Eight German Landrace pigs weighing 45–60 kg were used for the experiments. The pigs were housed at least 1 week before the experiments at the Department of Laboratory Animal Sciences of the Charité and had free access to standard chow and water. The study was conducted in accordance with the German legislation on the protection of animals and was approved by the local authorities (reference number G 0150/10, Landesamt für Gesundheit und Soziales, Berlin, Germany).
The following 5-mm laparoscopic sealing and cutting devices were compared with respect to their safety and efficacy: LigaSure® V (Valleylab Inc., Boulder, CO, USA), Harmonic ACE® (Ethicon Endo-Surgery, Cincinnati, OH, USA) and THUNDERBEAT® (Olympus Medical Systems Corp., Tokyo, Japan). For the LigaSure® (abbreviated as LS from hereon) and the Harmonic ACE® (HA), commercially available instruments were used according to the manufacturers’ instructions. For THUNDERBEAT® (TB), a prototype from Olympus Medical was used. The TB device integrates two energy modalities since it delivers both ultrasonically generated frictional heat energy and electrically generated bipolar heat energy simultaneously if used in the “seal-and-cut” mode. The design of the jaws is depicted in Fig. 1. In principle, bipolar heat energy is applied laterally and additional sealing and cutting is achieved by ultrasonic energy centrally (at the region of the white Teflon band, see Figs. 1 and 2). Additionally, a “seal” mode can be activated, leading to delivery of only bipolar energy. However, this mode was not evaluated in the present experiments since tissue division is not possible using this mode. Isolated use of the ultrasonic mode is not possible in this device. For practical reasons, two different devices were used on each animal. The sequence of devices and the application of the sealing instruments on each animal were randomized.
The following a priori hypotheses were investigated in the present experiments (sample size calculations were performed prior to the study based on the results of previous preliminary experiments):
The burst pressure of the TB in large arteries (≥5 mm) is superior to that of the HA.
The dissection speed of the TB in isolated vessels and compound tissue is superior to that of the LS.
The heat production of the TB is clinically comparable to that of the HA.
The primary end point was the mean burst pressure as the parameter of efficacy. Based on the preliminary data of a mean burst pressure of 750 mmHg (TB) and 500 mmHg (HA), a total of 43 seals per device were required to show a statistical significance on a 0.05 level with a power of 0.9. For differences in the speed of dissection of isolated arteries using a single activation, a sample size of 14 was required to detect a difference of 4.0 and 7.5 s. For repeated activation during dissection of compound tissue, a sample size of 8 was required to detect a difference of 20 and 30 s on the same power level. Heat production was investigated in a noninferiority design using a clinically relevant range of 150–250 °C, which is well within the range for different ultrasonic devices reported in the literature [8, 10]. Additional secondary end points such as histological width of the tissue seal were analyzed.
The cutting speed was measured during straight dissection of a defined length of the small bowel mesentery (10 cm). Thus, the cut mode of the THUNDERBEAT® was compared with the “max” mode (level 5, Table 1) of the Harmonic ACE®. The time until dissection and eventual seal failures were recorded. This procedure was repeated twice per animal and device, resulting in eight measurements per device. In addition, the time for sealing and dissection of the arteries was recorded. The carotid, lienal, femoral, iliac, popliteal, mesenteric, renal, axillar, and brachial arteries were used for determination of the cutting speed followed by burst pressure measurement or histological analyses (see below). For dissection of isolated arteries, the cut mode of the TB was compared with the “min” mode (level 3, Table 1) of the HA and the standard mode of the LS. Time was measured with standard digital stopwatches. Each process was measured simultaneously by two persons and the mean of both measurements was recorded. For all devices, measurement of time was started when the instrument was in place and ended with complete division of the respective tissue. For the bipolar clamp, this means the cumulative time of the sealing and the cutting process.
Burst pressure measurement
Burst pressure of the sealed vessel segments was measured ex situ. In brief, a catheter was introduced into the open end of the vessel segment and secured. Normal saline was infused into the arterial lumen at a fixed rate (Lambda VIT-FIT, LAMBDA Laboratory Instruments, Zurich, Switzerland), and the pressure was recorded by means of a pressure transducer (Greißinger Electronic GMH3150, Regenstauf, Germany). The maximum pressure (in mmHg) before leakage at the sealing site was defined as burst pressure. In case of leakage from a different site, the vessel was excluded from further analysis. All burst pressure measurements were performed by two persons, blinded for the respective study groups.
Thermal profiles during and after activation of the scissors were analyzed in detail using two different measuring methods. First, they were measured indirectly using an infrared camera (Variocam T, Jenoptik, Jena, Germany). To avoid reflections and disturbance of the measurement at the metallic parts, all metallic parts on the outer side of the instruments were blackened. Heat production was determined during a single activation of the different devices with dissection of mesentery. All instruments were activated until final tissue division. The devices were fixed at the shaft to avoid movement during measurement. The measured safety parameters were the maximum temperature at the outer side of the jaws and the time to decline to 60 °C after activation.
In addition, for confirmation of the temperatures obtained by indirect measurement (infrared camera), the temperature was directly measured using a thermosensor (K type thermocouple, Qilian Power Equipment, China). However, for technical reasons, the temperature was measured only inside the jaws by grasping the thermosensor with the jaws after cutting the 10 cm of mesentery. This measurement reflects the maximum inside temperature after a longer period of continuous/repeated activation. It might not be directly comparable with the maximum outside temperature, which is supposed to be slightly lower. Again, the maximum temperatures and the time to decline to 60 °C were measured.
First, a midline abdominal incision was performed from the xiphoid process to the symphysis. All experiments started with the measurement of the cutting speed in the mesentery in combination with the thermosensor measurement after cutting 10 cm of mesentery. Afterward, the small bowel was placed in the abdominal cavity and covered with moistened gauze to avoid drying. Isolation of various arteries for sealing and BP measurement was started peripherally. First, separate bilateral incisions were made medially at both hind limbs for dissection of the femoral and popliteal arteries. Next, the front limbs were used for preparation of the axillary and brachial arteries using separate bilateral incisions. This was followed by preparation and sealing of abdominal vessels. Finally, the carotid arteries were prepared bilaterally via a longitudinal median cervical incision. Arterial branches that potentially interfered with the BP measurements were ligated. Before sealing and division of the vessel, the external diameter was determined. The cutting devices were randomized and stratified for the diameter category as described above. Traction on the arteries was avoided during activation of the instruments. To ensure a maximum comparability of the different instruments, only one single seal and cut was used for these experiments. No additional sealing steps next to the cutting site at both ends were performed, although this may be done in clinical practice to increase the width and safety of the seal. In case of primary seal failure, a burst pressure of 0 mmHg was recorded for further analysis of the data. After finishing the arterial seals, thermal camera measurements were performed. The same standardized surgical workflow was used on all animals. All animals were euthanized after completion of all experimental procedures.
For histological evaluation, a total of 60 arterial seals in both vessel categories (20 per device) were collected. These specimens were not used for BP measurement and were immediately fixed in formalin. The samples were embedded in paraffin and serially cut in 5-μm sections. Thus, the cutting plane was placed rectangular to the seal for measurement of the sealing width. Staining with hematoxylin and eosin was performed using routine laboratory methods. The extent of adventitial collagen denaturation proximal to the tissue seal and the presence of gas formation caused by tissue boiling were evaluated qualitatively. The perpendicular width of tissue seal was measured in millimeters, beginning from the cut end of the vessel to the point where the vessel walls separated from each other.
For evaluation of the lateral thermal damage and potential damage to adjacent organs, the mesentery of the small bowel was dissected at 0.5 cm from the small-bowel wall. A non-heat-conducting spacer was used for the standardization of the distance. The tissue was fixed, embedded in paraffin, serially cut, and stained with hematoxylin and eosin. Histomorphological analysis investigated the presence of thermal damage to the small bowel. Thus, the relative number of samples with thermal damage and the histological depth of tissue necrosis were evaluated. All histological analyses were performed blinded for the respective study groups.
All values are given as mean and standard error of mean (SEM). For comparisons of continuous variables between groups, a one-way ANOVA was used followed by a Bonferroni post testing. For categorical variables, the χ2 test was used. Differences were considered significant if p was less than 0.05. All statistical analyses were performed using PASW 18.0 (SPSS, Inc., Chicago, IL, USA).