Introduction

Reperfusion therapy for acute ST-elevation myocardial infarction (STEMI) aims at early restoration of coronary circulation [1, 2]. It reduces infarct size and improves residual left ventricular (LV) function after STEMI. Primary percutaneous coronary intervention (PCI) is the choice of treatment for STEMI patients, especially when in cardiogenic shock [3–5]. In STEMI patients with cardiogenic shock, mechanical left ventricular support aims at improving the circulatory conditions of both the coronary arteries and other organs [6–9]. The circulatory response at microvascular level has been investigated in hemorrhagic and septic shock models, including ischemia–reperfusion injury [10–12]. In patients with heart failure and cardiogenic shock, disturbances of the microcirculation (MC) have also been described. However to date, little is known about the alterations in tissue perfusion and circulatory characteristics at the microvascular level, particularly in the acute phase of STEMI. Additionally, the effect of circulatory support on MC in STEMI patients after PCI has not been studied before [13–16]. Furthermore, the interference of the continuous flow of a circulatory support device with the normal vascular pulsatility is another unsolved issue.

We evaluated the microcirculatory response over time in several conditions with a non-invasive imaging modality called Sidestream dark-field (SDF, Microvision Medical, Amsterdam, The Netherlands) [17–19]. We compared MC between patients treated with and without LV circulatory support by an axial flow pump in STEMI patients after primary PCI. In concordance to the AMC MACH 2 study, we hypothesized that improvement of myocardial function and improved circulation with Impella support would have a positive effect on the peripheral MC and therefore tissue perfusion [20, 21].

Materials and methods

Patients and study design

The study was a sub-study of the non-randomized controlled AMC MACH 2 study and has been approved by the ethics committee of the Academic Medical Center (Amsterdam, the Netherlands) [20, 22]. As described earlier, we hypothesized that mechanical unloading of the left ventricle in STEMI patients treated by PCI may give the myocardium time to recuperate from ischemic stunning and may reduce infarct size. This may be particularly true in STEMI patients with cardiogenic shock or pre-shock, who have a large ischemic area at risk for necrosis. Prior to the initiation of a large trial in this patient category, as prolonged mechanical cardiac support with the novel Impella assist device was never evaluated in the setting of STEMI before, the MACH 2 study was designed to assess safety and feasibility of prolonged Impella support in a group of less severely hemodynamically compromised patients, which still might benefit from left ventricular unloading (i.e. anterior STEMI).

Written informed consent was obtained from each patient prior of inclusion. Patients between 30 and 80 years of age, who presented with a first anterior STEMI within 6 h after onset of symptoms were eligible for enrolment. Exclusion criteria were mainly deep cardiogenic shock, mechanical ventilation, blood transfusion in the previous 24 h, known hemoglobin diseases such as sickle cell or thallasemia, stroke or transient ischemic attack within the previous 4 weeks as well as exclusion criteria related to the ability to insert the Impella device such as: the presence of a mural thrombus in the left ventricle and severe aortic stenosis or the presence of a mechanical aortic valve. The last three consecutive patients with (Impella group n = 10) and the last three without circulatory support (non-support control group n = 10) of the MACH 2 study were selected for the MC sub-study. The Impella group patients had circulatory support for 3 days with the Impella LP2.5 immediately after primary PCI. The control group received routine care after primary PCI, including IABP therapy, as deemed necessary by the attending operator in one patient. Three healthy controls who matched for age and body mass index provided microcirculatory control data.

Impella LP 2.5

As described earlier, the Impella LP2.5 is a novel catheter (9 Fr) mounted micro-axial rotary blood pump (12 Fr), designed for short-term mechanical circulatory support. It is inserted through a femoral approach and positioned across the aortic valve into the left ventricle. Expelling aspirated blood from the left ventricle into the ascending aorta, the Impella is able to provide flow up to 2.5 L/min.

Data collection and echocardiography

Procedural and laboratory data, clinical history and admission medication were prospectively collected. Blood samples were taken for determination of troponin T, creatine kinase, creatine kinase-MB (CK-MB) and NT-proBNP at admission, post-PCI and every 6 h during the first 72 h of admission.

Global left ventricular ejection function (LVEF) in both groups was determined on transthoracic echocardiography (TTE) immediately after PCI (baseline) and at 3 days (after Impella removal). Baseline TTE also served to assess the presence of LV mural thrombus, which is an exclusion criterion for Impella therapy. A standard imaging protocol was used based on apical 4- and 2-chamber views. Analysis were performed by an independent core lab (DCRI echocardiography core laboratory, Duke University, Durham, NC, USA) blinded to treatment assignment.

Microcirculatory measurements and analysis

Assessment of the sublingual MC was performed using SDF, a hand-held microscope with light-emitting diodes (LEDs) arranged in a ring formation at the tip of the probe. These LEDs emit at a wavelength 550 ± 10 nm which is absorbed by hemoglobin in erythrocytes, which are visualized as dark flowing cells in the microvascular network consisting of small (0–20 μm), medium (21–50 μm) and large (51 μm and larger) vessels. In the Impella group baseline images of the MC are recorded after PCI but before Impella support. Thereafter, microcirculatory recordings are obtained immediately after start of support, at 24, 48 h of support and after device removal at 72 h. In the non-support control group microcirculatory assessment is performed at corresponding time points. The SDF device is gently placed on the sublingual region of the patients to avoid pressure artifacts. Subsequently, three to five sequences of steady images of various regions were obtained and stored as avi formatted video clips.

Analysis of the microvascular network is carried out by a validated scoring system described by De Backer et al. [23]. For differentiation of the quality or type of flow in the microvascular network the microvascular flow index (MFI) is determined. This score is based on quantification of the predominant type of flow per vessel type (small, medium, large) in four equal quadrants of the obtained image. Flow is characterized as absent (0), intermittent (1), sluggish (2), or normal (3). The MFI score represents the averaged values of the four quadrants [24].

In addition, the perfused vessel density (PVD), which is an estimate of the functional capillary density, is calculated. The PVD is calculated by the percentage of the vessel length of small vessels of all the vessels divided by the analyzed image area. Tissue perfusion or oxygen transport capacity by the MC is dependent on both functional capillary density (reflected by PVD) and blood velocity (reflected by MFI).

Statistical analysis

Data are presented as mean ± SD for continuous variables and as frequencies for categorical variables. Differences between the Impella group and control group in continuous variables were tested with the Student’s t test. Comparisons between baseline hemodynamic values, baseline LVEF, MFI, PVD, and subsequent measurements were performed in a paired fashion using the Student’s t test. All tests were two tailed and a P value of <0.05 was considered statistically significant. The statistical Package for the Social Sciences (SPSS Inc., Chicago, IL, USA; version 15.0) was used for statistical analysis.

Results

Patients

In total six consecutive patients from the MACH2 study were included in this study. Microcirculatory observations of three age and body mass index matched healthy controls were used as normalized standard. All six anterior STEMI patients were treated with primary PCI. After PCI, three patients received Impella support (Impella group) and three patients received routine care (non-support group). The baseline clinical characteristics of the study population are shown in Table 1. Patients of the Impella group had a higher risk profile. At admission the Impella group had higher serum levels of NT-proBNP, CK, CK-MB, peak CK-MB, troponin T, as well as a higher area under the curve of CK-MB and troponin T. In addition, baseline echocardiographic LVEF measurement, revealed a more depressed mean ejection fraction in the Impella group compared to the non-support group (i.e. 28 vs. 45%; P = 0.06). In overall, the Impella assisted patients were in worse clinical condition at baseline compared with the control group.

Table 1 Baseline characteristics
Full size table

Sublingual microvascular hemodynamics and LVEF

Adequate recordings were obtained at all times for all patients. Figure 1a shows an example of the MC at baseline in a patient from the Impella group. Figure 1b illustrates the MC of the same patient 48 h later during Impella support. In all patients visualization of the MC showed a network of small, medium and large vessels. The MC was assessed for MFI, representing the amount or quality of flow in the microvascular bed, and the PVD, estimating the functional capillary density. The MFI for both groups is depicted in Fig. 2. In the Impella group, the MFI is significantly more impaired at baseline compared to the non-support control group; the MFI scores respectively were 1.8 ± 0.24 and 2.5 ± 0.11 (P < 0.05). In other words, the flow of small capillaries is more stagnant in the Impella group at baseline. In the non-support control group no significant changes in MFI are observed during the observation period and the MFI remains below values of healthy controls. In contrast, in the Impella group an immediate improvement of the MFI is observed after initiation Impella support. The MFI increases from 1.8 ± 0.24 at baseline to 1.97 ± 0.27 (P = NS) directly after start of the Impella device, but does not become significant until after 24 h of assisted Impella support. This significant change in MFI from 1.8 ± 0.24 at baseline to 2.5 ± 0.19 (P < 0.05) at 24 h of support was sustained even after removal of the Impella at 72 h and remained significantly better than the MFI of non-support controls. With Impella support the quality of flow in the microvascular network did reach the values observed in healthy controls. In contrast to the MFI, the PVD was similar in both groups at baseline (Fig. 3). However whereas the PVD remained equal in the non-support control group, the PVD after 48 h was significantly enhanced in the Impella group, even above values of healthy controls. We found a significant increase of the PVD from 1.6 ± 0.1 to 1.9 ± 0.2 mm/mm2 (P < 0.05) in the Impella group after 48 h of assisted circulation, in comparison to the controls where the PVD remained respectively 1.6 ± 0.2 and 1.6 ± 0.3 mm/mm2 during the same time course. The mean PVD value of the healthy controls was 1.5 ± 0.2 mm/mm2.

Fig. 1
figure 1

a Microvascular network of a patient from the Impella group at baseline (i.e. after PCI, before initiation of support). b MC of a patient from the Impella group after 48 h of assisted circulation

Full size image
Fig. 2
figure 2

Microvascular flow index scores for individual patients (a) and for the Impella group and non-support group (b). Time points depicted are baseline, immediate after start of Impella support, at 24, 48 h of support and at 72 h after removal of the Impella device. Asterisk in b denotes the mean MFI observed in healthy controls. Dagger significant difference (P < 0.05) between Impella group and control group. Section symbol significant difference (P < 0.05) of value compared to baseline within Impella group

Full size image
Fig. 3
figure 3

Perfused vessel densities (estimate of functional capillary density) for individual patients (a) and for the Impella group and non-support group (b). Asterisk in b denotes the mean PVD observed in healthy controls. Dagger significant difference (P < 0.05) between Impella group and control group. Section symbol Significant difference (P < 0.05) of value compared to baseline within Impella group

Full size image

Hemodynamic and echocardiographic data

In the Impella group, LVEF improved from 28 ± 3% at baseline immediately after PCI to 38 ± 3% (P < 0.05) after 3 days (after Impella removal), whereas in the control group no significant improvement was observed. LVEF changed from 45 ± 4 to 37 ± 7% (NS). Thus, the improvement in MC in the Impella group paralleled the functional improvement of LV function.

Discussion

This is the first study describing a positive relationship between effective circulatory support and improvement in sublingual MC, paralleling LV function improvement after STEMI. Our data show that sublingual microcirculatory flow is closely related to LV function improvement. It would support the hypothesis that LV unloading and circulatory support may improve LV function. This phenomenon was demonstrated in STEMI patients with percutaneous LVAD support after PCI [20]. Perhaps SDF imaging of the sublingual microvascular bed may in a way mirror the condition of the myocardium among other parameters and it monitored the effect of adequate circulation and tissue perfusion.

An intact and adequate functioning microvascular network is essential for efficient tissue oxygenation, and consequently, normal end-organ functioning. In our case this also may have led to improved myocardial oxygenation and LVEF. As expected, the worse clinical status of the Impella group was reflected by a more impaired microcirculatory system. The observed microcirculatory dysfunction in STEMI patients may be due to reduced microcirculatory flow, despite seemingly sufficient circulation, especially in the non-support group. Normal MC depending on both the functional capillary density (reflected by PVD) and flow velocity/quality (reflected by MFI) as observed in healthy controls was only achieved in STEMI patients when treated with Impella LP2.5 support. The functional capillary density in both groups was at least at the level of healthy controls, though in the Impella group even above values of healthy controls. The latter may reflect recruitment of microcirculatory vessels in the Impella group. Moreover, the quality of microcirculatory flow (MFI) only in the Impella group reached values of healthy controls.

In other words, our STEMI patients receiving Impella LP2.5 support experienced microcirculatory benefit. In contrast, STEMI patients without support seemed hemodynamically more stable and had better microcirculatory conditions initially. However, in these patients MC did not improve in the first few days, which may reflect sustained microcirculatory distress. Their MC recovered over time but never reached the normal level of healthy persons during the observation period of 72 h. We may assume that despite steady recovery, STEMI patients without circulatory support and impaired MC are exposed to an increased risk of organ failure due to a priori suboptimal tissue perfusion. It has been reported that microvascular deterioration in septic patients, mainly expressed as decreased vascular density, is associated with development of multi-organ failure (MOF) [25, 26]. Besides microvascular hypodensity, several mechanisms such as endothelial cell dysfunction, altered vasomotor tone, reduction of red blood cell deformability and leukocytes adhesions to endothelial walls have a contributing role in progressive organ dysfunction [27, 28]. Furthermore, microcirculatory distress appears to be an independent predictive factor for the clinical outcome of patients in septic shock, and it has been reported that it is of crucial importance that re-establishment of the MC occurs within 24 h [28].

Remarkably, our study showed that restoration of the microvascular system was already observed after 24 h of circulatory support in hemodynamically compromised patients, simultaneous to the correction of systemic hemodynamics through the Impella LP2.5 assist device. Moreover, not only did an increase in the microcirculatory flow occur, but also this effect remained up to 72 h of support and was even persistent after removal of the Impella LP2.5. Another point of interest was the observation that the appearance and rating of the MC among normal healthy persons showed a very acceptable uniformity. Therefore, the sublingually obtained MC observations in healthy persons could be used as the normalized standard indeed.

Limitations

Our study has several limitations. First of all, there are limitations to the method used to assess MC. One may question whether the sublingual area is the representative site to monitor MC in general or in specific organs such as myocardial and renal tissue. It has been shown that sublingual MC alterations in patients with severe heart failure and sepsis were adequately related to outcome [28, 29]. Our study substantiates this observation in a different population. Second, acquisition and analysis of the SDF images are operator dependent. Despite a validating scoring system, the selection and quality of the images determine the quality of outcome. An important point of caution with the SDF imaging is the pressure-induced microvascular changes by application of the SDF microscope on tissue surfaces. Also the number of patients studied is small. In the recently initiated IMPRESS in STEMI trial comparing IABP versus Impella LP2.5 in pre-shock STEMI patients, MC assessment by means of SDF and its relationship with LV recovery will be one of the secondary endpoints.

Conclusion

Microcirculation assessed by SDF improved in STEMI patients treated with the Impella LP2.5 to levels observed in healthy persons and remained suboptimal after 72 h in patients without support. Sublingual SDF to assess MC may serve as a monitor of effective myocardial recovery after PCI and optimization of organ perfusion.