Introduction

In the US, routine canine sterilization is currently recommended at an early age in animals not intended for breeding. Recently there has been a growing interest in the potential role of gonadal hormones in the development of select canine morbidities. Some dog owners have become interested in exploring alternative sterilization procedures, such as ovary-sparing spay (OSS), in an attempt to minimize potential consequences of traditional gonadectomy (ie, ovariectomy [OVE] and ovariohysterectomy). Proposed gonadectomy consequences include an increased risk for urinary incontinence,^1–3 cranial cruciate ligament (CCL) disease,^4–7 hip dysplasia,^8,9 and certain neoplastic conditions such as osteosarcoma,^10,11 lymphoma,^9,12,13 transitional cell carcinoma,¹⁴ mast cell tumors,^9,12 and hemangiosarcoma.^9,12,13 Ovariohysterectomy has also been associated with the development of obesity^15–17 and behavioral issues.^15,18,19 Furthermore, some of the proposed effects seem more pronounced if gonadectomy is performed at an early age (prior to 1 year) and in certain breeds (ie, Golden Retriever, Vizsla, and Rottweiler).^9,11–13

However, there is a paucity of information regarding the outcome of these alternative gonad-sparing sterilization techniques. A relatively novel technique, OSS involves a complete hysterectomy while sparing the ovaries. This procedure has been developed to allow the maintenance of ovarian hormones, including estrogen and progesterone, while removing the uterus to prevent pregnancy and uterine disease such as pyometra. Currently, no evidence has been presented in regard to its efficacy in sterilization or development of stump pyometra or other sex hormone–related disorders, and it has been sparsely performed in specialty practice. Furthermore, to the authors’ knowledge, there are no previous studies that have compared laparoscopic-assisted OSS (hysterectomy; LapOSS) to laparoscopic OVE (LapOVE) regarding short- or long-term outcomes.

Therefore, our objective was to evaluate the incidences of orthopedic, behavioral, urinary/reproductive, neoplastic, or atopic disease processes for dogs undergoing either a LapOSS or LapOVE. We hypothesized that dogs undergoing elective LapOSS would have an increased incidence of reproductive disease such as stump pyometra in comparison to those dogs undergoing LapOVE.

Materials and Methods

Medical records of client-owned dogs presenting to Washington State University Veterinary Teaching Hospital between August 2013 and May 2020 for elective LapOSS or LapOVE were reviewed by the authors (MAO and CGS). The sterilization performed was elected by the owner and was not part of a randomized grouping by the authors. A multiple-choice client questionnaire was emailed to all clients in which complete medical records including history, physical examination, and surgery report were available (Qualtrics). Dogs with complete medical records and a complete client follow-up survey were included for analysis.

Laparoscopic-assisted ovary-sparing spay (hysterectomy)

The first portal was introduced using a threaded cannula, either 6 or 11 mm, (EndoTIP; Karl Storz Veterinary Endoscopy) through an incision on ventral midline about 1 cm caudal to the umbilicus with the Ternamian visual entry technique.²⁰ After confirmation of entry into the abdomen, the abdomen was insufflated to 8 to 12 mm Hg and a 5-mm 30° laparoscope was introduced to facilitate visual placement of additional cannulas. The second portal was placed halfway between the xiphoid and umbilicus and the third close (ie, 2 to 5 cm cranial) to the pubis. A brief abdominal explore was performed prior to the hysterectomy. The patient was placed in 10° Trendelenburg with a 15° tilt to 1 side to allow the uterine horn to be grasped, cauterized, and transected using a 5-mm vessel sealing device or laparoscopic curved Metzenbaum scissors with monopolar cautery (Endo Shears; Stryker) at the level of the proper ligament. The transection of the mesosalpinx was done close to the ovary to ensure that no cranial endometrial remnants remained. The caudal cannula was removed, allowing depressurization of the abdomen, and that portal was extended to allow exteriorization of the uterus. Appropriately sized polydioxanone (PDS II; Ethicon Inc) was used to ligate the cranial vagina immediately caudal to the cervix with 1 encircling and 1 transfixing ligature, and the cervix was transected, removing all uterine glandular tissue (Figure 1).²¹ After confirmation of adequate hemostasis, all cannulas were removed and portals closed in a routine fashion.

Intraoperative image of the fully exteriorized uterine body including the cervix. The red dotted line represents the area of ligation and transection caudal to the cervix (A). Examined cut surface of the uterine body and cervix. Black arrows demonstrate the distance between the glandular tissue of the uterine body and cut edge (B).

Citation: Journal of the American Veterinary Medical Association 261, 8; 10.2460/javma.22.12.0539

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Intraoperative image of the fully exteriorized uterine body including the cervix. The red dotted line represents the area of ligation and transection caudal to the cervix (A). Examined cut surface of the uterine body and cervix. Black arrows demonstrate the distance between the glandular tissue of the uterine body and cut edge (B).

Citation: Journal of the American Veterinary Medical Association 261, 8; 10.2460/javma.22.12.0539

• Download Figure
• Download figure as PowerPoint slide

Intraoperative image of the fully exteriorized uterine body including the cervix. The red dotted line represents the area of ligation and transection caudal to the cervix (A). Examined cut surface of the uterine body and cervix. Black arrows demonstrate the distance between the glandular tissue of the uterine body and cut edge (B).

Citation: Journal of the American Veterinary Medical Association 261, 8; 10.2460/javma.22.12.0539

• Download Figure
• Download figure as PowerPoint slide

Results

Complete medical records were obtained for 23 dogs undergoing LapOSS and 41 dogs undergoing LapOVE. Client surveys were completed for 17 of 23 dogs (73% response rate) in the LapOSS group and 16 of 41 dogs (39% response rate) in the LapOVE group. Therefore, a total of 33 dogs were included in the study: 17 LapOSS dogs and 16 LapOVE dogs. The mean follow-up time was 4.2 ± 1.8 years for the LapOSS group and 4.3 ± 2.0 years for the LapOVE group. At surgery, the mean age and body weight for LapOSS and LapOVE were 1.9 ± 2.0 years and 24.1 ± 10.7 kg and 2.1 ± 1.4 years and 26.3 ± 16.4 kg, respectively (Table 1). Two of the dogs in the LapOSS group were deceased at follow-up.

Breeds in the LapOSS group included Golden Retriever (n = 5), Labrador Retriever (2), Bernese Mountain Dog (2), mixed breed (2), Brittany Spaniel (1), German Shepherd Dog (1), Anatolian (1), Vizsla (1), Boxer (1), and Beagle (1).

Breeds represented in the LapOVE group included Labrador Retriever (n = 5), mixed breed (2), Alaskan Husky (1), German Shepherd Dog (1), French Bulldog (1), Yorkshire Terrier (1), Pudelpointer (1), English Mastiff (1) and Cane Corso (1).

Ten of 17 (59%) dogs in the LapOSS group and 6 of 16 (38%) dogs in the LapOVE group (P = .1694) had preexisting conditions including atopy (n = 7), urinary disease (5), orthopedic disease (2), gastrointestinal disease (2), behavioral issues (1), and subaortic stenosis (1). The distribution of preexisting conditions between groups is presented (Table 2).

Of the patients undergoing either sterilization procedure, 19 of 33 had heat cycles before treatment. Twelve dogs in the LapOSS group had experienced 1 (n = 5), 2 (3), 3 (1), or more (2) heat cycles prior to their procedure, and 7 dogs in the LapOVE group had experienced 1 (2), 2 (1), or 3 (4) heat cycles.

Seven LapOSS dogs and 6 LapOVE dogs had additional procedures performed during their sterilization surgery. Additional procedures were performed on a case-by-case basis as deemed appropriate by the attending surgeon or resident including laparoscopic gastropexy (n = 10), hiatal herniorrhaphy and esophagopexy (1), umbilical herniorrhaphy (1), brachycephalic airway surgery (1), mass removal (1), or episioplasty (1). Intraoperative complications included repeat laparoscopic entry (n = 1), moderate bleeding (1), and a splenic laceration (1). None of these resulted in a need for additional treatment perioperatively or conversion to an open laparotomy.

The follow-up times in the LapOSS and LapOVE groups were 4.2 ± 1.8 years and 4.3 ± 2.0 years, respectively (P = .9380) and overall ranged from 2 to 8 years. Of the 17 dogs undergoing LapOSS, 5 of 17 (29%) underwent an OVE at a later date. Three of these 5 OVE procedures were performed within 2 years of the LapOSS procedure, and 1 was performed 6 years following the original surgery. Reasons for OVE following LapOSS procedure noted included undesirable heat cycles (n = 2), mammary gland neoplasia (1), cystic ovary (1), and diabetes (1).

There were no behavioral changes seen in either group within 1 year of the sterilization procedure. One patient was noted to have aggressive behaviors (aggression toward people in the home) within 1 month following LapOSS sterilization, although this was also noted prior to the procedure. Three additional dogs began showing aggressive behaviors (dog reactivity in the home, cat reactivity) after 1 year, 1 in the LapOSS group and 2 in the LapOVE group (P = 1.0).

Three of 17 (18%) dogs in the LapOSS group developed orthopedic disease. One dog in the LapOSS group was diagnosed with a CCL tear, but this dog had an ovariectomy procedure performed 3 years prior to clinical signs of CCL tear due to diabetes. Two other LapOSS dogs developed biceps tendinopathy and patella luxation. In the LapOVE group, 2 of 16 (12.5%) dogs were diagnosed with orthopedic disease at the time of follow-up (P = 1.0). Two dogs developed pelvic limb lameness, of which 1 was diagnosed with a CCL tear.

Ten of 17 (59%) LapOSS dogs developed estrous cycle–related or clinical signs from the urinary tract or the reproductive tract following sterilization. Estrous cycle–related signs considered a normal postoperative expectation in LapOSS included heat cycles (n = 5), attraction of male dogs (4), clear vaginal discharge (3), increased vulvar licking (2), false pregnancy (2), and bloody vaginal discharge (1) and therefore were not included in comparison to the LapOVE group. Clinical signs indicating urogenital pathology included cystic ovary (n = 1) or urinary tract infection signs (1). None of the LapOSS dogs developed stump pyometra. Of the LapOVE group, 3 of 16 (19%) dogs had owner-reported concerns relating to the urogenital tract, of which none were considered indicative for urogenital pathology, including urine marking (n = 1), attraction of male dogs (1), clear vaginal discharge (1), or increased vulvar licking (1). None of these were showing heat cycles suggestive of ovarian remnants. The remainder of the dogs in the LapOVE group showed no reported urinary or reproductive concerns.

Five of the 17 (29%) LapOSS dogs developed neoplasia following sterilization including mammary gland carcinoma (n = 1), abdominal carcinomatosis (1), lymphoma (1), and benign masses (2; 1 hemangioma and 1 fibrovascular papilloma), while 2 (13%) dogs in the LapOVE group developed neoplastic disease including a mast cell tumor (1) and osteosarcoma (1; P = .3983). Mammary neoplasia thus developed in 1 of 17 (6%) LapOSS dogs and none of the LapOVE dogs (P = 1.0). However, the LapOSS dog with lymphoma developing had undergone a subsequent LapOVE only 1 year after hysterectomy and 4 years before cancer development. This dog was therefore considered to have developed neoplasia post-OVE. One dog in the LapOVE group died from osteosarcoma. A post hoc power analysis was performed using data on the proportions of dogs developing malignant neoplasia other than mammary tumor with intact ovaries, compared to those developing malignant neoplasia post-OVE. With 80% power and .05 significance, a sample size of 124 dogs/group may be needed to demonstrate a difference between groups. A similar analysis was performed on the development of mammary neoplasia, and a sample size of 118 dogs/group may demonstrate a difference between groups.

Of the 17 dogs undergoing LapOSS, 8 (47%) developed skin disease in comparison to 5 of 16 (31%) in LapOVE dogs (P = .4813). Atopic dermatitis and otitis were the most common disorders in both groups. Other integumentary diseases included hot spots, superficial pyodermatitis, allergic rhinitis, and an injection reaction.

Four of 17 (24%) LapOSS dogs experienced other various disorders, of which only 2 appeared having any possible relation with hormonal status: adrenal nodule and elevated ALT (n = 1) and diabetes (1). Within the LapOVE group, 3 of 16 (19%) dogs were displaying clinical signs of disease of which only 1 (polyuria and polydipsia) appeared potentially associated with endocrine/hormonal functions (P = 1.0).

Discussion

On the basis of information obtained from this small study, we concluded that stump pyometra did not develop in any of the dogs undergoing a complete hysterectomy and therefore does not seem to constitute a contraindication to performing this sterilization procedure. In addition, we were unable to detect an increased incidence of malignant neoplasia or other groups of disorders between the 2 study populations. Therefore, we rejected our hypothesis that LapOSS would lead to a higher incidence of reproductive tract diseases such as (stump) pyometra when compared to gonadectomy. Clearly demonstrated differences in development of specific orthopedic or neoplastic diseases could not be expected in a study of such limited subject numbers. However, we anticipated that our results could show important pilot data for further investigations of larger numbers of dogs. The limited development of diseases previously associated with ovarian hormone status^4,7,9–13 such as mammary neoplasia, cranial cruciate disease, or hemangiosarcoma/osteosarcoma were uncommon enough in our population that they may not be of strong clinical relevance for choosing sterilization method. However, the lack of preliminary and confirmed differences between the surgery types indicated that LapOSS may constitute a reasonable sterilization option, which has not been clearly demonstrated previously to the authors’ knowledge. Canine gonadectomy is the most performed surgical procedure worldwide due to the many benefits including population control and decreased incidence of reproductive disease, including pyometra and mammary neoplasia.^{9,13,15,22–29} A Swedish epidemiological study³⁰ demonstrated very high disease incidences of mammary tumors and pyometra in a population of 260,000 predominantly intact female dogs. The risk for a female dog to contract pyometra was 19%; mammary neoplasia, 13%; and either of these disorders, 30% by 10 years of age.³⁰ Therefore, proposed sterilization techniques for retaining the gonads and their associated hormones are highly controversial. Unfortunately, these newer alternative techniques for canine sterilization remain poorly investigated,^21,25,26,31 while at the same time are being embraced and requested by dog owners. Before veterinarians recommend alternative sterilization, outcome data to support such techniques in the short and long term are needed, which prompted us to perform this study.

In the authors’ experience, the main argument against LapOSS among veterinarians appears to be the fear of stump pyometra development, similar to that seen in ovarian remnant syndrome.^32–34 With a ligated cervix hindering the normal drainage of uterine secretions, the risk for stump pyometra is a highly relevant concern. Recently, 44% of 32 dogs with inadvertent remnant ovarian tissue showed such abnormal uterine tissue.³⁵ Among our 17 dogs with deliberately retained ovarian tissue, none displayed any uterine disease. It therefore seems reasonable to believe that complete hysterectomy indeed protected these dogs from the development of uterine disease. These authors want to stress the importance of removing all the endometrium. The uterine horns are immediately adjacent to the ovarian bursa, and the endometrium extends to overlap with the cervix.²¹ When performing a hysterectomy, it is imperative that no endometrium remnants are retained. This can be accomplished by transecting the mesosalpinx close to the ovary and ligating on the vaginal side of the cervix rather than the uterine body.²¹

Ovariectomy and OSS allow the retention of either the uterus or ovaries, respectively, and therefore a potential risk of development of disease within these organs. In agreement with previous studies, none of the ovariectomized dogs developed uterine pathology.^33,36 However, 1 dog in the LapOSS group developed disease of the ovary (ovarian cyst) that necessitated gonadectomy. Although only 1 dog in the LapOSS group developed mammary neoplasia, we were unable to make a definitive conclusion regarding the incidence of reproductive tumors in these patients given the small population size and limited follow-up time. A difference between groups may potentially be detected with a larger sample.

Overall, the incidence of ovariectomy following LapOSS was 29% (5/16), of which 2 of 16 dogs had the procedure performed due to the development of reproductive masses (mammary or the previously mentioned ovarian cyst). The recommendation to perform an ovariohysterectomy at the same time as removal of a mammary mass has been limitedly investigated, although sterilization is generally recommended. Studies have shown a decreased incidence of additional mammary masses after benign mammary mass removal and a proposed outcome benefit for specific subsets of mammary carcinomas.^37,38 One patient had an OVE 7 years following the original procedure after the development of diabetes, which subsequently reportedly resolved following surgery. Studies have revealed that removal of gonadal hormones via ovariohysterectomy in bitches can lead to the resolution of naturally occurring diabetes mellitus.^39–42 This resolution was likely attributed to the fact that progesterone-induced growth hormone secretion plays a major role in canine diabetes mellitus development during diestrus. However, the extent of growth hormone response to progesterone as well as tissue response to growth hormone and the glycemic/insulinemic responses to this phenomenon can be variable across dogs^43,44 due to many unclear variables, such as genetics, nutrition, and environment.^43–45

Studies have also been performed to further investigate the role of gonadal hormones on nonreproductive neoplasia, and a positive correlation has been found between female gonadectomy and the development of hemangiosarcoma, lymphoma, mast cell tumors, appendicular osteosarcoma, and transitional cell carcinoma of the bladder.^{11–13,46–48} In our study, 2 LapOSS patients developed nonreproductive neoplasia (mediastinal mass presumed to represent lymphoma and abdominal carcinoma), while 2 developed such in the LapOVE group (osteosarcoma and mast cell tumor). Furthermore, the LapOSS patient that developed lymphoma had undergone subsequent ovariectomy several years prior to lymphoma development. Due to the small number of dogs developing these tumors in this study, no conclusion can be drawn regarding a potential protective effect of sex hormones in nonreproductive neoplasia.

Previous literature has indicated a positive correlation between gonadectomy in certain breeds (Golden Retrievers, Labradors, and German Shepherd Dogs), with some of the more common orthopedic diseases, including CCL tear, hip dysplasia, elbow dysplasia, and osteoarthritis.^9–14 However, such correlation was not noted within our study population. Four dogs were noted to develop orthopedic conditions following their sterilization procedure: 3 dogs in the LapOSS group and 1 in the LapOVE group. There was 1 dog in each group that was diagnosed with a CCL tear, although the dog in the LapOSS group had an ovariectomy performed 3 years prior to any clinical signs or diagnosis of the CCL tear. Interestingly, the dog that developed a CCL tear in the LapOSS group had been without gonadal hormones for at least 3 years following a postoperative ovariectomy for inconvenient heat cycles within a year of the original procedure. This study was too small to draw any conclusions regarding gonadectomy and CCL tears, but our data could be useful for designing larger prospective studies to investigate the association between gonadectomy and the development of cranial cruciate disease.

Behavior changes were evaluated at 2 different periods in this study due to the potential for confounding factors such as environment, training, other diseases, or animal interactions potentially changing across the lifetime of an animal. Therefore, we specifically inquired about behaviors that changed immediately following the sterilization procedure (within 1 month) and those occurring within 1 year of the procedure. Behavioral changes following surgery were separated into aggressive and reproductive mediated to better assess for undesirable changes that may alter an owner’s decision in sterilization technique. One dog within the LapOSS group was noted to have aggressive behavior prior to the procedure that progressed following surgery and ultimately was the cause of humane euthanasia. Another dog within the LapOSS group, whose procedure was performed at 12 months of age, had aggressive behavior toward other dogs develop within 1 year following the procedure. These behaviors were able to be managed for many years with behavioral training and were not targeted toward entire or spayed females initially; however, they progressed into altercations with an intact housemate. Given the integral time and numerous factors previously shown to affect behavior throughout a dog’s lifetime, it is difficult to attribute the behavior to solely the maintenance of gonadal hormones, especially as previous work has indicated that aggression toward familiar people and dogs does not seem to be related to sex hormones.⁴⁹ However, gonadal hormones have been linked to an increase in boldness in both males and females,^15,18,50 which potentially could be responsible for the worsening of these dogs’ behaviors.

As expected, the majority of dogs in the LapOSS group exhibited reproductive behavior following their procedure including increased licking of the vulva, increased attraction by male dogs, false pregnancy, heat cycles, or vaginal discharge (clear or bloody). Interestingly, these behaviors or discharge were not seen in all LapOSS dogs as we expected. The lack of uterine tissue may explain the lack of bloody or clear vaginal discharge. The lack of additional signs associated with heat cycles, including attraction by male dogs, increased vulvar licking, or behavioral changes, could be related to poorly functional gonads. Hormone level analysis could have helped elucidate the function of the retained ovaries, but the scope of this study did not allow for such diagnostics. Additional studies investigating the functionality of retained gonads in LapOSS may be warranted to confirm that the gonads retain full function following removal of the remainder of the reproductive tract.

In both the LapOSS and LapOVE groups, 50% of the canines that had allergic skin diseases were diagnosed prior to their procedure. No difference between groups in relation to the development of skin disease was noted.

No dogs in our study went on to develop urinary incontinence. In a systematic review²⁴ published in 2012, it was noted that the evidence that gonadectomized females are at a higher risk for urinary continence was weak and that only dogs that are spayed before 3 months of age were at an increased risk. Age at sterilization did not appear to be a contributing factor for our dogs, but we were unable to fully evaluate whether dogs < 3 months of age would be affected differently as none of our dogs were < 4 months of age. The only dog that was spayed at the younger age of 4 months, illuminated the challenge of performing LapOSS in puppies, who have a very caudal location of the cervix. This makes it hard to exteriorize the cervix for ligation on the vaginal side. Therefore, we proceeded with a 6-month minimum age and > 9-month preferred age for LapOSS going forward after this dog. Ultimately, this limited the risk on the basis of previous research of urinary incontinence associated with a sterilization procedure.

The present study was affected by several important limitations. The main limitations of our study were its small size and the retrospective nature of the majority of the data. The latter entailed the possibility that not all postoperative diseases were documented in referral records or reported appropriately in owner follow-up surveys, potentially underestimating the frequency of disease. However, as many veterinarians seem to be primarily concerned by LapOSS potentially leading to stump pyometra, we consider our findings important to report despite these limitations. The lack of randomization was another limitation in this study. The decision to perform one procedure or another was elected by owners and therefore could not be assigned by the authors. Fortunately, the 2 groups were similar in age and weight, which at least minimized the effect of signalment variation on the outcome data.

Finally, our long-term follow-up would ideally extend through the entire lifespan of the dogs to capture all possible disease developments, most importantly neoplasia and orthopedic disease. The complete lack of follow-up data available for canine hysterectomy prompted us to report sooner rather than later to give the veterinary community information to base recommendations on, especially given the recommendations by large veterinary organizations for this procedure as an alternative to traditional spay. Our average follow-up time exceeded 4 years, and the average age of the dogs at follow-up was over 6 years of age, which hopefully provides useful outcome data. In comparison, dogs with inadvertent ovarian remnant syndrome have been shown to develop clinical signs at an average age of 4.5 ± 2.7 years.³⁵ Our follow-up time therefore seems appropriate.

On the basis of results of this study, we found no evidence that ovariectomized dogs or those dogs retaining ovaries and associated hormones have different risks of developing pyometra or other diseases analyzed in the short-term follow-up available. Therefore, we rejected our hypothesis and concluded that LapOSS appears to provide a safe method of sterilization, similar to that of LapOVE. Owners must be acutely aware and counseled prior to the procedure of the consequences of retained ovarian hormones and the associated heat cycles.

Supplementary Materials

Supplementary materials are posted online at the journal website: avmajournals.avma.org