Abstract
Background: Pyogenic flexor tenosynovitis is a closed space
infection involving the digital flexor tendon sheaths of the upper extremity
that can cause considerable morbidity. The purpose of the present report is to
describe the various risk factors leading to poor outcomes and to recommend a
clinical classification system for this condition.
Methods: We studied seventy-five patients with pyogenic flexor
tenosynovitis over a six-year period. The amputation rate and total active
motion were used as outcomes measures. The clinical factors influencing
outcomes were identified and analyzed.
Results: The five risk factors associated with poor outcomes were
(1) an age of more than forty-three years, (2) the presence of diabetes
mellitus, peripheral vascular disease, or renal failure, (3) the presence of
subcutaneous purulence, (4) digital ischemia, and (5) polymicrobial infection.
On the basis of the clinical findings and outcomes, three distinct groups of
patients could be identified, each with a progressively worse outcome.
Patients in Group I had no subcutaneous purulence or digital ischemia; these
patients had the best prognosis, with no amputations and a mean 80% return of
total active motion. Patients in Group II demonstrated the presence of
subcutaneous purulence but no ischemic changes; these patients had an
amputation rate of 8% and a mean 72% recovery of total active motion. Patients
in Group III had both extensive subcutaneous purulence and ischemic changes;
these patients had the worst prognosis, with an amputation rate of 59% and a
mean 49% return of total active motion.
Conclusions: We propose a three-tier clinical classification system
that can aid in prognosis and guidance in the treatment of pyogenic flexor
tenosynovitis of the upper extremity.
Level of Evidence: Prognostic Level II. See Instructions
to Authors for a complete description of levels of evidence.
Pyogenic flexor tenosynovitis is an aggressive closed-space infection of
the digital flexor tendon sheaths of the upper extremity that can cause severe
disability if not treated expediently and
appropriately1.
Kanavel classified it among the grave infections of the
hand2. The
prevalence of this infection has ranged from
2.5%3 to
9.4%4 in large
series of hand infections. Stern et
al.5 reported a high
rate (38%) of complications, including stiffness, persistent infection,
boutonniere deformity, and amputation. However, some authors have reported
good results with few complications in the treatment of this
condition6,7.
Michon8
classified flexor tenosynovitis into three stages. In the first stage, the
inflamed synovial sheath is distended with serous exudative fluid. In the
second stage, purulent fluid causes progressive distension of the tendon
sheath. The tendon is still viable. In the third stage, there is septic
necrosis of the tendon and the pulleys. The prognosis worsens with increasing
stage. Michon proposed that stages I and II could be treated with limited
incision with drainage and irrigation of the sheath, whereas stage III should
be treated with open débridement. Michon believed that the progression
of the disease is influenced by the virulence of the bacteria and the
patient's response to treatment. However, he did not identify any clinical
indicators of severity or preoperative factors predicting poor outcome. The
present study was designed to identify the preoperative risk factors
associated with poor outcomes and to propose a clinical classification system
for flexor tenosynovitis based on these factors.
With the approval of institutional authority, we reviewed all of the
operative records in our department's surgical logbook from 2001 to 2006. In
addition, we generated a list of patients with the diagnosis of suppurative
tenosynovitis (International Classification of Diseases [ICD], Ninth Revision,
code 7270) from the Medical Records Office. Seventy-five consecutive records
were found, and verbal consent was obtained from these patients for the use of
their medical reports. The diagnosis of pyogenic flexor tenosynovitis was made
clinically on the basis of the presence of one or more of the Kanavel
signs2 of fusiform
swelling, a semiflexed posture of the digit, pain on passive extension, and
tenderness along the tendon sheath. Where there was subcutaneous purulence, it
could often be detected as an area of boggy swelling. Occasionally, it could
present with purulent drainage from a preexisting puncture wound or an area of
skin necrosis. The presence of ischemic changes, including blistering, skin
necrosis, and delayed capillary refill, was similarly noted.
The study group included forty-seven male and twenty-eight female patients
with a mean age of forty-seven years (range, seventeen to seventy-four years).
All infections were community-acquired infections. The dominant hand was more
commonly involved (fifty patients; 67%). The middle finger was most commonly
involved (thirty-two patients; 43%), followed by the index finger (sixteen
patients; 21%) and the ring finger (thirteen patients; 17%). The average
number of surgical débridements was 1.5 (range, one to five), and the
mean time-interval to wound-healing and closure after the débridement
was five days (range, two to seventeen days). The mean duration of
hospitalization was nine days (range, five to fifteen days), and the mean
duration of follow-up was twenty-six months (range, fourteen to thirty-six
months).
We agree with Stern et
al.5 that the
complications of pyogenic flexor tenosynovitis include amputation and
stiffness of the surviving digits. Therefore, we specifically studied the
amputation rate and total active
motion9 as measures
of outcome related to the severity of the disease process at the time of
presentation. The total active motion for each digit was calculated by
modifying the formula used by Strickland and Glogovac for the grading of
digital
function10.
The total active motion for each finger was calculated as the sum of the
active ranges of motion of the metacarpophalangeal (MCP) joint, the proximal
interphalangeal (PIP) joint, and the distal interphalangeal (DIP) joint,
divided by 265°. The percentage of a normal 265° of motion at these
three joints (total active motion) was determined with use of the formula:
(active MCP + PIP + DIP flexion)/265° × 100%.
For the thumb, total active motion was calculated as the active range of
motion in the interphalangeal (IP) and metacarpophalangeal (MCP) joints,
divided by 140°. The percentage of a normal 140° of motion at these
two joints (total active motion) was determined with use of the formula:
(active MCP + IP)/140° × 100%.
The possible risk factors that were analyzed included age, comorbidities,
the presence of subcutaneous purulence, ischemic changes, and the number of
bacterial types that grew on culture. The patients were then classified into
three distinct groups according to the clinical findings and the rates of
amputation. We further analyzed the percentage recovery of total active motion
in each group.
Statistical analysis was performed with SPSS statistical software (version
11.0; SPSS, Chicago, Illinois). The box-plots demonstrating the age
distributions within the groups of patients with and without amputation were
analyzed. The age cut-off was determined by performing univariate analysis for
each age-group. Univariate analysis was performed with use of the chi-square
or Fisher exact test for comparison of proportions between two categorical
data. The Mann-Whitney U test was used to compare the nonparametric data
between two independent samples. The level of significance was set at p <
0.05.
Surgical Treatment of Infections
All operations were performed with the patient under general anesthesia
with use of tourniquet control, except in the cases of three patients who had
ipsilateral arteriovenous fistulae. In early stages of the disease, when the
infection was localized to the tendon sheath, we used the method of limited
incision with drainage and irrigation as described by
Neviaser6. In cases
in which there was concomitant subcutaneous purulence, we used the zigzag
(Bruner)
incision11.
Although many surgical approaches have been
described12,13
including the midaxial and serial transverse incisions, the use of a zigzag
incision allowed for better exposure, visualization of the tendon sheath in
its entirety, and proximal extension to assess the full extent of the
infection.
We practice excisional débridement in our institution. If the carpal
tunnel was involved, we performed a carpal tunnel release and débrided
the synovium. The wound was then irrigated with a copious amount of saline
solution and was left open with tulle gras dressing (Smith and Nephew, London,
England). The tulle gras dressing is useful as a form of continuous open
drainage. A bulky dressing was applied with the hand immobilized in the
functional position. Postoperatively, the hand was elevated, and daily
chlorhexidine irrigations and dressing changes were performed. On the third
postoperative day, a decision was made for either further débridement
or secondary closure.
Amputation was performed for thirteen (17%) of the seventy-five patients.
Amputation was performed at a mean of nine days (range, five to fourteen days)
after admission and a mean of three days (range, two to six days) after the
initial débridement. The mean percentage recovery of total active
motion of the surviving digits in the remaining sixty-two patients was 73%
(range, 4% to 100%). Table I
summarizes the demographic characteristics of the patients, their clinical
presentation, and the treatment that they received.
Time to Definitive Treatment
The mean interval from the onset of symptoms to the commencement of
intravenous antibiotics was three days (range, one to seventeen days), and the
mean interval from the onset of symptoms to surgery was six days (range, three
to twenty-one days). However, with the numbers available, there was no
significant correlation between the interval to definitive treatment and the
rate of amputation (p > 0.05).
Comorbidities
The rate of amputation among patients who were more than forty-three years
old was 24.4% (eleven of forty-five). Patients who were more than forty-three
years old had a 4.5 times higher risk of amputation (odds ratio, 4.5; 95%
confidence interval, 0.9 to 22.2; p = 0.041) than patients who were
forty-three years old or younger.
Thirty-nine patients (52%) had one or more of the medical comorbidities of
diabetes mellitus, peripheral vascular disease, and end-stage renal failure
requiring dialysis. The average number of comorbidities was 1.4 (range, zero
to three). The diabetic patients had had insulin-dependent diabetes mellitus
for more than five years and an elevated glycosylated hemoglobin (Hba1c) level
of >6.4% (the upper limit of the normal range for the Hba1c level at our
institution) at the time of admission. The comorbidities that were associated
with a higher risk of amputation included diabetes mellitus (rate of
amputation, 39% [ten of twenty-six patients]; p = 0.003), peripheral vascular
disease (rate of amputation, 71% [five of seven patients]; p = 0.003), and
renal failure (rate of amputation, 64% [nine of fourteen patients]; p =
0.002). We also analyzed other common comorbidities of hypertension,
specifically, hyperlipidemia and previous strokes, and found that, with the
numbers available, these factors were not significantly associated with a
higher rate of amputation (p > 0.05).
Clinical Findings
Thirteen patients (17%) had fever. None had systemic signs of infection
such as hypotension or tachycardia. The most common Kanavel sign was fusiform
swelling of the digit, which was found in 97% of the patients. This was
followed by pain on passive extension (noted in 72% of the patients) and a
semiflexed posture of the digit (noted in 69% of the patients). Tenderness
along the flexor sheath was a somewhat less common finding (noted in 64% of
the patients).
Thirteen patients (17%) underwent digital amputation. Subcutaneous
purulence was detected in fifty-one patients (68%). The amputation rate in
this group of patients was 24% (twelve of fifty-one), and these patients had a
6.8 times higher risk of amputation (odds ratio, 6.8; 95% confidence interval,
0.8 to 55.6; p = 0.039). Seventeen patients (23%) demonstrated signs of
ischemia, and these patients had a 25.6 times higher risk of amputation (odds
ratio, 25.6; 95% confidence interval, 5.7 to 117.6; p = 0.002).
Microbiological Findings
Table II presents the
microbiological findings associated with pyogenic flexor tenosynovitis based
on tissue specimens obtained during surgical débridement. One bacterial
species grew on culture of specimens from forty-three patients (57%); two
species grew on culture of specimens from fourteen patients (19%); and three
species grew on culture of specimens from one patient (1%). Bacterial cultures
were negative for seventeen patients (23%). Patients for whom cultures were
positive for two or more bacterial species had a 5.0 times higher risk of
amputation (odds ratio, 5.0; 95% confidence interval, 1.4 to 18.5; p = 0.035).
The most common organism was Staphylococcus aureus, which grew on
culture of specimens from thirty-two patients (43%). However, with the numbers
studied, no specific species or combination was associated with a higher risk
of amputation.
Trauma
Fifty-seven patients (76%) reported antecedent injuries involving the
fingers. The most common injury was a penetrating wound (prevalence, 61%;
forty-six of seventy-five patients) followed by blunt trauma (prevalence, 11%;
eight of seventy-five patients). Eighteen patients (24%) could not recollect
any injury. With the numbers available, neither a history of trauma nor the
type of trauma predicted severity or amputation.
Risk Factors
Univariate analysis showed that five factors were significantly associated
with an increased rate of amputation (p < 0.05). These included (1) an age
of more than forty-three years, (2) the presence of diabetes mellitus,
peripheral vascular disease, or renal failure, (3) the presence of
subcutaneous purulence, (4) ischemic changes at the time of presentation, and
(5) involvement of more than one bacterial type
(Table III).
Clinical Classification and Analysis
The patients were classified into three groups on the basis of the clinical
severity of pyogenic flexor tenosynovitis at the time of presentation
(Table IV). The patients in
Group I presented with Kanavel signs of tenosynovitis but no evidence of
subcutaneous purulence or ischemia (Fig.
1). The patients in Group II had concurrent localized subcutaneous
purulence but no ischemia (Fig.
2). The patients in Group III had concurrent extensive
subcutaneous purulence involving more than one phalangeal segment or spreading
circumferentially as well as signs of ischemia
(Fig. 3). The rate of
amputation was 0% in Group I, 8% (three of thirty-seven) in Group II, and 59%
(ten of seventeen) in Group III (p = 0.002). Among the surviving digits in
each group, the average recovery of total active motion was 80% (range, 34% to
100%) in Group I, 72% (range, 4% to 100%) in Group II, and 49% (range, 8 % to
68%) in Group III.
There have been numerous reports on the treatment of pyogenic flexor
tenosynovitis. Apart from the report by Michon, we are not aware of any
attempt to classify this condition according to severity and prognosis. The
present study of seventy-five consecutive patients allowed us to statistically
analyze possible risk factors for poor outcomes, particularly amputation and
stiffness. On the basis of our analyses, we were able to classify patients
with flexor tenosynovitis into three groups with distinct clinical features
and outcomes. Patients with more risk factors had more severe disease and a
greater risk of poor outcomes.
In this group of patients, amputation was performed for those who had
digital ischemia and residual infection despite the initial
débridement. The likelihoods of amputation and recovery of total active
motion were influenced by five factors: (1) an age of more than forty-three
years, (2) the presence of diabetes mellitus, peripheral vascular disease, or
renal failure, (3) the presence of subcutaneous purulence, (4) ischemic
changes, and (5) the growth of more than one bacterial species on culture of
specimens obtained at the time of surgery. However, we acknowledge that we
could not control for other variables, such as smoking and preexisting
stiffness resulting from arthritis, which might have affected the outcome.
The presence of diabetes mellitus, peripheral vascular disease, or renal
failure led to a higher risk of amputation. This finding is consistent with
the findings of Mann and
Peacock14, who
found high rates of amputation in diabetic patients with hand infection. It
also reaffirms the findings of Kour et
al.15, who noted an
increased risk of amputation in patients with hand infection who had both
diabetes and renal failure concurrently. We postulate that a combination of
immunosuppression16,17
and vascular disease leading to poor tissue perfusion contributed to the poor
prognosis in these patients.
The patients in Group I had no subcutaneous purulence or signs of ischemia.
They had good outcomes, with no amputations and with an average functional
return of 80% of total active motion of the affected digit.
Kanavel2 described
the four cardinal signs of acute flexor tenosynovitis, but not all of these
signs need to be present in the early course of the disease.
Kanavel2,
Pollen18, and Boles
and Schmidt19
believed that excessive tenderness along the tendon sheath was the most
reliable and reproducible sign.
Neviaser6 believed
that pain on passive extension is the most valuable sign and that it is the
only sign that presents early in the process. We noted that pain on passive
extension was a specific finding that occurred early in the infection and that
tenderness along the tendon sheath was a late sign of proximal extension.
The patients in Group II demonstrated the presence of subcutaneous
purulence. These patients had a poorer prognosis, with an 8% rate of
amputation and with an average recovery of 72% of total active motion.
Concurrent subcutaneous purulence was found in 68% (fifty-one) of the
seventy-five patients. There are two possible explanations for the presence of
subcutaneous purulence in patients with flexor tenosynovitis. As the disease
progresses, the pus distends the flexor sheath, which may rupture and
introduce infection into the subcutaneous plane. Subcutaneous purulence could
also be caused by the inoculation of the subcutaneous plane by penetrating
injuries, which was the mechanism of injury in 61% of our patients. Concurrent
subcutaneous purulence requires more extensive débridement, with an
increased potential for delayed resolution and mobilization, adhesion
formation, or devascularization of the digit.
The patients in Group III demonstrated the presence of extensive
subcutaneous purulence and ischemic changes. These patients had the worst
prognosis, with a 59% rate of amputation and only 49% recovery of total active
motion. The rapid accumulation of pus causes increased tissue compartment
pressure, which compromises the blood supply to the skin, causing blistering,
necrosis, and gangrene. Tissue ischemia and necrosis also decrease antibiotic
penetration, resulting in slow resolution of hand
infections4. The
combination of extensive débridement, delayed resolution, and tendon
ischemic damage, as in Michon stage III, results in poor function of any
surviving digits.
The presence of two or more pathogens resulted in a more serious infection
and was associated with an increased rate of amputation. With the numbers
available, the type of bacteria did not affect the final outcome.
Grinnell20 noted
that the prevalence of tendon slough was higher in cases of mixed infections.
Other reports have described mixed organisms especially in immunocompromised
patients such as intravenous drug abusers, patients with diabetes mellitus,
and patients taking
steroids1,3,21,22.
In the present study, diabetic and nondiabetic patients had similar
bacteriological findings. We did not find an increase in gram-negative
bacterial infection in this group as reported by Kour et
al.15.
This proposed classification system is based on preoperative clinical
assessment, and we believe that it can guide treatment and prognosis. We
propose early sheath irrigation as described by
Neviaser6 for
patients in Group I. In addition, we recommend open débridement for
patients in Group II because they have concomitant subcutaneous purulence that
will require more extensive débridement. Patients in Group III have the
worst prognosis. If the infection does not respond to the initial
débridement, amputation may be considered to shorten the period of
morbidity, especially when there are signs of digital ischemia and necrosis. A
future prospective study applying this treatment algorithm is needed to affirm
the validity of these recommendations. ?
Neviaser RJ. Infections. In: Green DP,
editor. Operative hand surgery. 2nd ed. New York: Churchill
Livingstone; 1988. p
1039-44.1039
1988
Kanavel AB. The treatment of
acute suppurative tenosynovitis—discussion of technique. In:
Infections of the hand; a guide to the surgical treatment of acute and chronic
suppurative processes in the fingers, hand, and forearm. 5th ed. Philadelphia:
Lea and Febiger; 1925. p
985-90.985
1925
Weinzweig N, Gonzalez M. Surgical
infections of the hand and upper extremity: a county hospital experience.
Ann Plast Surg. 2002;49:
621-7.49621
2002
[PubMed][CrossRef]
Glass KD. Factors related to the
resolution of treated hand infections. J Hand Surg [Am].
1982;7:
388-94.7388
1982
[PubMed]
Stern PJ, Staneck JL, McDonough JJ,
Neale HW, Tyler G. Established hand infections: a controlled, prospective
study. J Hand Surg [Am].
1983;8:
553-9.8553
1983
[PubMed]
Neviaser RJ. Closed tendon sheath
irrigation for pyogenic flexor tenosynovitis. J Hand Surg [Am].
1978;3:
462-6.3462
1978
[PubMed]
Juliano PJ, Eglseder WA. Limited
open-tendon-sheath irrigation in the treatment of pyogenic flexor
tenosynovitis. Orthop Rev.
1991;20:
1065-9.201065
1991
[PubMed]
Michon J. [Phlegmon of the tendon
sheaths]. Ann Chir.
1974;28: 277-80.
French.28277
1974
[PubMed]
American Society for Surgery of the
Hand. The hand, examination and diagnosis. 2nd
ed. New York: Churchill Livingstone; 1983.
1983
Strickland JW, Glogovac SV. Digital
function following flexor tendon repair in Zone II: a comparison of
immobilization and controlled passive motion techniques. J Hand Surg
[Am]. 1980;5:
537-43.5537
1980
Bruner JR. The zig-zag volar-digital
incision for flexor-tendon surgery. Plast Reconstr Surg.
1967;40:
571-4.40571
1967
[PubMed][CrossRef]
Green DP, editor. Operative hand
surgery. 2nd ed. Vol 3. New York: Churchill
Livingstone; 1988.3
1988
Tubiana R, McCollough CJ, Masquelet AC.
An atlas of surgical exposures of the upper extremity. London:
Martin Dunitz; 1990.
1990
Mann RJ, Peacock JM. Hand infections in
patients with diabetes mellitus. J Trauma.
1977;17:
376-80.17376
1977
[PubMed][CrossRef]
Kour AK, Looi KP, Phone MH, Pho RW. Hand
infections in patients with diabetes. Clin Orthop Relat Res.
1996;331:
238-44.331238
1996
[PubMed][CrossRef]
McMahon MM, Bistrian BR. Host defenses
and susceptibility to infection in patients with diabetes mellitus.
Infect Dis Clin North Am.
1995;9:
1-9.91
1995
[PubMed]
Bertoni AG, Saydah S, Brancati FL.
Diabetes and the risk of infection-related mortality in the U.S.
Diabetes Care. 2001;24:
1044-9.241044
2001
[PubMed][CrossRef]
Pollen AG. Acute infection of the tendon
sheaths. Hand. 1974;6:
21-5.621
1974
[PubMed][CrossRef]
Boles SD, Schmidt CC. Pyogenic flexor
tenosynovitis. Hand Clin.
1998;14:
567-78.14567
1998
[PubMed]
Grinnell RS. Acute suppurative
tenosynovitis of the flexor tendon sheaths of the hand. A review of one
hundred and twenty-five cases. Ann Surg.
1937;105:
97-119.10597
1937
[PubMed][CrossRef]
Eaton RG, Butsch DP. Antibiotic
guidelines for hand infections. Surg Gynecol Obstet.
1970;130:
119-22.130119
1970
[PubMed]
Spiegel JD, Szabo RM. A protocol for the
treatment of severe infections of the hand. J Hand Surg [Am].
1988;13:
254-9.13254
1988
[PubMed][CrossRef]