the scintigraphic anatomy of the equine sacroiliac joint

THE SCINTIGRAPHIC ANATOMY OF THE EQUINE SACROILIAC JOINT

CHARLOTTE ERICHSEN, DVM, MIETH BERGER, RN. PER EKSELL, DVM

The aim of this study was to describe the scintigraphic appearance and location of the sacroiliac joint (SI-joint) in dorsal and dorso 30" lateral views of the equine pelvis in a scintigram. The pelvis from 10 different horses was prepared by attaching plastic tubes along relevant margins of the bone and around the SI-joint. A radioactive solution was injected into the tubes, and scintigraphic images were acquired with a gamma camera. Five specific landmarks were identified on the left and right sides of the pelvis. There was no significant difference in the distance measurements between the left and right sides in the dorsal view. The SI-joint was roughly ovoid in shape, with its long axis angled to the midline. The most craniolateral point of the SI-joint (Y) was located at the middle (SD = 0.04) of a line drawn from the caudal margin of the tuber sacrale (TS) to the craniolateral margin of the tuber coxa (TC), which was called TS-TC. In 85% of the specimens, the long axis of the SI-joint (SILA) was located caudal to or intersecting TS-TC, and the mean angle between SILA and TS-TC was 16.4" (SD = 5.6"). The distance from the caudomedial point of the SI-joint (Z) to Y was 0.38 times (SD = 0.04) the length of TS-TC. The distance from the tuber sacrale to the medial border of the SI-joint (TS-X) was 0.15 times (SD = 0.02) the length of TS-TC. In the oblique view, the SI-joint appeared wider and was positioned more laterally compared to the dorsal view. The distance TS-Y increased to 0.61 times (SD = 0.046) TS-TC, but SILA/TS-TC was essentially unchanged. The mean angle between TS-TC and SILA increased to 20.9" (SD = 5.6"). TS-X increased to 0.31 times (SD = 0.037) TS-TC. The location of the SI-joint was very similar in all horses, which makes it possible to estimate the location of the SI-joint in a scintigram in which the tuber sacrale and the craniolateral margin of the tuber coxa are identified. Veterinary Radi- ology & Ultrasound, Vnl. 43, No. 3, 2002, p p 287-292.

Key words: scintigraphy, anatomy, equine, sacroiliac joint.

Introduction

UMBAR OR LUMBOSACRAL back pain in horses may cause L gait abnormalities and loss of performance.',2 The prob- lem is incompletely understood because this region involves many complex structures. The sacroiliac joint (SI-joint) is claimed to be one specific area causing pain. Manifestations of pain that are described to arise from the SI-joint are lameness and lack of impulsion,'.'p6 which result in poor performance in competitive horses.

The SI-joint is a synovial joint between the auricular surfaces of the wings of sacrum and ilium.7 The joint is located ventral to the iliac wing, which is covered by the gluteal musculature. The joint attaches the sacrum to the pelvis to form the pelvic ring, and forces originating from the pelvic limbs to propel the body forward are transmitted through this joint.8 The joint has a close-fitting and well-

From the Department of Large Animal Clinical Sciences (Erichsen). Norwegian School of Veterinary Science, 0033 Oslo, Norway and the Department of Clinical Radiology (Erichsen, Berger, Eksell), Swedish Uni- versity of Agricultural Sciences, 750 07 Uppsala, Sweden.

Address correspondence and reprint requests to Charlotte Erichsen, DVM, Department of Clinical Radiology, SLU, PO Box 7029, 750 07 Uppsala, Sweden.

Received June 9, 2001; accepted for publication September 26, 2001

developed joint capsule, which is surrounded by the ventral sacroiliac ligaments. The size and contour of the joint var- ies, but in most horses the shape has been described as a sock, with the convex border facing caudally and ~ e n t r a l l y . ~ There is little or no movement in this joint at the walk and trot, but a substantial degree of axial rotation has been found between the sacrum and tuber coxa at the canter.9 The movement in the SI-joint is still a subject of controversy, and the kinematics of this area are unclear. The joint is difficult to access by palpation, and intra-articular injection with local anestethics is difficult to perform. Because of this inaccessibility and anatomic features, the SI-joints are a diagnostic challenge.

Conventional radiography and linear tomography of the S1-joint has been described with the horse in dorsal recum- bency,'."' but nonspecific findings such as increased joint space, which was sometimes associated with pelvic asym- metry, were the only findings associated with "sacroiliac strain".2 Bone scintigraphy is used to detect areas of increased skeletal activity. Intense activity is always due to increased bone resorption or formation, and areas of increased radioactivity indicate bone disease. ' ' Scintigraphy has been reported to be an accurate and useful technique for identifying horses with SI-joint injury."'* To determine the

287

288 ERICHSEN ET AL. 2002

true anatomic location of abnormal findings is a prerequisite in diagnostic imaging. No anatomic studies have been per- formed to identify the anatomy of the SI-joint in scintigraphic images. The aim of this study was to describe the scintigraphic appearance of the SI-joint in dorsal and dorso 30" lateral views of the equine pelvis.

Material and Methods Ten specimens of the bony pelvis and sacrum were col-

lected (Table 1). The specimens were boiled to remove all soft tissue, and the sacrum was attached to the ilium in an anatomically correct position with use of metal screws. Plastic tubes with an internal diameter of 2.2 mm were placed along certain margins of the bone, and a solution containing 10 MBq/ml 99m-Tc was injected into the tubes (Fig. 1).

The prepared specimen was placed on a table with the pelvic floor horizontal and the tuberi ischii resting on a cushion. A gamma camera" with a low-energy general pur- pose (LEGP) collimator was used for an acquisition of 500,000 counts per image. The geometry of image acquisition was the same as when examining standing horses in a clinical situation. A straight dorsal view of the right and left SI-joints and a dorso 30" lateral view of the right SI-joint were acquired. All views included the tuber sacrale and the tuber coxa, and the images were evaluated by using a ded- icated nuclear medicine software program.? The radioactivity in the part of the tubes lying outside the structures of interest was masked and deleted. Five landmarks were rec- ognized in the image of the specimen (Fig. 2), and the measuring points were placed in the center of the tube activity at the landmark.

The line from the caudal margin of the tuber sacrale (TS) to the craniolateral margin of the tuber coxa (TC) was called the TS-TC line, and the line from the craniolateral (Y) to the caudomedial (Z) border of the SI-joint was called SILA. The most medial point at the border of the S1-joint was called point X. In both views the distances TS-TC, Y-Z, TS-Y, TS-X and the widest part of the SI-joint were measured three times,? and the mean of these distance measure-

"Picker tHermes

T A R L ~ I . Breed, Age, and Gender of the 10 Hor\es Whme Pelvises Were Examined

Number ot Specimen\ Breed Gender Age Range

_ _ .~

4 Swedish Gelding 5-1 1

1 Swedi\h Mare 16

I Swedi5h Unknown Unknown

4 Unknown Unknown Unknown

Warmblood

Warmblood

Warmblood

FIG. 1. Plastic extension tubes were taped to the specimen. Two longer tubes were placed along the margins of the iliac wings and around the tuberi sacrale. Two shorter tubes were attached along the articular borders of the SI-joints, at the ventral aspect of the iliac wing.

ments was calculated. The angle between TS-TC and SILA was measured$ three times, and the mean of these nieasure- ments was calculated (Fig. 3). In three specimens when the two lines did not meet, SILA was extrapolated to meet TS-TC to measure the correct angle.

To check if positioning of the pelvis when slightly tilted cranioventrally in the sagittal plane affected the location of SILA in the craniocaudal direction, one of the specimens was radiographed with two stiff metal wires placed to identify TS-TC and SILA. One longer wire was placed dorsally from the caudal margin of the tuber sacrale to the craniolat- era1 margin of the tuber coxa, and one shorter wire was placed along the long axis of the S1-joint at the ventral aspect of the sacrum. The plane of the SILA was located 8-10 cm lower to the TS-TC plane. The specimen was positioned as before, and radiographs were made in a straight dorsal view and caudodorsal oblique views in 5" , lo", IS", and 20" angles.

SNIH Image 1.62

FIG. 2. Scintigrwn of one of the rpecimena. The five anatomic landmarks and their measuring points are shown

V o t . 43, No. 3 SCINTICRAPHY OF THE EQUINE SACROILIAC JOINT 289

FIG. 3. The distances TS-TC, Y-Z. TS-Y, TS-X, and the width of the SI-joint were measured. The line between TS and TC was called TS-TC, and the long axis of the Sl-joint was called SILA. The angle between TS-TC and SILA was also measured.

Results

Dorsal View

There was no significant difference between the distance measurements of the right and left side of the 10 specimens; therefore, ratios were calculated with n = 20.

The craniolateral point Y of the S1-joint was found to be located at the middle of TS-TC in all specimens, and the most medial point of the SI-joint (X) was clearly separated from the tuber sacrale. The shape of the joint was ovoid with the long axis roughly 3.3 times the width of the joint (Fig. 4).

In five specimens point Y met TS-TC, and in one specimen point Z met TS-TC. In two specimens SILA did not meet TS-TC, and in the remaining 12 specimens SILA intersected TS-TC. The mean angle between SILA and TS- TC was 16.4" (SD = 5.6"). The calculated ratios of the dorsal view ( n = 20) are shown in Table 2.

Dorso 30" laterul view

Compared with the dorsal view, the S1-joint in the oblique view had a wider droplike shape, with the length roughly 2.5 times the width of the joint, and point Y was located more lateral along TS-TC toward the tuber coxa. Point X was also located more lateral in relation to the tuber sacrale (Fig. 5 ) .

In all specimens point Y was located caudal to TS-TC and SILA did not meet TS-TC. The mean angle between TS-TC and SILA was 20.9" (SD-5.6"). The calculated ratios of the dorso 30" lateral view (n = 10) are shown in Table 3.

The radiographs in four different caudodorsal oblique angles made the position of SILA in relation to TS-TC change in the craniocaudal direction, from intersecting TS- TC to lying caudal to it (Fig. 6).

Discussion There are no articles describing the scintigraphic anatomy

of the pelvis of the horse. Anatomy is routinely estimated by subjectively comparing the scintigram with a corresponding radiograph or photograph of the same region. The method used in this study to locate the S1-joint was chosen to obtain the most accurate anatomic information using the same imaging geometry as in a clinical situation. Scintigraphic reso- lution in the pelvic region of a horse is degraded by soft tissue attenuation, a large distance between bone and camera face, movement of the patient, and background activity. The scintigrams made in this preparation were not affected by most of these factors, which made it possible to accu- rately identify landmarks. By identifying certain landmarks and relating these to other structures, the anatomic location of the SI-joint could be determined in any image of the same view.

Breed, age, and gender of the specimens varied. The study was carried out under the assumption that breed variation did not exceed individual variation in respect to the shape of the ilium and sacrum. However, the shape and size of the pelvic canal is different between male and female horses, but this variation seems to affect the angulation of the shaft of the 0 s ilium and the shape of 0 s pubis and 0 s ischium, rather than the location and size of the SI-joint." By calculating the ratio between various distances the results are not dependent on the horse size, assuming the relationship between these anatomical landmarks are the same in all horses.

The location of the SI-joint in a scintigram was described on the basis of distribution of contrast medium within the SI-joint when the contrast medium was injected into the joint through the ilium in one live horse.' This procedure must be considered unreliable because of the anatomic features of the SI-joint, such as the deep location and narrow joint space. In the present study, a different approach was chosen to show the location of the SI-joint in a simulated scintigraphic image of a live horse. Specimens of the pelvis were used to give a true anatomical image, and these were positioned similarly to the position of the sacrum and ilium in the living horse. The results of the present study allow evaluation of anatomical landmarks and borders of the S1- joint with no other interference.

The use of plastic tubes to outline the SI-joint was not ideal because the stiffness of the tubes automatically smoothed the border around the joint, even though efforts were made to attach the tube firmly along the irregular articular border. The measurement of the width of the joint is, therefore, not precise, but other measurements were af-


FIG. 4. Scintigrams of the dorsal view of the right and left side of the pelvis of 10 specimens with a diagram showing the position of the long axis of the SI-joint (SILA) in relation to TS-TC.

fected very little by tube configuration. However, the method gives only an approximation of the width of the SI-joint in these two views, and to determine the more pre-

Landmark Description Ratio SD cise shape of the joint in a scintigraphic image, a different TS [o craniolaterai point oi sacruiliac joint (Y)/TS-TC O.SO 0.040 method should be chosen. SILADS-TC 0.38 0.040 It is reported that SI-joint surfaces are flattened and

spect, the dorso 30" lateral view should give an image with

TABLE 2. Calculated Ratios of the Dorsal View with n = 20

TS tO most medial point of sacroiliac joint (X)/TS-TC SILAlwidth of the sacroiiiac joint 3.31 0.40

0.15 0.020 angled approximately 30" to the dorsal plane.' In this re-

VOL. 43, No. 3 SClNTICRAPHY OF THE EQUINE SACROILIAC JOINT 29 I

the camera surface approaching a plane parallel to that of the joint surface. The results of this study show how tilting the camera head laterally alters the relative location of the SI-joint. The distance between the tuber sacrale and the medial border of the SI-joint was double the distance in the dorsal view, the craniolateral border (Y) of the joint was located slightly more lateral, and the joint was slightly wider, but the length of the joint was approximately the same.

The position of the SI-joint, described by SILA, in the craniocaudal direction in relation to TS-TC varied (Fig. 4). Some of this variability is probably anatomic, but some is due to variation in the craniocaudal tilt of the pelvis in relation to the camera. Because the tuber sacrale and the tuber coxa are located in a plane that is dorsal to the SI-joint, a relative cranial tilt of the pelvis would place the SI-joint more caudally in relation to the line TS-TC. Even though the position of the pelvis was standardized as much as possible, variation in its angle to the camera are likely to have occurred, just as when the living horse is examined. The

TABLE 3. Calculated Ratios of the Dorso 30 Lateral" Oblique View with n = 10

Landmark Description Ratio SD

TS to craniolaterd point of sacroiliac joint (Y)/TS-TC 0.61 0.046 SILA/TS-TC 0.35 0.045 TS t o most medial point of sacroiliac joint (X)/TS-TC 0.037 SILA/width of the wcroiliac ivint 2.55 0.37

0.31

FIG. 5. Scintigrams of the dorso 30" lateral view of the right side of the pelvis of 10 specimens with a diagram showing the position of the long axi5 of the SI-joint (SILA) in relation to TS-TC.

clinical relevance of the change in position of the SI-joint with small changes in camera-pelvis orientation should be minor when standardized examination procedures are used.

In a clinical situation, the most common views of the pelvis are the dorsal and dorsolateral oblique view^,'^-'^ and the region attributed to SI-joint uptake in the dorsal view is just lateral to the tuber ~ a c r a l e . ~ . ' ~ . ' ~ The SI-joint is actually located even more laterally, and a higher pixel

..............

FIG. 6. Diagram showing the positions of the two metal wires, which illustrate TS-TC and SILA, in the radiographs taken at the various caudodorsal angles.


natively, the higher pixel count could be due to other factors such as less attenuation because of muscle atrophy or motion artifacts. Smearing or a double image of the tuber sacrale due to the horse rocking from side to side during the acquisition is very likely to cause artifacts resembling increased activity lateral to the tuber sacrale. To avoid motion artifacts and increase the total pixel count, horses can be examined under general anesthesia, or motion correction of dynamic studies of standing horses can be done. It is obvi- ous that more work needs to be done to optimize the acquisition and evaluation of the equine pelvic scintigram.

Conclusion In clinical images it is possible to determine the location

of the S1-joint within certain limits. When the caudal margin of the tuber sacrale and the craniolateral border of the tuber coxa are both in the image, the position and angle of the ST-joint in relation to these landmarks in both the dorsal and

FIG. 7. A dorsal view of the equine pelvis of a live horse illustrating the TS-TC line. The oval dotted line indicates the border of the variation (k 2 SD) of the location of the SI-joint in this horse, based on the results from this study.

dorso 30” lateral view can be determined. Based on the results of this study the in location of the sl-joint was determined with two standard deviations (Fig. 7). It is only when one can be sure about the anatomic location of the SI-joint that correct evaluations, including quantitative analyses, can be done. count found in these reports just lateral to the tuber sacrale

may be in another structure than in the actual joint. If a true lesion in the bone causes the increased pixel count, the lesion might be a stress fracture in the 0s ilium or lesions located at the site of ligament or muscle insertions. Alter-

ACKNOWLEDGMENTS Charlotte Enchsen thanks Saren Johansson for his assistance in prepar-

ing the pelvises and Peter Lord for help to review the article in the final stages.

REFERENCES

I . Jeffcott LB. Symposium on back problems in the horse. 2. The diagnosis of diseases of the horse’s back. Equine Vet J. 1975:7(2):69-78.

2. Jeffcott LB. Disorders of the thoracolumbar spine of the horse-a survey of443 cases. Equine Vet J 1980;12:197-210.

3. Rooney JR, Delaney FM, Mayo JA. Sacroiliac luxation in the horse. Equine Vet J 1969;1:287-289.

4. Rooney JR. Sacroiliac arthrosis and “stifle lameness.” Mod Vet Pract 1977;58(2): 138-139.

S . Tucker RL, Schneider RK, Sondliof AH, Ragle CA, Tyler JW. Bone scintigraphy in the diagnosis of sacroiliac injury i n twelve horses. Equine Vet J 1998;30(5):390-395.

6. Jeffcott LB, Dalin G, Ekman S, Olsson SE. Sacroiliac lesions as a cause of chronic poor performance in competitive horses. Equine Vet J

7. Dalin G, Jeffcott LB. Sacroiliac joint of the horse. I . Gross mor- phology. Anat Histol Embryo1 1986:15( 1):80-94.

8. Nickel R, Schummer A, Wille K-H, Wilkens H. Knochen der Beck- engliedmal3e. In: Nickel R, Schummer A, Seiferle E, editors. Lehrbuch der Anatornie der Haustiere. 5th ed. Berlin und Hamburg: Verlag Paul Parey, 1984, p. 83-89.

9. Faber-Meinders M. Kinematics of the equine back during locomo- tion (dissertation). Utrecht University, 2000.

l985;17(2):1l1-118.

10. Jeffcott LB. Radiographic examination of the equine back case. J S Afr Vet Assoc 1975;46(1):133-136.

1 1 . Lamb CR, Koblik PD. Scintigraphic evaluation of skeletal disease and its application to the horse. Vet Radio1 1988;29( 1):16-27.

12. Attenburrow DP. The application of scintigraphy using a gamma camera in clinical equine practice. 2. Interpretation of results. Equine Vet Educ 1997;9(2): 103-109.

13. Nickel R, Schummer A, Wille K-H, Wilkens H. Verbindungen der Knochen der BeckengliedmaRe. In: Nickel R , Schummer A, Seiferle E, editors. Lehrbuch der Anatornie der Haustiere. 5th ed. Berlin und Han- burg: Verlag Paul Parey, 1984, p. 216-218.

14. Steckel RR. The role of scintigraphy in the lameness evaluation. Vet Clin N Am Equine Prac 1991;7(2):207-239.

15. Pilsworth RC, Sheperd MC, Herinckx, Holmes MA. Fracture of the wing of the ilium, adjacent to the sacroiliac joint, in Thoroughbred race- horses. Equine Vet J 1994:26(2):94-99.

16. Hornof WJ, Stover SM, Koblik PD, Arthur RM. Oblique views of the ilium and the scintigraphic appearance of stress fractures of the ilium. Equine Vet J 1996;28(5):355-358.

17. Gundel M, Schatzmann U, Ueltschi G. Ruckenprobleme beim Pferd: Ein Vergleich der bildgebenden Verfahren mit der klinischen Un- tersuchung. Pferdeheilkunde l997;6( I3):659-666.

the scintigraphic anatomy of the equine sacroiliac joint

Documents