Chronic infection and infected non-union of the long bones in paediatric patients: preliminary results of bone versus beta-tricalcium phosphate grafting after induced membrane formation

Chronic infection (CO) and infected non-union of the long bones are relatively rare conditions in paediatric patients. Large bone defects secondary to these conditions can be managed with the induced membrane technique. The technique requires grafting of the bone void, although it is not yet established what bone substitute is the best option. The aim of this work was to evaluate the outcome and efficacy of treatment in children with CO and infected non-union of the long bones using the induced membrane technique and bone (BG) versus beta-tricalcium phosphate (BTP) grafting.

Eight skeletally immature patients with CO and infected non-union of the long bones were treated surgically between 2010 and 2017 by a combination of resection of necrotic infected bone, debridement of surrounding soft tissue, osteosynthesis using a stable internal fixation when needed, and application of antibiotic-laden cement (ALC) spacer inducing new membrane before final bone reconstruction with bone substitutes: BTP in five cases, BG (allograft and/or autologous graft) in three cases. A second surgical step, once inflammatory markers had normalized, consisted of ALC spacer removal, application of BG or BTP graft and concomitant stable osteosynthesis, if needed, if this had not been done during the first surgical stage. All the patients underwent clinical, laboratory and imaging evaluation before and after surgery. Antibiotics were adjusted according to culture and sensitivity.

Mean patient age at time of diagnosis was 13 ± four years (range, 4–16) and all had at least a 12-month follow-up (range 12–60). Estimated time for induced membrane formation was significantly shorter in patients treated with BTP compared with BG: 3±1 vs. 10±2 (p = 0.02). This result was confirmed by multivariate analysis (p = 0.044) taking into account adjustment for age of patients and time after initial surgery. Time of final union was about 5.5 ± 4.1 months (range 2–66). At the last follow-up visit, bone had healed and all the patients had resumed daily living and sports activities.

The induced membrane technique with BG or BTP graft can achieve bone healing in large bone defects secondary to CO and infected non-union in children and adolescents. The choice of bone substitute is important. Our preliminary results show graft integration and bone healing can be expected sooner if BTP is used as bone void filler.