Matrix metalloproteinases and their tissue inhibitor after reperfused ST-elevation myocardial infarction treated with doxycycline. Insights from the TIPTOP trial

International Journal of Cardiology 197 (2015) 147–153

Contents lists available at ScienceDirect

International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard

Matrix metalloproteinases and their tissue inhibitor after reperfused ST-elevation myocardial infarction treated with doxycycline. Insights from the TIPTOP trial☆ Giampaolo Cerisano a,⁎, Piergiovanni Buonamici a, Anna Maria Gori b,c, Renato Valenti a, Roberto Sciagrà d, Betti Giusti b, Alice Sereni b, Silvia Raspanti d, Paolo Colonna e, Gian Franco Gensini a,b,c, Rosanna Abbate b, Richard Schulz f, David Antoniucci a a

Department of Heart and Vessels, Careggi Hospital, Florence, Italy Department of Experimental and Clinical Medicine, University of Florence, Italy c Don Carlo Gnocchi Foundation IRCCS, Florence, Italy d Nuclear Medicine Unit, Department of Clinical Physiopathology — University of Florence, Careggi Hospital, Florence, Italy e Division of Cardiology, Hospital Policlinico of Bari, Bari, Italy f Cardiovascular Research Centre, Mazankowski Alberta Heart Institute, Departments of Pediatrics and Pharmacology, University of Alberta, Edmonton, Alberta, Canada b

a r t i c l e

i n f o

Article history: Received 3 February 2015 Received in revised form 24 May 2015 Accepted 16 June 2015 Available online 18 June 2015 Keywords: Acute myocardial infarction Doxycycline Left ventricular remodeling Infarct size Metalloproteinases Tissue inhibitor of metalloproteinases

a b s t r a c t Background: The TIPTOP (Early Short-term Doxycycline Therapy In Patients with Acute Myocardial Infarction and Left Ventricular Dysfunction to Prevent The Ominous Progression to Adverse Remodelling) trial demonstrated that a timely, short-term therapy with doxycycline is able to reduce LV dilation, and both infarct size and severity in patients treated with primary percutaneous intervention (pPCI) for a first ST-elevation myocardial infarction (STEMI) and left ventricular (LV) dysfunction. In this secondary, pre-defined analysis of the TIPTOP trial we evaluated the relationship between doxycycline and plasma levels of matrix metalloproteinases (MMPs) and their tissue inhibitors (TIMPs). Methods: In 106 of the 110 (96%) patients enrolled in the TIPTOP trial, plasma MMPs and TIMPs were measured at baseline, and at post-STEMI days 1, 7, 30 and 180. To evaluate the remodeling process, 2D-Echo studies were performed at baseline and at 6 months. A 99mTc-SPECT was performed to evaluate the 6-month infarct size and severity. Results: Doxycycline therapy was independently related to higher plasma TIMP-2 levels at day 7 (p b 0.05). Plasma TIMP-2 levels above the median value at day 7 were correlated with the 6-month smaller infarct size (3% [0%–16%] vs. 12% [0%–30%], p = 0.002) and severity (0.55 [0.44–0.64] vs. 0.45 [0.29–0.60], p = 0.002), and LV dilation (−1 ml/m2 [from −7 ml/m2 to 9 ml/m2] vs. 3 ml/m2 [from −2 ml/m2 to 19 ml/m2], p = 0.04), compared to their counterpart. Conclusions: In this clinical setting, doxycycline therapy results in higher plasma levels of TIMP-2 which, in turn, inversely correlate with 6 month infarct size and severity as well as LV dilation. © 2015 Elsevier Ireland Ltd. All rights reserved.

1. Introduction

Abbreviations: STEMI, ST elevation acute myocardial infarction; pPCI, primary percutaneous coronary intervention; ECM, extracellular collagen matrix; MMPs, matrix metalloproteinases; TIMPs, tissue inhibitors of metalloproteinases; LV, left ventricular; LVEF, left ventricular ejection fraction; SPECT, 99mTc-sestamibi-gated single-photon emission computed tomography; LVEDVi, left ventricular end-diastolic volume index; 2D Echo, two dimensional echocardiographic. ☆ The study was not supported by any public grant or industry support. ⁎ Corresponding author at: Department of Heart and Vessels, Careggi Hospital, Largo Brambilla 3, I-50141, Florence, Italy. E-mail address: [email protected] (G. Cerisano).

http://dx.doi.org/10.1016/j.ijcard.2015.06.024 0167-5273/© 2015 Elsevier Ireland Ltd. All rights reserved.

The prevention of early left ventricular (LV) remodeling during ST elevation acute myocardial infarction (STEMI) is of critical importance, especially after a large STEMI [1]. Early remodeling leads to progressive LV dilation and dysfunction which affects outcome [2], and large STEMI patients are at increased risk for LV remodeling despite primary percutaneous coronary intervention (pPCI), and the current standard of drug therapy [1]. These results highlight the need of new therapeutic strategies to limit or prevent the remodeling process. The extracellular collagen matrix (ECM) plays a major role in postMI remodeling, and an activation of matrix metalloproteinases (MMPs) compared to their tissue inhibitors (TIMPs) has been linked

148

G. Cerisano et al. / International Journal of Cardiology 197 (2015) 147–153

to early ECM damage and LV dilation [3]. Accordingly, experimental studies have shown that gene deletion or pharmacological inhibition of MMP activity attenuates post-MI remodeling [4–8]. However, the structural changes of LV remodeling process have, at their base, both intra- and extracellular roots. Noteworthy, an imbalance between TIMPs and MMPs, particularly MMP-2, has been demonstrated intracellularly in isolated rat hearts subjected to ischemia and reperfusion injury [9]. This was found to be responsible for the acute contractile dysfunction and degradation of sarcomeric proteins [10–12] and, ultimately, for the infarct size [13]. Since infarct size is the major factor that promotes LV remodeling, the intracellular activation of MMPs could play a key role in this process. Unfortunately, the efforts to use synthetic MMP inhibitors in the clinical setting have failed due to the unanticipated adverse effects [14,15], and disappointing results on post-myocardial infarct remodeling process [16]. The rationale to use doxycycline to reduce post-MI remodeling is based on the fact that it is the most potent MMP inhibitor of the tetracycline class of antibiotics and exhibits MMP inhibition in vivo at blood levels lower than those required for its antibacterial effect [17]. It effectively crosses cell membranes, and accumulates preferentially at the site of tissue injury such as the damaged cardiomyocytes [18], making it appears to almost act as a smart drug [19]. Consistently, pre-clinical studies showed that doxycycline attenuates LV remodeling in a rat model of MI [20,21], prevents acute stunning in isolated rat-hearts subjected to ischaemia and reperfusion injury [12,22], and mimics the MMP-2 inhibition and infarct size reduction found with ischaemic post-conditioning [23]. Given the clinical availability and wellrecognized safety profile of tetracyclines as aforementioned, doxycycline use seemed to be an ideal starting point for clinical trials in STEMI patients. For the first time in the clinical setting, the TIPTOP (Early Short-term Doxycycline Therapy In Patients with Acute Myocardial Infarction and Left Ventricular Dysfunction to Prevent The Ominous Progression to Adverse Remodelling) trial [24] recently demonstrated that a timely shortterm therapy with doxycycline is safe and able to induce a significant decrease in LV dilation, infarct size and severity in patients successfully treated with pPCI for a first STEMI and LV dysfunction. In this secondary, pre-defined analysis of the TIPTOP trial, we set out to evaluate the time profiles of MMP and TIMP plasma levels. The results of this study may provide additional pathophysiological insights into the main study clinical findings. 2. Methods 2.1. Study design, patients and procedures The TIPTOP trial has been already described in detail [24]. In brief, it was a prospective, phase-2, single-centre, randomized, open-label controlled trial in which 110 patients older than 18 years with acute STEMI and LV ejection fraction (LVEF) b40% were randomly assigned in a 1:1 ratio to receive doxycycline or standard care. All patients were treated with pPCI, including stenting of the infarct-related artery, and received medical therapy for STEMI and LV dysfunction in accordance with standard and recommended practice. Doxycycline (Bassado; Pfizer Italia s.r.l.) was administered at 100 mg oral dose immediately after pPCI and then every 12 h for 7 days. The antimicrobial dose we used ensures plasma levels of doxycycline similar to those obtained with doses used in preclinical studies where doxycycline was shown to be effective in inhibiting MMPs and prevented abdominal aortic aneurysm [25] and post-infarction LV remodeling [20,21]. The Σ ST-segment elevation immediately before pPCI was considered as a surrogate of the area of myocardium at risk [26]. To evaluate the remodeling process two dimensional echocardiographic (2D Echo) studies were performed at baseline (immediately after pPCI) and at 6 months with a commercially available imaging system (Philips IE-33, Amsterdam, The Netherlands). 2D Echo and Doppler

studies were performed to obtain measurements of LV volumes, ejection fraction, mitral inflow E- and A-wave velocity, lateral mitral annulus E′ -and A′-wave velocity, and LV flow propagation velocity. Measurements were made using American Society of Echocardiography criteria [27]. According to the protocol of the main study [24], LV dysfunction was defined as LVEF b40%, as calculated by the on-site operator at the first echo examination in the coronary care unit immediately after pPCI. DICOM standardized images were recorded and stored on digital media, and sent to an independent, blinded, off-site core laboratory for analysis. The Core-Lab analysis was performed with a workstationbased system for 2D Echo visualization and image processing (Philips Xcelera R3.1L1, Amsterdam, The Netherlands). The mean Core-Lab inter-reader variability value for LV volumes analysis was 6.1% (ICC 0.90 with standard error 6.3 ml) and intra-reader variability was 4.1% (ICC 0.93 with standard error 3.8 ml). 99m Tc-sestamibi-gated single-photon emission computed tomography (SPECT) was performed and analyzed by two experts who were unaware of the treatment group assignments, to evaluate the final infarct size and severity at 6-month follow-up. Scintigraphic acquisition began 60 min after 99mTc-sestamibi injection (740 MBq) by the use of double-head gamma-camera equipped with high-resolution collimator, with a 180° rotation arc, 32 projections, 60 s per projection, 8 frames per heart cycle, and 64 × 64 matrices. Details about the modality of acquisition and analysis of scintigraphic data have been reported elsewhere [28]. To estimate the infarct size, perfusion defects were quantified as percentage of the left ventricular wall, with the defect threshold set at 60% of peak uptake. The infarct severity, as a measure of infarct transmurality, was defined as the lowest ratio of minimal to maximal counts in the short-axis slices evaluated for infarct size; therefore, the lower the severity index the greater the infarct transmurality. Furthermore, a coronary angiography was repeated at 6 months for the evaluation of infarct-related artery (IRA) patency. The patients were eligible for MMP/TIMP analysis when they consented for the main TIPTOP study. The study protocol was approved by local hospital ethical committee. Doxycycline was provided directly from the local hospital pharmacy, and the manufacturer had no role in the study. 2.2. Plasma collection Blood samples (5 cm3) were collected from a peripheral vein into chilled ethylenediamine-tetraacetic acid tubes. The samples were centrifuged (at room temperature at 1500 g for 15 min) and the supernatants were removed and the plasma aliquoted and stored at − 80 °C until assay. Plasma was used to measure MMP and TIMP profiles at baseline (early after PCI but prior to the administration of the first dose of doxycycline), and at post-MI days 1, 7, 30 and 180. 2.3. MMP and TIMP profiles For this study, representative MMPs from the different MMP classes were measured, specifically the interstitial collagenase MMP-8, the gelatinases (MMP-2 and MMP-9), and MMP-7 from the matrylisin subclass. The rationale for selecting these MMP types is that they have been identified in animal studies to be altered after MI and have been associated with matrix remodeling after myocardial injury [29]. All 4 known TIMPs (TIMP-1,-2,-3 and -4) were also measured because there is a growing awareness on their role on the post-MI remodeling process [30]. The quantification of MMP and TIMP species was performed using a bead-based multiplex immunoassay. Levels of different MMPs (MMP-2, MMP-7, MMP-8, and MMP-9) and TIMPs (TIMP-1, TIMP-2, TIMP-3 and TIMP-4) were determined using the Bio-Plex suspension array system (Bio-Rad Laboratories Inc., Hercules, CA, USA) and R&D Kits (R&D System, Milan Italy) following the manufacturer's instructions. The coefficient of variation of MMP and TIMP assays were 5.8% and 6.8%, respectively [31].

G. Cerisano et al. / International Journal of Cardiology 197 (2015) 147–153

2.4. Statistical analysis Discrete data were summarized as frequencies, whereas continuous data were summarized as mean ± SD or median and interquartile range, when appropriate. Differences in baseline characteristics between the two study groups were analyzed using the x2 test or Fisher's exact test for categorical variables, and the unpaired and paired, 2-tailed Student's t test or Mann–Whitney test for continuous variables. The temporal changes in MMP and TIMP plasma levels within the two study groups were analyzed by the Friedman's repeated measures test. Differences in MMP and TIMP plasma levels between the groups were tested using the Mann–Whitney test. Bivariate correlations between MMPs and TIMPs at each measurement time, and the outcome measures (6 month 99mTc-SPECT infarct size and severity, and change from baseline to 6 month [Δ] in echocardiographic left ventricular end-diastolic volume index [LVEDVi]) were performed using the Pearson's test, or Spearman's test in case of data that did not comply with normality (Shapiro-Wilk test). The magnitude of infarct size and severity, and Δ LVEDVi stratified on the basis of percentile levels of TIMP-2 plasma on the day 7 were evaluated by the Mann–Whitney test. To analyze the independent contribution of doxycycline on TIMP2 plasma levels on days 1 and 7, a stepwise multivariate analysis was undertaken; for data that was not normally distributed, logarithmic or square root transformations were performed. In addition to the baseline plasma level of TIMP-2 and concurrent plasma values of the main target of TIMP-2 (i.e. the MMP-2 levels on days 1 and 7, respectively), the other variables entered in the multivariated model are those that have been shown to be independently associated with TIMP-2 plasma values in the Global Registry of Acute Coronary Events (GRACE) risk study, nominally the age, baseline values of blood creatinine as expression of renal function, and smoking history [32]. All tests were two-sided, and a p value b 0.05 was considered statistically significant. Analyses were performed with SPSS software, version 19 (IBM Corp, Somers, NY). 3. Results 3.1. Baseline characteristics Baseline blood samples were available in 106 of the 110 (96%) patients enrolled in the TIPTOP trial, 52 patients in the doxycycline group and 54 patients in the control group. There were no significant differences in baseline characteristics of the two study groups (Table 1). The number of patients who were enrolled in the study and included in the analyses during the course of the study is summarized in Appendix A of the Supplementary Data. 3.2. MMP and TIMP profiles according to the doxycycline therapy The MMP plasma profiles measured over time are shown in Fig. 1. All the analyzed MMPs showed significant time-dependent changes (p b 0.001 for MMP-2,-7,-8 and p b 0.05 for MMP-9, respectively) in their circulating levels during the 6 month follow-up period. No significant differences were observed between groups at each evaluated time point. The TIMP profiles measured over time are shown in Fig. 2. As for the MMPs, all the analyzed TIMPs showed significant, timedependent changes (p b 0.001) in their circulating levels during follow-up. With the exception of TIMP-2, which showed significantly higher plasma levels (p b 0.05) at days 1 and 7 in the doxycycline group, no significant differences were observed between groups at each time point of evaluation. 3.3. The independent contribution of doxycycline on TIMP-2 plasma levels A stepwise multivariate analysis was performed to test whether the correlation between doxycycline and TIMP-2 on days 1 and 7 persists even after adjustment for the main factors potentially responsible for

149

Table 1 Baseline and procedural characteristics.

Age (years) Male sex Hypertension Diabetes Dyslipidemia Heart rate (beats/min) Systolic blood pressure (mm Hg) Σ ST-elevation pre PCI (mm) Mean Median Σ ST-elevation post PCI (mm) Mean Median LVEDVi (ml/m2) Mean Median LVESVi (ml/m2) Mean Median LVEF (%) Mean Median Left anterior descending artery Infarct related artery TIMI flow grades 0–1 Multivessel disease Procedural success IRA stenting Number of stents Stent length (mm) Abciximab use Ischemia time (min) Troponin I peak (ng/ml) Aspirin Thienopyridine (clopidogrel or prasugrel) Statin Βeta blocker ACE inhibitor/ARB antagonist Mineralcorticoid antagonists

Doxycycline group n = 52

Control group n = 54

p

70 ± 11 73 29 15 16 78 ± 15 131 ± 31

70 ± 12 65 18 27 10 81 ± 15 128 ± 27

0.98 0.40 0.18 0.10 0.40 0.25 0.54

17 ± 11 14 [9–21]

15 ± 9 13 [8–21]

0.49 0.68

6±5 5 [2–9]

6±4 5 [3–8]

0.98 0.80

49 ± 12 47 [40–56]

49 ± 14 47 [38–61]

0.99 0.91

31 ± 9 30 [24–35]

32 ± 11 31 [23–37]

0.59 0.93

38 ± 7 39 [32–43] 92 71 46 100 100 1.4 ± 0.6 24 ± 11 94 280 ± 180 227 ± 217 94 100 100 85 82 22

36 ± 7 36 [31–42] 87 67 52 100 100 1.4 ± 0.6 22 ± 11 93 240 ± 180 256 ± 185 96 100 100 87 80 27

0.22 0.21 0.53 0.68 0.56 1 1 0.76 0.35 1 0.97 0.47 0.67 1 1 0.78 0.80 0.73

∑ denotes sum, PCI primary percutaneous coronary intervention, LVEDVi left ventricular end-diastolic volume index, LVESVi left ventricular end-systolic volume index, LVEF left ventricular ejection fraction, TIMI thrombolysis in myocardial infarction, IRA infarct related artery, ACE angiotensin-converting-enzyme, ARB angiotensin II receptor blockers. Values are expressed as mean ± SD, or median [IQ range] or numbers (%).

the values of TIMP-2. Only on day 7 were TIMP-2 plasma levels independently related to doxycycline use, together with the concurrent values of plasma MMP-2. Day 1 values of plasma TIMP-2 were independently correlated only with the baseline plasma levels of TIMP-2 (Table 2).

3.4. Relationship between MMPs, TIMPs and 6 month infarct size/severity, and LV remodeling No significant correlations were observed for MMPs and 6 month TcSPECT infarct size/severity, and the change from baseline to 6 month in echocardiographic LVEDVi values (data not shown). On the contrary, significant correlations to 6 month 99mTcSPECT infarct size and severity, and echocardiographic Δ LVEDVi were found for some TIMPs (Table 3). Namely, TIMP-3 on day-30 was directly correlated only with 99mTcSPECT infarct severity, while TIMP-2 on day-7 was inversely correlated with all the outcome measures, i.e. with 99m TcSPECT infarct size and severity, and echocardiographic Δ LVEDVi. Of note, plasma TIMP-2 on day-7 was correlated with 99mTcSPECT infarct size (r = − .37, p = 0.001) and severity (r = .32, p = 0.003), and echocardiographic Δ LVEDVi (r = −.24, p = 0.03) even excluding the effect of the baseline values of LVEF (evaluated as a continuous variable). 99m

150

G. Cerisano et al. / International Journal of Cardiology 197 (2015) 147–153

Fig. 1. Time change in plasma matrix metalloproteinase (MMP). Changes of plasma MMP-2 (top left), MMP-7 (top right), MMP-8 (bottom left) and MMP-9 (bottom right) in patients with (blue line) or without (red line) doxycycline therapy. The number of patients evaluated at each time point in the two study-groups, respectively, is shown at the bottom of the figure.

3.5. Magnitude of infarct size/severity and LV remodeling stratified on TIMP-2 plasma levels The TIMP-2 plasma levels on day 7 were computed as median and then placed in relationship to 6 month measures of infarct size and LV remodeling. Plasma TIMP-2 levels above the median value were associated with smaller myocardial infarct size (3% [from 0% to 16%] vs. 12% [from 0% to 30%], p = 0.002) and severity (0.55 [from 0.44 to 0.64] vs. 0.45 [from 0.29 to 0.60], p = 0.002), and Δ LVEDVi (−1 ml/m2 [from − 7 ml/m2 to 9 ml/m2] vs. 3 ml/m2 [from − 2 ml/m2 to 19 ml/m2], p = 0.04) at 6-month follow-up, respectively (Fig. 3). Interestingly, the worst outcome occurred in patients with TIMP-2 values below the 25th percentile (infarct size: 27% [from 10% to 40%], infarct severity: 0.32 [from 0.24 to 0.49], Δ LVEDVi 9 ml/m2 [from −1 ml/m2 to 25 ml/ m2]). 4. Discussion The unique and significant findings from this study are twofold. First, in patients successfully treated with pPCI for a first STEMI and LV dysfunction, early treatment with doxycycline was independently associated with significantly higher plasma levels of TIMP-2. Second, a significant inverse relationship was observed between plasma levels of TIMP-2 on day 7 after index MI and the extent of infarct size and severity, and LV dilation at 6 month follow-up. 4.1. Early short-term doxycycline therapy in patients with STEMI: the TIPTOP trial The TIPTOP trial recently showed that a timely, short-term therapy with doxycycline, the most potent MMP inhibitor of the tetracycline class of antibiotics, is safe and able to induce a significant decrease in

LV dilation, infarct size and severity in patients successfully treated with primary PCI for a first STEMI and LV dysfunction [24]. Noteworthy, the anti-remodeling effect of doxycycline occurred in the ominous patients with a potentially greater reperfusion injury, such as those with a baseline TIMI flow grade ≤ 1 in the infarct-related artery, but not in those with a baseline TIMI flow grades 2–3 [33]. Although many tissue protective properties have been ascribed to tetracycline class of antibiotics, including anti-oxidant, anti-inflammatory and anti-apoptotic properties [19], to the plasma concentration achievable with the dose of doxycycline used in our study [25], it is plausible that MMP inhibition may be the main mechanism of its cardio-protective action [12,22]. 4.2. Time course of MMPs and TIMPs according to doxycycline therapy In this pre-defined secondary analysis of the TIPTOP trial, the assessment of plasma MMPs and TIMPs evidenced TIMP-2 differences with doxycycline therapy, marked by a significant increase in levels of TIMP-2 already after 1 day of therapy and then on day 7. However, on day 1 (i.e. only after 24 h of doxycycline therapy) plasma levels of TIMP-2 ultimately depend on their baseline values. Instead, on day 7 plasma levels of TIMP-2 still remain independently associated with doxycycline-therapy and its main concurrent substrate (MMP-2), while the baseline value of TIMP-2 is no longer significant. The mechanism contributing to the significant increase of plasma TIMP-2 in doxycycline treatment is unclear and deserves further investigation, however an up-regulation of TIMP-2 in relation to treatment with doxycycline has been observed in a mouse model of Marfan syndrome [34] and in other experimental settings [35,36]. Unlike other studies [37–40], we did not observe significant differences in MMP plasma levels according to doxycycline therapy. This may be due to methodological differences between studies. In any case, since we measured the plasma MMP concentration, we cannot exclude that doxycycline may have modulated post-translation MMP

G. Cerisano et al. / International Journal of Cardiology 197 (2015) 147–153

151

Fig. 2. Time change in plasma tissue inhibitor of metalloproteinase (TIMP). Changes of plasma TIMP-1 (top left), TIMP-2 (top right), TIMP-3 (bottom left) and TIMP-4 (bottom right) in patients with (blue line) or without (red line) doxycycline therapy. The number of patients evaluated at each time point in the two study-groups, respectively, is shown at the bottom of the figure.

activity. In glioma cell lines, for example, doxycycline reduces the activity of MMP-2 without changing its mRNA levels [41].

and infarct size, with a marked elevation in collagenases, resulting in excessive degradation of the ECM [43]. Because of its design, this study was unable to demonstrate any cause–effect relationships, but only correlations which, however,

4.3. Insights into the mechanisms of the beneficial effects of doxycycline therapy Recently, Manhenke et al. [42] observed an independent, inverse relationship between TIMP-2 on day 7 after index MI and 1 year infarct size, as assessed by cardiac magnetic resonance. In agreement with Manhenke et al. [42], in our study plasma levels of TIMP-2 on day 7 after STEMI correlated inversely with myocardial infarct size and severity, and LV dilation. Of note, patients whose plasma TIMP-2 levels increases slightly from day 1 to day 7 after index MI, and remains within the 25th percentile, have the largest infarct size/severity and LV dilation. This is in agreement with the results observed in a mouse MI model where TIMP-2 deficiency leads to larger LV chamber size Table 2 Stepwise multivariate analysis. Variable

Beta

p

95% CI

Contributors of TIMP-2 plasma levels on day 1 Baseline TIMP-2 .60

.000

.30–.53

Contributors of TIMP-2 plasma levels on day 7 MMP-2 day-7 .32 Doxycycline .26

.001 .006

.08–.29 .03–.14

The variables tested in the multivariate analyses, respectively, were age, smoke, baseline values of creatinine, baseline value of TIMP-2, concurrent values of MMP-2, and doxycycline therapy (for details see text in the statistical analysis section). TIMP-2 denotes tissue inhibitor of metalloproteinase-2, MMP-2 matrix metalloproteinase2.

Table 3 Bivariate correlations between TIMPs at each measurement time, and the outcome measures. Timing

TIMPs

Baseline Day-1 Day-7 Day-30 Day-180 Baseline Day-1 Day-7 Day-30 Day-180 Baseline Day-1 Day-7 Day-30 Day-180 Baseline Day-1 Day-7 Day-30 Day-180

TIMP-1

TIMP-2

TIMP-3

TIMP-4

6-month infarct size

6-month infarct severity

Δ LVEDVi

r

p

r

p

r

p

.15 .09 −.03 −.08 −.001 .03 −.09 −.23 −.16 −.14 .01 −.14 −.08 −.21 −.19 −.04 −.13 −.04 −.09 −.20

.15 .37 .79 .49 .99 .75 .41 .03 .14 .20 .90 .20 .47 .06 .10 .73 .23 .72 .42 .08

−.09 −.07 .07 .04 .03 −.05 .14 .26 .16 .13 .05 .19 .09 .24 .21 .04 .14 .06 .09 .22

.39 .49 .51 .69 .77 .65 .18 .02 .15 .26 .66 .08 .44 .03 .06 .71 .21 .60 .44 .05

−.06 −.03 −.12 .10 .004 −.09 .05 −.24 −.09 .008 −.02 −.03 −.05 −.009 −.02 −.01 −.03 .08 .08 −.05

.60 .81 .28 .35 .97 .42 .63 .03 .41 .94 .87 .76 .66 .93 .86 .91 .75 .47 .45 .69

TIMPs denote tissue inhibitor of metalloproteinases; Δ LVEDVi changes from baseline to 6 month in left ventricular end-diastolic volume index.

152

G. Cerisano et al. / International Journal of Cardiology 197 (2015) 147–153

Fig. 3. Relationship between extent of infarct size/severity and LV volumes with TIMP-2. Magnitude of infarct size (left), infarct severity (center) and Δ left end-diastolic volume index (LVEDVi) (right) in patients with plasma TIMP-2 values over (green box) or under (blue box) the 50th percentile. Middle hash of box indicates the median; 25–75th percentiles are represented by end caps of the box; the whiskers indicate the 10 and 90th percentiles. The infarct severity was defined as the lowest ratio of minimal to maximal counts in the short-axis slices evaluated for infarct size. Therefore, the lower the severity the greater the infarct transmurality.

allow for some interesting hypotheses. Among the four members of the TIMP family, TIMP-2 is the only one that can activate as well as inhibit MMPs [44]. TIMP-2 at low concentration forms a ternary complex with membrane bound MMP-14 and pro-MMP-2. A free MMP-14 then cleaves the propeptide of pro-MMP-2, which results in the generation of active MMP-2. However, higher concentrations of TIMP-2 neutralize MMP-14 activity and prevent the activation of pro-MMP-2 [45]. This loss of MMP-2 activity as a function of TIMP-2 supports our results, particularly if we consider the unique role [46] played by the activation of MMP-2 on ischemia/reperfusion injury through the rapid proteolysis of sarcomeric proteins [10–12]. Thus, it is possible to hypothesize that the favorable effect of doxycycline that we observed in the TIPTOP trial [24,33] may be due to the upregulation of TIMP-2 which, in turn, could lead to a favorable modulation of MMP activity during the early stage of a reperfused AMI at both extracellular and intracellular levels. 4.4. Study limitations The main limitations of this study coincide with those of the main study [24] (see Appendix B of the Supplementary data). Of these, the absence of an assessment with cardiac magnetic resonance imaging (MRI) deserves further comment. Without doubt MRI is the gold standard for noninvasive assessment of LV morphology/function, and tissue characterization [47]. This must be taken into account in the planning of new trials. However, since there is a significant correlation between MRI and 2D Echo data [48,49], and we have measured LV morphological and functional changes with the same technique, it is plausible to assume that an analysis with MRI would not have significantly affected the results of our study. On the other hand, the easy availability of 2D Echo makes it ideal for a very early baseline evaluation of more ill patients with AMI, such as those enrolled in our study. Likewise, MRI is more accurate than 99mTc-sestamibi SPECT for the assessment of small infarction and infarct transmurality. However, despite a lower spatial resolution, SPECT is a validated method for measurements of infarct size [50]. More specific limitations were that we analyzed only selected MMPs and therefore may have oversimplified the overall pattern of the complex interplay of MMPs and TIMPs. Furthermore, background therapies of the patients involved in the present study also were shown to affect collagen turnover. However, the therapy was similar in both groups, and, if any, changes related to doxycycline were observed beyond the

effects of background therapy [51]. Finally, we did not measure MMP activity, the various isoforms of MMP-2, or their noted post-translational modifications which affect their activity [52]. Therefore, the results are only for hypothesis generating purposes. In conclusion, in patients treated by primary PCI for a first STEMI with LV dysfunction, timely, short-term therapy with doxycycline can modulate higher plasma levels of TIMP-2 which, in turn, inversely correlate with the 99mTc-SPECT extent of infarct size and severity, and the echocardiographic extent of LV dilation at follow-up. Obviously, the confirmation of TIMP-2 as a key substrate of the favorable effect of doxycycline in this clinical setting calls for ad hoc studies. Study limitation of the main study. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.ijcard.2015.06.024. Conflict of interest The authors report no relationships that could be construed as a conflict of interest. References [1] L. Bolognese, A.N. Neskovic, G. Parodi, et al., Left ventricular remodeling after primary coronary angioplasty: pattern of left ventricular dilation and long-term prognostic implications, Circulation 106 (2002) 2351–2357. [2] J.N. Cohn, R. Ferrari, N. Sharpe, Cardiac remodeling: concepts and clinical implications: a consensus paper from an international forum on cardiac remodeling, J. Am. Coll. Cardiol. 35 (2000) 269–282. [3] B.I. Jugdutt, Ventricular remodeling after myocardial infarction and the extracellular collagen matrix. When is enough enough? Circulation 108 (2003) 1395–1403. [4] S. Matsumura, S. Iwanaga, S. Mochizuki, H. Okamoto, S. Ogawa, Y. Okada, Targeted delection or pharmacological inhibition of MMP-2 prevents cardiac rupture after myocardial infarction in mice, J. Clin. Invest. 115 (2005) 599–609. [5] J.S. Ikonomidis, J.W. Hendrick, A.M. Parkhurst, et al., Accelerated LV remodeling after myocardial infarction in TIMP-1-deficient mice: effects of exogenous MMP inhibition, Am. J. Physiol. Heart Circ. Physiol. 288 (2005) H149–H158. [6] F.G. Spinale, C.P. Escobar, J.F. Hendrick, et al., Chronic matrix metalloproteinase inhibition following myocardial infarction in mice: differential effects on short and long-term survival, J. Pharmacol. Exp. Ther. 318 (2006) 966–973. [7] D. Vanhoutte, M. Schellings, Y. Pinto, S. Heymans, Relevance of matrix metalloproteinases and their inhibitors after myocardial infarction: a temporal and spatial window, Cardiovasc. Res. 69 (2006) 604–613. [8] W.M. Yabrough, R. Mukherjee, G.P. Escobar, et al., Selective targeting and timing of matrix metalloproteinase inhibition in post-myocardial infarction remodeling, Circulation 108 (2003) 1753–1759. [9] C.J. Schulze, W. Wang, W.L. Suarez-Pinzon, J. Sawicka, G. Sawicki, R. Schulz, Imbalance between tissue inhibitor of metalloproteinase-4 and matrix metalloproteinases

G. Cerisano et al. / International Journal of Cardiology 197 (2015) 147–153

[10]

[11]

[12]

[13]

[14] [15]

[16]

[17]

[18]

[19]

[20]

[21]

[22]

[23]

[24]

[25]

[26]

[27]

[28]

[29]

[30] [31]

during acute myocardial ischemia–reperfusion injury, Circulation 107 (2003) 2487–2492. W. Wang, C.J. Schulze, W.L. Suarez-Pinzon, J.R. Dyck, G. Sawicki, R. Schulz, Intracellular action of matrix metalloproteinase-2 accounts for acute myocardial ischemia and reperfusion injury, Circulation 106 (2002) 1543–1549. M.A. Ali, W.J. Cho, B. Hudson, Z. Kassiri, H. Granzier, R. Schulz, Titin is a target of matrix metalloproteinase-2: implications in myocardial ischemia/reperfusion injury, Circulation 122 (2010) 2039–2047. G. Sawicki, H. Leon, J. Sawicka, et al., Degradation of myosin light chain in isolated rat hearts subjected to ischemia–reperfusion injury: a new intracellular target for matrix metalloproteinase-2, Circulation 112 (2005) 544–552. Z. Giricz, M.M. Lalu, C. Csonka, P. Bencsik, R. Schulz, P. Ferdinandy, Hyperlipidemia attenuates the infarct size-limiting effect of ischemic preconditioning: role of matrix metalloproteinase-2 inhibition, J. Pharmacol. Exp. Ther. 316 (2006) 154–161. B. Fingleton, Matrix metalloproteinase inhibitors for cancer therapy: the current situation and future prospects, Expert Opin. Ther. Targets 7 (2003) 385–397. N.A. Rizvi, J.S. Humphrey, E.A. Ness, et al., A phase I study of oral BMS-275291, a novel nonhydroxamate sheddase-sparing matrix metalloproteinase inhibitor, in patients with advanced or metastatic cancer, Clin. Cancer Res. 10 (2004) 1963–1970. M.P. Hudson, P.W. Armstrong, W. Ruzyllo, et al., Effect of selective matrix metalloproteinase inhibitor (PG-116800) to prevent ventricular remodeling after myocardial infarction. Results of the PREMIER (Prevention of Myocardial Infarction Early Remodeling) Trial, J. Am. Coll. Cardiol. 48 (2006) 15–20. L.M. Golub, S. Ciancio, N.S. Ramamamurthy, M. Leung, T.F. McNamara, Low-dose doxycycline therapy: effect on gingival and crevicular fluid collagenase activity in humans, J. Periodontal Res. 25 (1990) 321–330. J.B. Modrak, K.S. Rovang, Estimation of infarct size by determination of myocardial 3H-tetracycline accumulation in the coronary ligated rat, Res. Commun. Chem. Pathol. Pharmacol. 34 (1981) 149–152. M.O. Griffin, E. Fricovsky, G. Ceballos, F. Villarreal, Tetracyclines: a pleitropic family of compounds with promising therapeutic properties. Review of the literature, Am. J. Physiol. Cell Physiol. 299 (2010) C539–C548. F.J. Villarreal, M. Griffin, J. Omens, W. Dillmann, J. Nguyen, J. Covell, Early short-term treatment with doxycycline modulates postinfarction left ventricular remodeling, Circulation 108 (2003) 1487–1492. R.A. Garcia, K.V. Go, F.J. Villarreal, Effects of timed administration of doxycycline or methylprednisolone on post-myocardial infarction remodeling inflammation and left ventricular remodeling in the rat heart, Mol. Cell. Biochem. 300 (2007) 159–169. P.Y. Cheung, G. Sawicki, M. Wozniak, W. Wang, M.W. Radomski, R. Schulz, Matrix metalloproteinase-2 contributes to ischemia/reperfusion injury in the heart, Circulation 101 (2000) 1833–1839. M. Donato, V. D'Annunzio, B. Buchholz, et al., Role of matrix metalloproteinase-2 in the cardioprotective effect of ischaemic postconditioning, Exp. Physiol. 95 (2010) 274–281. G. Cerisano, P. Buonamici, R. Valenti, et al., Early short-term doxycycline therapy in patients with acute myocardial infarction and left ventricular dysfunction to prevent the ominous progression to adverse remodelling: the TIPTOP trial, Eur. Heart J. 35 (2014) 184–191. A.K. Prall, G.M. Longo, W.G. Mayhan, et al., Doxycycline in patients with abdominal aneurysms and in mice: comparison of serum levels and effect on aneurysm growth in mice, J. Vasc. Surg. 35 (2002) 923–929. R. Schröder, R. Dissmann, T. Brüggemann, et al., Extent of early ST segment elevation resolution: a simple but strong predictor of outcome in patients with AMI, J. Am. Coll. Cardiol. 24 (1994) 384–391. N.B. Schiller, P.M. Shah, M. Crawford, et al., Recommendations for quantitation of the left ventricle by two-dimensional echocardiography: American Society of Echocardiography Committee on Standards, Subcommittee on Quantitation of TwoDimensional Echocardiograms, J. Am. Soc. Echocardiogr. 2 (1989) 358–367. R. Sciagrà, A. Imperiale, D. Antoniucci, et al., Relationship of infarct size and severity versus left ventricular ejection fraction and volumes obtained from 99mTc-sestamibi gated single-photon emission computed tomography in patients treated with primary percutaneous coronary intervention, Eur. J. Nucl. Med. Mol. Imaging 31 (2004) 969–974. C.S. Webb, D.D. Bonnema, S.H. Ahmed, et al., Specific temporal profile of matrix metalloproteinase release occurs in patients after myocardial infarction: relation to left ventricular remodeling, Circulation 114 (2006) 1020–1027. F.G. Spinale, F. Villarreal, Targeting matrix metalloproteinases in heart disease: lessons from endogenous inhibitors, Biochem. Pharmacol. 90 (2014) 7–15. D. Inzitari, B. Giusti, P. Nencini, et al., MAGIC Study Group. MMP9 variation after thrombolysis is associated with hemorrhagic transformation in lesion and death, Stroke 44 (2013) 2901–2903.

153

[32] D. Kelly, I.B. Squire, S.Q. Khan, et al., Usefulness of plasma tissue inhibitors of metalloproteinases as markers of prognosis after acute myocardial infarction, Am. J. Cardiol. 106 (2010) 477–482. [33] G. Cerisano, P. Buonamici, R. Valenti, et al., Effects of a timely therapy with doxycycline on the left ventricular remodeling according to the pre-procedural TIMI flow grade in patients with ST-elevation acute myocardial infarction, Basic Res. Cardiol. (2014), http://dx.doi.org/10.1007/s00395-014-0412-2. [34] A.W. Chung, H.H. Yang, M.W. Radomski, C. van Breemen, Long-term doxycycline is more effective than atenolol to prevent thoracic aortic aneurysm in Marfan syndrome through the inhibition of matrix metalloproteinase-2 and − 9, Circ. Res. 102 (2008) 73–85. [35] W. Briest, A. Hölzl, B. Rassler, A. Deten, H.A. Baba, H.G. Zimmer, Significance of matrix metalloproteinases in norepinephrine-induced remodelling of rat hearts, Cardiovasc. Res. 57 (2003) 379–387. [36] M.M. Castro, E. Rizzi, L. Figueiredo-Lopes, J.E. Tanus-Santos, et al., Metalloproteinase inhibition ameliorates hypertension and prevents vascular dysfunction and remodeling in renovascular hypertensive rats, Atherosclerosis 198 (2008) 320–331. [37] D.L. Brown, K.K. Desai, B.A. Vakili, C. Nouneh, H.M. Lee, L.M. Golub, Clinical and biochemical results of the metalloproteinase inhibition with subantimicrobial doses of doxycycline to prevent acute coronary syndromes (MIDAS) pilot trial, Arterioscler. Thromb. Vasc. Biol. 24 (2004) 733–738. [38] B. Axisa, I.M. Loftus, A.R. Naylor, et al., Prospective, randomized, double-blind trial investigating the effect of doxycycline on matrix metalloproteinase expression within atherosclerotic carotid plaques, Stroke 33 (2002) 2858–2864. [39] J.B. Payne, L.M. Golub, J.A. Stoner, et al., The effect of subantimicrobial-dosedoxycycline periodontal therapy on serum biomarkers of systemic inflammation: a randomized, double-masked, placebo-controlled clinical trial, J. Am. Dent. Assoc. 142 (2011) 262–273. [40] C.J. Schulze, M.M. Castro, A.D. Kandasamy, et al., Doxycycline reduces cardiac matrix metalloproteinase-2 activity but does not ameliorate myocardial dysfunction during reperfusion in coronary artery bypass patients undergoing cardiopulmonary bypass, Crit. Care Med. 41 (2013) 2512–2520. [41] A. Wang-Gillam, E. Siegel, D.A. Mayes, L.F. Hutchins, Y.-H. Zhou, Anti-tumor effect of doxycycline on glioblastoma cells, J. Cancer Mol. 3 (2007) 147–153. [42] C. Manhenke, T. Ueland, B.I. Jugdutt, et al., The relationship between markers of extracellular cardiac matrix turnover: infarct healing and left ventricular remodelling following primary PCI in patients with first-time STEMI, Eur. Heart J. 35 (2014) 395–402. [43] V. Kandalam, R. Basu, T. Abraham, et al., TIMP2 deficiency accelerates adverse postmyocardial infarction remodeling because of enhanced MT1-MMP activity despite lack of MMP2 activation, Circ. Res. 106 (2010) 796–808. [44] J.M. Ngu, G. Teng, H.C. Meijndert, Mewhort, et al., Human cardiac fibroblast extracellular matrix remodeling: dual effects of tissue inhibitor of metalloproteinase-2, Cardiovasc. Pathol. (2014 Nov-Dec) 335–343, http://dx.doi.org/10.1016/j.carpath. 2014.06.003 Epub 2014 Jun 21. [45] H. Nagase, R. Visse, G. Murphy, Structure and function of matrix metalloproteinases and TIMPs, Cardiovasc. Res. 69 (2006) 562–573. [46] B.G. Hughes, R. Schulz, Targeting MMP-2 to treat ischemic heart injury, Basic Res. Cardiol. 109 (2014) 424, http://dx.doi.org/10.1007/s00395-014-0424-y. [47] S. Achenbach, M.G. Friedrich, E. Nagel, et al., CV imaging: what was new in 2012? J. Am. Coll. Cardiol. Img. 6 (2013) 714–734. [48] B.I. Gardner, S.E. Bingham, M.R. Allen, D.D. Blatter, J.L. Anderson, Cardiac magnetic resonance versus transthoracic echocardiography for the assessment of cardiac volumes and regional function after myocardial infarction: an intrasubject comparison using simultaneous intrasubject recordings, Cardiovasc. Ultrasound 7 (2009) 38, http://dx.doi.org/10.1186/1476-7120-7-38. [49] N. Mistry, S. Halvorsen, P. Hoffmann, et al., Assessment of left ventricular function with magnetic resonance imaging vs. echocardiography, contrast echocardiography, and single-photon emission computed tomography in patients with recent ST-elevation myocardial infarction, Eur. J. Echocardiogr. 11 (2010) 793–800. [50] R.J. Gibbons, Tc-99mSPECT sestamibi for the measurement of infarct size, J. Cardiovasc. Pharmacol. Ther. 16 (2011) 321–331. [51] B.I. Jugdutt, Pleiotropic effects of cardiac drugs on healing post-MI. The good, bad, and ugly, Heart Fail. Rev. 13 (2008) 439–452. [52] A.L. Jacob-Ferreira, M.Y. Kondo, P.K. Baral, et al., Phosphorylation status of 72 kDa MMP-2 determines its structure and activity in response to peroxynitrite, PLoS One 8 (8) (Aug 27 2013) e71794, http://dx.doi.org/10.1371/journal.pone.0071794.