The study was approved by the institutional ethics committee, and informed
consent was obtained from each patient. Synovectomies were performed on sixty
elbows in fifty-five patients from 1987 to 1994. There were two deaths; one
patient died as a result of cancer and one died from a perforated gastric
ulcer. Fifty-eight elbows in fifty-three consecutive patients (nine men and
forty-four women) with polyarticular rheumatoid
arthritis14 were
followed. The average time from the operation to the final examination was
thirteen years (range, ten to eighteen years). The ages of the patients at the
time of the operation ranged from thirty-one to sixty-four years (mean,
fifty-one years). The average durations of rheumatoid disease and elbow pain
were 7.6 years and 2.4 years, respectively. Thirty-two right elbows and
twenty-six left elbows were treated (see Appendix).
Eligibility criteria were elbow pain that had not responded to conservative
measures, including disease-modifying antirheumatic drugs or prednisone and
the avoidance of provocative activity, and radiographic changes that were
grade 2 or less according to the system of Larsen et
al.13. The
registered patients were assigned to one of two treatment groups (arthroscopic
or open synovectomy) with use of a minimization
method15, in which
assignment to arthroscopic or open synovectomy is selected to minimize any
imbalance between the two treatment groups with respect to the preoperative
arc of elbow flexion (=90° or <90°). The two operations were
performed by two senior surgeons (N.T. and H.S.) with each surgeon performing
his accustomed synovectomy (open or arthroscopic) to remove a possible bias in
surgical technique.
Arthroscopic synovectomy is carried out with the patient placed in the
lateral position with the involved side uppermost after the induction of
general anesthesia. The arm is supported with the forearm hanging free and the
elbow flexed 90°. A pneumatic tourniquet is used to exsanguinate the arm.
Twenty milliliters of sterile normal saline solution is injected to distend
the capsule. Four portals are used: mid-lateral, posterolateral,
anterolateral, and anteromedial. The mid-lateral portal, which is located in
the center of a triangle formed by the olecranon, the lateral epicondyle, and
the radial head, is useful as both a diagnostic and an operative portal for
the posterior compartment, including the olecranon, olecranon fossa, posterior
aspect of the capitellum, and posterior aspect of the radial head. The
posterolateral portal is established in the sulcus of the joint space
posterior to the mid-lateral portal for the insertion of instruments, such as
probes, graspers, and shavers. The anterolateral portal is established exactly
in the sulcus between the radial head and the capitellum anteriorly. The
anteromedial portal is then established with use of the switch-stick
technique. We use a 4.0-mm, 30° arthroscope with an attached camera, light
source, and fluid source. Anterior joint structures, such as the trochlea,
medial condyle, capitellum, radial head, and coronoid process, can be accessed
through the anterolateral and anteromedial portals. Proliferative synovial
tissue is excised to the extent possible with the grasper and the 3.5-mm
shaver with the side-cutting opening. Rotation of the forearm is important to
facilitate detection of residual synovial tissue behind the radial neck.
(Radial head resection was not performed in the current group of patients
because none had destruction of the radial head or the capitellum.) We do not
apply a splint, and the patient begins active motion of the elbow under the
supervision of a therapist and the physician on the day of the operation.
Open synovectomy is carried out with the patient placed in the supine
position with the arm flexed on an armtable after the induction of general
anesthesia. With use of tourniquet control, the elbow is approached with an
extended lateral incision from the proximal portion of the lateral epicondyle
to the posterior border of the ulna. The dissection is made between the
brachioradialis and the triceps muscles in the arm and between the extensor
carpi ulnaris and the anconeus in the forearm. The lateral collateral ligament
is released as a distally based flap. The joint capsule is incised, and the
proximal end of the ulna is mobilized to allow its subluxation medially on the
humerus until adequate access to the medial side of the joint is obtained.
Synovectomy is carried out until the lateral and medial compartments are
cleaned of proliferative synovial tissue involving the capsule. (In the study
patients, the radial head was not resected because the cartilage of the
radiocapitellar joint was not destroyed and motion of the elbow was not
restricted by the radial head.) The lateral ligament complex is reconstructed
with use of 1-0 Dacron sutures placed through 1.0-mm-diameter drill-holes in
the humerus. The tension on the suture line should be such that there is
unobstructed passive motion of the elbow but no subluxation. The skin and
other soft tissues are closed. A rubber Penrose drain is used for twenty-four
to forty-eight hours. Postoperatively, a long splint and compression dressing
are used until the soft-tissue reaction subsides (usually five to seven days
after surgery). Active and active-assisted motion exercises are then begun
under the supervision of a therapist.
The Mayo elbow performance
score5, which is
used to evaluate pain (maximum score, 45 points), motion (20 points),
stability (10 points), and daily function (25 points), was used to assess
clinical symptoms before the operation, at three to five years after the
operation (mid-term follow-up) and at the time of the latest follow-up (ten to
eighteen years after the operation). The radiographic findings were graded,
according to the method described by Larsen et
al.13, before the
operation and at the mid-term and latest follow-up evaluations. During the
assessment of the clinical symptoms, the elbows were covered with an elastic
bandage to blind the examiner. Both the clinical symptoms and the radiographic
changes were assessed separately by two trained examiners. The mean intraclass
correlation coefficient for intraobserver reproducibility was 0.93.
Statistical analysis was performed with one-way analysis of variance and
the chi square test. A probability value of <0.05 was considered to
indicate a significant difference.
No difference was detected between the two treatment groups (arthroscopic
or open synovectomy) in terms of age at the time of surgery, gender, duration
of disease and elbow pain, involved side, follow-up period, postoperative
dosage of prednisone, postoperative use of disease-modifying antirheumatic
drugs, preoperative pain level, range of motion, stability score, function,
Mayo elbow performance score, and preoperative radiographic Larsen grade (see
Appendix).
Clinical Assessment
For the patients who underwent arthroscopic synovectomy, the average pain
score at the time of the mid-term follow-up (34 points) and the latest
evaluation (23 points) had improved significantly (p < 0.001 and p <
0.01, respectively) compared with that at the preoperative evaluation (12
points). The mean arcs of flexion and forearm rotation were 82° and
99°, respectively, at the preoperative examination, 104° and 115°
at the mid-term evaluation, and 103° and 115° at the latest
evaluation. The mean stability scores preoperatively and at the mid-term and
latest evaluations in the arthroscopy group were 9 points. The average daily
function score of the upper limb at the mid-term evaluation (16 points) (p
< 0.05) and the latest evaluation (15 points) had improved compared with
the initial score (13 points). The mean Mayo elbow performance score at the
mid-term evaluation (78 points) and at the latest evaluation (67 points) had
increased significantly (p < 0.01 and p < 0.05, respectively) compared
with the initial score (50 points).
For the patients who underwent open synovectomy, the mean pain score at the
mid-term evaluation (37 points) and at the latest evaluation (31 points) had
also improved significantly (p < 0.001 and p < 0.01, respectively)
compared with that at the preoperative evaluation (12 points). The mean arcs
of flexion and forearm rotation were 83° and 100°, respectively, at
the preoperative evaluation, 95° and 114° at the mid-term evaluation,
and 99° and 112° at the latest evaluation. The mean stability scores
preoperatively and at the mid-term and latest evaluations in the open
synovectomy group were 9 points. The mean postoperative elbow function scores
at the mid-term and at the latest evaluations (15 points) were significantly
(p < 0.05) higher than the mean initial score (12 points). The mean Mayo
elbow performance score at the mid-term evaluation (78 points) and at the
latest evaluation (71 points) had increased significantly (p < 0.01 for
both) compared with the initial score (50 points). With the numbers available,
no significant difference was detected between the arthroscopic and open
synovectomy groups at the mid-term and at the latest evaluations with respect
to these measured parameters.
The overall clinical results in the elbows with a preoperative arc of
flexion of =90° in the arthroscopic synovectomy group were comparable
with those with a preoperative arc of flexion of =90° in the open
synovectomy group, although the mean scores for pain, function, and Mayo elbow
performance in the arthroscopy group were lower than those in the open
synovectomy group at both the mid-term and latest evaluations.
The elbows with a preoperative arc of flexion of <90° in the
arthroscopic synovectomy group had significantly (p < 0.05) better
function, on the average, than the elbows with a preoperative arc of flexion
of <90° in the open synovectomy group at the mid-term evaluation. The
mean scores for pain were lower but the mean arcs of motion and the mean
scores for function and Mayo elbow performance were higher in the arthroscopy
group than those in the open synovectomy group at the mid-term and latest
evaluations.
Recurrence of Synovitis and Reoperations
Clinically apparent synovitis recurred in six of the twenty-nine elbows
after arthroscopic synovectomy and in three of the twenty-nine elbows after
open synovectomy. The mean durations until recurrence were eight years after
arthroscopic synovectomy and nine years after open synovectomy.
In the fifteen patients with <90° of elbow flexion preoperatively in
the open synovectomy group, three elbows had become ankylosed with progression
of joint destruction, which reduced the ability of the patients to perform the
activities of daily living. Therefore, total elbow arthroplasty was performed
at eight years after surgery in two elbows and at nine years after surgery in
one elbow.
Postoperative Complications
One patient with =90° of elbow flexion in the arthroscopy group had
had transient neurapraxia of the radial nerve but function recovered
completely at three months after surgery, and another with <90° of
elbow flexion had a transient neurapraxia of the ulnar nerve but function
recovered completely at six months after surgery. In one patient with
<90° of elbow flexion in the open synovectomy group, a superficial
infection around a wound was observed at the mid-term evaluation, but no
infection was observed at the latest evaluation.
Radiographic Assessment (see Appendix)
Progression of joint destruction after arthroscopic synovectomy was the
same as that after open synovectomy. The mean Larsen grades after arthroscopic
or open synovectomy were 1.8 at the time of the mid-term follow-up and 3.1 at
the latest evaluation. No difference was found in the results for the elbows
with or without a preoperative arc of flexion of 90° after either
procedure.
Studies on open synovectomy of the elbow for the treatment of rheumatoid
arthritis have shown that 70% (twenty-eight) of forty
elbows16 to 88%
(twenty-two) of twenty-five
elbows17 evaluated
at the time of mid-term follow-up (average, five to seven years) and 67%
(fourteen) of twenty-one
elbows18 evaluated
at the time of long-term follow-up (average, fourteen years) had relief of
pain. Some elbows have had loss of flexion after open
synovectomy5,16-18.
Studies on arthroscopic synovectomy have shown that nine (64%) of fourteen
elbows8 to sixteen
(76%) of twenty-one
elbows9 evaluated at
the time of mid-term follow-up (average, four to eight years) had relief of
pain, and the arc of flexion improved or remained the same after arthroscopic
synovectomy8,9.
According to those studies, open synovectomy can effectively relieve pain for
a long period of time, but there have been no reports on long-term outcomes of
arthroscopic synovectomy, to our knowledge. However, when a restriction of
preoperative elbow movement is aggravated by open synovectomy, the ensuing
elbow stiffness can reduce the ability of the patient to perform the
activities of daily living. We believed that long-term assessment of pain
relief, including the relationship between preoperative and postoperative arcs
of motion, could help to determine the benefits of and the indications for
arthroscopic and open synovectomy.
The following three clinical points were clarified at the final follow-up
evaluation (average, thirteen years). First, eleven (48%) of the twenty-three
elbows in which arthroscopic synovectomy was performed and sixteen (70%) of
the twenty-three elbows in which open synovectomy was performed were only
mildly painful or not painful. However, with the numbers studied, there was no
significant difference between the overall clinical results of these two
procedures. Second, in elbows with a preoperative arc of flexion of
=90°, the clinical results of arthroscopic synovectomy were comparable
with those of open synovectomy, although the mean scores for pain, function,
and Mayo elbow performance at the time of the mid-term follow-up in patients
treated with open synovectomy were better. Third, in elbows with a
preoperative arc of flexion of <90°, arthroscopic synovectomy provided
significantly (p < 0.05) better function than open surgery at the time of
the mid-term follow-up, and the mean arcs of motion and the mean scores of
function and Mayo elbow performance in these patients were better than those
in patients treated with open surgery at the most recent followup evaluation.
We believe that arthroscopic synovectomy, in which only the proliferating
inflammatory synovial tissue is excised, provides a better balance between
pain relief and improvement in motion regardless of the amount of preoperative
elbow motion. Persistent improvement in pain relief can be obtained with open
synovectomy in which proliferative synovial tissue involving the capsule is
excised completely. However, patients with rheumatoid arthritis and
preoperative elbow stiffness have a higher risk of decreased motion following
open synovectomy.
Recurrent synovitis was observed in six (21%) of twenty-nine elbows after
arthroscopic synovectomy and in three (10%) of twenty-nine elbows after open
synovectomy. However, total elbow arthroplasty was necessary for three elbows
with a preoperative arc of flexion of <90° because of stiffness and
poor performance in the activities of daily living after open synovectomy.
In our series, complications were uncommon and were not severe. However,
the arthroscopic procedure should be performed by a trained and experienced
surgeon with careful attention to detail and a thorough knowledge of joint
anatomy, especially in an elbow with stiffness. Fibrous ankylosis of the elbow
seems to be a contraindication to the safe performance of arthroscopic
synovectomy.
In the present study, radiographs of the elbow showed equally extensive
deterioration following both procedures. Therefore, arthroscopic synovectomy,
as it has less surgical insult, might be more advantageous to these
patients.
The major factors that cause preoperative restriction in the range of
motion of the elbow are believed to be pain from synovitis, contracture of the
anterior capsule, and loss of joint congruence, including spur
formation, radial head deformity, and degenerative changes in the
radiocapitellar
joint19. All of our
patients had early disease (Larsen grade 2 or less) without a mechanical block
of elbow motion. In those with a preoperative arc of flexion of <90°,
arthroscopic synovectomy resulted in more improvement in elbow motion compared
with those who had undergone open synovectomy with capsulectomy, and no elbow
that had arthroscopic synovectomy required a total elbow arthroplasty to treat
ankylosis. Immediate (on the day of surgery) rehabilitation can be employed
following arthroscopic synovectomy, and this might be an important factor in
achieving the best long-term range of motion.
In conclusion, arthroscopic synovectomy for the treatment of elbows with
early rheumatoid arthritis is a reliable procedure that can effectively
alleviate pain and increase the range of motion, particularly in elbows with
preoperative stiffness. Our results suggest that one of the most favorable
indications for either arthroscopic or open synovectomy is a preoperative arc
of flexion of =90° in elbows with radiographic changes of Larsen grade
2 or less.
Tables presenting the demographic data and the clinical and radiographic
results on all study patients are available with the electronic versions of
this article, on our web site at
(go to
the article citation and click on "Supplementary Material") and on
our quarterly CD-ROM (call our subscription department, at 781-449-9780, to
order the CD-ROM). ?