FACFVMBS Ucsu63104
FACFVMBS Ucsu63104
FACFVMBS Ucsu63104
September 1988
GILBERT C. VAN CAMP, JR.
(1921-1988)
T h e A n i m a l R e p r o d u c t i o n Laboratory, C o l l e g e of Veterinary M e d i -
cine and B i o m e d i c a l Sciences wishes to express their sincere appreci-
ation to M r . Van C a m p for his generous contributions to t h e Equine
Sciences Program at C o l o r a d o State University.
EQUINE REPRODUCTIVE
ULTRASONOGRAPHY
September 1988
ACKNOWLEDGEMENTS
It w o u l d not have been possible to gather the material for this
bulletin without the outstanding contributions of our graduate stu-
dents. Their dedication to our program is sincerely appreciated. A
special thanks is extended to Dr. E. M . Carnevale for her valuable
assistance with the completion of this bulletin. Secretarial assistance
from Helen M a w h i n e y , Susan Lane and Sallie Varner was indispensa-
ble. W e are grateful to the C S U Experiment Station, Van Camp Foun-
dation, Abney Foundation, American Quarter Horse Association,
Arabian Horse Registry of America Inc, Hamilton-Thorn, Ed H.
H o n n e n and donations from Quarter Horse, Arabian and Thor-
oughbred breeders for funding for many of the studies reported
herein.
Note: All income from the sale of this bulletin, above printing costs,
will be used to aid in funding the equine teaching-research-service
program.
TABLE OF CONTENTS
LIST OF TABLES v
GLOSSARY xi
PREFACE xiii
Normal Anatomy — Uterus ductive tract is shown in Figures 1-6 to 1-9. Since the
A thorough knowledge of ultrasonographic anat- probe is generally held in a sagittal plane, images of the
omy and understanding of dynamic changes in the cervix and uterine body are longitudinally oriented
uterus is essential for ultrasonographic evaluation of with the cervix to the left of the ultrasonographic pic-
the mare's reproductive tract. The dynamic changes ture. The orientation of cranial on the right and caudal
visualized with ultrasonography mirror the ovarian on the left of the screen remains constant throughout
hormonal influences and aid in estimating reproduc- this bulletin. Uterine horns are seen in cross-section as
tive potential. the transducer is moved left or right. Depending upon
orientation of the horn at a given examination, it may
The relationship between orientation of a linear-
occasionally be necessary to manipulate the uterus in
array transducer and orientation of the mare's repro-
order to obtain a true cross-section.
Figure 1-6. Diagrammatic view of the mare's reproductive
tract.
Ultrasonographic characteristics of the uterus dur- estrus increased gradually between day -7 (2/14; day of
ing the anovulatory season, estrous cycle and preg- ovulation equals day 0), day -3 (11/16) and day -2
nancy may often be differentiated. During anestrus, (10/16), then declined between days -1 and +1 (0/12).
cross-section of uterine horns (Figure 1-10) and longi- At our laboratory, endometrial folds are graded from 0
tudinal section of uterine body (Figure 1-11) are often (no folds) to 3 (prominent endometrial folds). From a
flat and irregular and may contour closely to surround- preliminary study(5) involving 100 mare cycles, endo-
ing abdominal organs (Figure 1-10 and 1-11). In estrus, metrial folds were most prominent 1 or 2 days prior to
uterine horns are well-rounded and both horns and ovulation (Table 1-2). A change to a lower grade could
body commonly have an interdigitated pattern of be used to predict ovulation. For example, a change
alternating echogenic and nonechogenic areas (Fig- from grade 3 to 0 was concomitant with ovulation.
ures1-12and1-13; 7). The areas of decreased echogeni- Practitioners should be aware of the extent of change
city are the outer edematous portion of endometrial in height of endometrial folds between diestrus and
folds. The edema is due to effects of estrogen. This early estrus. Prominence of endometrial folds during
ultrasonograhic pattern is reminiscent of a sliced estrus (Figures 1-12 and 1-13) should not be considered
orange. Endometrial folds, generally, are visible at the pathologic. Before this observation became common
end of diestrus and become more prominent as estrus knowledge, some veterinary practitioners mistook
progresses, then decrease or disappear at approxi- prominent endometrial folds for endometritis, and
mately the time of ovulation which parallels estrogen unnecessarily treated mares. Ability to clearly observe
production. endometrial folds depends upon transducer frequency
In a recent study (3), ultrasonographic properties of and resolution of ultrasonographic equipment. On
the uterus of 16 mares were determined each day of occasion, impending abortion is suspected when, dur-
the cycle. Endometrial folds were not distinguishable ing routine scanning for pregnancy, the embryonic
during diestrus, and were prominent during estrus. vesicle is located in a uterus with prominent endomet-
The number of mares with images characteristic of rial folds.
Figure 1-12. Ultrasonographic image of a uterine horn from a Figure 1-13. Ultrasonographic image of the uterine body
mare in estrus. Presence of endometrial folds in this image is from a mare in estrus.
not an indication of endometritis. However, fluid (arrow)
within the lumen may be a sign of uterine infection.
Table 1-2. Number of Mares Displaying Endometrial However, endometrial folds (Figure 1-14) are not as
Folds in Relation to Day of Ovulation prominent as during estrus and may be associated with
Grade of increasing uterine tone.
endometrial Time from ovulation (days)
folds
-3 -2 -1 0" +2
Artifacts
0 73 27 33 64 95
1 3 17 24 24 4 Certain types or formations of tissue may cause
2 18 35 27 12 1 waves to bend (refract), bounce back and forth, or
3 6 21 16 0 0 re-echo (reverberate), or to become weakened or
aDay
entirely blocked. These distortions may be mistaken
of ovulation = day 0.
Adapted from (5). for normal or pathological structures or changes.
Fluid-filled structures such as follicles, embryonic vesi-
When a mare is in diestrus, individual endometrial cles and uterine cysts are common in the mare's
folds are less distinct, or not discernible, and the echo reproductive tract and are responsible for the most
texture is more homogeneous (Figure 1-7). When notable artifacts. An intense echogenic formation
scanning the uterine body, the uterine lumen is often beneath a fluid-filled structure is noted as an enhanced
identified by a hyperechogenic white line. This is due through-transmission artifact (Figure 1-15). This artifact
to apposition of endometrial surfaces, and probably is is common beneath images of follicles and embryonic
caused by specular reflection(3). In general, during vesicles. Sound waves passing through fluid are not as
diestrus the entire uterine portion of the reproductive attenuated as waves passing through adjacent tissue.
tract is well-circumscribed and defined. Therefore, there is a brighter echo beneath the fluid-
Ultrasonographic images of the pregnant uterus are filled structure when compared to echos of corres-
often identical to those of diestrus, with the exception ponding depth beneath adjacent tissues. Intensity of
that after day 16, endometrial folds may again appear.
Figure 1-14. Ultrasonographic image of a 20-day pregnancy Figure 1-15. Ultrasonographic image of enhanced through-
with endometrial folding. transmission artifact beneath the image of an embryonic
vesicle (arrow).
REFERENCES
1. Ginther, O.J. 1984. Ultrasonic evaluation of the 7. McKinnon, A.O., E. L. Squires and J. L. Voss. 1987.
reproductive tract of the mare: the single embryo. Ultrasonic evaluation of the mare's reproductive
J. Equine Vet. Sci. 4:75-81. tract: Part II. Comp. Cont. Educ. Pract. Vet.
2. Ginther, O.J. and R. A. Pierson. 1983. Ultrasonic 9:472-482.
evaluation of the reproductive tract of the mare;
principles, equipment and techniques. J. Equine
Vet. Sci. 3:195-201.
3. Ginther, O.J. and R. A. Pierson. 1984. Ultrasonic
anatomy and pathology of the equine uterus.
Theriogenology 21:505-515.
4. Ginther, O.J. and R. A. Pierson. 1984. Ultrasonic
evaluation of the reproductive tract of the mare:
ovaries. J. Equine Vet. Sci. 4:11-16.
5. McKinnon, A.O., E. L. Squires, E. M. Carnevale, L. A.
Harrison, D. D. Frantz, A. E. McChesney and R. K.
Shideler. 1987. Diagnostic ultrasonography of
uterine pathology in the mare. Proc. 33rd Ann.
Conv. A.A.E.P. pp.605-622.
6. McKinnon, A.O., E. L. Squires and J. L. Voss. 1987.
Ultrasonic evaluation of the mare's reproductive
tract: Part I. Comp. Cont. Educ. Pract. Vet. 9:
336-345.
Chapter 2
#
conceptus may also produce substances that are luteo-
tropic. For diagnosis of pregnancy at 10 to 12 days, a 5
or 7.5 MHz transducer is needed. However, due to
embryonic loss in early pregnancy, it is inappropriate
to discontinue the teasing program after initial exami-
nation for pregnancy. From a practical standpoint, the
first examination could be postponed until approxi-
mately 18 to 20 days post-ovulation, thus eliminating
scanning of mares that are destined to return to estrus.
One exception would be scanning of mares that have a
history of twinning or multiple ovulations. These
mares should be examined at day 12 to 15 post-
Figure 2-4. An early equine morula. ovulation to most effectively manage manual embry-
onic reduction. Frequency of subsequent scans will
depend on such factors as presence of twins, size and
quality of the vesicle and embryo, availability of the
mare and economics.
Days 17 to 22
The vesicle is spherical in shape before day 17. The
increase in size of vesicle and embryo is presented in
Tables 2-1 and 2-2, respectively. Others(6,7) have
reported similar data for the equine conceptus (Figure
2-20). The vesicle has a growth plateau between days 17
and 26, then growth resumes at a slightly slower rate
(6,7). After day 17, the vesicle is often irregular in shape
(Figures 2-10 and 2-11; 7).
Orientation is defined as rotation of the embryonic tois, which is initially recognized on day 24 (Figure
vesicle so the embryo proper is on the ventral aspect of 2-12) and, concurrent with its expansion, the contrac-
the yolk sac. On day 14, the vesicle is highly mobile. tion of the yolk sac. The interplay of growth between
Shortly after the end of the mobility phase (days 15 to these two fluid-filled structures results in the embryo
17), the dorsal uterine wall begins to enlarge and moving from the ventral (day 22; Figure 2-11) to the
encroach upon the yolk sac. Encroachment is enhanced dorsal (day 40; Figure 2-14) aspect of the vesicle. After
by increasing uterine tone. The disproportionate day 40 (Figure 2-15), the yolk sac degenerates and the
thickening and encroachment of the uterine wall on umbilical cord elongates from the dorsal pole, permit-
the vesicle, in addition to the massaging action of ute- ting the fetus to gravitate to the ventral floor where it is
rine contractions, cause the vesicle to rotate so the seen in dorsal recumbency from day 50 onwards (Fig-
thickest portion of the yolk sac (embryonic pole) ure 2-16). Apposition of yolk sac and allantois results in
assumes a ventral position (5). Hypertrophy of the ute- an ultrasonographically visible line normally oriented
rine wall is especially prominent on each side of the horizontally (Figure 2-13). On occasion, we have iden- |
dorsal midline. This probably accounts for the midline tified this junction in a vertical configuration (Figure ;
location of the apex of the triangularly-shaped vesicle, 2-17), and believe it has no deleterious effect on con- <
and the thinness of the uterine wall ventrally. The tinuing pregnancy. Twin vesicle walls, when in contact,
embryo is first detected ultrasonographically within generally appear as an ultrasonographically visible,
the vesicle at day 20 to 25, and is most commonly vertically oriented line (Figure 2-18). With knowledge
observed in the ventral position (Figure 2-11). The of the approximate stage of gestation and growth
heartbeat is commonly detected about day 22 (1), and characteristics of the conceptus, the clinician can dif-
is an important indicator of the embryo's well-being. ferentiate between the presence of an abnormally
oriented singleton (Figure 2-17) or two apposed yolk
Days 22 to 55
sacs (twins; Figure 2-18).
It is important that the ultrasonographer under-
stands and interprets clinically the growth of the allan-
Figure 2-10. Ultrasonographic image of an Figure 2-11. Ultrasonographic image of a Figure 2-12. Ultrasonographic image of a
18-day-old pregnancy. The embryonic ves- 22-day-old embryo (delineated by elec- 25-day-old embryo. Note developing allan-
icle has stopped moving and become tronic calipers). The heartbeat is commonly tois (arrow).
quite irregular. detectable about this time.
Figure 2-13. Ultrasonographic image of a Figure 2-14. Ultrasonographic image of a Figure 2-15. Ultrasonographic image of a
30-day-old pregnancy. Note developing 38-day-old fetus. The yolk sac has almost 45-day-old fetus. Note developing umbili-
allantois has pushed the embryo (arrow) regressed (arrow). cal cord (arrow).
dorsally, and regression of the yolk sac.
Figure 2-16. Ultrasonographic image of a 55-day-old fetus in Figure 2-17. Ultrasonographic image of a 26-day-old embryo
dorsal recumbency on the ventral uterine floor. with apposition of yolk sac and allantois giving the impres-
sion of a vertically oriented line.
uterine horn
NO. OF TIMES EMBRYO WAS IN
UTERINE BODY OR HORN
uterine body
10 11 12 13 14 15 16 17
DAY OF P R E G N A N C Y
mares was included for examination without the tech- by ultrasonography was low (ranging from 0 to 2.6%)
nician being aware of their breeding status. Two tech- on days 15 to 50 (Table 2-4). This was probably due to
nicians (1 and 2) experienced in both rectal palpation excessive straining by the mare or scanning too rapidly.
and ultrasonography collected data during 1982 and O n occasion, an embryo was found very close to the
1983. Two additional technicians (3 and 4), skilled in anterior cervix and it should be emphasized that this
rectal palpation but not ultrasonography, assisted in area should be scanned carefully. No false positives
collecting data during 1983. Presented in Table 2-3 is were recorded in 1982. During 1983, mares were
the number of mares diagnosed pregnant in 1982, and scanned either on the same days as in the 1982 study
the accuracy of ultrasonography for detection of (Group A) or on days 15, 25, 35 and 50 (Group B).
pregnancy. The number of mares available for scan- Results are presented in Table 2-5. The accuracy with
ning decreased from 15 to 50 days due to mares return- ultrasonography appeared similar for both years,
ing to estrus and(or) early embryonic losses. Accuracy although direct statistical comparisons probably would
ranged from 97.4 to 100%, and was similar (P > 0.05) at not have been valid. Of eight false negatives recorded
all stages of gestation (15 to 50 days). In contrast, the in 1983, seven were attributed to technicians 3 and 4.
accuracy of rectal palpation for pregnancy detection at Consequently, it appeared that knowledge of ultraso-
15 days was only 30% and 60%, respectively, for techni- nographictechnology as well as skill in rectal palpation
cians 1 and 2. By day 20, accuracy of ultrasonography were needed for accurate diagnosis. The accuracy of
and rectal palpation for pregnancy detection was sim- pregnancy detection at 15 days, reported in this study,
ilar (P > 0.05; 99.3 and 95.0%, respectively). However, using a 3.0 MHz, real-time scanner was extremely high
data were biased in that three categories were used for and comparable to previous reports (13,17).
data obtained by rectal examination (pregnant, open
and too early for diagnosis), whereas mares were
determined either pregnant or open with ultrasono-
graphy. The incidence of false negatives as determined
Table 2-3. Efficacy of Real-Time Ultrasonography
for Detection of Pregnancy (1982)
Stage of gestation (days)
15 20 25 30 35 40 50
No. pregnant
mares 167 159 158 151 143 138 132
Diagnosed
pregnant:
No. 165 158 154 150 142 138 132
% 98.7 99.3 97.4 99.3 99.3 100 100
Adapted from(14).
MANAGEMENT OF TWINS
AND EARLY EMBRYONIC DEATH
of both embryos between days 15 and 20. Thus, nonin- mon. It is possible that embryonic reduction was facili-
tervention resulted in 85.7% of the mares adjusting to a tated by competition for available nutrients. W h e n
single pregnancy by day 50. This natural, biologic, twin pregnancies fixed at opposite corpus cornual
embryo-reduction mechanism appears to be equally junctions, embryonic reduction was much less fre-
efficient, and in some cases more efficient, than treat- quent (15).
ments that have been suggested for elimination of one Manual crushing of one twin vesicle between 12 and
or more twins. More embryonic reductions occurred 30 days of pregnancy resulted in an extremely high
between days 25 to 30 and 30 to 35 (four in each time (96%) rate of single embryonic reduction (31). Fre-
period) than between days 15 to 20, 20 to 25, 35 to 40 or quent scanning is necessary to identify good separa-
40 to 50 (one in each time period). Unfortunately, tion of the vesicles (Figures 3-1 and 3-2). The smaller
number of mares at each time period was too small for vesicle should be manipulated into and crushed at the
meaningful statistical comparisons. Loss of pregnancy tip of one uterine horn. Occasionally, a distinct pop-
could be predicted, in the majority of cases, by measur- ping sensation is recognized when the vesicle is de-
ing size of the vesicle and(or) embryo between time stroyed. After day 16, the vesicles are likely to be fixed
periods. Generally, one conceptus continued to grow, at the corpus cornual junction and manual reduction
whereas the other vesicle and(or) embryo failed to may be more difficult, particularly if both are fixed on
grow or actually decreased in size between measure- the same side (Figure 3-3). Crushing of one vesicle is
ment periods. then performed in situ (31). Throughout the 1987 and
The exact mechanism for elimination of one of the 1988 breeding seasons at our laboratory, twins have
embryonic vesicles has not been elucidated, but we been successfully reduced by manipulation with the
believe its success is largely related to how the twin transducer. The ability to gently and accurately separ-
vesicles become fixed in the uterus. Twin vesicles fixed ate twins and manipulate one into the extremity of a
together at one corpus cornual junction approxi- uterine horn is greatly facilitated by monitoring with
mately 70% of the time (14). W h e n this occurred, suc- ultrasonography. The uterine horn is forced against
cessful reduction to a single pregnancy was very corn- the cranial and lateral margin of the pelvis and pres-
Figure 3-3. Two 16-day embryonic vesicles fixed together at Figure 3-4. Destruction of one twin vesicle with pressure
one corpus cornual junction. from a transducer.
sure increased by forcing the transducer down onto identify a discrepancy in size between the vesicles. If
the vesicle (Figure 3-4) until the vesicle is destroyed. there is a size difference, fetal reduction may have
Monitoring with ultrasonography and noninterven- already been initiated by the mare. When it becomes
tion until day 30 is another accepted method of treat- possible to determine fetal sex, this may become the
ment. If a mare has not adjusted to one pregnancy by criterion for fetal destruction. A 6-inch, 18-gauge
day 30, prostaglandins can be administered to termi- needle is inserted into the heart of the chosen fetus,
nate the pregnancy. Prostaglandins should be given while monitoring directly with ultrasonography, and
prior to formation of endometrial cups (days 35 to 40). potassium chloride is injected until cessation of heart-
If rebreeding is not desired or practical, induction of beat. This procedure has been performed in over 30
abortion can be delayed for as much as 70 to 80 days. mares, with the earliest pregnancy being 54 days and
When twins are diagnosed after formation of endo- the latest 168 days. However, twin pregnancies between
metrial cups (Figure 3-5), surgical intervention to 70 and 110 days of gestation are preferred (34). Approx-
remove one conceptus has been advocated (30). This imately 50% of the cases treated have resulted in
technique (30) was most successful when twins became mummification of one fetus, while the other pro-
fixed in opposite corpus cornual junctions (8 single ceeded to term. Unfortunately, the other sequel is
foals from 10 twin pregnancies) compared to when death and abortion of both fetuses within 30 days of
twins fixed together (0 foals from 8 twin pregnancies). the procedure. It was difficult to determine if these
A sophisticated technique involving cardiac punc- abortions were the natural tendency of mares to elimi-
ture of one co-twin and lethal injection of potassium nate twins (34), or a result of intrauterine invasion. No
chloride is currently being used in Kentucky (34). This adverse effects on health of mares have been recorded.
procedure is performed with either a 3 or 5 MHz sector However, this is not a stage of gestation when a large
transducer with a biopsy-needle guide attached. After percentage of abortions are expected from mares car-
tranquilization of the mare, the fetus is identified by rying twin pregnancies.
transcutaneous ultrasonography. After confirming that It is possible that methods will be developed that will
the mare has twin pregnancies, an attempt is made to allow one fetus to be carried successfully to term des-
Figure 3-5. Twin 45-day pregnancies. Note presence of two Figure 3-6. Small amount of fluid remaining at the corpus
umbilical cords (arrows). cornual junction after expulsion of the fetus.
pite twin pregnancy determined late in gestation. On tigation of embryonic losses between day 6, which is
three occasions, when premature lactation began at the first time an embryo can be recovered from the
approximately 8 mo of gestation, and supplemental uterus, and day 10, which is the first time the vesicle
progesterone was initiated and continued up to 2 wks can be consistently detected by ultrasonography. The
prior to expected delivery, a live foal was delivered incidence of EED prior to day 6 is unknown. However,
concurrent with a mummified fetus (35). Although the it was suggested that a major proportion of EED in
live foals were small, each nursed and continued to infertile mares occurred in the oviduct (4). The inci-
develop normally. W e have observed similar occur- dence of EED has been reported to be between 5 and
rences (37). However, further research is needed prior 30% of established pregnancies (8,24,36,38). In recent
to recommendation of widespread supplementation studies utilizing ultrasonography, it appeared that EED
of progesterone to mares with twin pregnancies. occurred in mares much earlier than previously re-
ported (16,46). Various causes and factors responsible
for EED in mares, apart from presence of twins, have
Early Embryonic Death
been suggested; which included nutrition (6,44,45),
Early embryonic death results in low reproductive plant estrogens and photoperiod (42), seminal treat-
performance of mares (5). Notifying the client that a ments (28), lactational stress and foal-heat breedings
valuable mare has undergone embryonic loss is a dis- (27), genital infections (8,21,27,36), chromosomal abnor-
tressing experience for a breeding manager or veteri- malities (7), hormonal deficiencies (44), anabolic ster-
narian. Clients enthusiastically support use of ultraso- oids (46), stress (43,44), failure of maternal recognition
nography for early pregnancy detection. However, not and deficiency of pregnant mare serum gonadotropin
all pregnancies continue to survive, even in normal (PMSG) production (1,2).
mares. Improvement in ultrasonographic equipment Migration of the conceptus was originally believed
has permitted investigation of early embryonic losses to be a contributing factor in EED (29). However, it is
between days 10 and 20 of gestation (46). This tech- now known to be a normal characteristic of the horse
nique, combined with embryo recovery, permits inves- conceptus (12). Lactating mares and mares bred during
Figure 3-7. Early embryonic death. Note disruption of placen- Figure 3-8. Early embryonic death. The embryo is being
tal membranes, and increased echogenicity of fetal fluid. expelled through the cervix.
foal heat have been reported to have a higher inci- tis. The high incidence of EED is not surprising, since
dence of EED than nonlactating mares (10, 26, 27). How- some of the mares had previously experienced EED or
ever, in a survey of 2,562 pregnancies in lactating and had failed to become pregnant during two cycles.
nonlactating mares, the incidence of EED was similar(3). Abnormal conditions of the uterus probably contrib-
Prior to the advent of ultrasonography, recognition uted to the higher incidence of EED. The relationship
and timing of EED was difficult. From data collected at between biopsy grade and pregnancy rates and(or)
our laboratory (46) over two breeding seasons involv- foaling rates is well-established (24,38). Breeders and
ing 356 mares diagnosed pregnant utilizing ultrasono- veterinarians should be aware of a greater discrepancy
graphy, the overall incidence of EED through day 50 between pregnancy rates and foaling rates for mares
post-ovulation was 17.3%. The majority (77.1%) of EED with a history of reproductive problems compared to
occurred prior to day 35 post-ovulation. During the normal, healthy mares.
period 15 to 35 days post-ovulation, a greater (P<0.05) Early embryonic death is diagnosed, in our labora-
incidence of EED occurred between days 15 to 20 tory, when an embryonic vesicle seen previously is not
(26.2%) and 30 to 35 (29.5%) post-ovulation compared observed on two consecutive ultrasonographic scans
to other time periods. Maternal recognition of preg- and(or) when only remnants of a vesicle are observed
nancy has been reported to occur between 14 and 16 (Figure 3-6). Ultrasonographic criteria for impending
days post-ovulation (20). In the study at our laboratory, EED are an irregular and indented vesicle, fluid in the
13.3% of mares pregnant at day 15 lost their pregnancy uterine lumen and vesicular fluid that contains echo-
by day 35 (46). It should be noted that formation of genic spots (Figure 3-7). Early embryonic death is sus-
endometrial cups occurs on approximately day 35. pected, particularly after day 30, when no fetal heart-
The highest incidence of EED (38.9%) recorded in beat is observed, there is poor definition of fetal
our laboratory has been for mares with previous his- structure, and the largest diameter of the fetal vesicle is
tories of infertility. These mares had been inseminated two standard deviations smaller than the mean estab-
for two cycles and failed to conceive, had poor uterine lished for that specific day of age, post-ovulation. Vesi-
biopsies or previously had clinical signs of endometri- cles increasing in size more slowly than normal are also
Figure 3-9. Early embryonic death. Note resorption of fetal Figure 3-10. Fetal fluid protruding into the cervix.
fluids.
characteristic of EED (16). Indications obtained by ducive to pregnancy maintenance (38). Unfortunately,
ultrasonographic scanning (16) of impending loss at recovery of embryos from infertile mares is low. Sup-
later stages include: failure of fixation, an echogenic plementation with progesterone to habitually abort-
ring within the vesicle, a mass floating in a collection of ing mares or mares with primary luteal inadequacy has
fluid and a gradual decrease in volume of placental been advocated (9). However, there is little experi-
fluid with disorganization of placental membranes mental evidence on the efficacy of this procedure(2).
(Figures 3-8 to 3-14). Perhaps the changes most likely to result in a decrease
Ultrasonographic scanning, during early pregnancy, in incidence of EED are improving management fac-
is an extremely useful management tool for pregnancy tors related to nutrition, environmental temperature, infec-
detection and determination of EED. However, if tious diseases and other stresses.
pregnancy rates are not reported until day 50, the
discrepancy between pregnancy and foaling rates
REFERENCES
decreases.
1. Allen, W.R. 1979. Maternal recognition of pregnancy
and immunological implications of tropho-
Prevention and Management of EED blast-endometrium interactions in equids. In:
There are few, if any, treatments to consistently Maternal Recognition of Pregnancy, Ciba Foun-
decrease incidence of EED, but artificial insemination dation Symposium 64, Excerpta Medica, New
can limit bacterial challenge to a mare's uterus, thus York, NY.
reducing potential losses from endometritis. In addi- 2. Allen, W.R. 1984. Is your progesterone therapy really
tion, any new information on causes and treatment of necessary? Equine Vet. J. 16:496-498.
endometritis should result in increased breeding effi- 3. Bain, A.M. 1969. Foetal losses during pregnancy in
ciency. The transfer of embryos from mares with poor the Thoroughbred mare: a record of 2,562 preg-
uterine-biopsy grades into normal recipient mares is nancies. N. Z. Vet. J. 17:155-158.
recommended to provide an environment more con-
Figure 3-11. Apparent increase in size of the amnionic cavity Figure 3-12. Decreased fetal fluids. Fetus is in the uterine
and loss of embryo outline. body instead of at the corpus cornual junction.
4. Ball, B.A., T. V. Little, R. B. Hillman and G. L. Woods. 11. Ginther, O.J. 1982. Twinning in mares: a review of
1986. Pregnancy rates at days 2 and 14 and esti- recent studies. J. Equine Vet. Sci. 2:127-135.
mated embryonic loss rates prior to day 14 in 12. Ginther, O.J. 1983. Mobility of the early equine
normal and subfertile mares. Theriogenology conceptus. Theriogenology 19:603-611.
26:611-619. 13. Ginther, O.J. 1983. Effect of reproductive status on
5. Ball, B.A. and G. L Woods. 1987. Embryonic loss and twinning and on side of ovulation and embryo
early pregnancy loss in the mare. Compend. attachment in mares. Theriogenology 20:383-395.
Cont. Educ. Pract. Vet. 9:459-471. 14. Ginther, O.J. 1987. Relationships among number of
6. Belonje, P.C. and C. H. van Niekerk. 1975. A review days between multiple ovulations, number of
of the influence of nutrition upon the oestrous embryos, and type of embryo fixation in mares.
cycle and early pregnancy in the mare. J. Reprod. Theriogenology 7:82-88.
Fert. Suppl. 23:167-169. 15. Ginther, O.J. 1988. Using a twinning tree for design-
7. Bishop, M.W.H. 1964. Paternal contribution to ing equine twin-prevention programs. J. Equine
embryonic death. J. Reprod. Fert. 7:383-396. Vet. Sci. 8:101-107.
8. Day, F.T. 1940. Clinical and experimental observa- 16. Ginther, O.J., D.R. Bergfelt, G. S. Leith and S. T.
tions on reproduction in the mare. J. Agri. Sci. Scraba. 1985. Embryonic loss in mares: incidence
30:244-261. and ultrasonic morphology. Theriogenology
9. Douglas, R.H., P. J. Burns and L. Hershman. 1985. 24:73-86.
Physiological and commercial parameters for 17. Ginther, O.J. and R. H. Douglas. 1982. The outcome
producing progeny from subfertile mares by of twin pregnancies in mares. Theriogenology
embryo transfer. Equine Vet. J. Suppl. 3:111-114. 18:237-244.
10. Fiolka, V.G., H.-J. Kuller and S. Lender. 1985. 18. Ginther, O.J., R. H. Douglas and J. R. Lawrence.
Embryonale Mortalitat beim Pferd. Mh. Vet.- 1982. Twinning in mares: a survey of veterinarians
Med. 40:835-838. and analyses of theriogenology records. Theri-
ogenology 18:333-347.
Figure 3-13. Disrupted placental membranes (arrows). Figure 3-14. Apparent increase in size of the amnionic cavity
with increased echogenicity of amnionic fluid.
19. Ginther, O.J., R. H. Douglas and G. L. Woods. 1982. 25. Kooistra, L.H. and O. J. Ginther. 1976. Termination
A biological embryo-reduction mechanism for of pseudopregnancy by administration of prosta-
the elimination of excess embryos in mares. glandin F2a and termination of early pregnancy by
Theriogenology 18:475-485. administration of prostaglandin F2a or colchicine
20. Hershman, L. and R. H. Douglas. 1979. The critical or by removal of the embryo in mares. Am. J. Vet.
period for the maternal recognition of pregnancy Res. 37:35-39.
in pony mares. J. Reprod. Fert. Suppl. 27:395-401. 26. Lieux, P. 1980. Comparative results of breeding on
21. Hoppe, R. 1968. The embryonic mortality in the first and second post-foaling heat periods. Proc.
mare. Proc. 6th Internat. Congr. Anim. Reprod. 26th Ann. Conv. A.A.E.P. pp.129-132.
A.I. 2:1573-1576. 27. Merkt, H. and A.-R. Gunzel. 1979. A survey of early
22. Hughes, J.P., G. H. Stabenfeldt and J. W . Evans. 1972. pregnancy losses in West German Thoroughbred
Clinical and endocrine aspects of the estrous mares. Equine Vet. J. 11:256-258.
cycle of the mare. Proc. 18th Ann. Conv. A.A.E.P. 28. Moberg, R. 1975. The occurrence of early embry-
pp.119-151. onic death in the mare in relation to natural ser-
23. Jeffcott, L.B. and K. E. Whitwell. 1973. Twinning as a vice and artificial insemination with fresh or
cause of foetal and neonatal loss in the Thor- deep-frozen semen. J. Reprod. Fert. Suppl.
oughbred mare. J. Comp. Path. 83:91-106. 23:537-539.
24. Kenney, R.M. 1978. Cyclic and pathologic changes 29. Moberg, R. 1976. The possible influence of site of
of the mare endometrium as detected by biopsy, pregnancy compared with site of ovulation on
with a note on early embryonic death. J.A.V.M.A. the incidence of early embryonic death in the
172:241-262. mare. Proc. 8th Internat. Congr. Anim. Reprod.
A.I. 4:610-612.
30. Pascoe, D.R. 1988. Personal communication. Oakey 39. Squires, E.L., R. H. Garcia and O. J. Ginther. 1985.
Veterinary Clinic. Oakey, Queensland, Australia. Factors affecting success of equine embryo trans-
31. Pascoe, D.R., R. R. Pascoe, J. P. Hughes, G. H. Sta- fer. Equine Vet. J. Suppl. 3:92-95.
benfeldt and H. Kindahl. 1987. Comparison of 40. Squires, E.L., A. O. McKinnon, E. M. Carnevale, R.
two techniques and three hormone therapies for Morris and T. M. Nett. 1987. Reproductive char-
management of twin conceptuses by manual acteristics of spontaneous single and double ovu-
embryonic reduction. J. Reprod. Fert. Suppl. lating mares and superovulated mares. J. Reprod.
35:701-702. (Abstr.). Fert. Suppl. 35:399-403.
32. Pascoe, R.R. 1979. A possible new treatment for 41. Squires, E.L., J. L. Voss, M . D. Villahoz and R. K.
twin pregnancy in the mare. Equine Vet. J. Shideler. 1983. Use of ultrasound in broodmare
11:64-65. reproduction. Proc. 29th Ann. Conv. A.A.E.P.
33. Pascoe, R.R. 1983. Methods for the treatment of pp.27-43.
twin pregnancy in the mare. Equine Vet. J. 15:40- 42. Swerczek, T.W. 1980. Early fetal death and infec-
42. tious placental diseases in the mare. Proc. 26th
34. Rantanen, N.W. 1988. Personal communication. Ann. Conv. A.A.E.P. pp.173-179.
P.O. Box 11849, Lexington, KY 40578-1849. 43. van Niekerk, C.H. 1965. Early embryonic resorption
35. Roberts, S.J. and G. Myhre. 1983. A review of twin- in mares. J. South. Afr. Vet. Assn. 36:61-69.
ning in horses and the possible therapeutic value 44. van Niekerk, C.H. and J. C. Morgenthal. 1982. Fetal
of supplemental progesterone to prevent abor- loss and the effect of stress on plasma progesta-
tion of equine twin fetuses the latter half of the gen levels in pregnant Thoroughbred mares. J.
gestation period. Cornell Vet. 73:257-264. Reprod. Fert. Suppl. 32:453-457.
36. Rossdale, P.D. and S.W. Ricketts. 1980. Equine Stud 45. van Niekerk, C.H. and J. S. van Heerdon. 1972.
Farm Medicine. 2nd ed. Lea & Febiger, Philadel- Nutrition and ovarian activity of mares early in
phia, PA. the breeding season. J. South Afr. Vet. Assn.
37. Shideler, R.K. 1987. Prenatal lactation associated 43:351-360.
with twin pregnancy in the mare: a case report. J. 46. Villahoz, M.D., E. L. Squires, J. L. Voss and R. K.
Equine Vet. Sci. 7:383-385. Shideler. 1985. Some observations on early embry-
38. Shideler, R.K., A. E. McChesney, J. L. Voss and E. L. onic death in mares. Theriogenology 23:915-924.
Squires. 1982. Relationship of endometrial biopsy
and other management factors on fertility of
broodmares. J. Equine Vet. Sci. 2:5-10.
Chapter 4
UTERINE PATHOLOGY
Forty-five mares were randomly assigned as they or any other data on uterine measurements. Control
foaled to one of three treatment groups: group 1 (n = mares were inseminated every other day with 500 mil-
15), controls; group 2 (n = 15), daily oral treatment with lion progressively motile spermatozoa (25) once a
0.044 mg/kg altrenogest for 8 days beginning the day > 35mm (1.38 in) follicle was detected. After withdraw-
after parturition and an injection of 10 mg of prosta- al of treatment, treated mares were inseminated using
glandin F2-alphab on day 9; group 3 (n = 15), daily oral identical criteria. Mares were inseminated during the
treatment with 0.044 mg/kg altrenogest for 15 days, first post-partum ovulatory period, regardless of
beginning the day after parturition. estrous behavioral signs.
Mares were teased daily with one or more stallions Pregnancy was diagnosed with ultrasonography 12
(25). Ultrasonographic (intrarectal 5 MHz linear array) to 14 days after ovulation and monitored every other
examination of each mare's reproductive tract was day until 20 days of gestation. Because the mares were
performed every other day beginning on day 3 and to be used as embryo donors in a companion study,
continuing until day 31 after parturition. Ultrasono- pregnancy was terminated with PGF2-alpha at day 20.
graphy was used to measure diameter of the uterus at Analysis of variance was used to determine differences
the tip and middle of the uterine horns and corpus between uterine size and mean interval to ovulation
cornual junctions, and to estimate quality (grades I to among groups. Chi-square analysis was used to evalu-
IV; Table 4-1) and quantity (ml) of uterine fluid. Follicu- ate pregnancy data.
lar growth and ovulation were monitored by ultraso- The previously gravid horn was larger than the non-
nography. The technician performing ultrasonogra- gravid horn for a mean of 21 days (range 15 to 25) after
phic examinations was unaware of group assignment, parturition. Uterine involution was most obvious at the
corpus cornual junction. When the results of three
ultrasonographic scans were similar, over a 5-day
a Regu-Mate, Hoechst-Roussel Agri-Vet,
period, the uterus was considered to be involuted. O n
Somerville, NJ 08876. the average, uterine involution was completed by day
bLutalyse, Upjohn Co., Kalamazoo, Ml 49001. 23 (range 13 to 29).
Figure 4-3. Grade III intrauterine fluid — dark grey, hypo- Figure 4-4. Grade IV intrauterine fluid — black, anechoic.
echogenic; few hyperechoic foci suspended in an anechoic
medium.
Quantity and quality of uterine fluid were not Fewer (P < 0.005) mares became pregnant when
affected by progestin treatment (Table 4-2). Number of uterine fluid was present during the first post-partum
mares with detectable uterine fluid decreased after ovulatory period (3 of 9, 33%), compared to when no
day 5 post-partum. Uterine fluid generally decreased fluid was detected (26 of 31,84%). Mares with uterine
in quantity and improved in quality between days 3 fluid during breeding did not have appreciably larger
and 15. uterine dimensions compared with those mares not
having fluid. There was no relationship between ute-
Table 4-2. Ultrasonographic Assessment of Uterine Fluid
rine size on day of ovulation and pregnancy rate. Ovu-
Quantity and Quality in 45 Mares After Parturition'
lations were delayed and pregnancy rates improved in
Post- No. of mares Mean uterine
partum with detectable fluid quality progestin-treated mares (Table 4-3). More (P < 0.05)
days uterine fluid (Grade I-IV) mares became pregnant (23/28, 82%) when they ovu-
3 34 2.1 lated after day 15, in the first post-partum ovulatory
5 37 2.8 period, than mares that ovulated before day 15 (6/12,
7 28 3.2
9 11 2.7 In this study (21), ultrasonography was useful in
11 6 3.3 detecting mares with postpartum uterine fluid. Further,
13 7 3.4
it could be used to aid in determining whether a mare
15 2 3.0
should be bred, treated, or not bred during the first
! Adapted from (21). post-partum ovulatory period. During estrus, uterine
fluid may be spermicidal and(or) an excellent medium
to support bacterial proliferation. When fluid is pres-
ent during diestrus, it may cause premature luteolysis
or EED (1). Quantity of uterine fluid during the first
post-partum ovulatory period appeared to be related
Table 4-3. Mean Interval from Parturition to Ovulation, Seven days after the last treatment, all mares remained
and Pregnancy Rates for Control and Progestin-Treated infected with either Streptococcus zooepidemicus
Mares Bred During the First Post-Partum Ovulatory Period and(or) E. coli. At this time, progesterone therapy was
Group I Group II Group III withdrawn, which resulted in recrudescence to cyclic-
Mean interval from ity. The hypothesis tested was: mares that had uterine
parturition to fluid after ovulation, determined by ultrasonography,
ovulation (days) 14.5a 18.2b 23c
would have a decreased pregnancy rate and(or) an
No. of mares
increased incidence of EED.
pregnant/
11/12 9/13 Mares were inseminated with 200 million progres-
no. mares bred 9/15
Pregnant (%) (60%)* (92%)y (69%)x,z sively motile spermatozoa every other day during
ab Significantly different (P < 0.10). estrus beginning on day 2 or 3 of estrus (25) or once a
bc Significantly different (P < 0.05). > 35 mm (1.38 in) follicle was detected. Ultrasono-
ac Significantly different (P < 0.01). graphy was used to monitor follicular changes, accumu-
x,y Pregnancy rates with different superscripts are significantly lation of intrauterine fluid and pregnancy. Mares were
different (P < 0.05). bred only during the first cycle after withdrawal of
y,z Pregnancy rates with different superscripts are significantly progesterone.
different (P < 0.10). Pregnancy rate on day 11 was 66% (37/56) compared
to stage of uterine involution, and was reduced or to 50% (28/56) on day 50. The high rate of EED between
eliminated by delaying the ovulatory period with days 11 and 50 (9/37, 24.3%) was expected, since all
progestins. mares had evidence of endometritis at the conclusion
Progestin treatment not only allowed time for elimi- of treatment with progesterone.
nation of uterine fluid before the first post-partum Fluid in the uterus during estrus did not influence
ovulation, but it also significantly delayed the first post- day-11 pregnancy rate (18/24,75%) compared to preg-
partum ovulation. Results of this study concurred with nancy rate of mares with no fluid (19/32, 59%). Preg-
those of others in which it was concluded that proges- nancy rates at day 11 were less when fluid was present 1
tin treatment delayed onset of the first post-partum or 2 days after ovulation (7/12,58.3%), but not signifi-
ovulatory period, but did not affect rate of uterine cantly different (P > 0.05) from pregnancy rates of
involution (16,17,27,29). Lower (P < 0.10) pregnancy mares with no fluid during estrus (30/44,68.2%). W h e n
rates in group 3, compared with group 2 (69 vs 92%; fluid was detected at both estrus and day 1 or 2 post-
Table 4-3) may have been related to effects of pro- ovulation, pregnancy rates were less at day 11 (4/11,
longed treatment with progestin on uterine defense 36.4%), but not significantly different (P > 0.05) from
mechanisms (9,33) in mares in which the uterus was rates in mares when fluid was not detected at all (13/21,
challenged with bacteria associated with foaling or the 61.9%).
post-partum period. Long-term progestin administra- No difference in incidence of EED was found be-
tion to normal, cycling mares has not been shown to tween days 11 and 50 for mares with fluid present
adversely affect fertility (31). during estrus (4/18, 22.2%) compared to no fluid
Since there were decreased pregnancy rates asso- observed during estrus (5/19,26.3%). A significant dif-
ciated with uterine fluid, and increased pregnancy ference (P<0.02) in rate of EED between days 11 and 50
rates as ovulation was delayed, we suggest both tech- was found between mares with fluid detected 1 or 2
niques could be used to manipulate breeding strate- days after ovulation (5/9,55.6%), and those that had no
gies and improve pregnancy rates of normal mares fluid during the same time period (2/28,14.3%).
bred during the first post-partum ovulatory period. Intrauterine fluid during estrus had no effect (P >
Effect of Intrauterine Fluid on Pregnancy Rate and 0.05) on day-50 pregnancy rate (15/24, 62.5%), com-
Early Embryonic Death. This study (20) was designed to pared to those mares that had no fluid detected (13/32,
determine the influence of intrauterine fluid on preg- 40.6%). Fluid 1 or 2 days after ovulation had a significant
nancy rate and early embryonic death. Endometritis effect (P < 0.05) on day-50 pregnancy rates (3/12,
was initiated in 60 progesterone-treated mares by 25.0%) compared to mares that had no fluid detected
either inoculation with a broth containing 5 x 107 (50 (25/44, 56.9%).
million) Streptococcus zooepidemicus organisms/ml W h e n fluid was detected during both estrus and 1 or
and(or) repeated uterine invasion during reproductive 2 days after ovulation, day-50 pregnancy rates were
evaluation. All mares received 150 mg of progesterone depressed (P < 0.05; 2/11,18.2%), compared to mares
in oil daily for 31 days. All reproductive evaluations that had fluid detected at either estrus or 1 or 2 days
were performed during this 31-day period. Mares after ovulation (14/24, 58.3%) and mares that had no
were divided into six treatment groups (n = 10 per fluid detected at any time (12/21, 57.1%).
group): Twenty-seven mares were randomly assigned to
Croup 1 — Uninoculated controls ultrasonographic examination on days 2, 4, 6, 8 and
Group 2 — Inoculated controls daily between days 10 and 20 post-ovulation. For this
Groups 3 to 6 — Inoculated plus various systemic group of mares, intrauterine fluid significantly
and intrauterine therapies. depressed pregnancy rates (P < 0.005) at day 50 (2/12,
16.7%), compared to pregnancy rates from those mares From this study (20), we concluded that only large
with no fluid detected during the same period (11/15, quantities (> 100 ml) of fluid during estrus affected
73.3%). recovery of embryos. However, it should be noted that
It was concluded from this study that: a) presence of grade I or II fluid was rare (n = 3 after ovulation, n = 32 in
intrauterine fluid during estrus in cycling mares did not estrus from both groups). For both groups combined,
affect pregnancy rates at either day 11 or 50; b) when the quality of fluid during estrus was grade I or II,
intrauterine fluid, 1 or 2 days after ovulation, did not grade III or IV, or fluid was not detected at all, embryo
affect day-11 pregnancy rates, but was associated with recovery was 13/32 (40.6%), 74/146 (50.7%) and 49/98
a significant increase in EED and reduced day-50 preg- (50%), respectively. W h e n the quality of fluid detected
nancy rates; and c) presence of intrauterine fluid dur- after ovulation (for both groups combined) was grade I
ing diestrus (days 1 to 20 post-ovulation) was associated or II, grade III or IV, or fluid was not detected, embryo
with a significant decrease in day-50 pregnancy rates. recovery was 1/3 (33.3%), 22/59 (37.3%) and 86/155
The discrepancy between effect of presence of fluid (55.5%), respectively. W h e n inflammation was asso-
in the uterus during estrus, compared to the previous ciated with decreased fertilization rate, the effect was
study, on mares after parturition may be partially expected to be most pronounced in mares with grade I
explained by the fact that mares in this study were all or II fluid. Unfortunately, the number of mares in this
diagnosed as having endometritis before cyclicity was group was quite low, which may have been the reason
initiated. Removal of exogenous progesterone resulted for not detecting a difference.
in a rapid return to estrus for most mares (mean 3.0 ± Despite the fact that critical studies on effect(s) of
2.0 days). Detection of fluid during estrus in these quality and quantity of intrauterine fluid during estrus
mares may not accurately reflect the normal popula- on fertilization rates are lacking, it appears that when
tion of mares. Another possibility is that intrauterine small quantities of intrauterine fluid (< 100 ml) are
fluid during the first post-partum ovulatory cycle may present during estrus and(or) after ovulation, embryo-
be a medium more conducive to bacterial growth after recovery rates are unaffected. It should not be assumed
breeding than fluid present during estrus in cycling that embryo-recovery rates are related to pregnancy
mares. rates in mares with uterine fluid. If intrauterine fluid
Effect of Intrauterine Fluid on Recovery of Embryos. does not affect fertilization, then its main effect may be
During the 1987 breeding season at our laboratory (20), on the developing conceptus and maintenance of the
ultrasonography was used to determine quality of ute- corpus luteum (CL). The fluid may result in embryonic
rine fluid (Table 4-1) and estimate quantity in two death or initiate release of prostaglandin, which des-
groups of normally cycling mares: group 1 — fertile troys the CL or prevents recognition of pregnancy.
experimental mares (n = 220 cycles), and group 2 — Diagnosis of Endometritis. There are numerous
infertile mares (n = 57 cycles). Data were recorded daily techniques to diagnose endometritis. However, no
during estrus and on day 1 and(or) 2 post-ovulation. technique is completely reliable. The common, cur-
Effect of quantity and quality of fluid and frequency rently accepted techniques are: a) rectal palpation, b)
and stage of cycle, at detection of fluid, on embryo vaginal-speculum examination, c) bacterial culture of
recovery was compared utilizing chi-square analysis. uterine contents, d) cytological examination of uterine
Embryo recovery was attempted either on day 6 or 7 contents, and e) endometrial biopsy.
post-ovulation from mares bred by artificial insemina- A study was conducted at our laboratory (20) to
tion with 200 to 500 million progressively motile sper- examine the efficacy of individual diagnostic tech-
matozoa (25). niques to predict endometritis. Sixty intact mares were
Group 1, normally cycling mares. Presence of fluid treated with progesterone for 31 days and 50 were
during estrus, or 1 or 2 days after ovulation, had no inoculated with a broth containing 5 x 107 (50 million)
effect on recovery of embryos (73/135, 54.1% and Streptococcus zooepidemicus organisms per milliliter.
40/79, 50.6%, respectively) compared to recovery of Reproductive evaluations were performed the day
embryos when no fluid was detected at estrus, or 1 or 2 progesterone treatment began, 13 days after progeste-
days after ovulation (44/85, 51.8% and 77/141, 54.6%, rone treatment began, and 2 and 7 days after the var-
respectively). In addition, quality of fluid had no effect ious therapeutic regimens. Thus, within the 60 experi-
on recovery of embryos. Quantity of uterine fluid mental mares, there were 240 examinations for diag-
altered recovery of embryos. Significantly fewer (P < nosis of endometritis. The following criteria were used
0.05) embryos were recovered when fluid volume during to assess degree of endometritis: a) ultrasonographic
estrus was estimated at > 100 ml (3/11, 27.3%) com- detection of intrauterine fluid, b) vaginal-speculum
pared to when fluid during estrus was estimated at < examination, c) cytological examination of uterine
100 ml (120/209, 57.4%). contents, d) culture of uterine contents, and e) acute
Group 2, infertile mares. Presence of fluid during and chronic inflammatory changes detected by endo-
estrus, or 1 or 2 days after ovulation also had no effect metrial biopsy. Each individual parameter was assigned
on recovery of embryos (20/50, 40.0% and 12/33, a score from 0 to 3 (Table 4-4). The total index score for
36.4%, respectively) compared to when no fluid was endometritis of 0 to 18, calculated from summation of
detected (2/7, 28.6% and 10/24, 41.7%, respectively). each component of the reproductive evaluation (0 to
3, for 6 diagnostic tests), was used as a standard to Table 4-5. Predictive Value of Individual Tests to
Diagnose Endometritis Based o n T w o Levels of a
determine when the mare had endometritis. To deter-
Total Index Score for Endometritis
mine the efficacy of each diagnostic test, a predictive
Total index Total index
value for each test was calculated (Table 4-5).
score score
It appeared that two conclusions could be drawn >5 >8
from this experimental model: a) a bacterial culture Score" Predictive value (%)
score of > 1 was not as accurate in predicting endo- Ultrasound > 1 98.6 94.4
metritis as other diagnostic tests, and b) ultrasonogra- > 2 99.2 95.5
phic detection of uterine fluid (score>1) was an accu- 3 100 97.1
rate indicator of endometritis. Vaginal- > 1 87.6 83.7
speculum > 2 95.5 94.1
Table 4-4. Score Criteria U s e d to Diagnose Endometritis examination 3 100 100
Cytology > 1 93.2 86.4
INDIVIDUAL COMPONENTS O F REPRODUCTIVE EVALUATION > 96.4
2 99.1
3 100 98.5
Ultrasonographic examination Speculum examination >
Culture 1 74.5 66.1
Score Criteria Score Criteria > 2 91.5 85.5
3 96.8 92.1
0 No fluid 0 Normal color, cervix
closed Endometrial > 1 98.6 92.5
1 Any grade IV fluid
biopsy — > 2 100 97.2
2 < 50 ml of grade I, II, 1 Cervix relaxed or
or III fluid open, mild reddening. acute changes 3 100 100
3 > 50 ml of grade I, II, no discharge Endometrial > 1 81.4 73.7
or III fluid 2 Moderate reddening + biopsy — > 2 97.6 95.2
discharge, cervix chronic changes 3 100 100
open
3 Severe reddening ±
1 Individual test.
discharge, cervix open
Uterine cytology
This study has several limitations: a) the model is
Uterine culture progesterone-dependent and may not accurately re-
Score Criteria
Score Criteria flect naturally-occurring endometritis. A proges-
0 Normal
1 Normal with occa- 0 No growth / light terone-dependent model may result in increased bac-
sional PMNs a growth, nonpathogen terial proliferation, decreased neutrophil numbers
2 PMNs 1-8/hpfb ± 1 Light growth /
and function, and decreased drainage of uterine con-
bacteria pathogen
3 Numerous PMNs* ± 2 Moderate growth / tents; and b) without an independent standard to
bacteria pathogen determine endometritis (i.e. used to compare the indi-
3 Heavy growth /
vidual tests against a positive or negative diagnosis of
pathogen +1 score
for extra pathogen endometritis), the predictive value in this study may
more accurately reflect each individual component's
Endometrial biopsy (acute) Endometrial biopsy (chronic)
influence on the total index score. However, we do not
Score Criteria Score Criteria believe these limitations detract from the usefulness of
0 Normal 0 Normal ultrasonography to diagnose endometritis. The appli-
1 Mild suppurative 1 Mild lymphocytic
inflammation, grade 1 inflammation, grade 1
cation of ultrasonography becomes even more appar-
2 Moderate suppurative 2 Moderate lymphocytic ent because it is noninvasive. All 10 uninoculated, con-
inflammation, grade 2 inflammation, grade 2 trol mares developed endometritis. They had a mean
3 Severe suppurative 3 Severe lymphocytic
inflammation, grade 3 inflammation, grade 3
total index score of 10.8, which was not significantly
different from inoculated mares with a mean of 11.1.
TOTAL INDEX SCORE OF ENDOMETRITIS (0-18)
Normal <3 Uterine contamination of uninoculated mares was
Mild 4-8 probably from repeated invasion to collect data. Thus,
Moderate 9-13
despite accepted hygienic techniques, invasion of the
Severe 14-18
uterus of progesterone-dominated mares resulted in
•"Polymorphonuclear leukocytes endometritis.
b High -power field
Uterine Cysts there are no reports on efficiency of these treatments,
endometrial curettage and saline lavage are frequently
Prior to ultrasonography, uterine cysts were most
applied to treat the primary problem, which would
commonly diagnosed from post-mortem examination,
appear to be lymphatic blockage.
and occasionally by rectal palpation (13). More recently
A study was conducted at our laboratory (20) to
they have been diagnosed by hysteroscopy (32) and
determine location of uterine cysts and the effect of
ultrasonography (11,19,20).
stage of estrous cycle on growth, shape, size, and gross
Cysts in the uterus are fluid-filled and apparently
and histopathological characteristics. Also, infrared
have two origins. The histological structures of uterine
radiation therapy was used to determine its effect on
cysts have been described (3,13). Endometrial cysts
the cysts. Ultrasonography was used to identify uterine
arise from endometrial glands (13), and are usually <
cysts in 12 cycling mares. Data were recorded on a total
10 mm (1.39 in) in diameter (13,17). Their incidence and
of 33 uterine cysts at 3-day intervals for 24 days. Nine-
significance is largely unknown. The second form of
teen cysts were < 10 mm (.39 in) in cross-sectional
uterine cysts are lymphatic in origin and generally are
diameter, eight cysts were between 10 (.39 in) and 20
larger than endometrial cysts. They are common in
mm (.79 in) and six cysts were> 20 mm (.79 in). Shape of
older mares (1), and have been associated with both
cysts varied from round to elongated. Single, multiple
normal and abnormal uterine biopsies (13). Size of
and apparently compartmentalized cysts were observed
uterine cysts may be indicative of origin. In a recent
(Figures 4-5 and 4-6). One cyst appeared to be located
report, large, endometrial, glandular cysts were sus-
transmurally. The uterus was divided into six segments,
pected, but were not substantiated with histological
left and right uterine horns, left and right corpus cor-
examination (32).
nual junctions and cranial and caudal uterine body.
No data has been reported on growth rate of uterine The number of cysts in each location was 5,5,4,4,4 and
cysts. Despite the occasional large cysts reported (32), it 11, respectively. There was no predisposition of a cer-
is unlikely that they grow at a similar rate as the early
embryonic vesicle (days 10 to 20). When visualized
with ultrasonography, cysts are commonly rounded,
with irregular borders, and occasionally are multiple or
compartmentalized. Movement of the early equine
conceptus (days 10 to 16), presence of specular reflec-
tion, spherical appearance and growth rate of the
embryo may aid in its differentiation from uterine
cysts.
The relationship between infertility and uterine
cysts is axiomatic. Cysts may impede movement of the
early conceptus, restricting the reported ability of the
vesicle to prevent luteolysis after day 10 (18). Later in
pregnancy, contact between the cyst wall and yolk sac
or allantois may prevent absorption of nutrients. This
may be more important when considering the report
that large uterine cysts are more commonly located at
the junction of the uterine horn and body (6,32), which
is the most common site of vesicle fixation (10). Finally,
cysts are commonly indicative of uterine disease. They
may reflect reproductive senility or be associated with
endometritis. It has been reported (1) that there is an
association between number of uterine cysts, age of
mare and endometrial biopsy.
The number of treatments proposed for uterine
cysts probably reflects inability of any individual treat-
ment to consistently be useful. Rupture of the fluid-
filled structures has been attempted via uterine-biopsy
forceps (5,13), surgery (28), fine needle aspiration (24)
and puncture via hysteroscopy (24,32). Electrocoagula-
tive removal of cysts has recently been described
(6,32). Although the number of treated mares was
extremely low (n = 6) and the mares were rigorously Figure 4-5. Multiple uterine cysts at the corpus cornual
junction.
selected for treatment, this technique may have future
application for individuals (6). Endometrial curettage
(28) and repeated lavage with warm saline (40 to 45°C =
104 to 113°F) have also been advocated (13). Although
Figure 4-6. Multiple, apparently compartmentalized, cysts in Figure 4-7. Hyperechogenic reflections slightly cranial to the
the uterine body surrounded by grade II uterine fluid. cervix, characteristic of air.
tain size cyst at any location. Cysts appeared to be mm (1.50 in), respectively, in cross-sectional diameter
distributed randomly throughout the uterus. Eight during the 24-day examination period.
mares were treated with infrared radiation therapyc Eighteen uterine cysts from eight mares were exam-
according to the manufacturer's recommendation. ined at necropsy. Endothelial cell lining was found in
Mares were treated during estrus and examined for a each cyst, irrespective of size, location, and amount of
24-day period. Four mares served as controls. There compartmentalization, including one cyst located
was no effect of treatment on size, location, shape or transmurally. It was concluded that all cysts detected
histology of cysts. Consequently, data on size, shape, ultrasonographically were lymphatic in origin. In addi-
location, changes associated with the estrous cycle and tion, 6 of the 8 mares (75%) had local distention of
histological examinations have been combined. lymphatic vessels. Five of eight mares (15/18, 83% of
There was no effect of stage of estrous cycle on size cysts examined at necropsy) had evidence of lympho-
of cysts. Mares in estrus had a mean cyst diameter of cytic infiltration, indicative of chronic endometritis.
12.0 mm (.47 in) compared to 11.9 mm (.47 in) for mares Circumscribed endometrial glands were recorded in
in diestrus. Cysts were occasionally more difficult to four mares. Small glandular cysts were found in the
discern during estrus, presumably due to presence of biopsy of one mare (post-mortem). However, these
endometrial folds. Presence or absence of intrauterine were not visualized by ultrasonography.
fluid did not affect size or ability to detect uterine cysts. W e concluded from this study that: a) uterine cysts,
One mare had three cysts that grew from 10 (.39) to 24 when detected by ultrasonography, were lymphatic in
mm (.94 in), 22 (.85) to 35 mm (1.34 in), and 25 (.98) to 38 origin; b) uterine cysts did not change rapidly in size or
shape, although they were more difficult to detect
during estrus; c) treatment with infrared radiation was
not effective; d) there was no consistent location for
uterine cysts; and e) uterine cysts were commonly
c Equine
Innovations Incorporated, P.O. Box 5375,
Loveland, Colorado 80538.
Figure 4-8. Hyperechogenic reflections due to air in the Figure 4-9. Ultrasonographic image of a hyperechogenic
uterine horn. body in the mare's uterus (arrow). The echo shadow is indica-
tive of dense tissue such as fetal bone.
associated with chronic, infiltrative, lymphocytic endo- 4-9) that was confirmed subsequently as the tip of a
metritis. uterine culturette.
Undoubtedly there are many other forms of less
commonly recognized uterine pathology such as ute-
Miscellaneous Uterine Pathology
rine neoplasia, abscesses and hematomas that will be
Recently we have identified other less commonly recognized as ultrasonography of the uterus becomes
recognized forms of uterine pathology, the most more routine.
common of which was air in the uterus. Air is recog-
nized as multiple, hyperechogenic reflections (occa-
SUMMARY
sionally a ventral reverberation artifact is present) and
it appears to be more prevalent slightly cranial to the Ultrasonography provides the opportunity to non-
cervix (Figure 4-7), although it can be present in the invasively examine the uterus to diagnose uterine
cranial body or uterine horns (Figure 4-8). Air, when pathology such as intrauterine fluid, air, uterine cysts
present < 24 hrs after artificial insemination, is consi- and debris. Intrauterine fluid detected during the first
dered normal. However, it is not expected to be found post-partum estrus has been associated with signifi-
in normal mares > 48 hrs after breeding. The observa- cantly decreased pregnancy rates. Large volumes of
tion of air in the uterus of mares that have not been intrauterine fluid detected during estrus have been
bred recently is an indication of pneumouterus and associated with decreased embryo recovery rates, and
reflects failure of the competency of the vaginal labia, fluid detected after ovulation has been associated with
vestibulovaginal sphincter and(or) cervix (21). increased EED. Ultrasonography can also be used to
On occasion, strongly echogenic areas in the uterine monitor therapeutic regimens for treatment of endo-
lumen are observed with a concomitant echo shadow, metritis. Ultrasonography was as accurate as any other
such as is seen with dense tissue like fetal bone. This diagnostic test for determining the presence of endo-
might be expected after mummification (11). We have metritis. In addition, results are available immediately
also identified a similar ultrasonographic image (Figure and it does not require invasion of the uterus.
REFERENCES 18. McDowell, K. J., D. C. Sharp, L.S. Peck and L. L.
Cheves. 1985. Effect of restricted conceptus mo-
1. Adams, G.P., J. P. Kastelic, D. R. Bergfelt and O. J.
bility on maternal recognition of pregnancy in
Ginther. 1987. Effect of uterine inflammation and
mares. Equine Vet. J. Suppl. 3:23-24.
ultrasonically-detected uterine pathology on fer-
19. McKinnon, A.O. and E. L. Squires. 1987. Equine
tility in the mare. J. Reprod. Fert. Suppl. 35:445-
ultrasound. 5th Annual Short Course Proceed-
454.
ings. Colorado State Univ. Anim. Reprod. Lab.
2. Andrews, F.N. and F. F. McKenzie. 1941. Estrus,
20. McKinnon, A.O., E. L. Squires, E. M. Carnevale, L. A.
ovulation and related phenomenon in the mare.
Harrison, D. D. Frantz, A. E. McChesney and R. K.
Res. Bull. #329. Univ. Missouri, Columbia, M O .
Shideler. 1987. Diagnostic ultrasonography of
3. Arthur, G.H. 1958. An analysis of the reproductive
uterine pathology in the mare. Proc. 33rd Ann.
function of mares based on post-mortem exami-
Conv. A.A.E.P. pp.605-622.
nation. Vet. Record 70:682-686.
21. McKinnon, A.O., E. L. Squires, L. A. Harrison, E. L.
4. Bain, A.M. 1957. Estrus and infertility of the Thor-
Blach and R. K. Shideler. 1988. Ultrasonographic
oughbred mare in Australasia. J.A.V.M.A. 131:179-
studies on the reproductive tract of postpartum
185.
mares: effect of involution and uterine fluid on
5. Baker, C.B. and R. M. Kenney. 1980. Systemic
pregnancy rates in mares with normal and delayed
approach to the diagnosis of the infertile or sub-
first postpartum ovulatory cycles. J.A.V.M.A.
fertile mare. In: Current Therapy in Theriogenol-
192:350-353.
ogy. Ed. D. A. Morrow. W. B. Saunders, Co., Phi-
22. McKinnon, A.O., E. L. Squires and J. L. Voss. 1987.
ladelphia, PA.
Ultrasonic evaluation of the mare's reproductive
6. Brook, D. and K. Frankel. 1987. Electrocoagulative
tract: Part II. Comp. Cont. Educ. Pract. Vet.
removal of endometrial cysts in the mare. J.
9:472-482.
Equine Vet. Sci. 7:77-81.
23. Merkt, H. and A. Gunzel. 1979. A survey of early
7. Bruner, D.W. 1951. Notes on genital infection in the
pregnancy losses in West German Thoroughbred
mare. Cornell Vet. 41:247-250.
mares. Equine Vet. J. 11:256-258.
8. Caslick, E.A. 1937. The sexual cycle and its relation
24. Neely, D.P. 1983. Equine Reproduction. Veterinary
to ovulation with breeding records of the Tho-
Learning Systems Co. Inc., Princeton Junction,
roughbred mare. Cornell Vet. 27:187-206.
NJ.
9. Evans,M.J.,J . M . Hamer, L. M. Gason, C. S. Graham,
25. Pickett, B.W., E. L. Squires and A. O. McKinnon.
A. C. Asbury and C.H.G. Irvine. 1986. Clearance
Procedures for collection, evaluation and utiliza-
of bacteria and non-antigenic markers following
tion of stallion semen for artificial insemination.
intra-uterine inoculation into maiden mares:
Colorado State Univ. Anim. Reprod. Lab. Bull.
effect of steroid hormone environment. Theri-
No. 03.
ogenology 26:37-50.
26. Piatt, H. 1973. Aetiological aspects of abortion in the
10. Ginther, O.J. 1983. Fixation and orientation of the
Thoroughbred mare. J. Compar. Pathol. 83:199-
early equine conceptus. Theriogenology 19:613-
205.
623.
27. Pope, A.M., D. L. Campbell and J. P. Davidson. 1979.
11. Ginther, O.J. and R. A. Pierson. 1984. Ultrasonic
Endometrial histology of post-partum mares
anatomy and pathology of the equine uterus.
treated with progesterone and a synthetic GnRH
Theriogenology 21:505-515.
(AY-24,031). J. Reprod. Fert. Suppl. 27:587-591.
12. Jennings, W.E. 1941. Some common problems in
28. Rossdale, P.D. and S. W . Ricketts. 1980. Equine Stud
horse breeding. Cornell Vet. 31:197-216.
Farm Medicine. 2nd ed. Lea & Febiger, Philadel-
13. Kenney, R.M. and V.K. Ganjam. 1975. Selected
phia, PA.
pathological changes of the mare uterus and
29. Sexton, P.E. and F.M. Bristol. 1985. Uterine involu-
ovary. J. Reprod. Fert. Suppl. 23:335-339.
tion in mares treated with progesterone and
14. Lieux, P. 1980. Comparative results of breeding on
estradiol 17-B. J.A.V.M.A. 186:252-256.
first and second post-foaling heat periods. Proc.
30. Squires, E.L., A. O. McKinnon and R. K. Shideler.
26th Ann. Conv. A.A.E.P. pp.129-132.
1988. Use of ultrasonography in reproductive
15. Loy, R.G. 1980. Characteristics of postpartum repro-
management of mares. Theriogenology 29:55-70.
duction in mares. Vet. Clin. North Am. Large
31. Squires, E.L., R. K. Shideler, J. L. Voss and S. K.
Anim. Pract. 2:345-359.
Webel. 1983. Clinical applications of progestins in
16. Loy, R.G., M.J. Evans, R. Pemstein and T.B. Taylor.
mares. Comp. Cont. Educ. Pract. Vet. 5:S16-22.
1982. Effects of injected ovarian steroids on repro-
32. Wilson, G.L. 1985. Diagnostic and therapeutic hys-
ductive patterns and performance in post-partum
teroscopy for endometrial cysts in mares. Vet.
mares. J. Reprod. Fert. Suppl. 32:199-204.
Med. 80:59-63.
17. Loy, R.G., J. P. Hughes, W.P.C. Richards and S. M.
33. Winter, A.J. 1982. Microbial immunity in the repro-
Swan. 1975. Effects of progesterone on reproduc-
ductive tract. J.A.V.M.A. 181:1069-1073.
tive function in mares after parturition. J. Reprod.
Fert. Suppl. 23:291-295.
Chapter 5
Introduction
Ultrasonography is useful for monitoring dynamic
follicular and luteal changes of equine ovaries, since it
permits rapid, visual, noninvasive access to the repro-
ductive tract. A 5 MHz transducer has greater resolu-
tion and is more suitable for evaluation of ovaries than
a 3 or 3.5 MHz transducer. Follicles as small as 2 to 3 mm
(.08 to .12 in) can be seen (5,7), and the corpus luteum
(CL) can usually be identified throughout its functional
life(16). Potential applications of ultrasonographic
examination of the ovaries include: a) estimating stage
of estrous cycle, b) assessing preovulatory follicles,
c) determining ovulation, d) examining the CL, and
e) diagnosing ovarian abnormalities and pathology.
follicular rupture (20,21). However, time sequence and In one study (17), various criteria such as percentage
follicular changes during ovulation are not well-charac- change in shape, size of follicle, echogenicity of follic-
terized. ular fluid and wall, and thickness of follicular wall were
Although stallion semen has been reported to sur- evaluated in their ability to predict time of ovulation.
vive for 5 days or longer in the mare's reproductive Size of the preovulatory follicle was as accurate as any
tract (13,15), it is more commonly accepted that a lapse method in determining ovulation time. Generally,
of > 48 hrs after breeding will result in decreased double preovulatory follicles ovulated after attaining a
numbers of viable spermatozoa and reduced fertility smaller maximum diameter than single preovulatory
(15). Use of frozen, cooled or poor quality semen may follicles.
markedly hinder life span of spermatozoa after insem- Thickening of the follicular wall occurs in most
ination. Although no critical studies have been per- preovulatory follicles prior to ovulation (Figure 5-7).
formed, the mare's oocyte probably begins to lose However, it generally occurs too early to be an adjunct
viability within 12 to 24 hrs after ovulation (15). In to predicting ovulation. Increased echogenicity of fol-
addition, semen deposited in the uterus after ovula- licular fluid is sometimes seen prior to ovulation (Fig-
tion requires time to reach the oviduct (site of fertiliza- ure 5-8), perhaps due to degeneration and subsequent
tion) and for capacitation. Breeding or insemination, shedding of granulosa cells from the follicular wall.
particularly with semen of reduced longevity, just prior This can be an indicator of impending ovulation,
to ovulation would maximize pregnancy rates and although it is neither common nor consistent enough
prevent overuse of an individual stallion. to be particularly diagnostic.
Accurate prediction of impending ovulation would In summary, the combination of softening of a large
allow for collection of mature equine oocytes for in follicle, particularly when associated with pain as
vitro fertilization or for gamete transfer from infertile determined by rectal palpation, and a substantial
mares (2,10,12). In addition, recently ovulated oocytes change in shape of the follicle, as detected with ultra-
or early cleavage embryos could be recovered from sonography, can be used to predict ovulation within a
the oviduct at specific times post-ovulation. 24-hr period for most mares.
Figure 5-6. Ultrasonographic characteristics of an irregularly- Figure 5-7. Ultrasonographic characteristics of thickening of
shaped preovulatory follicle. the follicular wall (arrow) as sometimes seen prior to ovula-
tion.
In the remaining mares, a tear or point was already ning and subsequently formed anovulatory hemor-
present in the follicular wall upon initial examination. rhagic follicles (AHF). This occurrence was possibly
In these mares, ovulation occurred within 3 to 27 min- due to season (11). It should be recognized that mares
utes. A tear or pointed appearance, observed with were chosen for this experiment based on follicles
ultrasonography just prior to ovulation, is likely due to with similar preovulatory characteristics. It is possible
breakdown of ovarian stroma and protrusion of the that all mares will not exhibit the same ultrasonogra-
follicle toward the ovulation fossa. These observations phic characteristics prior to ovulation. In addition,
probably parallel deterioration of the follicular wall, abnormal ovulations such as AHFs and luteinized,
stigma formation and protrusion of the basement unruptured follicles (9) may initially display a similar
membrane just prior to ovulation, as observed in other sequence of events as normal, preovulatory follicles
species (4). Witherspoon and Talbot (21) observed without ovulation. This was characteristic of two AHFs
deep layers of ovarian connective tissue that had rup- in the present study.
tured. Also, there were surface cuboidal and columnar Follicular fluid was not visualized in the oviduct or
cells in the area of the ovulation fossa after ovulation. uterine horn immediately after ovulation. Withers-
When the ovary was exposed by laparotomy, just prior poon and Talbot (21) demonstrated, by placing dye
to ovulation, tearing of a few strands of tissue on the within the preovulatory follicle, that follicular fluid
external surface and protrusion of the follicle into the entered the oviduct. However, due to a small oviductal
fossa was occasionally visualized (20,21). lumen, large volumes of follicular fluid and a rapid rate
Ovulation, defined as a rapid decrease in follicular of follicular collapse, it is unlikely that all fluid enters
size (Figures 5-1la to 5-1lj), occurred in an average of the oviduct.
42 sec (range 5 to 90). Little or no follicular fluid Failure to observe fluid in the oviducts during or
remained in the follicle after ovulation. Two mares after ovulation is not surprising. Firstly, the oviduct at
failed to ovulate within 12 hrs after initiation of scan- thefimbrial end is small (5 to 10 mm; .20 to .39 in) and a
Figure 5-10. A follicle just prior to ovulation. Note the pres- Figure 5-11a. The process of ovulation recorded by ultraso-
ence of a pronounced neck-like process of the follicular wall nography. Preovulatory follicle 50 min prior to beginning of
(arrow), and increased echogenicity of follicular fluid. ovulation.
sphincter-like structure would seem to preclude rapid It is important to note that these data may have been
entry of fluid necessary for dilation. Secondly, it is biased because only follicles with certain characteris-
assumed that fluid exit from the follicle is responsible tics were chosen for observation. It is quite possible
for passage of the oocyte and surrounding cumulus that follicles without these characteristics could have
oophorous into the oviduct. However, the fimbria has ovulated without observation.
fine cilia that may facilitate capture and transport of the
sticky, gelatinous cumulus oophorous into the ovi-
duct. W e have observed a pregnancy in a mare when
an oocyte was not recovered by aspiration from a
preovulatory follicle. In this case it appeared that des-
pite the loss of follicular fluid, the cumulus oophorous
and oocyte were captured by the fimbria, transported
Facing page:
into the oviduct and subsequently fertilized. Figure 5-11b. Preovulatory follicle 30 min prior to the begin-
Increased echogenicity of the follicular wall was vis- ning of ovulation. Note rent in dorsal follicular wall (arrow).
ualized in all follicles prior to ovulation (Figures 5-7
Figure 5-11c. Preovulatory follicle 20 min prior to beginning
and 5-8). Appearance of echogenic "spots" within the of ovulation. Note decrease in follicular size.
follicular fluid, probably due to dispersal of granulosa
cells, was noted in 7 of 13 follicles (54%; Figure 5-8). Figure 5-11d. Preovulatory follicle 15 min prior to beginning
of ovulation.
However, echogenic "spots" within follicular fluid, or
a bright, echogenic follicular border, were not consis- Figure 5-11e. The beginning of ovulation. Ovulation has
tently useful in predicting time of ovulation. A bright been defined as a rapid decrease in follicular size.
echogenic border, irregular shape and a tear in the Figures 5-11f to j. The process of ovulation over approxi-
follicular wall were predictive of imminent ovulation. mately 60 seconds.
Figure 5-11h. Figure 5-11i. Figure 5-11j.
Efficacy of Ultrasonography for ultrasonography in these circumstances allows one to
Determining Ovulation make a more accurate determination concerning ovu-
lation. In cases where ovulation is detected by ultraso-
In a study conducted in our laboratory (19), the
nography and not by palpation, it is most commonly
accuracy of rectal palpation and ultrasonography for
associated with a well-circumscribed developing struc-
detection of ovulation were compared. Thirty-four
ture, which although smaller, has palpation characteris-
normally cycling, nonlactating mares of light-horse
tics similar to that of a fluid-filled follicle. These struc-
breeds were used. Data collection began on day 2 of
tures either collapse and refill with blood to form a
estrus or when a > 35 mm follicle was detected by
corpus hemorrhagicum (Chapter 6) or apparently fail
palpation. The mares were palpated rectally and
to collapse completely.
scanned every 12 hours. Rectal palpation was per-
formed by one technician and ultrasonographic scan-
ning by a second, each unaware of the diagnosis made REFERENCES
by the other. A 3 MHz, linear array, real-time ultraso- 1. Bergin, W.C. and W. D. Shipley. 1968. Genital
nographic scanner was used in the study. Mares that health in the mare: observations concerning the
ovulated, based on palpation, but which had a second ovulation fossa. VM/SAC 63:362-365.
ovulatory-sized follicle, continued to be palpated until 2. Carnevale, E.M., A. O. McKinnon and E. L. Squires.
a second ovulation occurred or the mare returned to 1988. Effect of preovulatory follicular fluid aspira-
estrus. The length and width of each ovary as well as tion upon luteal function in the mare. Therioge-
diameter of each follicle was determined utilizing both nology 29:231. (Abstr.).
methods. The technician utilizing rectal palpation 3. Carnevale, E.M., A. O. McKinnon, E. L. Squires and
defined ovulation as absence of the follicle and a soft, J. L. Voss. 1988. Ultrasonographic characteristics
sometimes painful indentation. Ovulation based on of the preovulatory follicle preceding and during
ultrasonography was defined as a change in the echo- ovulation in mares. J. Equine Vet. Sci. (Accepted).
genic pattern characterized by disappearance of the 4. Espey, L.L. 1978. Ovulation. In: The Vertebrate
large, black, fluid-filled, nonechogenic structure and Ovary: Comparative Biology and Evolution. Ed.
the presence of an echogenic area. For this study, the R.E. Jones. Plenum Press, New York, NY.
results of ultrasonographic examinations were assumed 5. Ginther, O.J. and R. A. Pierson. 1983. Ultrasonic
to be an accurate indication of ovulation. evaluation of the reproductive tract of the mare;
In 24 of 34 mares, ovulation was detected at the same principles, equipment, and techniques. J. Equine
time by both methods. Three ovulations were detected Vet. Sci. 3:195-201.
by ultrasonography 12 hrs prior to detection by palpa- 6. Ginther, O.J. and R. A. Pierson. 1984. Ultrasonic
tion and five ovulations were detected with ultrasono- anatomy of equine ovaries. Theriogenology
graphy 12 hrs after detection by palpation. Thus, within 21:471-483.
±12 hrs, ovulation was detected by both methods in 32 7. Ginther, O.J. and R. A. Pierson. 1984. Ultrasonic
of 34 mares. In one mare, detection of ovulation by evaluation of the reproductive tract of the mare:
ultrasonography occurred 36 hrs prior to detection by ovaries. J. Equine Vet. Sci. 4:11-16.
palpation. In addition, ovulation was detected by ultra- 8. Hill, L.M., R. Breckle and C. B. Coulam. 1982.
sonography in one mare, but the technician failed to Assessment of human follicular development by
detect ovulation in this same mare by palpation. ultrasound. Mayo Clin. Proc. 57:176-180.
Although confirmation of ovulation by progesterone 9. McKinnon, A.O. and E. L. Squires. 1987. Ovarian
analysis was not performed in this mare, estrus ceased abnormalities. In: Equine Ultrasound. 5th Annual
approximately 2 days after the ultrasonographic detec- Short Course Proceedings. Colorado State Univ.
tion of ovulation. Anim. Reprod. Lab.
Double ovulations were detected by ultrasonography 10. McKinnon, A.O., E. L. Squires, E. M. Carnevale and
in 3 of 34 mares. These ovulations occurred on the G. E. Seidel, Jr. 1988. Heterogenous and xeno-
same ovary within 12 hours. These double ovulations genous fertilization of equine oocytes. Therioge-
were not diagnosed by rectal palpation. It may be that nology 29:278.
close apposition of two follicles on one ovary was 11. McKinnon, A.O., E. L. Squires and J. L. Voss. 1987.
responsible for failure of the palpator to recognize Ultrasonic evaluation of the mare's reproductive
both structures. The recognition of double ovulation is tract: Part I. Comp. Cont. Educ. Pract. Vet. 9:336-
important to prevent twin pregnancies, and in an 345.
embryo transfer program for correct scheduling of 12. McKinnon, A.O., M. B. Wheeler, E. M. Carnevale
recipient mares. and E. L. Squires. 1987. Oocyte transfer in the
Although ultrasonographic scanning of mares' ovar- mare: preliminary observations. J. Equine Vet. Sci.
ies should not be used as a replacement for rectal 6:306-309.
palpation, it is obvious there are occasions when this 13. Neely, D.P. 1983. Equine gestation. In: Equine
technique may be more accurate. Even with daily, Reproduction. Eds. D.P. Neely, I. K. M. Liu and R.
manual ovarian examination, there are times when B. Hillman. Veterinary Learning Systems Co. Inc.,
ovulation(s) cannot be accurately determined. Use of Princeton Junction, NJ.
14. Parker, W.A. 1971. Sequential changes of the ovu- 17. Pierson, R.A. and O. J. Ginther. 1985. Ultrasonic
lating follicle in the estrous mare as determined evaluation of the preovulatory follicle in the
by rectal palpation. Proc. Ann. Conf., College of mare. Theriogenology 24:359-368.
Veterinary Medicine and Biomedical Sciences, 18. Prickett, M.E. 1966. Pathology of the equine ovary.
Colorado State University, Fort Collins, CO. Proc. 12th Ann. Conv. A.A.E.P. pp. 145-154.
15. Pickett, B.W., E. L. Squires and A. O. McKinnon. 19. Squires, E.L., J. L. Voss, M. D. Villahoz and R. K.
Procedures for collection, evaluation and utiliza- Shideler. 1984. Use of ultrasound in broodmare
tion of stallion semen for artificial insemination. reproduction. Proc. Equine Ultrasound Short
Colorado State Univ. Anim. Reprod. Lab. Bull. Course, Colorado State Univ. Anim. Reprod. Lab.
No. 03. 20. Witherspoon, D.M. 1975. The site of ovulation in
16. Pierson, R.A. and O. J. Ginther. 1985. Ultrasonic the mare. J. Reprod. Fert. Suppl. 23:329-330.
evaluation of the corpus luteum of the mare. 21. Witherspoon, D.M. and R. B. Talbot. 1970. Ovula-
Theriogenology 23:795-806. tion site in the mare. J.A.V.M.A. 157:1452-1459.
50
Chapter 6
The corpus luteum (CL) is present during two-thirds required for accumulation of fluid and formation of
of the mare's estrous cycle, and for the first 6 mo of central clots (nonechogenic areas) was studied with
pregnancy (4). Progesterone, a primary hormonal pro- ultrasonography. Examinations were conducted at 15-
duct from the CL, has a multitude of functions, includ- min intervals for the first 2 hrs after ovulation, again at 8
ing initiation and maintenance of pregnancy. There- hrs and thereafter at 12-hr intervals for 5 days. In 2 of 10
fore, methods to evaluate the CL are extremely impor- mares, a nonechogenic area did not develop within
tant. Because of the position of the CL within the ovary, the luteal gland, and in one mare only a small central
palpation per rectum is of little value for identification area (0.5 cm2; 1.20 in2) was detected at 20 and 32 hrs,
and evaluation. However, ultrasonography has been and not thereafter. In five mares, a nonechogenic cen-
shown to be an effective and accurate means of identi- tral area developed within the luteal gland after the
fying this structure. Some of the reasons for ultrasono- expulsion of follicular fluid. The size of the nonecho-
graphic evaluation of corpora lutea are: genic area varied from 0.5 to 11.6 cm2 (.20 to 4.57 in2).
a) detect ovulation; For those mares with central nonechogenic areas,
b) evaluate CL formation; echogenic lines within the central area were detected.
c) determine size and characteristics of the CL; These were attributed to clotting and fibrinization of
d) determine if failure of a mare to display estrus is the contents. From the results of data collected in our
due to prolonged maintenance of a CL or absence laboratory (2), it appeared that, when CL evaluations
of a CL and follicular activity; were made on days 5 to 7 post-ovulation, the number
e) distinguish between anovulatory hemorrhagic of centrally nonechogenic CLs was lower (9.2%; n = 192
follicles, luteinized unruptured follicles or CLs; cycles) than that reported previously (3). In addition,
and the incidence of at least one centrally nonechogenic
f) determine if a mare has ovulated more than one CL increased with double ovulations (36%; n = 23 dou-
follicle. ble ovulations). However, from the results of more
After rupture of the follicle, a corpus hemorrhagi- recent data, a higher percentage of centrally non-
cum is formed as a transient phenomenon in the echogenic CLs was observed. This is dependent, to
development of the CL in the mare (1). However, it was some extent, upon days post-ovulation (Table 6-1).
demonstrated in a recent ultrasonographic study (3) Comparative studies, on duration of diestrus, concen-
that the equine luteal gland may involve two ultrasoni- trations of progesterone and fertility data should be
cally distinct luteal morphologies. Both types of luteal conducted to determine categorically if both morpho-
structures are uniformly echogenic on day 1. One logic types of CLs are normal.
type, classified as uniformly echogenic, is seen in Ginther (1) reported on the accuracy of detecting a
approximately 50% of the CLs and the percent of CL with ultrasonography. Location of the CL was estab-
echogenicity remains constant for the duration of dies- lished by daily palpation per rectum. Ultrasonographic
trus (Figures 6-1 to 6-4). The other, classified as cen- examinations were done by another technician
trally nonechogenic (corpus hemorrhagicum) devel- unaware of the site of ovulation. The ultrasonography
ops a nonechogenic center on day 0 or day 1 (Figures recorded location of the CL, or indicated that one was
6-5 to 6-7). The percentage of CLs considered echo- not found or that there was uncertainty about identifi-
genic was lowest on day 3, and increased linearly cation. The ultrasonographer was correct in 88% of his
throughout diestrus. In a subsequent study (5), the time examinations conducted on days 0 to 14 post-ovula-
Figure 6-1. Uniformly echogenic CL on day 0 (arrows). Figure 6-2. Uniformly echogenic CL on day 7 (arrows).
tion. In the remaining 12%, the ultrasonographer or developing luteal tissue. Although the ultrasono-
recorded the locations as uncertain. In addition, in all graphic properties of the mature CL are similar to
12 mares that were in estrus, the location of the CL was ovarian stroma, a CL can be distinguished by its
recorded as uncertain. From these results it appears defined borders. Ginther (1) found that the ultrasono-
that ultrasonography can be used to visualize a CL, graphic texture of the luteal gland was characterized
even if the site of ovulation is unknown. Therefore, by an echo pattern indicative of loosely organized,
ultrasonography is an extremely valuable diagnostic well-vascularized tissue, whereas ovarian stroma gen-
tool for determining the presence or absence of the erally yielded brighter echoes in a pattern representa-
CL. tive of dense tissue. Also, the majority of CLs had a
Visualization of the CL is much easier and more distinct mushroom or gourd shape.
accurate with a 5 MHz transducer and a quality ultra- In glands classified as centrally nonechogenic, the
sound scanner than a 3 MHz machine. Presented in nonechogenic area was first visible on day 0 or 1 post-
Figures 6-1 to 6-7 is a series of ultrasonographic images ovulation. These types of luteal structures were at their
and gross characteristics of CLs at various stages of greatest echogenicity on day of ovulation (75 to 100%
development. A 5 MHz transducer was used to obtain of the gland). This probably was due to the ultrasono-
these images. The ultrasonographic image is affected graphic properties of collapsed follicular walls. The
by amount of blood within the CL. Blood is nonecho- nonechogenic area, which was the central cavity,
genic, whereas luteal cells are echogenic. Generally, enlarged over days 1 to 3 due to enlargement of the
luteinization begins on the periphery of the structure blood clot (Figure 6-5). As the blood clot was resorbed,
and migrates medially. Normally, as the CL ages, blood that portion of the structure that was echogenic
is resorbed and a uniformly echogenic, luteal structure increased throughout the remaining portion of the
develops. Fibrin-like material can separate the blood cycle (Figure 6-6). In contrast, luteal glands that were
clot into areas of dark, nonechogenic sections contain- characterized as uniformly echogenic did not change
ing red blood cells, plasma and(or) perhaps follicular (Figure 6-3) throughout the cycle, except the bright-
fluid. Lighter areas may be indicative of fibrin strands ness (gray scale) changed throughout the life of the CL.
ti<itut
Figure 6-3. Uniformly echogenic CL on day 14 (arrows). Figure 6-5. Corpus hemorrhagicum or centrally nonecho-
genic CL (arrow) on day 1 post-ovulation.
REFERENCES
1. Ginther, O.J. 1986. Ultrasonic Imaging and Repro-
ductive Events in the Mare. Equiservices, Cross
Plains, Wl.
2. McKinnon, A.O., E.L. Squires and J.L. Voss. 1987.
Figure 6-6. Corpus hemorrhagicum on day 9 post-ovulation.
Ultrasonic evaluation of the mare's reproductive
The borders of the luteal structure are designated by arrows.
tract: Part I. Comp. Cont. Educ. Pract. Vet. 9:336-
345.
3. Pierson, R.A. and O.J. Ginther. 1985. Ultrasonic eva-
luation of the corpus luteum of the mare. Theri-
ogenology 23:795-806.
4. Squires, E.L., R.H. Douglas, W.P. Steffenhagen and
O.J. Ginther. 1974. Changes during the estrous
cycle and pregnancy in mares. J. Anim. Sci.
38:330-338.
5. Townson, D.H. and O.J. Ginther. 1988. The devel-
opment of fluid-filled luteal glands in mares.
Anim. Reprod. Sci. (In press).
OVARIAN ABNORMALITIES
Perhaps AHFs develop because of insufficient stimulus tion had not occurred. Surgical removal of two ovaries
for ovulation from gonadotropic releasing hormones. from two mares confirmed that ovulation into the ovu-
After the last ovulation of the year, mares may develop lation fossa had not occurred and, from prior ultraso-
a large follicle at the expected time, but the follicle nographic examination, it appeared that an atypical
does not ovulate and the mare enters the anovulatory corpus hemorrhagicum had formed (Figure 7-5). Both
season (1). mares ceased displaying signs of estrus within one day
of the suspected ovulation. Concentrations of proges-
terone levels were not available. These structures may
Luteinized, Unruptured Follicles
have been luteinized, unruptured follicles similar to
Although anovulatory estrous periods are very com- those in women and mice and may be associated with
mon during the anovulatory season, they are rare dur- senility. Luteinization without ovulation occurred quite
ing the ovulatory season (1). An incidence of 3.1% was commonly in pregnant mares in association with for-
reported in Thoroughbreds and Quarter Horses (6), mation of secondary CLs (13).
and even these may have been misdiagnosed because
palpation was used (Chapter 6).
Prolonged Maintenance of the Corpus Luteum
Luteinized, unruptured follicles have been reported
in women (5) and mice, but not in nonpregnant mares. Rectal palpation of the CL, although possible on
This phenomenon is thought to be associated with occasion (9), is generally unrewarding. Prolonged
reproductive senility. In 1986, a study (8) was initiated maintenance of the CL, resulting in pseudopregnancy
to recover embryos from oviducts of old, infertile (12) can be differentiated from an anovulatory or anes-
mares. On some occasions, when the oviducts were trous condition with a 5 MHz transducer. The CL is first
flushed 2 days post-ovulation, no embryos or unfertil- visible on day of ovulation (day 0) as a strongly echo-
ized ova were recovered and, from close examination genic, circumscribed mass of tissue(10). The echoge-
of the ovulation fossa, it appeared that recent ovula- nicity gradually decreases throughout diestrus.
Figure 7-3. Ultrasonographic image of an anovulatory, Figure 7-5. Ultrasonographic image of a luteinized, unrup-
hemorrhagic follicle. tured follicle. Note extreme irregularity and thickening of
the follicular wall indicative of luteal tissue (arrows).
Ovarian Neoplasia
The incidence of ovarian tumors is relatively com-
mon in mares when compared to other domestic spe-
cies. The incidence of ovarian tumors in horses has
Figure 7-6. Gross characteristics of a granulosa theca cell
tumor.