History

A 14-year-old Thoroughbred mare was admitted in mid-March to the Colorado State University Equine Reproduction Laboratory for routine foaling management. On the basis of the owner's breeding records, she was at 318 days of gestation, with her projected date of parturition in early April (340 days of gestation). Pregnancy had been confirmed by the referring veterinarian.

History of the mare revealed that she had given birth to 5 foals. Her past medical history was unremarkable, but she did undergo surgery as a 3-year-old to correct an undisclosed orthopedic condition.

Physical examination conducted at the time of admittance revealed that the mare was in good body condition and apparently healthy. She was administered a 4-way vaccine^a and a killed West Nile virus vaccine.^b In addition, the mare was dewormed by administration of a commercial anthelmintic.^c

A thorough reproductive examination was performed. No abnormalities were detected, and transrectal palpation and ultrasonography revealed an active, late-term fetus. Mammary glands of the mare were judged to be within typical limits, with no signs of milk distention or secretions from the teats (ie, waxing).

On day 329 of gestation, moderate mammary gland enlargement was detected. A sample of udder secretion was obtained, and the calcium carbonate concentration was measured with a commercial quantitation test kit,^d which yielded a value of 8.5 mg/dL. A device^e used to alert clinicians about onset of parturition was subsequently sutured to the cutaneous portion of the middle of the mare's vulva.

Udder secretions were obtained and evaluated each day until the mare foaled. On day 333 of gestation, the test yielded a calcium carbonate concentration of 50 mg/dL, which suggested that parturition was imminent.

During the late evening of day 333 of gestation, the mare appeared restless. At 1:30 AM on day 334 of gestation, the chorioallantoic membrane ruptured, which indicated the onset of stage 2 of labor. Parturition proceeded uneventfully (ie, no assistance was provided), and a healthy female foal was born at 1:40 AM.

Fetal membranes were passed 95 minutes after birth of the foal. At that time, the mare appeared bright and alert and was receptive to the foal. The mare did not have any signs of stress or discomfort during the immediate postpartum period. A small sample of colostrum was collected for evaluation, which revealed a specific gravity of 1.080. The foal appeared to have good vigor and began suckling < 1 hour after birth.

Approximately 20 hours after parturition, the mare was found recumbent in the stall, with a large quantity (approx 10 L) of fresh, bright-red blood exuding from her vulva. The mare had signs of severe pain and appeared anxious and weak. Physical examination revealed tachypnea (> 40 breaths/min) and tachycardia (> 100 beats/min), and mucous membranes of the mare were pale and white with a prolonged (> 3 seconds) capillary refill time. In addition, auscultation revealed a grade III/VI systolic murmur. The PCV was 30%.

Question

What is the most likely cause of the clinical signs and physical examination findings in this mare? Please turn the page.

Answer

Hypovolemic shock attributable to severe hemorrhage resulting from rupture of a major artery supplying blood to the uterus.

Results

The mare was administered 7.5 mg of detomidine hydrochloride,^f IV, in an attempt to provide analgesia and abate anxiety. A catheter was then inserted in her left jugular vein, and 2 L of frozen-thawed plasma and 7 L of fresh whole blood were administered by use of a blood filtration set. After plasma and blood (ie, colloid) administration, 10 g of aminocaproic acid^g diluted in 1 L of physiologic saline (0.9% NaCl) solution was administered IV. The mare stood for 2 to 3 minutes but then collapsed into lateral recumbency. The mare was then administered 500 mg of flunixin meglumine,^h IV, to provide additional analgesia and anti-inflammatory effects. The mare was observed closely throughout the remainder of the night, and the owner was advised of the guarded prognosis for the mare.

The subsequent morning, the mare's condition appeared to stabilize in that she did not have overt signs of pain or anxiety and no additional fresh blood was observed exuding from the vulva. However, the mare remained weak and would only stand intermittently for short periods. Oral administration of an antimicrobialⁱ was initiated, and additional doses of aminocaproic acid and flunixin meglumine were administered.

The foal suckled intermittently when the mare was standing. Serum IgG concentrations of the foal were evaluated, which yielded a qualitative value of approximately 800 mg/dL. Severity of the mare's condition suggested that she would be unable to provide proper nutritional support, so the foal was offered a commercial supplemental milk replacer^j every 2 to 3 hours. The foal remained bright and alert, suckled when the opportunity arose, and readily drank the milk replacer.

During the 24-hour period after treatment was initiated, the mare stood with greater frequency and appeared bright, alert, and responsive. During this time, the mare intermittently ate small quantities of food and passed small amounts of feces.

Approximately 24 hours after onset of the hemorrhagic episode, the mare's PCV was 12%. Serial samples were obtained for PCV analysis, which revealed an initial nadir followed by a gradual increase in PCV (on days 3, 8, and 14 after the hemorrhagic episode, PCV values were 11%, 13%, and 21%, respectively). Other than profound anemia, no other abnormalities were detected when CBCs were performed 1 and 3 days after the hemorrhagic episode.

Fourteen days after the hemorrhagic episode, a reproductive examination was performed to identify and evaluate the source of the hemorrhage. Vaginal examination with a speculum revealed that there was no laceration or lesion in the vaginal vault or external os of the cervix; however, a substantial quantity of old blood was evident on the floor of the vagina. Per rectal palpation revealed a large mass (approx 10 cm in diameter) in the region of the left uterine horn and broad ligament. Transrectal ultrasonography allowed us to examine the mass, which had characteristics highly suggestive of an organized hematoma (Figure 1). The location and association with the left broad ligament were consistent with the hypothesis that the most likely origin was the left middle uterine artery.

View Full Size

Transrectal ultrasonographic image of a large (approx 10 cm in diameter) mass associated with the left uterine horn and broad ligament of a mare obtained 14 days after a postpartum hemorrhagic episode. The mass was highly indicative of a hematoma resulting from rupture of the left middle uterine artery. Marks on the left side and top of the figure are at intervals of 1 cm.

Citation: Journal of the American Veterinary Medical Association 229, 10; 10.2460/javma.229.10.1571

• Download Figure
• Download figure as PowerPoint slide

Videoendoscopic examination was performed 18 days after the hemorrhagic episode. The endoscope was passed through the vulva into the vagina, which contained a moderate amount of dried blood. The endoscope was then passed through the cervix to allow examination of the lumen of the uterus. Multifocal areas of dried blood were evident throughout the endometrium. As the endoscope was passed into the left uterine horn, a large (approx 2.5 cm in diameter) and small (approx 1 cm in diameter) perforation separated by a thin strand of fibrous tissue were observed in the dorsolateral aspect of the left uterine horn (Figure 2). The videoendoscope was passed through the larger perforation, and a large organized hematoma was observed (Figure 3).

View Full Size

Videoendoscopic image of the dorsal aspect of the uterine lumen of the left uterine horn of a mare obtained 18 days after a postpartum hemorrhagic episode. Notice a large perforation (approx 2.5 cm [arrow]) in the endometrium was separated by a thin strand of fibrous tissue (arrowhead) from a small perforation (approx 1 cm).

Citation: Journal of the American Veterinary Medical Association 229, 10; 10.2460/javma.229.10.1571

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Videoendoscopic image of a large hematoma associated with the serosa of the left uterine horn of the mare in Figure 2. The image was obtained after the endoscope was passed through the large (approx 2.5 cm) perforation in the endometrium of the left uterine horn. The hematoma was organized, and no active bleeding was observed during the endoscopic examination. Notice the fibrin strands (arrows).

Citation: Journal of the American Veterinary Medical Association 229, 10; 10.2460/javma.229.10.1571

• Download Figure
• Download figure as PowerPoint slide

Analysis of clinical observations from transrectal palpation, transrectal ultrasonography, and videoendoscopy suggested that the hematoma was located within the broad ligament, with penetration into the uterine lumen. Consequently, a presumptive diagnosis of periparturient hemorrhage attributable to rupture of the left middle uterine artery was made; however, it was not possible to isolate or identify the precise site of arterial rupture.

Although parturition proceeded uneventfully, trauma to the uterine wall may have predisposed the mare to uterine rupture and subsequent hemorrhage. Because the hemorrhagic episode occurred > 20 hours after parturition, we hypothesized that the artery ruptured during parturition but that bleeding was initially confined to the left broad ligament. As the hematoma enlarged, blood disseminated along the broad ligament or under the serosal layer of the uterus. We further hypothesized that as the enlarging hematoma reached a compromised portion of the uterus, it perforated the myometrium and endometrium, which eventually led to hemorrhage extruding through the vulva.

Discussion

Hemorrhage during the peripartum period can be a serious and often life-threatening event in mares immediately before, during, or immediately after parturition. Rupture of the middle uterine, utero-ovarian, and external iliac arteries can lead to severe internal hemorrhage and death.¹ Other causes of hemorrhage during the peripartum period include uterine rupture, vaginal varicose veins, vaginal hematomas, cervical lacerations, and other forms of trauma to the perineal area.

With regard to arterial rupture, it has been hypothesized¹ that the aforementioned vessels undergo agerelated degeneration as a result of increased strain on the arterial walls, which leads to weakening of the vessel and formation of an aneurismal dilatation. Stress caused by the weight of a late-term fetus or uterine contractions during parturition can further compromise the endothelial lining of the arterial wall, thereby leading to blood dissemination and eventual arterial rupture. Interestingly, the right middle uterine artery has the highest frequency of rupture relative to other vessels.^1,2 This finding has led at least 1 author³ to speculate that this vessel is predisposed to rupture because the cecum displaces the uterus toward the left abdominal wall, which results in increased tension on the right broad ligament.

Copper deficiency has been proposed as a predisposing factor that can lead to arterial rupture in periparturient mares.⁴ Copper appears to play an important role in maintenance of elasticity of vascular walls, and serum concentrations of this trace mineral increase in pregnant mares during late gestation.⁵ In 1 study,⁴ mares that had fatal hemorrhage had significantly lower serum copper concentrations than did mares that did not hemorrhage.

An age predilection exists for arterial rupture in periparturient mares.³ Mares between 15 and 22 years of age are more likely to have a fatal rupture, whereas those between 9 and 15 years of age are likely to survive after an arterial rupture. A higher proportion of mares in the older age category hemorrhage into the peritoneal cavity, which results in hemorrhagic shock and death. Younger mares are more likely to have hemorrhage confined to the broad ligament and to survive the initial hemorrhagic event. Hemorrhage confined to the broad ligament or beneath the serosal surface of the uterus leads to the development of a hematoma.⁶ Hematomas may be an incidental finding during reproductive evaluations performed early during the post-partum period. However, hematomas in the broad ligament can rupture and cause death days to weeks after the initial hemorrhagic event.

Clinical signs associated with acute or peracute peripartum hemorrhage mimic those of colic or signs of abdominal pain.⁷ Physical examination findings often consist of tachycardia, tachypnea, sweating, and pale mucous membranes with a prolonged capillary refill time. These signs are often more severe in mares that hemorrhage directly into the peritoneal cavity than in mares with hemorrhage confined to the broad ligament or between the myometrium and serosal surface of the uterus.⁵ External hemorrhage is not often evident with arterial or uterine ruptures, but it is typically evident when there are cervical and perineal lacerations, vaginal varicosities, and vaginal or vulvar hematomas.

In addition to clinical signs and physical examination findings, certain procedures are useful to aid in diagnosing and characterizing the cause of periparturient hemorrhage in mares. These diagnostic modalities include transabdominal ultrasonography, transrectal palpation and ultrasonography, abdominocentesis, PCV analysis of serial samples, and vaginal examination with a speculum.^6,7 Per rectal palpation of the uterus and related structures often allows clinicians to discern a hematoma within the broad ligament, whereas transrectal ultrasonography can be a useful adjunct for confirming a hematoma and may also prove beneficial in determining the site of hemorrhage.

Abdominocentesis can be used to indicate hemoperitoneum, which is suggestive of an arterial or uterine rupture. The PCV of peritoneal fluid can range from 15% to 50% depending on the severity of the rupture.⁶ In instances of uterine rupture, subsequent analysis of peritoneal fluid may reveal intracellular and extracellular bacteria and high concentrations of leukocytes (ie, septic suppurative peritonitis). These findings are not as likely with an arterial rupture. Abdominocentesis was considered in the mare described here, but it was not performed because of the severity of the mare's condition and the risk of subjecting the mare to added stress and stimulation.

Measurement of PCV in serial blood samples obtained from a peripheral vein is helpful. The initial value may not be substantially decreased because of the effects of splenic contraction, thereby warranting sequential measurements to best evaluate the severity of hemorrhage or blood-loss anemia.

Finally, vaginal examination with a speculum may be warranted when hemorrhage from the vulva is observed. Vaginal examination may expose the source of hemorrhage, particularly when the cause of blood loss is varicosities, trauma, or hematomas in the vagina.

Several management strategies have been used for periparturient mares with arterial rupture. One strategy consists of keeping the mare quiet and confined to a stall that provides little or no auditory or visual stimulation.⁶ Initiation of fluid administration, whether in the form of colloid (eg, blood and plasma transfusions) or crystalloid (eg, physiologic saline solution or lactated Ringer's solution) solutions, has been advocated and successfully used by some clinicians.⁷ Others clinicians⁸ believe that these procedures can lead to unwanted stimulation and, thus, untoward effects on the mare; therefore, they recommend a more conservative management approach unless the patient is severely anemic.

Other strategies to control hemorrhage have included administration of hemostatic agents, such as naloxone and aminocaproic acid. Naloxone is a pure opioid antagonist whose efficacy has not been clearly characterized in horses with periparturient arterial rupture. However, it was recommended in 1 report⁶ that clinicians administer naloxone (8 to 32 mg/450 kg [8 to 32 mg/990 lb] diluted in 500 mL of physiologic saline solution) to periparturient mares affected by an acute hemorrhagic crisis. According to that report, administration of naloxone at this dose does not likely exert a hemodynamic effect but, rather, relieves anxiety. Aminocaproic acid inhibits fibrinolysis via its inhibitory effects on plasminogen activators. Administration of aminocaproic acid (loading dose of 40 mg/kg [18.2 mg/lb] diluted in 1 L of saline solution, followed by a maintenance dose of 10 to 20 mg/kg [4.5 to 9.1 mg/lb] diluted in 1 L of saline solution, q 6 h) may abate hemorrhage by stabilizing and supporting formation of blood clots.

Conjugated estrogens and formalin have also been used in the management of acute and uncontrollable hemorrhage in horses. The actions of conjugated estrogens (25 mg/kg [11.4 mg/lb]) and formalin (10% solution; 10 mL/450 kg [10 mL/990 lb] diluted in 1 L of saline solution) are unknown, and controlled studies to evaluate the efficacy of these agents for treatment of periparturient mares with acute hemorrhage are lacking.

Surgical intervention is usually not recommended in mares with periparturient arterial rupture because it is extremely challenging to locate and ligate the source of hemorrhage. Moreover, mares with active arterial hemorrhage are poor anesthetic candidates, and induction of general anesthesia may add to the hemodynamic instability of the patient.

To the authors' knowledge, there is only 1 report³ on the future reproductive performance of mares that had peripartum hemorrhage attributable to arterial rupture. In that report, attempts were made to breed 2 mares, each of which had a severe hemorrhagic event during the peripartum period. One of the mares conceived and foaled but subsequently died as a result of rupture of a middle uterine artery. Before that mare was bred, both broad ligaments were judged to be clinically normal via transrectal palpation. The other mare failed to conceive and continued to have a palpable mass in the broad ligament. On the basis of that report,³ it can be concluded that periparturient hemorrhage may predispose a mare to a hemorrhagic event during subsequent parturitions. However, little or no research has been conducted to characterize and carefully evaluate such a predisposition.

A second implication from that report³ is the belief that periparturient hemorrhage may have an adverse effect on future reproductive performance. Although anecdotal evidence exists that subsequent reproductive efficiency is only mildly affected in mares with a hematoma located in a broad ligament, there is a lack of published data to confirm or dispel this observation.

Outcome

During the subsequent 3 weeks after parturition, the mare's condition continued to improve. The PCV was measured routinely and continued to increase throughout the course of hospitalization. On days 19 and 23 after the hemorrhagic episode, the PCV was 24% and 26%, respectively. The mare was evaluated regularly for signs of abdominal pain and hypovolemia, but no abnormalities were detected after the initial hemorrhagic episode.

The mare and foal were discharged to the owner. Instructions for care included careful monitoring of the mare for signs of colic, anxiety, and shock. Because of the severity of the hemorrhagic episode and diagnostic findings, we also recommended that the mare should not be mated during that breeding season and possibly never again.

A follow-up conversation with the owner revealed that the mare was still alive 2 years after the hemorrhagic episode. The owner stated that the mare appeared to be in good health and that no subsequent attempts had been made to breed the mare.