- Email: [email protected]
Optimizing the experimental environment for dog telemetry studies
Journal of Pharmacological and Toxicological Methods 54 (2006) 141 – 149 www.elsevier.com/locate/jpharmtox
Original article
Optimizing the experimental environment for dog telemetry studies Anja Klumpp, Thomas Trautmann, Michael Markert ⁎, Brian Guth Department of Drug Discovery Support, General Pharmacology Group, Boehringer Ingelheim Pharma GmbH & Co. KG, Germany Received 2 February 2006; accepted 29 March 2006
Abstract Introduction: The objective of this study was to test the influence of housing conditions on hemodynamics during cardiovascular general pharmacological studies. Our goal was to optimize both the quality of the data through an optimization of the physiological conditions, as well as to ensure the dog's well-being in general pharmacological studies. Methods: Two groups of four dogs were equipped with radiotelemetry transmitters and continuously monitored in two different housing models. Model I consisted of 4 cages, two on each site of a corridor. Model II consisted of 4 cages positioned in a row, where the bordering cages were not separated with a metal plate. The physiological status of the dogs in the different housing models was based on the frequency of vocalizations and the average resting heart rate, as well as video monitoring. Results: The housing arrangement during the study had a remarkable effect on the hemodynamics measured. The hemodynamic parameters were best when the dogs were housed with their usual run mate. In this setting, they have impressively low average heart rates of about 60 bpm during the entire study, was associated with fewer vocalizations. Discussion: This study demonstrated that the quality of the acquired cardiovascular data for conscious dogs is dependent on the pen configuration and group make-up during a study. © 2006 Elsevier Inc. All rights reserved. Keywords: Methods; Refinement; Conscious dog; Telemetry; Housing conditions
1. Introduction The model preferred for cardiovascular studies according to the ICH S7a guideline is the conscious dog. The gold standard approach for studying conscious dogs is the use of a telemetry system to record cardiovascular parameters, i.e. blood pressure, left ventricular pressure, heart rate and ECG. The rationale for performing these studies in conscious dogs is based on the lack of possible influence of anesthesia on the parameters being measured. However, the conscious state is inherently responsive to environmental stimuli such that a higher spontaneous variability of the measured parameters must be taken into account. Thus, the quality of these types of cardiovascular studies is likely dependent on the degree of training of the animals used and the environment during a study. To show subtle drug-induced changes in heart rate, for example, it is essential to have a relatively constant heart rate without too
⁎ Corresponding author. Tel.: +49 7351 54 8727; fax: +49 7351 83 8727. E-mail address: [email protected] (M. Markert). 1056-8719/$ - see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.vascn.2006.03.010
many interruptions causing excitement (Guth, Germeyer, Kolb, & Markert et al., 2004). Furthermore, a stable heart rate is desirable to facilitate correct QT detection and heart rate correction (Meyners & Markert, 2004). An unstable heart rate therefore detracts from the detection of a possible prolongation of QT interval. The term refinement signifies the modification of any procedure, from the time a laboratory animal is born until its death, so as to minimize the pain and distress experienced by the animal and enhance its well-being. As mentioned by the members of the Joint Working Group on Refinement (Penny Hawkins et al., 2004), the ideal environment would allow for housing in stable, compatible groups. Optimizing animal welfare is not only important from the viewpoint of ethics; it is also a prerequisite for conducting good quality studies. The experience of pain and other stress will result in physiological changes (e.g. heart rate) that increase the variability of the experimental results and a loss of sensitivity. A high variability of heart rate, for example, is often caused by excited or stressed dogs. Thus, it is in the interest of good science to ensure that conditions regarding animal housing are optimized. Important lessons have been learned about how to
142
A. Klumpp et al. / Journal of Pharmacological and Toxicological Methods 54 (2006) 141–149
validate methods. The need to have predictive models available before validation is performed: the necessity to take the variability of the animal-based data which is to be used as the validation standard into account, and the need to have wellmanaged validation programs available (Purchase et al., 1998). Future progress will depend on the development of novel methods. The use of animals, especially dogs or monkeys, can only be justified, when clinically relevant data can be gained from few, but well-designed in vivo studies (Luft & Bode, 2002). This can only be assured when the dogs feel comfortable and aren't stressed which also reflects the housing conditions. Enrichment activities can include group housing when possible, rather than individual cages and use of materials similar to those found in their natural habitats. Once the study is in progress, it is important that the laboratory staff is well trained and competent in the handling of the species that is being used. One of the additional potential influences that may affect the hemodynamic status of the animals is the exact makeup of the group to be studied based on sex, family hierarchy and physical orientation to the other group members. It was the purpose of this study to assess the role of these factors on hemodynamic stability of the dogs, as an indicator of well-being and possible stress. 2. Methods 2.1. Animals Treatment of the animals followed the German Law on the Protection of Animals and was performed with permission from the State Animal Welfare Committee. Trained labrador dogs (male or female), at least 1.5 years of age, with body weights between 22 to 30 kg bred at Boehringer Ingelheim Pharma GmbH and Co. KG, Biberach were used. The dogs were group-housed as pairs of 2 in separate cages and had access to water ad libitum and were fed a standard dog diet once daily. They also had daily exercise periods of at least 1 h, each afternoon. To monitor the health of the animals, we collect blood samples every 3 months to determine the blood count and clinical parameters, including electrolytes and kidney parameters. The dogs were trained in the telemetry lab beginning 3 months before surgery, which was performed when the dogs were at least 1.5 years old. 2.2. Telemetry system The telemetry system used to measure cardiovascular parameters is manufactured by Konigsberg Instruments, Inc. (Pasadena, CA) and marketed by RMISS (Deleware). It consists of 5 major components: a) an implantable unit; b) a receiver (antenna) located in the animal's cage together with an amplifier; c) ambient pressure monitor to measure atmospheric pressure; d) a PC-based “base station” to receive and process the amplified signals; e) a PC-based data acquisition system (NOTOCORD Hem 3.5) to process signals. The implantable unit (“T27” total implant) consists of (1) two high fidelity pressure transducers (5.0 and 4.0 mm
diameter), (2) an ECG cable, (3) micro-power battery-operated electronics that process and digitize the information from the pressure transducers and the ECG lead, (4) a radio-frequency transmitter that sends the signals to the telemetry receiver, and (5) a battery. A small cable projecting from the transmitter contains a magnetic switch that allows the device to be turned on and off. Prior to implantation the zero value of the two pressure transducers are calibrated using a manometer and 250 mm Hg is set to 5 V. 2.3. One-time surgical implantation The transducers of the T27 implant were calibrated and the unit was sterilized using a low pressure ethylene oxide process prior to implantation. Dogs were anaesthetised with a combination of Rompun (xylazine hydrochloride, 1 ml/10 kg, i.v.) and Ketavet (ketamine hydrochloride, 0.7 ml/10 kg, i.v.) after premedication with Atropine sulfate (0.04 mg/kg i.m.) and ventilated with 66% O2 and 1%–1.5% isofluorane at a ventilation rate of 14/min. All procedures were performed under aseptic conditions using sterilized equipment. The dogs were placed in a lateral recumbency with the left side facing the surgeon. An incision was made between the fifth and sixth intercostal space. A small pocket was opened in the abdominal wall for implantation of the transmitter, battery housing, and induction switch coil. The cables with both pressure transducers and ECG leads extending from the ventrally implanted transmitter were guided subcutaneously to the lateral incision. The antenna was guided subcutaneously from the transmitter location towards the spine and then runs parallel to the spine for ∼25 cm. The initial ventral incisions required for battery and transmitter placement were closed. A left thoracotomy was performed between the fifth and sixth intercostal space to expose the left ventricle apex for insertion of the left ventricular Konigsberg transducer. The aorta pressure transducer was implanted next. The aortic transducer, which also served as one electrode of the ECG, was inserted into the thoracic aorta just below the aortic arch. The transducer was sutured into place and blood flow was restored. The lung was then inflated and the intercostal muscles were sutured closed and the pneumothorax evacuated. Chest incisions were closed. The gas anaesthesia was then turned off and dogs were allowed to wake up. Analgesics and antibiotics (Temgesic and Duphamox® LA, Amoxicillin-Trihydrat) were administered for 2 days (Temgesic) and 10 days (Duphamox) following the procedure to support a good recovery and to ensure that the dog has no postoperative pain, a transdermal plaster (fentanyl, 25 μg/h) was placed on the skin for 2 days. Dogs are allowed to recover for at least 21 days before experiments using test substances are initiated. 2.4. Runs in the telemetry lab Dogs used for studies in which the hemodynamic and ECG data are collected telemetrically are normally group-housed at
A. Klumpp et al. / Journal of Pharmacological and Toxicological Methods 54 (2006) 141–149
the company's animal's resource facility under standardized housing conditions that meet or exceed regulatory requirements. For conducting studies, however, the animals are transported to the telemetry laboratory having special rooms for the animals for the duration of the study and equipped with the appropriate antennae for collecting the transmitted signals as well as video monitoring. The dogs are trained over months in these runs and are adapted to them as secondary “home cages”. Two sets of 4 animal runs each are available in the telemetry lab. Whereas many aspects of these 2 areas are identical (see below), they do differ in subtle ways which might impact on study outcome. It is the comparison of these subtle differences in term of their affect on the hemodynamic status of the dogs which form the basis of this study. Common features: Both sets of runs provide each animal 4 m2 space and a cage height of 3 m. Room temperature and humidity is controlled (22 ± 2 °C; 60 ± 15% humidity) with a ventilation turnover of 13/hr (11/h is the required). They maintain a 12 h light (365 lx) and dark (3 lx, “moonlight”) cycle. Background “noise” is supplied during the lighted period in the form of a radio. Independent features: The physical layout of the 4 runs is different in the two telemetry labs. In the first (Model I), depicted schematically in Fig. 1, there are 2 cages on each side of a central corridor. Dogs in adjacent cages do not have visual contact due to a presence of a metal plate, but can see the dogs in the 2 opposing runs.
Cage 4
Cage 3
Cage 2
Fig. 2. Model II.
The frequencies of our dogs used in the studies are: 209.8 MHz, 200.9 MHz, 197.9 MHz, 178.0 MHz, 210.90 MHz, 205.99 MHz, 189.6 MHz, and 206.9 MHz. We used a specially designed floor antenna (Fig. 3) to improve the quality of the transmitted signals in that all 4 animals could be measured together as depicted schematically in Fig. 2. The diameters of the receiving antenna loops were calculated according to the following equation: k¼
c 240 000 m=s ¼ ¼ 1:2 m f 200 MHz
where c is the spreading velocity of the transmitted signal and f is the transmitted frequency (average over all frequencies used in the lab). The other factor examined in this study was the group makeup of the dogs used. For this, we compared a group of 4 males with a group of 4 females. It should be noted that an analysis of
Cage 4
FLOOR
Cage 1
Cage 2
Fig. 1. Model I.
Cage 1
FLOOR
The second telemetry lab has the 4 cages arranged in a row (Model II) (Fig. 2). In contrast to the other lab, the dividers between the cages allow visual contact and can be removed as needed to allow group housing, either as established pairs or all 4 together. As the telemetry device used for this study uses different transmitting frequencies, the animals could be studied when held together.
Cage 3
143
Fig. 3. Special design of the floor antenna.
144
A. Klumpp et al. / Journal of Pharmacological and Toxicological Methods 54 (2006) 141–149
Fig. 4. Housed alone in the old lab environment with individual runs and only visual contact to their usual run mate.
statistical power has shown that using a cross-over experimental design, a sample size of 4 gives a 90% power to detect a 10% change in the hemodynamic parameters measured. 2.5. Experimental design Two groups of dogs were used: Group I: consisting of 4 females Group II: consisting of 4 males.
Four variations of the run design in Model II were compared: 1. Housed alone without visual contact (Fig. 6) 2. Housed alone with visual contact to all other dogs (Fig. 7) 3. Run mates housed together with visual contact to others (Fig. 8) 4. All 4 dogs together (Fig. 9). The following parameters were used for comparison the physiological status in the different housing models:
Each group was studied in the 2 different housing models: Housing Model I Model I consists of 4 cages, two on each site of a corridor. In this model each dog has only visual contact to the dogs sitting opposite, because the bordering cages are separated by a metal plate (Figs. 4 and 5). Housing Model II Model II consists of 4 cages positioned in a row, whereas the bordering cages are not separated with a metal plate and additionally the floor was designed (for the dogs) with a rough surface and that is covered with wood shavings. In this model, the partition walls could also be removed, giving us the possibility of a new run design, in which dogs could be studied while either individually, or group-housed.
1. Time from first contact to the pen to baseline heart rate (internally defined as a HR = <60 bpm). 2. Mean (± S.E.M.) HR during the 22-h observation period. 3. Time after placebo treatment to baseline heart rate (internally defined as a HR = < 60 bpm). 4. Number of vocalizations during daytime measurements 5. Number of movement vs. sleeping phases. All studies used the same protocol. The dogs were placed into the cages at 7:00 a.m. and the telemetric implant was switched on. After a calming phase of 30 min the signal offset was performed (7:30 a.m.). Data were then collected for 2 h (7:30 a.m. – 9:30 a.m.) and is referred to as the “initial control”.
Fig. 5. Housed alone in the old lab environment with individual runs and only visual contact to their usual run mate.
A. Klumpp et al. / Journal of Pharmacological and Toxicological Methods 54 (2006) 141–149
145
Fig. 6. Housed alone in new lab without visual contact.
To simulate a drug application, the lab was then entered by a technician for a few minutes (9:30 a.m.). Data collection was then continued for the next 22 h. At 4:30 p.m. the dogs were fed with continuous collection of hemodynamic data.
Student's t-test. Statistical significance was accepted when p < 0.05. Using the error term of the analysis of variance as an estimate for the variability many-to-one t-tests comparing the dose groups of the test compound with the control group were added (Table 1). The evaluation was performed using the software package SAS 8.02 (SAS Institute Inc., Cary, USA).
2.6. Data acquisition and analysis 3. Results Digitized telemetry signals were processed by NOTOCORD (Notocord, Paris) software (Hem 3.5) on a beat-to-beat basis. The following parameters were continuously recorded for the duration of the experiment: aortic pressure (AP), left ventricular pressure (LVP), ECG lead II and body temperature. The hemodynamic and ECG parameters were calculated during the experiment and include: systolic, diastolic and mean aortic pressure, peak systolic and end-diastolic left ventricular pressure, LV dP/dt max and dP/dt min, heart rate; PQ-, QRSand QT-intervals. NOTOCORD software was used for acquisition of data whereas EXCEL was used for some basic data analyses. Data were summarized at predefined time points by calculating median values ± S.D. At each time point (every 10 min), a minimum of 400 sequential beats were used to calculate the median value. The summarized data are given as mean values ± SEM. The different base levels of the individual animals were taken into account by referring the measured values after administration of the test compound to the pre-treatment values. With these values, the standardised area under the curve (AUC divided by interval length) was calculated for the time intervals of interest. Comparisons between treatment and placebo were performed by one-way analysis of variance (ANOVA) followed by the
3.1. Model I In Housing Model I, although animals calmed down fast after disturbances (Graph 1), higher heart rates (70 bpm) and longer periods of vocalization (47 times per day, Table 1) indicated the restlessness of the dogs. This appeared to be more pronounced in females (group I). Females had a relatively high average heart rate of 75 bpm, higher variability between the subjects and more peaks where the dog's heart rate was increased during vocalization (Graph 1, circles). 3.2. Model II The dogs in Housing Model II were, in general, calmer than in Model I. Remarkable is the fact that they had very low heart rates. The dogs had heart rates between 50 and 77 bpm throughout the entire study, except when the “placebo application” was performed at which time the heart rate increased to 120 to 140 bpm. It reached ”pre-dose” values within few minutes, however (Table 1). In Housing Model II the barking periods were shorter (Table 1) and not as frequent (15 times per day). However, it has to be
Fig. 7. Housed alone in new lab with visual contact to all other dogs.
146
A. Klumpp et al. / Journal of Pharmacological and Toxicological Methods 54 (2006) 141–149
Fig. 8. Run mates housed together in new lab with visual contact to others.
mentioned that a difference between females and males was found (see below). 3.2.1. Group I (female) The females were very calm in all four variations of Model II, but especially when they were housed alone without visual contact where they almost never barked. In all variations, however, they behaved very calmly indicated by stable hemodynamic parameters throughout the study (Graph 1). Subjectively, however, the dogs appeared more comfortable when housed with their usual run mate (Table 1). 3.2.2. Group II (males) The behaviour of the males in the various run designs of the Housing Model II was very different compared to the females. They were calm, when they were housed alone without visual contact or when housed with their run mate with visual contact to all other dogs. The dogs were less calm when they were housed alone with visual contact or all 4 dogs together. In this case the males were very restless and they barked frequently. When they were housed alone with visual contact to all other dogs, instances of fighting were noted through the fence, possibly to determine their hierarchy. In contrast to the females, they were also very restless when they were kept all 4 together. Although they had the chance to determine their hierarchy, they were not calm. As a result, the measured parameters showed moderate fluctuations (Table 2). The best experimental environment seems to be when run mates were housed together with visual contact to others. In this environment they have the possibility to maintain their established hierarchy. After the hierarchy is determined, they
are calm. In this setting, they have impressively low average heart rates of about 50–60 bpm (Table 2). When drug application is simulated, the increase in heart rate is very moderate compared to when they are single-housed and the heart rate returns to pre-dose values much faster again. There is almost no inter-subject variability detectable. 4. Discussion This study demonstrated that the quality of the acquired cardiovascular data using conscious dogs is dependent on the pen configuration and group make-up during a study. To show drug-induced changes in heart rate, or other hemodynamic parameters, it is advantageous to have a stable heart rate with little spontaneous variability during the study (Guth et al., 2004). This can only be achieved when the dogs are unstressed, which reflects the holding conditions. A low frequency of vocalization episodes, which could be shown clearly in this study, is an additional marker of optimized housing conditions (Senn & Lewin, 1975). Additionally, less barking is thought to increase the welfare of dogs (Sales et al., 1997) since the dogs can produce barks of well over 100 dB. Based on the results of the study we conclude that the housing configuration during a study can have a substantial impact on the data collected. This is both in terms of the absolute values observed as well as on the variability over time. Housing animals during a study with their usual housing mate appears ideal. Furthermore we conclude that females were consistently calmer in this type of study demonstrating lower heart rates and less excitability. Use of females may therefore offer advantages to detect drug-induced hemodynamic effects.
Fig. 9. All 4 dogs together in new lab environment.
A. Klumpp et al. / Journal of Pharmacological and Toxicological Methods 54 (2006) 141–149
147
Table 1 Group I (females) Heart rate
No. of vocalizations
Average on day 1
Model I
Model II Model II Model II Model II
Housed alone in the old lab environment with individual runs and only visual contact to their usual run mate Housed alone in new lab without visual contact Housed alone in new lab with visual contact to all other dogs Run mates housed together in new lab with visual contact to others All 4 dogs together in new lab environment
Average on day 2 Day 1
Before noon In the afternoon In the evening At night Before noon
Before noon In the afternoon
73
71
55
49
51
15
32
50
60
57
52
55
5
5
55
62
59
52
55
11
12
60
68
64
60
58
9
7
55
67
60
52
56
8
7
SAP
No. of vocalizations
Average on day 1
Model I
Model II Model II Model II Model II
Housed alone in the old lab environment with individual runs and only visual contact to their usual run mate Housed alone in new lab without visual contact Housed alone in new lab with visual contact to all other dogs Run mates housed together in new lab with visual contact to others All 4 dogs together in new lab environment
Average on day 2 Day 1
Before noon In the afternoon In the evening At night Before noon
Before noon In the afternoon
128
124
119
115
117
15
32
125
119
114
108
122
5
5
122
117
109
104
113
11
12
118
118
116
118
118
9
7
127
120
118
120
111
8
7
This might offer a further benefit since some studies have shown that female gender is more susceptible to drug-induced QT-prolongation.
In 1959, William Russell and Rex Burch published “The Principles of Humane Experimental Technique” (Russell & Burch, 1959). They recommended that if animals were to be
GROUP I 140 Model I housed alone with visual contact to the dogs sitting opposite Model II housed alone without visual contact Model II housed alone with visual contact to all other dogs Model II run mates housed together with visual contact to others Model II all 4 dogs together
130 120
heart rate (min-1)
110 100 90 80 70 60 50 40
Median of 10 min 30 -1
0
1
2
3
4
5
6
time (hours) Graph 1. Heart rate of dogs (BPM) in the different housing conditions during the light-on period.
7
148
A. Klumpp et al. / Journal of Pharmacological and Toxicological Methods 54 (2006) 141–149
Table 2 Group II (males) Heart rate
No. of vocalizations
Average on day 1
Average on day 2 Day 1
Before noon In the afternoon In the evening At night Before noon Model I
Model II Model II Model II Model II
Housed alone in the old lab environment with individual runs and only visual contact to their usual run mate Housed alone in new lab without visual contact Housed alone in new lab with visual contact to all other dogs Run mates housed together in new lab with visual contact to others All 4 dogs together in new lab environment
70
65
53
53
6
19
48
68
66
54
57
14
15
58
77
82
65
64
24
23
55
70
69
61
61
5
6
61
70
65
56
65
9
8
No. of vocalizations
Average on day 1
Average on day 2 Day 1
Before noon In the afternoon In the evening At night Before noon
Model II Model II Model II Model II
Housed alone in the old lab environment with individual runs and only visual contact to their usual run mate Housed alone in new lab without visual contact Housed alone in new lab with visual contact to all other dogs Run mates housed together in new lab with visual contact to others All 4 dogs together in new lab environment
In the afternoon
52
SAP
Model I
Before noon
Before noon In the afternoon
139
129
117
115
121
6
19
131
132
118
123
132
14
15
137
140
127
126
128
24
23
136
134
122
123
128
5
6
139
143
128
132
137
9
8
used in experiments, every effort should be made to refine experiments so that they caused the minimum pain and distress. These guiding principles, the “3 Rs” of animal research, were initially given little attention. Progressively, however, they have become accepted as essential when animals are used in research (Flecknell, 2002). The three principles, of Replacement, Reduction and Refinement, have also proven to be an area of common ground for research workers since it has also been shown that the “3 Rs” can improve the quality of the science. Experiments should be carefully designed in order to reduce variation and to provide standardised optimum conditions of animal housing. Optimized conditions minimise unnecessary stress or pain and therefore produce better and more reliable data (Balls et al., 1995). The benefits of socialization with other dogs, especially long-term housing with the same run mate, has been demonstrated (Hubrecht, 1993). In other studies it was shown that stereotypic circling and barking was reduced and the dogs slept more and exhibited less stereotypic behaviour compared to single housed animals (Hetts et al, 1992). Moreover, good experimental design should include correct use of biostatistics. The appropriate number of animals necessitated for a given study type is crucial. A power calculation for a given study design should be included in every statistical evaluation (Markert, Klumpp, Trautmann, & Guth, 2004) since a lack of statistical power may require repeating the study. One should try to do reduce the variability of a given study design, since the power of a statistical
predictability is highly dependent on the intra-subject variation. 5. Conclusion Dogs acclimatize rapidly to a new lab environment. There is, however, an advantage in terms of hemodynamic parameter values and variability, when the dogs can be housed with their usual cage mates. This model should be preferred in respect to the refinement and improvement of the well-being of the dogs. References Balls, M., Goldberg, A. M., Fentem, J. H., Broadhead, C. L., Burch, R. L., Festing, M. F., et al. (1995). The three Rs: The way forward: The report and recommendations of ECVAM Workshop 11. ATLA Alternatives to Laboratory Animals, 23, 838–866. Flecknell, P. (2002). Replacement, reduction and refinement. ALTEX: Alternativen zu Tierexperimenten, 19, 73–78. Guth, B. D., Germeyer, S., Kolb, W., & Markert, M. (2004). Developing a strategy for the nonclinical assessment of proarrhythmic risk of pharmaceuticals due to prolonged ventricular repolarization. Journal of Pharmacological and Toxicological Methods, 49, 159–169. Hawkins, P., Morton, D. B., Bevan, R., Heath, K., Kirkwood, J., Pearce, P., et al. (2004). Husbandry refinements for rats, mice, dogs and nonhuman primates used in telemetry procedures: Seventh report of the BVAAWF/FRAME/RSPCA/UFAW Joint Working Group on Refinement, Part B. Laboratory Animals, 38(1), 1–10. Hetts, S., Clark, J. D., Calpin, J. P., Arnold, C. E., & Mateo, J. M. (1992). Influence of housing conditions on beagle behaviour. Applied Animal Behaviour Science, 34, 137–155.
A. Klumpp et al. / Journal of Pharmacological and Toxicological Methods 54 (2006) 141–149 Hubrecht, R. C. (1993). A comparison of social and environmental enrichment methods for laboratory housed dogs. Applied Animal Behaviour Science, 37, 345–361. Luft, J., & Bode, G. (2002). Integration of safety pharmacology endpoints into toxicology studies. Fundamental and Clinical Pharmacology, 16(2), 91–103 (Ref Type: Journal (Full)). Markert, M., Klumpp, A., Trautmann, T., & Guth, B. (2004). A novel propellant-free inhalation drug delivery system for cardiovascular drug safety evaluation in conscious dogs. Journal of Pharmacological and Toxicological Methods, 50, 109–119. Meyners, M., & Markert, M. (2004). Correcting the QT interval for changes in HR in pre-clinical drug development. Methods of Information in Medicine, 43, 445–450.
149
Purchase, I. F., Botham, P. A., Bruner, L. H., Flint, O. P., Frazier, J. M., & Stokes, W. S. (1998). Workshop overview: scientific and regulatory challenges for the reduction, refinement, and replacement of animals in toxicity testing. Toxicological Sciences, 43, 86–101. Russell, W. M. S., & Burch, R. L. (1959). The principles of humane experimental technique. The Principles of Humane Experimental Technique. Sales, G. D., Hubrecht, R., Peyvandi, A., Milligan, S., & Shield, B. C. (1997). Noise in dog kennelling: is barking a welfare problem for dogs? Applied Animal Behaviour Science, 52, 321–329. Senn, C. L., & Lewin, J. D. (1975). Barking dogs as an environmental problem. Journal of the American Veterinary Medical Association, 166, 1065–1068.
Original article
Optimizing the experimental environment for dog telemetry studies Anja Klumpp, Thomas Trautmann, Michael Markert ⁎, Brian Guth Department of Drug Discovery Support, General Pharmacology Group, Boehringer Ingelheim Pharma GmbH & Co. KG, Germany Received 2 February 2006; accepted 29 March 2006
Abstract Introduction: The objective of this study was to test the influence of housing conditions on hemodynamics during cardiovascular general pharmacological studies. Our goal was to optimize both the quality of the data through an optimization of the physiological conditions, as well as to ensure the dog's well-being in general pharmacological studies. Methods: Two groups of four dogs were equipped with radiotelemetry transmitters and continuously monitored in two different housing models. Model I consisted of 4 cages, two on each site of a corridor. Model II consisted of 4 cages positioned in a row, where the bordering cages were not separated with a metal plate. The physiological status of the dogs in the different housing models was based on the frequency of vocalizations and the average resting heart rate, as well as video monitoring. Results: The housing arrangement during the study had a remarkable effect on the hemodynamics measured. The hemodynamic parameters were best when the dogs were housed with their usual run mate. In this setting, they have impressively low average heart rates of about 60 bpm during the entire study, was associated with fewer vocalizations. Discussion: This study demonstrated that the quality of the acquired cardiovascular data for conscious dogs is dependent on the pen configuration and group make-up during a study. © 2006 Elsevier Inc. All rights reserved. Keywords: Methods; Refinement; Conscious dog; Telemetry; Housing conditions
1. Introduction The model preferred for cardiovascular studies according to the ICH S7a guideline is the conscious dog. The gold standard approach for studying conscious dogs is the use of a telemetry system to record cardiovascular parameters, i.e. blood pressure, left ventricular pressure, heart rate and ECG. The rationale for performing these studies in conscious dogs is based on the lack of possible influence of anesthesia on the parameters being measured. However, the conscious state is inherently responsive to environmental stimuli such that a higher spontaneous variability of the measured parameters must be taken into account. Thus, the quality of these types of cardiovascular studies is likely dependent on the degree of training of the animals used and the environment during a study. To show subtle drug-induced changes in heart rate, for example, it is essential to have a relatively constant heart rate without too
⁎ Corresponding author. Tel.: +49 7351 54 8727; fax: +49 7351 83 8727. E-mail address: [email protected] (M. Markert). 1056-8719/$ - see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.vascn.2006.03.010
many interruptions causing excitement (Guth, Germeyer, Kolb, & Markert et al., 2004). Furthermore, a stable heart rate is desirable to facilitate correct QT detection and heart rate correction (Meyners & Markert, 2004). An unstable heart rate therefore detracts from the detection of a possible prolongation of QT interval. The term refinement signifies the modification of any procedure, from the time a laboratory animal is born until its death, so as to minimize the pain and distress experienced by the animal and enhance its well-being. As mentioned by the members of the Joint Working Group on Refinement (Penny Hawkins et al., 2004), the ideal environment would allow for housing in stable, compatible groups. Optimizing animal welfare is not only important from the viewpoint of ethics; it is also a prerequisite for conducting good quality studies. The experience of pain and other stress will result in physiological changes (e.g. heart rate) that increase the variability of the experimental results and a loss of sensitivity. A high variability of heart rate, for example, is often caused by excited or stressed dogs. Thus, it is in the interest of good science to ensure that conditions regarding animal housing are optimized. Important lessons have been learned about how to
142
A. Klumpp et al. / Journal of Pharmacological and Toxicological Methods 54 (2006) 141–149
validate methods. The need to have predictive models available before validation is performed: the necessity to take the variability of the animal-based data which is to be used as the validation standard into account, and the need to have wellmanaged validation programs available (Purchase et al., 1998). Future progress will depend on the development of novel methods. The use of animals, especially dogs or monkeys, can only be justified, when clinically relevant data can be gained from few, but well-designed in vivo studies (Luft & Bode, 2002). This can only be assured when the dogs feel comfortable and aren't stressed which also reflects the housing conditions. Enrichment activities can include group housing when possible, rather than individual cages and use of materials similar to those found in their natural habitats. Once the study is in progress, it is important that the laboratory staff is well trained and competent in the handling of the species that is being used. One of the additional potential influences that may affect the hemodynamic status of the animals is the exact makeup of the group to be studied based on sex, family hierarchy and physical orientation to the other group members. It was the purpose of this study to assess the role of these factors on hemodynamic stability of the dogs, as an indicator of well-being and possible stress. 2. Methods 2.1. Animals Treatment of the animals followed the German Law on the Protection of Animals and was performed with permission from the State Animal Welfare Committee. Trained labrador dogs (male or female), at least 1.5 years of age, with body weights between 22 to 30 kg bred at Boehringer Ingelheim Pharma GmbH and Co. KG, Biberach were used. The dogs were group-housed as pairs of 2 in separate cages and had access to water ad libitum and were fed a standard dog diet once daily. They also had daily exercise periods of at least 1 h, each afternoon. To monitor the health of the animals, we collect blood samples every 3 months to determine the blood count and clinical parameters, including electrolytes and kidney parameters. The dogs were trained in the telemetry lab beginning 3 months before surgery, which was performed when the dogs were at least 1.5 years old. 2.2. Telemetry system The telemetry system used to measure cardiovascular parameters is manufactured by Konigsberg Instruments, Inc. (Pasadena, CA) and marketed by RMISS (Deleware). It consists of 5 major components: a) an implantable unit; b) a receiver (antenna) located in the animal's cage together with an amplifier; c) ambient pressure monitor to measure atmospheric pressure; d) a PC-based “base station” to receive and process the amplified signals; e) a PC-based data acquisition system (NOTOCORD Hem 3.5) to process signals. The implantable unit (“T27” total implant) consists of (1) two high fidelity pressure transducers (5.0 and 4.0 mm
diameter), (2) an ECG cable, (3) micro-power battery-operated electronics that process and digitize the information from the pressure transducers and the ECG lead, (4) a radio-frequency transmitter that sends the signals to the telemetry receiver, and (5) a battery. A small cable projecting from the transmitter contains a magnetic switch that allows the device to be turned on and off. Prior to implantation the zero value of the two pressure transducers are calibrated using a manometer and 250 mm Hg is set to 5 V. 2.3. One-time surgical implantation The transducers of the T27 implant were calibrated and the unit was sterilized using a low pressure ethylene oxide process prior to implantation. Dogs were anaesthetised with a combination of Rompun (xylazine hydrochloride, 1 ml/10 kg, i.v.) and Ketavet (ketamine hydrochloride, 0.7 ml/10 kg, i.v.) after premedication with Atropine sulfate (0.04 mg/kg i.m.) and ventilated with 66% O2 and 1%–1.5% isofluorane at a ventilation rate of 14/min. All procedures were performed under aseptic conditions using sterilized equipment. The dogs were placed in a lateral recumbency with the left side facing the surgeon. An incision was made between the fifth and sixth intercostal space. A small pocket was opened in the abdominal wall for implantation of the transmitter, battery housing, and induction switch coil. The cables with both pressure transducers and ECG leads extending from the ventrally implanted transmitter were guided subcutaneously to the lateral incision. The antenna was guided subcutaneously from the transmitter location towards the spine and then runs parallel to the spine for ∼25 cm. The initial ventral incisions required for battery and transmitter placement were closed. A left thoracotomy was performed between the fifth and sixth intercostal space to expose the left ventricle apex for insertion of the left ventricular Konigsberg transducer. The aorta pressure transducer was implanted next. The aortic transducer, which also served as one electrode of the ECG, was inserted into the thoracic aorta just below the aortic arch. The transducer was sutured into place and blood flow was restored. The lung was then inflated and the intercostal muscles were sutured closed and the pneumothorax evacuated. Chest incisions were closed. The gas anaesthesia was then turned off and dogs were allowed to wake up. Analgesics and antibiotics (Temgesic and Duphamox® LA, Amoxicillin-Trihydrat) were administered for 2 days (Temgesic) and 10 days (Duphamox) following the procedure to support a good recovery and to ensure that the dog has no postoperative pain, a transdermal plaster (fentanyl, 25 μg/h) was placed on the skin for 2 days. Dogs are allowed to recover for at least 21 days before experiments using test substances are initiated. 2.4. Runs in the telemetry lab Dogs used for studies in which the hemodynamic and ECG data are collected telemetrically are normally group-housed at
A. Klumpp et al. / Journal of Pharmacological and Toxicological Methods 54 (2006) 141–149
the company's animal's resource facility under standardized housing conditions that meet or exceed regulatory requirements. For conducting studies, however, the animals are transported to the telemetry laboratory having special rooms for the animals for the duration of the study and equipped with the appropriate antennae for collecting the transmitted signals as well as video monitoring. The dogs are trained over months in these runs and are adapted to them as secondary “home cages”. Two sets of 4 animal runs each are available in the telemetry lab. Whereas many aspects of these 2 areas are identical (see below), they do differ in subtle ways which might impact on study outcome. It is the comparison of these subtle differences in term of their affect on the hemodynamic status of the dogs which form the basis of this study. Common features: Both sets of runs provide each animal 4 m2 space and a cage height of 3 m. Room temperature and humidity is controlled (22 ± 2 °C; 60 ± 15% humidity) with a ventilation turnover of 13/hr (11/h is the required). They maintain a 12 h light (365 lx) and dark (3 lx, “moonlight”) cycle. Background “noise” is supplied during the lighted period in the form of a radio. Independent features: The physical layout of the 4 runs is different in the two telemetry labs. In the first (Model I), depicted schematically in Fig. 1, there are 2 cages on each side of a central corridor. Dogs in adjacent cages do not have visual contact due to a presence of a metal plate, but can see the dogs in the 2 opposing runs.
Cage 4
Cage 3
Cage 2
Fig. 2. Model II.
The frequencies of our dogs used in the studies are: 209.8 MHz, 200.9 MHz, 197.9 MHz, 178.0 MHz, 210.90 MHz, 205.99 MHz, 189.6 MHz, and 206.9 MHz. We used a specially designed floor antenna (Fig. 3) to improve the quality of the transmitted signals in that all 4 animals could be measured together as depicted schematically in Fig. 2. The diameters of the receiving antenna loops were calculated according to the following equation: k¼
c 240 000 m=s ¼ ¼ 1:2 m f 200 MHz
where c is the spreading velocity of the transmitted signal and f is the transmitted frequency (average over all frequencies used in the lab). The other factor examined in this study was the group makeup of the dogs used. For this, we compared a group of 4 males with a group of 4 females. It should be noted that an analysis of
Cage 4
FLOOR
Cage 1
Cage 2
Fig. 1. Model I.
Cage 1
FLOOR
The second telemetry lab has the 4 cages arranged in a row (Model II) (Fig. 2). In contrast to the other lab, the dividers between the cages allow visual contact and can be removed as needed to allow group housing, either as established pairs or all 4 together. As the telemetry device used for this study uses different transmitting frequencies, the animals could be studied when held together.
Cage 3
143
Fig. 3. Special design of the floor antenna.
144
A. Klumpp et al. / Journal of Pharmacological and Toxicological Methods 54 (2006) 141–149
Fig. 4. Housed alone in the old lab environment with individual runs and only visual contact to their usual run mate.
statistical power has shown that using a cross-over experimental design, a sample size of 4 gives a 90% power to detect a 10% change in the hemodynamic parameters measured. 2.5. Experimental design Two groups of dogs were used: Group I: consisting of 4 females Group II: consisting of 4 males.
Four variations of the run design in Model II were compared: 1. Housed alone without visual contact (Fig. 6) 2. Housed alone with visual contact to all other dogs (Fig. 7) 3. Run mates housed together with visual contact to others (Fig. 8) 4. All 4 dogs together (Fig. 9). The following parameters were used for comparison the physiological status in the different housing models:
Each group was studied in the 2 different housing models: Housing Model I Model I consists of 4 cages, two on each site of a corridor. In this model each dog has only visual contact to the dogs sitting opposite, because the bordering cages are separated by a metal plate (Figs. 4 and 5). Housing Model II Model II consists of 4 cages positioned in a row, whereas the bordering cages are not separated with a metal plate and additionally the floor was designed (for the dogs) with a rough surface and that is covered with wood shavings. In this model, the partition walls could also be removed, giving us the possibility of a new run design, in which dogs could be studied while either individually, or group-housed.
1. Time from first contact to the pen to baseline heart rate (internally defined as a HR = <60 bpm). 2. Mean (± S.E.M.) HR during the 22-h observation period. 3. Time after placebo treatment to baseline heart rate (internally defined as a HR = < 60 bpm). 4. Number of vocalizations during daytime measurements 5. Number of movement vs. sleeping phases. All studies used the same protocol. The dogs were placed into the cages at 7:00 a.m. and the telemetric implant was switched on. After a calming phase of 30 min the signal offset was performed (7:30 a.m.). Data were then collected for 2 h (7:30 a.m. – 9:30 a.m.) and is referred to as the “initial control”.
Fig. 5. Housed alone in the old lab environment with individual runs and only visual contact to their usual run mate.
A. Klumpp et al. / Journal of Pharmacological and Toxicological Methods 54 (2006) 141–149
145
Fig. 6. Housed alone in new lab without visual contact.
To simulate a drug application, the lab was then entered by a technician for a few minutes (9:30 a.m.). Data collection was then continued for the next 22 h. At 4:30 p.m. the dogs were fed with continuous collection of hemodynamic data.
Student's t-test. Statistical significance was accepted when p < 0.05. Using the error term of the analysis of variance as an estimate for the variability many-to-one t-tests comparing the dose groups of the test compound with the control group were added (Table 1). The evaluation was performed using the software package SAS 8.02 (SAS Institute Inc., Cary, USA).
2.6. Data acquisition and analysis 3. Results Digitized telemetry signals were processed by NOTOCORD (Notocord, Paris) software (Hem 3.5) on a beat-to-beat basis. The following parameters were continuously recorded for the duration of the experiment: aortic pressure (AP), left ventricular pressure (LVP), ECG lead II and body temperature. The hemodynamic and ECG parameters were calculated during the experiment and include: systolic, diastolic and mean aortic pressure, peak systolic and end-diastolic left ventricular pressure, LV dP/dt max and dP/dt min, heart rate; PQ-, QRSand QT-intervals. NOTOCORD software was used for acquisition of data whereas EXCEL was used for some basic data analyses. Data were summarized at predefined time points by calculating median values ± S.D. At each time point (every 10 min), a minimum of 400 sequential beats were used to calculate the median value. The summarized data are given as mean values ± SEM. The different base levels of the individual animals were taken into account by referring the measured values after administration of the test compound to the pre-treatment values. With these values, the standardised area under the curve (AUC divided by interval length) was calculated for the time intervals of interest. Comparisons between treatment and placebo were performed by one-way analysis of variance (ANOVA) followed by the
3.1. Model I In Housing Model I, although animals calmed down fast after disturbances (Graph 1), higher heart rates (70 bpm) and longer periods of vocalization (47 times per day, Table 1) indicated the restlessness of the dogs. This appeared to be more pronounced in females (group I). Females had a relatively high average heart rate of 75 bpm, higher variability between the subjects and more peaks where the dog's heart rate was increased during vocalization (Graph 1, circles). 3.2. Model II The dogs in Housing Model II were, in general, calmer than in Model I. Remarkable is the fact that they had very low heart rates. The dogs had heart rates between 50 and 77 bpm throughout the entire study, except when the “placebo application” was performed at which time the heart rate increased to 120 to 140 bpm. It reached ”pre-dose” values within few minutes, however (Table 1). In Housing Model II the barking periods were shorter (Table 1) and not as frequent (15 times per day). However, it has to be
Fig. 7. Housed alone in new lab with visual contact to all other dogs.
146
A. Klumpp et al. / Journal of Pharmacological and Toxicological Methods 54 (2006) 141–149
Fig. 8. Run mates housed together in new lab with visual contact to others.
mentioned that a difference between females and males was found (see below). 3.2.1. Group I (female) The females were very calm in all four variations of Model II, but especially when they were housed alone without visual contact where they almost never barked. In all variations, however, they behaved very calmly indicated by stable hemodynamic parameters throughout the study (Graph 1). Subjectively, however, the dogs appeared more comfortable when housed with their usual run mate (Table 1). 3.2.2. Group II (males) The behaviour of the males in the various run designs of the Housing Model II was very different compared to the females. They were calm, when they were housed alone without visual contact or when housed with their run mate with visual contact to all other dogs. The dogs were less calm when they were housed alone with visual contact or all 4 dogs together. In this case the males were very restless and they barked frequently. When they were housed alone with visual contact to all other dogs, instances of fighting were noted through the fence, possibly to determine their hierarchy. In contrast to the females, they were also very restless when they were kept all 4 together. Although they had the chance to determine their hierarchy, they were not calm. As a result, the measured parameters showed moderate fluctuations (Table 2). The best experimental environment seems to be when run mates were housed together with visual contact to others. In this environment they have the possibility to maintain their established hierarchy. After the hierarchy is determined, they
are calm. In this setting, they have impressively low average heart rates of about 50–60 bpm (Table 2). When drug application is simulated, the increase in heart rate is very moderate compared to when they are single-housed and the heart rate returns to pre-dose values much faster again. There is almost no inter-subject variability detectable. 4. Discussion This study demonstrated that the quality of the acquired cardiovascular data using conscious dogs is dependent on the pen configuration and group make-up during a study. To show drug-induced changes in heart rate, or other hemodynamic parameters, it is advantageous to have a stable heart rate with little spontaneous variability during the study (Guth et al., 2004). This can only be achieved when the dogs are unstressed, which reflects the holding conditions. A low frequency of vocalization episodes, which could be shown clearly in this study, is an additional marker of optimized housing conditions (Senn & Lewin, 1975). Additionally, less barking is thought to increase the welfare of dogs (Sales et al., 1997) since the dogs can produce barks of well over 100 dB. Based on the results of the study we conclude that the housing configuration during a study can have a substantial impact on the data collected. This is both in terms of the absolute values observed as well as on the variability over time. Housing animals during a study with their usual housing mate appears ideal. Furthermore we conclude that females were consistently calmer in this type of study demonstrating lower heart rates and less excitability. Use of females may therefore offer advantages to detect drug-induced hemodynamic effects.
Fig. 9. All 4 dogs together in new lab environment.
A. Klumpp et al. / Journal of Pharmacological and Toxicological Methods 54 (2006) 141–149
147
Table 1 Group I (females) Heart rate
No. of vocalizations
Average on day 1
Model I
Model II Model II Model II Model II
Housed alone in the old lab environment with individual runs and only visual contact to their usual run mate Housed alone in new lab without visual contact Housed alone in new lab with visual contact to all other dogs Run mates housed together in new lab with visual contact to others All 4 dogs together in new lab environment
Average on day 2 Day 1
Before noon In the afternoon In the evening At night Before noon
Before noon In the afternoon
73
71
55
49
51
15
32
50
60
57
52
55
5
5
55
62
59
52
55
11
12
60
68
64
60
58
9
7
55
67
60
52
56
8
7
SAP
No. of vocalizations
Average on day 1
Model I
Model II Model II Model II Model II
Housed alone in the old lab environment with individual runs and only visual contact to their usual run mate Housed alone in new lab without visual contact Housed alone in new lab with visual contact to all other dogs Run mates housed together in new lab with visual contact to others All 4 dogs together in new lab environment
Average on day 2 Day 1
Before noon In the afternoon In the evening At night Before noon
Before noon In the afternoon
128
124
119
115
117
15
32
125
119
114
108
122
5
5
122
117
109
104
113
11
12
118
118
116
118
118
9
7
127
120
118
120
111
8
7
This might offer a further benefit since some studies have shown that female gender is more susceptible to drug-induced QT-prolongation.
In 1959, William Russell and Rex Burch published “The Principles of Humane Experimental Technique” (Russell & Burch, 1959). They recommended that if animals were to be
GROUP I 140 Model I housed alone with visual contact to the dogs sitting opposite Model II housed alone without visual contact Model II housed alone with visual contact to all other dogs Model II run mates housed together with visual contact to others Model II all 4 dogs together
130 120
heart rate (min-1)
110 100 90 80 70 60 50 40
Median of 10 min 30 -1
0
1
2
3
4
5
6
time (hours) Graph 1. Heart rate of dogs (BPM) in the different housing conditions during the light-on period.
7
148
A. Klumpp et al. / Journal of Pharmacological and Toxicological Methods 54 (2006) 141–149
Table 2 Group II (males) Heart rate
No. of vocalizations
Average on day 1
Average on day 2 Day 1
Before noon In the afternoon In the evening At night Before noon Model I
Model II Model II Model II Model II
Housed alone in the old lab environment with individual runs and only visual contact to their usual run mate Housed alone in new lab without visual contact Housed alone in new lab with visual contact to all other dogs Run mates housed together in new lab with visual contact to others All 4 dogs together in new lab environment
70
65
53
53
6
19
48
68
66
54
57
14
15
58
77
82
65
64
24
23
55
70
69
61
61
5
6
61
70
65
56
65
9
8
No. of vocalizations
Average on day 1
Average on day 2 Day 1
Before noon In the afternoon In the evening At night Before noon
Model II Model II Model II Model II
Housed alone in the old lab environment with individual runs and only visual contact to their usual run mate Housed alone in new lab without visual contact Housed alone in new lab with visual contact to all other dogs Run mates housed together in new lab with visual contact to others All 4 dogs together in new lab environment
In the afternoon
52
SAP
Model I
Before noon
Before noon In the afternoon
139
129
117
115
121
6
19
131
132
118
123
132
14
15
137
140
127
126
128
24
23
136
134
122
123
128
5
6
139
143
128
132
137
9
8
used in experiments, every effort should be made to refine experiments so that they caused the minimum pain and distress. These guiding principles, the “3 Rs” of animal research, were initially given little attention. Progressively, however, they have become accepted as essential when animals are used in research (Flecknell, 2002). The three principles, of Replacement, Reduction and Refinement, have also proven to be an area of common ground for research workers since it has also been shown that the “3 Rs” can improve the quality of the science. Experiments should be carefully designed in order to reduce variation and to provide standardised optimum conditions of animal housing. Optimized conditions minimise unnecessary stress or pain and therefore produce better and more reliable data (Balls et al., 1995). The benefits of socialization with other dogs, especially long-term housing with the same run mate, has been demonstrated (Hubrecht, 1993). In other studies it was shown that stereotypic circling and barking was reduced and the dogs slept more and exhibited less stereotypic behaviour compared to single housed animals (Hetts et al, 1992). Moreover, good experimental design should include correct use of biostatistics. The appropriate number of animals necessitated for a given study type is crucial. A power calculation for a given study design should be included in every statistical evaluation (Markert, Klumpp, Trautmann, & Guth, 2004) since a lack of statistical power may require repeating the study. One should try to do reduce the variability of a given study design, since the power of a statistical
predictability is highly dependent on the intra-subject variation. 5. Conclusion Dogs acclimatize rapidly to a new lab environment. There is, however, an advantage in terms of hemodynamic parameter values and variability, when the dogs can be housed with their usual cage mates. This model should be preferred in respect to the refinement and improvement of the well-being of the dogs. References Balls, M., Goldberg, A. M., Fentem, J. H., Broadhead, C. L., Burch, R. L., Festing, M. F., et al. (1995). The three Rs: The way forward: The report and recommendations of ECVAM Workshop 11. ATLA Alternatives to Laboratory Animals, 23, 838–866. Flecknell, P. (2002). Replacement, reduction and refinement. ALTEX: Alternativen zu Tierexperimenten, 19, 73–78. Guth, B. D., Germeyer, S., Kolb, W., & Markert, M. (2004). Developing a strategy for the nonclinical assessment of proarrhythmic risk of pharmaceuticals due to prolonged ventricular repolarization. Journal of Pharmacological and Toxicological Methods, 49, 159–169. Hawkins, P., Morton, D. B., Bevan, R., Heath, K., Kirkwood, J., Pearce, P., et al. (2004). Husbandry refinements for rats, mice, dogs and nonhuman primates used in telemetry procedures: Seventh report of the BVAAWF/FRAME/RSPCA/UFAW Joint Working Group on Refinement, Part B. Laboratory Animals, 38(1), 1–10. Hetts, S., Clark, J. D., Calpin, J. P., Arnold, C. E., & Mateo, J. M. (1992). Influence of housing conditions on beagle behaviour. Applied Animal Behaviour Science, 34, 137–155.
A. Klumpp et al. / Journal of Pharmacological and Toxicological Methods 54 (2006) 141–149 Hubrecht, R. C. (1993). A comparison of social and environmental enrichment methods for laboratory housed dogs. Applied Animal Behaviour Science, 37, 345–361. Luft, J., & Bode, G. (2002). Integration of safety pharmacology endpoints into toxicology studies. Fundamental and Clinical Pharmacology, 16(2), 91–103 (Ref Type: Journal (Full)). Markert, M., Klumpp, A., Trautmann, T., & Guth, B. (2004). A novel propellant-free inhalation drug delivery system for cardiovascular drug safety evaluation in conscious dogs. Journal of Pharmacological and Toxicological Methods, 50, 109–119. Meyners, M., & Markert, M. (2004). Correcting the QT interval for changes in HR in pre-clinical drug development. Methods of Information in Medicine, 43, 445–450.
149
Purchase, I. F., Botham, P. A., Bruner, L. H., Flint, O. P., Frazier, J. M., & Stokes, W. S. (1998). Workshop overview: scientific and regulatory challenges for the reduction, refinement, and replacement of animals in toxicity testing. Toxicological Sciences, 43, 86–101. Russell, W. M. S., & Burch, R. L. (1959). The principles of humane experimental technique. The Principles of Humane Experimental Technique. Sales, G. D., Hubrecht, R., Peyvandi, A., Milligan, S., & Shield, B. C. (1997). Noise in dog kennelling: is barking a welfare problem for dogs? Applied Animal Behaviour Science, 52, 321–329. Senn, C. L., & Lewin, J. D. (1975). Barking dogs as an environmental problem. Journal of the American Veterinary Medical Association, 166, 1065–1068.