In 1854, Sir James Paget described an enlarged median nerve at
the entrance of the carpal canal with associated adhesions in a
patient who was managed with amputation of the arm because of unrelenting
pain and sensory dysfunction secondary to a fracture of the distal
part of the radius1-3. Putnam4, in 1880, and Pierre Marie and Foix5, in 1913, provided accounts of the
clinical picture of carpal tunnel syndrome, and Brain et al.6, in 1947, detailed the surgical treatment
of the condition in six patients. The surgical treatment of carpal
tunnel syndrome, however, was not generally accepted until the publication
of a series of articles by Phalen et al. beginning in 19507-12.
Currently, carpal tunnel syndrome is recognized as the most common
entrapment neuropathy13, affecting an estimated 1% of the
adult population of the United States14. Carpal tunnel release is
the most commonly performed operation in the field of hand surgery
and is also one of the most successful15,16.
A study in Maine revealed this procedure to be the sixth most common
outpatient ambulatory-surgery procedure, translating to an estimated
annual volume of 400,000 carpal tunnel releases in the United States17. Idiopathic carpal tunnel syndrome
accounts for the majority of cases12,18.
Surprisingly, despite the large number of diagnosed and treated
cases of this condition, controversy still exists regarding the
pathophysiology and treatment of idiopathic carpal tunnel syndrome.
For example, in contrast to the theory that spontaneous carpal tunnel
syndrome is caused by chronic tenosynovitis of the flexor tendons,
acute and chronic inflammation rarely have been found in studies
of the flexor tenosynovium in patients with carpal tunnel
syndrome who were treated with surgical release15,16,19-24.
In addition, the benefit of flexor tenosynovectomy combined with
open carpal tunnel release for the treatment of idiopathic carpal
tunnel syndrome is also undetermined. Some authors have advocated
routine flexor tenosynovectomy while others have not25.
The purpose of the present study was threefold: (1) to assess the
role of flexor tenosynovectomy in the operative treatment of carpal
tunnel syndrome on the basis of clinical outcome measures, (2) to
assess whether the gross or histologic appearance of the flexor
tenosynovium was correlated with preoperative or postoperative symptoms
or function in order to determine whether a group of patients who
might benefit from flexor tenosynovectomy could be identified intraoperatively, and
(3) to correlate the gross appearance of the flexor tenosynovium
with the histologic appearance.
The study group consisted of eighty-eight wrists in eighty-seven
patients (fifteen men and seventy-two women) who were treated surgically
for idiopathic carpal tunnel syndrome. The mean age of the patients
was fifty-eight years (range, sixteen to ninety-seven years). The
syndrome was bilateral in twenty patients (23%). Of the
forty-two patients (48%) who worked, thirteen did manual
labor and twenty-nine did nonmanual labor. Sixteen (18%)
of the eighty-seven patients were receiving Workers’ Compensation.
The most common preoperative symptoms included pain (fifty wrists;
57%), paresthesias (sixty-four wrists; 73%), and
weakness (twenty-six wrists; 30%).
The patients were evaluated preoperatively and were followed for
a minimum of twelve months postoperatively. Evaluation consisted
of a detailed history, a physical examination, a preoperative electrophysiological
study, and completion of a validated self-administered questionnaire
for the scaled assessment of the severity of symptoms and functional
status as described by Levine et al.26.
The responses on this questionnaire are assigned a score of 1 to
5 points, with 1 point as the best score and 5 points as the worst
score. The questionnaire was administered preoperatively and at
the final follow-up visit.
The physical examination included testing for the Phalen and Tinel
signs, manual muscle-testing, determination of the presence of visible
thenar atrophy, evaluation of two-point discrimination, and Semmes-Weinstein
monofilament testing. The criteria for the diagnosis of carpal tunnel
syndrome included a history of paresthesias in the median-nerve
distribution, nocturnal hand pain, positive findings on physical examination,
and a positive electrodiagnostic study. The criteria for the electrophysiologic
diagnosis included a median distal motor latency of >4.5
ms, a median sensory nerve conduction velocity of <50 m/s
across the carpal tunnel, or a lumbrical/interosseous
peak latency difference of >0.5 ms. Electromyographic studies
of the thenar muscles were utilized to demonstrate signs of denervation
and to measure the severity of median-nerve compression. The work-up
also included standard radiographs of the hand and wrist to rule
out osseous abnormalities. Patients with a history of inflammatory
arthritis, autoimmune disorders, thyroid anomalies, renal failure, acromegaly,
or tumors including myeloma were excluded from participation. Three
patients in each group had adult-onset diabetes mellitus.
The findings on physical examination included the Phalen sign
after less than forty-five seconds of testing in fifty-one wrists
(58%), the Tinel sign in forty-eight wrists (55%),
and thenar atrophy in thirty-two wrists (36%). The average
two-point discrimination was 6.5 mm preoperatively. Of the seventy-seven
patients who had Semmes-Weinstein monofilament testing, nine (12%)
showed diminished sensitivity to light touch, thirty-one (40%)
exhibited diminished protective sensation, and thirty-seven (48%)
showed a loss of protective sensibility.
The indication for surgery was the continuation of symptoms despite
a three-month trial of conservative management that included splinting
of the wrist in neutral. A cortisone injection (a mixture
of 0.5 mL of Celestone [betamethasone] and 1.5 mL
of 1% Xylocaine [lidocaine] without epinephrine, injected
1 cm proximal to the wrist crease and ulnar to the palmaris longus
tendon without elicitation of paresthesias) was offered to patients
whose presenting symptoms were of less than six months’ duration.
Patients who exhibited thenar atrophy or denervation on electromyography
were managed with surgery as soon as possible. All patients signed
an informed-consent document, and the study was approved by the
institutional review board.
In order for the study design to achieve at least 95% power with
the use of an unpaired t test with a large effect size of 0.8 and
an alpha (a) of 0.05, it was determined that at least forty samples
were required in each group (tenosynovectomy or no tenosynovectomy).
With a total of forty-four samples per group at the conclusion of
the study, these numbers were exceeded for both the paired and unpaired
t tests. At the time of surgery, the wrists were randomized to open
carpal tunnel release with or without tenosynovectomy. A computerized randomization
program was used for this purpose.
A standard open carpal tunnel release was then performed by one
of two surgeons (M.P.R. or R.J.S.) through an approximately 3 to
4-cm longitudinal palmar incision in line with the interspace between
the middle finger and the ring finger, stopping short of the volar
flexion crease of the wrist. Local anesthesia with sedation was
used in all cases. After open release, the gross appearance of the
flexor tenosynovium of all wrists was subjectively graded as mild,
moderate, or severe on the basis of the amount of tenosynovium present,
its apparent thickness, and the degree of hyperemia. Half of the
wrists were then treated with a flexor tenosynovectomy that involved removal
of all of the visible tenosynovium that could be withdrawn into
the incision by pulling the flexor tendons into the wound, with
care taken to protect the median nerve. The tenosynovium was typically
more abundant on the flexor digitorum superficialis tendons than
on the flexor digitorum profundus tendons. The tenosynovium specimens
were sent to the pathology department for immediate fixation in
10% buffered formalin. The specimens were embedded in paraffin,
cut into 5-m-thick sections, and stained with hematoxylin and eosin.
The histological appearance of the tenosynovium was then graded
as mild, moderate, or severe on the basis of the degree of hypertrophy,
hyperplasia, hyperemia, edema, and chronic inflammation. The grading
was done by one senior staff pathologist (M.P.) who was blinded
to all clinical information. Staining with Congo red for amyloid
was not done.
Statistical analysis was performed with use of the SAS statistical
package (SAS Institute, Cary, North Carolina) and included paired
and unpaired t testing, multiple analysis of variance, and determination
of Pearson correlation coefficients.
The symptom-severity and functional assessment scores improved
both in the group treated with tenosynovectomy (p < 0.0001
and p = 0.05, respectively; paired t test) and in the group
treated without tenosynovectomy (p = 0.0002 and p = 0.0004,
respectively; paired t test). Specifically, the mean symptom-severity
score (and standard deviation) improved from 3.0 ±
0.88 to 1.6 ± 0.68 points in the group treated with
tenosynovectomy and from 2.9 ± 0.64 to 1.6 ±
0.7 points in the group treated without tenosynovectomy. Similarly,
the mean functional status score improved from 2.6 ±
0.73 to 1.7 ± 0.71 points in the group treated with
tenosynovectomy and from 2.6 ± 0.69 to 1.6 ±
0.62 points in the group treated without tenosynovectomy. Furthermore,
with the numbers available, there were no significant differences
between the groups with regard to the symptom-severity or functional status
scores at the twelve-month follow-up (p = 0.77 and p = 0.44,
respectively; unpaired t test). No wound infections occurred, and
there were no differences between the groups with regard to scar
sensitivity, wrist motion, or finger motion.
The visual analysis of the tenosynovium resulted in a grading of
mild for twelve wrists, moderate for fifty-six wrists, and severe
for twenty wrists. The histologic analysis of the removed tenosynovium
revealed a grading of mild for ten wrists, moderate for twenty-four
wrists, and severe for ten wrists. Statistical analysis revealed
no significant correlation between the gross intraoperative findings
and the preoperative and postoperative symptom-severity scores (r = 0.038
and 0.038, respectively) or the preoperative and postoperative functional
assessment scores (r = 0.25 and 0.13, respectively). Similarly,
no significant correlation was seen between the histologic findings
and the preoperative and postoperative symptom-severity scores (r = 0.032
and 0.004, respectively) or the preoperative and postoperative functional
assessment scores (r = 0.086 and 0.021, respectively).
There was no significant correlation between the gross and histologic
grades (r = 0.34), although 63% of the specimens
that were graded as severe histologically were also graded as severe
visually.
The histologic examination demonstrated an absence of acute inflammation
as indicated by the absence of a neutrophilic exudate in all specimens.
Chronic inflammation (defined by the presence of substantial numbers
of lymphocytes, macrophages, or plasma cells), fibroblastic proliferation,
and granulation tissue also were not observed.
The pathophysiology of idiopathic carpal tunnel syndrome has
been viewed as the result of either a decrease in the size of the
carpal canal or an increase in the volume of the contents within
the canal. Previous studies have demonstrated no significant differences
between patients with and without carpal tunnel syndrome with regard
to the thickness, histological characteristics, and biomechanical
behavior of the transverse carpal ligament2,24,27.
Most investigators have believed that carpal tunnel syndrome is
caused by an increase in the volume of contents within the carpal
tunnel (either the tenosynovium or the lumbrical or superficialis
muscles) or is the result of a vicious cycle set in motion by swelling
within the tunnel that is caused by either external or internal
pressure and that leads to chronic tenosynovial thickening and fibrosis.
Lluch22, in an elegant study in
which a rabbit model was used, provided intriguing evidence that
tenosynovial thickening may be a consequence rather than a cause
of carpal tunnel syndrome. In that study, rabbits in which the volume
of the carpal canal had been surgically decreased with use of a
wire loop were noted to have tenosynovial changes that were quite
similar to those found in tenosynovial specimens from patients who
had had operative treatment of carpal tunnel syndrome22.
Many investigators have examined the flexor tenosynovium with
respect to its role in increasing the volume of contents within
the carpal tunnel. A nonspecific chronic tenosynovitis had long
been thought to contribute to this syndrome28,29;
however, several histologic studies have demonstrated that inflammatory
changes are extremely rare in specimens of removed tenosynovium
and that nonspecific fibrous changes and edema are usually noted
instead15,16,19-24.
The current study not only corroborated these histologic findings
demonstrating the absence of inflammatory changes in patients with
idiopathic carpal tunnel syndrome but also demonstrated no correlation
between the gross or histologic appearance of the tenosynovium and
the preoperative or postoperative clinical findings.
The main goal of our study was to assess the role of routine flexor
tenosynovectomy in the treatment of idiopathic carpal tunnel syndrome.
The rationale for routine tenosynovectomy is that if the edematous
and potentially bulky flexor tenosynovium is contributing to increased
volume within the carpal canal with resultant symptoms, then removal
of this tenosynovium should lead to increased space for the median
nerve in the carpal tunnel and improved relief from symptoms compared
with that achieved through ligament release alone. In fact, Melvin30 recommended tenosynovectomy through
a distal forearm incision without release of the transverse carpal
ligament, as the sole surgical treatment for carpal tunnel syndrome.
Alternatively, tenosynovectomy with release of the transverse carpal
ligament has been indicated for invasive, proliferative tenosynovitis
as seen in association with rheumatoid arthritis, granulomatous
infection, amyloid deposition in patients receiving dialysis, and
crystalline-deposition diseases such as gout.
The role of tenosynovectomy in the treatment of idiopathic carpal
tunnel syndrome, however, remains unclear. Freshwater and Arons31 compared two groups of patients
with idiopathic carpal tunnel syndrome: one group was treated with
a standard carpal tunnel release alone, and the other group was
treated with carpal tunnel release, tenosynovectomy, external neurolysis
of the median nerve, and intraoperative instillation of Kenalog (triamcinolone
acetonide). Those authors found no difference between the groups
with regard to signs, symptoms, or electromyographic data at the
two-year follow-up examination. In terms of the assessment of the
role of tenosynovectomy, however, that study was limited by the
additional variables of external neurolysis and the instillation
of cortisone in the group of patients who underwent tenosynovectomy.
Some investigators32,33 have
indirectly suggested a role for tenosynovectomy during carpal tunnel
release. Wheatley and Kaul32 reported
on five patients (six hands) who required a reoperation for persistent
symptoms after endoscopic release without tenosynovectomy. Fulminant
synovitis was identified at the time of the reoperation in all patients,
and the symptoms resolved with open release and synovectomy. Bloem
et al.33 found that patients who
had so-called post-carpal tunnel syndrome after open release often
had flexor tenosynovitis during reoperation. Overall, however, the
most frequent finding associated with repeated carpal tunnel release
is incomplete release of the ligament.
In the present study, we found that flexor tenosynovectomy provided
no added benefit in terms of clinical outcome compared with sectioning
of the ligament alone. We chose the outcome instrument designed
by Levine et al.26 because it
is a validated tool specifically designed to assess patients with
carpal tunnel syndrome. The questionnaire is self-administered,
and thus the results should be independent of the observer. We used
a prospective, randomized study design. In addition to our finding
that tenosynovectomy provided no additional benefit, we found that
the gross and histologic appearance of the tenosynovium had no predictive value.
The results of the present study support the theories proposed by
Phalen9 in his landmark article
of 1966. In that article, Phalen noted that "thickening
or fibrosis of the flexor synovialis within the carpal tunnel was
the most common cause of the syndrome, being found in 203 of 212
wrists." Later in the article, however, he stated
that "routine synovectomy is not advisable." Our
findings support the position that routine tenosynovectomy is unnecessary,
although not harmful. Flexor synovectomy has long been an adjunct
to open carpal tunnel release at our institution25,
and it was our impression that it was helpful. However, our data
have demonstrated otherwise, and we no longer perform routine tenosynovectomy
during open carpal tunnel release.
Note: The authors thank Obiwanne F.C. Ugwonali, MD, for his assistance
with the statistical analysis.