Comments from Peer Reviewers on Cloning Draft Risk Assessment

Comments from Peer Reviewers on the Draft Risk Assessment

Peer Reviewer 1

A Peer Review of:

Animal Cloning: A Draft Risk Assessment
Center for Veterinary Medicine
U. S. Food and Drug Administration
Department of Health and Human Services
7500 Standish Place
Rockville, MD 20855

Draft dated 02/14/2006

FORMAL CHARGE

Goal. The goal of this Peer Review is a written report reflecting the independent view of each peer reviewer that will be a thorough and meaningful assessment of the agency’s work product. When an agency generates a scientific assessment, it is presenting its scientific judgment about the accumulated evidence. The Peer Review report should distinguish scientific facts from professional judgments and provide input on the reasonableness of judgments made from the scientific evidence. The result should be an independent determination by each peer reviewer as to the appropriateness of (a) the assumptions made and hypotheses postulated, (b) the methodology utilized, (c) the quality and relevance of the data and information, (d) the accuracy of the analytic results, and (e) whether the conclusions reached are supported.

PEER REVIEW

Answers to the Specific Technical Questions.

Has the risk assessment adequately discussed whether cloning introduces any unique risks to either animal health or the consumption of food from clones or their progeny relative to current agricultural practices?

The discussion in Chapter IV, especially in section C. Implications of Epigenetic Reprogramming for Animal Health and Food Consumption Risks, quite adequately addresses the risks and the implications posed by the clones and their progeny to animal health and food composition relative to current agricultural practices. One of the most important aspects of this discussion is the issue of epigenetic normality and it effects. This has been well summarized in the following text, “It may be, as many have suggested (Wilmut 2002; Jaensich et al 2004), that no clone is completely “normal” with respect to its epigenetic profile. Although this is an important point for assessing the overall safety of the cloning process for any particular species, the relevance of “epigenetic normality” to food consumption risks is unclear. Further, because similar abnormalities have been noted in animals produced using other ARTs, the issue of defining normality becomes significantly more complex. It may be that normality encompasses a range on a continuum, and that animals that are healthy, meet appropriate developmental and behavioral milestones, and reproduce to bear healthy young are “normal”, regardless of their epigenetic status. The most compelling conclusions that can be made about food consumption risks, then, are drawn from assessments of the health status of the animals and the composition of food products derived from them, and not from gene expression studies.”

With regard to the progeny of clones, the draft points out that, “Progeny of animal clones, on the other hand, are not anticipated to pose food safety concerns, as natural mating resulting from the production of new gametes by the clones is expected to reset even those residual epigenetic reprogramming errors that could persist in healthy, reproducing clones (Tamashiro et al. 2002; Yanigamichi 2002; NAS 2003, Fulka et al. 2004).” This aspect is supported by research data but needs to be monitored in cloned livestock. Further, it may be necessary to establish the long term epigenetic effects on the progeny and grand progeny of clones.

Spelling above: Yanigamichi should be Yanagamachi.

Specific Comments: Chapters I-IV

Page 3: Cloning is also NT with embryonic cells. Paragraph 1.

Page 10: Edible products from perinatal clones may pose very limited human food consumption risks? Evidence?

Page 22: “boar semen does not freeze well?” This is not completely true as great strides have been made in boar semen freezing methodologies and litter sizes now approach those available with fresh semen.

Page 23: In the US, AI is performed mainly by technicians and producers not by veterinarians.

Page 26: Oocyte maturation times for swine are longer approximately 42-48 hours.

In section 4, I would not use pierce the ovum but penetrate the zona and fuse with the ovum.

Page 30: Batchelder, 2005 is Batchelder et al., 2005, this mistake occurs many times in Chapter V as well as throughout the document.

Page 41-43: This is a good overview of risk assessment.

Page 43: In the section, “ (2) Determining the degree to which existing data address questions of animal health or food consumption risk.”, the conclusion is correct “few reports directly addressed food safety.”

Page 43-44: One problem is that only information published in, “peer-reviewed journals, or otherwise made available to the agency by companies engaged in cloning, with explicit permission for release to the public.” , was used in the risk assessment. This is a potential dilemma as the public maybe unaware of all the potential associated risks because they do not have access to all pertinent information.

Page 44: The section, “Characterizing residual uncertainties persisting following a review of the existing data.” is on target and appropriate.

Page 44-45: It is appropriate to consider transgenic clones separately in the risk assessment because of the different risk space because of the transgenic event.

Page 45: The lack of data on food safety is problematic and therefore must be indirectly assessed “the health status of the animals producing food would have to contribute to both the animal health and food safety components.” This is acceptable if the datasets are appropriately analyzed.

Page 49: The draft report states, “In general, the Center has relied on integrated physiological measurements to survey animal health, although it is likely that genomics, proteomics and metabolomics will see increased use for such purposes in the future. At the time that this risk assessment was prepared, however, these methods had not been sufficiently developed and validated to allow them to be used as survey tools.” The CVM has not used genomics, proteomics and metabolomics to assess animal health risks. They have held others to this standard why is it not being used in this analysis? I understand why these have not been used but a specific set of criteria and standards need to be established by CVM for risk assessment in clones and transgenics.

Page 50: The “Two-Pronged Approach to Assessing Food Consumption Risks” is appropriate.

Page 66: The following statement is appropriate and accurate, “The most compelling conclusions that can be made about food consumption risks are drawn from assessments of the health status of the animals and the composition of food products derived from them, and not from gene expression studies.”

Page 89: This is a significant conclusion, “The most important implication of the mouse clone literature for domestic livestock clones is the observation that anomalies noted in clones are not transmitted to their progeny.”, borne out by the experimental evidence. And supports with experimental evidence, the following statement in the draft, “It is also consistent with the limited but consistent observations of healthy, fully functional progeny born to domestic livestock clones. Thus, the empirical evidence supports the assertion that “Progeny of animal clones, on the other hand, are not anticipated to pose food safety concerns, as natural mating resulting from the production of new gametes by the clones is expected to reset epigenetic reprogramming errors that could persist in healthy, reproducing clones” (NAS 2002a).”

Page 100: I do not believe that the following statement is true, “Relatively few studies have been published on embryo and fetal loss in livestock animals bred by AI or natural service.” The following are references that address this issue:

Pope GS, Hodgson-Jones LS. Use of plasma progesterone levels in an assessment of embryonic loss in dairy cattle. Vet Rec 1975;96: 154.

Kummerfeld HL, Oltenacu EA, Foote RH. Embryonic mortality in dairy cows estimated by nonreturns to service, estrus, and cyclic milk progesterone patterns. J Dairy Sci. 1978 Dec;61(12):1773-7.

Bulman DC. A possible influence of the bull on the incidence of embryonic mortality in cattle. Vet Rec 1979;105: 420-422.

Bulman DC, Lamming GE. The use of milk progesterone analysis in the study of oestrus detection, herd fertility and embryonic mortality in dairy cows. Br Vet J 1979;135: 559-567.

Collier RJ, Beede DK, Thatcher WW, Israel LA, Wilcox CJ. Influences of environment and its modification on dairy animal health and production. J Dairy Sci 1982;65: 2213-2227.

Smith MF, Nix KJ, Kraemer DC, Amoss MS, Herron MA, Wiltbank JN. Fertilization rate and early embryonic loss in Brahman crossbred heifers. J Anim Sci 1982;54: 1005-1011.

Sreenan JM, Diskin MG. Early embryonic mortality in the cow: its relationship with progesterone concentration. Vet Rec 1983;112: 517-521.

Pope WF. Uterine asynchrony: a cause of embryonic loss. Biol Reprod 1988;39: 999-1003.

Shore LS, Rios C, Marcus S, Bernstein M, Shemesh M. Relationship between peripheral estrogen concentrations at insemination and subsequent fetal loss in cattle. Theriogenology 1998;50: 101-107.

Dunne LD, Diskin MG, Sreenan JM. Embryo and foetal loss in beef heifers between day 14 of gestation and full term. Anim Reprod Sci 2000;58: 39-44.

Santos JE, Thatcher WW, Chebel RC, Cerri RL, Galvao KN. The effect of embryonic death rates in cattle on the efficacy of estrus synchronization programs. Anim Reprod Sci. 2004 Jul;82-83:513-35.

Has the Risk Assessment adequately identified the hazards and characterized the risks relating to animal health?

The discussions in Chapter V, quite adequately addresses the risks and the implications posed by the clones and their progeny to animal health relative to current agricultural practices. One of the most important aspects of this discussion is the limited data that is available and the need for further collection and analysis of samples from clones of all stages of development, physiological conditions (growing, gestating, lactating, senescence) as well as diverse chronological ages. This is an aspect that needs to be addressed by funding agencies and made a priority by regulatory agencies.

The assessment indicates that it appears from the analysis that regardless of species, animal clones during the juvenile to prepubertal age do not show increased morbidity or mortality over animals produced by natural mating or assisted reproduction. The data also supports the conclusion that the majorities of these animals that survive the neonatal period grow and develop normally. The data also support the conclusion that reproductively mature cattle clones show no incidence of increased health risk. But again, the data set is limited. No conclusions could be drawn regarding an analogous situation in sheep, goat and swine clones due to insufficient data.

There is not sufficient data to assess adverse health risks in mature and breeding age clones. The available data suggests "that there are no apparent risks to the health of maturing animals from cloning." Empirical observations have indicated that there is no increased risk of health problems in the progeny of clones as compared to conventionally produced animals.

Specific Comments: Chapter V

Page 113: The CVM has adequately identified that the lack of complications in goats may be due to small sample sizes. More data on goat cloning needs to be obtained.

Page 113: The summary: "Summary of Outcomes Noted During the Pregnancy and Parturition Node (Developmental Node 1) Studies performed to date indicate that health problems observed in pregnancies carrying animal clones are not unique; similar problems are well documented in pregnancies produced by IVP and ET, and the same birth defects are sometimes seen in animals that are naturally bred. Based on a review of the literature, the SCNT process in cattle and sheep is associated with increased incidences of early pregnancy loss or later-term spontaneous abortion of clone embryos and fetuses. Early embryo loss seems to be related to in vitro culture conditions, which may cause abnormal development and early embryo/fetal death in both SCNT and IVP pregnancies. Failure of epigenetic reprogramming may also play a role in these losses for SCNT embryos (see Chapter IV). The impact of such events on the health of the dam is dependent on the stage of pregnancy when loss occurs. Losses due to defects in the embryo or failure to implant do not pose a hazard to the dam in early stages of pregnancy, whereas mid- and late-term spontaneous abortions may pose a health hazard to individual females if they are unable to completely expel the fetus and its associated membranes." , is appropriate and accurate.

Page 132: The summary: "Summary of Outcomes Noted During the Perinatal Period (Developmental Node 2) As with pregnancy data, information from the perinatal period indicates that cattle and sheep clones are at the greatest risk of morbidity and mortality, compared with goats and swine. Also as observed in the pregnancy data, the abnormalities noted in animal clones are not unique to animals derived by SCNT; similar outcomes have been observed in other ARTs, albeit at lower rates. Most of the information on neonatal mortality comes from cattle. Neonatal death rates for cattle clones currently average approximately 20%. Dystocia may be the most influential factor on calf mortality, due to trauma of difficult labor and emergency C-section; however, abnormal organ development also appears to play in important role. At this time it is not useful to attempt to quantify the risk of neonatal death in lambs due to the small numbers of animals generated by SCNT and the variability among laboratories performing cloning in sheep. However, it appears that mortality for IVP and SCNT produced lambs was higher relative to that experienced by lambs produced by AI and natural mating.

The major clinical finding associated with these observed outcomes appears to be a complex of clinical signs collectively known as LOS, which has been described in calves and lambs produced by ET, IVP, BNT, and SCNT. Some of the clinical signs/reported may be directly related to fetal oversize, constraints of the surrogate’s uterine capacity, and dystocia during labor. Other signs, such as respiratory, cardiovascular, hepatic and renal (kidney) abnormalities do not appear to be related to intrauterine effects, and may occur even among calves within the normal range of birth weights for their breed, but are considered part of the syndrome due to the frequency of co-occurrence. Hence, LOS may be a misnomer, but the term has become familiar to scientists working in the ART field. The causes of LOS remain unclear, but may be related to in vitro culture conditions and other factors, such as incomplete reprogramming of the somatic cell nucleus (see Chapter IV).

With the exception of one pig clone born with anal atresia, no other reports of frank deformities have been noted for this time period in non-transgenic swine clones, although birth weights may be lower in swine clones relative to non-clone comparators. The single study reporting high mortality rates in non-transgenic swine clones reported clinical signs that may be related to various causes, including infectious disease, which cannot be ruled out based on the available data.

Goats do not appear to exhibit cloning-related adverse outcomes.", is appropriate and accurate.

Page 147, CVM identified that the peer group, "post-pubertal, near-adult and adult females may not provide an appropriate comparison for pre-pubertal, rapidly growing males.", is not an adequate comparator for normal range of blood chemistries.

Page 150: From the data presented the following statement, "Most prepubertal cattle, swine and goat clones appear to grow and develop normally following the early neonatal period as demonstrated by reports on health status and laboratory measurements presented in the available published data and other reports on health status supplied by private companies.", appears appropriate.

Page 161: With the limited data presented, the following statement is justified,

“The limited data suggests that there are no adverse effects on the reproductive health of cattle clones, although this tentative conclusion must be tempered by the small number of available studies.” However, the CVM should encourage groups involved in cloning to collect further data on the reproductive health and performance of clones. This should become a funding priority for FDA and/or USDA to support such evaluations.

Page 164: I agree with the statement that, “convincing data on clones addressing the issue of premature aging are not currently available.”, however this data needs to be collected over the next ~ 5-10 years to adequately answer this question. Premature aging does not appear to be problematic for animal health based on the available data.

Page 170: I agree with the conclusions that,” Insufficient time has elapsed since the first domestic livestock clones were born to make any reliable observations on maturity, aging, or the lifespan of these animals. Reports on telomere lengths in animal clones are highly variable (see Box V-1), appear to be tissue dependent, and may not be reliable predictors of lifespan. As most female food animals are not maintained to old age, the risk of increased health problems or decreased longevity, if any exist, would be primarily to male animals kept as breeding stock.” More data collection and research in this area should be encouraged and supported.

Page 173: The Summary of the Outcomes for Clone Progeny is accurate and appropriate. There is no evidence that the progeny of clones have any increased health risks over conventionally produced animals.

Page 173-175: Conclusions – The reasons for the inability to perform a strict quantitative analysis of the risk of SCNT to animal health are clearly described and explained. The number of animals evaluated is small and the rates of adverse outcomes appear to be decreasing. Because of these factors it is suitable to conclude that, “it is not appropriate to perform a quantitative analysis of “relative risk” until these rates have stabilized.” This should be re-evaluated in the next five (5) years as more data becomes available.

The conclusions are accurate and appropriate based on the data examined and the literature reviewed. Further analysis, though desirable, is not currently possible because of the lack of specific information such as the short time the technology has been available and the limited data and scope of data from some species.

Has the Risk Assessment adequately identified the hazards and characterized the risks relating to food consumption?

The assumptions of this portion of the risk assessment, “This Chapter of the Risk Assessment is focused on food safety concerns, and assumes that any clones or their products would be subject to the same local, state, and federal laws and regulations as conventional food animals or their products. These assumptions exclude animals that would fail FSIS inspection and therefore not enter the human food supply (although they might be rendered). It also assumes that any hazards arising from the consumption of products derived from animal clones would result from epigenetic dysregulation of the genome of the developing animal, as described in Chapter IV.”, and “Because much of the focus of this analysis is the identification of subtle hazards in otherwise healthy-appearing animals, the Critical Biological Systems Approach (CBSA) evaluates animal health data on as fine a level of resolution as possible. This includes individual animals or even individual analytes per animal in order to have a sensitive screen for adverse outcomes (and thus food consumption risks).”, are reasonable and scientifically appropriate assumptions.

The identification of applicable peer-reviewed papers is appropriate and justified. The use of similar methodology, as described in Chapter V, using developmental node-specific groupings is scientifically sound and justified with regard to the analysis of bovine clones. This is not possible with the other species because of scarcity of appropriate data for assessment.

Inclusion of “the most comprehensive survey of the health status of cattle clones that has been assembled” from Cyagra, Inc., increases the accuracy and confidence in this Risk Assessment. Furthermore, the uniqueness of this data set is described in the following text from the assessment draft, “The information provided by Cyagra differs from that presented in the peer-reviewed literature for several reasons:

The data were collected specifically to address issues raised in this risk assessment, and thus are not part of a hypothesis-testing study, or written to provide examples of novel or unusual events;
They have not been peer-reviewed outside CVM (to the Center’s knowledge);
They include individual animal data; and
They are far more extensive with respect to the number of clearly non-transgenic animals evaluated (n=78 surviving and tracked animals), and the number of observations on individual animals than any other study or series of studies from a particular laboratory. “

The availability of such information allows much more detail analysis and assessment of potential risks to food consumption. This increases the confidence in the risk assessment.

Specific Comments: Chapter VI

Page 179-214: The relevance of this text to Food Consumption Risk assessment seems difficult to understand. I am unable to determine it value in the risk assessment.

PAGE 201: The following text, “Ninety percent of the total clinical chemistry values of the Clones were within the range of values exhibited by the Comparators, and 90 percent of the hematology values were within the Comparator range. Twenty-seven of the 33 analytes (substances that were measured, such as sodium, cholesterol, or liver enzyme activity) had either no differences or one difference relative to the Comparators (Chart E-101). The remaining six analytes tended to be more variable between Clones and Comparators. Liver values (AST, GGT, cholesterol, bile acids (hBA)) were lower in several clones, for reasons likely related to the placental/umbilical abnormalities, or transitions from fetal to adult circulation. GGT levels were also low relative to the Comparators, probably related to blood sampling prior to colostrum intake, whereas Comparators were administered colostrum prior to blood draw. None of the out-of-range values of these analytes poses any particular concern for food safety, as they are relatively close to the Comparator range.”, is valuable in the risk assessment as it allows comparison of the clones and “normal” animals. This is scientifically appropriate and sound in the context of this risk assessment. However, I still maintain that direct sampling of the tissues/edible products for compositional analysis is an important part of this risk assessment.

Page 211: The information described in this text is again a valuable component of the risk assessment, “Measurements of analyte levels in the entire 1-6 month old cohort were generally very close to those measured in the Comparators (Chart E-201). In aggregate, 96% of the total analyte values for Clones were within the range of the Comparators.”

Page 214: The conclusions from the section, “(g). Summary for Juvenile Developmental Node in Bovine Clones (Developmental Node 3) With the exception of visible physical anomalies that were detected, individual animal and analyte review of the data indicated no differences between clones and conventional animals that reflect any food consumption hazards in clones. Clones that may be physiologically “unstable” at birth appear to normalize all of the measured variables within two months of birth (Chavatte-Palmer et al. 2002; Cyagra 2003). Some juvenile clones succumb to the sequellae of LOS. Surviving clones appear to grow normally, and careful evaluation of the laboratory results indicates that the clones’ physiology reflects normal, appropriate responses to ongoing growth and developmental signals, and that they are functionally indistinguishable from non-clones.”, is appropriate based on the data and results summarized. Again, direct testing of product composition is a more robust strategy for examining food safety of these animals.

Page 221: I agree, the Aoki data would be more helpful in the assessment than the small numbers presented in the abstract. It really does not give much information that is useful.

Page 222: This is the type of useful information needed for the food consumption assessment, “Heyman et al. (2004) reported that first lactation milk yields (9,341 ± 304 kg vs. 8,319 ± 1800 kg for a 305 day lactation) and somatic cell counts (SCC), which are a measure of mammary gland health) for three female Holstein clones were similar to those of three age-matched non-clone comparators. Somatic cell counts for both clones and comparators (116 ± 103 x 103 vs. 113 ± 50 x 103) were well below the level indicative of subclinical mastitis (1,000 x 103), and the SCC limit cited by the Pasteurized Milk Ordinance for fluid milk entering commerce. “

Page 223: The Summary Statement for Reproductive Development and Function in Bovine Clones (Developmental Node 4) is appropriate based on the available information. Reproductive function is in delicate balance with animal health and thereby indicates normal physiological function. This however still does not get at food safety risks directly but is a good indirect measure of safety.

Page 227: The conclusions in the section, Summary Statement for Post-Pubertal Maturation in Bovine Clones (Developmental Node 5) are appropriate based on the data. The statement, “Individual animal reviews indicated no health problems, or changes in physiological parameters that would indicate a food consumption risk that would not be detected in existing food safety regulations (e.g., mastitis in milking cows). “ is also justified.

Page 232: Summary Statement, the data from swine studies in developmental nodes 1 and 2 is limited and does not allow reliable risk assessment. However, the statement, “piglet clones appear to be normal and healthy.”, is justified based on the available data.

Page 238; Again the data is limited for swine clones but the conclusion that “none of the swine clones exhibited any adverse outcomes that have not been observed in conventionally bred and reared swine.”, is justified based on the assessment of available data.

Page 239: The final conclusions for swine clones on limited data sets are reasonable. The statement, “The identified abnormalities in the Archer et al. (2003a) (parakeratosis) and Viagen dataset (lung adhesion) are abnormalities normally seen in case noted does not pose a food consumption risk, as the affected skin from the carcass would be condemned at the slaughterhouse, and would not enter the food supply. The apparently normal status of the clinical measurements indicates that the clones in this study possess the same physiological functions and behaviors as their conventional counterparts, and thus are not likely to pose a greater food consumption risk than conventional swine.”, is appropriate based on the available data. It is recommended that collection of additional data for swine clones be made a priority.

Page 242: The inability to draw conclusions regarding the food safety of sheep clones is due to the lack of s substantial data set for analysis. The conclusion of the draft is appropriate.

Page 246: Based on the available data, the following conclusion, “Based on the data reviewed, there do not appear to be any anomalies present in the goat clones that would have a direct impact on the safety of food products derived these animals.”, is justified and supported scientifically by the data.

Has the risk assessment adequately addressed whether the edible products derived from animal clones and their progeny are as safe to eat as the edible products derived from their conventional counterparts?

One of the major issues with this section of the draft risk assessment is the limited data available for analysis. The assessment is based on “there are now several studies that have evaluated the composition of the milk and meat of both cattle and swine clones, and one large study that has evaluated the composition of the meat of the progeny of swine clones.” The collection of additional data in the future would increase the confidence in the risk assessment. I however, strongly agree and the data is supportive of the idea that “there is no a priori reason to expect that SCNT will introduce any new, potentially toxic substances into the milk or meat of otherwise healthy animals, the remaining food safety concerns addressed whether subtle changes have occurred that would alter the presence of important nutrients.”

Page 248: It is appropriate to identify the nutrients in meat or milk whose alterations would likely affect the overall diet as well as the major and moderate contributors to the total daily diet of meat and milk consumers.

Page 250: The nutrients shown in Table VI-14 are appropriate components for the compositional analysis of meat and milk. However, the carbohydrate composition of milk should also be considered as it is a major nutrient of milk. This is confirmed by the first paragraph on page 251, “bovine milk and milk products (excluding butter) provide approximately ….. and 5% of the carbohydrates consumed by the US population.”

Page 256: One problem with the analysis of meat is that as stated, “There are no full chemical characterizations for meats.” This will make assessment difficult unless the physiological status, age and samples taken are standardized. Even with that said there may be considerable variation in the chemical composition between and among the various meat samples form ‘normal” and cloned animals.

Pages 258-260: The analysis of the data from Walsh et al., 2003 and the conclusion in the draft of the following, “No significant differences were noted between clones and non-clones with respect to the concentrations of the individual milk proteins that were sampled. No significant differences (p> 0.05) were observed when the gross composition of milk from Holstein clones and Holstein non-clones was compared over the course of the entire lactation cycle (Table VI-17).”, is accurate and appropriate.

Page 262: the Japanese Research Institute obtained similar results and conclusions for Animal Science in Biochemistry and Toxicology provided a report entitled “Investigation on the Attributes of Cloned Bovine Products”. They showed that, “Milk constituents were compared between ordinary cattle, BNT clones, and SCNT clones. The results are reported as the mean of samples obtained three and six weeks after parturition and provided in Table VI-17. No biologically significant differences were observed between any of the groups for the parameters tested.”

Page 263: The examination of allergenicity showed no differences between milk from “ordinary cattle, BNT clone cattle, and SCNT clone cattle.” “Based on these two studies, the authors conclude that there were no biologically or statistically significant differences in the allergenic potential of milk from ordinary cattle or BNT or SCNT clones.” This is an appropriate conclusion.

Page 268: I agree with the conclusions of the milk composition risk assessment, “Summary Statement on Composition of Milk from Clones Based on the available data, milk from cow clones does not appear to differ significantly in composition from milk from non-clones. Small differences have been noted between clones and comparators, but given the different diets and husbandry conditions of these animals, it is difficult to determine with certainty whether the small changes seen in some components were a function of the diet, handling, or related to cloning. In summary, none of the small reported differences in any of the studies indicate any concern for food safety.”

Page 282: The conclusions from the CVM based on the available data are appropriate, scientifically sound and justified. I strongly agree with the following statement, “None of the characteristics that we examined differed in any biologically significant way between the clone and comparator, and none identified any potential nutritional or other hazards. Based on this review, CVM concludes that the data support the operating hypothesis underlying the Compositional Analysis approach, that is, meat and milk from clones and their progeny do not differ materially from that derived from conventional counterparts, and do not pose any additional food consumption risks relative food from conventional animals.”

Pages 283-286: The explanation of how the conclusions were drawn the approaches used for drawing those conclusions is clearly explained.

Page 286: The conclusions from CVM on Food Consumption Risks of Cattle Clones were: “Edible products from perinatal bovine clones may pose some very limited human food consumption risk.”. They further went onto conclude that based on the Cyagra data “Laboratory measures of key physiological functions do not appear to indicate that surviving animals are very different from conventional newborns. It is therefore unlikely that food consumption risks have been introduced into these animals.” They did qualify this last statement in stating that, “The uncertainty associated with the preceding statement is relatively high, however, for the following reasons. First, postulated differences in epigenetic reprogramming between perinatal clones and comparators suggest that some subtle hazards may have been introduced into these animals. Second, the relatively poor condition of many of these perinatal clones also precludes the conclusion that no food consumption risks, such as nutritional imbalances, are present. Therefore, given that perinatal clones may differ from comparator animals of the same age, at this time, the Center concludes that they may pose a very limited nutritional risk for consumption as food. Rendering these clones will not to pose such risks in animal feed or to humans consuming animals fed material derived from the clones.” The conclusions of CVM are appropriate based on the data examined. The large-scale consumption of perinatal bovine clones is questionable and the effects of potential “nutritional imbalances” on food safety are not described in detail.

Page 287: I do agree with the statement under the section -Risk Hypothesis Statement for Perinatal Bovine Clones, that “At this time there is insufficient information to move from Hypothesis 2 (Clones are Different) to Hypothesis 1 (Clones are the Same), even though the available data neither identify nor predict the presence of food consumption hazards (and subsequent risks) from these very young clones. The uncertainties in the data are relatively high and lead the Center to have a relatively low degree of confidence in the safety of edible products from perinatal bovine clones. We note, however, that it is highly unlikely that clones of this age group would be consumed for food.”

In section i. Edible products from juvenile bovine clones pose no additional food consumption risk(s) relative to corresponding products from contemporary conventional comparators. I agree with the following conclusion, “The underlying biological assumption for this developmental node is that if any anomalies were to be found in the youngest clones and those animals were to survive to be healthy adults, the juvenile developmental node would be a period of equilibration and normalization. The data appear to be consistent with such a hypothesis.”

Further, I agree with the conclusions in the last paragraph, which states, “None of the physiological measures taken, including both clinical chemistry and hematology, indicated any food consumption hazards.” The CVM has drawn the appropriate conclusions for the available data in this portion of the draft.

Page 288: The analysis of the data by CVM in the section-Risk Hypothesis Statement for Juvenile Bovine Clones, is consistent with the data examined and the published results. I agree with the conclusions in the following text, ”The assessment began at the position of Hypothesis 2, but the scientific evidence has moved the assessment from Hypothesis 2 to Hypothesis 1 for maturing juvenile clones. The weight of the evidence and the underlying biological assumptions lead the Center to conclude that there would not likely be any additional risk from the consumption of food from healthy juvenile clones relative to corresponding products from their conventional comparators. The consistency of these observations across all of the data for juvenile bovine clones makes the uncertainty associated with this judgment relatively low, and provides the Center with a relatively high degree of confidence in judgments regarding the health of (and consequent food safety of edible products derived from) this age cohort of bovine clones.”. I agree with the conclusion of CVM that there would not likely be any additional risk from the consumption of food from healthy juvenile bovine clones.

Page 289: The following text presents the accurate and appropriate scientific conclusions based on the data and results examined, “All of the reports on the composition analysis of meat or milk from bovine clones show that there are no biologically significant differences in the composition of milk derived from clone and non-clone cattle. Additionally, data from one report that show no difference in allergenic potential for meat or milk derived from clone cattle compared to meat or milk from non-clone comparators. Similarly, neither meat nor milk from clone or non-clone cattle induced mutations in a mutagenicity assay (Japan 2004). Finally, none of the reports identified an endpoint that would pose a hazard for human consumption.” Based on the present analysis of the available data it would be difficult to imagine that food from adult clones that “would pose a hazard for human consumption”.

Further, the conclusions in the section -Risk Hypothesis Statement for “Adult” Bovine Clones, are appropriate based on the data examined. And the following text is supports the change in position from Hypothesis 2 to Hypothesis 1, “The assessment began at the position of Hypothesis 2: that animal clones may appear to be copies of the donor animal, but that the process of cloning may have introduced subtle hazards that could pose food consumption risks. As presented above, however, the weight of the evidence has moved the assessment from Hypothesis 2 to Hypothesis 1 (Clones are the same as their sexually-derived counterparts). Extensive and consistent empirical evidence, including epigenetic, physiological, and health data on individual animals and compositional analysis of milk and meat derived from individual animals, indicate that adult bovine clones are biologically equivalent to their contemporary comparators. Therefore, evidence confirming the health of the animals produced via similar methods, and evidence confirming the compositional similarity of meat and milk from clone and non-clone cattle indicates that there is no additional risk from the consumption of edible products from these animals relative to sexually-derived comparators. The consistency of the observations provide the Center with a high degree of confidence in judgments regarding the health of (and food safety of edible products derived from) this age cohort of bovine clones. We note that given the economic considerations involved, it is not likely that many adult clones would enter the food supply as meat at this stage of the technology, unless they had suffered a non-treatable injury or old age. Milk products from lactating female bovine clones, however, could be introduced into the food supply.”

Page 291: I agree with the CVM’s analysis of and conclusions based on the available data for swine clones in section-Risk Hypothesis Statement for Swine Clones. The following text is consistent with the analysis of the data, “ Based on both underlying biological assumption and confirmatory data, CVM concludes that consumption of food from healthy adult swine clones would not pose an additional risk above consumption of their conventional counterparts. The data from Archer et al. (2003 a,b) is particularly compelling as it includes data on behavior, epigenetic reprogramming, and physiological measurements at two time points in the development of these clones. Likewise, data from Viagen includes information on growth, reproduction, carcass and meat composition, indicating that swine clones are not materially different from age-matched, genetically related swine. In this case, the Center finds itself at an intermediate Risk Hypothesis Level of “1 minus”, or relatively high certainty based on biological plausibility, consistency of observations among different and compelling datasets, and consistency with responses observed across other clone species. “ There appear to be no additional risks posed by “that consumption of food from healthy adult swine clones would not pose an additional risk above consumption of their conventional counterparts”.

Page 292: At this time the conclusions of CVM regarding the risks posed by consumption of food products from sheep clones (Risk Hypothesis Statement for Sheep Clones) is appropriate as, “At this time there is insufficient information to support Hypothesis 1; Hypothesis 2 must be the default position with respect to potential food consumption risks from sheep clones. CVM was not able to find any studies providing specific evidence to show that sheep generated by SCNT are healthy and normal, and would therefore pose no additional food safety concerns beyond those of their conventional counterparts.” It does seem highly unlikely that food products from sheep clones would pose any additional food consumption risks than would cattle clones but because of the lack of appropriate data, the CVM’s conclusions are prudent and should err on the side of scientific conservatism. Additional data needs to be collected and this should be encouraged and made a funding priority.

Page 292: The CVM’s conclusion, “Edible products from goat clones pose no additional food consumption risk(s) relative to corresponding products from contemporary conventional comparators.”, is scientifically justified and appropriate.

Further, the justification outlined in the section-Risk Hypothesis Statement for Goat Clones, is appropriate. The following text supports CVM’s position.

“Although the assessment began at Hypothesis 2, based on the underlying biological assumptions stated for the other clone species, consistency of responses with other species of clones, and a small but relatively rich dataset, CVM concludes that Hypothesis 1 more appropriately represents the conclusions regarding the food safety of goat clones. CVM places particularly high weight on the study of reproductive function, as it is one of the most complex physiological pathways to coordinate. The consistency of appropriate reproductive function, even in a small cohort of animals, adds to the confidence that can be placed in the judgment that these animals are as normal and healthy as their sexually-derived counterparts. Based on this finding, edible products from goats are not anticipated to pose more of a food consumption risk than their sexually-derived counterparts. Further, given the data on the normal reproductive function of these animals, and a preliminary report of normal reproductive function of one male offspring of a male goat clone, CVM has more confidence in the empirical demonstration that clone progeny are as healthy as other sexually-derived animals.”

Page 294: Based on the data, I agree with the CVM’s conclusion that “Edible products derived from the progeny of clones pose no additional food consumption risk(s) relative to corresponding products from other animals.” The following supporting justification is appropriate and consistent with the analysis,

“Progeny of clones, from the first sexual breeding of a clone through subsequent generations, will likely provide the overwhelming majority of clone-derived food products (both meat and dairy) in the US. The underlying biological assumption for health of progeny animals is that passage through the process of creating the cells that ultimately become ova and sperm naturally resets epigenetic signals for gene expression. This process is thought to effectively “clear” the genome of incomplete or inappropriate signals. The rationale for this assumption has been developed in Chapter IV, and dominates the conclusion that edible products from any clone progeny pose no additional food consumption risk(s) relative to those from any other sexually reproduced animals. It has been supported by detailed empirical evidence both in the mouse model system, which clearly indicates that phenotypic alterations noted in the parent clones are not passed to their sexually-derived progeny, and observations on the health and meat composition of progeny of livestock clones. In addition, the extensive information provided by Viagen on the progeny of clone swine provides direct data on the health of these animals and on the composition of meat derived from them. The swine data support the underlying biological assumption that the progeny of clone animals are essentially indistinguishable from the comparable progeny of non-clone animals.

We therefore concur with the high degree of confidence that the outside scientific community (NAS 2002 a,b) places in the underlying biological assumption, and conclude that consumption of edible products from clone progeny would not pose any additional food consumption risk(s) relative to consumption of similar products from sexually- derived animals.”

Pages 293-294: The conclusions outlined in the section-Summary of Risk Hypotheses are supported by the data and its analysis. The text in the following sections, “The current weight of evidence suggests that there are no biological reasons, either based on underlying scientific assumptions or empirical studies, to indicate that consumption of edible products from cattle, pigs, or goat clones that would pass FSIS inspection poses a greater risk than consumption of those products from their non-clone counterparts. The level of certainty is highest for bovine clones, followed closely in degree of certainty by swine and, and then goat clones. The lack of species specific data for sheep clones precludes an evaluation of the risk for consumption of sheep clones at this time. Consumption of edible products from the progeny of clones poses no additional risk(s) relative to those from other sexually-derived animals, based on underlying biological assumptions and compelling evidence from the mouse model system and the Viagen dataset on the health of clone progeny and their meat composition. No food safety concerns were raised in the study of the composition of milk or meat from bovine clones. The level of confidence that may be placed in these overall conclusions is quite high, although additional data can always increase confidence.”, are appropriate and scientifically supported.

Page 294-298: Additional considerations are described in the sections contained on these pages. The CVM has described possible risk due to Potential Allergenicity. The data that was examined supports “the lack of a unique allergic response to milk derived from clone cattle.” This is a reasonable and scientifically supported conclusion based on the available data.

The issue of potential Microbiological Effects was also addressed by CVM. Although there is little experimental data to support their conclusions, “It is likely that bacterial shedding from food animal clones poses no greater risk than that posed by conventional food animals. The complexity of the intestinal microflora makes this an extremely difficult question to address directly. Indirect evidence of normal intestinal microflora, however, can be inferred from the health status and growth characteristics of the animal clones, suggesting a normal microflora population.” The conclusion is reasonable based on the health status and normal growth patterns but is not directly supported by data. This is an aspect that needs further study, which should be straightforward given the current state of methodology in microbiology.

The issue of Unanticipated Effects was also discussed by CVM. The following text is appropriate, “This risk assessment has attempted to identify the range of potential hazards and risks that could be generated as the result of SCNT in domestic livestock species. Although it may be possible for a healthy clone to express some proteins inappropriately, the same argument can just as easily be made for sexually-derived animals. At this time, there is no validated method for determining small differences in protein constituents in foods, and even if such methodologies existed, the question would still remain as to how to interpret them--what foods would be used as comparators, and what degree of variability would be considered to pose a risk (NAS 2004)?

Finally, the issue of the hypothetical dysregulation of endogenous substances that may pose a hazard by virtue of increased dose should be addressed. The primary concern in this case is the up-regulation of small molecules that may retain bioactivity in the bodies of the human (or animal) food consumer, usually by virtue of the lack of degradation in the intestinal tract. For example, levels of endogenous substances that have posed some public concern in the past (e.g., estrogen and IGF-I) have been evaluated in bovine clones, and based on those data, there is no reason to expect that the levels of these substances in clones would pose any food consumption risks for humans.” There may be other unanticipated effects but these are not obvious at the present time.

CVM does accurately point out that changes in the technology used to produce clones might significantly change the risk assessment outcome. Changes in technology need to continue to be monitored. The follow section in the draft is an appropriate conclusion, “Significant changes in cloning technology, especially those accompanied by donor nucleus or oöcyte treatment regimens introducing new hazards into the overall process, would significantly increase the uncertainty associated with our judgments regarding the degree of risk that could accompany the resulting clones and clone food products. Without a careful evaluation of the animals arising from such methods, it would not be appropriate to speculate on the relative safety of the process from either an animal health or food safety perspective.” Additional research is warranted as cloning technology matures and advances.

The final issue of, How Much (Information) Is Enough?, has also been addressed in the draft. The following text describes the CVM’s thinking in this regard, “The question of determining when sufficient data have been collected in order to allow high confidence in risk-based decisions regarding edible products from animal clones is difficult to determine in the abstract. In practice, the answer is “it depends on what questions you ask, and how the data answer those questions”.

Because the nature of the technology has generally precluded generating large datasets on clones with good statistical power, CVM constructed a systematic approach to frame the appropriate questions (hazard identification), evaluated the available data (hazard characterization), and attempted to characterize resulting risk (probability of harm given that exposure occurs). This weight of evidence approach allows for the evaluation of the data from the CBSA and Compositional Analysis prongs of the Risk Assessment as part of an overarching whole. The conclusions from this risk assessment represent the judgment of CVM veterinarians, animal scientists, toxicologists, and risk assessors. The underlying assumptions for clones and their progeny were that the animals needed to meet all relevant federal, state, and local laws and regulations for conventional animals, and the food products derived from clones or their progeny also had to meet relevant federal, state, and local laws and regulations.

When considered across the Developmental Node spectrum, the data on the health of livestock clones were remarkably consistent across species, despite initial anomalies that appear to be species-specific. For example, although LOS may be more prevalent in cattle and sheep, most surviving animals normalize initial anomalies and become “healthy and normal.” This consistency has increased the value of even small datasets (e.g., goats), and has contributed significantly to the judgments regarding the health of these clones and their suitability as food sources. In addition, CVM evaluated a number of reports on the composition of meat and milk from clones and their progeny. No biological important or safety-relevant differences were noted when compositions were compared to standard databases or contemporary comparator controls. If anything, these data confirm the rather wide variability in the composition of meat and milk eaten on a daily basis. In summary, no toxicological hazard of concern for the human consumer has been identified in any of the reported studies. Although additional data from other sets of animals, particularly in other species routinely used for food, could be useful in increasing the confidence that may be placed in overall judgments regarding food safety, the weight of the evidence at this time is sufficient for the agency to draw the conclusions it has made in this Risk Assessment with reasonable certainty.” I agree with the final statement, there is sufficient to support the conclusions made in this risk assessment.

Pages 301-305: Summary and Conclusions section. The conclusions in this section are consistent and supported by the data examined in this risk assessment. The CVM has made a scientifically sound analysis of the data and its conclusions are appropriate. Their overall conclusions are: “For Animal Health: SCNT results in an increased frequency of health risks to animals involved in the cloning process, but these do not differ qualitatively from those observed in other ARTs or natural breeding. The frequency of live normal births appears to be low, although the situation appears to be improving as the technology matures. Cattle and sheep exhibit a set of clinical signs collectively referred to as LOS that do not appear to be present in swine or goats. Surrogate dams are at risk of complications from birth if the fetus suffers from LOS, or from accumulation of fluid in the cavities of the placenta (hydrops). Clones exhibiting LOS may require additional supportive care at birth, but can recover and mature into normal, healthy animals. Most clones that survive the perinatal period are normal and healthy as determined by physiological measurements, behavior, and veterinary examinations. Progeny of animal clones also have been reported as normal and healthy.

For Food Consumption Risks: Extensive evaluation of the available data has not identified any food consumption risks or subtle hazards in healthy clones of cattle, swine, or goats. Thus, edible products from healthy clones that meet existing requirements for meat and milk in commerce pose no increased food consumption risk(s) relative to comparable products from sexually-derived animals. The uncertainties associated with this judgment are a function of the empirical observations and underlying biological processes contributing to the production of clones. There is less uncertainty about the health of clones as they age and have more time to exhibit the full range of functionality expected of breeding stock. Edible products derived from the progeny of clones pose no additional food consumption risk(s) relative to corresponding products from other animals based on underlying biological assumptions, evidence from model systems, and consistent empirical observations.”

Pages 307-320: the Glossary is very useful and appropriate.

Appendix A -Risk and Safety Assessment Primer for Animal Cloning is quite helpful in describing the methods and intent of a risk assessment such as this.

Appendix C -Comparison of Outcomes Among Assisted Reproductive Technologies (ARTs) provided a good benchmark for comparison of risk assessment of clones versus other reproductive technologies in livestock.

Appendix H: The Comprehensive Veterinary Examination provided a helpful outline of the issues and parameters regarding the assessment of animal health.

The CVM examined a great deal of data in developing this risk assessment. This is a major strength of this document.

PEER-REVIEW SUMMARY

In summary the Animal Cloning: A Draft Risk Assessment, in my opinion, has adequately examined the risks associated with SCNT with regards to animal health and food consumption. It has pointed out specific aspects where its conclusions are drawn on and supported by data that allow both high and low levels of confidence in the assessments. One particularly important issue is that further data need to be collected in a numbers of areas that I have pointed out in the course of my review. Data on cloned sheep is especially lacking and needs to be supplemented if a recommendation regarding cloned sheep is to be made by the CVM. The CVM has done a thorough and conscientious job in developing this draft risk assessment for animal clones. The staff of the CVM is to be commended for their efforts.

Peer Reviewer 2

Review of Animal Cloning: A Draft Risk Assessment

Parameters of Review:

The following factors were considered and discussed in this review of the Animal Cloning: A Draft Risk Assessment; a) the appropriateness of the assumptions made and hypotheses postulated, b) the methodology utilized, c) the quality and relevance of the data and information, d) the accuracy of the analytic results, and e) whether the conclusions reached are supported.

Review:

The draft Risk Assessment on Animal Cloning, as prepared by the Center for Veterinary Medicine –FDA, provides a very comprehensive review of the technologies involved in somatic cell nuclear transfer (SCNT) and information available on clone animals in those species used for food. In this draft risk assessment the reviewers consider SCNT and it impact on animal health and any potential risks related to consumption of food products derived from clones. SCNT is placed in context of other advanced reproductive technologies currently utilized in agricultural animal breeding. Furthermore, the methodologies and assumptions used to identify potential harms and to assess associated hazards and risks are clearly delineated in considerable length.

The CVM has developed a framework for evaluation of information that involves the division of the life cycle of animal clones into five functional developmental nodes. These nodes include: Developmental Node 1- initial production of the SCNT embryo, early embryonic and fetal development (gestation); Developmental Node 2 - the perinatal period, including late gestation, birth, and the first critical days after birth; Developmental Node 3 - Juvenile development and function; Developmental Node 4 - Reproductive development and function; and Node 5 - Post pubertal maturation. This framework provides a logical approach for division of the available information into comprehensible units that reflect real phases within the animal’s life and also the stages of animal production and management.

This framework is used in the assessment of risks to animal health and for identification of food consumption risks. The risk assessment then further subdivides the available information and considers the risks as they applied to different agricultural species, cattle, swine, sheep and goats.

The risk assessment questions whether or not there is any harm or associated risks to the animals, either the recipient that carries the clone fetus through gestation to term or the animal clone, itself. As is readily apparent from the published literature, the process of cloning does increase the risk of adverse heath outcomes, relative to conventional animals. None of the adverse outcomes are unique to cloning, but are observed at a higher frequency in the clone animal production. The risk assessment clearly identifies the risks to animals involved in cloning and relates these risks to the framework of developmental nodes. The highest risks are associated with Nodes 1 and 2; once a clone animal has reached puberty risks derived from its production method appear to diminish. Clone progeny do not appear to have any increased risks as compared to progeny of animals produced by conventional techniques.

The risk assessment also considers the safety of food products derived from animal clones and their progeny. It is a particularly difficult challenge to assess the potential harms and associated risks of products (milk, meat) derived from clone animals and their progeny. In a more traditional risk assessment the potential harm is clearly identified; i.e., an added component to fortify and/or enhance a food or drug to treat a disease. This risk assessment excludes any clones with transgenic modifications, so there is no exogenous transgene product to consider. The products, therefore, are derived from an animal whose uniqueness results from having been made through a modification of reproductive techniques. While some may consider somatic cell nuclear transfer a rather extreme modification, it is in fact, the simple exchange of one unmodified nucleus for that of another. In the case of SCNT, the exchange is between a somatic cell nucleus and an oocyte nucleus. Both the zygotic nucleus (if the oocyte had been fertilized) and the somatic cell nucleus contain a full complement of the DNA code that contains essentially the same coded information for the enzymes, structural proteins, and other components that make up a cell and consequently the whole animal. This fact, that both contain a full set of coded information, reduces the question to that of whether or not the oocyte/early embryo can correctly process the somatic cell nucleus such that the result is an animal which is normal physiologically and, thus, is basically healthy. The risk assessment covers these issues in chapters discussing epigenetic changes and subsequent affects on embryonic development and clone health.

The risk assessment then considers the question “Are the milk and meat derived from cloned animals altered and, if so, does that alteration contribute any harm?” The CVM-FDA has made the assumption that if an animal is apparently healthy and within the normal range of basic physiological and metabolic functions, then products derived from that animal produced through SCNT should pose no greater risk than products derived from animals produced by more traditional reproductive methods. This is a logical assumption. Data collected from several sources and for several species indicates that those clone animals that survive gestation and birth and that appear healthy and normal also have physiological and metabolic parameters that fit within the normal range.

Some might question if the DNA processed slightly differently due to altered epigenetics of clone could result in altered proteins. Could these slight alterations produce an allergen. While this may be possible, it is unlikely that any such alteration would be unique to clones but also would occur in animals processed by more traditional reproductive techniques, although at lower incidence. Thus any person, who would be potentially allergic to that allergen, is likely to already avoid that food product. Furthermore, data from the one available study suggestions that there is no alteration in the digestibility or allergenicity of clone food products (meat/milk).

Summary:

Overall, the draft risk assessment is a comprehensive review of animal cloning. In the case of risks to animal health, the harms are identified and the risk assessment characterized. As concluded in the draft risk assessment, animals involved in the cloning process are at increased risk of adverse heath outcomes, although none of these risks are unique to the cloning process. Furthermore, the variability in the incidence of outcomes reported by different labs and for different species would suggest that these adverse outcomes are associated with technical difficulties and may be overcome through technical advances and improvements in the ability to assess donor cell suitability and embryo viability.

In the case of food safety of products derived from clones and clone progeny, the risk assessment concludes that consumption of food products derived from healthy appearing animals that are produced through SCNT, and their progeny, provide no greater risk than consumption of products derived from healthy appearing animals produced by more traditional breeding techniques. This conclusion was based on the consideration of available data on the health of animal clones, the composition of meat and milk from those animals and corresponding information on clone progeny. The conclusions of the risk assessment were developed following a thorough and logical examination of the technology and with comprehensive consideration of potential harms and associated risks.

In summary and as a response to the Specific Technical Questions:

Has the risk assessment adequately discussed whether cloning introduces any unique risks to either animal health or the consumption of food from clones or their progeny relative to current agricultural practices?
Yes. Throughout the risk assessment, the question as to whether there are any unique risks is considered. As discussed in the risk assessment, any health issues are likely to arise from modifications in epigenetic control mechanisms. While such modifications can affect physiological functions and, thus, the health of the animal, it is the increased incidence of adverse outcomes rather than uniqueness to clone animals that is of concern. As stated in the risk assessment, regulations exist that would remove unhealthy animals from food production.
Has the risk assessment adequately identified the hazards and characterized the risks relating to animal health?
Yes. Information obtained from public and private sources were utilized to identify hazards and characterize risks related to animal health. Increased risks to clone embryos and their recipients (surrogate dams) include completion of gestation and clone survival during the perinatal period.
Has the risk assessment adequately identified the hazards and characterized the risks relating to food consumption?
The cloning process provides a challenge for identifying hazards and characterizing risks as the process does not involve treating food products or food animals with exogenous components. The process is an alternative method of reproducing an animal. Therefore, the question is ‘Does modifying of the method of animal reproduction introduce any alteration in the food product derived from that animal?’ The risk assessment considers this question and makes the assumption that if an animal is healthy and has physiological parameters that fall within the normal range, then food products derived from that animal should pose no greater risk than that of animals produced by other reproductive techniques. This assumption is logical and is supported by data from the studies that analyzed components of milk and meat derived from animal clones and controls.
Has the risk assessment adequately addressed whether the edible products derived from animal clones and their progeny are as safe to eat as the edible products derived from their conventional counterparts?
Yes. See response to specific technical question 3.

Peer Review 3

Review of the CVM Animal Cloning: A Draft Risk Assessment (2/14/2006)

Third Reviewer

Overview :

In this document, the CVM has attempted to provide a risk assessment covering animal safety and human food safety of clones produced by somatic cell nuclear transfer (SCNT) techniques. Traditional risk assessment requires the identification of specific hazards associated with an activity, process or chemical, and estimation of the incidence of these adverse outcomes. Hazards must be viewed as possibilities of adverse outcome, and are usually based on the observation of such adverse outcomes either “naturally” or by testing. Risk is the probability of an adverse outcome associated with a defined exposure to a hazardous activity, process or chemical. Risk may be actuarial (observed, measured) or predictive (estimated or extrapolated) in nature. Whatever the case, risk is a probability of an adverse outcome under very specific conditions. Risks (probabilities) cannot be qualitative as they are numbers. The idea of doing a “qualitative” risk assessment is inherently erroneous. What CVM has done in this document is not a qualitative risk assessment, but a comparative risk assessment. This is especially true for the animal safety portion of the document, in which the incidence of specific adverse outcomes of SCNT cloning is compared to the incidence of these adverse outcomes for other assisted reproductive technologies (ARTS). This is an entirely appropriate way to compare the technologies and to see where SCNT cloning ranks in relation to the other ARTS.

While the ARTS have been compared from the point of view of “safety”, in reality they have been compared (at least at the very early life stages) by considering the incidence of success (success rate, failure rate) and SCNT technology has been similarly assessed and compared to the other ARTS. The fact that SCNT cloning is being done despite the low success rate, indicates that it appears to be a desirable and economically acceptable technology. The types of hazards (LOS, hydrops, etc) are not unique and can be managed with standard veterinary care.

A similar assessment method has been used for estimating the risks which may be associated with consuming the products of SCNT cloned animals. In this case however, there are no identifiable, observed hazards associated with eating SCNT cloned products. The CVM has appropriately attempted to identify potential hazards, and utilize these in the risk assessment process. Since none have actually been observed, it has been impossible for the CVM to establish any numerical incidence for comparison. The best that can be done in this case is to recognize that there is a reasonable certainty of no harm from eating cloned animal products, and the CVM has done this by utilizing a standard of “no additional risk”. The CVM might have included some probability estimates that any individual might actually consume cloned animal products, or the percentage that such clone-derived foods might make up of an individual’s diet based on the number of clones which might be slaughtered for food, and the amount of milk from cloned animals which might be consumed. Similar to estimates which have been made concerning the ingestion of an injection site, such data would establish a probability figure which would be of value for estimating the probability of exposure. This would be useful even though the hazards remain hypothetical.

The CVM has collected all publicly available data as well as some private data for this comparative assessment. Since most of the studies which are cited were not specifically done to address the animal safety or food safety questions, there are some gaps in the desired data, but CVM has identified everything which is currently available. In this reviewer’s opinion, the data they have uncovered are sufficient to support their conclusions as to the comparative safety of SCNT cloning to food animals, and to consumers of foods derived from SCNT clones and their progeny.

Specific concerns:

Food Consumption Risks (Chapter VI)

On page 178 the documents states: “ Thus, although some of the data in this chapter reprises information previously addressed in Chapters IV and V, the methods by which the data were evaluated differed. ”

In fact, most of the data presented are indistinguishable from the data presented on the same developmental nodes in the animal safety section. The inclusion of repetitious information previously covered is not justified, and for the most part is it virtually irrelevant. Cell culture conditions, and Zona pellucida coverage again is irrelevant to human food safety. Most of pages 181-200 could be cut completely without being missed.

p 179 Inclusion of some data on transgenics

“As indicated previously within the Risk Assessment and detailed in Appendix D, some of the animals on which reports are provided are somatic cell nuclear transfers of transgenic cells, thereby actually being reports on transgenic animal clones. These have been included in the food consumption risk assessment when they provide corroborative information, and the transgenic status of the animals has been indicated when that information is available .”

At every other place in this document, the focus has been on SCNT clones. Inclusion of data on transgenic animals, even as corroborative data, could tend to confuse the issue. While these data do not indicate any hazards, I think this information should not have been included in the human food safety portion, it could muddy the waters.

p 283 - Uncertainties.

“Proving the negative or, in the absence of its proof, identifying the conditions under which concerns have been minimized to levels considered “acceptable” becomes the goal of a comprehensive risk assessment.”

This should be deleted, you cannot prove a negative at all, so what this assessment should be doing is to assess the hazards and then to put some estimates of incidence on them, if possible. Additionally, this statement mixes risk management strategies with the risk assessment process, which should not be done.

“In fact, certainty of prediction is unattainable in science. In its absence, risk assessment can provide a systematic approach for bounding the “risk space” in which risk managers operate by allowing assumptions and uncertainties to be clearly identified. Especially for new technologies in which uncertainty may be high, the “bounded framework” risk assessment process allows decision makers (both risk assessors and risk managers) to draw conclusions based on the data. Then, by explicitly addressing uncertainties, identifying biases, scientifically defensible (or alternatively, policy-based) judgments can be made about acceptable risk levels. ”

“Risk space” and "bounded framework" terms are used but not really defined, therefore leaving the reader thinking that these are well defined when in fact they are not. These are actually aspects of risk management, and another place where management creeps into the assessment process. Additionally, one cannot use science to determine acceptable risks; one can only use science to estimate hazards and incidence. Establishing acceptable risk is policy-based, and while scientific information is used to arrive at an acceptable level of risk, science has nothing to do with the policy process.

P 304-305 Summary and Conclusions

On pages 304-305 there is a discussion about residual uncertainty which presents information about the possible sources of uncertainty, but no information about the magnitude of these sources of uncertainty.

As a general comment about the food safety assessment, it appears that the CVM has not addressed one avenue which could supply some quantitative data for estimating the potential exposure of consumers to foods derived from cloned animals. It should be possible to estimate the number of cloned animals (or the amount of milk from cloned animals) which might enter the food supply. From these data, one could then calculate consumption factors (how much of these products might be ingested). This is perhaps the only avenue where CVM can provide some quantitative data related to food safety. The Economic Research Service provides data on the numbers of animals slaughtered for food each year, (approximately 35 million cattle, 100 million swine, 3 million sheep in 2003) and milk production figures are also available.

Appendix A: Risk Assessment Primer

There are some problems with what is said here in relation to the NAS process and documents. The CVM document refers to the NAS 2002 report “Animal Biotechnology: Science-based Concerns” for a description of a risk assessment process which differs from the 1983 process, and presents a table (A-1) comparing these to support a change in paradigm for looking at animal biotechnology. In actuality, the NAS procedure listed in the 2002 Animal Biotechnology Report was specific for doing environmental risk assessment of transgenic organisms. In the 2002 report, it specifically states in relation to the 1983 four-step process: "These steps do not apply well to GE organisms in the environment because dose-response and exposure assessments are intended to apply to substances that can be quantified in discrete amounts and that cannot reproduce themselves.”

This reviewer believes that the 1983, four-step process is perfectly applicable to assessing the risks of SCNT cloning to animals and to consumers of SCNT clone-derived foods.

Appendix A: page 5. Basically in this appendix, CVM recognizes that it is not possible to do a real risk assessment. “When performing a risk analysis, it is critically important to distinguish between a hazard and the potential risk(s) that may result from exposure. A hazard can be defined as an act or phenomenon that has the potential to produce an adverse outcome, injury, or some sort of loss or detriment. These are sometimes referred to as harms, and are often identified under laboratory conditions designed to maximize the opportunity to detect adverse outcomes. Thus, such observational summaries are often referred to as “hazard identification” or “hazard characterization”. Risk, as previously discussed, is the conditional probability that estimates the probability of harm given that exposure has occurred. In a qualitative assessment such as this, however, risks can be discussed only within a qualitative context, and no quantitative interpretations should be made.” Put this right up front in the document, and change the word “qualitative” to “comparative”.

Use of the CBSA approach for evaluating the risks of consumption of foods derived from SCNT clones is relevant only for the last three developmental nodes (from juvenile through maturation). Developmental node two, which is the perinatal period through the first few days of life, is not really relevant to the food safety evaluation. Animals which fail to thrive during this period will not be eaten, nor will the survivors as they are much too valuable to be turned into veal cutlets. The CVM has included this node on page 10 of the Executive Summary, stating that perinatal clones may pose some very limited human food consumption risk. This seems unwarranted considering that these animals would not be eaten, and CVM recognizes this in the document.

In several places in the document, the issue of uncertainty is mentioned and the idea of characterizing the uncertainty presented. In no place in the document however, is there a systematic presentation of this characterization except to explain the possible sources of uncertainty. This is probably because the uncertainty cannot be characterized any further than that. This is expected since most of the potential hazards are hypothetical rather than actuarial, particularly in relation to food safety.

Animal Safety Assessment:

(a) Assumptions made and hypotheses postulated

This section of the report is thorough and well done, and the comparisons with other ARTS are appropriate. There is essentially no other way to accomplish this except in comparison to the other methods.

(b) The methodology utilized

The CBSA approach is an appropriate and organized way to look at the comparative incidence of adverse outcomes for the various ARTS.

The data are relevant; however they are not always as complete as that one would wish to have when making such comparisons. The inclusion of more complete, unpublished data is very helpful and useful for evaluating the success rate and adverse outcomes associated with SCNT in the various species.

(d) Accuracy of the analytic results

There have been no actual probabilities calculated, only comparisons of specific adverse outcomes and the rates of these adverse outcomes associated with the ARTS.

(e) Conclusions.

The conclusion, that SCNT cloning, in comparison to other ARTS, has an increased incidence of adverse outcomes is supported by the data that have been presented. The conclusion that these adverse outcomes are not different from those seen with other ARTS is also supported by the data.

Food Safety Assessment:

(a) Assumptions made and hypotheses postulated

Both of the methods used to assess the food safety question are based on the assumption that food derived from healthy animals will be safe for human consumption. This is an acceptable assumption which is unlikely to be contested. The converse of this assumption is the assumption that food derived from un-healthy animals or animals with anomalies will be unsafe for human consumption. This is also an acceptable assumption for the purposes of this document, as is the assumption that such animals would not enter the commercial food chain.

(b) The methodology utilized

The CBSA approach is less adaptable to the food safety question. It appears that this approach was used since it was successfully used for the animal safety assessment, but it is not really relevant and probably not even necessary. The compositional analysis method, for which there are now some excellent data is the most relevant and important approach confirming the expected finding that meat and milk from clones does not differ compositionally from meat and milk from non-cloned animals.

The compositional analysis data are of high quality and great relevance. These are the most important and most relevant data available to address the food safety question.

(d) Accuracy of the analytic results

No actual probabilities have been calculated as no actual hazards have been observed.

(e) Conclusions.

The conclusions are well supported by the data. The lack of data on sheep should not be disturbing and the inference that foods derived from cloned sheep are also safe for consumption would be quite appropriate based on the data from the other ruminant species.

Recommendations :

This reviewer has several recommendations for this document to enhance its readability and increase its value to public understanding.

Acknowledge that this is a comparative risk assessment, and succinctly define, right at the beginning of the document, the terms hazard and risk, and the RA process.
For the animal safety section, create a table or tables in which the observed adverse outcomes associated with the ARTS can be easily compared. The tables which are included in this document are cumbersome and difficult to use for comparison. For those ARTS for which there are incidence data, include these in the table, as well as the number of observations. Such a table would allow a reader to compare the ARTS without having to page back and forth through the document. If desired, this could be done using separate tables for the various life stages used in the CBSA approach, and even for each species if necessary. The main idea is to promote readability of the data and understanding of the findings. This document is a thorough review and comparison of these technologies which has not been done before, thus it is very important that it be reader friendly. As it stands, it is not, and is actually quite difficult to work through. Move the difficult information to the appendices.

The reviewers have been asked to address four, specific questions, listed below. The questions and this reviewer’s answers are included below:

Specific Technical Questions.

Has the risk assessment adequately discussed whether cloning introduces any unique risks to either animal health or the consumption of food from clones or their progeny relative to current agricultural practices?

Yes, this assessment has done a thorough review of the available data and has not found any indications that SCNT cloning introduces any unique hazards to the animals or human food safety. The hazards found have all been reported previously (animal health) or none have been identified at all (human food safety).

Has the Risk Assessment adequately identified the hazards and characterized the risks relating to animal health?

The comparative approach using data from the other ARTS is thorough and current. More data will be forthcoming in the future so that this could be addressed again at another time if any currently unrecognized and unique hazards are documented.

Has the Risk Assessment adequately identified the hazards and characterized the risks relating to food consumption?

Hypothetical hazards have been addressed adequately, and CVM has been careful to address only those hypothetical hazards which might be expected to occur. Risks have not been characterized and cannot be characterized for the human food safety aspects since the hazards are all hypothetical.

Has the risk assessment adequately addressed whether the edible products derived from animal clones and their progeny are as safe to eat as the edible products derived from their conventional counterparts?

CVM has appropriately concluded that food derived from cloned animals will not pose any additional risk to human consumers.

Summary Review of “Animal Cloning: A Draft Risk Assessment”

Executive Review Summary:

Three reviewers were asked to review FDA Center for Veterinary Medicine (CVM) document titled “Animal Cloning: A Draft Risk Assessment” dated February 14, 2006. All of the reviewers agree that the CVM has provided an exhaustive review of data on somatic cell nuclear transfer (SCNT) cloning which are currently available in the published literature on animal safety and human food safety. In addition, some highly relevant unpublished data were also provided to CVM to assist with the review. The methodology used in the evaluation of these data was also considered by all of the reviewers to be appropriate for the animal safety assessment. For the human food safety evaluation, two reviewers questioned aspects of the Critical Biological Stages Approach used for human food safety evaluation. All agreed that the compositional analysis approach and supporting data were by far, the most important and relevant for addressing the issue of human food safety.

All of the reviewers agree that the conclusions drawn by the CVM with respect to animal safety and human food safety are supported by the data presented. The continued collection of data on SCNT clones of all food animal species is strongly encouraged, particularly for sheep for which data supporting human food safety are largely lacking.

All comments provided by each reviewer are included below in the peer-review summary section. These comments have been reorganized by the peer review coordinator, but none of the comments have been edited (except for minor formatting changes); thus some may statements may appear to be redundant. The charge to the reviewers is included below, as are the specific technical questions which the reviewers were asked to address. The final section includes some specific recommendations from the reviewers which are not included elsewhere.

Peer Review Charge

Peer R eviewers are charged with clearly identifying and characterizing any scientific uncertainties and ensuring that the potential implications of the uncertainties for the technical conclusions drawn are clear. Where reviewers identify scientific uncertainties, they are requested to suggest ways to reduce or eliminate those uncertainties and to comment on how the government should address the uncertainties which the Draft Risk Assessment identifies.

The Peer Review report should distinguish scientific facts from professional judgments and provide input on the reasonableness of judgments made from the scientific evidence. The result should be an independent determination by each peer reviewer as to the appropriateness of (a) the assumptions made and hypotheses postulated, (b) the methodology utilized, (c) the quality and relevance of the data and information, (d) the accuracy of the analytic results, and (e) whether the conclusions reached are supported.

Specific Technical Questions. Peer Reviewers are charged with providing a broad evaluation of the overall Draft Animal Cloning Risk Assessment. In addition, they are asked to respond to the following Specific Technical Questions:

Has the risk assessment adequately discussed whether cloning introduces any unique risks to either animal health or the consumption of food from clones or their progeny relative to current agricultural practices?
Has the Risk Assessment adequately identified the hazards and characterized the risks relating to animal health?
Has the Risk Assessment adequately identified the hazards and characterized the risks relating to food consumption?
Has the risk assessment adequately addressed whether the edible products derived from animal clones and their progeny are as safe to eat as the edible products derived from their conventional counterparts?

PEER-REVIEW SUMMARY

Introduction

Animal Safety Assessment:

(a) Assumptions made and hypotheses postulated

(b) The methodology utilized

The CBSA approach is an appropriate and organized way to look at the comparative incidence of adverse outcomes for the various ARTS.

(d) Accuracy of the analytic results

There have been no actual probabilities calculated, only comparisons of specific adverse outcomes and the rates of these adverse outcomes associated with the ARTS.

(e) Conclusions.

Human Food Safety

(a) Assumptions made and hypotheses postulated

(b) The methodology utilized

The compositional analysis data are of high quality and great relevance. These are the most important and most relevant data available to address the food safety question.

(d) Accuracy of the analytic results

No actual probabilities have been calculated as no actual hazards have been observed.

(e) Conclusions.

Specific Technical Questions.

Has the risk assessment adequately discussed whether cloning introduces any unique risks to either animal health or the consumption of food from clones or their progeny relative to current agricultural practices?

Yes. Throughout the risk assessment, the question as to whether there are any unique risks is considered. As discussed in the risk assessment, any health issues are likely to arise from modifications in epigenetic control mechanisms. While such modifications can affect physiological functions and, thus, the health of the animal, it is the increased incidence of adverse outcomes rather than uniqueness to clone animals that is of concern. As stated in the risk assessment, regulations exist that would remove unhealthy animals from food production.

The discussion in Chapter IV, especially in section C. Implications of Epigenetic Reprogramming for Animal Health and Food Consumption Risks, quite adequately addresses the risks and the implications posed by the clones and their progeny to animal health and food composition relative to current agricultural practices. One of the most important aspects of this discussion is the issue of epigenetic normality and it effects. This has been well summarized in the following text, “It may be, as many have suggested (Wilmut 2002; Jaensich et al 2004), that no clone is completely “normal” with respect to its epigenetic profile. Although this is an important point for assessing the overall safety of the cloning process for any particular species, the relevance of “epigenetic normality” to food consumption risks is unclear. Further, because similar abnormalities have been noted in animals produced using other ARTs, the issue of defining normality becomes significantly more complex. It may be that normality encompasses a range on a continuum, and that animals that are healthy, meet appropriate developmental and behavioral milestones, and reproduce to bear healthy young are “normal”, regardless of their epigenetic status. The most compelling conclusions that can be made about food consumption risks, then, are drawn from assessments of the health status of the animals and the composition of food products derived from them, and not from gene expression studies.”

Spelling above: Yanigamichi should be Yanagamachi.

Has the Risk Assessment adequately identified the hazards and characterized the risks relating to animal health?

Yes. Information obtained from public and private sources were utilized to identify hazards and characterize risks related to animal health. Increased risks to clone embryos and their recipients (surrogate dams) include completion of gestation and clone survival during the perinatal period.

The next several paragraphs address the conclusions of the risk analysis described in chapter V. Cattle and sheep used as surrogates dams for SCNT embryos have increased incidence of late gestational complications, but at a lower frequency than with other ARTs such as IVP. The risk to surrogate swine and goats bearing cloned fetuses does not appear to be increased compared to the general population. There is an increased level of uncertainty associated with these conclusions because of small size of the data set. Further, there is an increased incidence of “mortality and morbidity in perinatal calf and lamb clones compared with calves and lambs produced using other ARTs. In cattle and sheep, the increased risk appears to be a function of LOS. Survival of these clones appears to be a function of both the severity of the clinical signs and neonatal management. The available information suggests that morbidity and mortality is not increased in perinatal swine and goat clones; however, the limited dataset in these species increases the uncertainty associated with this conclusion.” This is problematic but does not diminish the conclusions based on the available data. It is always better to have more information but the information analyzed appear to be adequate for the conclusions that were drawn. Has the Risk Assessment adequately identified the hazards and characterized the risks relating to food consumption? Hypothetical hazards have been addressed adequately, and CVM has been careful to address only those hypothetical hazards which might be expected to occur. Risks have not been characterized and cannot be characterized for the human food safety aspects since the hazards are all hypothetical. Has the risk assessment adequately addressed whether the edible products derived from animal clones and their progeny are as safe to eat as the edible products derived from their conventional counterparts? CVM has appropriately concluded that food derived from cloned animals will not pose any additional risk to human consumers.

The data also support the conclusion that reproductively mature cattle clones show no incidence of increased health risk. But again, the data set is limited. No conclusions could be drawn regarding an analogous situation in sheep, goat and swine clones due to insufficient data.

There is not sufficient data to assess adverse health risks in mature and breeding age clones. The available data suggests “that there are no apparent risks to the health of maturing animals from cloning.” Empirical observations have indicated that there is no increased risk of health problems in the progeny of clones as compared to conventionally produced animals.

Has the Risk Assessment adequately identified the hazards and characterized the risks relating to food consumption?

The cloning process provides a challenge for identifying hazards and characterizing risks as the process does not involve treating food products or food animals with exogenous components. The process is an alternative method of reproducing an animal. Therefore, the question is ‘Does modifying of the method of animal reproduction introduce any alteration in the food product derived from that animal?’ The risk assessment considers this question and makes the assumption that if an animal is healthy and has physiological parameters that fall within the normal range, then food products derived from that animal should pose no greater risk than that of animals produced by other reproductive techniques. This assumption is logical and is supported by data from the studies that analyzed components of milk and meat derived from animal clones and controls.

The assumptions of this portion of the risk assessment, “This Chapter of the Risk Assessment is focused on food safety concerns, and assumes that any clones or their products would be subject to the same local, state, and federal laws and regulations as conventional food animals or their products. These assumptions exclude animals that would fail FSIS inspection and therefore not enter the human food supply (although they might be rendered). It also assumes that any hazards arising from the consumption of products derived from animal clones would result from epigenetic dysregulation of the genome of the developing animal, as described in Chapter IV.”, and “Because much of the focus of this analysis is the identification of subtle hazards in otherwise healthy appearing animals, the Critical Biological Systems Approach (CBSA) evaluates animal health data on as fine a level of resolution as possible. This includes individual animals or even individual analytes per animal in order to have a sensitive screen for adverse outcomes (and thus food consumption risks).”, are reasonable and scientifically appropriate assumptions.

Inclusion of “the most comprehensive survey of the health status of cattle clones that has been assembled” from Cyagra, Inc., increases the accuracy and confidence in this Risk Assessment. Furthermore, the uniqueness of this data set is described in the following text from the assessment draft, “The information provided by Cyagra differs from that presented in the peer-reviewed literature for several reasons:

The data were collected specifically to address issues raised in this risk assessment, and thus are not part of a hypothesis-testing study, or written to provide examples of novel or unusual events;
They have not been peer-reviewed outside CVM (to the Center’s knowledge);
They include individual animal data; and
They are far more extensive with respect to the number of clearly nontransgenic animals evaluated (n=78 surviving and tracked animals), and the number of observations on individual animals than any other study or series of studies from a particular laboratory.”

The availability of such information allows much more detail analysis and assessment of potential risks to food consumption. This increases the confidence in the risk assessment.

Has the risk assessment adequately addressed whether the edible products derived from animal clones and their progeny are as safe to eat as the edible products derived from their conventional counterparts?

CVM has appropriately concluded that food derived from cloned animals will not pose any additional risk to human consumers.

Yes. See response to specific technical question 3.

Specific Comments and Concerns:

Specific Comments: Chapters I-IV

Page 3: Cloning is also NT with embryonic cells. Paragraph 1.

Page 10: Edible products from perinatal clones may pose very limited human food consumption risks? Evidence?

Page 22: “boar semen does not freeze well?” This is not completely true as great strides have been made in boar semen freezing methodologies and litter sizes now approach those available with fresh semen.

Page 23: In the US, AI is performed mainly by technicians and producers not by veterinarians.

Page 26: Oocyte maturation times for swine are longer approximately 42-48 hours.

In section 4, I would not use pierce the ovum but penetrate the zona and fuse with the ovum.

Page 30: Batchelder, 2005 is Batchelder et al., 2005, this mistake occurs many times in Chapter V as well as throughout the document.

Page 41-43: This is a good overview of risk assessment.

Page 43: In the section, “ (2) Determining the degree to which existing data address questions of animal health or food consumption risk.”, the conclusion is correct “few reports directly addressed food safety.”

Page 43-44: One problem is that only information published in, “peer-reviewed journals, or otherwise made available to the agency by companies engaged in cloning, with explicit permission for release to the public.” , was used in the risk assessment. This is a potential dilemma as the public maybe unaware of all the potential associated risks because they do not have access to all pertinent information.

Page 44: The section, “Characterizing residual uncertainties persisting following a review of the existing data.” is on target and appropriate.

Page 44-45: It is appropriate to consider transgenic clones separately in the risk assessment because of the different risk space because of the transgenic event.

Page 45: The lack of data on food safety is problematic and therefore must be indirectly assessed “the health status of the animals producing food would have to contribute to both the animal health and food safety components.” This is acceptable if the datasets are appropriately analyzed.

Page 50: The “Two-Pronged Approach to Assessing Food Consumption Risks” is appropriate.

Page 89: This is a significant conclusion, “The most important implication of the mouse clone literature for domestic livestock clones is the observation that anomalies noted in clones are not transmitted to their progeny.”, borne out by the experimental evidence. And supports with experimental evidence, the following statement in the draft, “It is also consistent with the limited but consistent observations of healthy, fully functional progeny born to domestic livestock clones. Thus, the empirical evidence supports the assertion that “Progeny of animal clones, on the other hand, are not anticipated to pose food safety concerns, as natural mating resulting from the production of new gametes by the clones is expected to reset epigenetic reprogramming errors that could persist in healthy, reproducing clones” (NAS 2002a).”

Page 100: I do not believe that the following statement is true, “Relatively few studies have been published on embryo and fetal loss in livestock animals bred by AI or natural service.” The following are references that address this issue:

Pope GS, Hodgson-Jones LS. Use of plasma progesterone levels in an assessment of embryonic loss in dairy cattle. Vet Rec 1975;96: 154.

Kummerfeld HL, Oltenacu EA, Foote RH. Embryonic mortality in dairy cows estimated by nonreturns to service, estrus, and cyclic milk progesterone patterns. J Dairy Sci. 1978 Dec;61(12):1773-7.

Bulman DC . A possible influence of the bull on the incidence of embryonic mortality in cattle. Vet Rec 1979;105: 420-422.