• Decrease font size
  • Return font size to normal
  • Increase font size
U.S. Department of Health and Human Services

Animal & Veterinary

  • Print
  • Share
  • E-mail

Section Contents Menu

Safety & Health

Appendix C: Comparisons of Outcomes Among Assisted Reproductive Technologies (ARTs)

Although there have been several studies comparing the outcomes of somatic cell nuclear transfer (SCNT) with various other assisted reproductive technologies, it is important to note that most of these evaluated data once other technologies had matured and were well-integrated into agricultural practice. The following summary provides an overview of several studies comparing the outcomes of four key ARTs. Comparison of success rates from SCNT with these ARTs may not be entirely appropriate due to the relative newness of SCNT technology. However, a review of the available studies indicates a trend of increasing adverse outcomes with increasing technological assistance; specifically, the increased rate of pregnancy failure, late gestational complications and problems associated with Large Offspring Syndrome (LOS) are most commonly associated with in vitro manipulation of the embryo. Table C-1 presents outcomes noted in various studies of artificial insemination (AI), in vivo produced embryo transfer (ET), in vitro produced embryos (IVP), blastomere nuclear transfer (BNT), and SCNT.

Table C-1. Outcomes noted among studies for various ART in cattle, swine and sheep.

Developmental Node1 Gestational Period ART Outcome Reference Comments
Node 1 Early conceptus, early embryo prior to completion of organogenesis (gd 42 in cattle) IVP, BNT Higher rate of embryonic death than AI or in vivo produced embryos Reichenbach et al. 1992;
Kruip and den Daas 1997;
Wells et al. 1998;
Hasler 2000;
Cattle and sheep
    IVP Pregnancy loss following transfer of IVP or in vivo produced embryos prior to gd2 21 or within 2 weeks of transfer Farin and Farin 1995
McMillan et al. 1998
    IVP Increased total length of conceptus from IVP embryos 2X that of in vivo produced at gd 12 and 17 Farin et al. 2001
Lazzari et al. 2002
    IVP Gd 16 IVP conceptuses shorter than in vivo Bertolini et al 2002 Likely reflects survival status during critical time of maternal recognition
    IVP 19% of conceptuses from IVP blastocysts degenerated by gd 17 Farin et al. 2001  
    IVP Altered embryonic disc morphology affected by culture medium for IVP embryos Fischer-Brown et al. 2005  
    IVP Pregnancy rates 45% or higher in dams receiving IVP embryos Hasler 2000;
van Wagtendonk-de Leeuw et al. 2000;
Lane et al 2003
Factors affecting outcome include embryo culture system, embryo quality, embryo evaluator, number of embryos implanted, synchrony with dam’s estrus cycle, fresh vs. frozen embryos
    ET Higher embryo survival rate when embryos are transferred fresh rather than frozen-thawed Spell et al. 2001  
    AI Embryo loss ~30% by gd 30in beef and dairy cattle Smith et al. 1982;
Sreenan and Diskin 1983;
Dunne et al. 2000;
Santos et al. 2004
    AI Embryo loss by gd 21 associated with high plasma estrogen levels on day of insemination Shore et al. 1998 Possible estrogenic effect of legume in diet
Node 1 Late embryonic/ early fetal period (days 30-90) ET, IVP Embryo loss for in vivo produced embryos < 5% (from 2 months – term) King et al. 1985
Hasler et al. 1987

Embryo loss for IVP embryos higher

  1. 13% after gd 40
  2. 10.7-13.1%
  3. 24% total between gd 53-calving, with more between gd 50-80
Hasler et al. 1995;
Agca et al. 1998;
Hasler 2000;
Block et al. 2003


Depending on medium

    AI, ET, IVP Pregnancy rates at gd 22 not different among groups. At gd 42, pregnancy rates similar between AI and ET, but increased embryo loss in IVP compared to AI and ET Drost et al. 1999  
    IVP Abnormal development of allantoic membranes and cavity in placentae of IVP embryos gd 30-90 Peterson et al. 2000 Abnormal placental development and reduced placental blood membrane development
    IVP Abnormal placentome and blood vessel morphology between gd 70-222 Miles et al. 2004
Miles et al. 2005
    IVP Gd 61 and older fetuses heavier than ET fetuses; altered fetal organ growth; excessive amniotic fluid Sinclair et al. 1999  
    IVP Gd 70 altered angiogenesis and placental morphometry;
modified synthetic oviductal medium (mSOF) compared with medium with serum had fewer placentomes, low placental fluid volume and lower fetal weight: placental weight ratio; Placentomes (cotyledon tissue) had decreased density of blood vessels, decreased expression of angiogenic factor mRNA and vascular endothelial growth factor (VEGF)
Miles et al. 2005
Farin et al. 2006
    AI Fetal loss by gd 44 30-40% in swine pregnancies Vonnahme et al. 2002 Fetal survival related to placental efficiency
    AI Embryonic/fetal loss varies from 10 to 20% between gd 28 and 80 in beef and dairy cattle Pope and Hodgson-Jones 1975;
Kummerfeld et al. 1978;
Bulman and Lamming 1979;
Lucy 2001
Progesterone levels in dams’ milk may be normal through first 30 days of pregnancy, followed by sudden drop
    AI Embryonic/fetal loss 11 to 44% by gd 50 in beef cattle Bulman 1979 Attributed to bull
Node 1 Late gestation ET,
Compensation in vascular beds of IVP bovine embryos; Compared with in vivo embryos, IVP had decreased fetal villi, binucleate cell volume densities in placentomes. Proportional volume of blood vessels in maternal caruncles increased in IVP group. Ratio of blood vessel volume density: placentome surface area increased. Miles et al. 2004 Theorized to compensate for increased fetal size and need for increased nutrients and gas exchanges, but increased vascular blood network at level of placentome
    IVP IVP fetuses show increased glucose and fructose in fetal plasma levels; increased placental surface area Bertolini et al. 2004  
    IVP Hydroallantois frequency in IVP pregnancies (1/200) higher than in “normal” pregnancies (1/7,500) Hasler et al. 1995  
    IVP, SCNT Pregnancy loss higher in SCNT than IVP embryos; 50-100% loss gd 30-60; placentae hypoplastic and reduced cotyledonary development Hill et al. 2000;
Chavette-Palmer et al. 2002;
Heyman et al. 2002;
Edwards et al. 2003;
Lee RS et al. 2004
    BNT Late gestation abortions, stillbirths, underdeveloped fetuses for gestational age, edema, hydronephrosis, testicular hypoplasia, skull and heart malformations; lack of udder development in dams Wells et al. 1998  
    IVP, SCNT Broad distribution of fetal and neonatal body weights for both IVP and SCNT-derived embryos; shifted to “heavy” relative to in vivo embryos Wilson et al. 1995
Kruip and de Daas 1997
Farin et al. 2001
Miles et al. 2005
Two competing explanations: (1) “normal” for these animals may be heavier than for in vivo produced embryos, or (2) a proportion of animals shifts weight distribution of population
      Adaptation to small changes in biochemical parameters and morphology Sangild et al. 2000  
    IVP, ET Increased gestation length, dystocia, perinatal mortality, fetal edema, altered organ development, abnormal limbs Kruip and den Daas 1997;
Behboodi et al. 1995;
van Wagtendonk-de Leeuw et al. 1998;
Farin et al 2001;
Bertolini and Anderson 2002;
Edwards et al. 2003;
Rerat et al. 2005
Frequency and severity of abnormalities: IVP>ET>AI
    AI 55.9% of abortions due to infection Santos et al. 2004  
Node 2 Perinatal IVP, BNT Increased birth weight,
Increased crown-rump length; increased mortality and physical deformities
Behboodi et al. 1995;
Wilson et al. 1995;
Walker et al. 1996;
Rerat et al. 2005
    IVP Perinatal mortality in IVP ranges from 2.4-17.9%, due to dystocia associated with large fetuses Hasler et al 1995;
van Wagtendonk-de Leeuw et al. 1998;
Block et al. 2003
Lower in heifers than in cows
IVP and SCNT fetuses have cerebellar hypoplasia, respiratory distress, and heart enlargement Schmidt et al. 1996;
van Wagtendonk-de Leeuw et al. 2000;
Chavette-Palmer et al. 2002
    IVP Altered expression of mRNA for non-imprinted myostatin and glyceraldehydes-3-phosphate in IVP fetuses Crosier et al. 2002  
    IVP, SCNT Altered expression of mRNA or protein in IVP and SCNT placentae for VEGF, peroxisome proliferators activated receptor ?, leptin, bovine placental lactogen, transforming growth factor (TGF) ß1, 2, 3, TGF- ß receptor, major histocompatibility class I antigens Davies et al. 2004;
Miles et al. 2004;
Ravelich et al 2004;
Miles et al 2005;
Ravelich et al. 2005
    IVP, SCNT Expression of demethylating enzymes DMT 1, 3a altered in IVP and SCNT preimplantation embryos Wrenzycki et al. 2004  
    BNT Birth weight range 26.4 to 67.3 kg; slow to stand, poor suckling behavior, flexor tendon deformities, hypoxemia, hypoglycemia, acidosis, hypothermia; altered metabolic hormones (thyroxine, triiodothironine, and insulin) Garry et al. 1996  
    BNT Calving rate ~50% using high quality embryos; some very large calves (up to 70.5 kg); contracture of limbs and spine, cardiac and skull deformities noted in a few calves; high rate of dystocia (52/100); hydroallantois observed in four cows Willadsen et al. 1991  
    IVP, SCNT Increased incidence of dystocia and C-section deliveries for IVP pregnancies compared to AI/NM; lack of contractility and other signs of labor in ewes; higher mortality among IVP and SCNT compared to AI/NM Ptak et al. 2002 Sheep
    AI Heat stress reduces birth weight and passive transfer of immunity and results in low IgG concentration in calves Collier et al. 1982 High levels of glucocorticoids accelerates “gut closure”
Nodes 2-3 Postnatal IVP Increased feed intake and growth rate Rerat et al. 2005  
    IVP Altered glucose and electrolyte metabolism compared to AI persisting through early juvenile period Rerat et al. 2005  

1 For the purposes of this table, Developmental Node 1 is divided into three stages of pregnancy: early embryo, late embryo-early fetal, and late gestation.
2 Gd= gestation day or day of pregnancy.

A. Successes and Failures of AI, IVP, and ET

Success of AI depends on a variety of factors, including health of the female and timing of insemination relative to ovulation. In dairy cattle, conception rates to AI following spontaneous estrus have declined from approximately 55 percent in the 1950s to 45 percent in the late 1990s. The use of hormones to synchronize estrus for timed AI has further reduced conception rates to approximately 35 percent. The reasons for this apparent reduction in dairy cow fertility are not clear, although a number of factors have been cited as possibly contributing to the phenomenon, including increased milk production (resulting in increased stress and reduced availability of nutrients for reproductive function), increased average herd size (resulting in fewer person-hours spent observing cows for estrus behavior), nutrition, herd health, inbreeding, and environmental pollution (Lucy 2001). Embryo loss has been estimated to occur at a rate of 10 to 20 percent in dairy cattle (Lucy 2001) and as high as 30 percent in beef cattle (Dunne et al. 2000), and generally occurs prior to 30 days gestation. Fetal losses in swine pregnancies can be as high as 40 percent following AI (Vonnahme et al. 2002). The reasons for these losses in utero are not always apparent. Lucy (2001) noted that embryo loss may occur even in cases where the developing embryo appeared normal. However, in swine, fetal loss appears to be related to the size and efficiency of the placenta (Vonnahme et al. 2002).

Betts and King (2001) noted that the developmental competence (an embryo’s ability to progress through normal cell division and development) of IVP and cultured embryos was low. Using in vitro procedures (as published up to 2001), less than half of inseminated bovine oöcytes reached blastocyst stage, and of those that did, many did not implant or attach following transfer. Betts and King (2001) noted that chromosomal abnormalities such as aneuploidy and polyploidy played a fundamental role in most of these embryonic deaths.

The evolution of IVP technology in cattle can be observed by comparing early studies (conducted prior to 2002) with more recent publications. Studies using IVP embryos during the mid- to late 1990s (Behboodi et al. 1995; Farin and Farin 1995; Hasler et al. 1995; Walker et al. 1996; Drost et al. 1999; Sinclair et al. 1999) noted relatively high rates of embryo loss and LOS among fetuses and neonatal calves. In contrast, several more recent studies using IVP embryos have indicated few or no problems (Chavatte-Palmer et al. 2002; Heyman et al. 2002; Bertolini et al. 2004; Rerat et al. 2004). However, high embryonic mortality and placental abnormalities may still be observed with IVP in some labs (Miles et al. 2004; Miles et al. 2005).

Embryo transfer, in which oöcytes are fertilized in utero then removed and transferred to surrogates, has become a commercially viable technology (See Chapter II), and is generally more successful than IVP. In a study by Drost et al. (1999), initial pregnancy rates, as determined by blood progesterone levels at gestation day 22, were similar among cows bred by AI, ET or IVP. However, by gestation day 42, embryo loss among cows receiving IVP embryos was higher than either AI or ET, while pregnancy rate was similar between cows bred by AI compared to those receiving ET embryos. The success of ET may be affected by treatment of the embryo prior to transfer and synchrony between the surrogate and the embryo donor (Pope 1988; Spell et al. 2001). According to Spell et al. (2001) fresh embryos had a higher rate of survival than embryos that had been frozen then thawed prior to transfer. Embryo survival was also higher when surrogates had been in estrus within 12 hours of the embryo donor (Spell et al. 2001).

In order to follow fates of client-owned pregnant cows carrying IVP-derived pregnancies in a commercial ET operation, Hasler et al. 1995 noted that for the first 100 transfers, 24 ended in pregnancy loss before 100 days of gestation. The success rate improved the subsequent year, however, with only 7 percent of IVP-derived pregnancies spontaneously aborting. They compared these results to 5.3 percent of ET pregnancies aborted between two and seven months of gestation in an earlier study.

In a comparison of AI and IVP, Behboodi and coworkers (1995) noted an increased incidence of dystocia and Cesarean sections (C-section) for IVP derived pregnancies compared to AI in a small group of cattle (8/13 IVP-derived pregnancies vs. 7/71 AI pregnancies requiring C-section). Birth weights of calves derived from IVP embryos were higher than calves produced by AI, likely contributing to the observed increase in dystocia among dams carrying IVP-derived pregnancies. Sinclair et al. (1999) also observed large IVP-derived fetuses with altered development and excessive amounts of amniotic fluid. In that study, nine of 13 fetuses (69 percent) derived from embryos co-cultured with granulosa cells (a type of cell found in the ovary) and one of six embryos (17 percent) incubated in synthetic oviductal fluid (SOF) plus steer serum were oversized, while embryos that had been incubated with SOF alone produced normal sized fetuses. Bovine embryos cultured for three to five days post-fertilization also were associated with increased dystocia due to oversized calves in a study by Walker et al. (1996). (See discussion of influence of culture conditions on success rates in Chapter IV).

Farin and Farin (1995) collected bovine IVP and ET fetuses from beef heifers at seven months gestation and compared development between the two groups. Fetuses from the IVP group were heavier than their ET counterparts (18.6 ±1.1 vs. 15.4 ±0.8 kg), had greater heart girths (56.5 ± 1.2 vs. 52.4 ± 0.9 cm) and weights (139.7 ± 8.3 vs. 116.2 ± 5.8 g), and greater long bone lengths (23.1 ± 0.6 vs. 21.3 ± 0.4 cm).When organ and skeletal measures were compared on a per kilogram body weight basis, however, IVP fetuses had consistently smaller skeletal measures than ET fetuses. Internal organ weights per unit of body weight were not different between the two groups of calves. The authors concluded that IVP fetuses were undergoing abnormal and disproportionate development compared to ET fetuses. It should be noted that the most rapid period of prenatal growth in cattle is during the last two months of gestation (months 8 and 9) (NRC 2001), which would have occurred after these pregnancies were terminated.

Young and Fairburn (2000) noted that both IVP and embryo culture have resulted in abnormal phenotypes, including up to two-fold increases in birth weight (LOS), excess amniotic fluid, hydrops fetalis1, altered allometric organ growth2, advanced fetal development, placental and skeletal defects, immunological defects, and increased perinatal death.

Markette et al. (reviewed by Farin et al. 2001) observed that 54.7 percent of ET recipients were pregnant at 60 days gestation, with the majority of pregnancies lost prior to day 24 of gestation. In a large study, King et al. (1985) reported that the incidence of pregnancy loss in 1,776 embryo transfer recipients was 3.15 percent from 2 to 3 months of gestation, and 2.14 percent between 3 to 7 months. These mid- and late-gestation spontaneous abortions were not influenced by embryo age, embryo quality, time between embryo collection and transfer, asynchrony of recipient with donor estrus, donor age, ovarian response to gonadotropin treatment, or whether or not the donor had a history of infertility, according to the authors. In most studies, pregnancy loss during the fetal period (day 42 to 280 of gestation in cattle) was greater following transfer of embryos produced in vitro than that for embryos produced in vivo. Mid- to late-gestation spontaneous abortion of about 7 to 13 percent has been reported for recipient cattle carrying fetuses derived from IVP embryos, and in some studies pregnancy loss has been considerably higher (Farin et al. 2001).

Conversely, Bertolini et al. (2004) compared fetal development in in vivo and IVP cattle pregnancies and reported no significant difference between groups for pregnancy rates (20/53 and 36/112 for control and IVP groups respectively) and fetal losses after day 45 (2/20 and 3/36 for control and IVP groups respectively). They did report that fetal losses between gestation days 30 and 44 were 3.4-fold higher (P<0.05) in the IVP group (17/36) than in controls (4/20). Also in contrast to earlier studies, Bertolini et al. (2004) reported that their measurements of conceptus physical traits for both in vivo produced controls and IVP pregnancies on days 90 and 180 demonstrated allometric proportionality between fetal body size and body weight with no physical deformities observed in any fetus.

In a review of research on early embryo development, Gardner and Lane (2005) stated that the environment of the preimplantation embryo has a profound effect on the physiology and viability of the conceptus. Among the many factors that can influence development of IVP embryos, they cite the use of serum products as an important contributor to developmental abnormalities in cultured embryos. These authors state: “Mammalian embryos are never exposed to serum in vivo…Rather, serum is a pathological fluid, the composition of which is greatly undefined and varies enormously with source…serum induces premature blastulation in domestic animal embryos…affects embryo morphology…and leads to perturbations in ultrastructure…and energy metabolism.” Other factors that may influence development of embryos in vitro include ammonia, oxygen, inadequate nutrients, and freezing (Gardner and Lane 2005).

Rerat et al. (2004) compared the perinatal health characteristics of IVP and AI cattle and observed no differences in post-natal mortality or viability. Calves in this study were generally healthy with the health status of IVP calves at birth and during the first 112 days of life similar to that of AI calves. Clinical traits such as heart rate, rectal temperature, and respiratory rate were nearly identical in both groups. At birth, measurements indicative of growth performance such as potassium, 3,5,3’-triiodothyronine (a metabolic hormone), and thyroxine concentrations were lower in IVP than in AI calves. Postnatally, IVP calves had a faster growth rate than AI calves under conditions of identical nutrient intake.

Sakaguchi et al. (2002) induced twinning in Japanese beef cows by transferring one or two in vivo fertilized embryos into AI bred cows. Fetal dystocia occurred in 7 of 14 twin parturitions, in which some twin calves appeared to enter the uterine cervix at the same time, but no single parturition was accompanied by dystocia. The incidence of retained placenta was significantly higher in the twin parturitions (10/14; 71 percent) than in the single parturitions (2/22; 9 percent). These complications are known to occur with natural twins in cattle, however, and may not be directly related to ET technology. The incidence of retained placenta in healthy, single calf-bearing dairy cows is approximately 5-15 percent, (slightly lower in beef cows) and is increased when there are twins. The expected incidence of dystocia is 10-15 percent in first-parity animals, and 3-5 percent in mature cows (Merck Veterinary Manual Online 2002).

B. Outcomes for BNT, Fetal- and Adult-Cell SCNT

Although success rates for various types of cloning have improved, they are still highly variable across studies. In earlier studies, generally less than 10 percent of all NT embryos transferred to recipients were born alive (Wells et al. 1999). Some of these early studies noted that both blastomere and somatic cell NT clones appeared to have the same low success rate and exhibited many of the same problems, such as poor or dysfunctional placentation and LOS (Stice et al. 1996; Wells et al. 1998). Stice et al. (1996) reported that no fetuses derived from BNT survived beyond day 60 of gestation. Wells et al. (1998) reported a 64 to 80 percent pregnancy loss during the attachment phase for clone fetuses derived from an embryonic sheep cell line, while a further 43 percent of pregnancies were lost in the last trimester, such that 11 percent of embryos survived to term (12/112). In contrast, Le Bourhis et al. (1998) reported 9/30 transferred male bovine BNT clones developed to calving, while 6/27 female BNT clones resulted in live calves. Heyman et al. (2002) compared development and survival of BNT, fetal and adult NT clones to IVP-derived embryos under the same culture conditions. Pregnancy loss from 90 days of gestation to calving were 43.7 percent for adult and 33.3 percent for fetal SCNT, compared to 4.3 percent for BNT clones, while none of the IVP-derived pregnancies were lost. Pace et al. (2002) reported 75 percent pregnancy loss of adult (some transgenic) SCNT embryos throughout pregnancy.

Results from these studies may reflect the evolution of NT technology over time. Embryonic or BNT cloning and IVP success rates appear to have improved. Although losses remain high for the newer SCNT technology, success rates for this technology also have improved over time. It remains to be seen what progress may be made in further reducing pregnancy loss and other risks associated with SCNT.

C. Conclusions regarding outcomes for ARTs

Based on the studies reviewed, there appears to be a general trend indicating that the frequency of embryo/fetal loss and abnormal pregnancy outcomes increases with increasing manipulation of the embryo and in vitro culture. This trend is evident, even when maturity of the technology is considered. Causes of embryo/fetal loss are not always evident, but late gestational complications (hydrops and dystocia) and fetal/neonatal abnormalities (skeletal and organ deformities, oversize, metabolic alterations) have all been noted in ET, IVP, BNT and SCNT. The frequency of these outcomes varies somewhat among laboratories, but has the general trend ET<IVP<BNT<SCNT. These data support a conclusion that SCNT falls on a continuum of ARTs, and that the adverse outcomes noted with SCNT are not unique, but are of concern due to their increased frequency.


 1Accumulation of fluid in the entire body of the newborn.

 2 Allometric growth refers to differences in the rate of growth of a particular organ or part in relation to the rest of the organism. An example of normal allometric growth is the legs of a newborn foal (horse) in proportion to its body size; the legs are long and out of proportion to the rest of the body. As the foal ages, the body grows (and fills out) more rapidly than the legs, so that in adult horses the legs appear proportional to the rest of the body. Altered allometric growth in the context of ARTs has resulted in enlarged hearts and undersized kidneys, as well as other organs, which are not appropriately proportioned to the rest of the body.