Aims This is a prospective randomised study which compares the radiological and functional outcomes of ring and rail fixators in patients with an infected gap (> 3 cm) nonunion of the tibia.
Patients and Methods Between May 2008 and February 2013, 70 patients were treated at our Institute for a posttraumatic osseocutaneous defect of the tibia measuring at least 3 cm. These were randomised into two groups of 35 patients using the lottery method. Group I patients were treated with a ring fixator and group II patients with a rail fixator. The mean age was 33.2 years (18 to 64) in group I and 29.3 years (18 to 65) in group II. The mean bone gap was 5.84 cm in group I and 5.78 cm in group II.
The mean followup was 33.8 months in group I and 32.6 months in group II. Bone and functional results were assessed using the classification of the Association for the Study and Application of the Method of Ilizarov (ASAMI). Functional results were also assessed at six months using the short musculoskeletal functional assessment (SMFA) score.
Results The bone result was excellent, good, fair and poor in 21, 12, 0 and 2 in group I; and 14, 15, 3, and 3 in group II, respectively. The functional results were excellent, good, fair, poor and failure in 16, 17, 1, 0 and 1 in group I; and 22, 10, 0, 3 and 0 in group II, respectively. Both fixator systems achieved comparable rates of union and functional outcomes. The rate of deep pintract infection was significantly higher in the rail fixator group but patients found it more comfortable.
Conclusion We recommend the use of a ring fixator in patients with a bone gap of more than 6 cm. Patients with a bone gap up to 6 cm can be managed with either a ring or rail fixator.
Cite this article: Bone Joint J 2016;98B:1399–1405.
Defects of bone and softtissue in the leg may be the result of highenergy trauma, surgical debridement, and treatment of bone tumours. They are managed using a number of different techniques which include the application of antibiotic impregnated beads; cancellous bone grafting; fibrin sealants mixed with antibiotics; tibiofibular synostosis: local softtissue rotational flaps; free microvascular softtissue flaps and bone transplant.1,6
However, the treatment of these infected -tibial osseocutaneous defects is demanding because of the associated problems of deformity, shortening, joint stiffness and osteopenia.5
Distraction osteogenesis has the potential to treat these problems simultaneously. Bone transport can be undertaken using either a ring or monolateral fixator or by inserting an intramedullary nail.3,4,7,12 The disadvantages of a ring fixator are that it may be cumbersome, heavy, complicated both for surgeon and patient and may not be well tolerated by the patient when applied for a long period of time.1,7,9,10 In contrast, a monolateral rail fixator has been shown to be less cumbersome for the patient, easier to apply and consequently more surgeon and patientfriendly.11
To date, there have been few studies which compare the outcomes of ring and monolateral fixators.12
The aim of this study was to compare ring and monolateral rail fixators in patients with an infected nonunion of the tibia in which there was a bone defect of more than 3 cm by assessing the rate of union, the radiological changes and the functional outcome of the patients.
Patients and Methods
We carried out a prospective randomised study of 70 patients (62 males and eight females, mean age 31.3 years; 18 to 65) with an infected posttraumatic osseocutaneous defect of the tibia who presented to our institute between May 2008 and February 2013. The study was approved by Institutional Review Board and Ethical committee. We included only those patients who had a bone gap of at least 3 cm after radical resection of necrotic bone and had a current or past history of active infection. Of these, 46 patients presented on admission with a bone gap of at least 3 cm and 24 had at least a 3 cm gap as a result of surgery. Patients were considered to have an active infection if there was a discharging sinus or if a pathogenic organism was isolated from the wound. We excluded patients who had evidence of nonunion of the tibia for reasons other than trauma, had an insensate foot, were aged > 65 years or < 18 years, had a pathological fracture, a periarticular nonunion or any medical or skeletal condition which might affect bone healing. All patients gave their informed consent to the study. They were then randomised into two groups using the lottery method. Patients who presented with a bone gap of 3 cm or more were randomised preoperatively while those who had a bone gap of at least 3 cm after radical resection of bone were randomised intraoperatively. Group I consisted of 35 patients who were treated with the ring fixator system and group II, 35 patients treated using the rail fixation system. Details of these two groups are shown in Table I.
Using the classification of Paley et al,13 12, two, 13, seven, and one patient in group I and six, zero, 12, 12 and five patients in group II had type A1, A2, B1, B2 and B3 respectively (Table II).
No patient had a fixed deformity. Two patients in group I had a stiff nonunion without deformity (Paley type A21). No statistically significant difference was seen between the two groups in terms of age, gender, mode of trauma, side and site of fracture, associated injuries, time between injury and application of frame/fixator, nature and the number of previous surgeries and length of followup.
Operative treatment included thorough softtissue debridement, the removal of any implant, exposure of the fracture site, radical resection of necrotic or sclerotic bone down to healthy bleeding bone and opening of the medullary canal. The transporting bone fragment end was cut until coverage of skin was possible so that transporting the bone fragment carried skin over it during the process of distraction. After achieving a vascular environment at the fracture site the fixator frame was applied and corticotomy undertaken. Cultures of the infected tissue, obtained at the time of debridement, dictated the choice of post-operative antibiotic treatment.
The mean bone gap was 5.84 cm (3 to 9) in group I and 5.78 cm (3 to 10) in group II (p = 0.87). The mean size of the wound was 5 cm x 2.5 cm (1 to 9 x 1 to 4) in group I and 4 cm x 2.2 cm (1 to 8 x 1 to 4) in group II. The standard ring fixator frame included four rings, with four rods and a mean of ten wires per patient (9 to 11). During insertion of the wires in the ring fixator group, the muscle being penetrated was stretched to ensure maximum joint movement. The frame was applied as described by Ilizarov.14 Schanz pins were inserted by hand for the rail fixation systems to decrease the risk of thermal necrosis and pin loosening. The rail fixation system included a mean of nine tapered halfpins of 6mm diameter per patient. All patients had a corticotomy at one level. In group I, this was carried out proximally in 24 patients, distally in eight patients and at the diaphysis in three. In group II the corticotomy was proximal in 28 patients and distal in seven. Acute docking at the fracture site was carried out in two patients in group I and one patient in group II. Fibular osteotomy or resection (three patients in group I and four patients in group II) was carried out when needed to correct deformity or acute shortening. Splitskin grafting was undertaken in four patients in group I and six patients in group II.
Joint movement and partial weightbearing with crutches or a walker was encouraged, depending on the patient’s degree of compliance and pain. Distraction, at an initial rate of 1 mm per day in four increments, was started on day 7. All patients underwent compression at the fracture site and distraction osteogenesis at the corticotomy site. Broadspectrum antibiotics were given for five days and thereafter for another ten days, the choice of antibiotic being determined by the culture and sensitivity. Antibiotics were prescribed during followup as per requirements, which may include treatment for pintract infections, or persistent discharge from the wound. All patients were reviewed monthly. Complications such as muscle contractures, axial deviation, premature consolidation, delayed consolidation, refracture and pintract infections were recorded. Pintract infections were classified according to the system as suggested by Paley15 into grade 1, grade 2 and grade 3. Grades 1 and 2 infections were grouped together for statistical analysis and termed superficial infections. Grade 3 infections were termed deep infections. If necessary, the frame was adjusted under anaesthesia to correct malalignment. Adjustment of the fixator also included freshening of the bone edges and excision of interposed soft tissue. The frames were retained until adequate consolidation and fracture union was achieved or when the patient insisted that it was removed because of pain or a complication during treatment such as infection, and aseptic loosening. Patients were supplied with a PVC brace for six months after removal of the fixator.
Two independent observers assessed the outcomes (interobserver agreement Kappa 0.78). The results were recorded as bone and functional results using the classification of the Association for the Study and Application of the Method of Ilizarov (Table III).4 The fracture was considered united when there was radiological evidence of bridging trabeculae- across three of four cortices on anteroposterior and -lateral radiographs, when there was no motion at the fracture site after loosening of the frame and when the patient was able to walk without pain at the fracture site.
The external fixator index16 was calculated by dividing the time in the frame (in months) by the length of the regenerated bone. The short musculoskeletal functional assessment (SMFA) score was administered six months postoperatively with patients still in their fixator.17 No patient was lost to followup.
Data were analysed using IBM SPSS statistics 20 software (IBM Inc., Armonk, New York). The chisquared and independent Student’s ttest were used. A p value < 0.05 was considered statistically significant.
There were no intraoperative complications.
The mean operating time was 86.1 minutes (60 to 120) in group I and 67.8 minutes (45 to 90) in group II (p < 0.0001). The mean time in external fixation was 11.1 months in group I and 10.5 months in group II (p = 0.46). The external fixator index was 2.12 months/cm in group I and 2.11 months/cm in group II (p = 0.94). The mean followup was 33.8 months in group I and 32.6 months in group II.
Fractures united primarily in 27 patients (77%) in group I (Fig. 1) and 28 patients (80%) in group II (Fig. 2) and, after fixator adjustment with freshening of edges, in an additional seven patients in group I and five in group II. One patient had premature removal of the ring fixator in group I due to excessive pain and went on to amputation. The rail fixator was converted to a ring fixator in two patients in group II for reasons given below, but union was eventually achieved in both patients. All corticotomy sites healed.
The mean gain in bone length was 5.2 cm (3 to 9) in group I and 5.1 cm (3 to 8) in group II. A limb length discrepancy > 2.5 cm persisted in three patients in group I and six patients in group II (p = 0.47). Deformity > 7° persisted in nine patients in group I and seven patients in group II (p = 0.77). No patients in either group had residual infection at the fracture site. Four patients in each group had a loss of > 15° knee extension. There were 14 patients in group I and eight in group II who had a loss of more than 15° in ankle dorsiflexion (p = 0.19).
Three wires broke during treatment in two patients in group I. Superficial pintract infections occurred in 19 patients in group I and 17 patients in group II (p = 0.81). Deep pintract infections occurred in two patients in group I and 13 patients in group II (p = 0.0027).
Correction of deviation of the transporting fragment was required in seven patients in group I and five patients in group II. The transported bone carried overlying skin with it thus avoiding flap reconstruction. One patient in group II had premature consolidation. One patient in each group had a refracture after removal of the frame. Two patients in group I had severe pain requiring admission for pain control.
The corticotomy site became infected in one patient in group II. Recorticotomy was performed after treatment of the infection and the rail fixator was converted to ring fixator. Another patient in group II had aseptic loosening of the distal pins and the rail fixator was converted to ring fixator. The results of both these patients were considered to be poor: however, both patients ultimately achieved union with a ring fixator. We did not analyse these patients in the ring fixator group and the outcome of patients before the application of ring fixator was assessed (intention to treat analysis). There was no statistically significant difference in bone and functional results between the two groups as per ASAMI criteria. Comparison of bone and functional results in patients with a bone gap up to 6 cm and a bone gap of more than 6 cm is depicted in Tables IV and V. The mean SMFA score at six months postoperatively was 38.2 (30.4 to 48.4) in Group I and 34.8 (28.2 to 44.5) in Group II (p = 0.0014). The mean Functional Index in Group I was 36.1 (28.7 to 47) in Group I and 32.8 (26.5 to 42.6) in Group II (p = 0.003). Bothersome index assesses how much the patient is bothered.17 The mean Bothersome Index17 was 43.5 (35.41 to 52.08) in Group I and 40.2 (33.33 to 52) in Group II (p = 0.0048). A total of 60 complications occurred in group I patients (mean of 1.71 complications per patient) and 70 in group II patients (mean of two complications per patient).
The aim of treatment in infected nonunion is to obtain solid bony union and eradication of infection while maintaining maximum functional use of the extremity.1 This is achieved by radical debridement down to bleeding bone,10 which is essential even when using the Illizarov technique.4,7,9,12,13,18,19
The results of our study are comparable with others reported in literature (Tables VI and VII).3,4,10,16,20,24 Union was achieved primarily in 77% of patients in group I and 80% of patients in group II. Further procedures including adjustment of the fixator, removal of interposed soft tissue and freshening of the bone ends resulted in union in an additional 14% to 20% patients in each group. In our study no patient needed bone grafting although it has been reported as necessary in up to 83% patients undergoing bone transport.16,25 We advocate early freshening of the fracture ends with removal of any interposed soft tissue after achieving docking as Iacobellis, Berizzi and Aldegheri12 have recommended.
The Ilizarov method is a comprehensive approach to all aspects of chronic tibial nonunion and simultaneously addresses the problems of deformity, shortening, defects, infection and osteoporosis.13,14,18,26 We found that the mean operating time was significantly less with a rail fixator than a ring fixator suggesting that the former is easier to apply. Iacobellis et al12 also compared circular and mono-lateral fixators and found thatcircular fixators were heavier, more uncomfortable for patients and associated with longer medication times. The Limb Reconstruction System (LRS Pitkar, Pune, India) is a uniplanar dynamised external fixator, which is lightweight and easy to construct.11 Unfortunately it is more difficult to correct a threedimensional deformity with a uniplanar external fixator than with Ilizarov ring fixator.11 Circular frames also confer greater stability and more flexibility than uniaxial devices.1
In our study the rates of superficial infection were comparable in both groups. Chaddha et al16 reported pintract infections in 80% of patients treated with a ring fixator. Wani et al26 also found pintract infection to be the most common complication associated with an Ilizarov ring -fixator. Pintract problems are related to pin to skin motion, the amount of soft tissue between skin and bone and the diameter of the pin used.15 Although the infection is superficial and mild in most instances, infection increases the risk of wire loosening: because of the weight borne by the external fixator this can cause the frame to become unstable.26 We found that the rate of deep pintract infection was significantly less in the ring fixator group (5.7%) than in the rail fixator group (37%). This may be related to larger diameter of pin used in rail fixation systems. We also noticed that ankle stiffness was more common in the ring fixator group than in the rail fixator group. Baruah2 reported that equinus deformity was less common in the halfpin group. Significant joint stiffness has been reported in up to 63% patients who have a ring fixation system applied.4,10,16
We found that there was no statistically significant difference in the bone and functional results between the two groups when assessed using the ASAMI criteria in patients with bone gap of up to 6 cm. Therefore in patients with a bone gap of up to 6 cm, both types of fixator achieved -comparable results.
In total, 13 patients in group I and 11 patients in group II had bone gap of more than 6 cm. More patients in group II had some residual problem (infection/deformity/shortening) than in group I, indicating a better bone result from using the ring fixation. The incidence of infection was also higher in group II than in group I which indicated an inferior bone result in group II in patients with a bone gap of > 6 cm. Similarly, an excellent functional outcome was less common in group II than in group I. Thus, in patients with a bone gap of more than 6 cm a ring fixator is more likely than a rail fixator to achieve excellent bony union.
The SMFA scores were measured at six months postoperatively when patients were still in their fixator frames so as to detect the function and level of comfort patients experienced. We were able to show that the mean SMFA score, mean Functional Index and mean Bothersome Index17 were significantly higher in group I than in group II at this stage.
The limitation of our study is the small number of patients in each group: we did not carry out a power calculation to determine the sample size. As these conditions are relatively rare, we felt that a sample size of 70 patients would be a reasonable number.
In conclusion, both fixator systems achieve comparable, reliable rates of union and functional outcome in infected gap nonunions of the tibia and help in treating infection, shortening, bone and loss of soft tissue simultaneously without flap coverage.
We recommend the use of the ring fixator in patients with a bone gap of more than 6 cm: patients with a bone gap of up to 6 cm can be properly managed with either system.
Take home message: Infected nonunion of tibia with bone gap more than 6 cm should be treated with ring fixators.
R. Rohilla: Performed surgeries, Data analysis, Writing the paper.
J. Wadhwani: Data collection, Data analysis, Writing the paper.
A. Devgan: Performed surgeries, Writing the paper.
R. Singh: Performed surgeries, Writing the paper.
M. Khanna: Data collection, Data analysis.
No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article.
This article was primary edited by S. P. F. Hughes and first proof edited by A. C. Ross.
- Received January 23, 2016.
- Accepted June 3, 2016.
- ©2016 The British Editorial Society of Bone & Joint Surgery