Nerve palsies following total hip arthroplasty have been
reported by a number of authors1-18.
These reports have mainly described problems due to trauma during
the operative approach4,6-8,11,16,
the effects of lengthening of the lower extremity6,9,11,12,17,
compression secondary to cement protrusion5,13 or
hematoma formation7, and broken
trochanteric wires1.
All nerves about the hip, including the sciatic1-3,5-7,
obturator13, femoral8,9,11,14,17, peroneal7,8,11,17, and superior gluteal nerves4, have been involved. Most authors1,5,7,13 have described motor and sensory
deficits, with little or no mention of pain5,7-9,11,12,17.
To our knowledge, there have been no previous descriptions of cases
in which pain was the sole, or even the primary, symptom. The unique
feature of our case is the fact that pain was the only subjective problem
and there was no objective evidence of any motor or sensory deficit.
The prevalence of sciatic nerve palsy has been reported to be
13% (thirty-one of 243) after primary hip replacement,
5.2% (nine of 172) after arthroplasty for congenital hip
dysplasia, and 3.2% (twenty-two of 694) after revision
total hip arthroplasty11. Femoral
neuropathies, according to some reports, are as common as sciatic
neuropathies; the prevalence has been reported to be as high as
4.7% (seven of 150) after total hip arthroplasty and 5.5% (three
of fifty-five) after revision operations, with an overall prevalence
of 2.3%14. Palsies of
the peroneal, obturator, and superior gluteal nerves are less common.
Reduction maneuvers, retraction, and leg positioning during preparation
of the canal are the points during the operation when the sciatic
nerve is most vulnerable to injury, and the greatest changes in
evoked potentials during intraoperative monitoring have been observed at
these times2,16.
We report the case of a patient who had painful sciatic and femoral
nerve neuropathies, which started immediately postoperatively and
resolved after a revision operation was performed to match the contralateral
limb length. Clinical and radiographic examinations, performed to
compare limb lengths, revealed a 25-mm increase in length on the involved
side prior to the revision operation. The findings on motor and
sensory examinations were normal. All preoperative and postoperative
findings were documented by electromyography and nerve-conduction
studies.
A sixty-nine-year-old man who had had a left total hip arthroplasty
for osteoarthritis at another institution on October 22, 1996, presented
to us for a semi-urgent office visit three weeks postoperatively.
The patient had been hospitalized for five days, and he described
a satisfactory stay except for severe pain and muscle spasms in
the left limb, which had been present from the time of the operation.
The patient described the pain as radiating down the anterior
aspect of the thigh to the lateral part of the knee and over the
top of the foot. He was not able to sleep for more than ten minutes
at a time. A few days prior to the office visit, he had noted swelling
of the limb and had visited the emergency room of another local
hospital, where a contrast venogram was interpreted as negative
for deep-vein thrombosis. He was being treated with a postoperative
protocol of Coumadin (warfarin) for prophylaxis against deep-vein
thrombosis, and the international normalized ratio, on November
16, 1996, was 1.2. Oxycodone-acetaminophen had been prescribed at
the time of his discharge from the hospital, but ingestion of two
tablets every four hours had not relieved the pain.
Clinical examination revealed an obvious limb-length discrepancy,
with the side that had been operated on being longer. An inequality
of 25 mm was determined by measuring the differences in knee position
with the patient sitting on the examination table. There was a mild
amount of edema throughout the lower extremity; the incision had
a trace of erythema, but it was well healed. The patient had a strong
dorsalis pedis pulse and no increase in symptoms on the straight-leg-raising test.
The findings on motor and sensory examinations were normal.
Radiographic examination revealed a well-positioned cementless
acetabular component with a single screw for added fixation and
a cementless porous-coated femoral component (Fig. 1). The distance
from a horizontal line drawn between the inferior aspect of the
left and right ischia to the tip of the lesser trochanter showed
an increase of 28 mm on the involved side. The preoperative radiographs
were not available to us; however, the patient believed that the
limb lengths had been equal prior to the hip replacement.
The results of electrodiagnostic testing were consistent with
both chronic and acute neuropathic changes of the left sciatic nerve.
There was also evidence of active and chronic neuropathic changes in
the femoral nerves, with the changes on the left, involved side
being much worse than those on the right. Lumbar paraspinal electromyography
demonstrated increased insertional activity at the third and fourth
lumbar levels on the left and at the fourth and fifth levels on
the right. After the suspected etiology of the symptoms was discussed with
the patient, he agreed to a revision of the femoral component in
the hope of correcting the limb-length discrepancy and eliminating
the presumably neurogenic pain.
Eight weeks after the primary replacement, the femoral component
was revised with use of a long-stem cementless implant, which was
more deeply seated within the femoral canal. A trochanteric advancement
was carried out simultaneously to restore abductor muscle tension
and stability. The trochanteric fragment was fixed with a Dall-Miles claw-and-cable
system (Stryker-Howmedica-Osteonics, East Rutherford, New Jersey).
Measurements made on the postoperative radiographs, as described
above, showed the length of the affected limb to be within 4 mm
of that of the contralateral limb (Fig. 2).
Postoperatively, the patient did extremely well. The pain and
muscle spasms in the left lower extremity resolved completely within
the acute postoperative period. Repeat electrodiagnostic
studies, performed nine months after the revision operation, showed spontaneous
resolution of the acute changes in the left sciatic nerve since
the time of the previous study, with no additional active denervation.
There also was evidence on electromyographic and nerve-conduction
studies that the left femoral nerve had marked interval improvement,
with normal spontaneous activity and some mild chronic motor axonopathic
changes.
At the time of the last follow-up examination, twenty-eight months
after the revision operation, there had been no recurrence of pain,
spasms, or burning in the left lower extremity.
Nerve palsies after total hip arthroplasty may be associated
with mechanical changes of the environment surrounding the affected
nerve, whether the cause is a mass effect1,13 or
overlengthening of the lower extremity6,9,11,12,17.
When a posterior approach is used for the hip exposure, care should
be taken, when the external rotator muscles are retracted, to prevent
direct trauma to the nerve. During an anterolateral approach, similar
caution should be exercised to protect the femoral nerve.
There have been several reports3,5,6,9 documenting
ranges of lengthening of the lower extremity that are safe with
regard to the sciatic nerve. An intraoperative method for measuring
the change in length with the trial components in place is helpful
in determining limb length. The technique of the senior author (K.A.K.)
involves placement of a unicortical large-fragment screw above the
superior rim of the acetabulum. The screwdriver is placed in the
hex head of the screw, and the distance from the shaft of the screwdriver
to a mark made with the cautery at the vastus tubercle on the lateral
aspect of the greater trochanter is then measured. The initial length,
before the femoral head is dislocated and the femoral-neck cut is
made, is recorded and is compared with the measurement made after
the trial reduction. After implantation of the prosthetic femoral
component, a final check of the limb length is made to ensure proper
seating of the component in the proximal aspect of the femur. This
technique has been useful in documenting the change in the length
of the lower extremity in the operating room during total hip replacement.
The acute neuropathy in our patient seems to have been related
to the approximately 25 mm of limb-lengthening. To our knowledge,
this complication has not been reported previously. No signs of
nerve palsy, such as motor weakness or loss of sensation, were evident,
but electromyography and nerve-conduction studies clearly showed
local alteration of nerve function. This suggests that the pain
was neurogenic and presumably was due to traction on the femoral
and sciatic nerves caused by overlengthening. When the patient reported
that there was no history of diabetes or alcohol abuse, the chronic changes
noted on the original electrodiagnostic studies also suggested the
possibility of an underlying peripheral neuropathy. With an underlying chronic
inflammatory process of the sciatic and femoral nerves, an acute
traction event may have resulted in enough microvascular injury
to the nerve to have caused a painful neuritis as opposed to a loss
of motor and sensory function. The increased lumbar paraspinous
insertional activity, noted at the third and fourth lumbar levels
on the left and at the fourth and fifth levels on the right, suggests
an underlying problem in the lumbar spine, such as spinal stenosis,
a condition that could lead to a so-called double-crush syndrome
and that might explain the severe symptoms associated with the limb-lengthening.
This report highlights the need for concern about limb-length
changes associated with total hip arthroplasty as well as the need
for a reliable intraoperative technique for limb-length measurement.