Severely comminuted intra-articular fractures of the distal
end of the radius are challenging to treat. Restoration of congruity
of the articular surface is the most critical factor for a good
functional result1-7. Restoration
of radial length (the distance from the radial styloid process to
the distal head of the ulna), radial tilt angle, and volar tilt
angle is also important1-6,8.
Failure to achieve and maintain nearly anatomic restoration can
lead to degenerative arthritis, distal radioulnar and midcarpal
instability, and ulnar impaction syndrome, with resultant pain,
decreased motion and strength, and poor function3,5,9-14.
Since 1990 we have addressed the problem of loss of postoperative
stability during fracture-healing by using a combination of internal
and external fixation for the most severely comminuted intra-articular
fractures of the distal end of the radius. The indication for this
treatment protocol is a fracture classified as AO type C3 on preoperative
radiographs because, in our judgment, these fractures have a high
probability of collapse during healing. The aim of the protocol
is to obtain nearly anatomic reduction with internal fixation and
to maintain the reduction by applying an external fixator to assist
in neutralizing forces across the radiocarpal joint during fracture-healing15.
We report the results of a retrospective study of the use of
open reduction and combined internal and external fixation of AO
type-C3 fractures in seventeen patients who were followed for a
minimum of two years.
We reviewed retrospectively hospital and clinical records of
all patients who underwent surgical treatment of an intra-articular
fracture of the distal end of the radius at the University of Miami
School of Medicine/Jackson Memorial Medical Center from
January 1990 to August 1994. Seventy-eight patients with a total
of eighty-one fractures were identified. Fifty-six fractures were
classified as AO type B2, B3, C1, or C2. Twenty-five AO type-C3 fractures
were treated with open reduction and combined internal and external
fixation according to our protocol.
We examined the records and radiographs of the twenty-five patients
(twenty-five fractures) who were treated with the combined technique.
In all patients, the combined technique was applied after attempts
to manage the fracture by closed reduction alone had failed. The
specific radiographic criterion for considering closed reduction
a failure was more than a 2-mm step-off of the distal articular
surface of the radius1,6. Data
on age, gender, handedness, mechanism of injury, fracture characteristics,
concomitant ipsilateral injuries of the forearm and wrist, systemic
injuries, type of fixation, surgical procedures, postoperative management,
and complications were obtained from the patients’ records.
Informed consent was obtained from the patients in accordance with
the guidelines of the United States Food and Drug Administration
and the sponsoring institution.
Seventeen patients were available for follow-up evaluation at
least two years after the injury. Data for eight male patients with
a mean age of thirty-nine years who were lost to follow-up within
two years after the injury were not included in our analysis but
were examined separately. All eight of these patients were treated
with a single surgical approach. External fixation was applied to
all eight fractures and, in addition, Kirschner wires were used
for seven; a dorsal buttress plate, for four; and a volar buttress
plate, for four. Three corticocancellous and two cancellous bone
grafts were used for five fractures. Seven of the eight patients
were followed until after the fixator was removed, at a mean of
six weeks, and three patients were noted to have a total of four
complications at the last follow-up visit. One of these patients
had poor rotational motion that required an arthroplasty of the
distal radioulnar joint, one had poor digital motion and also required
removal of a plate that was causing pain, and one had limited radioulnar
motion.
Demographic Data
The seventeen patients who returned for long-term follow-up evaluation
included fourteen men and three women who ranged in age from twenty-seven to
fifty-nine years (mean, forty-three years). Sixteen patients presented
initially at our institution, and one was referred secondarily.
The dominant wrist was injured in five patients and the nondominant
wrist, in twelve. The initial injury resulted from a high-energy
mechanism in eleven patients: seven were injured in a fall from
a height; one, in a motorcycle accident; one, in an airplane crash;
one, in an automobile-pedestrian accident; and one, in an explosion.
The injury resulted from a lower-energy mechanism in six patients:
three fell while walking or running, two were injured in a bicycle accident,
and one was injured while roller-blading. Eleven patients had a
heavy-duty job, and six had a lighter-duty job.
Concomitant Injuries
Concomitant ipsilateral injuries of the wrist and forearm in
seven patients included an isolated tear of the scapholunate ligament
in three patients, an open wound at the site of the fracture of
the distal end of the radius in three patients (a Gustilo16 grade-I fracture in two patients
and a Gustilo grade-II fracture in one patient), a forearm compartment
syndrome in three patients, a fracture of the ulna in two patients,
a fracture of the scaphoid and capitate in one patient each, and
tears of the scapholunate and lunotriquetral ligaments and the volar
capsule in one patient. Injuries to other areas in four patients
included a fracture of the patella and a subdural hematoma in one
patient; a fracture of the acetabulum in one patient; a testicular
injury and severe burns to the face and torso in one patient; and
fractures of the contralateral humeral shaft, the femur, and the
tibial plateau in one patient.
Surgical Technique
Open reduction combined with internal and external fixation was
performed, if possible, within one week after the injury. In patients
with systemic injuries, the surgery was performed as soon as the patient’s
condition had stabilized. In patients who presented with an acute
carpal tunnel syndrome, the carpal tunnel was released emergently
and open reduction and internal and external fixation was performed
during the same operation. In patients with an open fracture or
massive swelling that precluded early open reduction and internal
fixation, external fixation alone was performed initially and then
definitive open reduction and internal fixation was done when the
wound was clean and the swelling had decreased.
A general anesthetic was administered to thirteen of the seventeen
patients, and an axillary block was used in four. The external fixator
was initially placed, with use of standard technique10,17. An open technique was used for
placement of the fixator pins. Pins were placed in the neck and
base of the second metacarpal18.
The most distal pin in the radius was placed at least 2 cm from
the fracture site, and the proximal pin was placed at least 2 cm
proximal to the distal pin. Two types of external fixator were used, depending
on the surgeon’s preference. An AO fixator (Synthes USA,
Paoli, Pennsylvania) with 4.0/3.0-mm Schanz pins was used
in sixteen patients, and a double-stacked frame was constructed
for stability10. An Orthofix wrist
external fixator (Orthofix, Richardson, Texas) with 3.0 to 3.3-mm
tapered pins was used in one patient, and the frame determined the pin
placement in the second metacarpal and the radius2.
Maximum manual traction was placed on the fixator to help to reduce
the fracture and to correct radial length and displacement through
ligamentotaxis19. The reduced
position of the fracture fragments was checked with fluoroscopy,
and it was determined, in all seventeen patients, that use of external fixation
alone was inadequate to reduce articular comminution and to correct
and maintain radial length.
The surgical approach depended on the direction of the displacement
of the fracture and the location of the comminution4. The approach was dorsal for thirteen
fractures with dorsal displacement and comminution, and an anterior
Henry approach was used for three fractures with volar displacement
and comminution1. One fracture
required both approaches because the severe comminution and instability
involved both the volar and the dorsal cortex2 (Figs. 1-A and 1-B).
In the dorsal approach, a straight incision is used to release
the extensor retinaculum at the third extensor compartment and to
allow retraction of the extensor pollicis longus tendon7. The fourth compartment is raised
off the distal end of the radius subperiosteally and is retracted ulnarly,
and the second compartment is raised and is retracted radially,
exposing the fracture site4. The
dorsal capsule is released off the dorsal surface of the radius
in a "T" configuration to allow visualization
of the articular fragments1. Every
attempt is made to avoid dissection in the fourth compartment to
minimize scarring of the extensor tendons. Upon closure, the extensor
retinaculum is repaired and the extensor pollicis longus is positioned
superficial to the retinaculum.
The volar approach is made between the flexor carpi radialis
tendon and the radial artery2,7.
Dissection is extended down to the pronator quadratus, and the muscle
is freed from the volar aspect of the distal end of the radius,
exposing the fracture4,6. The
volar capsule is not dissected off the volar aspect of the radius
in order to maintain the strong ligamentous support2. Release of the volar articular fragments
from the soft tissue should be limited to the extent necessary to
allow visualization and to ensure anatomic reduction of the articular
surface4.
To correct radial shortening and malalignment, the articular
surface should be elevated as a unit to the level of the scaphoid
and lunate and then reduced6.
During this process the carpus is used as a template to judge the
anatomic position of the distal fragments.
After the articular fragments had been anatomically reduced,
fourteen fractures required stabilization with various internal
fixation devices depending on fragment size. Eleven fractures with
smaller fragments were stabilized with 0.045-mm Kirschner wires,
two with larger fragments were stabilized with 3.5-mm cortical screws1,12, and one was stabilized with a
0.045-mm Kirschner wire and a 3.5-mm cancellous screw. Articular
stabilization of the other three fractures was provided by a 3.5-mm
distal radial plate (Synthes USA); a dorsal plate was used in one
patient, and a volar plate was used in two others.
Defects in eleven fractures were filled with bone graft to provide
stability and to prevent collapse1,6,12,19,20.
Indications for bone-grafting were an AO type-C3.2 or C3.3 fracture
in which defects in the metaphysis alone or in the metaphysis and
diaphysis were noted after reduction of the articular surface. Eight
fractures had metaphyseal defects alone, and three fractures had
combined metaphyseal and diaphyseal defects. The type of bone graft
depended on the size of the defect. Corticocancellous iliac crest
bone graft was used for five fractures with larger defects. The
graft was placed in the defect with the cortex positioned so that
it offered optimal resistance to displacement or collapse of the
articular fragments—that is, posteriorly in a dorsally
displaced fracture and anteriorly in a volarly unstable fracture.
Cancellous bone graft was then packed into the remaining spaces
of the fracture. Of six fractures with smaller defects, five were
treated with cancellous bone and one was treated with bone substitute
(Collagraft; Zimmer, Warsaw, Indiana).
A Synthes AO 3.5-mm distal radial plate (Synthes USA) was applied
to eleven fractures to buttress the cortex for added stability4,19. A dorsal plate was used in two
patients, a volar plate was used in eight patients, and plate fixation of
both cortices was performed in one patient who had extensive comminution
of both the anterior and the posterior metaphyseal and diaphyseal
cortices (Figs. 1-C, 1-D, 1-E, and 1-F).
The external fixator was maintained across the wrist postoperatively
to provide slight distraction and to unload the radiocarpal joint.
Approximately 30% to 50% of the distraction force
was released at the end of the operative procedure. Satisfactory reduction
with slight distraction of the radiocarpal joint was confirmed with
fluoroscopy or plain radiographs10.
The external fixator remained until bone-healing was demonstrated
on plain radiographs; it was removed at a mean of seven weeks (range,
five to nine weeks) postoperatively. Bone-healing was determined
radiographically by the appearance of bridging trabeculae across
the fracture site and clinically by the fracture site being nontender
to palpation. Ten of the patients had the Kirschner wires removed
when the fixator was removed. The wires were removed earlier from
two patients who had pin-track-related problems.
Twenty-eight concomitant procedures performed on the ipsilateral
wrist and forearm included excision and débridement of
an open fracture, open reduction and internal fixation of an open
fracture of the ulnar shaft, and repair of intercarpal ligament injuries
with bone anchors and pinning or with pinning alone. Carpal tunnel
release was performed in fifteen patients who had had signs and
symptoms of median-nerve compression preoperatively or were at risk
for compression because of extensive swelling of the wrist and forearm1,9. Three patients who had an acute
forearm compartment syndrome were treated with immediate fasciotomy.
Two of them later had delayed primary wound closure, and one had
split-thickness skin-grafting.
The time from the injury to the definitive operation ranged from
one to fifteen days (mean, seven days). Two patients with severe
associated injuries had a staged approach that included emergency
stabilization of systemic injuries, placement of external fixation
across the wrist to stabilize the fracture of the distal end of
the radius, and decompression of the carpal tunnel. Excision and
débridement was performed on an open ulnar fracture in
one of these patients and on a grade-II open fracture of the distal end
of the radius in the other. Open reduction and internal fixation
of the distal end of the radius was delayed until the condition
of each of these two patients had stabilized and the wounds were
clean, nine and six days after the injury. The two patients with
a grade-I open fracture of the distal end of the radius had open
reduction and internal fixation at the same operation as the initial
excision and débridement.
Postoperative Management
All seventeen patients began active and passive range-of-motion
exercises of the hand, forearm, elbow, and shoulder on the day after
the operation4. To prevent finger
flexion contractures, volar static splints were worn at night for
approximately one to two weeks.
Fracture Classification and Follow-up Assessment
Fracture classification and radiographic assessment were performed
retrospectively by three raters—two fellowship-trained
full-time hand surgeons and a chief resident in orthopaedics—who independently
measured the parameters on each radiograph. Two of the raters were
blinded to the identity of the patients. Objective and subjective clinical
results were determined from a physical examination and an interview
conducted by one of us (R.A.R.) at the time of final follow-up.
Fracture classification: Fractures were categorized
on preoperative radiographs (without traction or computed tomographic scans)
within the AO type-C3 classification and according to the Frykman
system5-7,21-23. Type-C3 fractures,
the most severe within the AO system, are divided into three subclassifications according
to whether the comminution involves the articular surface (C3.1),
metaphysis (C3.2), or diaphysis (C3.3)3,19,22.
An injury score was assigned to each fracture on the basis of the
number of fracture fragments and the number of intercarpal injuries:
1 point was assigned for each fragment, and 1 point was assigned
for each injury7.
Objective assessment: At the final follow-up
examination, patients were tested for range of motion, for grip
strength with a Jamar grip dynamometer (Sammons Preston, Bolingbrook,
Illinois), and for pinch strength with a pinch dynamometer. Measurements
of the contralateral side served as controls7.
A sensory evaluation with Semmes-Weinstein monofilament testing
was performed. The wrist and the distal radioulnar joints were palpated
for tenderness.
Subjective assessment: At the final follow-up
visit, a questionnaire was given orally regarding the level of pain,
return to work, type of work, and participation in sports activities.
Patients were asked if they were very satisfied, satisfied, or not
satisfied with the result of the surgery and whether they would
undergo the surgery again under similar circumstances.
Assessment of radiographic parameters: Standard
guidelines were utilized to determine selected radiographic parameters3,4,7,15,24,25, which were analyzed
retrospectively on anteroposterior and lateral radiographs made
at four time points: preoperatively before any reduction maneuvers,
immediately after the operation, just before removal of the external
fixator, and at the final follow-up evaluation. Values for the contralateral
side were determined from anteroposterior and lateral radiographs
made at the final follow-up examination7.
Assessment of arthritis: The severity of arthritis
at the radiocarpal joint was determined retrospectively on anteroposterior
and lateral radiographs made preoperatively and at the final follow-up
examination. Arthritis was graded as none, mild, moderate, or severe
according to the system of Knirk and Jupiter5.
Clinical rating systems: The outcome for each
patient was evaluated with use of two scoring systems. The clinical
scoring system of Green and O’Brien26,
as modified by Cooney et al.27,
provides a score based on subjective and objective clinical data
(pain, work activity, range of motion, and grip strength). The demerit-point
system of Gartland and Werley28,29,
as modified by Sarmiento et al.30,
provides a score based on subjective and objective clinical and
radiographic data (amount of residual deformity, pain, range of
motion, grip strength, and complications). Each system allows grading
of the final outcome as excellent, good, fair, or poor27,28,30.
Statistical Analysis
The independent t test (two-tailed) or one-way analysis of variance
was used to test for differences between groups. The Pearson correlation
coefficient was used to examine for relationships between variables.
Chi-square analyses were conducted on nominal variables. The probability
of a type-I error (alpha) was set at £0.05 for all statistical
analyses. Kappa and intraclass correlations (3,1) were used to examine
the interrater reliability31.
The mean duration of follow-up of the seventeen patients was
thirty months (range, twenty-four to sixty-one months).
Fracture Classification and Outcome
The values provided by the three raters for fracture classification
and radiographic parameters were not averaged because interrater
reliability was poor (kappa values of <0.5). For this report,
we used only the values supplied by one of us (R.A.R.), a full-time
hand surgeon.
Fracture Classification
Within the AO type-C3 classification, six of the seventeen fractures
were subclassified as 3.1; eight, as 3.2; and three, as 3.3. Three
fractures were classified as Frykman type VII and fourteen, as Frykman
type VIII. The mean injury score (and standard deviation) was 4.2
1.25 points (range, 3 to 8 points). Three fractures were open;
two of them were classified as Gustilo grade I and one, as Gustilo
grade II.
Objective Outcome
The mean values for range of motion, grip strength, and pinch
strength for the seventeen patients are given in Table I. The mean total
arc of flexion-extension at the final follow-up evaluation was 93°
(range, 0° to 140°), 72% of the value for the uninjured
side; the mean total arc of supination-pronation was 159° (range, 115°
to 180°), 91% of the value for the uninjured side; and
the mean total arc of radioulnar deviation was 44° (range, 0° to
64°), 71% of the value for the uninjured side.
The mean grip strength on the injured side was 31 kg (range,
12 to 43 kg), 73% of the grip strength on the uninjured
side. The mean pinch strength, known for fifteen patients, was 8
kg (range, 5 to 12 kg), 80% of that on the uninjured side.
At the final follow-up evaluation, all patients demonstrated
normal sensibility, which ranged from 1.65 to 2.83 on the Semmes-Weinstein
monofilament test.
Subjective Outcome
Sixteen patients were very satisfied with the outcome, and one
patient was not satisfied. The rating of very satisfied, even when
given by patients with wrist stiffness and/or moderate
pain, reflected the patients’ appreciation of the substantial
recovery that they achieved after a severe wrist injury. All seventeen
patients indicated that they would have the surgery again under
similar circumstances. Seven patients reported no pain at the final
follow-up examination, and seven had only mild or occasional pain.
Two patients reported moderate but tolerable pain. The patient who
was unsatisfied with the outcome reported severe, intolerable pain.
He had been incarcerated after removal of the external fixator and
had not been given an opportunity for physical therapy. Three patients
described pain at the distal radioulnar joint and had tenderness
on palpation of the joint. Two of these patients had had an ulnar
styloid fracture on preoperative radiographs.
Fourteen patients had returned to work by the time of the final
follow-up evaluation, and ten had resumed their preinjury employment.
Eight of the eleven patients who had held a heavy-duty job before
the injury returned to work; five of them had no restrictions and
three had some restrictions. The mean time until the patients returned
to work was ten months (range, 0.25 to thirty months). Of the three
patients who were not working at the time of follow-up, one was
permanently disabled because of multiple medical problems unrelated
to the wrist fracture, one was able to work but could not find employment,
and one was incarcerated.
Twelve patients reported that they had participated in sports
before sustaining the wrist injury. By the time of the final follow-up
evaluation, all twelve had returned to sports, including football,
fishing, billiards, golf, and skiing. Eight of them had returned
to the same sports activities in which they had participated before
the injury. The mean time until the patients returned to sports
was ten months (range, two to thirty months).
Radiographic Outcome
At the final follow-up examination, fifteen patients had <2
mm of articular step-off, thirteen had <3 mm of total articular
incongruity, and twelve had 10 mm of radial length (Table II). Using repeated-measures
analysis of variance, we found a significant improvement in eight
parameters (p < 0.002 for all) when radiographs made immediately
postoperatively were compared with those made preoperatively. With
the numbers available, no significant change in the mean scapholunate
angle, carpal height index, or revised carpal height index could
be detected. Also, we detected no significant differences in eight
parameters when the final follow-up radiographs were compared with
those made immediately postoperatively. However, significant differences
in the mean revised carpal height index, ulnar variance, and tilt angle
of the distal part of the radius (p < 0.05) were found
(Table II).
Comparison of the injured and uninjured wrists at the final follow-up
evaluation demonstrated, with the numbers available, no significant
differences in the mean radial tilt angle, radial length, ulnar
variance, scapholunate angle, carpal height index, or revised carpal
height index. Statistical comparisons of the injured and uninjured
wrists were not conducted for the mean articular gap, articular
step-off, total articular incongruity, or shortening because the
values for the uninjured wrists were zero; however, the mean values
for the injured wrists were within normal limits (Table II). The mean
tilt angle of the distal end of the radius for the injured wrists
at the final follow-up visit was significantly less than that for
the uninjured wrists (p < 0.05).
Arthritis
Three of the seventeen patients had no radiographic signs of
arthritis at the final follow-up visit, ten had mild arthritis,
and three had moderate arthritis. One patient, who had demonstrated
mild arthritis on preoperative radiographs, had severe arthritic changes
with joint incongruity, osteophyte formation, and bone-on-bone contact
at the final follow-up evaluation5.
Outcome of Evaluations with Clinical Rating
Systems
According to the modified clinical scoring system of Green and
O’Brien26, the functional
result was excellent for one patient and good for four (29% good
or excellent results); the result was fair for eight patients and
poor for four. The mean score (and standard deviation) was 66.5
17.3 points (range, 25 to 90 points). The mean pain score was 20
points (range, 0 to 25 points), and the mean return-to-work score
was 21 points (range, 0 to 25 points).
According to the demerit-point system of Gartland and Werley28,29, eight patients had an excellent
result and two had a good result (59% good or excellent
results); four patients had a fair result, and three had a poor result.
The mean demerit-point-system score was 7.6 7.8 points (range,
0 to 22 points).
Complications
Five of the seventeen patients had a total of six complications
(two major and four minor9). In
one patient with an AO type-C3.2 fracture, loss of reduction and
collapse of the fracture at six weeks led to the development of
moderate arthritis at the radiocarpal joint. A wrist fusion and
a Sauvé-Kapandji procedure was performed four months after
the injury to eliminate pain and to improve function. Although the
patient had a poor result according to both clinical rating systems
at the final follow-up visit, sixty-one months after the injury, he
had only occasional pain and a stable wrist and he was very satisfied
with the ultimate outcome.
Reflex sympathetic dystrophy developed in one patient; it resolved
with physical therapy after removal of the fixator at nine weeks.
The patient had a fair result according to both clinical rating systems.
At the final follow-up visit, forty-four months after the injury,
the patient reported occasional pain and was very satisfied with
the result. He had returned to his preinjury sports activity and to
his previous job as a counselor.
Four minor complications involving Kirschner wires (tendon irritation
or superficial infection) were noted in three patients. All symptoms resolved
after removal of the wires.
Effect of Preoperative Factors on Surgical
Technique
We analyzed whether the mechanism of injury, injury score, Frykman
classification, AO type-C3 subclassification, or presence of associated
injuries influenced the type of fixation used (Kirschner wires,
screws, or a buttress plate), the number of fixation methods used,
or the use of a bone graft in the seventeen patients. Chi-square
analysis demonstrated a significant relationship between the AO type-C3
subclassification and the use of bone-grafting (p < 0.03).
The need for a bone graft increased with increasing severity of
comminution. No other significant relationships were noted.
Effect of Radiographic Factors on Outcome as Measured
by Clinical Rating Systems
We examined the relationship of radiographic parameters and the
final outcome. The Pearson correlation coefficient indicated that
total articular incongruity immediately after the operation showed a
moderately strong correlation with the final outcome according to
both rating systems (r = 0.70 and p < 0.002 for
the outcome according to the demerit-point system; r = 0.74
and p < 0.0008 for the outcome according to the Green and
O’Brien system). As the total articular incongruity increased,
the final outcome worsened. There was a moderate correlation between
both the articular step-off and the articular gap immediately after
the operation and the final outcome (r = 0.64 and 0.61,
respectively, and p < 0.01 for the outcome according to
the demerit-point system; r = 0.72 and 0.61, respectively,
and p < 0.01 for the outcome according to the Green and
O’Brien system).
Effect of Duration of External Fixation and
Degree of Distraction on Outcome Parameters
Chi-square analysis showed that neither the number of weeks that
the external fixator had been in place nor the degree of distraction
that had been applied with the fixator, as measured by the carpal
height index and the revised carpal height index, influenced the
final outcome according to the demerit-point system or the Green
and O’Brien clinical assessment. Moreover, neither of these
factors influenced the range of motion, grip strength, or score
in the pain category of the Green and O’Brien system.
The findings of our retrospective study confirm the observation
by others that anatomic restoration of the articular surface is
a critical part of the operative treatment of AO type-C3 fractures
and has a direct influence on the final outcome1,5,6,19.
Bradway et al.19 and Knirk and
Jupiter5 showed that >2
mm of articular incongruity (step-off) was associated with a high
prevalence of post-traumatic arthritis and poorer functional results.
We attribute the maintenance of fracture reduction throughout
bone-healing to two factors. The first was augmentation of fracture
fixation by the external fixator3,4,6,19.
Placement of the external fixator at full distraction before beginning
the surgical approach aided fracture reduction and improved radial
length and alignment4,10, while
removal of 30% to 50% of the distraction force
at the end of the operation avoided overdistraction. As in other
series in which external fixation was used2,10,21,32,
the distraction force across the wrist was maintained (there was
essentially no change in the mean carpal height index or the mean
revised carpal height index between radiographs made immediately
after the operation and those made just prior to removal of the
external fixator). The fixator neutralized the compressive forces
across the joint, preventing collapse of the fracture during the
initial two to four weeks of healing, when bone resorption and early
new-bone formation are maximal.
The second factor was our use of bone graft in eleven of the
seventeen fractures1,2,7,11,19-22,
to provide structural support and to accelerate the healing process.
We attained satisfactory results using corticocancellous iliac crest
bone graft for added structural support in five fractures with large defects.
An important detail was the placement of the corticocancellous bone
graft into the defect with the cortex positioned so that it provided
maximal resistance to displacement or collapse of the articular
fragments12. The mean time to
bone union for the five fractures in which corticocancellous bone
graft had been used was seven weeks, only one week longer than that
for the six fractures in which bone substitute or cancellous bone
alone had been used.
The use of an external fixator alone, or in conjunction with
percutaneous or limited internal fixation, for unstable fractures
of the distal end of the radius has produced good or excellent results3,10,15,17-21,33-37. Intraoperative
assessment of each of our patients showed that the external fixator
alone failed to reduce the severe articular comminution adequately or
to correct and maintain length through ligamentotaxis14,38.
We removed the fixators five to nine weeks (mean, seven weeks)
after the operation, earlier than in several other series2,10,15,17,33,37. Perhaps we were able
to do so because we had used bone graft in conjunction with the
fixation, as others have reported7.
Earlier removal of the fixator allowed us to begin range-of-motion
exercises and to avoid complications commonly associated with the
prolonged use of external fixators3,9,10,14,18,21,33,38.
The use of open reduction combined with internal and external
fixation has been reported previously in selected cases4,6,19,22,23,39-41. Bass et al. presented
the results of the combined two-incision technique in a group of
AO type-C3 fractures of the distal end of the radius2. Good results were achieved, with
a total range of wrist motion and grip strength that were 80% and 83% of
the values for the uninjured side, respectively, despite the external
fixator having been maintained for a mean of twelve weeks.
The seventeen patients in our series sustained the most severe
type of comminuted intra-articular fracture of the distal end of
the radius (AO type C3), yet the objective and subjective results
are comparable with those in many series that have included less
severe fractures1,2,4-6,19,22.
Bradway et al. reported on sixteen patients with an AO type-C2 or
C3 fracture treated by open reduction and internal fixation, augmented
by external fixation in five patients19.
Patients had a mean of 110° of flexion-extension, and the mean grip
strength was 75% of that on the normal side. Jupiter and
Lipton described thirteen AO type-C1 or C3 fractures that were treated
by open reduction and internal fixation, with supplemental external
fixation used in three cases4.
The mean total arc of flexion-extension was 74% of the
value for the uninjured side, and the mean grip strength was 76% of
the value for the uninjured side.
Limitations of our study are the small number of patients, the
retrospective design, and the lack of a control group. A randomized,
prospective study of a group treated with the protocol and a control group
treated with internal fixation alone would have been necessary to
assess the value of supplemental external fixation.
We believe that severely comminuted intra-articular (AO type-C3)
fractures of the distal part of the radius should be treated by
open reduction and combined internal and external fixation. Our
series demonstrates that the combined technique, supplemented by
bone-grafting and plate fixation as needed, is a satisfactory treatment
that can lead to a high rate of return to work and sports, a high
level of patient satisfaction, and a low rate of complications.