Afifty-eight-year-old man with a history of heavy alcohol consumption but
no other major medical problems that would suggest a lack of immunocompetence
presented with progressive pain in the right hip of two years' duration.
Radiographs demonstrated
Ficat-Arlet3
Stage-IV osteonecrosis of the hip. Total hip arthroplasty was recommended to
the patient, but he declined to undergo that procedure. Instead, he chose to
receive an intra-articular corticosteroid injection as a temporizing measure
for symptomatic relief, and he was referred to a physiatrist who routinely
performed intra-articular hip injections. The injection was performed in a
surgery center under aseptic conditions and fluoroscopic guidance. Iodinated
contrast medium was used to confirm needle placement in the joint prior to
injection of the corticosteroid. The patient had a decrease in the symptoms
for a few weeks, but within two months the pain had returned, he used a cane
full-time, and he had difficulty working. Still undecided about undergoing
total hip arthroplasty, he returned to his orthopaedic surgeon and requested
an alternative temporizing solution. He received an injection of the
viscosupplement hylan G-F 20 (Synvisc; Biomatrix, Ridgefield, New Jersey) in
the standard 2-mL dose commonly prescribed for knee osteoarthritis. Again, the
injection was performed under aseptic conditions by the same physiatrist,
using fluoroscopic guidance. This time the patient did not experience any
relief, and within two to three weeks the symptoms had worsened so that he
could no longer work or walk. No more injections were given. The pain
continued to worsen, severe swelling developed in the right lower extremity,
and the patient had low-grade fevers and night sweats. Five weeks after the
hylan G-F 20 injection, radiographs and a computed tomography scan (Figs.
1-A,
1-B, and 1-C) demonstrated
massive destruction of the hip joint and an abscess of the joint tracking
superiorly and enveloping the right hemipelvis. The patient was referred to
our institution for further treatment. He complained of severe pain and had
erythema and swelling over the anterior aspect of the hip. Attempts at passive
hip motion caused substantial pain.
Aspiration of the right hip yielded gross purulence. We explored the hip
through an anterior approach that extended over the anterior iliac crest to
gain access to the iliacus abscess superiorly. More than 2 L of pus was
drained; the organism was identified as Peptostreptococcus. The necrotic
femoral head and acetabulum were débrided. The hip joint communicated
directly with the iliacus abscess. Drains were placed into the iliacus cavity
and the hip joint, and the wound was closed. Treatment with culture-specific
intravenous antibiotics was begun.
The patient was discharged to a skilled nursing facility but returned one
week later with a wound dehiscence. He was treated with resection of the
infected remnant of the femoral head and neck and additional
débridement of the acetabulum. After reaming of the acetabulum and
proximal part of the femur, an antibiotic-impregnated cement spacer was placed
(Fig. 2) and the anterior and
posterior wounds were once again closed over drains. Enteric gram-negative
rods and gram-positive cocci were grown on culture of intraoperative
specimens. Repeat incision and drainage procedures were performed, and the
anterior incision was packed open with iodine-soaked sponges. A vacuum-suction
sponge device was then utilized to bring the wound edges together, but a
split-thickness skin graft was ultimately required for definitive coverage of
the wound.
The patient was discharged to a skilled nursing facility while still being
treated with organism-specific intravenous antibiotics. He returned eight
weeks later, at which time there was no clinical or laboratory evidence of
persistent infection; the patient underwent removal of the
antibiotic-impregnated cement spacer and conversion to a total hip
arthroplasty with uncemented components. There were no more complications. Six
months after the arthroplasty, the patient had no hip pain or clinical
evidence of infection. He used a cane when walking long distances.
In 1997, the United States Food and Drug Administration approved two
viscosupplements, Synvisc (Biomatrix, Ridgefield, New Jersey) and Hyalgan
(Sanofi, New York, NY) as medical devices for use in patients with
osteoarthritis of the
knee4. In 2001,
Supartz (Smith and Nephew Orthopaedics, Memphis, Tennessee) was similarly
approved4.
Viscosupplementation has been used for the treatment of hip osteoarthritis.
Peyron and Balazs5
reported that nineteen of twenty patients responded to a single injection of
20 mg (2 mL) of Healon (Pharmacia, Uppsala, Sweden). Bragantini and
Molinaroli6 treated
fifty hips with weekly injections of Hyalgan and found that the pain had
decreased and the range of motion was improved in 68% after three months. In a
study of fifty-three patients with osteoarthritis of the hip treated with
Synvisc, Chevalier et
al.7 reported that
50% were pain-free at one month, as assessed with use of a visual analogue
pain scale. Lanzotti et
al.8 reported that
all of thirty patients who had received five Hyalgan injections at weekly
intervals had improved scores for pain on a visual analogue scale. Brocq et
al.9 found that
thirteen of twenty-two patients had a 50% improvement in the Lequesne
index10 three
months after one or two injections in the hip. We found no published reports
that supported the use of viscosupplementation for treatment of osteonecrosis
of the hip.
A review of the use of viscosupplementation with hylan G-F 20 for patients
with osteoarthritis of the knee revealed that adverse events had occurred
after 2.7% of 1537 injections and in 8.3% of 336
patients11. The
most common adverse events were local reactions consisting of pain, warmth,
and swelling. Gout, pseudogout, and granulomatous synovitis have also been
reported. In contrast, studies of the use of viscosupplementation for
osteoarthritis of the hip have shown a rate of adverse events of 10% to
12%7,9.
Septic arthritis after hyaluronan or hylan injections appears to be very rare;
there were no reported cases of sepsis in the two largest series published to
date11,12.
A low prevalence of septic arthritis has been reported after corticosteroid
injection. The usual organism is Staphylococcus aureus.
Hollander13
reported an infection rate of 0.072% in a series of 250,000 injections in 8000
patients. Gray et
al.14 reported a
rate of 0.001%. In our review of the literature, we found a single reported
case of septic arthritis of the hip after multiple injections of hyaluronate
and
glucocorticoids15.
We report a case of septic arthritis and an intrapelvic abscess following a
single intra-articular injection of Synvisc to treat degenerative arthritis
associated with end-stage osteonecrosis of the hip even though the injection
was performed by an experienced physician using proper technique and
meticulous aseptic
precautions16.
Hylan G-F 20 may resorb relatively slowly after injection and act as a
foreign body, causing either a local inflammatory reaction or an infection if
bacterial contamination occurs. Although the benefits of viscosupplementation
have been well documented, adverse effects have also been reported. The
present case suggests that, if an inflammatory reaction occurs after
viscosupplementation injection, infection should be considered a possible
cause.