Treatment of aseptic diaphyseal nonunion of the lower extremities with exchange intramedullary nailing and blocking screws without open bone graft

Abstract

Objective:  To report the preliminary results of the treatment of aseptic diaphyseal nonunion of the lower extremities with exchange nailing plus blocking screws.

Methods:  Between June 2005 and September 2008, twelve patients with diaphyseal nonunion in the lower extremities (femur in five patients and tibia in seven; hypertrophic nonunion in eight patients and atrophic nonunion in four) were treated by reaming, exchanging the original intramedullary nail with a larger one, and using blocking screws, and the therapeutic effect assessed.

Results:  All patients were followed up for 1–2 years (average, 1.5 years) to assess union. Bony union was achieved in all patients within 4.7–13.5 months (average, 7.8 months). All patients were pain free without any complications by the last follow‐up.

Conclusion:  The therapeutic method of exchanging the nail combined with blocking screws is effective for aseptic nonunion of the lower extremity after intramedullary nailing.

Introduction

Intramedullary nailing is a widespread technique for the treatment of femoral and tibial diaphyseal fractures. Nevertheless, whether diaphyseal nonunion should be reamed or not is frequently disputed. There are a number of modalities for treating long bone nonunion, including additional plates and screws, exchanging the nail, external fixation, and bone grafting. Generally, exchanging the nail is the preferred technique for stabilizing long bone fractures with diaphyseal nonunion. Successful results have been reported for nonunions of the tibial and femoral shafts ¹ , ² .

However, not all authors have achieved similar results. The union rate has been reported to range from 72% to 100% in the femur and 76% to 96% in the tibia ³ , which suggests that exchanging the intramedullary nail alone might be far from solving all problems in the treatment of nonunion. Here we propose a modified method of exchanging the intramedullary nail combined with blocking screws for the treatment of aseptic nonunion in the lower extremity after intramedullary nailing.

Materials and methods

General information on the patients

Between June 2005 and September 2008, twelve patients with nonunion in the lower extremities were enrolled in this study, including ten men and two women aged from 28 to 45 years (average, 35.6 years). Nonunion was defined as neither clinical nor radiological evidence of healing of fracture at least 9 months after the primary intramedullary nailing. There were five femoral diaphyseal nonunions (one isthmic nonunion with deformity and four infra‐/supra‐isthmic nonunions) and seven tibial diaphyseal nonunions (infra‐/supra nonunions). Nonunion was classified based on the configuration of the main fracture gap. There were eight hypertrophic nonunions and four atrophic nonunions. The bone defects in the four atrophic cases were less than 50% in circumference and 2 cm in length. The coronal deformity was 15° in one tibial nonunion and 14° in one femoral nonunion. No rotational deformity was found in any case.

The primary fractures were classified according to the AO‐classification (Arbeitsgemeinschaft fur Osteosynthesefragen‐ classification) as follows: type A in four, type B in seven, and type C in one. All five femoral fractures were closed injuries, while of the tibial fractures four were closed and three open. The cause of injury included traffic accidents in seven cases, falling from a height in three, and lesser falls in two. The primary intramedullary nailing included interlocking nailing in eleven cases and expandable nailing in one. The mean time from the primary fracture to our intervention was 19 months (range, 9–36 months).

All cases had no history of infection and no patients had elevated erythrocyte sedimentation rates or elevated C‐reactive protein before exchange of their intramedullary nails.

Operative techniques

For nonunions without deformity, after removal of the initial nail we used sequential hand reamers to clear the nonunion site for easier passage of a guide wire. Then we over reamed the medullary canal under the guidance of the wire to 2.5–3 mm larger in diameter than the removed nail. A nail 1–2 mm in diameter larger than the original one was inserted (six tibial and five femoral nails were manufactured by Smith and Nephew Richards, Memphis, TN, USA; one tibial nail was made by Chuang‐sheng, Chang Zhou, China. Thereafter, a C‐arm photographic check was performed in two orthogonal planes to ensure: (i) an adequate fit between the nail and the canal; (ii) sufficient length of nail in the shorter distal or proximal fragment; and (iii) correct alignment of the extremity. All nails were locked statically. No less than two locking screws were placed in the shorter fragment. Finally, two blocking screws were placed adjacent to the nail on the coronal plane according to the potential translation direction of the shorter fragment. In all cases, the placement of blocking screws was performed under the guidance of image intensification using a Kirschner wire (Shanghai Medical Instruments, Shanghai, China) as a drill. When the Kirschner wire met the nail during drilling, it was tilted slightly and further advanced until it had perforated the opposite cortex. Then a cortical screw was substituted for the Kirschner wire. The two blocking screws were placed as far apart as possible in the shorter fragment to obtain maximal stability.

For nonunions accompanied by deformity, as shown in Fig. 1, in which the path of the original nail had caused fracture malalignment, we placed thick Kirschner wires to aid reaming after the passage of the guide wire thus creating a new, correctly aligned path. The location of the Kirschner wire was determined based on displacement of the shorter fragment. In tibial nonunion with deformity, we recommend placing the Kirschner wire before hand reaming in order to avoid perforation of the cortex in the proximal fragment. Cortical screws were substituted for the Kirschner wire as a support device before final insertion of the nail. The use of blocking screws guides the reamer in the correct direction and prepares the right path for the nail.

Figure 1.

This femur fracture had a tendency toward lateral displacement and angulation (a). Lateral radiograph showing the locking screw has broken. (b) Anteroposterior radiograph showing the locking screw has broken. (c) Anteroposterior radiograph after revision with interlocking nail and blocking screws showing the deformity has been corrected. (d) Postoperative lateral radiograph.

One distal tibial nonunion with deformity required fibular oblique osteotomy because the deformity could not be corrected in a closed manner.

Postoperative management

On discharge from the hospital, all patients were allowed to exercise lying down without weight bearing. Patients were not permitted to increase their weight bearing until bridging callus was apparent on their X‐ray image, which usually did not occur until at least 6 weeks after the operation. The nonunion was considered healed if there was no tenderness at the fracture site, full weight bearing was tolerated without pain, and the radiograph demonstrated bridging callus across the nonunion site on orthogonal views. No patients were lost to follow‐up prior to achieving union.

Results

The average time of follow‐up was 18 months (range, 12–24 months). All patients achieved union without a secondary procedure. The overall union rate was 100%.We defined union of a fracture clinically as having no pain, no tenderness, and no need of aids to assist ambulation; and radiographically as a solid callus with sufficient cortical density to connect both fragments. The mean healing time was 7.8 months (range, 4.7–13.5 months). By the time of union the deformity in the coronal plane was 0° in the two cases who had coronal deformities preoperatively. There were no wound infections or malunions (with angulation >5°, rotational deformity >10°, or shortening >2 cm).The ranges of movement of the hip, knee and ankle were close to normal in all cases.

Case

A 45‐year‐old man sustained a distal tibial fracture in a traffic accident. He was treated by open reduction, steel wire cerclage, and tibial interlocking intramedullary nailing. One year after the initial operation dynamization was performed due to nonunion. The patient was then allowed full weight bearing. Unfortunately the nail broke one year after the dynamization procedure. The X‐ray film showed nonunion and a 15° varus deformity in the coronal plane. He underwent removal of the device, fibular oblique osteotomy, closed reduction with blocking screws aid, and fixation with a TriGen tibial nail (Smith and Nephew Richards, Memphis, TN, USA). Eventually the nonunion healed seven months after the revision surgery (Fig. 2).

Figure 2.

(a) Preoperative radiographs of a 45‐year‐old male patient who had sustained a distal tibial fracture, which was then treated by open reduction, steel wire cerclage, and a tibial interlocking intramedullary nailing. (b) Radiographs two years after the initial operation showing the nail has broken due to nonunion. (c) Post‐revision surgery radiograph with exchange nailing and blocking screws. (d) Radiographs 7 months after revision surgery with exchange nailing and blocking screws (arrow) showing the nonunion has healed.

Discussion

With development of the interlocking technique, the indications for intramedullary nailing have been expanded from merely isthmic tibial and femoral fractures to include fractures in the distal and proximal regions of the long bones. In recent years, intramedullary nailing has become the standard treatment for diaphyseal tibial and femoral fractures. However, nonunion or delayed union is still frequently reported.

The usual factors determining the occurrence of nonunion or delayed union are: (i) the primary state of the injury; (ii) the morphology and location of the fracture; (iii) the pathophysiological state of the patient; and (iv) the therapy for the fracture. The first three factors are beyond the surgeons' control. As for the fourth factor, the so‐called iatrogenic factors including poor reduction, inadequate immobilization, and excessive soft tissue stripping should be cautiously avoided. Unfortunately, some of these treatment principles are occasionally neglected as in the case described above. Once nonunion is detected, a successful treatment strategy for long bone aseptic nonunion should be based on three goals: (i) achieving mechanical stability of the bone fragments; (ii) augmenting the potential for bone healing; and (iii) correcting any alignment deformity. The question is whether there is an ideal salvage method for nonunion.

Exchange nailing has been used successfully for the treatment of aseptic nonunions in the femur and the tibia. It offers both mechanical and biological advantages ³ . A large nail can provide greater bending rigidity and strength than a small nail. Furthermore, reaming of the medullary cavity not only clears the obliterated ends of the bone fragments, which facilitates formation of vascular anastomoses, but also introduces new active tissue, consisting of medullary pulp and bone chips formed during reaming, into the site of nonunion. These tissues provide a source of cells and growth factors. Reaming of the medullary canal increases periosteal blood flow and stimulates the formation of new bone. For aseptic nonunion (whether hypertrophic or atrophic) of the lower extremity without substantial bone loss, exchange nailing seems to be the most appropriate method with a reunion rate ranging from 72% to 100% in the femur and 76% to 96% in the tibia ³ . Despite the high success rate, not all cohorts achieve equivalent results, which suggests that exchange nailing alone is not enough.

Recently, Niedźwiedzki et al. reported unfavorable results for exchange nailing ⁴ . He claimed that additional procedures are required. We made a critical analysis of the X‐ray images in his paper and found that all three failed cases had supra‐isthmic or infra‐isthmic nonunion. We concluded that instability was possibly the main reason for the revisions failing.

Similarly, stabilization of proximal and distal fractures is associated with a high incidence of malalignment and nonunion. Various factors contribute to the problems including: (i) a striking mismatch between the implant and the metaphysis with no nail–cortex contact; (ii) muscular forces decreasing the stability of the juxta‐articular fragment and displacing the fracture; and (iii) a single‐plane interlock resulting in translation of the nail along the interlocking screws ⁵ . Mechanical stability can usually be improved by increasing the length of the nail and the number of interlocking screws. Interlocking screws have been proved to be good at controlling the limb length and rotation of the fragments under a static mode. However, interlocking screws alone do not provide enough stability in cases with a tendency toward lateral displacement and angulation, which is usually the case with supra‐isthmic or infra‐isthmic fractures.

Blocking screws, placed adjacent to the nail, physically block transverse nail translation, and increase the mechanical rigidity of the bone–implant combination ⁵ . Blocking screws provide a point of support for the short fragment on the side of the nail, making it possible to control displacement and angulation in a dynamic mode. The efficacy of blocking screws in femoral and tibial fractures has been proved by various studies ⁶ , ⁷ . Encouraged by successfully using blocking screws in fresh fractures, we applied the blocking screw technique to the stabilization of infra‐isthmic and supra‐isthmic nonunion without deformity to maintain stability, and achieved the preliminary success reported in this study.

Blocking screws not only maintain stabilization, but also aid reduction during surgery. When the plan is to use a nail again in the situation of nonunion with deformity, there is a strong tendency for the new nail to slip into the preformed path of the old nail despite the fact that new starting points have been created. Furthermore, in tibial nonunions, the ends of the fragments are frequently sclerotic and misaligned. As a result of the misalignment, attempting to open the sclerotic part of the proximal fragment can result in perforation during hand reaming ⁵ . Blocking screws are placed to prevent the nail from reentering the old nail path. By narrowing the medullary canal in the metaphysis, a tight mechanical fit between the nail and the canal is provided and the wrong nail path is blocked. During insertion of the new nail, the nail abuts the blocking screws and therefore enters the new nail path with correct alignment ⁵ . In our study, we used thick Kirschner wires to guide the nail centrally into the short fragment. This ‘canal control’ technique played a crucial role in the correction of deformity (shown in the operative techniques).

The indications for open bone grafting during exchange nailing remain unclear, and no consensus can be found in the literature ⁸ , ⁹ . An autograft is considered ideal for grafting procedures, delivering a combination of osteogenic stem cells, osteoinductive growth factors, and an osteoconductive scaffold. However there are limitations regarding donor site morbidity. An allograft, on the other hand, holds the risk of disease transmission. Synthetic graft substitutes lack osteoinductive or osteogenic properties. Composite grafts combine scaffold properties with biological elements but they are costly. Some surgeons use a graft when there is pre‐operative evidence of osteoporosis, but no significant difference in the healing rates, compared to those who did not receive a graft, has been proved ⁸ . These authors concluded that the use of open bone graft was not required to achieve good results ⁸ .

Many authors agree that exchange nailing is indicated both in hypertrophic and atrophic long bone nonunions ³ , ⁹ . Exchange nailing for an avascular nonunion may stimulate a healing response and augment mechanical stability. Exchange nailing for a hypertrophic nonunion augments mechanical stability, which is the main requirement for achieving osseous healing. But in the presence of bone loss (bone loss >50% of the cortical diameter or length >2 cm), or comminuted fractures, nonunions do not respond favorably to exchange nailing alone. Based on suggestions in the literature ³ , ⁹ , ¹⁰ and our clinical results (union rate 100%), we assume open bone graft is necessary only when (i) the initial revision procedure has failed; (ii) there is nonunion with marked bone loss or comminuted fractures. As described in the case report above, we carried out a fibular osteotomy and removed the failed tibial wire cerclage and broken nail, but we did not use a bone graft as well, as some surgeons recommend.

Our retrospective study has some limitations, the principal ones being the small number of study subjects and the lack of a control group. Nonetheless, our study suggests that treatment of nonunion of the lower extremities with exchange nailing and blocking screws provides preliminary good results.

In summary we recommend the method of nail exchange combined with blocking screws for the treatment of selective aseptic nonunion of long bones. We believe it to be especially useful for infra/supra‐isthmic nonunion with or without deformity. Although no consensus so far has been reached in the literature, we assume that the aforementioned technique is not appropriate for nonunion with massive bone defect (more than 50% of the circumference or 2 cm in length ⁸ ), or nonunion with rotational deformity. Under such circumstances other techniques, including open bone graft and an additional plate, should be considered.

Disclosure

No author or related institution has received any financial benefit from research in this study.