A long-term porcine model for measurement of

A long-term porcine model for measurement of gastrointestinal motility Joerg Schnoor1, Stefanie Bartz1 , Bernd Klosterhalfen2, Werner Kuepper3, Rolf Rossaint 1 & Juliane K. Unger1 1

Department of Anaesthesiology, University Hospital, RWTH Aachen, Germany, 2 Department of Pathology, University Hospital, RWTH Aachen, Germany and 3 Central Animal Experimentation Facility, Department of Veterinary Medicine, University Hospital, RWTH Aachen, Germany

Summary Animal models have become an essential tool in the investigat ions of gut motilit y under experimental conditions. To determine the in¯uence of various anaesthetic drugs on the motility pattern of the gastroduoden al tract, a new long-term model has had to be developed for allowing measurements in conscious and unrestrained as well as in sedated and analgosedat ed pigs. Since mechanical ventilation in¯uences gut motility, it was necessary that this anim al model enabled the investigation of the effect of drugs causing sedation and analgosedat ion during spontaneous breathing. Seven male, castrated pigs, German landrac e, 32±40 kg bodyweight (BW) were investigat ed in this study. After habituati on of the pigs to local housing conditions over 5 days, the animals were trained over 4 days to prepare for experimental situations and investigators. Pigs were inserted with a central venous catheter and with percutaneous enterogastrostomy (PEG ) under general anaesthesia. Intestinal motility was measured by intralum inal im pedancom etry. T he catheter was introduced over the PEG into the stom ach and positioned into the duodenum by duodenoscopy. Measurements were done in conscious, unrestrained pigs and with sedated, and analgosedated anim als on subsequent days. T he habituati on and training of the pigs to the investigators and for the laborat ory conditions took between 7 and 9 days. T he initial anaesthesia protocol for the instrumentation using remifentanil =propofol led to pyloric spasm and was thus unsuitabl e for duodenal intubation with an endoscope. In contrast, a combination of ketam ine =propofol enabled this procedure. It was practicable to measure gut motility in conscious, unrestrained pigs. Spontaneous breat hing was suf®cient under propofol sedat ion and analgosedation using fentanyl-propofol. Systematic ally local application of polividon iodine in the area of the subcutaneous catheters avoided the necessity of using systemic prophylactic antibiot ics. In conclusion, the habituation and training for 9 days enabled the measurement of gut motility by intraluminal impedancometry in conscious pigs. T he insertion of the catheter was done during general anaesthesia using a combination of propofol and ketamine. For the future determination of gut motility performed under general anaesthesia, each sedation and analgosedat ion concept has to be evaluated to see whether it allows spontaneous breathi ng or whether mechanical ventilation is necessary. Keywords Chronic instrumentation; anim al model; impedancometry; percutaneous gast roenterostomy; duodenoscopy; spontaneous breathing; sedati on; analgosedation C o rre spond e nc e to : Jo e rg Sc hno o r MD, De partm ent o f Ana esthe sio lo gy, Unive rsity Ho spital, RWTH Aa ch e n, Pa uw e lsstr. 30, D-52074 Aa c he n, G e rm a ny E-m a il: jo erg.schno o r@gm x.de Accepted 7 October 2002

# Laboratory Animals Ltd. Laboratory Animals (2003) 37, 145–154

146

Schnoor et al.

To understand intestinal motility in man, different animal models have become an essential tool for experimental investigations. Porcine models are widely used in biomedical research (Giardino e t a l. 1997 ). T his is due to their anatom ical and physiological similarity to humans (Noah 1995 ). Initially, the standardization of laboratory anim als and their housing were the main objectives of experimental animal science, and then the principles of anim al `well-being’ and the advan tages of `re®nement’ becam e increasingly recognized (Derrel e t a l. 1997 ). In order to obtai n physiological baseline values of gastrointestinal motility in the pig, it is very important to develop an approach which is feasibl e for investigations in conscious animals which avoid mental or physical distress as well as for systematic measurements under different types of sedati on (Kadirkam anathan e t a l. 1999). T herefore, the aim of our study was to develop a long-term porcine model, which had to ful®l the following: T he habituation of pigs to their investigators and laboratory conditions. T he use of chronic instrumentation with an intraluminal im pedance catheter for measurements of gastroduodenal motility patterns. T he measurement of motility patt erns in conscious, unrestrained, and healthy pigs without mental or physical distress.

Measurements in spontaneously breat hing pigs under sedation and analgosedation. Gentle intravenous induction of the planned sedation and analgosedation for measurements without any in¯uences from premedication with neuroleptic agents. No need for system ic antibiotic prophylaxis because of chronic instrum entat ion.

Materials and methods T he investigation was approved by the local Institutional Animal Use Committ ee. We studied seven male, castrated pigs, German landrace, between 32±40 kg BW.

Expe rim e nta l sc h e d ule (Fig 1) Pigs were allowed to habituate to local housing conditions over 5 days. Over the following 4 days the animals were trained to stay in the laboratory with all investigators and technical equipment. Eac h pig was instrumented at day 10 and measurements started at day 11 in the conscious and unrestrained anim al (M1). Measurements under sedation (M2) and analgosedat ion (M3 ) were performed on days 12 and 13. Ha b itua tio n a nd tra ining T he pigs were housed individually in a single cage for at least 9 days before instrum entat ion.

Fig 1 Experimental schedule with animal housing in single cages for habituation starting from day 1. Training of the animals was performed from day 6 through to day 9. After instrumentation on day 10, measurements were taken from conscious pigs at day 11 (M1) and from sedated (day 12, M2) or analgosedated pigs at day 13 (M3) Laboratory Animals (2003) 37

Porcine model for measurement of gastrointestinal motility

Social contact s with their fellow penmat es were possible by visual contact , vocalizati on, and olfacti on. An inner partition of plexiglas 1 wall prevented the others from direct contact and catheter dislocation. Room temperature was provided within a range of 18±22 C. Relati ve humidity was within 40±70% . Daily rhythm was adjusted by dark and light cycle of 12 h each. T he pigs had free access to water and were fed two times a day with standard diet for maintenance (Muskator 1 , Muskat orWerk, Germany). From the ®fth day each pig was transposed in a smaller cage into the laboratory for 8 h. During this phase the pigs were fed with standard diet. Since the study protocol includes an enteral nutrition during sedation and analgosedat ion, the anim als were additi onally fed with Biosorb1 Multi Fibre (Pfrimmer Nuticia GmbH, Germany). T his was done in order to prevent diarrhoea after feeding with unusual nutrition. T he anim als were habituated to the investigators and to isolati on from other pigs up to the ninth day (see Fig 1). T his time was used to build up a familiar at mosphere for reassuring the pigs if necessary. T he smaller cage measured about two times body length and several tim es body width (76.5 cm6 122 cm678 cm ). T he animals had the opportunity to turn around freely and to lie down with stretched legs. T he behaviour was documented by video. In case an anim al showed continuous restlessness and anxiety, the option of excluding the pig from the study had the approval of the local Institutional Animal Use Committee.

Instrum e nta tio n Instrumentation was performed on day 10. Eac h pig was fasted overnight but had free access to water. Premedication was realized by using 4 mg=kg BW azaperone (Stresnil 1 , Janssen-Cilag GmbH, Germany) and 0.5 mg atropinsulfate (Atropin 1 , B. Braun Melsungen, Germany) i.m. Twenty minutes later 10 mg=kg BW ketam ine hydrochloride (Ketam in 1 , Sano®-Cefa Gm bH, Germany) was injected i.m. After insertion of an intravenous line into an ear vein, general anaesthesia was induced intravenously by 2±4 mg=kg BW propofol (Propofol 1 ,

147

Parke-D avis, Germany) and was maintained by continuous infusion of 3±12 mg=kg BW =h propofol and 0.17±0.5 mg=kg BW =min remifentanil (Ultiva1 , GlaxoWellcome, Germany), as performed in humans (Morgan & Mikhail 1996, T hees e t a l. 2001 ). T he pigs underwent orotracheal intubation and were mechanically ventilated with a ServoVentilat or 900D, oxygen =air (FiO 2 0.3), tidal volume 10 ml =kg BW, respiratory rat e 13±18 breat hs =min without positive end-expiratory pressure (PEEP). Both oxygen saturation (SaO 2) between 94±99% measured by pulse oxym etry on the pig tail and the expiratory carbon dioxide partial pressure (pET CO 2) between 35±38 mmHg measured by intermittent venous blood gas analysis could be secured. Blood pressure was measured noninvasively on the hind limb (Riva-R occi; AS =3 Compact 1 , Dat ex-Engstrom, Finland and Dura CuffTM, Infant Cuff, Criticon, US) using the A. plantaris medialis for measurements. Mean arterial blood pressures (MAP), heart rate (HR), SaO 2, and pET CO 2 were assessed at baseline (tim e 0) and continuously every 15 min during the investigation period (for monitoring see Fig 2). T he animals were inserted with a central venous catheter (CVC, silicone tube, 7 mm outer diameter, 4 mm inner diam eter) into the left jugular vein via venae sectio. T he cat heter was tunnelled subcutaneously towards the paravertebral bac k. Sutures were disinfected using polividon±iodine solution (Betaisadona 1 , Mundipharm a GmbH, Germany). Body temperature (Temp) was controlled twice a day as well as the white blood cell count (WBC ) out of CVC. Fluid replacement was done with 2 ml =kg BW =h of crystal loid solution during surgery. T he percutaneous enterogastrostomy (PEG, GastroPEG 1 , B. Braun Melsungen, Germany) was done as is usually performed in humans (Ponsky & Gauderer 1989 ). After endoscopical identi®cation of the lesser gastri c curvat ure, an over-the-needle cannula was inserted into the stomach (Karl Storz GmbH & Co., Germany; Olym pus Optical GmbH & Co. KG, Germany). A guide inserted through the cannula was grasped by the gastroscope and led out orally to chain up Laboratory Animals (2003) 37

148

Schnoor et al.

Fig 2 Procedures and monitorings compared to investigation period. HR ˆ heart rate; MAP ˆ mean arterial blood pressure; SaO2 ˆ oxygen saturation; pETCO2 ˆ partial pressure of expiratory carbon dioxide; TEMP ˆ body temperature; HCT ˆ haematocrit; WBC ˆ white blood cell count; impedance ˆ measurement of motility by intraluminal impedance technique; enteral nutrition ˆ enteral nutrition via percutaneous enterogastrostomy (PEG)

the catheter for enterogastrostomy. A skin incision of 2 cm at the point of punction helped to pull the PEG through the abdom inal wall. T he end of the PEG was tunnelled subcutaneously towards the paravertebral bac k. Sutures were disinfected by polividon± iodine solution (for PEG and impedance catheter, see Fig 3). T he im pedance catheter (2 mm diameter, 100 cm length, ¯exible, 16 ring electrodes distribut ed over the distal 32 cm, Helmholtz Institute, Germany) was introduced along the PEG into the stomach and grasped by the gast roscope. After passing the pylorus by the gast roscope the catheter was positioned under visual control with 14 electrodes intubating the proximal duodenum. Impedance signals were stored online on a personal computer (PC ) before the impedance tracings were visually analysed on a PC screen (Fi g 4). Locally-developed software was used for data acquisition, analysis, and graphic presentation. Laboratory Animals (2003) 37

Fig 3 Percutaneous enterogastrostomy (PEG) with inserted catheter for intraluminal impedancometry. A haemostasis valve=side port for use with 7–7.5 Fr. catheters was connected to the end of the PEG tube. This procedure enabled an air-tight condition between the GIT and the outer surrounding


149

Fig 4 Impedance tracings (16 channels) which demonstrate gastric motility waves (channels 1–2), and duodenal motility wave (channels 7–16). Channels 3–6 are positioned in the pyloric area

Finally, the CVC and PEG were placed in a catheter bag, which was sutured at the bac k (Unger e t a l. 2000 ). After instrumentation was ®nished, postoperative analgesia had already been started under general anaesthesia with ®rst dosages of metam izolnatrium 20 mg=kg BW i.v. (Novalgin 1 , Hoechst, Germany) and carprofen 4 mg=kg BW s.c. (Rimadyl 1 , P®zer GmbH, Germany). T he pigs were extubated under spontaneous breathing and monitored for the following 4 h. Overnight the anim als had free access to water and were fed during the evening with standard diet.

Mea sure m e nts o n d ay 11 (M1, c o nsc ious pigs) After morning feeding and drinking, the pigs were transported to the laboratory. T he intravenous line and the impedance catheter were connected to the signal transducer for online measurements of central venous pressure and gut motility. Fluid replacement was done with 2 ml =kg BW =h of crystalloid solution during the investigat ion period. Measurements and video control were started for the following 8 h. Four hours aft er

feeding standard diet, enteral nutrition was started for one hour via PEG with a common substrat e 2 ml =kg BW =h (Biosorb1 , Multi Fibre, Pfrim mer Nuticia GmbH, Germany). At the end of the measurements video control and infusion were stopped. CVC and PEG were stored in the catheter bag. To prevent postoperative pain and its consequent distress, continuous analgesia was maintained with daily intravenous injections of metamizol (Novalgin 1 , 20 mg=kg BW i.v.) via the CVC. Body temperature and WBC were controlled in the sam e way as the day before. T he anim als were returned overnight to the pig sty with its standard housing conditions.

Me a sure m e nts o n d a y 12 (M2, se d a te d pigs) After morning feeding and drinking, the animals were transported to the laboratory. An intravenous line was connected and sedation was started with a bolus of 2 mg=kg BW propofol initially, followed by continuous infusion of 3±12 mg=kg BW =h propofol. T he impedance catheter was connected to the signal transducer and measurements were started. Fluid replacement Laboratory Animals (2003) 37

150

Table 1

Schnoor et al.

Parameters monitored for the control of the physical state HR=min

Instrumentation (n ˆ 7) 92.0 Measurement day 1 M1 (n ˆ 2) Measurement day 2 144.9 M2 (n ˆ 3) Measurement day 3 108.2 M3 (n ˆ 3)

WBC ( 103 =ml)

MAP (mmHg)

SaO2 (%)

RR=min

TEMP ( C) HCT

11.7

77.1

10.8

97.9

0.9

15.4 15.4

3.7 0.9

37.2 39.8

1.6 0.2

30.0 30.7

1.2 6.3

15.6 24.0

2.5 13.1

14.6

94.7

15.6

96.8

2.5

15.2

4.8

39.3

1.5

29.3

2.8

24.7

7.4

13.3

79.8

17.4

98.8

1.1

17.3

4.0

39.4

1.3

28.0

3.6

21.4

4.4

With respect to the haemodynamic state, blood pressure, heart rate, respiratory rate and oxygen saturation were measured. To control animals with respect to postoperative infections and gastral bleeding body temperature, haematocrit, white blood cell count were determined. Values of haemodynamic monitoring, respiratory state and white blood cell count are expressed as means SD. HR ˆ heart rate, MAP ˆ mean arterial blood pressure, SaO2 ˆ oxygen saturation, RR ˆ respiratory rate, TEMP ˆ body temperature, HCT ˆ haematocrit, WBC ˆ white blood cell count

was done with 2 ml =kg BW =h of crystalloid solution. Haemodynam ic param eters (MAP, HR, SaO 2) were monitored for the 8 h investigation period. Tem p and WBC were controlled as usual. When necessary, the pigs received 2±4 l oxygen via a face mask to keep the level of oxygen saturation (SaO 2) continuously at between 94±99% . Enteral nutrition was started as in day 11. At the end of the measurements, metamizol and carprofen were injected for analgesia. After recovery from sedation the animals were returned to the pig sty.

Mea sure m e nts o n d ay 13 (M3, a na lgo se d a te d pigs) After a bolus of 2 mg=kg BW propofol, analgosedat ion was initially performed by continuous infusion of 3±12 mg=kg BW =h propofol and 2±8 mg=kg BW =h fentanyl hydrogencitrat (Fentanyl 1 -Janssen, Janssen, Germany). Procedures and monitoring were similar to the day before (Table 1). At the end of the measurements, anim als were killed by an overdose of barbit urate and pathohistological examinations were performed.

Results For results and modi®cations see Table 2.

Ha b itua tio n All the animals were included in the study. None had to be excluded from the study because of mental distress. It took between 8 and 9 days to habituate and train the pigs to Laboratory Animals (2003) 37

the investigators and for the laborat ory conditions. At the end of this phase the pigs’ behaviour was attenti ve and playful, with several sleeping phases of 10±30 min.

Instrum e nta tion For instrumentation general anaesthesia was initially maintained by continuous infusion of propofol and remifentanil. Clinical signs of an inadequate depth of anaesthesia, such as elevated blood pressure or heart rate and animal movements required higher dosages of propofol while the doses of remifentanil remained unchanged. Duodenal intubation by endoscope was not possible in four cases because of pyloric spasms. Despite repetitive application of butylscopolamin 2±4 mg i.v. (Buscopan 1 , Boehringer, Germany) pyloric spasms persisted. T hese pigs had to be excluded from measurements of duodenal motility on days 12 and 13. T herefore, we decided to use (S)ketam ine hydochloride ((S)-Ketanest 1 , Parke-D avis, Germany) instead of remifentanil for the subsequent experiments. After orotracheal intubation, anaesthesia was maintained by propofol and (S)-ketanest BW =h without any signs of an inappropriate depth of anaesthesia. We had no further problem s with duodenal intubation. T he pathohistological examination of the ®rst pig revealed a bac terial colonizat ion of the CVC and PEG. T herefore, with all further animals, polividon±iodine (5 ml) was introduced into the subcutaneous tunnels. T his procedure successfully prevents bac terial infections.


151

Table 2 Basic model and the resulting changes which became necessary for a successful experimental procedure Days

Basic model

Requirements

Habituation Instrumentation General anaesthesia

Practicable Propofol: 3–12 mg=kg BW=h Remifentanil: 0.17–0.5 mg=kg BW=min (S)-ketamine (initially not used) Sutures

None 15.7 4.2 mg=kg BW=h 0.29 0.08 mg=kg BW=min

Local disinfection Heparinization of CVC Measurement day M1 Measurement day M2 Sedation Measurement day M3 Analgosedation

(Initially not planned) Practicable

0.5 mg=kg BW=h Sutures, subcutaneously tunnels, catheters CVC heparinization None

Propofol: 3–12 mg=kg BW=h

12.2

5.0 mg=kg BW=h

Propofol: 3–12 mg=kg BW=h

13.2

5.7 mg=kg BW=h

BW ˆ body weight, CVC ˆ central venous catheter

After instrumentation of the ®rst pig we found subcutaneous emphysema of the area around the PEG tunnel. T he PEG cat heter is made from silicone which enables transfer of oxygen to the surrounding tissue. Since the pressure in the stomach is raised during gastroscopy due to act ive gas insuf¯ations, gas transfer through the silicone wall of the PEG was increased. T herefore, we ®lled the PEG catheter with water in order to reduce gas transfer out of the stomach into the subcutaneous tissue. Subcutaneous emphysema could be prevented in further experiments using this procedure.

Mea sure m e nts o n d ay 11 (M1) Measurements of motility in conscious and unrestrained pigs were feasible. As during the habituation on day 9, the pig’s behaviour was attentive and playful, with several sleeping phases as well. The pigs showed no signs of pain, distress, or loss of body weight over the investigation period. Under pathohistological examination we found thrombotic clotting at the end of the CVC in two pigs which prevented blood aspiration. After heparinization of the CVC with 400 IU i.e. blood could always be aspirated in subsequent experiments. Mea sure m e nts o n d ays 12 (M2) a nd 13 (M3) Sedation was maintained with propofol. Analgosedation was maintained with

propofol and fentanyl. While the pigs were sleeping on the left side, four pigs had to receive oxygen via a face mask to keep saturation above 94% . For values of noninvasive blood pressures and respiratory states see Table 1.

Discussion We developed a new approach for measuring intestinal motility in pigs using the intraluminal im pedance technique. T he main keyst ones of this approach are (a) habituation and training, (b) chronic instrumentation under general anaesthesia, and (c ) measurements in conscious and in sedated or analgosedated pigs. T his approach allowed: T he habit uation of the pigs to investigators and for laborat ory conditions. T he use of chronic instrumentation with an intraluminal impedance cat heter for measurements of gast roduodenal motility patterns. T he measurement of motilit y patterns in conscious, unrestrained, and healthy pigs without mental or physical distress. Measurements in spontaneously breathing pigs under sedation and analgosedat ion. Gentle intravenous induction of the planned sedation and analgosedation for measurements without any in¯uences from premedication with neuroleptic agents. Laboratory Animals (2003) 37

152

Schnoor et al.

No need for systemic prophylactic antibiotics because of chronic instrumentation.

Th e ne e d o f a po rc ine m o d e l w ith unre stra ine d pigs T here are many factors which can affect intestinal motilit y. Beside pain and stress, a large number of pharmacological substances, commonly used in humans, in¯uence motility patt erns (Zander 1998). To investigate these effects systemat ically, a reliable animal model which enables the examinati on of motility patterns in conscious anim als with and without the administration of drugs is needed. Porcine models are frequently used in experimental medicine. Comparabl e anatomic and physiological conditions enable the investigators to use monitoring methods which are establi shed in man (Dodds 1982 ). Porcine models are frequently used and are favoured by a number of different training centres (Noah 1995). And yet porcine models in conscious but unrestrained anim als are still a challenge for experimentation (Kadirkam anat han e t a l. 1999 ). Ha b itua tio n a nd tra ining T he pigs were kept in single cages for 9 days for habituat ion and training. Since the pigs had to undergo chronic instrumentation it was necessary to keep them in single cages in order to avoid dislocation or damage of the catheters by other pigs. Furthermore, the cage walls had to be prepared so that the instrumented pig had no chance of dislocating the catheter using its surroundings. However, it was also most im portant to reduce the impairment of the mental state as much as possible. Plexiglas 1 addressed both concerns: safety with respect to the catheters and social contact between the pigs for their mental welfare. Communication with other pigs was possible visually, by vocalizati on and olfact ion. For the last 4 days they were daily transported to the laborat ory in a smaller cage for habit uation with the investigators, room and noises. T he duration of stay started on day 6 and was planned for 8 h. On day 6, the pigs were excited and restless. T herefore, with regard to animal welfare, Laboratory Animals (2003) 37

training was ®nished after 2±4 h. On the following days (7±9) they began to be reassured and to investigate the environment with curiosity. On day 8, training phases were extended to 8 h and in most cases it was possible to make the pigs lie down by brushing their abdom inal walls. To occupy the pigs’ minds, we offered them ball s and towels to play with. Our experience was that habituati on for 9 days, with a 4-day training phase included, was necessary and was long enough to enable the measurement of intestinal motility in conscious and unrestrained pigs on the following day.

Instrum e nta tion In order to be able to investigate the effects of sedatives and analgesics on motility patterns, we established in our protocol the new intraluminal impedance technique. It was im portant to develop a technique which allowed measurements over several days with fully awak e and unrestrained pigs, therefore it was necessary to avoid using an oral cat heter technique. We decided to use a PEG for guiding the impedance catheter into the duodenum. Under general anaesthesia, intestinal intubation was accomplished using a standard gastroscope or coloscope, as has been previously described (Gholson e t a l. 1990 ). Duodenal positioning of the impedance catheter was realized under visual control. For successfully duodenal positioning in the case of 35 kg BW pigs, an endoscope of 130±150 cm in length was needed. We used a gastroscope (130 cm length, Karl Storz GmbH & Co. KG, Germany) or a coloscope (176 cm length, Olym pus Optical Co. GmbH, Germany). T he latter was more convenient with regard to the usage of the working channel because repeated loops were necessary for duodenoscopy. T he catheter bag prevented skin irritat ion by its wide ranging ®xation on the bac k. A cord-retrac table leash helped tighten up intravenous lines and im pedance catheters while the pigs were freely moving on measurement day M1 (Unger e t a l. 2000 ). To enable measurements of gastrointestinal motility on the day after instrumentati on, general anaesthesia for the


instrumentation was ®rst performed with propofol and remifentanil. T he in¯uence of propofol on motility patterns seems to be negligible (Behrendt & Kuhlen 2000 ). A disturbanc e of physiological motility on the following day caused by remifentanil was not expected, because remifentanil is an opioid with a very short context-sensitive half-life (Jensen e t a l. 1992 ). Propofol dosages used in clinical routines for the maintenance of general anaesthesia in humans were unable to secure an appropriate depth of anaesthesia. Propofol was continuously infused by motorpump within 10±20 mg=kg BW =h, as described for pigs by other investigators (Graham e t a l. 1988 ). T he continuous infusion of remifentanil was also regulated by checking for clinical signs of inappropriate depth of anaesthesia. As expected, recovery time was short. T he pigs awoke about 10 min after termination of the propofol and remifentanil application. As well as the persistence of pyloric spasms, we observed a distended duodenum under gastric air insuf¯at ions. We interpreted these clinical signs as indicat ors of reduced intestinal motility, most likely induced by remifentanil. We decided to use ketam ine instead of remifentanil for general anaesthesia. T he following duodenal intubations under ketam ine were practicable in all four cases. T he pigs evoked between 20 and 45 min after cessation of the continuous infusion of (S)ketanest 1 . Irrespective of the type of analgesics used, free intak e of standard diet and water was possible after extubation without any problems. Postoperative analgesia using metamizol and carprofen were suf®cient. No reduction of body weight was found during the 3 days of investigati on. All the anim als showed an increase of body weight of about 2 kg through the 4 experimental days. Beside other monitored parameters (Tem p, WBC, HR, MAP, RF and defecation ) the increase of BW indicated an acceptable physical state. T he inserted CVC enabled us on consecutive days to initiate sedation and analgosedat ion directly, without any distress or injection of neuroleptic agents which probably in¯uence motilit y patterns (Behrendt & Kuhlen 2000 ). Injections of

153

¯uids and drugs were possible throughout the total investigation period, alt hough aspiration of blood was impossible on later days in two cases because of clot form at ion at the end of the catheter, as found at ®nal pathohistological exam ination. We found no other signs of thrombotic complications. After heparinization of the catheters, distinct clot formation at the end of the cat heter could be prevented and blood sam pling could be realized throughout the experiment. We avoided any antibiotic therapy because of possible interactions with motility patterns (Bozkurt e t a l. 2000, Koutsoumbi e t a l. 2000 ). T herefore, we carefully disinfected all sutures, catheters and subcutaneous tunnels. After colonizati on was found in the subcutaneous tunnel of the ®rst pig, we specially introduced polovidon±iodine (5 ml) into the subcutaneous tunnels. T hereafter, we did not ®nd any signs of infection on further pathohistological examination.

Im pe d a nc o m e try T he im pedance technique permits the determination of oesophageal and duodenal motility patterns in humans (Silny e t a l. 1993, Nguyen e t a l. 1995 ). It dem onstrates the sam e features with regard to durati on, propagation directions, and contract ion frequency as recorded by manometry (Nguyen e t a l. 1999). In contrast to manometry, the im pedance technique promotes measurements independent of surrounding pressure. Disregarding the used technique for measurements this approach might promote duodenal intubation of any ¯exible catheter ( 2 mm outer diameter) to investigate gastrointestinal motilit y in conscious unrestrained pigs. T his study focused on habituati on and instrum entation in a chronic porcine model; and results of gast roduodenal motility are beyond the scope of this study and will be described separately. Me a sure m e nts und e r se d a tio n a nd a na lgo se d a tio n During sedation and analgosedati on, the pigs were kept sleeping in ¯ank recumbency as when they were sleeping naturall y. Laboratory Animals (2003) 37

154

T herefore, dosages had to be found for tolerating positioning with suf®cient spontaneously breat hing. Both measurements under sedati on and analgosedati on with continuous infusions of propofol with or without fentanyl were practicable. T he additi onal fentanyl applicat ion on M3 did not decrease the dose of propofol infusions. In consequence, spontaneous breathing rates were reduced and it was necessary to use 2±4 l =min oxygen via a face mask to ensure oxygen saturat ion was kept above 94% . In conclusion, percutaneous enterogastrostomy using general anaesthesia with propofol and ketam ine allows duodenal intubation with an impedance catheter. T he approach developed which includes habituation and training, promotes long-term measurements of motility patterns in conscious and unrestrained as well as in sedat ed and analgosedat ed pigs. Ac k no w le d gm e nts T his study was supported by START, RWT H Aachen, Germany, and B Braun Melsungen AG, Melsungen, Germany.

References Behrendt W, Kuhlen R (2000) Auswirkungen von AnaÈsthesie und Analgosedierung auf die Darmfunktion. C h irurgisc h e G a stro e nte ro lo gie 16, 1±6 Bozkurt A, Deniz M, Yegen BC (2000) Cefaclor, a cephalosporin antibiotic, delays gastric emptying rate by a CCK-A receptor-m ediated mechanism in the rat. British Jo urna l o f Ph a rm a c o lo gy 131, 399±404 Derrel CJ, Rager DR, Calpin JR (1997) Animal wellbeing: III. An overview of assessm ent. La b o ra to ry Anim a l Sc ie nce 47, 580±5 Dodds WJ (1982) An effective mass-sc reening program for animal models of the inherited blood disorders. Pro gre ss in C linica l a nd Bio lo gica l Re se a rch 94, 117±32 Gholson CF, Provenza JM, Silver RC, Bacon BR (1990) Endoscopic retrograde cholangiography in the swine. A new model for endoscopic training and hepatobiliary research. G a stro inte stina l End o sc o py 36, 600±3 Giardino R, Fini M, Orienti L (1997) Laboratory animals for arti®cial organ evaluation. Inte rna tio na l Jo urna l o f Arti® cia l O rga ns 20, 76±80

Laboratory Animals (2003) 37

Schnoor et al.

Graham MR, Thiessen DB, Mutch AC (1988) Left ventricular systolic and diastolic function is unalter during propofol infusion in newborn swine. Anesth e sia & Ana lgesia 86, 717±23 Jensen AG, Kalman SH, NystroÈm PO, Eintrei C (1992) Anaesthetic technique does not in¯uence postoperative bowel function: a comparison of propofol, nitrous oxide and iso¯urane. C a na d ia n Jo urna l o f Anesth e sia 39, 938±43 Kadirkamanathan SS, Yazaki E, Evans DF, Hepworth CC, Gong F, Swain CP (1999) An ambulant porc ine model of acid re¯ux used to evaluate endoscopic gastroplasty. G ut 44, 782±8 Koutsoumbi P, Epanomeritakis E, Tsiaoussis J, Athanasakis H, Chryson E, Zoras O, Vassilakis JS, Xynos E (2000) T he effect of erythrom ycin on hum an esophageal motility is mediated by serotonin receptors. Am e rica n Jo urna l o f G a stro e nte ro lo gy 95, 3388±92 Morgan GE, Mikhail MS (1996) C linic a l Ana esth e sio lo gy, 2nd edn. London: Appleton & Lange Noah MD (1995) An established porc ine model for animate training in diagnostic and therapeuti c ERCP. End o sc o py 27, 77±80 Nguyen HN, Silny J, Wuller S, Marschall HU, Rau G, Matern S (1995) Chyme transport patterns in hum an duodenum , determined by multiple intraluminal impedancometry. Am eric a n Jo urna l o f Ph ysio lo gy 268, G700±8 Nguyen HN, Silny J, Matern S (1999) Multiple intraluminal electrical impedancometry for recording of upper gastrointestinal motility: current results and further implications. Am e rica n Jo urna l o f G a stro ente ro lo gy 94, 306±17 Ponsky JL, Gauderer MW (1989) Percutaneous endoscopic gastrostomy: indications, limitations, techniques, and results. Wo rld Jo urna l o f Surge ry 13, 165±70 Silny J, Knigge KP, Fass J, Rau G, Matern S, Schumpelick V (1993) Veri®cation of the intraluminal multiple electrical impedance measurement for the recording of gastrointestinal motility. Jo urna l o f G a stro inte stina l Mo tility 5, 107±22 T hees Ch, Frenkel Ch, Hoeft A (2001) Remifentanil in der NeuroanaÈsthesieÐeine multizentrische Anwendungsbeobachtung. AnaÈ sth e sio lo gie & Inte nsivm ed izin 42, 205±11 Unger JK, Gerlach JC, Juhr NC, Rossaint R (2000) Development of a special catheterbag to enable arti®cial organ evaluation in concious, unrestrained pigs: technical note. Inte rna ti o na l Jo urna l o f Arti® c ia l O rga ns 23, 268±74 Zander JF (1998) DarmmotilitaÈtsstoÈ rungen bei IntensivpatientenÐUrsachen und therapeutische MoÈglichkeiten. In: Re fre sh e r C o urse : Ak tue lle s Wisse n fuÈ r AnaÈ sth esiste n (Deutsche Akademie fuÈr AnaÈsthesiologische Fortbildung, eds). Berlin: Springer Verlag, pp 69±82