6
Animal-Welfare Issues Related to the Ascites Method for Producing Monoclonal Antibodies

Availability of Data

Few published data are available on animal-welfare issues related to the ascites method for producing mAb. Most references simply state that animal-welfare problems are related to this method but do not provide supporting data. For instance, a report on mAb production published by the European Center for the Validation of Alternative Methods (ECVAM) (Marx and others 1997) states that "the main disadvantage of the ascites method is that it is extremely painful for the animals used, due to the following: a) the injection of primer; b) the resulting peritonitis caused by the primer; c) abdominal tension; d) the invasive tumors which result." Such statements are generally not supported in the literature. Most of the cited papers contain scant data and tend to be circular and to cite each other or to cite a review of the earlier literature by McGuill and Rowan (1989) or a short paper by Amyx (1987) that contains only opinion and no data.

Almost 10 years ago, McGuill and Rowan (1989) concluded their article on the refinement of mAb production and animal well-being with a statement that "research is necessary to provide a better understanding of the stresses endured by animals under protocols designed to maximize production of mAb. The effects of large intraperitoneal tumor masses, of frequent and large-volume paracentesis, and of pristane priming must be studied to provide a scientific basis for developing procedures that minimize animal distress and suffering." Very little has been done to develop such data since then.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 33
6 Animal-Welfare Issues Related to the Ascites Method for Producing Monoclonal Antibodies Availability of Data Few published data are available on animal-welfare issues related to the ascites method for producing mAb. Most references simply state that animal-welfare problems are related to this method but do not provide supporting data. For instance, a report on mAb production published by the European Center for the Validation of Alternative Methods (ECVAM) (Marx and others 1997) states that "the main disadvantage of the ascites method is that it is extremely painful for the animals used, due to the following: a) the injection of primer; b) the resulting peritonitis caused by the primer; c) abdominal tension; d) the invasive tumors which result." Such statements are generally not supported in the literature. Most of the cited papers contain scant data and tend to be circular and to cite each other or to cite a review of the earlier literature by McGuill and Rowan (1989) or a short paper by Amyx (1987) that contains only opinion and no data. Almost 10 years ago, McGuill and Rowan (1989) concluded their article on the refinement of mAb production and animal well-being with a statement that "research is necessary to provide a better understanding of the stresses endured by animals under protocols designed to maximize production of mAb. The effects of large intraperitoneal tumor masses, of frequent and large-volume paracentesis, and of pristane priming must be studied to provide a scientific basis for developing procedures that minimize animal distress and suffering." Very little has been done to develop such data since then.

OCR for page 33
Animal Welfare Issues Related To Ascites Method Some cell lines, especially unoptimized or aggressive cell lines, produce signs in mice that can be interpreted as indicating pain, but it is not clear from the literature that ascites production itself is necessarily associated with pain. Gebhart (personal communication) evaluated pain in ascites production and found that possible pain associated with intraperitoneal administration of pristane in mice is no more significant than that observed with the administration of intraperitoneal saline. In humans, intraperitoneal administration of drugs is generally perceived as mildly uncomfortable but not distressful or painful. In clinical experience in humans, abdominal tumors are rarely painful unless they invade the intestinal organs; and ascites fluid accumulation is not painful, although a large accumulation can cause distress and discomfort because it interferes with respiration. When ascites fluid accumulation does become uncomfortable, tapping of the fluid is perceived as a welcome relief-not painful. Therefore, in mice it is probably more important to perform needle taps frequently to avoid ascites fluid accumulation sufficient to cause distress. However, some cell lines produce clinical signs in mice indicating distress, including anorexia, rapid breathing, hunched posture, hypothermia, and decreased activity. Those signs can be observed in any sick mouse, regardless of the cause, and are usually seen with poor mAb-producing, more-aggressive cell lines. (It is important to note that this depends on the cell line.) Also, cell lines can produce various pathologic changes in mice that vary with the cell line and may be associated with pain or distress (Jackson and others 1999a). Methods For Measuring Pain Or Distress In Laboratory Rodents Clearly, data are required if one is to determine objectively whether the ascites method for producing mAb causes pain or distress in mice, but it is unclear what such data would have to be. There is no reliable scientific measurement of pain or distress in rodents. Although that statement is true with regard to our ability to know with certainty the levels of pain or distress that an animal is experiencing, there is some information on what might or might not cause pain or distress (NRC 1992). Current guidelines suggest that if a procedure might be expected to cause pain in humans, we should expect it to cause pain in animals. Very little other guidance is given, and, as stated above, there are no broadly accepted techniques for assessing the degrees of pain or distress in animals. In most instances, the standard approach is to assume that an animal is in distress if it is experiencing significant weight loss (such as a weight loss of greater than 15 or 20%) or if it stops taking food and water. Those measures might be useful as indicators of severe distress, but there is a shortage of data on weight gain and

OCR for page 33
loss in rodents that are undergoing different procedures. In any case, weight loss is not useful for assessing mice with ascitic fluid, in that they all gain weight regardless of degree of distress. More subtle behavioral indices are needed to provide an indication whether the animals are experiencing mild to moderate pain or distress. Such subtle methods might include observing an animal under red light. Rodents are generally more active under red light, and declines in activity will be easier to identify than when the animals are relatively quiescent under bright light. Another possibility might be the use of a formal behavioral index, such as the disturbance index of Barclay and others (1988), which permits measurement of relatively subtle changes in behavior that appear to be related to an animal's level of disturbance. In developing the disturbance index, the investigators noted that departures from normal behavior could be produced by relatively minor procedures. They assumed that the extent of behavioral change is related directly to the severity of pain or distress. They found that the amount of exploratory behavior of a rat or a mouse placed in a strange cage was very predictable. Increases or decreases in an animal's exploratory behavior during a period of 1 1/2–3 hours after a test treatment could be converted into a disturbance index that quantified the departure from normal exploratory behavior. In one of the tests conducted on the animals, the investigators found that intraperitoneal injections of mineral oil resulted in a statistically significant increase in the disturbance index after 6 days (Barclay and others 1988). No further work on the disturbance index has been done, and the index has not been correlated with other signs of pain or distress, such as physiologic or clinical changes. Therefore, it remains unclear whether the information gathered with the index can provide a means of classifying mild, moderate, and severe pain or distress. In recent years, a few investigators, such as David Morton (University of Birmingham, UK) have begun to explore other approaches to assessing the distress experienced by laboratory animals. In 1985, Morton and Griffith published a paper outlining how one might use behavioral and clinical signs to assess animal distress. The ideas in the paper have been applied in a number of cases and can be reduced to five types of observations: At-a-distance-observations: Behavior Appearance Observations requiring handling: Body weight, temperature, and so on Clinical signs Provoked behavior Morton has promoted the use of score sheets individualized for specific experimental procedures or projects. Preliminary data from those score sheets

OCR for page 33
and sheets developed by other investigators (reported at the recent Conference on Humane Endpoints in Zeist, the Netherlands, November 23–25, 1998) are beginning to provide stronger evidence of the experience of pain or distress in laboratory rodents and rabbits. The sheets provide a significant amount of data that have proved to be of benefit to both animal care staff and investigators. In summary, biomedical scientists have not recorded much evidence of pain or distress in laboratory rodents undergoing the ascites method of mAb production. That could be because they have not looked carefully or because sensitive techniques have not been developed to measure signs that might indicate the presence of pain or distress. The committee concludes that the possibility of distress in the mouse ascites method, particularly around the time of removing ascitic fluid (Jackson and others 1999a; Detolla 1998), should limit the use of the mouse ascites method as a routine method for producing mAb. Priming Priming of the peritoneal cavity is often accomplished through an intraperitoneal injection of pristane. Pristane is believed to act by inducing granulomatous reactions and interfering with peritoneal fluid drainage (Amyx 1987). Amyx (1987) suggested that large volumes of pristane injected intraperitoneally into mice are associated with weight loss, a hunched appearance, and lack of activity and that these clinical signs can be minimized by lowering the dose of pristane while achieving the desired effect. In spoken comments at a colloquium on recognition and alleviation of animal pain or distress (Amyx 1987), Amyx reported that the usual dose of 0.5 ml of pristane produces strong distress symptoms and that a smaller dose of 0.2 ml produces milder symptoms. Research on the optimal amount of pristane that would produce the greatest yields of ascites showed no statistically significant differences in the volumes produced in mice given preinjections of 0.1, 0.2, and 0.5 ml of pristane (Brodeur and others 1984; Hoogenraad and Wraight 1986). Hoogenraad and Wraight (1986) found that mice given 0.5 ml of pristane yielded an average of 9.72 ml of ascites fluid, and mice given 0.1 ml yielded an average of 9.65 ml. Alternatives to pristane have been evaluated. Gillette (1987) reported that of seven agents studied as primers, only incomplete Freund's adjuvant (IFA) produced results comparable or superior to those of pristane (it is suggested in some of the literature that IFA substantially reduces the 2-week waiting period between priming and hybridoma-cell injection). Although stress caused by these priming agents was not evaluated, mice primed with IFA survived for more taps once ascitic fluid was produced than did mice primed with pristane. In addition to the clinical signs described by Amyx (1987) and the Netherlands Code of Practice for the Production of Monoclonal Antibodies (Anon. 1989), two papers by Jackson and others (1999a, b) address the clinical and pathologic features of mAb production in mouse ascites. These papers also provide

OCR for page 33
some data on the effects (or lack of apparent effects) of pristane priming on an animal. Reports from Kuijpers in the Netherlands (Anon. 1989) state that mice treated with pristane showed various stages of peritonitis 10–20 days after treatment, with various amounts of abdominal fluid. There appears to be little dispute that pristane causes inflammation and blocks lymph flow in the peritoneal cavity, but there is some dispute about whether it causes animal distress. Pristane does induce a lupus-like syndrome in mice 4–8 weeks after intraperitoneal injection, which affects animal welfare (Satoh and Reeves 1994; Richards and others 1998). Jackson and others (1999a) note that peritonitis is known to cause abdominal pain in animals and in humans; they therefore find it reasonable to speculate that the induction of granulomatous peritonitis after intraperitoneal injection of pristane might cause pain or distress in mice. However, Jackson and others (1999a) fail to provide any clinical data to support the speculation. Clinical abnormalities were not observed in any of their test animals during the 2-week period after an injection of 0.5 ml of pristane. In unpublished observations from a blinded study, Gebhart observed intraperitoneal injections of 0.2 ml of pristane or 0.2 ml of saline into 12-week-old male BALB/c mice (Charles River) and was unable to distinguish by observation mice receiving pristane from those receiving saline. No writhing was produced by either pristane or saline, and there was no discernable nociceptive effect of either injection other than what would be attributed to penetration of the peritoneal cavity by a 25-gauge needle. As the data from the disturbance index paper (Barclay and others 1988) suggest, however, discomfort can develop slowly and not be apparent for several days (it took 6 days for mineral oil to produce an observable change in the disturbance index). When observed a week after injection of either pristane or saline, BALB/c mice appeared normal and healthy. Their coats were clean and smooth, and activity was normal (mice were housed two per cage). In previous experience (Ness and Gebhart 1990; Ness and Gebhart 1988), abdominal visceral discomfort in rodents is typically associated with a guarded posture and reduced activity. A disturbance index was not used, but it was not possible to distinguish mice that received pristane and mice that received saline a week after the injections. Perhaps a more provocative test or index would reveal some differences, but no differences were apparent in this mouse strain. In this regard, there are likely to be differences in effect and response between different strains of mice. The BALB/c mice used here are relatively docile and easy to handle. In future studies attempting to assess disturbances or distress in mice, the strain should be considered. Those observations suggest that injection of pristane does not produce pain or cause distress. Ascites After priming, a hybridoma cell-suspension is injected into the peritonea] cavity of the research animal. This leads to the development of a tumor, to the

OCR for page 33
accumulation of ascitic fluid, and to abdominal distension. Factors that affect animal welfare during this step of mAb production include the number of hybridoma cells injected, the type of tumor that develops (disseminated tumor or solid, invasive mass), and the effects of the ascites itself. Brodeur and others (1984) reported that survival times of mice reflect the number of hybridoma cells injected intraperitoneally. Mice given 2.5 × 107 cells survived an average of 8.5 days; mice given 3.2 × 106 cells survived an average 12.7 days. In this experiment, ascites formed sooner with larger cell inocula, but larger cell inocula shortened survival times. Current standard protocols usually call for the injection of 106 cells, which leads to longer survival times. In the study by Jackson and others (1999a), most of the animals survived 17–20 days. As indicated above, mild ascites in humans causes discomfort and distress but is generally painless (Burnside and McGlynn 1987). However, patients with massive ascites that causes a tense abdomen are frequently unable to walk, and they experience abdominal discomfort, indigestion, and heartburn (Mauch and Ultmann 1985; Pockros and Reynolds, 1986). Mauch and Ultmann (1985) reported that elevation of the diaphragm due to ascites is associated with dyspnea, orthopnea, or tachypnea. It therefore seems reasonable to assume that mice with large accumulations of ascitic fluid experience discomfort and distress. Clinical and pathologic data of Jackson and others (1999a) and reports by Kuijpers in the Netherlands (Anon. 1989) support that assumption. Kuijpers reported clinical signs in animals that include weakness, dehydration, anemia, and apathy. Kuijpers reported that without intervention animals die within 2–4 weeks from cachexia, dehydration, and complications associated with the tumor. Kuijpers also stated that "the abdominal wall is stretched in a relatively short time, which causes pain." Pathologic findings reported by Kuijpers include fluid in the thorax and abdominal cavity, adhesions of the abdominal organs, empty digestive tract, expanded spleen, and pale kidneys and liver. In addition, tumors filled the abdominal cavity, particularly near the lymph nodes in the mesentery. It should be stressed that these mice had particularly aggressive hybridomas, which invaded the peritoneal surfaces and intestinal walls. Jackson and others (1999a) monitored animals daily and reported clinical signs that included roughened hair coat, hunched posture, progressive increase in abdominal distension, decrease in activity, palpable abdominal masses, thin appearance, and dehydration (see table 1). The onset of those signs was related to the accumulation of ascitic fluid, which was determined by visible abdominal distension and increasing body weight, and the severity of the abnormalities increased with time. No such abnormalities were observed in the control mice. Necropsy revealed such pathologic findings as gross lesions, hemoperitoneum, abdominal adhesions, and disseminated intra-abdominal tumor or solid tumor masses, which might interfere with normal organ function. Although there were statistically significant differences among the hybridoma cell lines in the incidence, severity, and rate of progression of the abnormalities, the general progres-

OCR for page 33
Table 1 Incidence of Clinical Abnormalities in Ascitic Mice     No. CAF1 Micea No. SCID Mice     Control A B C Db Control E Number of Animals 12 20 20 20 20 11 20 Daily observations                 Rough hair coat 0 20 20 20 18 0 20   Hunched posture 0 20 20 20 18 0 20   Animal died after tap 1 0 0 0 1 0 0 0   Animal died after tap 2 0 0 1 3 0 0 0 Observations during posttapping period               Hunched posture 0 19 17 19 6 0 10   Decreased activity 0 19 17 19 8 0 9   Tachypnea 0 19 4 19 3 0 11   Pallor 0 19 7 19 10 0 3 a A, B, C, D, and E designate different hybridoma cell line groups. b Two mice in Group D produced no ascites and remained clinically normal. Source: Adapted from Jackson and others, 1999a. sion was similar. The different rates of ascites development had a statistically significant effect on survival and, ultimately, on the amount of antibody produced. Harvesting Ascitic Fluid Timing of harvesting, frequency of harvesting, and the effect of fluid harvesting on the animal all appear to be critical for the welfare of the animal. Brodeur and others (1984) reported that draining ascitic fluid as soon as it accumulates is necessary to reduce mortality. The UKCCCR guidelines (UKCCR 1998) recommend that ascitic-fluid volumes should not exceed 20% of the baseline body weight before abdominal paracentesis. Clinical data from Jackson and others (1999a) provide insight into the animal-welfare aspects of harvesting ascites. In their study, five different hybridomas were each injected into 20 mice. Abdominal paracentesis was performed when moderate abdominal distension was visible (the mean weight gain for all test mice was 17.4%). Each mouse was tapped a maximum of three times. However, any animal with persistent, severe clinical abnormalities was euthanized. During the 30 minutes immediately after the tap, the following clinical abnormalities were noted in some animals in each group: roughened hair coat, hunched posture, decreased activity, tachypnea,

OCR for page 33
dyspnea, and pallor most evident on the muzzle and ears (see table 1). Generally, those signs were transient and mild to moderate. However, in 19 of the animals from four of the five groups, after the first or second tap, symptoms were persistent and severe; five died, and the remaining 14 were euthanized (see table 1). Survival rates of the mice for tap 1 were 90–100%; for tap 2, 85–100%; and for tap 3, 35–100%. No clinical abnormalities were observed in control animals after a sham paracentesis. Jackson and others (1999a) conclude that the abnormal clinical signs observed after paracentesis were compatible with circulatory shock. Some IACUCs have established policies (see table 2) that require tapping of abdominal fluid to be performed under anesthesia and to be followed by injection of 1–2 ml of warmed saline to minimize posttapping hypovolemic shock. It is not clear that anesthesia use does reduce overall distress (for example, anesthetic administration causes handling stress), and it is possible that hypovolemic reactions can be minimized if the fluid is drawn off slowly (as is recommended for removing ascitic fluid from human patients—generally, up to 5% of body weight over 4 hours). However, if more than one tapping is to be permitted, the clinical signs observed by Jackson and others (1999a) after tapping indicate that some level of fluid replacement might alleviate some distress. There was some discussion at the ILAR working-group meeting of whether it was preferable for animal welfare to conduct fewer taps, which would mean that more mice would be needed to produce the required amount of antibody, or whether one should perform several taps and reduce mouse use by 50–75%. It is evident from the Jackson and others (1999a) paper that the clinical signs and evidence of animal discomfort and distress increase sharply once ascites has developed to the point where fluid can be harvested. Thus, one can reduce animal use; but if distress is high and increasing steadily in the 4–6 days from the first to the third and terminal tapping, then the extra taps are likely to be accompanied by substantial distress. If the animals are tapped just once under terminal anesthesia, the extent of distress will be minimized, but more animals will be put through (the procedure. The overriding criterion should be animal distress. Therefore, the number of taps should be limited and varied according to animal welfare and characteristics of the hybridoma being used. Some hybridomas seem to cause little distress (Jackson 1999a), and multiple taps could be allowed. Determining the number of taps depends on careful monitoring of the animal; if signs of distress appear, a terminal tap should be performed. A number of IACUCs have adopted a 0.2 ml volume limit on the pristane injection based on recommendations made by McGuill and Rowan (1989) (table 2). Members of the committee did not object to this as a guideline, but several reported that, in some strains of mice, 0.2 ml might not be sufficient to produce ascites and that as much as 0.5 ml might be required. There are some questions about the suitability of IFA as a substitute for pristane, and the committee recommends that IACUCs determine the reaction of animals to IFA before permitting its routine use. Most IACUCs with public policies recommend that

OCR for page 33
Table 2 University Policies for Production of Monoclonal Antibodies   Columbia Penn State Stanford U. Arizona UC Davis U. Iowa UMDNJ U. Minnesota UTK Test Cell Lines for Murine Virus — X X — — X — — — Priming If pristane used, not to exceed 0.25 ml Pristane as low as 0.1 ml or IFA = 0.5 ml — Minimum necessary CFA or pristane to produce ascites Pristane not to exceed 0.2 ml; agents other than pristane must be justified Pristane not to exceed 0.2 ml; IFA as good or better Pristane or IFA preferably; 0.1–0.2 ml, maximum 0.5 ml 0.50 ml maximum pristane, CFA, or IFA; up to 0.75 ml for large mice Pristane not to exceed 0.2 ml; agents other than pristane must be justified Needle size for tap — 18–22 gauge 18 gauge or smaller 21 gauge or smaller 18–22 gauge 18–20 gauge 18 gauge or smaller 20 gauge or smaller — No. tappings 3, then sacrifice Maximum 3, last terminal — — 3, last terminal 2, last after euthanasia maximum 3, preferably 2, last after euthanasia — 2 within 3 days of one another, then euthanize Monitoring after inoculation of tumor cell line Daily Daily 3 times/week for first week, then daily Daily 3 times/week for first week, then daily # daily At least daily Daily 2 times/day after observation of ascites

OCR for page 33
  Columbia Penn State Stanford U. Arizona UC Davis U. Iowa UMDNJ U. Minnesota UTK Replacement fluid after ascites harvesting — — — Administration of saline to help prevent shock — 1–2 ml of saline subcutaneous — — Anesthesia during tapping — Used to minimize P&D from tapping (methoxyflurane) Anesthesia used for training of new personnel or students Anesthesia preferable Anesthesia used for new personnel — 12 — —

OCR for page 33
animals be monitored daily after inoculation of the hybridoma cells. The committee believes that this is an appropriate standard with the proviso that IACUCs require that investigators pay special attention to animals after ascites develops and abdominal distension occurs. Most IACUCs limit the number of taps to two or three (with the final tap being terminal). The number of taps is a critical issue for animal welfare, and the investigator should pay particular attention to the condition of the animals. If signs of distress, such as hunched posture and roughened coat, are observed, the animals should be euthanized immediately; the committee believes that the primary task of the IACUC should be to limit animal pain and distress rather than animal numbers. There are too few data to develop a formal recommendation on either the use of anesthesia during the tap or the administration of saline replacement after the tap. The committee urges IACUCs to collect data on their experience with mAb production and animal welfare and make them available to others at meetings and through appropriate publications and discussion groups. Feeder Cell Harvesting and Serum Supplements for in Vitro Hydridoma Culture The use of FBS is accompanied by animal-welfare costs. The harvesting of such fetal serum has raised concerns about the welfare of the animals from which it is obtained, but there are few descriptions in the published literature of such operations. It has been suggested that hybridoma cell cultures require feeder cells harvested from mice peritoneal cavities or mouse embryoblast serum (Harlow and Lane 1988, pp. 220–221). The prevalence of the use of feeder cells is unknown, and the extent to which serum and feeder-cell supplements are required was disputed during the 1-day workshop held by the committee. Summary of Animal Welfare Issues Data from the study by Jackson and others (1999a) suggest that the clinical condition associated with the production of ascites generally worsens in association with progressive tumor growth, continuing ascites production, and repeated abdominal paracentesis. Clearly, there is a lack of data on animal-welfare issues related to mAb production by the ascites method. The article by Jackson and others (1999a) is the only published study in the peer-reviewed literature of the last decade that provides clinical data on how this procedure affects the experimental animal. Lack of conclusive data is evident when one looks at university policies for this procedure (see table 2). There are no best-practice guidelines regarding animal welfare; variations are apparent from university to university for the procedures of priming, number of taps, monitoring of animals, anesthesia, and use of replacement fluid.

OCR for page 33
An additional animal-welfare concern, although not directly related to the ascites method, is the use of FBS during in vitro mAb production. It should not be assumed that in vitro procedures are inherently more humane; the use of fetal bovine serum raises questions with regard to the methods for collecting the serum, in that the serum might be obtained under circumstances that may lead to distress for the animals (McGuill and Rowan 1989).