Abstract
We describe a very cheap, simple and effective dynamic external fixator for treatment of pilon fractures of the proximal interphalangeal joint. At final follow-up, nine such fractures had regained an average range of motion of 79° (range, 65–90°). There was high patient satisfaction and there were no serious complications.
INTRODUCTION
The word pilon is derived from Latin (a “pounder”) (Stern et al., 1991). Although the word has become synonymous with a compression fracture of the distal tibia, it can be used to describe any hinge joint fracture which is characterized by fragmentation with central depression and splaying of one of the concave articular surfaces. In the hand it is most commonly used to describe injuries to the proximal interphalangeal joint. These fractures mostly occur with a severe loading force to the tip of the finger. This causes compression and hyperextension of the proximal interphalangeal joint resulting in metaphyseal comminution and depression of the central articular surface of the middle phalanx.
Pilon and other intra-articular fractures of the proximal interphalangeal joint may be treated by a variety of techniques with variable results. These include extension blocking splints (McElfresh et al., 1972), external fixation devices using a force couple (Agee, 1987), a “S” Quattro splint (Fahmy, 1990), longitudinal traction (Schenk, 1986), dynamic spring fixators (Allison, 1996), internal fixation, advancement arthroplasty (Eaton and Malerich, 1980), extension blocking with Kirschner wires (Viegas, 1992) and cerclage wire fixation (Weiss, 1996).
It is now recognized that the pilon fracture is a distinct entity which requires special consideration when planning treatment. The problem is two-fold: firstly, the proximal interphalangeal joint fracture fragments are too small to be reconstituted anatomically with open procedures. Secondly, the fibroblastic reaction around the proximal interphalangeal joint leads to long-term stiffness unless joint motion is maintained throughout the healing period (Salter et al., 1980). We have used a very cheap, freely available and effective technique for trealing these fractures. This system utilizes two Kirschner wires bent to form a dynamic external fixation device, which is preloaded to maintain traction and thus keep the fracture reduced while allowing motion. This method is a modification of the dynamic external fixation proposed by Gaul and Rosenberg (1998) and later used by Hynes and Giddins (2001).
PATIENTS AND METHODS
A prospective study was conducted on the use of dynamic external fixation for pilon fractures of the proximal interphalangeal joint. Between 1999 and 2002, eight patients with nine fractures (one patient had injured two fingers) were treated. The demographic details of the patients are shown in Table 1. The average age of the patients was 43 (range, 19–72) years. There were six men and two women. All eight patients were available for final review.
SURGICAL TECHNIQUE
All of our patients elected to have a general anaesthetic for the procedure. Under X-ray control, each fracture was reduced with manual traction and manipulation. A smooth 1.2 mm Kirschner wire was then drilled through the centre of the head of the proximal phalanx, perpendicular to its long axis, using X-ray control to ensure its placement at the centre of rotation of the proximal interphalangeal joint while avoiding the collateral ligaments. A second wire was then inserted parallel to the first through the head of the middle phalanx (Fig 1). Extra-long Kirschner wires (30 cm) should be used for this procedure. Using both wires as joysticks, further fracture manipulation was carried out. In some cases, a very fine Kirschner wire was inserted from the dorsal surface of the middle phalanx to try and realign the major articular fragments and dislodge the central depressed fragment.
The proximal wire is bent into a tight hook on both sides of the finger, about 5 mm from the skin. The distal wire is then bent 90° in the frontal plane and then shaped into a parabolic curve in the saggital plane which runs towards the proximal wire. The ends of this wire are then twisted into tight hooks, which are engaged into the hooks on the proximal wire after preloading the curve of the wire to create tension and distraction across the fracture (Fig 1). A large bulky dressing is applied over the wire ends during the first 24–48 h, but is removed when mobilization is commenced.
The patients were followed-up at weekly intervals in the clinic until removal of the fixator at 4 weeks. They were encouraged to start passive and assisted active mobilization within 48 h (Fig 1). All patients received physiotherapy from a specialist hand physiotherapist.
RESULTS
We reviewed all eight patients with pilon fractures of the proximal interphalangeal joint at an average follow-up of 2.2 (range, 2–3) years. Two fractures were due to a direct hit from a cricket bat, three fractures occurred while playing rugby/football, three from a fall down a flight of stairs and one during an assault. Objective analysis revealed an average arc of motion of 79° (range 65°–90°) with an average fixed flexion deformity of 8° (range, 0–20°). The average delay from injury to the operation was 6 (range, 2–10) days. At final review, subjective analysis revealed that all patients were pain free while performing their activities of daily living. One patient (a retired manual labourer) complained of moderate-to-severe pain with heavy work. There were no other major complaints, only minor complaints of inability to wear rings in all eight patients and mild-to-moderate cold intolerance in two patients. All patients returned to their previous jobs and levels of physical activity, but all had experienced difficulty in carrying out activities of daily living while the fixator was attached. However, the overall satisfaction rate was high with seven of the eight patients rating the result as good to excellent and the other rating it as fair. There were no cases of superficial or deep infection. Our final radiographic analysis showed that the overall contour of the articular surface was fair although the central depression remained (Fig 2). There was definite post-traumatic arthritis in only one instance.
The main complication of this procedure was uncoupling of the fixator assembly due to inadequate bending of the wires for hook formation. This occurred in two cases. In the first instance (case 1), it occurred during a dressing change at 48 h and was managed by immediate recoupling of the fixator and tightening of the hooks. In the second instance (case 4), the patient presented 1 week after neglected disengagement of the fixator. It was felt that reconstitution of the fixator at this late stage of treatment would not alter the final result, and therefore the fixator was removed and the joint was mobilized. It was evident from these two cases that this complication could be easily avoided by proper bending of the wire ends for hook formation and taking care to ensure that the hooks were adequately tight.
DISCUSSION
In the last two decades, there has been a renewed interest in the use of external fixators for the treatment of intra-articular fractures. In hand surgery, this has led to the development of a number of dynamic external fixation devices for the treatment of difficult proximal interphalangeal joint fractures including pilon fractures. These devices range from simple coiled springs to very complex devices such as the compass hinge (Inanami et al., 1993; Suzuki et al., 1994; Allison, 1996; Bain et al., 1998). Functionally, both types of fixators allow active and passive motion of the affected joint while maintaining traction. However, a comparison of the final outcome does not show superior results with the usage of more complex devices (Krakauer and Stern, 1996). Therefore, there is merit in applying techniques that are cost effective, technically simple and can be widely used.
In the treatment of pilon fractures, a dynamic external fixator brings together the beneficial effects of ligamentotaxis and early joint motion that have been used successfully in the treatment of other intra-articular fractures. It helps in the reshaping of the articular surface and does not interfere with the tenuous blood supply of the fracture fragments. It may also promote cartilage repair and prevent joint fibrosis which would restrict movement.
Stern et al. (1991) first emphasized the importance of identifying pilon fractures which have a characteristic mechanism of injury, a fracture configuration with dorsal and palmar fragments of the middle phalanx and no joint dislocation or subluxation. They performed a comparative analysis of three different forms of treatment: splintage, internal fixation and traction. With an average follow-up of 1 year, the best results were achieved with traction using a Banjo frame. In the internal fixation group, 75% of the patients achieved a satisfactory outcome with a comparable range of motion, but 25% of the patients in this group required proximal interphalangeal joint arthrodesis due to complications, including infection and loss of reduction. Extension blocking splintage produced the least successful results, with some degree of pain in all cases and the highest incidence of degenerative arthritis and restricted joint motion. It is important to note that, irrespective of the type of treatment, the final range of motion at the distal interphalangeal joint was reduced. However, this was least affected in the external fixator group.
Innovative techniques have been devised to apply ligamentotaxis and allow early joint motion. Schenck (1986) used a large ring (the size of a pizza pan) attached to a forearm extension and attached the affected finger to this by rubber bands. This construct allowed passive finger flexion and extension at regular intervals and an average 87° arc of motion was achieved. However, the device was cumbersome and had to be worn for a mean of 7 weeks.
A simple type of dynamic external fixator was devised by Gaul and Rosenberg (1998) and was successfully used to treat two pilon fractures. This fixator consisted of two fine K-wires: the distal wire passed through the distal half of the middle phalanx and was bent to 90° to run along the finger while the proximal wire was drilled across the head of the proximal phalanx. The ends of both wires were then bent into hooks and engaged to allow distraction and provide lateral stability during motion and at rest. Hynes and Giddins (2001) also used a fixator made from a similar configuration of K-wires. Eight patients with pilon fractures were acutely treated with the modified Gaul’s fixator and their results were good with an average range of motion from 12° to 88°. The fixators were retained for an average of 6 weeks. However, two of these patients developed sepsis at the proximal wire/bone interface. Other authors have also reported a high incidence of pin track sepsis around the proximal wires which can lead to serious complications (Bains et al. 1998; De Soras et al., 1997).
We feel that high rates of pin site infection may occur for three reasons. Firstly, the use of straight K-wires will result in static longitudinal traction and any active or passive motion will result in rotation of the wire at the proximal bone/wire interface, rather than the wire coupling, leading to loosening and sepsis. This was demonstrated by Allison (1996), who used a dynamic fixator made of stainless-steel spring wire with windings hooked around 2 K-wires. None of their patients developed any pin site infections. The reason for this could be that, during finger motion, the spring moved independently and did not interfere with the proximal and the distal wire–bone interfaces. The second reason for high infection rates in the previous series may be that the cancellous bone of the proximal phalangeal metaphysis is not strong enough to resist the torque generated at the bone–wire interface. This problem is further compounded if the wire is not placed in the centre of rotation of the proximal interphalangeal joint and interferes with the attachments of the collateral ligaments (Hastings and Ernst, 1993). This again leads to wire loosening and sepsis. The third reason why other studies may have experienced high infection rates is that the fixator was unnecessarily retained for up to 6 weeks: this may increase the risk of pin site infection. With these points in mind, the design of our fixator was created and the duration of fixation was shortened.
To prevent proximal pin tract sepsis, we also elected to reverse the configuration of the fixator so that the hinge lay at the proximal wire, allowing motion without interference to the wire/bone interface in either phalanx. The distal wire, which formed the long lever arm, was given a dorsal parabolic curve and was preloaded to maintain constant distraction and thus allow improved biomechanical distribution of forces and impart springiness to maintain fracture reduction during motion.
Our average follow-up of 2.2 years is not long enough to predict the final incidence of osteoarthritic changes in the proximal interphalangeal joint. However, it was encouraging to note that only one case showed signs of early articular degeneration. This was the case where the fixator was removed prematurely due to neglected unhooking of the K-wires. Overall, we feel that this is a simple and cost-effective technique that produces results comparable to other dynamic fixation devices. Both junior and senior surgeons can easily be taught to assemble this fixator which utilizes only freely available components.