A novel bone graft technique combined with plating for aseptic recalcitrant long bone nonunion

Although conservative treatment and surgical techniques have greatly improved, the incidence of nonunion remains at close to 5–10% [1]. The incidence rate of nonunion in most long bones is > 10% [2]. Bone nonunion that meets the diagnosis displays poor self-healing and normally requires therapeutic intervention. In recent years, many treatment methods of bone nonunion have been reported, including plate and intramedullary nail with bone graft, external fixation, biological stimulation, and pulsed ultrasound [3‐7]. The prognosis for patients who received these treatments was satisfactory. Therefore, the majority of nonunions can be effectively managed with these techniques.

Clinical orthopedic doctors still encounter patients with recalcitrant long bone nonunion. This is frequently accompanied by unstable soft tissue coverage, poor osteogenesis, and severe disability. Long-term nonunion leads to patients displaying depression, anxiety, reduced psychological resilience, and other psychological traumata. Repeated treatment also represents a huge economic burden to society and the patients’ families [8, 9]. Therefore, it is necessary to ensure the early healing of nonunion.

The authors’ previous research reported that the nonunion ends were determined as consisting of one osteogenesis deactivation zone (scar tissue and hardened sequestrum) and two osteogenesis activation zones (normal porosis tissue). The authors used a new bone graft technique termed the bone-forming channel technique, which spans the osteogenesis deactivation zone to bridge the osteogenesis activation zones [10]. Combined with this technology, this study adopted a new surgical strategy and achieved good results in clinical treatment [11‐13]. In the present study, the authors retrospectively analyzed the patients with recalcitrant nonunion to whom this surgical method was applied, with the belief that this surgical strategy can have a good therapeutic effect on such patients, and to provide some data for clinical treatment.

In this retrospective study, 36 patients met the criteria, of whom 2 patients were lost to follow-up, such that finally 34 patients (24 males and 10 females) were enrolled, with a mean age of 49.8 ± 12.3 years (range, 29–75 years). The demographic and clinical data are shown in Table 1.

There were three patients with ulnar nonunion, one patient with ulnar and radius nonunion, nine patients with humeral shaft nonunion, one patient with proximal humerus nonunion (Fig. 2), two patients with supracondylar humerus nonunion, ten patients with femoral shaft nonunion, one patient with supracondylar femoral nonunion, one patient with tibiofibular nonunion, five patients with tibial nonunion (Fig. 3), and one patient with clavicular nonunion. A total of 15 patients sustained an injury from tumbling, 5 from motorcycle accidents, 5 from crashes, 4 from a heavy object, 2 from a high fall, and 3 from a machine, and 4 patients had open wounds (Table 1). Among the classification of nonunion, 11 displayed hypertrophic nonunion, 9 an atrophic nonunion, and 14 synovial pseudarthrosis. All patients received had received one or more regular treatments, and the mean number of previous treatments was 1.82 ± 0.87 times (range, 1–5 times). The mean duration since the first treatment was 14.1 ± 7.7 years (range, 5–37 years).

Patients were followed-up for a mean of 35.6 ± 22.0 months (range, 13–78 months). In total, 32 patients displayed union after 6.8 ± 2.4 months (range, 5–18 months), while 2 patients still had nonunion, a healing rate of 94.1%. The VAS score was 0.7 ± 1 at the final follow-up. The functional results showed that 19 patients were excellent, 11 patients were good, 2 patients were poor, and 2 patients did not heal with the implants intact. Five patients had a > 10° coronal angular deformity, four patients had bone shortening of ≥2 cm, two patients had peripheral nerve injury, and one patient had mild pain at the bone donor site. There was no plate breakage or loosening or incision site infection. The patients’ details of prior treatments and functional results of follow-up are presented in Tables 1 and 2.

There are many risk factors for bone nonunion, including age, smoking, type II diabetes, energy of injury, severity of soft tissue injury, infection, degree of mechanical stability, inherent degree of vascularization around the fracture, and numerous medications [16, 17]. Some studies have shown that the incidence of nonunion is higher in younger subjects, possibly because high-energy injuries usually occur in younger patients [18, 19]. In this study, nine patients had humeral shaft nonunion and ten patients had femoral shaft nonunion. These two fracture sites are normally caused by a high-energy injury. In studying the injury mechanism, > 50% of the patients had suffered a high-energy injury. Therefore, the energy at the time of injury determines the probability of fracture healing.

At present, there is no unified standard approach for bone nonunion fixation, which is mainly determined according to the bone nonunion location, soft tissue conditions, general condition, and infection [20]. External fixation is mostly used in infected nonunion or patients with poor soft tissue [4, 21]. It has the advantages of providing stimulation to the nonunion site percutaneously and simultaneously controls juxta-articular or ‘emmental’ bone fragments, facilitating staged complex bone and soft tissue reconstruction. Intramedullary nailing is most often used to treat fresh femoral and tibial shaft fractures. Some studies have confirmed that intramedullary nailing is effective in treating aseptic long bone nonunion [20, 22]. It has the advantage of minimal exposure of the nonunion end, which may improve bone union and early weight-bearing [23]. However, both external fixation and intramedullary nailing cannot control rotation, and do not allow auto bone grafting. Moreover, because each patient in the present study had suffered a long period of nonunion, most of them required complete synovial pseudarthrosis debridement and deformities correction during surgery, thus external fixation and intramedullary nails may not have been suitable for the patients in the present study.

In the present study, each recalcitrant nonunion had a nonunion duration longer than 10 years and had received one or more failed previous treatments, and the treatment was challenging for surgeons. Therefore, whether the bone had the ability to heal required consideration, as did whether the factors resulting in nonunion could be reversed by surgery. Due to sufficient stability, fracture ends compression, deformities correction, and bone graft permission, LCP combined with auto bone grafting has been successfully reported for the treatment of nonunion by some authors, with a bone healing rate of up to 100% [24‐27]. A detailed preoperative examination and intraoperative observation was conducted in the present study to identify those patients who displayed aseptic bone nonunion. LCP and bone grafting were applied in these patients. This method was not used for those suspected of low-grade infectious nonunion [28]. A positive outcome of this study was that LCP fixation ensured good alignment of nonunion ends after debridement of deformities or synovial pseudarthrosis. In the final follow-up, five patients displayed a coronal angle deformity > 10°, there was no obvious rotation deformity, and the deformity did not affect the function. Another advantage of this study was that the bone-forming channel technique ensured that the autogenous structural iliac bone extended over the osteogenesis deactivation zone and bridged the osteogenesis activation zones, maximizing the osteoinduction and osteoconduction effects of the bone graft, resulting in bone healing. Furthermore, this technology had been successfully used in previous studies, and achieved a satisfactory outcome with a bone union rate of 100% [10‐13]. To the best of our knowledge, this is the first time this technique has been applied to treat recalcitrant nonunion. Moreover, some patients were fixed by two LCPs, with the second LCP fixation in front of the bone graft to potentially avoid bone graft migration, thereby improving the bone union rate. Additionally, two LCPs ensured an absolutely stable mechanical environment of the nonunion ends, such that patients might start early functional exercise without an external brace, resulting in good limb function. At the final follow-up, 19 patients had excellent functional results and 11 patients had good functional results, and all patients with bone healing could return to work.

In this study, the bone healing rate was 94.1%, with non-healing in only two patients, of whom one had previously received five failed interventions. This clearly demonstrated the necessity to apply effective treatment measures the first time when presented with nonunion. Jiang et al. treated five patients with refractory postoperative diaphyseal femur fracture nonunion. All the patients had been previously treated with all the standard surgical procedures for fracture nonunion [27]. All the patients achieved bone union following a combination of double LCP plating and an autologous fibula graft. However, the case numbers of this study were small, and fibular harvest may have caused severe complications. Mittal et al. reported 12 patients with recalcitrant nonunion who were treated with double LCP plate fixation combined with decorticated and cancellous bone graft, which was similar to our surgical strategy [29]. All the patients healed within a mean follow-up of 14 months. This study highlighted the considerable effectiveness of osteoperiosteal decortication in the treatment of fracture nonunion and that the LCP provides fixation without negative effects on periosteal vascularization.

In the present study, the method of using an iliac crest bone graft (ICBG) was improved. As early as 1956, Nicoll proposed an ICBG method termed ‘cancellous insert grafts’ to treat patients with long bone nonunion [30]. The classic method was to use a solid block of cancellous iliac bone to bridge the gap in long bones, and in some cases, cortical iliac bone was used as a lid to cover the cancellous grafts. This method was applied in 27 cases, and union was achieved in all of them in a mean time of 14 weeks. In recent studies, the ICBG normally involved splitting of the iliac bone grafts into longitudinal strips followed by implantation around the fracture site [31, 32]. Although these methods are effective, the bone-forming channel technique further improves the biological environment at the bone graft site. Our method with slotting at the proximal and distal regions of the fracture and implantation of the entire structural iliac bone block with cortex and cancellous bone into the slotted area ensures a tight and accurate fit. The combination of Judet decortication and double LCP fixation maximized the fracture union rates.

The present study was limited by its small number of cases and the lack of a control group. This study did not compare its surgical strategy with other treatments. Because this study was retrospective, the fracture healing time could only be judged according to the results of the patient review, which may have been inaccurate. Despite the limitations of the study, the results demonstrated that this novel technique of autologous iliac bone grafting has a good healing rate in patients with recalcitrant bone nonunion. In future research, our sample library will be further expanded, which will provide greater evidence, bringing hope to patients with recalcitrant bone nonunion and reducing the burden of this disease on society and patients’ families.

This study recommends its described surgical strategy for aseptic recalcitrant long bone nonunion. Judet decortication preserved the blood supply of the fractured ends, double LCP provided mechanical stability, and the bone-forming channel technique provided a favorable biological environment. This method can improve the fracture union rates and rescue many patients with aseptic recalcitrant long bone nonunion. Although the outcome of some patients was not optimal, fracture healing still greatly improved the situation for patients with long-term bone nonunion.