FDA Home Page | CVM Home Page | CVM A-Z Index | Contact CVM | Site Map | FDA Centennial

<<back

Appendix F: The Viagen Dataset

A. Background

Viagen, Inc., a privately held company engaged in cloning swine and cattle for agricultural purposes, also responded to CVM’s request for additional data on clones and their progeny (for a similar discussion on cattle, see Appendix E: The Cyagra Dataset). Viagen designed two studies to evaluate the health, growth, and meat composition of swine clones, fertility of boar⁸¹ clones, and health, growth and meat composition of clone progeny vs. age-matched, genetically related, artificial insemination (AI)-derived comparator animals. Various scientists from CVM with expertise relevant to these studies provided peer-review comments to Viagen on the overall experimental design, but did not formally approve or disapprove the studies. All of the raw data that CVM received from Viagen is found at the end of this Appendix.

As described in their report, the company’s objectives in running these studies were to

1. Compare the health of swine clones vs. AI-derived pigs of a related genetic line;

2. Compare the biological characteristics, including laboratory measurements and meat composition, of swine clones vs. AI-derived swine;

3. Assess the reproductive characteristics of clone boars vs. AI-derived boars; and,

4. Compare the biological characteristics, including laboratory measurements and meat composition, of the progeny of clone boars vs. progeny of AI-derived closely-related boars (nuclear donors for clones or progeny of nuclear donor).

B. Experimental Design

Study 1 Overview: Evaluation of Clones

Survival, health, growth and meat characteristics of seven clones (6 “Hamline”⁸² and 1 Duroc (a common breed of swine used in U.S. commercial pork production), all barrows⁸³) and 15 conventional barrows (all Hamline) were followed from 50 days after birth through slaughter at approximately 6 months of age. Clones were assigned to this study shortly after they were weaned. Age-matched comparator⁸⁴ pigs were selected from litters sired by the Hamline nuclear donor boar in a conventional breeding (AI) program.

An additional group of four clones (three Hamline and one Duroc) and three AI-derived boars (the Hamline nuclear donor and two AI sons of the Duroc nuclear donor) were used for evaluation of semen quality and fertility. The progeny generated from these boars were used in another experiment to study whether progeny of clones were materially different in growth, health or meat characteristics from conventional swine. Because of the small number of clones, these animals were evaluated on a case-by-case basis and simple means generated when appropriate, rather than applying statistical methods to the data.

Figure F-1: Design of Viagen Study 1

Study 2 Overview: Evaluation of Clone Progeny

In the second experiment, survival, health, growth, and meat composition of progeny derived from clone boars (three Hamline and one Duroc) and either their nuclear donor (Hamline) or two sons of the nuclear donor Duroc boar were compared. Gilts⁸⁵ sired by six different boars were randomly assigned according to their sire (such that each maternal grandsire was represented in each breeding group to control for effect of maternal grandsire) to each clone or comparator boar, and bred by AI. Sixty-eight gilts farrowed (gave birth) under standard commercial procedures, yielding 402 total progeny from clone boars and 300 total progeny from comparator boars. This included 284 progeny from Hamline clones, 118 progeny from the Duroc clone, 61 progeny from the Hamline donor sire (Hamline comparator), and 239 progeny from Duroc donor sire sons (Duroc comparators).

Figure F-2: Design of Viagen Study 2

1. Study 1: Clones vs. Comparators

Clone pigs were obtained by Caesarean section (C-section) on the day of or the day before the sows’ predicted farrowing date. Clone piglets were kept under highly biosecure⁸⁶ conditions prior to initiating the experiment. Clones were bottle fed commercial milk replacer and did not receive colostrum at the time of birth. Comparator pigs, however, were born naturally (i.e., vaginally) in a standard commercial farrowing house, and suckled their dams until weaning. By the start of the test period (~50 days of age), the clones were similar in weight to the comparator pigs (16.4 vs. 15.3 kg for clones and comparators, respectively). In addition, comparator barrows were castrated during the first week of life, as per normal industry practice, but clone barrows were castrated at a later age (~30 days old). Pigs were assigned to this study shortly after they were weaned. Comparator barrows were selected from litters sired by one of the nuclear donor boars in a conventional (AI) breeding program. At the start of Study 1 clones and comparator animals were housed in the same facility. Target slaughter weight was approximately 122 kg.⁸⁷

a. Animal Health, Hematology, Clinical Chemistry, and Urinalysis

Health records from date of birth were submitted for clones, and included birth weight, heart rate, respiration, body temperature, body weight, and daily health observations. All health problems and treatments were recorded. Necropsy reports were provided for clones that died or were euthanized. Blood samples of clones and comparators were collected at birth and again either approximately one week prior to slaughter, or immediately after animals were slaughtered. Urine samples were collected after slaughter (by bladder puncture). Blood hematology, clinical chemistry, and urine values were assayed as for the Cyagra dataset at the same independent testing laboratory (Cornell University’s Animal Health Diagnostic Laboratory, see Appendix E).

b. Boar Semen Evaluation

Semen samples were evaluated in two commercial boar lines (Hamline and Duroc). The Hamline included three clone boars and the sexually derived nuclear donor boar for the clones as a comparator. The Duroc line included one clone boar and two comparators that were AI-derived progeny of the donor (semen from the Duroc nuclear donor was unavailable). Clones and comparator animals were housed at separate facilities. Semen was collected at least once a week from each boar, was diluted with a commercial semen extender to yield 3 X 109 sperm per dose, and cooled to 19°C per standard industry practice. Semen from clones was evaluated pre-freezing and post-freezing (post-shipping) to determine quality. Comparator boar semen was only evaluated post-shipping. Standard tests such as sperm concentration, percent total motility, percent progressive motility, and number of head/tail abnormalities were evaluated.

c. Farrowing Rate

Following semen evaluation, semen from the four clone boars and three comparator boars were used to evaluate pregnancy (number of females diagnosed pregnant divided by number of females bred) and farrowing (number of females giving birth to a litter divided by the number of females bred) rates. Three of the clone boars were bred to 12 gilts each and one clone boar was bred to 13 gilts (total of 49 gilts bred). For the comparator boars, two were bred to 14 gilts each and one was bred to 12 gilts (a total of 40 gilts bred). All gilts were bred via AI twice, once on the day of observed estrus and approximately 24 hours later.

2. Study 2: Progeny of Clone Boars vs. Progeny of Conventional Boars

Breeding resulted in 36 litters (402 total progeny) from the clone boars and 25 litters (300 total progeny) from the comparator boars. Reproductive outcomes evaluated included litter size, birth weight, number of stillborns, number of dead/destroyed animals, abnormalities, and nipple counts (another measure of breeding fitness often taken at birth). Animals were slaughtered when they had reached the target weight of approximately 122 kg. Average age at slaughter was 169.3 days for progeny of comparators and 174.9 days for progeny of clones.

a. Statistical Analysis

i. Methodology

To aid CVM’s analysis of the very large sample sizes in Study 2, the Center’s biostatisticians produced data summaries to facilitate veterinary and scientific evaluation of the outcomes collected from the progeny. Two types of summaries were used: contingency tables and boxplots. Definitions and examples of both are given below.

Contingency tables present data in a row and column format and are often used to summarize the results of discrete variables. A variable can be defined as discrete if its responses are limited to a few predetermined values. For example, a discrete variable from Study 2 is “Abnormality at birth” which is described according to four possible discrete outcomes: Normal, Atresia, Spraddle legs, or Other. For those progeny for which this variable was recorded, the distributions of “Abnormality” for the live-born progeny of animal clones and progeny of comparators are displayed in the following contingency table (Table F-1).

By contrast, continuous variables are those whose responses can take on a continuum of values. Graphical displays are useful in summarizing the responses from continuous variables. The boxplot is one type of graphical display. An example of a continuous variable from Study 2 is “animal weight.” The boxplots of birth weights from the progeny experiment are provided as an example below.

Figure F-3: Boxplot of birth weights of progeny of clones and comparators (in kgs)

In Figure F-3, the boxplot on the far right includes all of the birth weights recorded for clone progeny and comparators. To create a boxplot, the birth weight data was ordered from smallest to largest. The bottom edge of the box was drawn at a value of birth weight for which 25 percent of the observations have a lower value, which is called the “25th percentile”1 (or first quartile) of the sample. Similarly, the top of the box was drawn at a value of birth weight for which 75 percent of the observations have a lower value, which is called the “75th percentile”⁸⁸ of the sample. The value of the observation associated with the 75th percentile minus the value of the observation associated with the 25th percentile is called the interquartile range (IQR). Within the box, a plus sign is used to indicate the sample mean and the line from side to side of the box indicates the sample median which is the value of the observation associated with the 50th percentile below which half the observations lie. Any observation values which lie above the “upper fence,” a point equal to 1.5 times IQR plus the value associated with the 75th percentile or below the “lower fence,” a value equal to the value of the observation associated with the 25th percentile minus 1.5 times IQR, are individually plotted on the graph and are called extreme values. (Note the actual fence values are not plotted.) The whiskers stemming out from the boxes represent the observation just below the upper fence and the observation just above the lower fence. In the example above, the boxplots, from left to right, are for progeny of Hamline clones, progeny of Hamline comparator animals, progeny of Duroc clones, and progeny of Duroc comparator animals.

Boxplots do not require assumptions about the form of the underlying continuous data distributions and are uniformly applicable across a range of different distributions. When boxplots are used to describe a variable sampled from a single population and having a normal distribution, they tend to identify about 1 percent of values above and 1 percent of values below the box as “extreme.”

When the underlying distribution has more data in the tails than a normal distribution, then more than 1 percent of the distribution is identified as extreme. To evaluate the Viagen data, values from the progeny of clones were compared to the ends of the whiskers in the boxplots for the values from the progeny of comparators.

ii. Utility of Statistical Analysis as a Tool for Assessment of Outcomes

Results (numerical values) for any outcome from the clone progeny were compared to the ends of the whiskers for the distribution of values from progeny. If a result fell within the range defined by the box and whiskers, it was considered “within the same range as the comparators” and therefore posed no significant concern for the health of the animal or the composition of the food product. If a result for a clone fell outside the range identified by the whiskers for the comparators, or was identified because of its value in a contingency table, the appropriate CVM scientist (e.g., veterinarian, animal scientist, chemist, toxicologist) traced back all of the data for that the individual animal, and conducted a thorough examination of all of the relevant clinical information to draw a conclusion regarding the biological or health significance of the value.

In addition, to provide consistency with the presentation of the analysis developed for the Cyagra dataset (Appendix E), box charts were constructed. As described in that Appendix, black boxes or up or down arrows were assigned to values for each clone progeny by the reviewing veterinarian or scientist on the basis of this multi-step review procedure.

b. Specific Methods for the Analysis of Animal Health, Hematology, Clinical Chemistry, and Urinalysis Data

Progeny of clones and comparators were farrowed during a three week period from July 7-27, 2004, at the Meat Animal Research Center, USDA, Clay Center, Nebraska. All available animals were sampled (serum and whole blood) in July, October, and January. Blood samples were obtained between 7:30 and 10:30 AM. After collection, samples were placed on ice until centrifugation at 900 x g for 10 min at 4°C. Serum was collected and stored at -20°C until analysis for estradiol-17ß (E2), and Insulin like growth factor-I (IGF-I). These time points represented early, middle and late stages in the life of the test subjects, which mirrors the lifespan of animals used for commercial pork production in the U.S. When considered in the context of the CBSA approach used in this Draft Risk Assessment, these time points roughly correspond to the Perinatal (July) and Juvenile (October and January) developmental nodes, as most pigs used for food are slaughtered before they reach sexual maturity. July and October blood samples were collected from the cranial vena cava. In January, blood and urine samples were harvested postmortem. Pigs were individually identified by numbered ear tags, which did not reveal to caretakers or personnel taking samples whether the animals were the progeny of clones or comparators.

c. Meat Composition, Carcass Characteristics, and Meat Quality Assessments for Clones, Comparators, and Progeny.

As in standard US pork production, swine were slaughtered when they reached approximately 122 kg in weight. Samples were collected from the latissimus dorsi muscle (the muscle that comprises the large, round-shaped muscle next to the vertebrae end of the ribs that makes up the boneless portion of the “rib eye” roast or steak), frozen, and held at -20oC for approximately two weeks prior to shipping to a commercial laboratory for compositional analysis. Samples were identified by a nine digit code so that personnel at the analytical laboratory did not know which samples belonged to comparators, clones, or progeny of clones. Muscle samples (approximately 0.45 kg, deboned) were held frozen at -20oC for three to 20 days after receipt at the laboratory before being partially thawed (10 – 18°C overnight) and homogenized using a commercial grade meat grinder with a #12 blade and passed through a 1/8 inch screen. Homogenized meat samples were analyzed for fatty acid and amino acid profiles, cholesterol, vitamins B6, B12, niacin, calcium, iron, phosphorus, and zinc. Analyses were carried out according to the Association of Analytical Communities International (AOAC) methods for amino acids (982.30), calcium, iron, phosphorus and zinc (965.17 and 985.01), cholesterol (994.10), fatty acid profile (996.06), niacin (944.13), vitamin B12 (952.20) and vitamin B6 (961.15). With the exception of vitamins B6, B12, niacin, and cholesterol, values were reported as percentages. Vitamin B6, niacin and cholesterol were reported as mg/100 g of sample, and vitamin B12 was reported as mcg/100 g of sample. Percentages were based on grams of nutrient per 100 grams of homogenized meat sample.

Economically important traits for the swine carcass evaluation include animal live weight, dressing percentage, carcass length, loin muscle area, backfat depth, extent of muscling, and firmness. These are described in the following paragraphs.

The normal range for dressing percentage⁸⁹ is 68-77 percent (average of 72 percent) with a minimum carcass length of 76.2 cm. The minimum acceptable area for loin muscle is 11.4 cm2 and the maximum back fat thickness over the last rib (10th or 11th rib) is 3.8 cm.⁹⁰

Three degrees of pork carcass muscling are recognized in the pork grading standards or USDA⁹¹ carcass muscling score: muscle score #1 = thin (inferior), muscle score #2 = average, and muscle score #3 = thick (superior). Scores of either 2 or 3 are considered acceptable under USDA inspection standards. Muscle firmness is a subjective score determined by visual observation and physical handling of the meat. Muscle firmness is measured on a three-point scale: score #1 = soft, score #2 = firm, and #3 = very firm. As with muscling score, a firmness score of 2 or 3 is considered desirable.

C. Results

1. Study 1
 
a. Survival

Seven clones and 16 comparators were raised by Viagen Inc. The seven clones began the study at 50 days of age, and all seven survived until the end of the observation period (approximately 195 days of age); however, two clones were euthanized due to chronic health problems at the end of the study. Thus, five clones were slaughtered as per normal industry practice. One of these clones was diagnosed with a lung adhesion post mortem, which, by USDA standards would have led to condemnation of the carcass. Similarly, 15/16 comparators reached the end of the observation period (approximately 171 days of age), with one comparator euthanized approximately four months prior to the end of the study because of a chronic respiratory condition.

Because litters of clones were born approximately six weeks apart, they were separated into two groups of animals and matched with comparators born at approximately the same time. In the first group, five clones and 11 comparators were raised and studied. In this first group, one clone (#21, farrowed 10/13/03) and one comparator (#025435) did not thrive. Clone #21, a Duroc, was euthanized at the end of the study (May 27, 2004) and was not included in the final carcass evaluations because of low weight. However, hematology and clinical chemistry values for this clone were used for analysis of health data. Non-clone comparator #025435 was euthanized on December 4, 2003 because of a respiratory condition that was not improving.

Viagen raised two clones and five comparators in the second group. Clone #23, a Hamline, farrowed on 11/23/03, was described as a “poor doer” and was euthanized at the end of the study. All comparators from this group survived to slaughter.

b. Animal Health, Growth, Blood Clinical Chemistry, Hematology, and Urinalysis

Clones were smaller on average at birth than their conventionally bred counterparts (1.1 vs. 1.7 kg) (Table F-2). By the start of the test period (~50 days), however, clones were similar in weight to their comparators (16.4 vs. 15.3 kg).

Animal health records indicated that six of the seven clones developed scours (severe diarrhea) approximately two months after birth, which would have been shortly after they were moved from the biosecure environment. The affected clones were #18, 19, 20, 21, 22 and 23. One of these animals (Clone #22) was diagnosed at approximately four months of age with influenza and secondary bacterial septicemia, was periodically treated with antibiotics throughout the experimental period, and was eventually euthanized at the end of the study. One of the comparator animals (#025435) was diagnosed at approximately one month of age with influenza and secondary bacterial septicemia. This animal was periodically treated with antibiotics throughout the experimental period and was eventually euthanized. The clones weighed less at slaughter and took an average of 27 days longer to reach the target slaughter weight. As a result, the calculated average daily (weight) gain (ADG) for clones was lower than for comparators (0.63 vs. 0.93 kg/day). The clones were slaughtered later than comparators because of the greater length of time needed for them to achieve the target weight, and also because the slaughter facility could only accommodate them on a few specified days. This latter point was because clones are currently withheld from the food supply, and the facility preferred to slaughter the clones on days when no other swine were being processed, in order to prevent accidental mixing of the clones with conventional animals.

As previously mentioned, clones and comparators were reared under different conditions prior to starting the experiment. Clones were delivered by C-section, did not receive colostrum, were fed a commercial milk replacer, were castrated at an older age, and were raised in a biosecure environment prior to initiation of the study. The clones were then housed along with the comparator animals in an environment similar to commercial swine facilities, where they were exposed to a typical range of pathogens. The comparators were farrowed vaginally following natural initiation of labor in a farrowing house similar to commercial swine operations, and the progeny were allowed to nurse off their dams. This disparity likely accounts for some of the differences in growth and health observed. The sudden change from a biosecure environment, with a low immune challenge, to a more conventional barn with a high immune challenge placed considerable stress on the immune systems of the young clones. This stress likely led to the higher incidence of scouring in this group, and may have resulted in decreased food consumption or absorption, and increased energy expenditure to combat illness.

Body temperature, heart and respiration rates were measured in clones, but not in comparators. Because of the absence of comparator data, observations for these measurements are put in the context of common veterinary practice or reference texts or manuals.

Body Temperature (BT). Clones #18, 19, 20, 22, and 24 were measured for BT 2-9 times per day for the first eight days of life (no records on BT for other clones were available). Body temperature increased over this eight day period from an average of 100.38 ± 0.18ºF (37.99°C) on Day 1 to 101.89 ± 0.21ºF (38.83°C) on Day 8. The low BT was 97.7º F (36.5°C), and occurred in clones #22 and 24 early in the study (Days 1 and 2 post farrowing). The high temperatures were 104.6ºF (40.33°C), 103.8ºF (39.89°C), and 103ºF (39.44°C) and occurred in one clone (#24), twice on Day 1, 1, and 8 respectively. The Merck Manual’s Reference range for swine BT is 101-103°F (38.33 – 39.44°C). The BT indicates an apparent appropriate response to autonomous BT regulation (adjustment to post-natal environment) after delivery from the uterus where BT is not self-regulated, but controlled by the intra-uterine environment. As mentioned in Chapter V, newborn pigs typically cannot regulate their own BT. Generally, it is recommended to maintain the environment of newborn pigs at a temperature of 86-93°F (30-34°C) for the first week to allow for this adaptation period (Merck Veterinary Manual Online).

Heart rates (HR). The heart rates (HRs) of clones also were measured 4-12 times per day for the first eight days of life. Heart rate decreased over time from 194.33 ± 1.42 to 178.11 ± 1.18 beats per minute (bpm). The highest value was 220 bpm for clone #18 on Day 1; the lowest value is 100 bpm for clone #22 on Day 1. This decrease is an appropriate response for neonates. Heart rates decrease with age as fetal circulation (ductus venosus,⁹² foramen ovale,⁹³ and ductus arteriosus⁹⁴ closure) normalizes and the neonate develops autonomic control (Medline, Medical Encyclopedia⁹⁵). There were no values from the non-clone group for direct comparison. These values should be evaluated with caution because the very act of ausculting (listening to) the thorax of an animal that is seldom handled can raise HR artifactually. In fact, it is possible that the reduction in HR over time represents the pigs becoming more accustomed to handling. Data on HR for non-clone neonatal swine are not routinely reported; however, one study (Foster et al. 2001) recorded HR of day-old piglets before and after insertion of a breathing tube. Prior to insertion of the tube, average HR was 190.75 ± 36.45 bpm. In another study (Aaltonen et al. 2003) examining the effects of mecomium aspiration and asphyxia in piglets aged 10 to 12 days. Average HR for the control (untreated) group was 203 ± 23 bpm. Given these data for comparison, the HR of clone pigs during the first eight days of life fell within or slightly below the range of non-clone pigs of similar ages.

Respiration. The same clones as above were evaluated 5-12 times per day for the first eight days of life. Respiration rates decreased over time, ranging from 87.56 ± 0.71 breaths/minute on Day 1 to 70.29 ± 0.79 breaths/minute on Day 7. As with HR, data on swine neonatal respiration rates are not commonly reported. However, one study (McDeigan et al. 2003) measured respiratory responses in pigs from 3 to 7 days of age before and after exposure to either saline (control) or Escherichia coli endotoxin. Average pre-treatment respiration rate for control piglets was 44 ± 5 breaths/minute in this study, and thus considerably lower than the average respiration rates of the clones in the Viagen study. The lowest daily average respiration rate among the neonatal clones in the Viagen study was 69.6 breaths/min on day 7 (Clone #18). The cautions mentioned for interpreting HR data should also be applied to interpreting respiration rates. Also, animals delivered by C-section have a transient tachypnea (rapid breathing) because the pulmonary surfactant and fluid is not distributed and squeezed out of the lungs as it would in a vaginal birth (Medline, Medical Encyclopedia).

Daily Health Records. Clones #18, 19, 20, 21, 22, and 23 all had scours (severe diarrhea) while on study. Clone #18 herniated a loop of bowel post-castration, which was repaired, and the pig responded well. This pig had an inguinal hernia, sometimes referred to as a “busted pig,” which is not uncommon in swine (Gatphayak et al. 2005). As previously mentioned, clone #21 was described as a “poor doer” (unthrifty, poor weight gain) and was euthanized because of low body weight. He had chronic scours and a skin abscess on his neck. Clone #22 had scours twice while on study. He developed sepsis (a blood infection), and had a low body weight at slaughter. Clone #23 had many health problems, including scours, respiratory disease, cyanotic (bluish) skin color, and grew slowly. He was euthanized because of his low body weight. Clone #24 had some respiratory difficulty toward the end of the study but responded to therapy. One comparator pig (#025435) was euthanized for non-responsive diarrhea and pneumonia early in the study. Another comparator (#027446) had one reported health problem of a swollen dewclaw that responded to therapy.

Although these clones had more health problems than comparators, it is important to note that differences in early rearing are likely to have contributed to these outcomes. First, the clones were not suckled on their dams, but were fed commercial milk replacer. This deprived them of the passive immunity provided by transfer of maternal antibodies in the colostrum. Secondly, pigs raised in a biosecure environment would not have developed antibodies to common environmental pathogens, and would be at increased risk of developing infections until their immune systems adjusted to the environmental challenges when they were moved to lessbiosecure conditions at day 50 of age. The comparators, on the other hand, were born in conditions equivalent to commercial swine breeding, and were allowed to suckle on their dams, thereby allowing their immune systems to develop along more conventional lines. We also note that even though some of the clones developed illness, most (5/7) were able to respond to
treatment. Further, the hernia observed in clone #18 is not unique to clones, but is considered common among male pigs in the general swine population (Merck Veterinary Manual Online).

Clinical Chemistry and Pathology. Clinical pathology is the term generally used for laboratory findings that includes clinical chemistry, hematology, and urinalysis among others. As described in Appendices E and G, clinical chemistry and hematology responses are best evaluated in the context of the whole animal, including its age, species, breed, husbandry, geographic location, reproductive status, and the laboratory performing the analysis. Laboratory findings complement the subjective physical diagnosis of the patient by providing objective information for the process of differential diagnosis, monitoring treatment, and formulation of a prognosis. “Abnormal” laboratory measurements and examinations are often defined as those values lying outside the limits of the reference range. Determining what constitutes “normal” is more complex than simply comparing an individual value to a reference range derived from a sample of a representative population. In this study, the firm provided serum chemistry data, white cell counts, red cell counts, and urinalysis using a rapid “dipstick” test. Because the number of pigs in the study was low, a statistical analysis of the comparator ranges was not appropriate. Values from clones were compared to those from the comparators.

Across all three sampling periods, 89 percent (281/315) of clones’ hematology measurements were within the range of the comparator population. At the beginning of the study, 80 percent of the clones’ hematology values (84/105) were within the range of the comparators. This percentage improved at the next sampling period, approximately two months later, such that 88 percent of the clones’ values (92/105) were within range. For clinical chemistry, 76 percent of the clones’ values (236/315) were within the comparator range across all time periods. As with hematology, the percentage of clone values within the range of comparators was lower at the beginning of the trial, but this ratio improved as the animals grew. At the beginning of the study, 63 percent (66/105) of the clones’ values were within the comparator range. By the second sampling period, 83 percent (88/105) of the clones’ values were within the comparator range. By the end of the trial it appeared that clones for the most part had stabilized, with 84 percent of the hematology values and 98 percent of clinical chemistry values within the comparator range (Table F-4).

Seven comparators had elevated WBC counts with no apparent health problem. Clone #20 had a high WBC (32,000 cells/µl) during an episode of diarrhea. This is an appropriate clinical response. One clone (#18) and one comparator (#025457) had low platelets at the end of the study. Red blood cell size for both clones and comparators indicated a microcytosis (low MCV). Because the clones and comparators were similar, this value is likely related to the laboratory and not a difference in health.

Clone #21 had elevated BUN levels at the start and middle of the study and elevated liver and muscle enzymes at the start and end of study. He did not show an increase in WBC, and had a slight increase in urine glucose at the end of the study. These findings are consistent with a “poor-doing” animal. Animals in a negative energy balance (due to inadequate nutrition, food intake or utilization) will start to mobilize muscle protein for energy once their body energy stores are depleted. Blood urea can elevate with increased protein digestion or mobilization. CK (an intracellular enzyme for muscle) will also increase with muscle break down. The elevated liver enzymes indicate some insult to the liver. The cause of these health problems is not evident from the data presented.

Clone #23 exhibited an increase in blood glucose at the start of the study (January) and bile acids at the mid-point of the study (March). By the end of the study (May), clone #23 had glucose and bile acid values within the range of the comparators. Clone #23 was also a “poor doer” (poor weight gain), but had no other clinical pathology abnormalities evident. Although increased blood glucose levels may indicate metabolic disturbances such as diabetes, the increased glucose levels in this study (in blood and once in urine) appear to have been transient and resolved spontaneously. They therefore do not appear to be indicative of an insulin deficiency or resistance. Bile acids are an indicator of liver function (see Appendix G). If the bile acid value is consistently elevated, liver insufficiency may be indicated. However, bile acid levels are also dependent on when the animal was fed, and increase after a meal. Because information on the relationship between blood draws and feeding times was not provided, no conclusion can be made about the single measurement which on its own does not suggest a health problem.

Urinalysis. Urine values were determined by dipstick. There was apparently no Specific Gravity (SG) determination using a refractometer or sedimentation evaluation, so SG was measured only by the dipstick method, which is less accurate than these other methods. With these caveats in mind, two clones and one comparator had glucose in their urine. The percent of clones and comparators with blood and protein in their urine were similar. Seven out of 15 urine samples from clones were positive for blood (47 percent) vs. 16/ 44 (36 percent) samples from comparators. For protein, 6/15 samples from clones (40 percent) vs. 20/44 (45 percent) samples from comparators were positive. Color, turbidity, SG, bilirubin, pH, ketone, nitrite, and leukocytes were all similar between clones and comparators. The presence of blood in the urine may have been due to the fact that these samples were taken after the pigs had been slaughtered.

Organ weights. The clones’ body weights were lighter on average than comparators at slaughter (117 vs. 128.8 kg). Clones #21 and 23 were so light (86.2 and 89.8 kg, respectively), they were not included in the organ weight analyses. Kidneys were smaller for clones vs. comparators on a percent of body weight basis. However, heart, liver, lungs, and spleen were similar as a percent of body weight for clones vs. comparators. With no clinical chemistry indicators of renal insufficiency in clones, the smaller kidney weights are interesting but not conclusive. These findings indicate no appreciable differences between clones and comparators in organ size as a percent of body weight.

IGF-1 and E2. In general, the clones had lower levels of IGF-I than the comparators after birth and before slaughter (Table F-5). However, the values were within the range observed in the comparator group, except for one clone (Clone #19), that had lower levels at the beginning of the study. This clone had levels of IGF-I at the limit of assay sensitivity (low value). Before slaughter, clones #19 and #20 presented levels of IGF-I that were close or below the detectable values by the analytical method.

Although, critical illness can modify and has an important impact on the human endocrine system (Sartin et al. 1998), these animals did not show health-associated conditions at the time of sampling that could explain the low levels of IGF-I. Additionally, the IGF-I axis (the glands, organs and hormones associated with IGF-I), as well as cortisol and gonadal steroids, are endocrine determinants of the growth potential of animals (Mauras and Haymond 2005). Nevertheless, the body weights of the animals with low IGF-I (clones #19 and #20), were not lower than the control animals. This observation suggests that the reduction in IGF-I levels seen before slaughter for these two clones is not biologically important and perhaps is associated with the handling of samples or time of sampling relative to when the animals ate. In addition, limited information has been provided describing the handling and storage of samples in the current experiment, and hemolysis and storage of blood samples have been shown to reduce detectable concentrations of hormones using the assay technique employed in this study (Reimers et al 1991). Therefore, the possibility that the low levels of IGF-I in clones #19 and #20 may be due to hemolyzed serum used for the hormone analysis cannot be excluded.

Clones had slightly lower levels of E2 than the comparator group; this trend was observed during the perinatal period and before slaughter (Table F-6). However, the E2 values of clones were within the range of the comparators, but one comparator animal (#025461) displayed elevated levels of E2, as a result, the mean values of the comparator group increased. This animal also presented the highest rate of weight gain throughout the study. The body weight at slaughter was higher than the other comparators. If the data for this animal is removed from the comparator group, the apparent difference between animal clones and comparators is considerably diminished. Considering the above observations it can be concluded that the slight reduction in the levels of E2 in the clones was not biologically relevant.

Conclusions for Health and Growth of Viagen Clones. The results of this study are limited by the small number of animals and the design of the experiment. However, results of this experiment indicate that even though the clone barrows were subjected to a significant immunological challenge after moving from more biosecure conditions to more standard housing conditions, most clones were able to respond appropriately to this stress. This immunological challenge potentially slowed the growth of clones early in this experiment and thus may have resulted in a delay in their maturation.

c. Carcass Characteristics

Carcass characteristics were provided on the four Hamline clones and 15 comparator barrows followed in Study 1; these are summarized in Table F-7. Hot carcass weights⁹⁶ averaged 85.91 and 90.68 kg for clone and comparator barrows, respectively, with the exception of the animals that were excluded (see discussion in next paragraph). Carcass lengths were 82.4 and 84.5 cm for clones and comparators, respectively. Dressing percentages were 70.1 and 70.2 percent for clones and comparators, and were similar across breed groups. Backfat thickness over the first rib, tenth rib, last rib, and lumbar vertebra were slightly greater for comparators than clones which may, in part, be due to the heavier body weight of comparator barrows at the time of slaughter. Qualitative characteristics including USDA carcass muscle score, firmness, and marbling were similar across groups and are illustrated in Table F-7. All animals received score 2 for carcass muscle. All of the clone and comparator barrows had marbling scores of either 1 or 2.

The Duroc clone barrow carcass was condemned at slaughter due to a lung adhesion, and thus data relating to growth and carcass characteristics from this animal were not included for these variables. The rationale for excluding the carcass for this analysis is that in some cases, lung adhesions are due to bacterial infection, and animals fail to thrive, (i.e., their growth and carcass characteristics are adversely affected). Two other clones (#21 and 23) were approximately 45.45 kg lighter than any of the other animals in the experiment at the time of slaughter, and for this reason were excluded from carcass evaluation.

Measurements of pH at 24 hours post-slaughter on the longissimus dorsi muscle were similar. Loin eye area for meat cuts for clones and comparators were quite similar at 116.33 ± 10.16 and
11.76 ± 11.18 cm, respectively.

d. Meat Composition Analysis for Clones and Comparators

Meat composition data were available for the five clones (four Hamline and one Duroc) and 15 comparator animals (all Hamline). Because there were no differences between the Duroc and Hamline clones, data were pooled. All of the values examined (amino acids, fatty acids, cholesterol, nutritionally important vitamins and minerals (see discussion in Chapter 6) were remarkably similar, and no biologically relevant differences were noted (Table F-8).

Conclusions for Carcass Characteristics and Meat Composition of Clones. No remarkable differences were observed between clones and comparators for any of the characteristics evaluated. The small differences in backfat thickness and marbling are likely due to the lighter weight of clones vs. comparators at slaughter. Differences in meat nutrient composition were very small and likely not biologically relevant. The lack of biologically relevant differences in the food composition values between muscle of swine clones and comparators supports the conclusion that there is no additional food consumption risk from the consumption of muscle from swine clones compared to their conventional counterparts.

e. Semen and Breeding Evaluation

Sperm concentration, percent total motility (percent of sperm in motion), percent progressive motility (evaluates forward movement), and number of head/tail abnormalities were similar for the four clones and the comparator boars. The clones did not appear to have a reduction in semen quality relative to the comparator boars. The lowest post-shipment semen concentration, percent motility, and percent progressive motility were observed in the Duroc line comparator boars.

f. Farrowing Rate

The farrowing rate for the gilts inseminated with semen from clones and comparators was 73.5 percent vs. 62.5 percent, respectively. Both these rates are below the industry average of approximately 80 percent. The PigCHAMP swine industry record keeping system farrowing rate average for the first quarter of 2005 was 79.1 percent (Olson⁹⁷ 2005). It should be noted that the PigCHAMP rate included all parities (the number of times a sow has farrowed), and the rate would be expected to be lower for gilts. The lower farrowing rates of the pigs in this study could be attributed to many factors including the feed intake of the gilts as well as the ability of technicians to detect gilts in heat. Farrowing rate is generally more attributable to the female as opposed to the male member of the breeding pair, and takes into account such factors as ovulation rate (number of oöcytes released per estrus), age/parity of the sow, and environmental factors (Hafez and Hafez 2000). Additionally, the rate for the comparator-bred group was low relative to the clone-bred group, which was largely attributable to the gilts bred by one boar whose farrowing rate was 41.7 percent. This boar (Hamline nuclear donor) was reported to be five years old at the time of semen collection (relatively old for a breeding boar), which may have contributed to the low farrowing rate for gilts inseminated by his semen. As the clone-bred group had a higher farrowing rate than the comparator-bred group, and the farrowing rate of the clone-bred group was only slightly lower than industry averages, cloning does not appear to negatively impact AI-based boar reproductive performance.

The mean litter size for the progeny of the clone boars was 10.94 pigs and the median was 11.5 pigs. For the pigs from the comparator boars, the mean litter size was 11.76 pigs and the median was 12.0 pigs. The mean litter size for the U.S. in the first quarter of 2005 according to PigCHAMP records was 10.66 pigs per litter (Olson 2005). Although the comparator boars had a higher mean litter size, both groups were near the U.S. average for litter size. Litter size appeared to be more variable in gilts inseminated by clones: 11 percent of the litters from the clone boars had more than 14 pigs, whereas none of the comparator litters had more than 14 pigs. Further analysis of litter size showed that the clone boars also had a higher frequency of litters with less than ten pigs. Approximately 27.8 percent (10/36) of the litters from the clone boars had fewer than ten pigs in the litter. For the comparator group, only one litter (out of 25 litters or 4 percent of all comparator litters) had fewer than ten pigs. Many factors can affect litter size in pigs including the estrus cycle in which the gilts are bred, genetics, nutrition, management, environment, and ovulation rate of the gilt (Aherne and Kirkwood 1985). Although many of these factors were controlled (nutrition, parity, management, and environment), other factors besides cloning cannot be ruled out as contributors to the wider distribution of litter size for the litters from the clones. Finally, as semen characteristics appear to be similar between the clones and their comparators, the differences in litter sizes were most likely due to gilt or uncontrollable management variation such as breeding cycle or feed intake.

Conclusions for Reproductive Performance of Clones. There were no apparent differences in semen quality between clones and comparator boars. Farrowing rate was slightly higher among gilts bred by semen from clone boars, although the difference between clones and comparators could be traced to the Hamline nuclear donor, which was older than any of the other boars used in this study. Average litter size was similar for clones and comparator boars, and farrowing rate and litter size for clones were similar to national averages. Therefore, reproductive performance does not appear to be affected by the cloning process in these animals.

2. Study 2: Progeny of Clones vs. Comparators

a. Survival

A summary of the survival data from this study is presented in Table F-11. At the start of the study, there were 310 (295 made it to slaughter) live-born clone-derived progeny and 251 (243 made it to slaughter) live-born comparator-derived progeny. The percentage of mummified pigs (dead, desiccated fetuses) at farrowing was 3.3, 2.8, 1.7, and 0 percent for the progeny from Hamline comparator, Hamline clones, Duroc comparators, and Duroc clone boars, respectively. In both breeds of pigs (Hamline and Duroc) the percentage of mummified pigs was slightly higher in the comparator group than in the clone-derived pregnancies. In this study, litters from Hamline clones and comparators and Duroc comparators had higher rates of mummies in their litters than the U.S. average (0.2 percent) (SwineReproNet)⁹⁸; however, the percentage of mummies in the litters derived from the Duroc clone was similar to the U.S. average. A substantial number of pigs were lost around the time of birth, ranging from 17.0 percent-31.4 percent, and in each breed, these losses were slightly higher in the group comprised of progeny derived from clones. Most of these losses were due to the categories of “weakness” or “unknown causes.” Further analysis of the data indicated that an entire litter of 13 progeny from a Duroc clone boar was lost on July 15, 2004, shortly after birth. Reasons provided for the loss of this litter included “unknown” (n=7) and “weak” (n=6). The accompanying animal health records note, however, that between July 13 and 15 sows were stressed due to both high temperature and humidity in the farrowing house. The records also indicate that C- section was performed on one of the heat-stressed gilts, and the gilt and her 13 pigs subsequently died. If these 13 progeny from the one litter that died are removed from the evaluation, the differences in survival rate for progeny from clones and from comparator boars are slight and inconsequential.

The percentage of stillborns among progeny of clones (36/394) and the progeny of the comparator boars (27/294) was the same (9 percent). This level is within the estimates of industry averages for stillborns (range 5 -10 percent (SwineReproNet)). Sixteen of the 36 litters from the clone boars had at least one stillborn pig (44 percent), while 13 of the 25 litters from the comparator boars had at least one stillborn pig (53 percent). Therefore, the frequency of stillborns in litters from the clone boars was lower than the comparator group. The litters of the clone Hamline boars had an average of 0.8 ± 0.3 rate of stillborn per litter, and were virtually identical to the comparator Hamline boars (0.8 ± 0.07). Similarly, the stillborn rate for the Duroc clone boar litters was 1.4 ± 0.4 per litter, while the rate for the comparator Duroc boar litters was 1.1 ± 0.3 stillborns per litter. According to PigCHAMP, these rates are similar to the average stillborn rate of 0.93 pigs per litter for the U.S. swine industry records for the first quarter of 2005 (Olson 2005). Therefore, cloning does not appear to affect the stillborn rate component of reproductive performance of pigs for either genetic line.

One clone progeny pig was destroyed due to a deformity (DDFR). One comparator pig was destroyed due to an injury (DINJ). The destroyed pigs represent 0.25 percent and 0.3 percent for the progeny of the clone and comparator boars, respectively. The deformity of the destroyed clone progeny was not described and may more accurately fit into the “other abnormalities” category. Adding an additional pig to this category for the progeny of the clone boars does not increase the incidence rate to a level that warrants concern (see discussion on abnormalities). Injuries during nursing are usually due to sow overlays. Consequently, the injured pig reported in its own category may typically be included in the overlay category as discussed below. Adding this pig to the overlay category does not substantially change the frequency of overlays and therefore does not warrant changing the conclusion relative to overlays as discussed below.

The rates of abnormalities in both clone and comparator derived progeny were similar to industry observed levels. Progeny of the clone boars had an abnormality rate of 2.5 percent (10/394), including anal atresia (lack of opening of the anus) (1/394, 0.3 percent), spraddle legs (leg weakness) (4/394, 1 percent), and “other” (5/394, 1.2 percent). Three of the 295 offspring of the comparator boars had abnormalities (1 percent), all of which were recorded as having spraddle legs. The incidence and cause(s) of spraddle legs is not well documented or fully understood, but it may involve several factors including genetics, management, slick flooring, mycotoxins, and a virus or combination of viruses (Goodband et al.1997). Furthermore, the incidence of spraddle legs was the same for both groups. Therefore, the occurrence of spraddle legs does not appear to be related to cloning. Of the two other classes of abnormalities observed in the progeny of clones, all have previously been reported in the literature on swine reproduction. The incidence of anal atresia in the general swine population has been reported to range from 0.1-1.0 percent (Wiedemann et al. 2005). Further investigation of the “other” category revealed that these pigs were from two litters. Three pigs from one litter were described as having “typical leg abnormalities possibly associated with overcrowding in the uterus.” These pigs were from a litter of 18 pigs, which would be considered large compared to industry average (Vonnahme et al. 2002). The other two pigs were from another litter, and were described as having short legs. The frequency of miscellaneous abnormalities in newborn pigs has been reported by Spicer and coworkers (1986) as 1.2 percent, and included cleft palate, anal atresia, renal hypoplasia, hydrocephalus and accidental death. Because similar abnormalities have been reported in the swine industry at a similar frequency to that of this study, the rate of “other” abnormalities in this study is not a high concern.

Seven of the 394 progeny (2 percent), all from the same litter (#339) derived from one of the clone boars were disposed of for unknown reasons. Furthermore, of the seven clone-progeny pigs that were categorized as “disposed of because they were weak” (2 percent), six were from the same litter (#339) as the pigs “disposed of for unknown reasons.” Only one other progeny of the clone boars was disposed of for being weak. No comparator pigs were disposed of for unknown reasons or because of weakness. The fact that 4 percent of the pigs from the clone boars were destroyed due to weakness and unknown reasons compared to 0 percent of comparators could be a matter for concern. However, this effect results primarily from observations from one litter, and the daily observation records indicate that heat stress was a problem in the farrowing house when this litter was delivered. Further, given most of the pigs in this litter were disposed on the day they were born, with only two pigs living for three days, it appears that heat stress of the dam and/or pigs may have contributed to the loss of this litter. The consequent disposal of this litter resulted in the relatively high rate (4 percent) of “unknown” and “weak pig” disposals for the progeny of the clone boars. The data do not, however, fully account for the removal and subsequent care of the pigs in this litter, and this interpretation should be considered preliminary. We also note that unknown disposals or weak pig disposals do not appear to be a problem across litters for the clone progeny group, which would be expected if cloning were a primary contributing factor to the incidence of weak pigs in the progeny of clones.

The frequency of overlays (death due to the sow lying on top of the piglets) for pigs in litters from the clone boars and the litters from the comparator boars were 8.2 percent and 5.0 percent respectively. Industry estimates for pig deaths due to crushing by the sow are between 4.8 and 18 percent (Lay et al. 1999). The frequencies of overlays for both groups in this study are near the low end of the estimated range of deaths due to crushing. Crushing has been reported to be related to several factors including the genetics, activity level of the sow and sow housing (Hay et al. 2002). Secondary factors that potentially influenced the number of pigs that die due to the dam lying on pigs included environmental conditions at the time of farrowing and the number of litters that were born on a single day. The data indicate that 315 of the 688 pigs were born on one of six days where the heat index was above 104ºF (40°C). The daily sow/litter observation records indicate that high temperatures in conjunction with poor cooling may have contributed to the number of crushing deaths, as the sows attempted to find a cooler and more comfortable position. The records do not differentiate between the two groups of gilts, and gilt housing was the same for both groups in this study. Therefore, any differences in the number of crushing deaths in this study were probably due to differences in sow activity. Also, the litter observation records indicate normal growth and behavior for all of the pigs and therefore, there is no evidence to indicate an increase in susceptibility to crushing of pigs in the litters from the clone boars.

The total number of disposed pigs (stillborns, destroyed, overlays, unknown deaths and weak pigs) was 21 percent for the pigs from the clone boars and 14 percent for the pigs from the comparator boars. The difference in the two groups is primarily due to the unknown and weak pig disposals and a higher rate of overlays. These categories have been discussed previously. Also, there was a 3.0 percent loss of pigs post-weaning in pigs derived from Duroc comparators and a 1.8 percent loss of pigs obtained from Hampshire clone boars. Based on these data there is no evidence to suggest that progeny of clones are at increased risk for mortality compared to AI-derived pigs.

Initial plotting of the relationship between age and body weight suggested that, although the mean length of time to slaughter was similar for progeny within the four groups, there might be a broader range in the length of time that it took to reach slaughter weight. These data indicate that there was a similar range, from 144 to 210 days, in the length of time that it took for progeny from conventionally bred and clone boars to attain their slaughter weight. Furthermore, the data indicate that there were only small differences in ADG between the four groups of pigs between birth and slaughter or weaning and slaughter. When ADG was calculated over 4 week intervals there were no significant differences prior to 20 weeks of age. The data indicated that slightly less than 50 percent of the progeny from each of the four groups was slaughtered prior to week 24 and that only 27 animals remained on the experiment at week 28. The small differences in percentage of animals could be attributed to the small number of animals remaining in the experiment at this point than to any one of the groups.

Because of its retrospective survey nature, and its smaller size, the Cyagra dataset was evaluated using a slightly different procedure (see description in Appendix E). In that case, values from approximately breed- and age-matched comparators were used to establish a range against which values obtained from clones were compared (The Cyagra approach). To determine how outcomes would differ if the Viagen data had been analyzed by the same method used to evaluate the Cyagra dataset, both approaches were applied to the clinical chemistry and hematology variables from Viagen Study 2. The following tables (Tables F-12a, F-12b and F-13) summarize the proportion of values that were identified as outliers for further examination by both procedures. The boxplot procedure identified a slightly higher percentage of values for further examination than the procedure used to describe the Cyagra data. However, the boxplots identified similar percentages of outlier values for progeny of comparators as for progeny of clones.

b. Growth, Hematology, Clinical Chemistry, and Urinalysis

i. Growth

The birth weights of clone and comparator progeny were similar. To establish a base population for the comparator group of pigs (n = 267), mummified and stillborn pigs were excluded. A birth weight outlier analysis of the progeny of the clones and comparator boars (n=617) indicated that two progeny of clone boars were lightweight outliers (0.59 and 0.64 kg). One comparator pig was a lightweight outlier (0.41 kg) and one comparator pig was a heavyweight outlier (2.31 kg). The boxplots for birth weights showed a similar distribution for the progeny of the clones and comparator boars. Additionally, the mean birth weight for the offspring of the clone boars was 1.5 kg and the mean for the comparator group was 1.45 kg. The median for both groups was 1.5 kg. The similarity in birth weights, the birth weight distributions and the low frequency of outliers between the progeny of the clone boars and those of the comparator boars indicate that birth weight is not a health concern for progeny of clones.

The progeny from both clones and comparator boars had similar nipple counts with similar distributions of the counts. Nipple counts are important because they are genetically transmitted and indicative of the number of offspring a female can feed. Most of the pigs in this study had between 12 and 16 nipples (95.7 percent and 98.4 percent for the clones (389/394) and comparators (282/294), respectively). The industry standard is for at least six functional teats per side (total 12) of the underline of a gilt (Ahlschwede and Kuhlers 1992). The progeny of the clone boar group had no pigs with fewer than 11 nipples while the progeny of the comparator group had three pigs with 10 or fewer nipples. Nipple counts for the progeny of clone boars are therefore within normal bounds of these pig populations.

Growth characteristics of these animals were also analyzed, with the data evaluated for potential outliers prior to statistical analysis or plotting. Forty-two outliers were identified in the dataset containing body weight measurements (n =2,966). Six of these outliers represented body weights taken prior to the death of 4 animals. Nine of the outliers were data points associated with animals derived from both clone and comparator sires that lost weight near the end of the experiment. No reason for the weight loss was provided, and these animals were excluded from the final analysis. The mean slaughter weights were, 121.7, 119.6, 120.2, and 121.9 kg for progeny derived from the Hampshire comparators, Hampshire clones, Duroc comparators, and Duroc clones, respectively. Body weights at the time of slaughter ranged from 108.8 to 134.7 kg for progeny from the Hampshire comparator boar, 97.4 to 135.7 kg for progeny from Hampshire clones, 97.0 to 136.2 kg for progeny from Duroc comparator boars, and 108.0 to 137.0 kg for progeny from the Duroc clone boar. The mean number of days from birth to slaughter was 173.7, 174.7, 168.2, 175.5 days for progeny obtained from Hampshire comparator, Hampshire clones, Duroc comparators, and the Duroc clone, respectively.

Tabulation of the calculated average daily gain (ADG) for progeny from clones and comparators at various time points throughout their lives is shown in Table F-15. Average daily gains from birth to slaughter were 0.69, 0.68, 0.71, and 0.69 kg/day for progeny derived from Hampshire comparator, Hampshire clones, Duroc comparators, and Duroc clone boars, respectively. Similar finds were observed for ADG between weaning and slaughter (0.76, 0.73, 0.77, and 0.74 kg/day, respectively). Body weights were also measured every 4 weeks with only small differences between the progeny of clones and comparators.

ii. Hematology, Clinical Chemistry, and Urinalysis

First Blood Sampling⁹⁹ (July). The results of blood clinical chemistry and hematology for July, when the progeny were between three and 30 days old, are in Charts F-1 and F-2, respectively. More that 94 percent of these values showed no differences between the progeny of clones and comparators. We identified the following variables to be of interest for clone progeny: hemolysis, lipemia, percent saturation, ALT, AST, CK, Lipase, SDH, LUC, MCHC, and MPV. This is because greater than 5 percent of the clone values were outside the comparator range and therefore would be more that we would expect by chance. In approximately 5 percent of the clone progeny, values for hemolysis, lipemia, percent saturation, ALT, AST, CK, Lipase, SDH, LUC, MCHC, and MPV were outside the comparator range. To determine whether these values indicated concerns for the health of the animals, we compared the amount of variability between the clone progeny and the comparator progeny groups. Following that analysis, we determined that the values for hemolysis, percent saturation, ALT, SDH, LUC, MCHC, and MPV had a similar amount of variability and did not warrant any further concern. The values for lipemia, AST, CK, and lipase were out of range more frequently for clone progeny and required further consideration.

Hemolysis and lipemia can be considered artifacts based on sample handling or drawing samples from animals that have been fed recently, respectively, and can have a significant adverse impact on the quality of other blood data values (Duncan and Prasse 2003). Hemolysis, either from poor collection technique, age of the sample, or poor handling of the sample once in the laboratory, has a serious effect on many blood chemistry tests, including dramatic effects on the enzymes alkaline phosphatase, GGT, and CK. In general, it is advisable to draw a new sample if gross hemolysis is noted. Lipemia (a measure of the amount of fats in the blood) can become elevated if animals have their blood drawn shortly after eating. The increased levels of lipids can falsely raise Na, K, and Cl levels and artifactually lower AST and ALT levels (Shanahan 2004). For that reason, lipemia is not considered a health related variable in food animals (Duncan and Prasse 2003). Creatine kinase (CK) is an enzyme found in muscle tissue and to a lesser extent in liver cells, and elevations are often indicators of muscle injury (including muscle damage during venipuncture) or hemolysis (Duncan and Prasse 2003) (See Appendix E). It is unclear if the small elevations that were noted were due to those reasons, or injury near the time of handling. AST is an enzyme found in liver cells and muscle cells. It has a longer half-life than CK. Although eight more clone progeny (19/242 or 7.8 percent) had elevated AST compared to the non-clone progeny (11/163 or 6.8 percent), neither the level of increase nor the number of animals with increased levels were sufficiently high to indicate a real biological difference. Lipase is a pancreatic enzyme that breaks down fat; its elevation can indicate pancreatic inflammation (pancreatitis). A 2-3 fold increase, however, is considered the threshold for further evaluation to determine whether pancreatic inflammation is responsible. (Duncan and Prasse 2003). The upper level is 93.5 U/l and average elevation is 132 U/l. In pancreatitis, there also is usually an increase in another enzyme (amylase) to corroborate the condition (Duncan and Prasse 2003). Amylase values in this study are within range. As values in this study represented less than a 2 fold increase, we did not consider them biologically relevant. We therefore concluded that there are no biologically relevant differences in blood values between clone progeny and comparator progeny at this point in their development.

Second blood sampling (October). The results of blood clinical chemistry and hematology for October, when the animals were approximately 12 to 15 weeks old, are in Chart F-3 and F-4, respectively. In comparing the clone to non-clone progeny variability, the CK and basophil values were similar enough to conclude no difference. For the chemistry values in clones, indirect bilirubin/total bilirubin and bile acids had more outliers compared to non-clone progeny. Bilirubin is a breakdown product from the hemoglobin of senescent (old) RBCs. The liver processes this by conjugating the product to a salt and making it water soluble. Elevations in bilirubin can indicate reduced hepatic function (Duncan and Prasse 2003). The clones in this study had low bilirubin. There is no known cause for low bilirubin. For this reason, this finding was considered not clinically relevant. Bile acids were elevated in 25 clone progeny and 13 comparator progeny. Bile acids can be artifactually elevated in response to eating. This value may also indicate some hepatic insufficiency (insufficient number of liver cells to perform the metabolic functions of the liver). If the liver were adversely affected, we would expect to find other corroborating analytes to confirm this possibility. There are no other analytes to confirm hepatic insufficiency in these animals.

Hematology values for basophils for progeny of clones and comparators were similar enough not to warrant further discussion. More clone progeny had lower MCH (mean corpuscular hemoglobin) and MCV (mean corpuscular volume) values and more clone progeny had higher RBC values than the comparators. RBCs are elevated in 12 clone progeny and only 2 comparator progeny. Elevations in RBC can be from excitement (splenic contraction), hydration status (dehydration causes an increase in RBC), or an absolute polycythemia (true increase in production). The cause here is unknown but does not seem to indicate a health problem. MCH is a value derived by dividing hemoglobin by the RBC number. Because the RBC number is high, the MCH must be low. A decrease in mean MCV can mean an iron deficiency. This is usually accompanied by anemia. Anemia is defined as a reduced number of RBCs or decreased Hematocrit/ Packed Cell Volume (PCV). Because there was an increase in these values, its significance is minimal.

Third blood sampling (January). The results of blood clinical chemistry and hematology for January, when the animals were approximately 24 weeks old, are in Charts F-5 and F-6, respectively. Values with similar variation between clone and comparator progeny are Mg++, LUC, and RDW. These values require no further discussion.

Sodium:potassium (Na:K) ratio is a value derived from the sodium concentration and compared to the potassium concentration. Neither Na+ nor K+ were significantly different. Their ratio is used to determine adrenal function (to detect Addison’s Disease) in small animals. Its significance in pigs is not listed in clinical pathology texts as being clinically relevant (Duncan and Prasse 2003). ALT, AST, BA, CK, and SDH are analytes with significance for liver and muscle tissues. Because we have seen elevations in these enzymes before and discussed them above, we decided to determine if there was clinical relevance to the increase in clone progeny. As discussed, BA may be increased depending on when the blood was drawn in relation to a meal. Elevations in values for analytes with significance for liver and their effect on body weight are discussed in the next section. Daily health observations were not available. As stated in Appendix E and elsewhere, one can only evaluate lab tests in the context of a complete clinical picture.

In the January hematology, hematocrit and RBCs were elevated in progeny of clones. Elevated hematocrit and RBC values are rarely an adverse health issue. The MCHC (mean corpuscular hemoglobin concentration) had as many high values as low values, which reduces its significance as an indicator of a health problem. The clone progeny had 21/138 (15.2 percent) animals with elevated segmented neutrophils (segs) vs. 4/84 (4.8 percent) for the comparator progeny. Segmented neutrophils are elevated in response to bacterial exposure. With no daily clinical health observations, it is difficult to interpret this observation. However, this may be an appropriate response to some challenge in the pig’s environment because no other analytes indicative of active infection (over all white cell count, banded neutrophils, globulin) are elevated.

No differences in the levels of IGF-I (Fig.1), and E2 (Fig. 2) in progeny of swine clones versus comparator animals were found at slaughter. Male progeny of animal clones and their comparators have similar levels of E2. The levels of E2 were slightly, but not significantly, diminished in the female progeny of clones vs. the comparators. This minor decrease was considered to be part of the normal variation in blood levels that may occur depending on reproductive status of the female and time of day. The reproductive status of the female animals was not provided, however, these animals were slaughtered at approximately six months of age, when swine are generally still pre-pubertal. Nevertheless, the levels of E2 in the progeny of swine clones do not differ significantly from the comparators.

Figure F-4 depicts the levels of IGF-I in the offspring of swine clones and comparators (female and males) at slaughter. CL=offspring clones, F=female, M=male. Values are mean ± SEM, the numbers above the bars = number of samples (animals) per group.

Figure F-5 depicts the levels of Estradiol-ß in offspring of swine clones and comparators (female and males) at slaughter. CL=offspring clones, F=female, M=male. Values are mean ± SEM, the numbers above the bars = number of samples (animals) per group.

Urinalysis. Urine samples were harvested after slaughter, which may account for the presence of blood in some samples. Only one clone progeny had protein in its urine. This is not unrealistic by random chance and not a health issue.

The distribution of pH values is similar between the clone progeny and the comparator progeny and indicates normal urine variation. This is especially true for animals on an herbivorous diet which is typical of current swine management practice (Duncan and Prasse 2003). No animals in this experiment had glucose in the urine.

Conclusions for Animal Health of Progeny of Clones. Although there was a higher death loss among progeny of clones in this study, most of this loss can be attributed to a single litter farrowed by a heat-stressed sow that did not survive. Causes of death (e.g., stillbirth, overlay, weakness) were similar to national statistics for commercially raised swine, and there were only minor differences between groups. Few animals were noted with abnormalities in either group, and the rates and types of abnormalities were similar to national statistics for commercially raised swine. Growth rates from birth to weaning for progeny of clones and comparators were similar. Differences were noted in both the early (neonatal) and mid-trial (early juvenile) blood values between progeny of clones and comparators in Study 2. The differences during the neonatal period were few and minor. The clone progeny values were considered to be within the range of variation for a normal population of neonatal animals. There are some differences between the clone progeny and comparator progeny during the second blood sampling (early juvenile period). The values for analytes with significance for liver for this second sampling period offered mixed results, none of which confirm liver abnormalities. The blood cell values for this second sampling are also inconsistent, offering no indication of blood cell abnormalities. There are increases in liver-function associated analytes in this dataset (late juvenile period). The other values indicate no negative health impact on progeny of clones.

c. Carcass Characteristics

Given the large variation in live weight at the time of slaughter, one might anticipate that many of the post-slaughter carcass characteristics, such as marbling and backfat thickness, would also vary considerably. Carcass characteristics are provided in Table F-17. Hot carcass weights were 79.9, 79.4, 79.0, and 81.2 kg for progeny from Hamline comparator, Hamline clones, Duroc comparators, and Duroc clone boars, respectively. Carcass length was also similar, 82.7, 81.6, 82.3, and 81.5 cm, respectively. The first rib values were 22.2, 23.4, 23.8, and 25.9 mm for progeny from Hamline comparator, Hamline clones, Duroc comparators, Duroc clone boars, respectively, whereas the last rib values were 16.0, 16.9, 17.4, and 19.0 mm, respectively.

All animals were given score 3 for carcass muscle score and 2 for firmness. Ninety-three percent of the progeny from the clones and comparator boars had marbling score within the 2 to 4 range. Measurements of pH at 24 hours post-slaughter on the longissimus muscle were similar. Loin eye area for meat cuts for progeny from Hamline comparator, Hamline clones and comparators, as well as Duroc comparators and clones were also very similar. In summary, all of the carcass characteristics evaluated were similar between the offspring of clones and comparators.

d. Meat Composition from the Progeny of Clones and Comparators

Table F-18 provides the comparison of key nutrients between the progeny of clones and their comparators. Data were reported for 412 swine of which 242 were the progeny of clones and 163 were the progeny of comparator boars. The primary comparison was made between the reported nutrient concentrations of these two groups. A secondary comparison was made to reference swine muscle values currently in the food supply (USDA Food Composition Data for pork, fresh, composite of trimmed retail cuts (loin and shoulder blade, separable lean and fat, raw), USDA National Nutrient Database for Standard Reference Release 18)). The latter comparison is less tightly controlled than the comparison with the comparator, largely due to the differences in cuts, and the unknown nature of the breed(s) of swine used in the USDA dataset.

The composition of the meat from the progeny of clones and comparators indicates that the meat samples were indistinguishable at the level of the key nutrients evaluated. Only two values (alanine and erucic acid) of 56 (0.04 percent) were not virtually identical, less than would be expected by chance alone. Neither of these differences is biologically significant.

Comparing the meat composition of either the progeny of clones or the comparators to the USDA values reveals that neither is as closely comparable to that dataset as they are to each other. For example, values for niacin and vitamin B12 from the progeny of both clones and comparators were higher than the USDA values for a similar type of swine muscle (shoulder blade and loin), while virtually identical to each other. Other nutrients that differ from the USDA database include palmitic acid (16:0), palmitoleic acid (16:1), stearic acid (18:0), oleic acid (18:1), linoleic acid (18:2), arachidonic acid (20:4), and niacin. The levels of these six fatty acids are higher in the database than in the progeny of clones and comparators. Little variability was observed between the other values in the nutrient profiles of the USDA database and those obtained from the progeny of clones and comparators. The differences between the nutrient concentrations in progeny of clones and comparators compared to USDA database may be due to diet, swine genotype, or storage stability effects. The important conclusions from the two comparisons, however, are that (1) there are virtually no differences between the progeny of clones and comparators, (2) the closely genetically related comparators are a better reference point than the USDA data base, and (3) none of the differences pose a food safety concern. These data suggest that there is no increased risk for humans to consume muscle from the progeny of swine clones.

Conclusions for Carcass Characteristics and Meat Composition for Progeny of Clones and Comparators

Although some minor differences in backfat thickness were noted for progeny of clones vs. comparators, they have no significance for food safety. The increased values for niacin and B12 and decreased values for six fatty acids compared to USDA values were similar for progeny of clones and comparators in this experiment, and may reflect differences in diet, genotype, or sample handling compared to the national average. Because these values were similar between the two groups involved in this study, there is no increased risk associated with meat from progeny of clones vs. contemporary comparators. All other meat composition values were similar between groups, indicating no increased risk associated with meat from progeny of clones vs. contemporary comparators.

D. Conclusions from the Viagen Dataset

1. Study 1: Clones vs. Comparators

a. Animal Health

The interpretation of the results of the study comparing conventionally bred and clone barrows is limited because it was not initiated until the animals were approximately 50 days of age, clones were raised under different conditions prior to initiating the experiment, and there were a limited number of clones. The conventionally derived barrows were selected based on two criteria: (1) their sire was one of the donor boars for SCNT; and (2) the pigs were similar in age and weight to the clone barrows. Retrospective evaluation of the birth weights for the seven clone barrows indicated that these animals were smaller at birth than their conventionally bred counterparts. The growth rate data would suggest that clone barrows grew as well as conventionally bred animals prior to weaning, as these animals reached the same body weight at around 50 days of age. However, after the clones were moved to the more conventional rearing facility to be raised with the AI-derived barrows, it took the clone barrows on average 27 days longer to reach their slaughter weight, and the clone barrows were on average 18.2 kg lighter when they were slaughtered than the conventionally bred barrows.¹⁰⁰ Furthermore, three of the seven clone barrows were not processed at the end of the experiment. One of the clones was condemned at slaughter due to a lung adhesion and the other two animals were approximately 45 kg lighter than their counterparts. The health records for the clones indicate that these animals developed several health issues including scouring. These clones were born and maintained under highly biosecure conditions until the beginning of the study (at approximately 50 days of age), were potentially premature at delivery, and were deprived of colostrum. Thus, moving these animals to a conventional production system could have had a dramatic effect on their growth rate. However, four of the seven clone barrows responded appropriately and overcame the pathogenic challenge.

b. Food Safety

The most significant difference between the comparator and clone barrows at slaughter was a trend for higher backfat thickness in the conventionally bred animals, consistent with the observation that the conventionally bred animals were heavier at the time of slaughter than clones.

Data were presented on the key nutrient levels of latissimus dorsi muscle from swine clones and comparators. Fifty-six nutrients were measured in tissue samples of five clones (one animal was euthanized for health reasons) and 15 comparator swine. There were limitations in the usefulness of the data due to the study design and data reporting. A comparison of reported values to reference swine muscle values (USDA National Nutrient Database for Standard Reference Release 18) was possible for only four nutrients because the other nutrients were reported as percentages. Values for niacin and vitamin B12 in both clones and comparator swine were above USDA values for a similar type of swine muscle (shoulder blade and loin), but were similar to each other. Values for cholesterol and vitamin B6 were similar to the USDA values. Little variability between reported values was observed when the data were examined by nutrient.

The lack of variability observed in the food composition values between muscle of swine clones and comparators supports a conclusion that there is no additional risk in the human consumption of muscle from swine clones. However, limitations in the study design, reporting of data, and the elevated niacin and vitamin B12 concentrations in both clone and comparator muscle compared to the USDA reference values diminish the confidence of this conclusion. A more definitive comparison between the food composition of clone and comparator swine muscle could be made if more of the analyses could be compared to values in reference (USDA) swine muscle.

2. Progeny of Clone Boars vs. Progeny of Comparator Boars

This experiment was designed to determine whether progeny from clones performed as well as progeny from comparator boars and if food products from progeny of clones would pose any additional risk relative to corresponding products derived from comparator animals. Data were provided on 300 and 402 progeny derived from comparator and clone boars, respectively.

a. Animal Health

Although there was a higher percentage loss of pigs at birth for progeny derived from clone boars, this difference was primarily due to the loss of an entire litter of 13 pigs. Secondary factors that potentially influenced the number of pigs that died due to the dam lying on pigs included environmental conditions at the time of farrowing and the number of litters that were born on a single day. Although there was a slight increase in the percentage of progeny of clones that were crushed by their dam, there was a similar number of pigs/litter crushed by their dams among comparators, and these values were similar to values that are commonly found within the swine industry. Thus, there do not appear to be any differences in the survival of progeny from clone boars when compared to progeny from conventionally derived boars. There was a similar percentage of mummified pigs presented at birth. The survival and growth rate data do not show any animal health concerns for progeny of clone boars when compared to progeny from conventionally derived boars. Hematology, clinical chemistry and urinalysis values for clone progeny were considered to be within the range of variation for a normal population of animals.

b. Food Safety

Data were presented on the food composition of latissimus dorsi muscle from the progeny of swine clones and controls. Fifty-eight nutrients were measured in the tissue samples of 242 AI-derived comparators and 163 clone progeny. A positive aspect of the study is the large numbers of test swine and the numerous nutrients analyzed. Negative aspects of the study are the lack of method performance data in the test matrix (meat), the choice of latissimus dorsi as the test matrix instead of a retail pork cut and/or another edible tissue, and the lack of storage stability data.

The nutrient concentrations of clone progeny and comparator swine are very similar. A few nutrients did have differences in the variability and distribution of values between clone progeny and comparator swine. We evaluated the differences, and determined they were minor and not biologically significant.

Most of the nutrient concentrations were similar to USDA reference values. Six fatty acids had lower concentrations in both the clone progeny and comparator swine compared to the USDA values. One B vitamin was higher in the clone progeny and comparator swine than the USDA value. The difference between these nutrient concentrations in clone progeny and comparator progeny compared to USDA values may be due to effect of diet, swine breed, or storage stability effects on method performance.

Based on the lack of difference in the nutrient concentrations between muscle of progeny from AI-derived comparator and clone boars, we conclude that there is no increased risk for humans to consume muscle from the progeny of clone swine. The current study provides no information regarding the food composition of other swine edible tissue (liver, kidney, fat). Therefore, food safety conclusions about muscle cannot be extrapolated to other edible tissues of swine.

E. Addendum

On January 5, 2006, Viagen Inc. faxed several pages of data from the re-assay of samples which they had identified as outliers using the criteria outlined for the Cyagra dataset (Appendix E). Samples from 15 clone progeny were re-assayed because the values for specific nutrients were > 10 percent above or below the range of values for comparators. Table F-20 provides a comparison of the original and re-assay values for the nutrients assayed by animal.

CVM conducted a follow-up analysis of the data using the new values and found they had only very minor effects on the average nutrient values for clone progeny. The reanalyzed means are presented in Table F-21, and compared to the original means for progeny of clones and comparators.

Of the 22 nutrient values that were reanalyzed, 12 means were unchanged compared to the original values. Nine values were changed (0.01 g/100g) compared to the original values. Only the mean for Vitamin B12 differed by more than 0.01 g/100 g; however, the reanalyzed mean was similar to the mean for comparators (0.93 ± 0.34 vs. 0.97 ± 0.28 g/100 g). All 22 values were similar to values for comparators. None of these minor changes affect the assumptions regarding the safety of meat from clone progeny.

Table and Charts (pdf)

⁸¹ A boar is a reproductively intact male pig.

⁸² “Hamline” refers to a specific crossbred line of swine used by Viagen, Inc. This line was developed by crossing various breeds, including Duroc, European Landrace, Pietran, and European Large White swine.

⁸³ A barrow is a castrated male pig.

⁸⁴ Although this study is a more controlled experiment than the retrospective review of the Cyagra clones, the word “comparator” is used for consistency with the discussion of Cyagra data, rather than the more common term “control.”

⁸⁵ A gilt is a young female pig that has not given birth (termed “farrowing”).

⁸⁶ “Biosecure” connotes conditions that, although not completely sterile, are designed to prevent introduction of most disease-causing organisms. Because animals’ immune systems produce antibodies in response to specific stimuli (antigens), animals raised under biosecure conditions lack the full battery of antibodies needed to respond to common pathogens found in conventional environments (although they may have some antibodies if they have been vaccinated, these antibodies will be specific for the antigen used in the vaccine). It can take up to two weeks for an animal to develop a sufficient concentration of antibodies to a specific antigen following exposure.

⁸⁷ Swine are typically slaughtered in groups. Usually, all swine raised together in a pen will go to slaughter on the same day once the average weight of the animals within that pen has reached a pre-set target weight. The date of slaughter can be affected by the availability of transportation and the capacity of the slaughter facility.

⁸⁸ A percentile is defined as (1/(total sample size))*100

⁸⁹ Dressing percentage refers to the proportion of the live animal that ends up as the carcass used for food production, or Carcass Weight / Live Weight X 100.

⁹⁰ http://www.animalrangeextension.montana.edu/Articles/Swine/2-Det-quant.htm

⁹¹ http://ars.sdstate.edu/AnimalEval/Swine/SwineGrade.htm

⁹² The ductus venosus carries oxygenated blood from the umbilical vein to the caudal vena cava, bypassing the fetal liver.

⁹³ The foramen ovale is a hole between the right and left atria in the fetal heart, for the purpose of by-passing the fetal lungs in utero.

⁹⁴ The ductus arteriosis carries oxygenated blood from the main pulmonary artery to the descending aorta, bypassing the fetal lungs.

⁹⁵ http://www.rashaduniversity.com/mrashad/neonphys.html

⁹⁶ Hot carcass weight refers to the weight of the eviscerated carcass post exsanguination, but prior to chilling.

⁹⁷ www.pigchamp.com/overview5.asp

⁹⁸ http://www.traill.uiuc.edu/swinerepronet/paperDisplay.cfm?ContentID=6266

⁹⁹ Appendix H has detailed descriptions clinical chemistry values and what they measure.

¹⁰⁰ As mentioned previously, because animal clones are voluntarily withheld from entering the food supply, the slaughter facility could only make certain dates available to slaughter these animals. Thus, the clones were slaughtered as a group on one of the pre-selected, available dates once their average weight approached the target weight.

	CVM A-Z Index | Contact CVM | About CVM | Site Map FDA Home Page | Search FDA Site | FDA A-Z Index | Contact FDA | Privacy | Accessibility | HHS Home Page