Immediate Corticocancellous Bone Autografting in Segmental Bone Defects of the Hand

INTRODUCTION

High-energy crush and avulsion injuries to the hand often result in type III (A, B and C) open fractures, which may include arterial injury, segmental bone loss and extensive soft tissue damage (Freeland and Jabaley, 1988). The classification system described by Gustilo (1987) for open tibial fractures can be applied to the upper extremities to help guide management and prognosis. Type IIIA fractures include any open fracture resulting from high-energy trauma but with adequate soft tissue cover despite, extensive skin flaps raised by the injury. Type IIIB fractures have inadequate soft tissue coverage with periosteal stripping and type IIIC fractures are associated with an arterial injury requiring repair. These complex injuries require a treatment plan that re-establishes normal bony anatomy and promotes early healing and rehabilitation to maximize function. The benefits of early bone grafting in treating hand fractures are recognized throughout the medical literature. These include faster bony union time, less soft tissue contracture and less immobilization, which ultimately lead to better hand function (Calkins et al., 1987; Gonzalez et al., 1993, 1998; Rinaldi, 1987; Segmüller, 1981; Sundine and Scheker, 1996).

Early bone grafting has evolved from being a secondary procedure to a delayed primary procedure (Calkins et al., 1987; Freeland and Jabaley, 1988; Gonzalez et al., 1993, 1998), and, ultimately, to a primary bone grafting procedure (Rinaldi, 1987; Stahl et al., 1999). Both Rinaldi (1987) and Stahl et al. (1999) described successful primary bone grafting of autologous corticocancellous segmental defects of the hand. In their series, Stahl et al. (1999) described three patients who underwent immediate corticocancellous bone grafting for hand defects ranging from 0.5 to 2.0 cm. No infections were observed, and all patients experienced bony healing with good cosmetic results. These authors advocated immediate bone grafting if prerequisites such as adequate wound débridement and appropriate soft tissue coverage can be guaranteed. Rinaldi (1987) reported one case of immediate post-traumatic bone grafting in his series of 21 patients. In this case, an unsalvageable amputated index ray was used to reconstruct the avulsed middle phalanx of the middle finger. No infection occurred, and results 9 years after surgery were fair, considering the severity of the original injury. Successful immediate bone grafting in combination with emergency free-flap procedures to reconstruct complex upper extremity injuries have been described by Ninkovic et al. (1995), Scheker et al. (1993), Sundine and Scheker (1996) and Tropet et al. (1997).

The purpose of this retrospective analysis was to determine the effectiveness of immediate autologous corticocancellous bone grafting as a treatment for type III hand fractures with segmental bone defects caused by high-energy crush and degloving injuries.

PATIENTS AND METHODS

Seven patients with type III open hand fractures with segmental bone loss presented for treatment from November 1987 to February 2000. The fractures were caused by high-energy avulsion or crush injuries and all of them had associated soft tissue injuries. The study was performed with the approval of the institutional review board of the Christine M. Kleinert Institute of Hand and Microsurgery, Inc., Louisville, Kentucky.

All seven patients in this study were male. The mean follow-up time was 14 (range 3–35) months. The mean patient age was 39 (range 22–58) years. Four of the seven patients experienced an injury to their dominant hand. Type III fractures resulted in segmental bone loss of the phalanges in five patients (Figs 1 and 2) and of the metacarpals in two patients. Combined fractures to more than one digit or metacarpal occurred in five of the seven patients. Multi-level fractures with intraarticular involvement occurred in four of the seven patients (see Table 1).

The mechanism of injury was digital crush in five patients and dorsal degloving in the metacarpal area in two patients. The two patients with dorsal hand degloving injuries suffered comminuted metacarpal fractures with associated extensor tendon avulsion and dorsal skin loss in the metacarpal region (cases 4 and 5, Table 1). Five patients had bone loss in more than one area. All fractures requiring bone grafting were associated with severe soft tissue injuries (Table 2). These injuries consisted of one or more of the following: digital devascularization (one case), complete digital artery division (five cases), complete digital nerve division (three cases), complete extensor tendon laceration (three cases), complete extensor tendon avulsion (two cases) and dorsal skin degloving of the hand requiring lateral arm free-flap cover (two cases).

Surgery was performed within 24 hours of the initial injury in all cases. First, the patient’s extremity was examined in the emergency room and assessed for sensation, vascularity, soft tissue injury and bony injury. Either an axillary block or general anaesthesia was used. Pre-operative and intraoperative antibiotics as well as tetanus prophylaxis were administered to all patients. Our treatment protocol, which has been recommended by several authors (Lister and Scheker, 1988; Scheker et al., 1993; Sundine and Shecker, 1996), consisted of radical wound débridement under tourniquet control with excision of all non-viable or questionably viable tissue.

Débridement was begun at the periphery of the wound, out of the zone of injury, to identify a clear plane between injured and uninjured tissue. This approach is likened to the extirpation of a tumour in which the wound is excised en bloc, leaving behind only healthy tissue. Radical débridement also allows better identification of the extent of injury. Small bony fragments providing no structural support were débrided, whereas bony fragments with viable periosteum or attached articular cartilage were left in place. Débridement of all questionably viable tissue is critical to convert a contaminated wound into a clean one, minimizing the risk of infection, as originally expounded by Godina (1986). Once débridement was considered complete, the tourniquet was released and all areas that did not demonstrate obvious perfusion were re-excised under tourniquet control. Following débridement, the wound was examined to establish the extent of injury and evaluate the tissues which would require repair or replacement.

Segmental bone defects were reconstructed primarily with an iliac crest corticocancellous bone graft in seven patients (see Figs 1–5). In one patient, bone was harvested from the distal radius (case 1, Table 3). Internal fixation of phalangeal fractures and bone grafts was achieved using one or more of the following: Mini-condylar plates (1.5 mm) 0.045 inch K-wires, intraosseous wires or 2.0 mm lag screws. Fixation of metacarpal fractures and bone grafts was achieved using one or more of the following: 2.7 mm plates and screws, cerclage wires or 0.045 inch K-wires.

Primary wound closure was achieved in all cases and was an absolute prerequisite for open reduction and internal fixation with immediate bone grafting. Primary wound closure was possible in five patients, whereas two patients required a lateral arm free flap for skin cover. Both cases requiring flaps followed a dorsal degloving injury in the metacarpal region (case 5, Figs 6–9).

The details of treatment and recovery for each patient are shown in Table 3. The results of treatment were interpreted according to clinical and radiological time of bone healing, in which clinical union was defined as the absence of pain during fracture manipulation on physical examination. Radiological bony union was defined as trabeculae seen crossing the fracture interface in both the anterior–posterior and lateral plain X-ray views.

Other outcomes measured included total active motion (TAM), grip strength and complications, including the rate of re-operation or amputation, rate of infection and time to return to regular work duty.

The mean TAM was measured in all affected digits. Range of motion data from the unaffected hand were unavailable for comparison. Therefore, the standard TAM values defined by the American Association of Orthopaedic Surgeons were used as the control. TAM is defined as the total active range of motion, minus extensor lag, for the metacarpophalangeal, proximal interphalangeal and distal interphalangeal joints combined. A one-sample t-test was used to compare the mean range of motion of the study group against a control value of 260° as defined by the American Association of Orthopaedic Surgeons, and a P-value of<0.05 was considered to be statistically significant.

Grip strength was measured and compared with that of the uninjured, contralateral hand using the Jamar Dynamometer (Sammons Preston Rolyan, Jackson, Michigan, USA) set at the second ring and with the elbow flexed at 90°. All recordings were made by an independent hand therapist at final follow-up. A paired sample t-test was used for statistical analysis, and a P-value of<0.05 was considered to be statistically significant.

Two-point discrimination of all affected digits was also evaluated by an independent hand therapist at final follow-up. Testing for two point discrimination was measured on the palmer–radial and palmer–ulnar aspects of the fingertips.

RESULTS

Eleven of the 12 bone grafts performed were successful, with a final bone union rate of 92%. The mean time to bony union was 18 (range, 11–28) weeks. The only case of bony non-union occurred following a crush injury to the index finger which had segmental bone loss to the middle phalanx with fracture extension into the distal interphalangeal joint. This was a complex injury compounded by associated extensor tendon and digital artery lacerations which were also repaired. One patient required bone stimulation 6 weeks after surgery because of delayed union and achieved clinical union after 3 months.

The final range of motion, grip strength and two-point discrimination in all injured digits is described in Table 4. The average TAM in bone-grafted digits was 178±53° at final follow-up compared with the normal control value of 260° (American Association of Orthopaedic Surgeons) (P = 0.001). Only two of the seven patients in this series had a TAM that was comparable with that of the control (cases 5 and 7). Such decreases in range of motion can be expected following major type III open fractures to the hand.

The mean grip strength was 29.8±10.3 kg in the injured hand compared with 63.4±11.3 kg in the contralateral unaffected hand and this difference reached statistical significance (P = 0.002).

The average two-point discrimination was 6.7 mm for the radial digital nerve and 7.2 mm for the ulnar digital nerve measured at the palmer fingertip level of the salvaged injured fingers.

Only patients with combined extensor tendon and complete digital artery division at initial injury required revision amputation or joint fusion. None of the patients with either skin and extensor tendon loss or neurovascular injury alone required further surgery. Indications for revision surgery included severe proximal interphalangeal joint flexion contracture and stiffness (three digits). Revision amputation at the level of the initial fracture site was required in one finger in each of two patients; case 3 at 14 days and case 7 at 543 days following initial surgery. One of these patients (case 3) had experienced a severe crush injury to the middle and ring fingers and both injured digits required corticocancellous bone grafting and soft tissue reconstruction. Two weeks following his initial surgery, this patient underwent revision amputation of his left ring finger at the proximal phalangeal level for a non-viable finger despite two previous attempt at revascularization. The amputated ring finger had initially required corticocancellous bone grafting, extensor tendon repair and venous and arterial repair with interpositional vein grafting for revascularization. On the second postoperative day, the ring finger became cyanotic and the patient was returned urgently to the operating room for arterial and venous anastomotic revision and vein grafting which later failed. Malrotation of the middle finger was also noted in this patient following treatment but he refused further corrective surgery. The second patient (case 7) who underwent revision amputation of his index finger had suffered an open comminuted fracture with segmental bone loss to the middle phalanx of his index and ring fingers, as well as a displaced proximal phalangeal fracture of the middle finger which required internal fixation. He also had extensor tendon and digital artery division to the index finger and an extensor tendon laceration to the middle finger. He developed a middle phalangeal bony non-union and distal interphalangeal joint instability, which were treated by Acutrak screw fixation and distal interphalangeal joint fusion, respectively. Despite aggressive therapy, he continued to have limited function and requested an amputation. One patient (case 6) had his corticocancellous bone grafting fixation hardware removed and underwent secondary proximal interphalangeal joint arthroplasty because of joint stiffness. This was followed by fusion, at the patient’s request, because of a lack of useful function. His initial soft tissue injuries also included extensor digitorum communis and digital artery divisions. Although the two lateral arm free flaps survived, one of the flap donor sites required secondary skin grafting because of a delay in wound healing.

The infection rate in this series was 0%.

The average time of return to regular work, including the recovery time after secondary surgeries, was 5 months and 21 days (range 1–11 months). All patients were able to return to their pre-injury employment without any limitations.

DISCUSSION

The rationale for primary bone grafting of complex hand injuries was derived from maxillofacial surgery, where immediate bone grafting following trauma is advocated. Hallock (1995) preferred early bone grafting and soft tissue reconstruction to multi-stage procedures in gunshot injuries to the face to avoid secondary and often uncorrectable deformities caused by soft tissue contracture. Successful primary one-stage management of Gustilo type III bony injuries also has been described in the treatment of lower extremity injuries. Tropet et al. (2000) treated five patients with type IIIB open tibial fractures. They described an aggressive emergency management of these injuries, including radical débridement, intramedullary nailing and iliac crest bone grafting. In their series, only one superficial infection occurred, with no non-unions, and the mean healing time was 8.5 months.

Immediate bone reconstruction, combined with radical débridement and primary soft tissue reconstruction provides the most favourable circumstances for healing and functional recovery of the seriously injured hand (Chen et al., 1992; Freeland and Jabaley, 1988; Godina, 1986; Gupta et al., 1999; Lister and Scheker, 1988; Ninkovic et al., 1995; Scheker, 1994; Scheker et al., 1993; Sundine and Scheker, 1996; Tropet et al., 1997). According to Stahl et al. (1999), immediate stable internal fixation combined with corticocancellous bone grafting of missing bone preserves normal bony length and allows intrinsic muscles to function under normal tension. It also optimizes venous and lymphatic return, reduces the risk of soft tissue oedema and eliminates dead space, which can preclude the risk of infection. Restoring the normal integrity of the skeleton, through corticocancellous or cancellous bone grafting, also provides a construct for soft tissue reconstruction.

All 12 bone grafted digits in this series were salvaged. Two digits later required secondary amputation (cases 3 and 7) and one digit required proximal interphalangeal joint fusion (case 6) because of limited function and pain. These three cases were all severe crush injuries involving at least a digital artery (with or without devascularisation) and tendon injury. The only non-union in this series was also found in this group (case 7) and required amputation because of pain and stiffness, despite healing following open reduction and internal fixation with an Acutrak screw. Limited perfusion of the traumatized soft tissue envelope, compounded by vascular injury, may increase the risk of non-union. As expected, poorer function and range of motion were also found in this high-risk group. A better outcome was seen in the four other bone graft cases (cases 1, 2, 4 and 5), which were associated with either a combined skin and tendon injury (cases 4 and 5) or a neurovascular injury alone (cases 1 and 2). In selective cases, severe crush injuries involving segmental bone loss associated with two or more soft tissue components could warrant primary amputation. This is especially true when vascular injury results in digital devascularization or when near total amputation is present. In retrospect, cases 3 and 7 could have been considered for primary amputation to facilitate rehabilitation and expedite an earlier return to work. However, the benefits of digital salvage in severe cases should be individualized and not overshadowed by consideration of high risks of functional impairment and pain. This can be attested by the fact that all patients in this series were able to return to their pre-injury employment with no pain at final follow-up.

Complex injuries require an equally complex and aggressive management plan to maximize function and limit long-term impairment. Our treatment protocol included radical wound débridement, primary bone grafting and immediate soft tissue reconstruction. This study illustrates that primary bone grafting of segmental hand defects resulting from high-energy crush or avulsion injuries is now a reliable option. Immediate reconstruction can be done safely as long as radical débridement and adequate soft tissue cover of bone is also performed. The 92% bony union rate, 0% infection rate, and 100% return to pre-injury employment found in this series is encouraging and supports the use of primary bone grafting in an acute setting. The major advantage of primary reconstruction is the ability to provide early and aggressive postoperative rehabilitation to lower the morbidity often associated with these devastating injuries.