Greater Trochanter of the Hip: Attachment of the Abductor ...

Musculoskeletal Imaging

Christian W. A. Pfirrmann, Greater Trochanter of the Hip:
MD Attachment of the Abductor
Mechanism and a Complex of
Christine B. Chung, MD Three Bursae—MR Imaging
Nicolas H. Theumann, MD and MR Bursography in
Debra J. Trudell, RA Cadavers and MR Imaging in
Donald Resnick, MD Asymptomatic Volunteers1

Index terms: PURPOSE: To evaluate trochanteric anatomy with magnetic resonance (MR) im-
Femur, MR, 449.121411, 449.12143 aging, bursography, MR bursography, and anatomic analysis.
Hip, anatomy, 449.92
Hip, MR, 449.121411, 449.12143 MATERIALS AND METHODS: T1-weighted and fat-saturated T2-weighted (trans-
verse, sagittal, coronal, and coronal oblique planes) MR imaging of the greater
Published online before print trochanter was performed in 10 cadaveric hips and 12 hips of asymptomatic
10.1148/radiol.2211001634 volunteers. Three bursae comprising the trochanteric bursa complex were injected,
and conventional radiography and MR imaging were performed. The specimens
Radiology 2001; 221:469 – 477 were sectioned for anatomic analysis, corresponding to the MR imaging planes.
Tendon attachments and bursal localization were related to the facets of the greater
Abbreviations: trochanter.
SGMeB ϭ subgluteus medius bursa
SGMiB ϭ subgluteus minimus bursa RESULTS: The bony surface of the greater trochanter consists of four facets:
TrB ϭ trochanteric bursa anterior, lateral, posterior, and superoposterior. The gluteus medius muscle attaches
to the superoposterior and lateral facets. The gluteus minimus muscle attaches to
1 From the Department of Radiology, the anterior facet. The trochanteric bursa covered the posterior facet and the lateral
Veterans Administration Medical Cen- insertion of the gluteus medius muscle. The subgluteus medius bursa was located in
ter, San Diego, Calif. Received Octo- the superior part of the lateral facet, underneath the gluteus medius tendon. The
ber 9, 2000; revision requested De- subgluteus minimus bursa lies in the area of the anterior facet, underneath the
cember 5; revision received April 27, gluteus minimus tendon, medial and cranial to its insertion, and extends medially
2001; accepted May 2. Supported by covering the distal anterior part of the hip joint capsule. The trochanteric bursa is
the Swiss National Science Founda- delineated with fat on both sides and can be seen on transverse nonenhanced
tion. Address correspondence to T1-weighted images as a fine line curving around the posterior part of the trochanter.
C.W.A.P., Department of Radiology,
University Hospital Balgrist, Forch- CONCLUSION: MR imaging and bursography provide detailed information about
strasse 340, CH-8008 Zurich, Switzer- the anatomy of tendinous attachments of the abductor muscles and the bursal
land (e-mail: christian@pfirrmann.ch). complex of the greater trochanter.
© RSNA, 2001
The hip joint, much like the glenohumeral joint, boasts one of the widest ranges of motion
Author contributions: in the human body, with its greater trochanter serving as the main attachment site for
Guarantors of integrity of entire study, strong tendons, facilitating complex movement such as postural gait (1). This complex
C.W.A.P., D.R.; study concepts, D.R.; motion is achieved with the sophisticated attachment architecture of the abductor mech-
study design, C.W.A.P., D.R.; literature anism in the trochanteric surface and its three interposed bursae (2– 4). Despite the
research, C.W.A.P.; clinical studies, importance of the integrity of the greater trochanteric structures for normal gait (5),
C.W.A.P., D.J.T.; data acquisition, reports describing the tendon insertions and the bursal anatomy are sparse. Furthermore,
C.W.A.P., C.B.C.; data analysis/inter- to our knowledge, the magnetic resonance (MR) appearance of the tendon attachments in
pretation, C.W.A.P., N.H.T.; statistical the greater trochanter and even the precise localization of the three main trochanteric
analysis, C.W.A.P.; manuscript prepa- bursae have not been described.
ration, C.W.A.P., N.H.T.; manuscript
definition of intellectual content, 469
C.W.A.P., C.B.C., D.R.; manuscript ed-
iting, C.W.A.P.; manuscript revision/
review, D.R., C.B.C.; manuscript final
version approval, D.R.

Figure 1. Schematics of the proximal part of the femur in the frontal
view (left image), lateral view (middle image), and posterior view (right
image) display (a) the four facets: the anterior facet (AF), lateral facet
(LF), posterior facet (PF), and superoposterior facet (SPF); (b) osseous
attachment sites of the gluteus medius (GMe) and gluteus minimus
(GMi) tendons; and (c) locations of the bursae: TrB, SGMeB, and SGMiB.

Although pain over the lateral aspect oblique transverse orientation. The pos- two signals acquired, field of view of 14 ϫ
of the hip has been commonly attributed terior facet is the most posterior aspect of 14 cm, and matrix of 512 ϫ 256. Subse-
to a trochanteric bursitis, the spectrum of the trochanter. It borders the lateral and quently, transverse and coronal T2-
pathologic abnormalities about the hip superoposterior facets and has a curved weighted fat-saturated fast spin-echo im-
has broadened with the identification of shape in a mediolateral direction. ages were obtained with the following
entities such as rotator cuff tears of the parameters: 4,000/120, an echo-train
hip, a term referring to a tear of the glu- Cadaveric Specimens length of eight, section thickness of 3
teus medius or minimus tendon (6,7). mm, three signals acquired, field of view
Despite similar clinical presentations, Ten fresh human hips were harvested of 14 ϫ 14 cm, and matrix of 512 ϫ 256.
treatment of tears or bursitis can be dif- from eight nonembalmed cadavers (two
ferent, emphasizing the need for accurate women, six men; mean age at death, 75 Bursography
diagnosis. years; age range at death, 67– 81 years).
The specimens consisted of a hemipelvis After acquisition of the initial MR im-
The objectives of our cadaveric and hu- and the proximal portion of the femur, ages, bursography was performed (C.W.A.P.)
man volunteer study were to (a) describe including intact soft tissues. Radiography in nine cadaveric hips. The first cadaveric
the normal MR imaging appearance of was performed on all specimens to exclude hip underwent anatomic analysis without
the attachment sites of the abductor abnormalities from previous trauma, sur- bursography. In eight cadaveric hips, a sin-
mechanism and (b) to analyze the loca- gery, or severe osteoarthritis. The speci- gle bursa (trochanteric bursa [TrB]) (n ϭ 3),
tion, extent, and appearance of the bur- mens were immediately deep-frozen at the subgluteus medius bursa (SGMeB) (n ϭ
sal complex of the greater trochanter. Ϫ40°C (Bio-Freezer; Forma Scientific, Mari- 2), and the subgluteus minimus bursa
etta, Ohio). All specimens were allowed to (SGMiB) (n ϭ 3) were injected. In one ca-
MATERIALS AND METHODS thaw for 24 hours at room temperature daveric hip, TrB and SGMeB were injected
prior to MR imaging. simultaneously. The injections were per-
Analysis of Femoral Bones formed by one of the authors (C.W.A.P.)
MR Imaging with fluoroscopic guidance, with the spec-
Femoral bones derived from 10 human imen in a prone position for the TrB and in
skeletons were visually inspected to de- MR imaging was performed with a the supine position for the SGMeB and the
termine the presence of osseous land- 1.5-T MR imager (Signa; GE Medical Sys- SGMiB. The needle trajectories for bursog-
marks of the greater trochanter that tems, Milwaukee, Wis) with two receive- raphy of the different bursae are displayed
represented sites of musculotendinous only 5-inch gradient coils centered later- in Figure 2 as dotted lines. In each case, a
attachments (1,3,4,8 –10). The following ally over the palpable greater trochanter. 20-gauge, 0.91 ϫ 8.89-mm spinal needle
anatomic observations comprise the ba- T1-weighted MR images were acquired in (Terumo Medical, Somerset, NJ) was used.
sis for the remaining methodology. Four the transverse, sagittal, coronal, and
distinct facets could be differentiated in coronal oblique (parallel to the femoral The expected position of the TrB is im-
all trochanters (Fig 1); the anterior facet is neck) planes centered at the level of the mediately posterior to the cortex of the
located on the anterolateral surface of the greater trochanter. Sequence parameters posterior facet of the greater trochanter.
trochanter. It is oval and shares a medial were 500/22 (repetition time msec/echo The needle was inserted from a postero-
border with the intertrochanteric line. time msec), section thickness of 3 mm, lateral approach, targeting the posterior
The lateral facet has an inverted triangu- facet at the level where the trochanter is
lar shape, with the most caudal portion
(the tip of the triangle) comprising the
palpable part of the trochanter at physi-
cal examination. The posterior superior
border is in contact with the superopos-
terior facet; this facet forms the most cra-
nial part of the trochanter and has an

470 ⅐ Radiology ⅐ November 2001 Pfirrmann et al

verse (n ϭ 4) planes. The tendon inser-
tion sites in the greater trochanter were
recorded by two musculoskeletal radiolo-
gists (C.W.A.P., C.B.C.) by means of con-
sensus. On each section, the bursal spaces
were slightly distended, allowing docu-
mentation of the extent and topographic
location of each bursa. The size of the
bursa was measured with a ruler.

Asymptomatic Volunteers

Twelve hips (six right, six left) in 12
asymptomatic volunteers (five men,
seven women; mean age, 35.1 years; age
range, 21–51 years) were imaged. Institu-
tional review board approval and in-
formed consent were obtained prior to
MR imaging. Only hips with no history
of trochanteric pain or other local symp-
toms or trauma to this region were in-
cluded in the study. The MR imaging pro-
tocol was the same as described previously.
The imaging was performed with a 1.5-T
imager (Siemens Symphony; Siemens
Medical Systems, Erlangen, Germany).

Figure 2. Schematic of the anatomy on sectional images. Sagittal, Analysis of Images
transverse, coronal image through the anterior part of the greater
trochanter, and coronal image through the posterior part of the All 22 MR studies without injection in
greater trochanter (a,b,c,d, respectively). The dotted lines display the the bursae (cadaveric hips, n ϭ 10; vol-
needle paths for bursography. AF ϭ anterior facet, G. Medius ϭ gluteus unteers, n ϭ 12) were analyzed indepen-
medius, G. Minimus ϭ gluteus minimus, LF ϭ lateral facet, oe ϭ dently by two musculoskeletal radiolo-
obturator externus, oi ϭ obturator internus, p ϭ piriformis muscle, gists. Reader 1 (C.W.A.P.) was involved in
PF ϭ posterior facet, SPF ϭ superoposterior facet. the anatomic study, and reader 2 (C.B.C.)
was not involved in the anatomic study
best palpated. The SGMeB was accessed trast material (Omnipaque 350), and one and received only a description of the
through a lateral approach with the nee- part 15% concentrated solution of gela- facet anatomy and a diagram of the bur-
dle inserted in the coronal plane, tilted tin were slowly injected with fluoro- sal topography (Fig 2). All sequences used
30° inferiorly (Fig 2b). The target was the scopic monitoring into the TrB (6.3–9.5 before bursography were evaluated. Im-
upper part of the lateral facet, 3 mm un- mL), SGMeB (1.0 –2.5 mL), and SGMiB age analysis included both facet and bur-
derneath the cranial end of the trochan- (1.5–2.5 mL). MR imaging of the nine sal identification. Detectability of the
ter (Fig 2b, 2d). For the injection of the cadaveric hips was performed within 30 facet was rated from one to four on a
SGMiB, the needle was inserted from an minutes after the injection by using T1- continuous scale: 1, not detectable; 2,
anterolateral approach, targeting the an- weighted fat-saturated spin-echo se- poor detectability; 3, fair detectability;
terior facet at a point 3 mm beneath the quences in the transverse, coronal, and and 4, good detectability. The best plane
cranial end of the greater trochanter and sagittal planes, with the same parameters for the detection of each facet was re-
just medial to the intertrochanteric line. as previously described. corded. The visibility of the three bursae
In all cases, the needle was advanced un- on images before enhancement was rated
til bony contact was felt. MR and Anatomic Comparison by using the same scale, and the best
plane was also recorded. For the cadav-
The position of the tip of the needle After imaging, all cadaveric hips were eric studies, the dimensions of all facets
was verified with a test injection of a immediately positioned, with the hip were measured. The exact orientation of
small amount of iodinated contrast ma- joint in neutral position, frozen at Ϫ65°C the abductor tendons attaching to the
terial (Omnipaque 350; Nycomed Amer- for at least 24 hours, and subsequently trochanter was recorded, and the diame-
sham, Princeton, NJ). After the verifica- cut with a band saw into 3-mm-thick sec- ter of the tendons 3 mm proximal to
tion of intrabursal localization, a mixture tions that corresponded in thickness and their insertion was measured.
of one part 4-mmol/L gadopentetate orientation to those of the MR images:
dimeglumine (Magnevist; Schering, Ber- coronal (n ϭ 3), sagittal (n ϭ 3), or trans- RESULTS
lin, Germany), one part iodinated con-
Facet Anatomy of the Greater
Trochanter on MR Images

The results regarding the detectability
of facets on MR images are summarized

Volume 221 ⅐ Number 2 Greater Trochanter of the Hip: MR Imaging and Bursography ⅐ 471

TABLE 1
Detectability of the Facets of the Greater Trochanter on MR Images

Detectability* Reader 1 Reader 2

Facet Type Reader 1 Reader 2 Best Plane No. of Cases (%)† Best Plane No. of Cases†

Cadaveric study 4 4 Transverse 9 (90) Transverse 9 (90)
Anterior 3.6 4 Coronal 10 (100) Coronal 10 (100)
Lateral 3.2 4 Sagittal Sagittal 10 (100)
Superoposterior 3 4 Sagittal 9 (90) Sagittal 10 (100)
Posterior 10 (100)
4 4 Transverse Transverse 10 (83)
Human study 3.4 4 Coronal 11 (92) Coronal 12 (100)
Anterior 3.2 4 Sagittal 12 (100) Sagittal 12 (100)
Lateral 3.4 4 Sagittal 11 (92) Sagittal 12 (100)
Superoposterior 12 (100)
Posterior

* Mean of the assigned values for the detectability of the trochanteric facets on a continuous four-point scale: 1, not detectable; 2, poor detectability;
3, fair detectability; and 4, good detectability.

† Number of cases for which this plane was chosen for best detectability. Value in parentheses is the percentage.

in Table 1. The detectability of all facets No tendon insertion was present on this TABLE 2
in the cadaveric and living human hips facet, which was covered by the TrB. Dimensions of the Facets and Bursae
was fair to good for both readers. A sche- of the Greater Trochanter in the
matic of the cross-sectional appearance Cadaveric Study: Abductor Cadaveric Study
of the facets is displayed in Figure 2. Ta- Attachments in the Greater
ble 2 displays the dimensions of the fac- Trochanter Structure and Mean size Range
ets in the cadaveric hips. Dimension (mm) (mm)
The attachment sites of the gluteus
The anterior facet was usually best seen medius and minimus muscles in relation Facet 25.8 17–38
on transverse images (Fig 3a). It was also to the facet anatomy are displayed in Fig- Anterior 29.6 22–40
easily identified on anterior coronal and ure 1b. Mediolateral
sagittal images (Fig 3b). This facet corre- Craniocaudal 20.3 12–40
sponded to the insertion of the gluteus The gluteus medius attachment could Lateral 36.8 17–50
minimus tendon. The anterior border be divided into three parts. The main ten- Anteroposterior
was formed by the intertrochanteric line don arose from the central posterior por- Craniocaudal 15.2 10–24
just posterior to the capsular insertion of tion of the muscle and had a strong in- Superoposterior 17.0 10–22
the hip. On transverse images, a distinct sertion covering the superoposterior Mediolateral
tubercle was usually present at this inser- facet (Fig 5). The main tendon in the Anteroposterior 25.5 20–40
tion point (Fig 4b, 4c). transverse plane was flat and oval, with a Posterior 28.4 20–48
coronal to coronal oblique orientation. Mediolateral
The lateral facet was easily identified This tendon was thicker medially than Craniocaudal 42.5 38–47
on the coronal and coronal oblique im- laterally. The mean anteroposterior di- 47.5 43–55
ages through the posterior half of the ameter was 19.3 mm (range, 11–26 mm), Bursa
trochanter (Fig 3c). It had to be carefully and the mean mediolateral diameter was Trochanteric 16.2 12–22
distinguished from the anterior facet, 7.0 mm (range, 4 –12 mm). The lateral Mediolateral 20.0 14–22
which had a similar configuration in the part of the gluteus medius tendon inser- Craniocaudal
coronal plane but was clearly a more an- tion was obliquely orientated in the lat- Subgluteus 22.7 22–24
terior structure (Fig 2, c, d; Fig 3b). On eral facet from posterior to anterior, and medius 19.0 13–22
transverse images, the lateral facet ap- this tendon continued anteriorly cover- Anteroposterior
peared broader at its superior extent, be- ing the insertion of the gluteus minimus Craniocaudal
ing consistent with its inverted triangular tendon. The lateral part of the gluteus Subgluteus
shape. The superior border was in contact medius tendon arose from the undersur- minimus
with the superoposterior facet, which face of the muscle and was usually thin Mediolateral
constituted the most cranial part of the or almost purely muscular in nature (Fig Craniocaudal
trochanter. The lateral facet was identi- 3c). The anterior tendon was surrounded
fied best on sagittal images. This facet by and attached to the gluteus minimus ficial part of the muscle fascia. The mor-
was covered completely by the main ten- tendon (Fig 3a) but was not visible mac- phology of the tendon in the transverse
don of the gluteus medius muscle (Fig roscopically. plane was flat and biconcave. The main
3d). tendon usually was separated from the cor-
The gluteus minimus attachment could tical bone by a thin linear region with a
The sagittal plane was best for visualiz- be divided into two distinct components. signal intensity isointense to the muscle.
ing the posterior facet (Fig 3d). The con- The main tendon demonstrated a consis- The mean anteroposterior diameter at 3
tour was usually straight on sagittal im- tent insertion in the anterior facet of the mm above the insertion site was 17.4 mm
ages and curved on transverse images. On trochanter on its lateral and inferior aspect. (range, 12–23 mm), and the mean medio-
transverse images, the border between In contrast with the gluteus medius mus- lateral diameter was 4.0 mm (range, 2– 6
the lateral facet and the posterior facet cle, the main tendon arose from the super- mm). The transverse orientation of the ten-
was sometimes not well defined (Fig 3a). don at the level of the acetabular roof was
coronal oblique and parallel to the iliac

472 ⅐ Radiology ⅐ November 2001 Pfirrmann et al

teus minimus insertion was a muscular in-
sertion in the ventral and superior capsule
of the hip joint.

Figure 3. Facet anatomy on T1-weighted MR images (500/22). (a) Human study. Transverse Cadaveric Study: MR Anatomy of
image demonstrates the anterior (thin black arrows), lateral (black arrowheads), and posterior the Trochanteric Bursae
(white arrowheads) facets. The gluteus minimus tendon (large white arrow) is attached to the
anterior facet and is surrounded by the anterior part of the gluteus minimus muscle (thick black The localization of the bursae is shown
arrows). The TrB (small white arrows) can be identified as a thin line. (b– d) Cadaveric study. in Figures 1 and 2. Table 2 summarizes
(b) Coronal image through the anterior part of the greater trochanter shows the anterior facet the mean size of each bursa. The results
(arrowheads) with the gluteus minimus tendon (curved arrow) attached to it. (c) Coronal image regarding the detection of bursae on
through the posterior part of the greater trochanter displays the lateral facet (arrowheads), the nonenhanced MR images are shown in
lateral part of the gluteus medius tendon (curved arrow), and the iliotibial tract (straight arrow). Table 3.
(d) Sagittal image shows the superoposterior facet (black arrowheads), with the main tendon
(curved white arrow) of the gluteus medius muscle attached to it, and the posterior facet (white The TrB was the largest of the three
arrowheads). The piriformis tendon (straight white arrow), the obturator internus tendon bursae. It covered the posterior facet, the
(straight black arrow), and the obturator externus tendon (curved black arrow) are also seen. distal lateral part of the gluteus medius
tendon (lateral facet), and the proximal
wing. The tendon orientation twisted ap- during its course to the insertion in the part of the vastus lateralis insertion. The
proximately 90° in an inward direction anterior facet. The second part of the glu- bursa was located beneath the gluteus
maximus muscle and the iliotibial tract
(Figs 4c, 5). The TrB did not extend over
the anterior border of the lateral facet.
This bursa was lined by a small layer of
fat on both sides. Therefore, it could be
routinely identified on nonenhanced
transverse MR images as a fine linear
structure paralleling the posterior facet
(Figs 3a, 5d).

The SGMeB was found deep to the lat-
eral part of the gluteus medius tendon (Fig
6). The SGMeB covered an area of the su-
perior part of the lateral facet. Its superior
extent was marked by the tip of the tro-
chanter; its anterior extent, by the lateral
facet; and its posterior and inferior extent,
by the tendinous insertion of the gluteus
medius muscle. In one case, this bursa
communicated with the SGMiB. The col-
lapsed bursa could not be reliably identi-
fied on most nonenhanced MR images.

The SGMiB was found in the area of the
anterior facet (Fig 4). The bursa lay beneath
the gluteus minimus tendon, medial and
superior to its insertion. The bursa ex-
tended medially, covering the distal ante-
rior portion of the capsule of the hip joint
(Fig 4a). Its close relationship to the joint
capsule resulted in an intraarticular injec-
tion in one cadaver on the first attempt at
bursography. A more lateral position al-
lowed the entry of the needle into the
bursa. The SGMiB was sometimes detected
as a fine linear structure adjacent to the
gluteus medius tendon on nonenhanced
transverse MR images (Fig 4b). In most
cases, however, these two structures were
not distinguishable.

Human Study: MR Anatomy of the
Trochanteric Bursae

The results regarding the detection of
bursae on nonenhanced MR images of
the volunteers are shown in Table 3. The

Volume 221 ⅐ Number 2 Greater Trochanter of the Hip: MR Imaging and Bursography ⅐ 473

TrB could be seen in almost all cases. graphic anatomy. An understanding of Figure 4. Cadaveric study. (a) Transverse T1-
Only in one volunteer with increased the anatomy of the four facets of the weighted fat-saturated MR image (500/22) ob-
muscle mass and little fat was the bursa greater trochanter is essential because tained after bursography shows the SGMiB
not seen. The detectability and visibility each facet has specific tendinous attach- (black and white arrowheads) deep to the main
rates were in the same range for the ca- ments and specific nearby bursae. In ca- tendon of the gluteus minimus (curved black
daveric study and the human study. daveric specimens, the facets are easily arrow) and extending medially to merge with
delineated at gross inspection, and their the hip capsule. Contrast material is present in
DISCUSSION appearance is also constant on cross-sec- the hip joint space (curved white arrow).
tional MR images. (b) The corresponding nonenhanced trans-
The greater trochanter of the femur dem- verse T1-weighted MR image (500/22) shows
onstrates a complex but consistent topo- It is generally believed that muscular the SGMiB (black arrowheads) as a hypoin-
attachments to the bone demonstrate tense thin line medial to the main tendon of
the gluteus minimus (curved black arrow). The
tubercle (straight white arrow) at the medial
end of the anterior facet is the attachment site
of the hip capsule. (c) The corresponding
transverse anatomic section shows the dis-
tended SGMiB (black arrowheads) and TrB
(white arrowheads). Note the close relation-
ship between the SGMiB (black arrowheads)
and the hip joint space (curved arrow). The
osseous tubercle (straight arrow) with the at-
tachment of the hip capsule can be easily iden-
tified. (d) Coronal T1-weighted fat-saturated
MR image (500/22) through the anterior part
of the trochanter after SGMiB bursography
shows the relation of the contrast material in
the bursal space (curved arrow) and the con-
trast material in the joint space (straight ar-
row), which had resulted from a previous in-
advertent injection. (e) The corresponding
coronal anatomic section shows the gluteus
minimus muscle and tendon (white arrows)
and the SGMiB (black arrowheads) deep to the
gluteus minimus tendon. The anterior part of
the gluteus medius attachment (black arrow)
covers the gluteus minimus tendon.

four consistent zones composed of ten-
don, unmineralized fibrocartilage, min-
eralized fibrocartilage, and Sharpey fibers
(11). Although little variability is noted
in the presence of these zones, wide vari-

474 ⅐ Radiology ⅐ November 2001 Pfirrmann et al

Figure 5. MR images (500/22) of the TrB in cadaveric study. (a) Trans-
verse, (b) coronal, and (c) sagittal T1-weighted fat-saturated MR images
obtained after bursography of the TrB (curved white arrow). (d) Transverse
T1-weighted nonenhanced MR image shows TrB (white arrowheads) and
SGMiB (black arrowheads) as thin hypointense lines, the gluteus minimus
tendon (curved arrow), and the iliotibial tract (straight arrow).

ability can be noted in the extent in dif- called rotator cuff tears of the hip (6,7). tendon tears in the abductor mechanism,
ferent tendons. This variability can make This term refers to a tear in one or more all lesions were found in the gluteus me-
identification of distinct tendinous fibers of the tendons of the abductor mecha- dius tendon. However, 10 lesions ex-
difficult, as in the case of the gluteus me- nism of the hip, similar to that seen in tended into the gluteus minimus tendon.
dius and minimus attachments, which rotator cuff tears of the shoulder. Bunker
have large curving insertion sites in the et al (6) described the typical appearance Another reported series (13) detailed
trochanteric surface, each with strong of this tear as a circular or oval defect in the imaging findings of tears and avul-
tendinous as well as macroscopically ev- the gluteus minimus tendon, which ex- sive injuries of the gluteus medius muscle
ident, almost purely muscular attach- tended posteriorly in the lateral part of specifically. Four of six patients had a
ments (3,4,10). Although differences in the gluteus medius tendon. In this re- bony avulsion of the trochanteric tip at
the function of both of these gluteal mus- port, the tear appeared to originate in the the site of the attachment of the main
cles have been emphasized in the litera- gluteus minimus tendon (6). gluteus medius tendon, corresponding to
ture (1,3,8), there are no specific clinical the superoposterior facet. All four pa-
tests that can differentiate between dys- In a second series by Kagan (7), the tients showed atrophy of the gluteus me-
function of these muscles. rotator cuff tears of the hip were located dius muscle as an associated finding (13).
in the anterior part of the gluteus medius This analysis of previous reported finding
Recently, an expanded spectrum of tendon. During surgical exploration, appears to indicate a broad spectrum of
pathologic abnormalities affecting the signs of trochanteric bursitis and gluteal lesions that affect different parts of the
soft tissues of the hip has been noted, tendon tears often coexisted. In a re- abductor mechanism.
with such findings including not only ported series (12) of 35 patients with MR
trochanteric bursitis but also the so- imaging evidence of either tendinosis or Although inflammation of the syno-
vial lining of bursal cavities is well docu-
mented, such bursitis may lead to non-
specific symptoms, resulting in missed
diagnoses. Bursal cavities are frequently
underestimated in their importance, and
they even have been called “the Cin-
derellas of the body” (2). Bursae are
found where tendons move against each
other or glide over bony surfaces (1).
There are about 20 bursae (1,2) about the
hip joint, with variable extent and prev-
alence. The three bursae that are located

Volume 221 ⅐ Number 2 Greater Trochanter of the Hip: MR Imaging and Bursography ⅐ 475

about the greater trochanter and that are TABLE 3
investigated here are reported most com- Detectability of the Bursae of the Greater Trochanter on
monly (1). Some reports describe a sepa- Nonenhanced MR Images
rate subgluteus maximus bursa that lies
distal to the TrB (14). In the literature, Detectability (yes/no) Mean Detectability* Best Plane
the subgluteus maximus bursa is also
used synonymously with the TrB. In our Bursa Reader 1 Reader 2 Reader 1 Reader 2 Reader 1 Reader 2
analysis, only one bursa was seen in this
location. Cadaveric study 10/0 10/0 3.1 2.6 Transverse Transverse
TrB 2/8 1/9 1.2 1.2 Coronal Coronal
In the radiologic and orthopedic liter- SGMeB 4/6 2/8 1.6 1.4 Transverse Transverse
ature (14 –19), only the TrB has received SGMiB
attention. To our knowledge, our report 11/1 11/1 3.1 3.5 Transverse Transverse
represents the first description of the Human study 1/11 0/12 1.2 1.0 Coronal Coronal
cross-sectional anatomy and MR appear- TrB 4/8 5/7 1.5 1.6 Transverse Transverse
ance of the bursal complex of the tro- SGMeB
chanter. The TrB almost always was seen SGMiB
on standard transverse T1-weighted spin-
echo MR images as a thin hypointense * Mean of the assigned values for the detectability of the trochanteric bursae on a continuous
line, because it is surrounded by fat on four-point scale: 1, not detectable; 2, poor detectability; 3, fair detectability; 4, good detectability.
both sides (Fig 3a). The SGMiB is depicted
less frequently; such depiction relates ei- Figure 6. Cadaveric study. (a) Coronal T1-weighted fat-saturated MR image (500/22) through
ther to a separation of the bursa from the the lateral facet of the greater trochanter after bursography shows the SGMeB (arrowheads) deep
gluteus minimus tendon by a fat layer or to the lateral part of the gluteus medius tendon and muscle (arrows). (b) On the corresponding
to a medial extension of the bursa (Fig anatomic section, the SGMeB (arrowheads) is filled with gelatin.
4b). Depiction of this bursa is improved
when it is distended either with an effu- gest that the injection site should be All of our injections were performed in
sion or injection of a contrast agent. slightly more posterior, with the needle cadavers of elderly patients. The extent
approaching the greater trochanter from and appearance of the bursae may be dif-
Lateral hip pain is often a challenging a posterolateral position. Fluoroscopic ferent in a younger population. Further,
diagnostic and therapeutic problem, guidance with contrast enhancement al- only limited clinical data were available
with a wide range of differential diagnos- lows verification of intrabursal injection. regarding the cadaveric specimens, and
tic considerations that include articular Fluoroscopic or CT guidance also allows clinical symptoms of the trochanteric re-
disease, soft-tissue alteration, and re- accurate localization of the two other gion could not be completely excluded.
ferred pain from a distant pathologic smaller bursae for diagnostic or therapeu-
condition. The spectrum of pathologic tic purposes. In conclusion, MR imaging and bur-
abnormalities includes classic trochan- sography provide detailed information
teric bursitis; degeneration and tears of This study has some limitations. It is about the anatomy of the tendinous at-
the tendons of the gluteus medius and well known that bursal cavities show tachments of the abductor muscles and
minimus muscles; arthritis of the hip; in- great variability. Owing to the small sam- the bursal complex of the greater trochan-
sufficiency fractures of the proximal por- ple, this issue cannot be appropriately ter of the femur. A detailed understanding
tion of the femur, a spinal source of pain; addressed with the results of our study. of the anatomy of the trochanteric facets is
and other nerve entrapment syndromes;
as well as poorly defined syndromes such
as pseudotrochanteric bursitis, adiposis
dolorosa, tensor fasciae femoris syn-
drome, abductor muscle strain, and myo-
fascial pain (19). Many of these disorders
lead to similar clinical symptoms, and
imaging studies such as routine radiogra-
phy and bone scintigraphy may be unre-
warding. MR imaging potentially may
provide useful information to initiate the
appropriate treatment.

The injection of the trochanteric bur-
sae is technically easy to perform. In our
small series, enhancement of all three
bursae was successful. The main purpose
for an intrabursal injection is related to
diagnostic or therapeutic injection of lo-
cal anesthetics and corticosteroids in pa-
tients with suspected trochanteric bursi-
tis. The classic injection site is described
as the point of maximum tenderness or
the point where the greater trochanter is
most palpable (18,20). Our results sug-

476 ⅐ Radiology ⅐ November 2001 Pfirrmann et al

important for understanding the structure, 7. Kagan A II. Rotator cuff tears of the hip. 14. Little H. Trochanteric bursitis: a common
function, and pathologic abnormalities Clin Orthop 1999; 368:135–140. cause of pelvic girdle pain. Can Med As-
that are seen commonly in this region. soc J 1979; 120:456 – 458.
8. Gottschalk F, Kourosh S, Leveau B. The
References functional anatomy of tensor fasciae latae 15. Steinbach LS, Schneider R, Goldman AB,
and gluteus medius and minimus. J Anat Kazam E, Ranawat CS, Ghelman B. Bursae
1. Williams RL, Warwick R. Gray’s anatomy. 1989; 166:179 –189. and abscess cavities communicating with
Philadelphia, Pa: Saunders, 1980; 390–394, the hip: diagnosis using arthrography
477–482, 600– 603. 9. Armbuster TG, Guerra J Jr, Resnick D, et and CT. Radiology 1985; 156:303–307.
al. The adult hip: an anatomic study. I.
2. Bywaters EGL. The bursae of the body. The bony landmarks. Radiology 1978; 16. Jaovisidha S, Chen C, Ryu KN, et al. Tu-
Ann Rheum Dis 1965; 24:215–218. 128:1–10. berculous tenosynovitis and bursitis: im-
aging findings in 21 cases. Radiology
3. Beck M, Sledge JB, Gautier E, Dora CF, 10. Nazarian S, Tisserand P, Brunet C, Muller 1996; 201:507–513.
Ganz R. The anatomy and function of the ME. Anatomic basis of the transgluteal
gluteus minimus muscle. J Bone Joint approach to the hip. Surg Radiol Anat 17. Swezey RL. Pseudo-radiculopathy in sub-
Surg Br 2000; 82:358 –363. 1987; 9:27–35. acute trochanteric bursitis of the subglu-
teus maximus bursa. Arch Phys Med Re-
4. Duparc F, Thomine JM, Dujardin F, et al. 11. Resnick D, Niwayama G. Entheses and habil 1976; 57:387–390.
Anatomic basis of the transgluteal ap- enthesopathy: anatomical, pathological,
proach to the hip-joint by anterior and radiological correlation. Radiology 18. Shbeeb MI, O’Duffy JD, Michet CJ Jr,
hemimyotomy of the gluteus medius. 1983; 146:1–9. O’Fallon WM, Matteson EL. Evaluation of
Surg Radiol Anat 1997; 19:61– 67. glucocorticosteroid injection for the
12. Kingzett-Taylor A, Tirman PF, Feller J, et treatment of trochanteric bursitis. J Rheu-
5. Kumagai M, Shiba N, Higuchi F, Nish- al. Tendinosis and tears of gluteus medius matol 1996; 23:2104 –2106.
imura H, Inoue A. Functional evaluation and minimus muscles as a cause of hip
of hip abductor muscles with use of mag- pain: MR imaging findings. AJR Am J 19. Traycoff RB. “Pseudotrochanteric bursi-
netic resonance imaging. J Orthop Res Roentgenol 1999; 173:1123–1126. tis”: the differential diagnosis of lateral
1997; 15:888 – 893. hip pain. J Rheumatol 1991; 18:1810 –
13. Chung CB, Robertson JE, Cho GJ, 1812.
6. Bunker TD, Esler CN, Leach WJ. Rotator- Vaughan LM, Copp SN, Resnick D. Glu-
cuff tear of the hip. J Bone Joint Surg Br teus medius tendon tears and avulsive in- 20. Raman D, Haslock I. Trochanteric bursi-
1997; 79:618 – 620. juries in elderly women: imaging findings tis: a frequent cause of “hip” pain in rheu-
in six patients. AJR Am J Roentgenol matoid arthritis. Ann Rheum Dis 1982;
1999; 173:351–353. 41:602– 603.

Volume 221 ⅐ Number 2 Greater Trochanter of the Hip: MR Imaging and Bursography ⅐ 477