Advantages and disadvantages of AST and BT from previous literature
Currently, Ilizarov reconstructions, the Masquelet technique, vascularised and non-vascularised bone grafts and bone substitutes are the main methods to treat tibia defects [2, 8, 9, 17-22]. However, bone transfer is the preferred technique for the treatment of infected tibial bone defects [1, 4, 23, 24]. Ilizarov reconstruction techniques include two main clinical treatment protocols: bone transport and acute shortening and gradual lengthening [11, 25]. Bone transport is a safe and reliable approach of tackling segmental tibia bone defect. It can simultaneously repair bone defect and soft-tissue defect. It has the advantages of quick wound healing, shortened treatment duration, less bone grafting and reliable treatment efficacy [26, 27]. However, postoperative complications are common, such as bone exposure and bone non-union usually along with axis deviation in long segmental bone transport, consolidation of newly formed bone is poor, delayed union or non-union can occur at the docking site, and pin track infection and screw loosening, stiffness of the knee and ankle joint foot drop can occur [5]. Many studies of the ASD technique have shown that it has obvious advantages and can significantly shorten the time of union [24, 28-30]. It reduces or closes the wound, effectively reduces the soft-tissue tension, and reduces the incidence of postoperative bone infection, bone exposure, osteonecrosis and soft-tissue necrosis; it is especially suitable for patients with large wounds [31-37]. However, according to two studies [38, 39], it may cause vascular and nerve injury and require more bone grafts and a limited shortening distance. At present, there are many comparative studies on bone transport and acute shortening technique in the treatment of infected tibial bone defects, but no conclusion has been reached. As far as the author knows, the present study is the first meta-analysis about the issue.
Outcome analysis
EFI
The present meta-analysis showed that AST was superior to BT from the aspect of the EFI, which is an effective index to evaluate the treatment of bone defect and non-union with the Ilizarov technique, which is closely related to age, pathological characteristics, osteotomy position, elongation speed and bone defect length [40, 41]. Many studies reported that the EFI of the bone transport group ranged from 0.87 to 2.8 months/cm [4, 5, 23, 42] compared to 1.2–2.5 months/cm in the acute shortening group [24, 30, 32, 33, 35, 36, 39, 43]. In the present study, significant difference was detected in the two groups in terms of the EFI (P <0.05) (Fig. 4), which means the EFI of the BT group was higher than that of the AST group. This result indicated the advantage of the AST group in treatment duration, which is consistent with the current mainstream literature.
Bone union time
A bone defect area is always filled with soft tissue just because bone ends cannot reach the docking site in time, which may consequently affect bone union time. The AS technique can bring forward and solve the problem of non-union because it shortens the duration of bone defect ends’ contact and performs bone grafting at an early stage [12]. Kemal et al. reported on a study of 24 patients with mean defects of 7.01 cm. They reported an average bone union time of 275.5 ± 70.6 days [5]. A study of 31 cases reported the mean time to union as 40.1 weeks (12.6–80.7 weeks) [32]. The mean healing index in another study was 30 days/cm [33]. MP Magadum et al. described the mean lengthening achieved as 10 cm, and mean union time was 6.3 months in a study of 27 patients with infected non-union and large bone defects in the tibia [30]. In the meta-analysis, no difference was detected in the two groups, according bone union time (P >0.01). Some studies show that multiple-level bone transport can significantly decrease bone union time [44]. Results may also be affected by the different bone transport modalities used in the included studies. Some studies believe that docking site union is the key factor that affects the whole therapeutic time, and the AS technique is more advantageous in shortening the docking site union time [24, 28-30]. Therefore, the bone union time of the AS group may be shorter. However, bone union may be affected by many factors, such as the severity of the original injury and infection, the length of bone defect and other factors. In addition, the number of studies included was small, so the results should be critically considered.
Bone grafting
Four studies included in this study reported the bone grafting data, and the results showed that the difference between the two groups was statistically significant (P <0.05), which means the AST group needed more bone grafting. At present, most of the research claims to perform bone grafting at the docking site to reduce the bone union time [25]. According to previous literature reports, the bone grafting rate of the BT group ranged from 14.3% to 40% [1, 23, 45] compared to 20–43% in the AST group, which is consistent with the present study [2, 11, 36].
Bone and functional results
All the eligible trials applied the Association for the Study and Application of the Method of Ilizarov (ASAMI) criteria to assess bone and functional results [30]. An excellent rate range from 64% to 83% in the BT group was documented, [4, 19] compared to 53–100% in the AST group [30, 34-36]. Kemal et al. reported bone union of 95.8% and 12 (50%) cases had excellent radiological results [5]. No difference was detected in the two groups, according to bone results (P >0.01). This suggested that both groups were at the same risk for delayed union, malunion and non-union. Due to the limited number of references, there may be some bias in the results, so it is necessary to include more high-quality literature for further analysis and evaluation to draw a more accurate conclusion. All five eligible studies described the detail of functional results, and the result showed that a significant difference was found in the two groups (P >0.01). Studies illustrated that excellent functional results ranged from 38% to 58% [4, 5, 19] in the BT group compared to 60–86% [30, 35, 36] in the AST group. The functional results mainly depended on professional guidance of functional exercise, prevention of needle penetration too close to the joint, adoption of methods to correct the existing ankle deformity and so on. Although AS has the advantage of earlier wound closure and avoiding a flap graft, the shortened tendon becomes relaxed and prone to foot drop [12].
Complications
Pin track infection and screw loosening are the most common complications in external fixation, and usually the final results will not be affected by these complications. In terms of reports, limb length discrepancy, permanent nerve and vascular damage, vascular crisis, re-fracture and newly formed bone infection are rare, and thorough debridement is the key step to controlling bone infection [4]. Studies reported that complications in the BT group ranged from 8.3% to 100% [1, 4, 23, 42] and 9% to 100% in the AST group [32, 33, 36, 37, 43]. Sarah et al. recommended the use of Doppler ultrasonography to assess distal pulses as necessary and to choose the appropriate shortening method according to the soft-tissue and wound condition.[38] Three studies published the data about complications, and one significant difference was found in the two groups according to the complications (P >0.01). Because the included studies did not describe the types and detailed statistical data of the complications, this study could not carry out subgroup analysis, so there may be some bias in the results.
Strengths and limitations
To the best of our knowledge, this is the first meta-analysis to compare the safety and efficacy between bone transport and AST for the treatment of infected tibial bone defects. Moreover, low heterogeneity analysis and publication bias was detected in the meta-analysis. There were certain limitations in the present study. First, all of the eligible studies were retrospective studies, and the sample size was small; most studies were performed in a single centre, which may lead to a certain degree of bias. Therefore, part of the conclusions should be treated with caution. Second, the results may be affected due to the different inclusion and exclusion criteria and measurement indicators of each study. Third, the included literature lacked standardised and unified standards for the recording of variables, especially the external fixation index and bone union time, which resulted in many variables could not be combined for analysis. Fourth, in the five studies, further fixation after removal of external fixation was different, including nail, plate and plaster. The shortening methods were also disparate; immediate shortening or gradual shortening were applied in different studies, and the external fixation types included monolateral fixators and ring fixators. All of these selections may induce heterogeneity and impair the reliability of the conclusion. Therefore, further study based on large-size, multi-centre clinical RCTs, which apply unified and correct scoring system, evaluation indicators and random methods of blinding, is still necessary in the future for achieving higher-level evidence for clinical treatment.