A thirty-two-year-old right-handed male landscaper and snowplow
operator went to his primary-care physician with a one-month history
of pain, heaviness, and positional weakness in his right shoulder.
He was treated with limitation of activity, nonsteroidal anti-inflammatory
medication, and shoulder exercises for five months prior to referral
to our office.
Physical examination, performed six months after the onset of symptoms,
revealed isolated weakness of external rotation of the shoulder
with slight atrophy of the infraspinatus muscle. Electromyography
revealed selective partial denervation of the infraspinatus muscle,
and nerve-conduction studies showed evidence of delayed conduction
to the infraspinatus muscle, indicating nerve entrapment.
Magnetic resonance imaging demonstrated an ovoid mass, 1.3 cm
long, with homogeneous fluid-like signal intensity in the spinoglenoid
notch (Fig. 1).
The infraspinatus muscle appeared brighter than the other muscles
on T2-weighted fast-spin-echo images, consistent with denervation.
The posterosuperior aspect of the glenoid labrum was abnormal, suggesting
a labral tear contiguous with the glenoid labral cyst.
The patient elected to undergo magnetic resonance image-guided
needle aspiration of the cyst. The hospital’s Human Research
Committee granted permission for the procedure, and written informed
consent was obtained from the patient. The aspiration was performed
in an open-configuration 0.5-T superconducting magnetic resonance
system (Signa SP; General Electric Medical Systems, Milwaukee, Wisconsin).
The magnet has a double-doughnut configuration with a 56-cm-wide
vertical space between the magnet rings that permits direct access
to the patient during imaging. In addition to standard multiplanar
images, this system can acquire continuously updated images, which
can be manipulated with use of an integrated optical tracking system
that monitors the position and trajectory of a needle-holder (Pixsys
3000 Flashpoint Position Encoder; Image-Guided Technologies, Boulder,
Colorado). With this optical tracking system, the scanner obtains two-dimensional
images along the needle path or orthogonal to it9.
These images are displayed on a monitor within the magnet’s
vertical gap. Thus, the needle location and trajectory are visualized
on images acquired in near-real time while the intervention is being
performed.
Method of Aspiration
The patient was positioned prone, and a small, flexible transmit-receive
surface coil was placed over the right shoulder. Fast-gradient-recalled
echo images (repetition time, 20 msec; echo time, 9.8 msec; flip
angle, 90°; matrix, 256 128; 1 NEX) were obtained continuously
during needle placement. The entry point and angle of the needle
were chosen by observing the projected needle path on the continuously acquired
images while scanning with an empty needle-holder against the skin.
When the projected needle path, displayed as an interrupted line
on the images, passed through the lesion without intersecting any
critical structures or bone, the entry point was marked on the skin.
The right shoulder, with the magnetic resonance surface coil in
place, was then prepared and draped under sterile conditions. An
18-gauge, 5-cm-long magnetic resonance-compatible needle (EZ-Em;
Westbury, New York) was mounted into a needle-holder, and the needle
was advanced into the lesion under direct visualization on the continuously
acquired interactive magnetic resonance images (Figs. 2-A and 2-B). After confirmation
that the needle was within the labral cyst on images acquired in
three orthogonal planes relative to the needle, the lesion was aspirated.
Initially, only a scant volume of gelatinous, clear, yellow fluid
could be extracted. A second 18-gauge needle was then placed in
tandem to the first needle. Aspiration through the second needle
yielded an additional 0.4 mL of fluid. Axial T2-weighted fast-spin-echo
images showed nearly complete resolution of the lesion (Figs. 3-A and 3-B). One milliliter
of steroid suspension (Celestone Soluspan; Schering, Kenilworth,
New Jersey) was injected immediately superficial to the site of
the labral cyst. Fentanyl (50 g) and Versed (midazolam; 0.5 mg)
were administered intravenously during the procedure for conscious
sedation. To ensure safe recovery from the medications, the patient
was observed for two hours in our post-procedure holding area prior
to discharge.
Postoperative Course
The patient returned to work within a few days, by which time his
symptoms had resolved completely. Analysis of the aspirate demonstrated
hypocellular fluid containing scattered histiocytes, characteristic
of a ganglion cyst.
One year later, the patient remained asymptomatic and had no muscle
weakness or other signs of suprascapular neuropathy on physical
examination. Magnetic resonance images showed no recurrence of the
glenoid labral cyst and complete resolution of the infraspinatus
muscle edema (Fig. 4).
Three years and seven months after treatment, the patient was still
asymptomatic and had normal findings on physical examination.
Suprascapular neuropathy is an uncommon cause of shoulder discomfort
and disability. It most commonly results from acute trauma or from
overuse of the shoulder due to repetitive movements that exert traction
on the suprascapular nerve as it passes through the suprascapular
or the spinoglenoid notch6,10-15.
Less commonly, mass lesions at either the suprascapular or the spinoglenoid
notch may compress the nerve as it passes through these locations.
Epidemiology
While mass lesions are an uncommon cause of suprascapular nerve
compression, glenoid labral cysts are the most common of the mass
lesions reported to cause this abnormality. Antoniadis et al.16, in a report of twenty-eight consecutive
cases of suprascapular neuropathy in twenty-seven patients who were
seen over a ten-year period, described only three cases of mass
lesions causing nerve compression; all of the lesions were ganglia and
were in the spinoglenoid notch. Fritz et al.1 reported
on twenty-six patients who had suprascapular nerve compression caused
by mass lesions that were identified by magnetic resonance imaging.
Twenty patients (77%) had a glenoid labral cyst, five (19%)
had a malignant soft-tissue mass, and one (4%) had a hematoma
secondary to a scapular fracture. Glenoid labral cysts may also
be present without suprascapular nerve compression1,4. Fritz et al.1 found
one glenoid labral cyst in twenty-five asymptomatic, healthy subjects
who had unilateral magnetic resonance imaging.
A total of sixty-three glenoid labral cysts were reported in sixty-two
patients in three series1,2,4.
The mean age of the patients was thirty-six years (range, sixteen
to sixty-eight years). The cysts were much more common in men (89%)
than they were in women (11%). In two of these studies2,4, the maximum cyst diameter was
between 3 and 70 mm. The mean cyst diameter, reported in one study4, was 11 mm. Forty-nine cysts (78%)
were located in the spinoglenoid notch; three (5%), in
the suprascapular notch; four (6%), in both the suprascapular
and the spinoglenoid notch; and seven (11%), at another
site1,2,4.
Etiology
Some authors have suggested that there may be a causal relationship
between glenoid labral cysts and glenoid labral tears2,4,6,17. Tirman et al.4 were the first, to our knowledge,
to describe this relationship, in a report on twenty-five glenoid
labral cysts with associated labral tears that were visualized on
magnetic resonance imaging studies. Those authors postulated that
glenoid labral cysts are analogous to meniscal cysts in the knee
and result from a collection of joint fluid pumped through tears
in the glenoid labrum or capsule. Moore et al.2 found
labral tears in ten of eleven patients who had a glenoid labral
cyst and had had an arthroscopic procedure. In that study, the labral
lesions were not visualized on magnetic resonance imaging studies.
Treatment
On the basis of six cases and a review of the literature, Fehrman
et al.18 proposed an algorithm
for the treatment of glenoid labral cysts that cause nerve compression
at the spinoglenoid notch. Those authors recommended that, if nonoperative
therapy fails, arthroscopic evaluation and treatment of the intra-articular
abnormalities, followed by open excision or percutaneous aspiration
of the cyst, should be performed18.
Iannotti and Ramsey7 reported
that three patients who had a labral cyst of the spinoglenoid notch
were treated successfully with arthroscopic decompression of the
cyst into the joint followed by labral repair. Moore et al.2 reported that six of eleven patients
required open resection of a labral cyst in addition to arthroscopic
repair of the labrum because the cyst could not be adequately treated
through the arthroscope. Arthroscopy has been suggested as the preferred method
of treatment of these lesions2,6,19.
While arthroscopic decompression or open resection has been the
treatment of choice for glenoid labral cysts2,6,7,18,
several surgeons have initiated the use of computerized tomography
or ultrasound for performing percutaneous needle aspiration under
direct visualization1,4,8,20-22.
The role of minimally invasive modalities for the treatment of these
cysts is still being investigated.
Results
There is very little information on the natural history of untreated
labral cysts. Fritz et al.1 reported
on one patient who had a 1.5-cm-diameter cyst at the spinoglenonid
notch. The cyst was unchanged in size on the magnetic resonance
imaging study made at ten months, and the patient remained asymptomatic
at twenty months. In two other reported cases, a glenoid labral
cyst that was diagnosed with magnetic resonance imaging studies
could not be found at the time of the operation and therefore was
presumed to have resolved spontaneously1,2.
The results of nonoperative treatment of glenoid labral cysts have
been inconsistent. Fritz et al.1 documented
two cases of spontaneous resolution of glenoid labral cysts, with
a decrease in symptoms, after nonoperative treatment. Moore et al.2 reported on six patients with a glenoid
labral cyst at the spinoglenoid notch who had been treated with
rest, physical therapy directed at rotator-cuff strengthening, and
work restrictions involving avoidance of repetitive above-the-shoulder
activities. After one year, four patients had reported no change
in the symptoms and two had shown improvement. One of the latter
two patients returned to full activity after a period of limited
activity, and the other patient continued to limit his activity
because of pain. The findings on magnetic resonance imaging studies
mirrored the clinical results, with nearly complete resolution of
the cyst in the patient who returned to full activity and with no
change in the other patient.
Previous authors have reported on ten patients who had computerized
tomography image-guided needle aspiration (four patients)1,20 or ultrasound image-guided needle
aspiration (six patients)4,8,21,22.
A successful outcome was reported for eight of these ten patients
after two months to two years of clinical follow-up. Four of the
six patients who had been treated with aspiration with the aid of
ultrasound imaging had relief of pain, and three also had a decrease
in muscle atrophy; two of the six had recurrence of the labral cyst
within six months after treatment4,8,21,22.
All four patients who underwent aspiration with the aid of computerized
tomography had resolution of the symptoms1,20.
The duration of follow-up for some of these patients may have been
too short to observe a delayed recurrence of the cyst.
Excision or aspiration of the labral cyst may be followed by reaccumulation
of cyst fluid because the underlying abnormality of the capsulolabral
complex has not been repaired. Additionally, symptoms related to
the labral abnormality may persist. Cyst recurrence has been reported
in two patients who had an open procedure in which the labral abnormality
was not repaired2,4 as well as
in two others in whom only needle aspiration was performed4,21.
Comparison of Minimally Invasive Treatment
Options
Ideally, image-guidance should provide rapid, multiplanar images
of the needle, cyst, and surrounding structures, so that the needle
can be followed throughout the procedure and adjustments can be
made to obtain the optimal trajectory. Magnetic resonance imaging
provides better soft-tissue contrast for visualizing labral cysts
than does computerized tomography4.
While ultrasound is useful for delineating cysts, the ultrasound
beam cannot penetrate bone, thus blinding the operator to portions
of the anatomy. Magnetic resonance imaging does not have this limitation.
Magnetic resonance image-guided processes performed with use of
our method and ultrasound image-guided processes are interactive,
since the plane of imaging, which contains the instrument and the
target, is controlled by the operator and changes during the procedure9,23. Computerized tomography-guided
procedures are not interactive, as needle adjustments are typically
made between scans.
The magnetic resonance-image guided method that we used inherently
shows the needle and the cyst on each of the images, without use
of an additional mechanical needle-guide or stereotactic frame.
Since the magnetic resonance images obtained during needle placement
are acquired essentially in real time, the resultant loss of signal
to noise and spatial resolution is apparent when these images are
compared with conventional, diagnostic images. Although the quality
of the rapidly obtained images is suboptimal for diagnosis, it is
adequate for needle guidance.
Implications of Magnetic Resonance Image-Guided
Therapy
Magnetic resonance image-guided interventional therapy is a rapidly
growing technique that is being used to augment a variety of procedures
in almost all divisions of medicine. We have been involved in the
development of magnetic resonance image-guided interventional procedures
for the last five years. During this time, we have performed twenty-five
magnetic resonance image-guided orthopaedic procedures, including three
needle aspirations of ganglia (one each in the shoulder, knee, and
ankle), one needle aspiration of an acetabular paralabral cyst,
two injections of steroid into the iliopsoas bursa, fifteen needle
biopsies of soft-tissue masses (five of which have been reported
previously9), and four image-guided
open excisional biopsies of bone lesions. Currently, most needle
procedures can be performed within approximately one hour of magnet
time, which includes the time required for sterile draping of the
patient.
The cyst described in the current report was small; the volume estimated
on magnetic resonance imaging was about 1 mL. We aspirated 0.4 mL
of this material and assumed that the remainder extravasated into
the surrounding soft tissues. While it is possible that the ganglion
would have resolved spontaneously, we thought that this was unlikely
since the symptoms had persisted for five months despite nonoperative treatment.
The complete recovery and long-term success seen in our patient,
and previous reports of successful needle-aspiration therapy, suggest
that percutaneous needle aspiration may be the only treatment required
for some patients who have a glenoid labral cyst. Since image-guided
needle aspiration is less invasive than the operative alternatives,
this procedure may be indicated for patients who have glenoid labral
cysts that do not respond to nonoperative treatment. Should a cyst
recur following initial needle aspiration, the procedure could be repeated
or an operative treatment may be performed. If, despite successful
percutaneous treatment of the cyst, the underlying abnormality of
the capsulolabral complex proves to be the cause of persistent symptoms,
arthroscopic labral repair can be performed subsequently.
Although the magnetic resonance imaging system that we used is
not widely available, these techniques are transferrable to other
open-configuration systems23,24.
While an integrated optical tracking system makes needle guidance
easier, the procedure can be performed easily under magnetic resonance-image
guidance with use of any open-configuration magnet combined with
the standard biopsy technique employed for computerized tomography,
assuming that the magnetic resonance system produces images of adequate
quality.
The cost of the magnetic resonance image-guided aspiration procedure
has not yet been determined by the third-party payers. The cost
of ultrasound image-guided and computerized tomography image-guided
aspiration procedures at our hospital is $147 and $470,
respectively, as defined by the Medicare Part-B 1999 fee schedule.
If we assume a higher cost for a magnetic resonance image-guided
procedure that is proportional to the difference between the diagnostic
and procedural costs of computed tomography, the magnetic resonance
procedure will likely cost between $850 and $1000.
This compares favorably with the $3500 cost, at our hospital,
for the arthroscopic procedure that would have been required for
this patient.
We have found magnetic resonance image-guided needle aspiration
to be a simple procedure with minimal associated morbidity, essentially
no different than that associated with an arthrogram or an injection
of a joint or the bursa. The risk of infection should be similar
to that associated with arthrography, which has been reported to
be less than 0.003%25.
In our experience, intravenous conscious sedation may be necessary
for procedures involving deep lesions, but this adds little risk
of morbidity. At our institution, magnetic resonance image-guided
procedures are conducted as a team effort with close collaboration
between departments. Simple procedures, such as needle biopsies,
are performed by a radiologist after consultation with the referring
surgeon. Image-guided open procedures are performed by the surgeon
in conjunction with a radiologist who monitors and interprets the
images and an anesthesiologist who delivers general anesthesia.
We believe that magnetic resonance image-guided aspiration of
glenoid labral cysts of the shoulder represents a safe, minimally
invasive alternative to operative treatment, especially in patients
without shoulder instability. This technique offers advantages compared
with other methods of image guidance.