Laparoscopic Oviductal Embryo Transfer and Artificial Insemination in Felids – Challenges, Strategies and Successes


Author's address (for correspondence): WF Swanson, Center for Conservation and Research of Endangered Wildlife, Cincinnati Zoo & Botanical Garden, Cincinnati, 45220 OH, USA. E-mail:


Embryo transfer (ET) and artificial insemination (AI) are potentially invaluable techniques for the propagation and management of genetically valuable domestic cat and endangered nondomestic cat populations. Many of the challenges that impair the effective application of ET and AI in felids may be overcome by using laparoscopic oviductal (LO) approaches. LO-ET and LO-AI are minimally-invasive procedures, requiring only two small skin incisions for insertion of a laparoscope and grasping forceps into the abdominal cavity to permit visualization and catheterization of the oviduct for embryo or semen deposition. With concurrent improvements in embryo culture systems and ovarian synchronization protocols, LO-ET has proven effective over the past decade for propagation of laboratory cats, cat models of hereditary disease and nondomestic cats. To date, viable offspring have been produced following LO-ET of non-frozen and frozen-thawed IVF-derived embryos in eight cat hereditary disease models and two nondomestic cat species, the ocelot and sand cat. LO-AI with low sperm numbers (c. 2–8 million motile) has shown similar efficacy to LO-ET, resulting in high pregnancy percentages (50–70%) following insemination of gonadotropin-treated domestic cats. Multiple kittens also have been produced in two hereditary disease models following LO-AI with frozen semen, and both ocelot and Pallas' cat kittens have been born after LO-AI with freshly-collected semen. The application of LO-ET and LO-AI to felids has resulted in substantial improvement in the efficiency of assisted reproduction for genetic management of these invaluable domestic cat and wild cat populations.


Over the past 30 years, assisted reproductive techniques, such as embryo transfer (ET) and artificial insemination (AI), have shown tremendous promise in felids for propagation of companion animals, laboratory cats of biomedical importance and endangered nondomestic cat species (Pope et al. 2006; Swanson 2006; Howard and Wildt 2009). The ultimate success or failure of assisted reproduction in cats is dependent on the ability to produce viable offspring in an efficient and humane manner for population or species management. Beginning with the first successful AI and ET procedures in domestic cats in the 1970s (Sojka et al. 1970; Kraemer et al. 1979), multiple studies have demonstrated the general feasibility of using various approaches for assisted reproduction, including vaginal AI, transcervical AI and ET, and intrauterine AI and oviductal AI/ET via laparotomy (Table 1). Although each of these methods have resulted in birth of live kittens, practical application in domestic cats has been hindered by poor efficiency in some instances and the requirement for invasive surgical intervention for performing intra-abdominal reproductive procedures. These limitations become amplified with the attempted extrapolation of research findings in domestic cats to the propagation of nondomestic felids, a group of species in which our understanding of basal reproductive physiology is even more restricted and animal welfare issues are an even greater concern than in domestic cats.

Table 1. Production of viable offspring in the domestic cat using different approaches for artificial insemination (AI) and embryo transfer (ET)
VaginalSojka et al. (1970)Not applicable
TranscervicalChatdarong et al. (2007)Swanson and Godke (1994)
LaparotomyTsutsui et al. (2000)Kraemer et al. (1979)
LaparoscopyHoward et al. (1992)Not reported
LaparotomyTsutsui et al. (2001)Goodrowe et al. (1988)
LaparoscopyConforti et al. (2011)Swanson et al. (2001)

The application of laparoscopy to reproductive studies in felids has been invaluable for helping to alleviate some of these animal welfare concerns while allowing access to intra-abdominal reproductive organs through a minimally-invasive, minimally-traumatic approach (Wildt et al. 1977). Without laparoscopy, the extrapolation of assisted reproduction to the genetic management of non-domestic cats likely would be unattainable in the near future. Laparoscopic methods for oocyte collection and intrauterine insemination have been used extensively over the past 20 years with numerous cat species (Goodrowe et al. 1988; Howard et al. 1992; Howard and Wildt 2009). More recently, laparoscopic approaches have been developed and applied in cats for gaining access to the oviduct, specifically to conduct laparoscopic oviductal embryo transfer (LO-ET) and artificial insemination (LO-AI) procedures (Swanson et al. 2001; Conforti et al. 2011; Lambo et al. 2012). In going LO, our primary objectives were to overcome some of the anatomical, behavioral, and physiological challenges intrinsic to ET and AI in felids and begin producing genetically-valuable offspring in both domestic and nondomestic cats with improved efficiency while minimizing procedural animal welfare concerns.


The primary challenges to conducting ET and AI in felids include anatomical constraints, behavioral factors, physiological variation among species and individuals, and a relatively poor understanding of feline reproductive biology. Female domestic cats average just 3–4 kg in body mass and this small size creates anatomical constraints not typically encountered in larger domesticated mammals. Cats have a constricted vaginal lumen, connected by an obliquely orientated cervical os to an elongated bicornuate uterus, and relatively small ovaries partially enclosed by ovarian bursae (Nickel et al. 1979). These small dimensions and narrow luminal diameters create logistical difficulties in penetrating the cervix and accessing both distal uterine horns with any confidence, and in allowing thorough evaluation of reproductive tract morphology. Behavioral factors that can complicate assisted reproductive procedures, especially in nondomestic cats, may include overt aggression, intractability, and the occurrence of ‘silent’ estrus in some females. Although vaginal AI may be conducted without sedation in many domestic cats, all other approaches require anesthesia of the female, potentially interfering with neuroendocrine reflexes and behavioral responses intrinsic to ovulation and sperm transport. The inability to reliably determine estrus (or luteal) status in many females based solely on their behavior confounds scheduling of procedures to correspond to either natural or gonadotropin-induced estrus. Other challenges are related to felid reproductive physiology including seasonality, variable ovulatory mechanisms (spontaneous or induced), and limited sperm production and quality in some species and/or individuals. Within a laboratory research colony, the impact of seasonality may be lessened by using controlled light cycles and fecundity can be optimized through selective breeding, but colony housing also may increase the tendency for spontaneous ovulation in many females. With nondomestic felids, reproductive physiology is even more variable and not subject to consistent environmental or selective influences across zoological institutions. The limited availability of adequate numbers of viable, highly motile spermatozoa, especially in small-sized cats, impairs the potential for conducting either vaginal or uterine AI procedures (Table 2). Lastly, our poor understanding of in vivo sperm transport and storage mechanisms, in vitro embryo developmental requirements and appropriate ovarian synchronization methods reduces the likelihood of success with many AI or ET procedures. For example, with uterine ET, one primary concern has been the potential adverse effects of prolonged in vitro culture on embryo viability and health of resulting offspring. Although improvements in culture medium have been reported and blastocysts can be routinely produced after 6–8 days of culture, pregnancy success after transfer of uterine stage embryos (morulae or blastocysts) remains poor (Pope et al. 2006; Herrick et al. 2007). The outcome of both ET and AI also depends on the use of ovarian synchronization protocols to simulate a suitable oviductal or uterine environment that supports conception and embryo/fetal development. Alterations in standard exogenous gonadotropin protocols has allowed some improvement (Magarey et al. 2005), but further research and refinement are necessary.

Table 2. Relative merit of vaginal, uterine and oviductal artificial insemination (AI) with freshly-collected semen in small cat speciesa
Cat speciesTotal sperm/ejaculate (×106)bVaginal AI/ejaculatecUterine AI/ejaculatedOviductal AI/ejaculatee
  1. a

    Adapted from Swanson et al. 2007.

  2. b

    Mean (±SEM).

  3. c

    Based on 40 × 106 motile sperm per AI.

  4. d

    Based on 10 × 106 motile sperm per AI.

  5. e

    Based on 2 × 106 motile sperm per AI.

Fishing cat56.4 ± 17.71420
Ocelot129.4 ± 39.421050
Black-footed cat28.4 ± 6.30210
Sand cat43.5 ± 11.00315
Pallas' cat25.2 ± 3.20210


Using LO-ET and LO-AI, most anatomical barriers can be bypassed and the ovaries, oviducts and uterus directly visualized to assess ovulatory responses and any reproductive tract pathology (Wildt et al. 1977). Laparoscopic aspiration of in vivo matured ovarian follicles followed by in vitro fertilization has allowed generation of cat embryos in more than 20 felid species, and live offspring have been produced following ET of IVF-derived embryos in nine species to date (domestic cat, wild cat, tiger, ocelot, serval, caracal, fishing cat, black-footed cat, sand cat). Most of these earlier ET procedures, however, required laparotomies to gain access to the oviducts or uterus. With LO-ET, embryos can be transferred directly into the oviducts at an early cleavage stage (2–4 cells), limiting the duration of in vitro culture to 24–30 h and lessening the negative influences of a suboptimal in vitro environment, while eliminating the need to conduct more invasive laparotomy procedures.

Laparoscopic intrauterine AI has been used successfully to produce offspring in nine cat species (domestic cat, leopard cat, cheetah, tiger, puma, snow leopard, clouded leopard, ocelot, tigrina), although pregnancy percentages have been low for most species. One primary limitation with laparoscopic intrauterine AI (as well as vaginal and transcervical AI) is the requirement for fairly high numbers of motile sperm (>10 million/AI) to ensure sperm penetration through the uterotubal junction into the oviducts to achieve fertilization (Table 2). With LO-AI, much lower sperm numbers may be used, allowing multiple AI procedures to be conducted with a single ejaculate (Table 2). Because the need for extensive sperm transport in vivo is eliminated, LO-AI also permits insemination with sperm samples of lower quality, such as frequently occurs after cryopreservation and thawing.

Our laboratory began exploring the LO-ET approach in domestic cats in the late 1990s (Swanson et al. 2001) and have continued refining this methodology over the past 12 years for LO-ET and, more recently, LO-AI. The standard laparoscopic approaches used for LO-AI and LO-ET are similar, both requiring general anesthesia of the female, placement in dorsal recumbency and sterile preparation of the abdominal region, followed by tilting of the surgical table (i.e. Trendelenburg's position) and insertion of sterile laparoscopic instruments. For AI and ET, only two small skin incisions are needed for placement of cannulae – one for mid-ventral abdominal insertion of a 7–10 mm laparoscope and the second for right lateral insertion of specialized grasping forceps (Fig. 1). To obtain an optimal design for the grasping forceps, a surgical instrument company (MDS Incorporated, Brandon, FL, USA) was consulted to construct stainless steel – plastic composite forceps with the following attributes: tip (tapered, serrated jaws), shaft (5 mm diameter, 30 mm length, rotational on central axis), and handle (retaining clip, self-holding in closed position). For effective cold sterilization, easy dismantling and reassembly of the primary forceps components (tip, shaft, handle) also was a requirement.

Figure 1.

Laparoscopic grasping forceps [with a tapered serrated tip (inset photo), long rotational shaft and self-closing handle] and an accessory trocar/cannula for laparoscopic oviductal artificial insemination (LO-AI) and laparoscopic oviductal embryo transfer (LO-ET) procedures

For LO-ET, the grasping forceps are used to secure the craniomedial edge of the bursa covering either ovary and the bursa is extended and everted to visualize the abdominal oviductal ostium (Fig. 2a). A polypropylene i.v. catheter (20 g, 5 cm length; Sherwood Medical Co. St. Louis, MO, USA) is placed through the ventral abdominal wall dorsal to the ovary and inserted (c. 2 cm) through the oviductal ostium into the distal oviduct. Polyethylene tubing (25 cm length, PE10; Bectin Dickinson Co. Sparks, MD, USA), attached to a blunted 30 g (1.25 cm) needle and 1 ml syringe, is passed through the catheter and, with continued forward pressure on the tubing, the catheter is completely withdrawn from the oviduct. Embryos (n = 3–7, depending on species), contained in 5–10 μl of culture medium at the distal end of the transfer tubing, are expelled deep into the oviductal lumen with slight air pressure from the syringe. For LO-AI, the procedure is similar, except a shorter i.v. catheter (18 g, 3.2 cm length; Terumo Medical Corporation, Elkton, MD, USA) is placed through the ventral abdominal wall dorsal to the ovary for passage of the insemination needle. For AI, a blunted steel needle (22 g, 6.8 cm length), derived from the stylet within an i.v. catheter (20 g, 5.0 cm length; Sherwood Medical Co.), is attached to a 1 ml syringe and used to aspirate a low volume (10 μl) of highly-concentrated semen (c. 100–400 × 106 motile sperm/ml). The AI needle is passed though the oviductal ostium and the semen is deposited deep within the oviduct using slight air pressure. Surgical closure of each skin incision requires just 1–2 sutures and a small amount of tissue adhesive, and cats typically return to normal activities shortly following anesthetic recovery. During the past 20 years, we have conducted more than 1000 laparoscopic procedures in domestic cats and over 300 laparoscopic procedures in nondomestic cats with just one procedure-related mortality (caused by respiratory compromise during laparoscopy) occurring within each study group and without a single post-operative infection.

Figure 2.

Laparoscopic view of (a) oviductal embryo transfer using polyethylene tubing (arrow) passed through a polypropylene i.v. catheter for embryo deposition and (b) oviductal artificial insemination using a blunted, stainless steel needle (arrow) for sperm deposition


Over the past decade, LO-ET has been used extensively in our laboratory for reproductive studies with domestic cats, and for propagation of cat models of hereditary disease and nondomestic cat species. Our research in domestic cats determined that a modified gonadotropin regimen, using 100 IU eCG with 1000 IU pLH, produced consistent ovulatory responses, and superior synchronization compared to the standard eCG/hCG regimen (Magarey et al. 2005). Laparoscopic transfer of non-frozen IVF embryos (5 embryos/female) into the oviducts of this recipient type resulted in 87% (13/15) becoming pregnant with c. 60% of transferred embryos developing to term in pregnant females (Magarey et al. 2005; G. M. Magarey, J. B. Bond, H. L. Bateman, J. R. Herrick, W. F. Swanson, unpublished data). Subsequent LO-ET studies have focused on using this approach for propagation of the 24 cat hereditary disease models stored as frozen semen or embryos within our biological resource bank. These disease models have been banked in collaboration with seven veterinary and medical schools as an alternative to maintaining perpetual living colonies. LO-ET of non-frozen or frozen-thawed IVF embryos into eCG/pLH-treated recipients has allowed production of viable kittens for eight cat hereditary disease models (mucopolysaccharidosis I and VI, spinal muscular atrophy, Chediak-Higashi syndrome, lipoprotein lipase deficiency, Niemann-Pick type C, porphyria, progressive retinal atrophy). Most of these kittens were returned to collaborating veterinary schools to reestablish breeding colonies or for ongoing studies (Swanson et al. 2012). This LO-ET approach also has been extrapolated to propagation of nondomestic felids with the production of multiple pregnancies and viable offspring in the ocelot (Leopardus pardalis) and sand cat (Felis margarita). In research with ocelots in the US and Brazil, a total of five pregnancies and five term kittens have been born from 10 LO-ET attempts with frozen-thawed IVF embryos, and, in studies with sand cats in the United Arab Emirates, one pregnancy and two kittens were born following LO-ET of non-frozen embryos into four recipients.

Our initial study investigating the application of LO-AI in domestic cats assessed the relative fertility of low numbers (c. 1 million motile/site) of freshly-collected spermatozoa deposited within the oviduct vs the distal uterus (Conforti et al. 2011). Females (n = 16) were inseminated in one oviduct with spermatozoa collected from one male and in the contralateral uterine horn with spermatozoa collected from a second male, and then spayed 20 days post-AI for paternity testing of resulting fetuses. Overall, 69% (11/16) of females conceived, with most (73%, 36/49) of the fetuses produced by spermatozoa deposited within the oviduct. A subsequent LO-AI study assessed the relative fertility of freshly-collected semen (1 million motile) vs semen frozen in a novel soy-lecithin based cryoprotectant (c. 2 million motile; Lambo et al. 2012). Half (8 of 16) of the inseminated females conceived and subsequently gave birth to 36 kittens, including eight kittens produced by semen cryopreserved in the soy medium. Using slightly higher sperm numbers (4–8 million motile total), LO-AI with frozen semen also has been used to propagate two cat disease models (hypertrophic cardiomyopathy, mucopolysaccharidosis I). LO-AI of nine females resulted in six pregnancies (67%) and the generation of 25 offspring (Swanson et al. 2012; W. F. Swanson, H. L. Bateman, J. Newsom, C. Lambo, unpublished data). Lastly, LO-AI with freshly-collected semen has been used in US zoos to produce one pregnancy and one viable kitten in the ocelot and one pregnancy and three viable kittens in the Pallas' cat (Otocolobus manul).


The application of LO-ET and LO-AI, in conjunction with modifications in ovarian synchronization methods and embryo culture systems, has resulted in meaningful improvements in pregnancy percentages and offspring production in both domestic cats and nondomestic cat species. The development of these minimally invasive approaches have allowed us to overcome some of the intrinsic challenges associated with assisted reproduction in felid species while minimizing procedural discomfort and animal welfare concerns. Continued refinement, especially further investigation of recipient synchronization and cat semen cryopreservation methods, may be necessary to obtain greater efficiency for routine use with genetic management of valuable domestic and wild cat populations.


The author thanks CREW post-doctoral fellows (Genevieve Magarey, Valeria Conforti, Colleen Lambo, Jason Herrick), and research/cat keeper staff (Helen Bateman, Jackie Newsom, Rachel Carpenter, Carla Mascari) for their tremendous contributions to the completion of the studies described in this paper. These studies were funded, in part, by the Institute of Museum and Library Services and the National Institutes of Health (which were not involved with the execution of this research).

Conflicts of interest

None of the authors have any conflicts of interest to declare.

Author contributions

The author conceived and wrote this review paper.