It is well recognized that unstable fractures of the distal part of the
radius may require operative treatment to restore alignment and that failure
to restore alignment often leads to wrist and forearm
dysfunction1-4.
One particularly challenging fracture pattern has been that associated with
complex fragmentation of both the articular surface and the metaphysis (Figs.
1-A,
1-B,
1-C,
1-D,
2-A,
2-B, and
2-C). In the Comprehensive
Classification of
Fractures5, such
fractures are classified as subgroup C3.2. Extensive metaphyseal comminution
makes it more difficult to restore the alignment of the distal fragments and
leaves them with little or no bone-to-bone contact to help prevent loss of
alignment. External fixation with or without ancillary Kirschner wires or bone
graft may be unable to prevent settling in this circumstance. A single dorsal
or volar plate may not provide adequate stability, and the distal fragments
may displace in the direction opposite to the plate. Previous reports have
described the use of combined internal and external
fixation6,7,
temporary bridge plating of the
wrist8, and even
primary wrist
arthrodesis9 for
these complex fractures.
It has been our preference to use a combined dorsal and volar exposure and
plate fixation for the treatment of these complex fractures, with the
rationale that this approach can allow better access to and control of the
fragments. The plates can cradle the articular fragments, compressing them
together and providing improved support at the metaphyseal level. The concern
is that the combined dorsal and volar dissection might increase stiffness of
the wrist and digits, lead to increased complications such as reflex
sympathetic dystrophy, and compromise blood supply to the bone, thereby
contributing to a risk of nonunion or infection. We reviewed our experience
with combined dorsal and volar plate fixation of complex fractures of the
distal part of the radius in order to better characterize the advantages and
disadvantages of this technique. In particular, we were interested in the
rates of nonunion and osteonecrosis.
Three surgeons (D.R., J.B.J., and K.J.P.) at two medical centers managed
twenty-nine consecutive patients who had a subgroup-C3.2
fracture5 with
combined dorsal and volar plate fixation. The group-C3 designation indicates
the presence of more than one articular fracture line. All of our patients had
a coronal split in the articular surface as described by
Melone10,11.
The subgroup-C3.2 designation indicates that the metaphyseal fracture line is
also comminuted but does not extend into the diaphysis. The classification of
these fractures was based on radiographs, computed tomography scans, and
operative exposure. Trumble et al. categorized metaphyseal comminution as
dorsal, volar, radial, and
ulnar12; all four
areas were involved in these patients. Four patients had inadequate follow-up
and could not be contacted, leaving twenty-five patients available for
evaluation. The Human Research Committee approved a retrospective review of
medical records and also approved contacting patients to invite them to return
for a physical and radiographic examination to extend the period of follow-up
when necessary.
The study group included sixteen men and nine women with an average age of
forty-six years (range, twenty-six to seventy-two years) (see Appendix).
Twelve right hands, nine of which were dominant, and thirteen left hands, none
of which were dominant, were involved. Twenty-one patients were employed at
the time of injury. Specifically, three patients worked in jobs involving
heavy labor, three worked in technical jobs, six worked at a desk, two worked
as teachers, and one each worked as a truck driver, an artist, a police
officer, a medical assistant, an attorney, a postman, and a dentist. Two
patients were retired, and two were homemakers.
The mechanism of injury was a simple fall from a standing height for nine
patients, a high-energy fall for fourteen patients (including a fall from a
height for ten patients, a fall during sports for two patients, a fall on ice
for one patient, and a fall down stairs for one patient), and a motor-vehicle
accident for two patients. Associated injuries of the ipsilateral wrist
included a displaced fracture of the scaphoid waist in two patients and a
complete rupture of the scapholunate interosseous ligament in two patients.
One other fracture was associated with a Gustilo and
Anderson13 type-2
wound on the ulnar side of the wrist and a traumatic extensor pollicis longus
laceration at the site of the fracture. Five patients had progressive sensory
dysfunction of the median nerve that was thought to be due to an acute carpal
tunnel syndrome.
Twenty patients had an associated fracture of the distal part of the ulna.
In nineteen patients, this fracture involved the base of the ulnar styloid (Q
modifier 1 according to the Comprehensive Classification of
Fractures5). The
patient with an open wound had a fracture of the ulnar head (Q modifier 4) and
complete avulsion of the soft tissues from the distal part of the ulna,
including the triangular fibrocartilage complex
(Fig. 2-A).
One patient had a fracture of the contralateral radial head and a severe
soft-tissue injury of the contralateral leg. Another patient had a fracture of
the contralateral clavicle. In the remaining twenty-three patients, the
fracture was an isolated injury.
Six patients were treated after the failure of a previous operation to
restore alignment, including external fixation augmented with Kirschner wires
(three patients) and dorsal plate fixation (three patients) (Figs.
1-B and
1-C). One of the patients who
initially had had external fixation also had had a second operation for dorsal
plate fixation before referral to us. For these six patients, the average
interval between the initial injury and the index operation for dorsal and
volar plate fixation was twenty days (range, nine to thirty-five days) and the
average interval between the most recent previous operation and the index
operation was ten days (range, four to eighteen days). In the remaining
nineteen patients, the initial treatment consisted of closed reduction and
immobilization in either a splint or a cast.
Operative Techniques
Temporary intraoperative external fixation or skeletal distraction was used
to facilitate reduction in fourteen patients. In four patients, the fixator
was left in place for three weeks (two patients), four weeks (one patient), or
six weeks (one patient) after surgery in order to avoid the need for a
circumferential dressing in the hopes of limiting digital edema and
stiffness.
Fractures were first exposed dorsally through a longitudinal incision
centered over Lister's tubercle. The extensor pollicis longus was mobilized
from the third dorsal compartment, was transposed radially, and was left in
the subcutaneous tissues at the end of the procedure. The radial wrist
extensors were retracted radially. An attempt was made to keep the fourth
dorsal compartment intact by elevating it subperiosteally in the ulnar
direction.
The articular surface was visualized and manipulated through a dorsal
capsulotomy. Impacted central articular fragments were repositioned and were
supported with bone. In four patients, bone from the iliac crest was used;
however, in most patients there was sufficient local bone for this purpose.
Angular stable buttress pins (screws that engage a threaded hole in the plate)
were used to help support the articular fragments in seven patients.
Major fragments were realigned and fixed provisionally with 0.045 or
0.062-in (0.11 or 0.16-cm) Kirschner wires. The implants that were placed on
the dorsal surface of the radius included a p-shaped
plate14 in eighteen
patients, a 2.0-mm plate in two patients, and a 3.5-mm T-shaped plate in five
patients; all plates were manufactured by Synthes (Paoli, Pennsylvania).
On the volar side, a volar-radial exposure as described by
Henry15 was used
for sixteen patients and a volar-ulnar carpal tunnel-releasing
exposure16 was used
for nine patients (including all five patients with acute carpal tunnel
syndrome). A 2.0, 2.4, or 3.5-mm T-shaped plate (Synthes) was applied to the
volar surface of the distal part of the radius. In twenty-two of the
twenty-five patients, the volar plate incorporated angular stable pins that
engaged the plate by means of threads in the screw-holes.
In four patients with a very small volar fragment, either a 24-gauge
stainless steel wire (two patients) or a 2-0 braided nonabsorbable suture (two
patients) was passed through the volar radiocarpal ligament attachments to
better control this
fragment17. In two
patients, a percutaneously inserted 0.062-in (0.16-cm) Kirschner wire that had
been used to secure the radial styloid fragment was left in place for four
weeks.
Associated injuries included two scaphoid fractures (both of which were
treated with screw fixation) and two scapholunate interosseous ligament tears
(one of which was treated with open repair and stabilization with 0.062-in
[0.16-cm] Kirschner wires and one of which was treated with a temporary screw
that was removed six months later).
Postoperative Management
The wrist was supported in a volar splint or an external fixator for two to
six weeks; an external fixator was used for patients who had substantial
swelling. Active-assisted exercises of the digits and forearm and use of the
involved limb for light daily activities were encouraged immediately after
surgery. Swelling in the digits was treated with frequent active exercises,
elevation of the limb, and compressive dressings, including wrapping of the
fingers with self-adherent elastic bandages or the application of compressive
gloves. The splint or external fixator was removed and wrist exercises were
initiated six weeks after the injury (or earlier if the patient had achieved
full digital and forearm motion and had limited swelling). Resistive exercises
were initiated once healing was established on radiographs, usually at about
ten weeks after the injury.
Complications and Subsequent Operations
Two patients had a tendon rupture that was related to the implant: in one
patient the flexor pollicis longus ruptured on the sharp radial edge of the
volar plate, and in the other patient the extensor digitorum communis to the
ring finger ruptured as a result of chronic irritation caused by the dorsal
plate. The flexor pollicis longus was reconstructed with an intercalary
palmaris longus tendon graft, and the extensor digitorum communis to the ring
finger was reconstructed with an extensor indicis proprius transfer; both
reconstructions were performed at the time of plate removal. Twenty-one
patients had plate removal within the study period because of pain or surgeon
preference. Two patients had a carpal tunnel release at the time of hardware
removal.
Evaluation
The patients were evaluated by the treating surgeon after all subsequent
procedures. Motion and grip strength were recorded, posteroanterior and
lateral radiographs were made, and a final rating was assigned according to
the system of Gartland and
Werley18 and the
Mayo modification of the system of Green and
O'Brien19. The
final alignment on radiographs was characterized by measuring the orientation
of the articular surface on the posteroanterior and lateral radiographs, the
ulnar variance, and the maximum articular step or gap. Radiographic signs of
arthrosis were rated according to the system of Knirk and
Jupiter3.
After an average duration of follow-up of twenty-six months (range,
fourteen to forty-eight months), the final range of motion of the wrist
averaged 54° (range, 30° to 70°) of extension, 51° (range,
25° to 80°) of flexion, 27° (range, 5° to 30°) of ulnar
deviation, 18° (range, 10° to 25°) of radial deviation, 79°
(range, 30° to 90°) of pronation, and 74° (range, 20° to
80°) of supination. The average grip strength was 70 lb (31.8 kg) (range,
20 to 110 lb [9.1 to 49.9 kg]), or 78% (range, 45% to 100%) of that of the
contralateral, uninjured limb. No patient had stiffness of the fingers or
thumb or development of a complex regional pain syndrome.
Radiographs that were made at the time of the most recent follow-up
revealed an average of 2° of dorsal angulation (range, 20° of volar
angulation to 15° of dorsal angulation) of the articular surface of the
distal part of the radius on the lateral radiograph, an average of 21°
(range, 14° to 30°) of ulnar inclination of the articular surface on
the posteroanterior radiograph, an average of 0.8 mm of positive ulnar
variance (range, 1 mm of negative ulnar variance to 3 mm of positive ulnar
variance), and an average of 0.7 mm (range, 0 to 2 mm) of articular
incongruity.
All fractures healed. Seven patients had development of radiographic signs
of arthrosis during the study period. These changes were rated as grade 1 in
five patients and as grade 2 in two patients according to the scale of Knirk
and Jupiter3. There
were no radiographic signs of devitalization of the fracture fragments.
According to the relatively forgiving rating system of Gartland and
Werley18, the
functional result was excellent for thirteen patients, good for eleven, and
fair for one. According to the more stringent modified system of Green and
O'Brien19, the
functional result was excellent for five patients, good for five, fair for
fourteen, and poor for one.
The common feature of the fractures treated in this series was a
combination of complex articular and metaphyseal comminution. The
Comprehensive Classification of
Fractures5
distinguishes complex articular fractures of the compression (nonshearing)
type as those that are associated with more than one articular fracture line.
Group-C1 and C2 fractures have a simple split, either between the lunate and
scaphoid facets or in the coronal plane. Group-C3 fractures are those that are
associated with additional fracture lines or comminution. As described by
Melone10,11
and confirmed by
others4,20-22,
group-C3 fractures usually involve a coronal split in the lunate facet of the
distal radial articular surface. Melone and others have emphasized the need
for operative exposure and internal fixation of the volar lunate facet
fragment10,11,16,20.
When there is limited metaphyseal comminution (as in subgroup-C3.1 fractures),
the dorsal lunate facet fragment and the radial styloid can be manipulated and
percutaneously pinned and protected with an external
fixator20; however,
when there is more extensive metaphyseal comminution (as in subgroup-C3.2 and
C3.3 fractures), the realigned fragments have little or no osseous support and
may require more stable fixation. In subgroup-C3.3 fractures, the
extra-articular comminution extends to the diaphysis. In this situation,
control of the articular fragments can be so difficult that some surgeons have
suggested temporary plate fixation across the wrist
joint8.
In patients with subgroup-C3.2 fractures, a plate can be used to bridge the
metaphyseal comminution, but it can be difficult to gain reliable hold of the
articular fragments. If a plate is placed on one side of the bone, the
fragments tend to collapse to the opposite side. The use of combined dorsal
and volar plate fixation enhances support of the articular fragments by
compressing them between the two plates. In particular, volar plate fixation
is increasing in
popularity23,24,
but it may not be suitable for the treatment of subgroup-C3.2 fractures. The
occasional need for combined dorsal and volar plate fixation for very complex
fractures has been described in other
series7. The use of
angular stable buttress pins or screws is useful, particularly when bone
quality is poor or there is extensive articular fragmentation. Alternative
forms of treatment that have been suggested include combined internal and
external
fixation6,7
and even primary wrist
arthrodesis9.
In our experience, we have found that combined dorsal and volar plate
fixation of subgroup-C3.2 fractures can achieve healing with good alignment.
The complications that we have noted have been primarily related to the
implants. Nonunion, osteonecrosis, digital stiffness, and dystrophic pain
syndromes are uncommon. The functional result can be surprisingly good for
patients with such complex fractures. The objective results of treatment in
the present series are comparable with those in other series of similarly
complex fractures, in which the average arc of flexion and extension has been
reported to range from 93° to 105° and the average grip strength has
been reported to range from 73% to 83% of that in the opposite
limb6,7,25.
Plate fixation of the distal part of the radius continues to be associated
with the disadvantages of tendon irritation and occasional tendon rupture.
These complications are more commonly associated with dorsally applied plates.
Given the advantages of combined dorsal and volar plate fixation in terms of
maintaining reduction and facilitating rehabilitation, the need for a second
procedure for implant removal seems reasonable.
Nearly all patients had a satisfactory result according to the relatively
forgiving rating system of Gartland and
Werley18, which may
reflect the restoration of functional motion and radiographic alignment.
However, 60% of the patients had an unsatisfactory result according to the
more stringent criteria of the modified scale of Green and
O'Brien19, which
may better reflect the residual difficulties experienced by these patients. In
spite of successful healing and adequate radiographic alignment, the extensive
injury to the articular surface and soft tissues inevitably compromises the
ultimate result.
The term "complex articular fracture of the distal part of the
radius" is used inconsistently. The present study focused on fractures
that were associated with complex fragmentation of both the metaphyseal and
articular areas of the distal part of the radius (subgroup C3.2 according to
the Comprehensive Classification of
Fractures5). We
believe that it is difficult to treat these fractures with either a dorsal or
a volar plate alone but that combined dorsal and volar plate fixation can
achieve healing in reasonable alignment. The frequent need for implant removal
is justified given the complexity of these injuries and the associated
potential for severe compromise of wrist, forearm, and hand function.
A table showing pertinent clinical and radiographic information on each of
the twenty-five patients is available with the electronic versions of this
article, on our web site at
(go to the article citation and click on "Supplementary Material")
and on our quarterly CD-ROM (call our subscription department, at
781-449-9780, to order the CD-ROM).
Note: The authors thank Todd Morgan, MD, and Amit Gupta, MD, for
their help with earlier versions of this data set, presentation, and
manuscript.