Animals

The measurements were performed in accordance with the (blinded) legislation governing animal studies following the “Guide for the Care and Use of Laboratory Animals” (National Research Council, 8th edition, 2011) and the “Directive 2010/63/EU on the Protection of Animals Used for Scientific Purposes” (EU Official Journal, 2010). Official permission was granted from the governmental animal care and use office (blinded).

Measurements were made in four female German Landrace swine (Gerd Heinrichs, Heinsberg-Karken, Deutschland; mean weight of 55.1kg ± 6.5kg, mean age of 5.3 months ± 0.5 month). The animals were housed under controlled environmental conditions (20°C ± 1°C, 12:12h light/dark cycle). Before starting the experiments, there was an acclimatization period of 2 weeks. Animals were kept in groups of 2 to 4 and received food and water ad libitum.

The animals received premedication by intramuscular injection of atropine (Atropinsulfat, B. Braun Melsungen AG, Melsungen, Germany), azaperone (Stresnil 40mg ad. us. vet.; Sanochemia Pharmazeutika AG, Neufeld, Austria) and ketamine (10% Ketavet ad. us. vet., Zoetis Deutschland GmbH, Berlin, Germany) [7]. Intubation was followed by mechanically ventilation with an oxygen-air mixture. Anaesthesia was maintained with continuous intravenous infusion with propofol (Propofol 2% MCT Fresenius; Fresenius Kabi Deutschland GmbH, Neuss, Germany). For analgesia fentanyl (Fentanyl-Janssen 0,5mg, Janssen-Cilag GmbH, Neuss, Germany) was continuously administered. Vital functions were monitored during the entire experiment. All animals were treated with ASS (Aspirin 500 mg) and heparin (3000 IU, Heparin-Natrium-5000; ratiopharm GmbH, Ulm, Germany) intravenously to prevent blood coagulation during interventions.

The narcotized animals were punctured at the femoral artery followed by insertion of an 8F sheath for further endovascular procedures. The experiments were acute trials, animals were euthanized by intravenous injection of 0.5- 1ml/kg body weight natrium-pentobarbital (Narcoren 16g/100ml; Merial GmbH, Hallbergmoos, Germany).

Blood pressure measurement

First, we assessed the vasculature of the upper limbs via DSA. We then injected clots in the branches of the upper limb arteries following published procedures and we measured the diameters of the occluded vessels [8, 9]. We then measured the intra-arterial blood pressure 5–10 mm proximal and distal to the clot with a Trevo microcatheter (Trevo Pro 18, Stryker, Kalamazoo, CA, USA) using a pressure transducer (MEMSCAP SP844), an AdInstruments Powerlab 16/35 workstation, and LabChart 8 Software (AdInstruments, Dunedin, New Zealand). The pressure transducer was connected to the microcatheter via a standard 3-way valve (Discofix C; Braun, Melsungen, Germany) and tubing (Original-Perfusor line 150 cm; Braun). Continuous flush through the catheter was interrupted during measurements via the 3-way valve. First, we assessed the pressure immediately before the clot. Then, we passed the clot with the microcatheter (n = 46) or the microwire (n = 2) first and confirmed the correct position of the microcatheter with a superselective angiography series [6]. We then measured the pressure behind the clot. We measured the blood pressure until a stable value of blood pressure was reached, which took approximately 5 to 30 seconds distal to the clot. Afterwards, the microcatheter was pulled back through the clot to its proximal end and the blood pressure was measured again following the same procedure. Overall, we measured pressures in 48 cases.

Data and statistical analysis

Blood pressure was recorded continuously with a sampling rate of 1000 Hz. For data analysis we used the MATLAB R2017a software (The Mathworks, Natick, USA). We applied a high pass filter (Butterworth, 3^rd order, cut-off frequency (f_C) = 1 Hz) for suppressing signal wander induced by e.g. breathing and a third order low pass filter (Butterworth, 3^rd order, f_C = 40 Hz) for filtering high-frequency noise and movement artefacts.

Blood pressure was defined as the average of all values acquired over a period of 5 seconds. The arterial pulse pressure was defined as the magnitude of the difference between the systolic maximum and the diastolic minimum blood pressure averaged over a period of 5 seconds. Whenever the typical arterial pulse curve with its systolic maximum and diastolic minimum was not detectable, this was noted as a loss of the pulse curve (Fig 1). In accordance with Sorimachi et al. we defined pressure loss as falling below the cut-off value at 30 mmHg [4, 5]. As an alternative definition of pressure loss, we postulated that there was pressure loss whenever the arterial blood pressure behind the clot was below two standard deviations of the blood pressure proximal to the clot. As our stroke model does not allow quantifying the collateralization of vessels, we hypothesized that large arteries are collateralized better than small arteries and calculated the relationship between vessel size and pressure.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Arterial pulse wave proximal and distal to the clot.

Upper row (a and b): case with maintained arterial pulse wave. Pressure curves before (a) and behind (b) the clot. Mean arterial pressure and pulse pressure before (a) and behind (b) the clot are 100 mmHg and 86 mmHg and 4.4 mmHg and 0.5 mmHg, respectively. Lower row (c and d): case with loss of arterial pulse wave. Pressure curves before (c) and behind (d) the clot. Mean arterial pressure and pulse pressure before (c) and behind (d) the clot are 98 mmHg and 10 mmHg and 7.3 mmHg and 0 mmHg, respectively.

https://doi.org/10.1371/journal.pone.0210572.g001

Continuous parametric variables are presented as mean ± standard deviation, non-parametric variables as median with interquartile range. After testing for normal data distribution with the Shapiro-Wilk test, Mann-Whitney U tests were used to determine whether there is a difference in either blood pressure or arterial pulse wave proximal and distal to the clot, respectively. We also examined the relationship between vessel diameter and loss of blood pressure (pressure loss and pulse pressure loss) by cross-tabulation using Pearson’s χ2 tests and Pearson’s correlation coefficient. Values with an α level < .05 were considered as statistically significant. All statistical analyses were calculated in SPSS V.23 software (IBM, Armonk, New York, USA).

Limitations

The porcine stroke model in the upper limbs does not allow the quantification of collateral status and is therefore one of the major limitations of our study. Nonetheless, our results show that pressure pulse measurements are feasible and may serve as a foundation for further studies. A further limitation is the fact that we conducted our experiments with only one microcatheter, as mechanical properties such as the rigidity of the catheter may have an impact on our results [18]. Another difficulty was the choice of a suitable high and low pass filter. For baseline wander removal we used a Butterworth high pass filter which is both computationally fast and comparably accurate to more complex filter models [19]. Due to the simple filter design it will be easy to expand this method into real time analysis applications. We obtained the best results in terms of signal intensity and suppression of baseline wander at a cut off frequency of 1 Hz. Nevertheless, it was not possible to remove interfering signals completely (Fig 1). We also selected a suitable cut-off frequency for the low-pass filter by trial and error to suppress noise without removing the desired signal. Frequencies of 40 Hz or higher were removed by the filter.