A seventeen-year-old boy was involved as an unrestrained, front-seat
passenger in a high-energy, head-on motor-vehicle collision. Paramedics found
the boy trapped under the dashboard and required more than thirty minutes to
extricate him. On arrival at our institution, the patient was tachypneic
(respiratory rate, 33 breaths/min), tachycardic (heart rate, 105 beats/min),
and hypotensive (blood pressure, 76 mm Hg) but was alert and oriented and had
a Glasgow Coma Scale score of 15. Radiographs revealed multiple long-bone
fractures, including a type-I open comminuted fracture of the right femoral
shaft (Fig. 1-A), according to
the system of Gustilo and
Anderson14, and
closed diaphyseal fractures of the right tibia and fibula, right humerus, and
left femur. Computed tomography demonstrated hemoperitoneum, an adrenal
hematoma, and hepatic and renal lacerations; there were no pelvic or
acetabular fractures and there was no evidence of hip dislocation or
subluxation. Both lower extremities had diminished pulses, but the right lower
extremity was cooler and had delayed capillary refill when compared with the
left. Both lower extremities had palpable femoral and nonpalpable popliteal,
dorsalis pedis, and posterior tibial pulses. The latter pulses all were weakly
perceptible on Doppler ultrasound examination.
The patient was taken urgently to the operating room, where he underwent
irrigation and débridement and application of an external fixator for
the open right femoral fracture, closed reduction and external fixation of the
right tibial and left femoral fractures, and closed reduction and splinting of
the right humeral fracture. Given the hemodynamic status of the patient and
the uncertainty regarding perfusion of the lower extremities, the
lower-extremity fractures were stabilized initially with external fixation to
minimize blood loss and operative time. The patient was given four units of
packed red blood cells, as well as 3300 mL of crystalloid solution to maintain
hemodynamic stability in the operating room. Angiography was performed within
twelve hours after stabilization of the long-bone fractures to evaluate
persistently diminished pulses in both lower extremities. The catheter was
inserted into the right common femoral artery, and the guidewire was
manipulated into both external iliac arteries. There were no known
complications from the procedure, and the angiogram revealed no evidence of
vascular injury (Fig. 1-B). The
patient was returned to the operating room forty-eight hours later for repeat
irrigation and débridement of the open right femoral shaft fracture,
open reduction and internal fixation of the right humerus, and retrograde
intramedullary nailing of the left femur. Two days later, a third irrigation
and débridement of the right femur was performed, followed by
retrograde intramedullary nailing of the right femur and antegrade
intramedullary nailing of the right tibia.
The right femoral shaft fracture was stabilized through a retrograde
approach. After the performance of a medial para-patellar arthrotomy, a
guide-pin was placed in the distal aspect of the femur under fluoroscopic
guidance and a reamer was used to establish the starting hole at the anterior
edge of the intercondylar notch. The fracture was reduced, and a beaded
guidewire was placed across the fracture site. The external fixator was
removed, and the femoral canal was reamed sequentially to 11.5 mm in diameter
to facilitate passage of the nail. A titanium nail, 400 mm long and 10 mm in
diameter, was inserted manually without difficulty. Because the canal was
reamed 1.5 mm wider than the diameter of the nail, insertion did not
necessitate striking the nail with a mallet. The nail was cross-locked
proximally and distally (Fig.
1-C).
Postoperatively, the patient remained non-weight-bearing on either lower
extremity for a period of six weeks. Weight-bearing status was advanced when
radiographic and clinical evidence of healing was demonstrated. The patient
had a good range of motion of the right knee and was walking without pain and
without the aid of crutches ten weeks postoperatively. A hip examination at
that point elicited no pain during full active and passive range of
motion.
Seven months after the operation, radiographs of the right femoral
diaphysis showed continued fracture-healing. Radiographs of the hip, however,
demonstrated changes in the femoral head that were consistent with
osteonecrosis (Fig. 1-D).
Although the patient remained asymptomatic, the diagnosis of osteonecrosis of
the femoral head was confirmed on magnetic resonance imaging
(Fig. 1-E). Sixteen months
after the injury, the right hip remained pain-free; however, radiographs
showed further progression of osteonecrosis
(Fig. 1-F) and physical
examination revealed a decrease in the range of internal rotation. The patient
was evaluated to determine whether he should be treated with a vascularized
free fibular graft to manage the osteonecrosis; however, because he was still
relatively asymptomatic, he opted for a period of close observation.
We report a case of osteonecrosis of the femoral head following the
use of retrograde intramedullary nailing to stabilize a femoral shaft fracture
in a skeletally mature adolescent with multiple injuries. To our knowledge,
this finding has not been described in the literature. Osteonecrosis following
antegrade nailing of the femur in pediatric and adolescent patients has been
described1-10
and is believed to be the result of iatrogenic injury to the lateral
epiphyseal branches of the medial circumflex femoral artery. These vessels,
which lie adjacent to the piriformis fossa, are subject to direct trauma
during the preparation of the femoral canal and during antegrade insertion of
the intramedullary
nail6,13.
The case that we reported, which involved a retrograde technique only,
demonstrates that other factors may contribute to the development of
osteonecrosis of the femoral head in the trauma setting.
Case reports have documented osteonecrosis of the immature femoral head in
the absence of femoral head fracture, femoral neck fracture, or dislocation of
the
hip3,15.
Two cases of osteonecrosis of the capital femoral epiphysis following trauma
were presented by
Stougard15. One
patient, a thirteen-year-old boy, had a fracture of the greater trochanter
only; the other patient, an eleven-year-old boy, sustained a direct blow to
the hip without radiographic evidence of injury and had subsequent development
of osteonecrosis15.
Buford et al. presented a case in which osteonecrosis developed in both hips
of a twelve-year-old
boy3. Although on
one side a femoral shaft fracture had been treated with antegrade
intramedullary fixation 1.5 years prior to the development of osteonecrosis,
osteonecrosis also developed in the contralateral hip, which had no known
injury.
Osteonecrosis occurs when the capital femoral epiphysis or femoral head
fails to receive enough blood to support its viability. Therefore, any
condition that disrupts the blood flow to the epiphysis or femoral head
potentially can cause osteonecrosis. Theoretically, in the case presented
here, screws introduced during proximal locking of the femoral nail could have
injured the deep branch of the medial femoral circumflex artery, the main
nutrient vessel supplying the femoral head. Gautier et al. demonstrated,
however, that the deep branch of the medial femoral circumflex artery
invariably crosses the femur posteriorly at least 10 mm, and usually 16 to 20
mm, proximal to the superior aspect of the lesser
trochanter16. In
our patient, the most proximal locking screw was secured at the inferior level
of the lesser trochanter; thus, iatrogenic injury would be unlikely to occur.
Additionally, on the basis of the angiogram
(Fig. 1-B), the branches of the
medial femoral circumflex artery appear to be remote from the proximal locking
screw (Fig. 1-C).
In a patient who has been involved in a motor-vehicle accident, the cause
of osteonecrosis of the femoral head is most likely to be direct trauma to the
deep branch of the medial femoral circumflex artery due to subluxation or
dislocation of the hip or to a fracture of the femoral neck. Hip dislocations
and femoral neck fractures frequently are not initially recognized when an
ipsilateral femoral shaft fracture is
present17.
Preoperative and postoperative radiographs and computed tomography scans
clearly demonstrated no evidence of fracture or dislocation of the hip or
femoral neck in our patient. The possibility remains that the patient
sustained a subluxation of the femoral head that spontaneously reduced prior
to his arrival in the emergency department. Recently, eight cases of traumatic
hip subluxation in football players were
reported18. In each
instance, there was a triad of injury, including a posterior acetabular lip
fracture, hemarthrosis, and a disruption of the iliofemoral ligament. Even
without hip subluxation, the force of the trauma alone, which may cause an
intracapsular hematoma, could theoretically induce injury to the arterial
supply to the femoral
head19. On the
basis of the findings in thirteen children with a fracture of the femoral
neck, Song et al. suggested that intracapsular hematoma after hip trauma may
result in increased intracapsular pressure, thereby compromising perfusion of
the femoral
head19.
Jones and Sakovich championed the idea of fat emboli as a potential cause
of osteonecrosis, and their experiments revealed that fat emboli became
localized in the subchondral bone of the femoral head within a period of
twenty-four hours to five weeks following injection of a vegetable
triglyceride in
rabbits20. They
further demonstrated that fat emboli transiently occluded blood flow in the
small vessels. Subchondral capillaries are deemed more prone to blockage
because small Haversian canals limit the amount of vasodilation, which is
necessary to allow flow past the fat
droplet20. Our
patient had several episodes during which fat emboli could have been released
into circulation. Despite early fixation, the four long-bone fractures
probably created showers of fat emboli. Additional embolism of fat could have
occurred during both the reaming of the canal and the insertion of the
nail.
Interestingly, osteonecrosis of the femoral head has been associated with
angiography of the pelvis in which the femoral artery was
cannulated21. Obaro
and Sniderman reported the case of a forty-one-year-old woman with
osteonecrosis of the femoral head following alcohol embolization of the right
medial femoral circumflex artery after a failed prior internal iliac artery
ligation to control atraumatic pelvic
hemorrhage21.
Despite the lack of embolization during our procedure, the possibility
remains, although it is unlikely, that the femoral artery may have been
injured during cannulation.
The final potential cause of and/or contributor to the development of
osteonecrosis in our patient was the overall medical condition of the patient,
or, more specifically, the blood-volume status during the initial
resuscitation period. Throughout the first hour of hospitalization, he
sustained hypovolemic shock secondary to multiple long-bone fractures and
lacerations of both the liver and the kidney. Furthermore, the patient
required substantial fluid resuscitation, consisting of four units of packed
red blood cells and >3 L of crystalloid solution, in the operating room on
the night of the injury. Although a literature search failed to yield any
documentation of hypovolemia as a source of osteonecrosis, we believe that
prolonged periods of decreased perfusion secondary to low circulating blood
volume could result in osteonecrosis of the femoral head. The amount of time
that bone can tolerate such an insult remains unknown.
The development of osteonecrosis of the femoral head in an adolescent
following intramedullary nailing of the femur can be devastating. This case
demonstrates that, in the setting of polytrauma, factors other than iatrogenic
damage of the vasculature during the preparation and antegrade insertion of
the nail should be considered in the etiology of femoral head osteonecrosis.
?