Patient population and study protocol
We studied a total of 21 patients with history of CAD and multiple prior interventions (PCI, CABG, or both) who suffered from disabling angina despite individually optimized medical therapy (Table 1).
An N-13 NH3 PET scan at rest and under adenosine was performed for clinical reasons to reevaluate a repeat mechanical revascularization. If reintervention had been found inappropriate, the viable myocardial wall with the leading perfusion abnormality was identified for CSWT. The baseline PET scan was performed 21 ± 36 days before onset of therapy. A repeat PET scan was performed 43 ± 27 days after the last CSWT session for clinical reasons, to verify the therapeutic effect on regional perfusion and to decide upon continuation of therapy.
Canadian Cardiovascular Society (CCS) scale was assessed in all patients before and after therapy. Additionally, echocardiography and, if feasible, ergometry were performed.
Patients gave their written informed consent to CSWT and the clinical PET studies. The data analysis was approved by the local ethics committee of the Ruhr-University Bochum (Reg. No. 60/2013).
Image acquisition and stress testing
Patients were investigated with an ECAT-951 R scanner (CTI/Siemens Medical systems) and, after a change of the device, with a Biograph mCT (Siemens, Erlangen, Germany). Directly before the PET acquisition, a transmission scan was performed with an ECAT-951 R scanner and a low-dose CT with the Biograph mCT for attenuation correction.
Adenosine was infused intravenously at a constant rate of 0.14 mg/kg/min over 6 min. During the stress phase, heart rate and blood pressure were recorded every 2 min, starting with the onset of the adenosine infusion until completion of after 6 min. Mean arterial blood pressure was calculated from the average values of all four time points and minimal coronary resistance as mean arterial blood pressure/global perfusion.
Two minutes after the onset of the adenosine infusion, on average 600 MBq N-13 NH3 were injected as an intravenous bolus. Image acquisition over 15 min was started simultaneously with the bolus injection. A consecutive set of 20 frames was reconstructed for quantification of perfusion.
Quantitative perfusion analysis
Quantification of the N-13 NH3 scans was based on an irreversible two-compartment model . Corrections for fractional blood volume, partial volume effects and spillover activity from left-ventricular blood pool to tissue were calculated as described elsewhere . A validation of the model in humans had been performed by the argon inert gas technique beforehand . The quantification procedure delivered 20-segment parametric polar maps of MBF. Segments with a fractional blood volume >0.5 were excluded from further analysis. Such large values are only explainable by spillover from the ventricle and mostly referred to the basal segments of the septum. Furthermore, segments with a resting MBF < 50 mL/min/100 g were regarded as scarred and also excluded from the analysis. Global perfusion was calculated as the average of all myocardial segments .
The CSWT target zone was assigned to the anteroseptal wall in 8 cases, to the inferolateral wall in 6, to the inferior wall in 6, and to the apex in 1 case. The reference wall was defined as the remote wall opposite to the CSWT target zone. In the case of the apical CSWT target, the anterior wall served as the reference. For quantitative analysis, perfusion values of the segments of the parametric polarmaps representing either the target or the remote wall were averaged.
Shock wave therapy device, procedure and protocol
A detailed description with illustration of the device and the CSWT protocol has been given elsewhere . In brief, shockwaves were applied with a commercially available shock wave generator (Cardiospec, Medispec, Gaitersburg, USA) under echocardiographic guidance. The energy flux density was adjustable between 0.03 and 0.2 mJ/mm2, with a focus size of 8–9 mm and a length of 25 mm to ensure transmural coverage. The focus could be set up to a depth between 4 and 15 cm. During the procedure, patients lay in supine position. ECG, blood pressure and vital signs were monitored continuously.
The target region previously defined with the PET scan was adjusted with the ultrasound probe. Shock wave release occurred with an R-wave ECG trigger.
The CSWT scheme was performed over 3 months (93 ± 34 days) in blocks of 3 sessions applied every other day during the treatment weeks, with 6 weeks intervals (44 ± 25 and 42 ± 15 days, respectively). The target region previously defined with PET was divided into a basal, midcavity, and apical zone. During each treatment, these zones were targeted with the ultrasound probe, progressing from the basal zone in week 1 (sessions 1–3) to the midcavity zone in week 2 (sessions 4–6), and the apical zone in week 3 (sessions 7–9). Patients received 300 shocks (when starting) to 500 shocks (after verification of tolerance) per treatment session. Thus, during the 9 shockwave sessions, a total of 2700–4500 individual discharges were delivered. Troponin I was controlled several hours after each application.
Resting two-dimensional echocardiography was performed at baseline and follow-up with a 1.7/3.3 MHz multifrequency probe in harmonic imaging mode using commercially available echo equipment (Vivid E9, General Electric, Horten, Norway). The classical two apical views (2- and 4-chamber views) were applied to measure biplane LVEF and both the endsystolic and enddiastolic volumes.
Exercise stress testing
The bicycle test was performed in the upright position (Ergometer 900 ERG, GE Medical Systems). Depending on the patient’s condition, ergometry was started at 25 W or 50 W, and was increased by 25 W each minute. The test was stopped when typical chest pain, significant ST depressions, or exhaustion of the leg muscles occurred.
For the sample size estimation, a power of 90 % and a significance criterion of 0.05 were chosen. The minimum expected difference between the two means and the standard deviation were estimated to 10 mL/min/100 g each. Accordingly, about 20 patients had to be enrolled .
Data are given as mean value ± standard deviation. In the first step, the paired parameters were tested for normal distribution with the Kolmogorov–Smirnov test. As all parameters were normally distributed, post hoc comparisons were performed with a paired t test.
The contingency between change in CCS class (unchanged or better) and change in perfusion (increase <5 % and increase ≥5) was examined using Fisher’s exact test.
Differences were considered statistically significant at values <0.05 (two-sided). For the analyses, the statistical software package IBM SPSS (version 20) was used.