Improving the Lives of Laboratory Animals

WHAT'S THE DOSE, DOC?

Last November the Institutional Animal Care and Use Committee at the University of Washington, Seattle, suspended immunologist Chen Dong from laboratory animal research. Among the irregularities cited: snipping the tails of mice without anesthesia. Tail snipping is a common practice; foregoing anesthesia is not.

Dominick J. Wells of the Faculty of Medicine, Imperial College London, who studies the effects of tail biopsy on laboratory mice, says he doubts Dong's is an isolated case. He notes that molecular biologists did not use anesthesia in the 1980s when they first began snipping tails to collect DNA samples for genotypic analysis.

Yet contrary to popular belief, says Wells, the practice is harmful, no matter how small the excision, because even two-week-old mice have bone growth in the tail. Possible alternatives include a tiny ear snip or a mouth swab, but mice fight the latter, Wells reports, and there's no proof that ear snips are less painful. "We need to do more research to determine what is going on as opposed to what we think is going on," he says.

Alicia Z. Karas, assistant professor of clinical sciences, Tufts University School of Veterinary Medicine in Boston, says withholding pain relief is common after surgery. Investigators either do not recognize when an animal is in pain, she says, or dismiss the effectiveness of analgesia as unproven in mice.

Data do exist, however. Pain researcher Jeffrey S. Mogil, McGill University, Montreal, has conducted extensive studies of the relative effects of analgesia in mice. "I am literally the only person in the world who would collect such data. If I haven't gotten to [an analgesic], no one else has," says Mogil, who has determined the responses of 12 mouse strains to acetaminophen. Even on his own campus, he says, veterinarians are largely unaware of his work. Mogil's explanation: His papers appear not in veterinary publications but usually in the journal Pain.

The fundamental principles for care of laboratory animals were spelled out in a seminal book, The Principles of Humane Experimental Technique.¹ Advocating the "three Rs" for humane treatment of laboratory animals – replacement, reduction, and refinement – zoologist William Russell and microbiologist Rex Burch advised researchers to avoid animal testing wherever possible, minimize the number of animals tested, and refine procedures to minimize suffering.

To some extent, researchers have taken note. Testing of nonhuman primates has declined considerably in recent years, as has work with dogs, cats, rabbits, and hamsters.² Mice numbers have increased, however, due to the expanding use of transgenic animals, and fish are almost as prevalent as mice in Canadian research.

According to a report by Clément Gauthier, executive director of the Canadian Council on Animal Care (CCAC), pain and distress are declining in lab animals.² Since 1996 the percentage of all animals used in experiments that cause pain, without resort to anesthesia, analgesics, or tranquilizers, has fallen from 12% to 9% in the United States and from 5% to 3% in Canada.

But is the pain relief adequate for those animals that receive it? Penny Hawkins of the Research Animals Department of the UK's Royal Society for Prevention of Cruelty to Animals surveyed 137 scientists, veterinarians, and animal technicians on pain management at 28 laboratories from June 1999 to April 2001. No one refused to participate, and virtually all were sensitive to the possibility of pain and suffering resulting from their experiments, according to her report.³

Pain relief practices were reported as considerably varied. Only five facilities had written policies requiring routine administration of postoperative analgesia. At the other facilities pain relief was standard, but with exceptions. Three laboratories did not give analgesia to rats being fitted with headpieces. At another, mice received no pain relief before or after embryo transfer. One researcher did not provide analgesia to frogs having oocytes removed surgically, because he could not find the correct dose in the literature.

"I believe that it is very hard, if not impossible, to say how much laboratory animals suffer as a whole ... their suffering is likely to be underestimated, often, ironically, because people do not want to think that they are causing animals pain or distress," says Hawkins.

FEELING NO PAIN

Courtesy of Sue VandeWoude

Female Swiss Webster laboratory mice provided with three types of structural environmental enrichment.

Attention to pain management is a relatively recent phenomenon, according to Paul A. Flecknell at the University of Newcastle, UK. He credits animal welfare groups but also speculates that full-color advertisements for animal analgesia played a role. "It's difficult to know why things have changed, but they certainly have," he says "When I started working with lab animals in the late seventies, there wasn't a great deal of interest."

Veterinarian Sue VandeWoude and her colleagues at Colorado State University in Fort Collins recently published species-specific pain scales for canines and large animals after orthopedic surgery, for rabbits and birds after humeral orthopedic surgery, and for rats in an arthritis study.⁴ Animals do show pain, "but you need to know what to look for," she says. "A lot of people aren't as in tune with some of the rodent species as they should be."

Yet even when investigators are sensitive to animal suffering, pain can be difficult to discern. Evolution has taught many small animals to hide pain from predators. Many freeze when people approach them. Some keep still during the day because they are nocturnal. Flecknell advises skeptics to fix a video camera on their animal cages at night and observe carefully. His lab's extensive research in rats, for example, shows that they will arch their backs like cats or tense their stomach muscles after abdominal surgery. "We think these behaviors mean the rat is in pain," he says, especially since pain relief medication stops those behaviors.

Even fish are not immune. Though a US researcher concluded in 2002 that fish lack the brain structures or functional equivalents for feeling pain,⁵ British investigators have found nociceptors in the lips of rainbow trout. According to their studies, injection of bee venom or acetic acid causes trout to avoid eating and to act as if they perceive pain.⁶ Consequently, Gilly Griffin, the CCAC's director of guidelines development, reports that draft guidelines refer to the "potential to experience pain" and oblige investigators "to mitigate or minimize potential pain and distress whenever feasible, consistent with good scientific practice."⁷

Courtesy of Kay Stewart

Hard plastic baby rattles are placed in rabbit cages to provide the animals an opportunity for exploration and play. The rabbits toss and push the rattles around the cage.

For many species, good scientific practice is difficult to define. VandeWoude says her institution has an aggressive policy requiring administration of analgesia and euthanasia at early time points. An investigator may request an exception, however, on grounds that analgesia would interfere with scientific outcomes. "A lot of pain relievers we use are opioids, which are not inconsequential in physiological effects," says VandeWoude.

When investigators think through the possible consequences of analgesics, they often drop their objections, Flecknell says. "Scientists say 'I don't want to give an analgesic because it will upset my study,"' he says. "But unalleviated pain will cause a whole lot of physiological and pathological changes. Giving analgesics may give you more accurate results."

Transgenic mice are another growing concern, as unintended consequences of genetic manipulation can alter physiology and change behaviors. Marilyn Brown, director of animal welfare and training at Charles River Laboratories, a leading animal breeder headquartered in Wilmington, Mass., says its staff is trained to look for unexpected changes such as weight loss or skin ulceration. Company policy calls for consultation with veterinarians and the researcher for whom the animal has been bred. Often, Brown says, the animal can be made comfortable, but not always. An animal that cannot eat because of malformed teeth would probably be euthanized.

PROBLEMS WITH ENRICHMENT

Enrichment is also problematic in a breeding setting. Male mice are famously aggressive toward each other, and housing them together can lead to brutal battles. Whether adding enrichment in the form of more hiding places would decrease fighting or make the animals more territorial is an open question, says Brown, so the company uses standard housing and separates fighting animals.

Courtesy of Paul Flecknell & J. V. Roughan

Some researchers have trouble recognizing pain in their laboratory subjects. But back-arching has been shown to be a sign of pain after abdominal surgery in rats.

Cohabitating rabbits can be vicious, too. Timo Nevalainen of the Universities of Kuopio and Helsinki in Finland called off an experiment in paired housing for rabbits after 140 days because of fighting. Yet on the plus side, he observed that the paired animals varied less from each other in growth measures than singly housed rabbits. If the fighting could be overcome, paired housing might lower the number of animals needed to prove a statistical endpoint.

Not all animals should be kept from doing what comes naturally. Rats, for instance, establish social order through combat, says William Greenough, who favors enrichment but says it must be better defined and its consequences, both psychological and physical, better understood. Rats in enriched environments have lower body weight, longer bones, more synapses per neuronal cell, increased capillary capacity, and other changes that could have profound experimental repercussions, he says.

Greenough notes that Richard J. Smeyne at St. Jude Children's Research Hospital in Memphis, Tenn., found that a strain of C57Bl/6 mice bred for sensitivity to the toxin MPTP lost that trait when they were raised in enriched housing. Smeyne says he used a 1 x 1 m cage with mazes, exercise wheels, and 14 companion animals instead of a standard 30.5 x 20 cm, five-mouse cage.

Nevertheless, neither Smeyne nor Greenough is ready to abandon enrichment. To the contrary, Smeyne speculates that animals in enriched environments might be more like normal humans than animals used for drug research. Recalling how deprivation affected the physical and mental states of children in Rumanian orphanages, he questions whether animals raised in standard environments are dysfunctional and, therefore, poor experimental models. "Maybe that's what we're testing our drugs on," he says. "Maybe we have to do [testing] in an enriched environment."

For Timo Nevalainen, the answers will come with caution and common sense. "The question is not whether ... to enrich the environment. The question is how," says Nevalainen, past president of the Federation of European Laboratory Animal Science Associations. "We have to make sure a [change] is really an improvement. Then we have to make sure it does not interfere. ... We always have to guard the integrity of the science."

Maintaining that balance will surely factor into the debate when animal care-givers convene later this year in Nantes, France, for the International Council for Laboratory Animal Science (ICLAS) Meeting for Harmonization of Guidelines. On June 13 and 14, meeting participants will consider the possibility of international recognition for standards of care. Says ICLAS president Gilles Demers, "We are trying to assure that everywhere in the world people doing science are doing good science according to good animal practice."

Jane Salodof MacNeil macneiljs@earthlink.net is a freelance writer in Groveland, Mass.