 6, and 12 months, and then yearly for 3 years total. Routine angiographic follow-up at 13 months was pre-specified for all patients in the biomarkers substudy, with the exception of patients in whom stent thrombosis occurred or bypass graft surgery was performed within 30 days. Patients with documented restenosis before 13 months, or with clinically driven angiography after 6 months but before 13 months, were also considered to have met the angiographic follow-up requirement. If a patient had documented restenosis before 13 month, the pre-intervention in-stent late loss measurements were used.

Biomarkers substudy

A total of 501 patients randomly assigned to receive PES in the angiographic follow-up cohort were enrolled in the formal biomarker substudy. Venous blood samples were obtained at study enrollment before stent implantation, hospital discharge, 30 days and 1 year. A total of 26 inflammatory and thrombotic biomarkers were measured.

Biomarker assays

Biomarker values were determined by Alere Inc., San Diego, CA, using either Luminex or microtiter immunoassay methods. A detailed description methods and assays used is provided in the supplementary appendix.

Study endpoints and definitions

Primary endpoints of interest for the present study included in-stent late-loss and ischemia-driven target lesion revascularization (TLR). Quantitative coronary angiography was performed by an independent core laboratory (Cardiovascular Research Foundation, New York, NY). Late loss was defined as the difference between minimal luminal diameter post-procedure and minimal luminal diameter at follow-up. TLR was considered to be ischemia-driven if there was stenosis of at least 50 % of the target lesion, with ischemia as documented by a positive invasive or noninvasive functional study, ischemic changes on the electrocardiogram, or symptoms referable to the target lesion, or in the absence of documented ischemia, if there was stenosis of at least 70 % as assessed by quantitative coronary analysis at the independent core laboratory.

Net adverse clinical events (NACE) was defined as the composite of major adverse cardiac events (MACE) or major bleeding not related bypass graft surgery. MACE was defined as the composite of all-cause death, reinfarction, target vessel revascularization for ischemia, or stroke. Anemia was defined using WHO criteria as a hematocrit value at initial presentation <39 % for men and <36 % for women [11]. Creatinine clearance was calculated at baseline by the Cockcroft–Gault equation [12].

Statistical analysis

Categorical variables are presented as percentages and were compared with the Fisher’s exact test. Continuous variables are presented as medians with interquartile ranges, and were compared using Mann–Whitney U test. The Kaplan–Meier method was applied to estimate outcomes which were compared by log-rank test. For the purpose of the current analysis, patients were divided into tertiles according to biomarker values at admission, at 30 days, and according to differences in biomarker measurements between admission and 30 days. Associations between biomarker tertiles and in-stent late loss were investigated using the Kruskall–Wallis test. Associations between biomarker tertiles and ischemia-driven TLR were investigated using the χ2 test, and p for trend values are given. Biomarkers that were significant in univariate analysis were included in multivariate analysis to adjust for lesion length, reference vessel diameter, and diabetes mellitus. Multivariate Cox proportional hazards was used for the endpoint of TLR, and multivariate linear regression was used for the endpoint of in-stent late loss. Stepwise methods were used to create the models with entry and exit criteria set at p = 0.10.

Results

Patients

Figure 1 depicts the patient flow in the biomarker substudy. Of 2,257 patients randomized to receive PES, angiographic follow-up was intended in 1,308, 501 of whom (38.3 %) were enrolled in the biomarker substudy. Table 1 shows baseline demographic, angiographic and procedural characteristics in the cohort of PES-randomized patients enrolled in versus not enrolled in the biomarker substudy. Patients in the biomarker substudy had a slightly higher creatinine clearance, lower rates of prior MI and Killip class >1, but slightly higher rates of prior congestive heart failure. Complete assessment of all pre-specified biomarkers at admission was achieved in 454 patients (91 %).

Fig. 1
figure 1

Patient flow chart. HORIZONS-AMI harmonizing outcomes with revascularization and stents in acute myocardial infarction, DES drug-eluting stent, BMS bare metal stent

Full size image
Table 1 Demographic, angiographic and procedural characteristics in patients randomized to treatment with PES according to enrollment in the biomarker substudy
Full size table

Table 2 shows the 3-year clinical and 13-month angiographic outcomes in patients randomized to treatment with the PES stratified according to enrollment in the biomarker substudy. Patients in the biomarker substudy had lower rates of NACE, major bleeding and mortality at 3-year follow-up. The 3-year ischemia-driven TLR rate was similar in patients in the biomarker substudy (9.1 %) and patients not in the substudy (9.6 %, p = 0.63). There were also no significant differences in the rates of binary angiographic restenosis or in-stent or in-lesion late loss between both the groups at 13-month angiographic follow-up.

Table 2 Clinical and angiographic outcome in patients randomized to treatment with PES according to enrollment in biomarker substudy
Full size table

Relationship between biomarkers and late loss

Table 3 shows mean in-stent late loss at 13-month angiographic follow-up according to admission biomarker tertiles. Two biomarkers were identified which showed an inverse relationship with in-stent late loss. Low levels of Interleukin-6 (IL-6) and low levels of placental growth factor (PLGF) were associated with higher in-stent late loss at 13-month angiographic follow-up.

Table 3 Mean in-stent late loss at 13-month angiographic follow-up according to biomarker tertiles at admission
Full size table

Associations between in-stent late loss and 30-day biomarker, and differences between 30-day and admission biomarkers are shown in Table 4. Lower values of Interleukin 1 receptor antagonist (IL-1ra), matrix metalloproteinase 9 (MMP9), myeloperoxidase (MPO) were associated with greater late loss. Moreover, a decline between admission and 30 days in biomarker values of IL-1ra, MMP9, and monocyte chemotactic protein-1 (MCP-1) were associated with greater late loss.

Table 4 Mean in-stent late loss at 13-month angiographic follow-up according to biomarker tertiles measured at 30 days and tertiles of changes in biomarkers between admission and 30 days
Full size table

After multivariate adjustment, only MPO at 30 days and a decline of MCP-1 between admission and 30 days were independent predictors of lower in-stent late-loss (Table 7).

Relationship between biomarkers and ischemia-driven TLR

Table 5 shows 3-year ischemia-driven TLR rates according to admission biomarker tertiles. Three biomarkers showed an inverse relationship with ischemia-driven TLR. Lower levels of IL-6, PLGF and cardiotrophin-1 (CT-1) were associated with higher rates of ischemia-driven TLR at 3 years.

Table 5 3-Year ischemia-driven TLR rates according to biomarker tertiles measured at admission
Full size table

Table 6 shows 3-year ischemia-driven TLR rates according to 30-day biomarker tertiles and tertiles of differences between admission and 30-day biomarkers. Low values of IL-6 at 30 days were also associated with greater rates of ischemia-driven TLR.

Table 6 3-Year ischemia-driven TLR rates according to biomarker tertiles measured at 30 days, and tertiles of changes in biomarkers between admission and 30 days
Full size table

After multivariate analysis, only CT-1 at admission was an independent predictor of a lower TLR rate at 3-year follow-up (Table 7).

Table 7 Independent predictors of 3-year TLR and in-stent late loss at 13-months
Full size table

Discussion

This formal biomarker substudy from the large-scale HORIZONS-AMI trial identified a number of biomarkers that were associated with angiographic and/or clinical restenosis after PES implantation in STEMI. Low levels of IL-6 and PLGF at admission were associated with both higher in-stent late loss at 13-month angiographic follow-up and higher rates of ischemia-driven TLR at 3-year follow-up. Additionally, low admission levels of CT-1 were associated with higher rates of ischemia-driven TLR. At 30-day follow-up lower values of Il-1ra, MMP9, and MPO, and a decline relative to admission in IL-1ra, MCP-1, and MMP9 were associated with higher in-stent late loss. Finally, low values of IL-6 at 30 days were associated with ischemia-driven TLR. After multivariate adjustment, only MPO at 30 days and a decline of MCP-1 between admission and 30 days were independent predictors of lower in-stent late-los, and CT-1 was the only independent predictor of lower TLR.

The cytokines IL-1ra, IL-6, CT-1, and MCP-1 have been shown to contribute to atherosclerotic plaque development and plaque destabilization in animal models [13, 14]. However, little is known about their significance in the pathology of restenosis. Interestingly, IL-6 has also been found to have atheroprotective effects. For example, systemic IL-6 deficiency in a murine model enhanced atherosclerotic plaque formation [15]. It should be noted that there was only a weak association between IL-6 on admission and in-stent late loss with a univariate p-value of 0.49. Thus, further investigation in other studies is warranted. After multivariate analysis, CT-1 remained an independent predictor of TLR. CT-1 is a cytokine that signals via leukaemia inhibitory factor receptor gp130-dependent pathways and was originally described to induce cardiomocyte growth and survival [16]. This is the first study to identify CT-1 as a potential marker to predict clinical restenosis and further studies are needed to confirm or disprove our findings.

PLGF, a member of the vascular endothelial growth factor family, has been shown to stimulate arterial intimal thickening and macrophage accumulation in carotid arteries of cholesterol-fed rabbits [17]. In humans, PLGF has shown potential to predict death or myocardial infarction in patients with acute coronary syndromes [18]. Previous animal studies have reported a critical role for MMP9 in cell migration and intimal growth [19, 20]. Therefore, it has been hypothesized that MMP9 might play a key contributing role to the occurrence of restenosis, a theory which was not confirmed in this present study. Finally, 30-day plasma levels of MPO, a leukocyte-derived enzyme that catalyzes the formation of a number of reactive oxidant species and is associated with atherosclerosis in animal models, were also inversely related to restenosis in the current study [21].

This is the first study investigating the potential utility of biomarkers to predict restenosis in the setting of primary PCI with DES. Interestingly, all the aforementioned biomarkers showed an inverse relationship with angiographic and/or clinical restenosis in the HORIZONS-AMI biomarker substudy, both when measured at admission and when measured at 30-day follow-up. These findings may seem counter-intuitive as higher plasma levels of inflammatory markers were not associated with clinical and/or angiographic restenosis. Studies in the BMS era have reported an association between higher plasma C-reactive protein (CRP) levels and restenosis, while prior studies in the DES era reported the absence of an association between CRP and restenosis [22, 23]. The results of the current analyses also suggest that inflammation after DES implantation may have little relationship to restenosis.

In our study, only low levels of IL-6 and PLGF at admission were associated with both higher in-stent late loss and ischemia-driven TLR. The fact that biomarkers that are predictors for late loss are not automatically predictors for TLR can be explained by the fact that not every significant stenosis also causes anginal complaints, and may therefore not be revascularized, even if there is a considerable amount of late loss.

The early detection of patients at high risk of restenosis may be clinically relevant as it may help identify patients who might benefit from targeted pharmaceutical intervention or alternative therapies to reduce restenosis. For example, a recent study has suggested that triple antiplatelet therapy with cilostazol, a phosphodiesterase III inhibitor, in addition to aspirin and a thienopyridine may result in lower rates of restenosis in patients receiving DES [24, 25].

Limitations

Several limitations of this study need to be mentioned. As we measured 26 different biomarkers and evaluated each three times (admission, 30-day, and the difference between admission and 30-day), the possibility of spurious significant results due to chance cannot be excluded. Formal statistical correction for multiple comparisons was not applied. However, we consistently found lower rates of several inflammatory biomarkers to be associated with both angiographic and clinical restenosis, whether measured at admission, 30 days, or as the difference between 30 days and admission. Second, as PES were exclusively in HORIZONS-AMI, these results don’t apply to other DES, such as rapamycin-eluting stents. Third, prior studies have shown that there may be gender related differences in biomarker release during acute coronary syndromes, as only 22.5 % of patients in the biomarker substudy were women, we were not able to investigate gender differences in the current study. Finally, all patients underwent primary PCI for STEMI, and these results may therefore not be applicable to patients undergoing elective stent implantation.

Conclusions

Low values of IL-6, PLGF, and CT-1 measured at admission, IL-1ra, IL-6, MPO, and MMP9 measured at 30 days, and a decline between admission and 30 days in IL-1ra, MCP1, and MMP9 were significantly associated with clinical and/or angiographic restenosis after PES implantation for STEMI. Larger prospective trials to confirm and validate the utility of these biomarkers are warranted.