Valgus Intertrochanteric Osteotomy Without Fibular Strut Grafting for Femoral Neck Nonunion in Children: A...

Paediatric femoral neck nonunion is uncommon but notoriously difficult to treat because of shear forces across the fracture line and the tenuous blood supply to the femoral head. Although vascularised or non-vascularised fibular strut grafting is classically recommended, valgus intertrochanteric osteotomy (VITO) alone realigns mechanical forces and may obviate the need for grafts. When the nonunion coexists with established avascular necrosis (AVN), the decision is even more controversial. We present a single-centre series of children treated with stand-alone VITO, i.e. without any bone graft augmentation, and analyse union rates, functional recovery, limb-length discrepancy (LLD) correction, and the rationale for choosing osteotomy even in hips that already showed AVN.

Between December 2020 and July 2024, 18 consecutive children (< 16 years) with radiographically proven femoral neck nonunion underwent VITO. Patients were stratified a-priori into two age cohorts (Group A < 10 years, Group B ≥ 10 years) and characterised for Delbet fracture type, presence of trochanteric override, and pre-operative coxa vara (neck-shaft angle < 120°). Five hips (28%) had established AVN (Kalamchi-MacEwen grades I-III). Pre-operative planning on calibrated radiographs provided the Pauwels (shear) angle from which the closing-wedge angle (Ω) was calculated (Ω = measured angle - target 25-30). The templated sheet guided intra-operative cuts. LLD was measured clinically with the block test and radiographically from the anterior-superior iliac spine to the medial malleolus.Osteotomies (planned correction ≈ 42°) were stabilised with either a double-angled blade plate (11 hips) or a dynamic hip screw/valgus plate (7 hips). Minimum follow-up was 12 months (mean 25.4 ± 4.5). Consecutive sampling, pre-specified eligibility criteria, a single-team protocol, > 95% prospectively captured baseline data, subgroup/sensitivity analyses, and zero loss to follow-up limited bias.

Mean NSA improved from 89.4 ± 3.5° to 131.6 ± 3.7°, and the Pauwels angle fell from 59.3 ± 4.8° to 30.2 ± 2.9°, achieving the biomechanical target in every case. All fractures united; mean union occurred at 14.5 ± 1.5 weeks for the fracture site and 10.6 ± 1.0 weeks for the osteotomy. HHS rose from 46.5 ± 4.5 ("poor") pre-operatively to 83.3 ± 3.0 ("good") at final review. LLD decreased from 3.0 ± 1.2 cm to 2.1 ± 1.5 cm. Complications were limited to mild residual Trendelenburg gait in 6 hips (33%); no implant failure, loss of correction, infection, or new avascular-necrosis cases were observed. Outcomes were comparable between fixation constructs, although blade plates were favoured for corrections > 45°.

VITO alone achieved union in all cases with substantial functional recovery and low morbidity, indicating that fibular strut grafting is not necessary in paediatric femoral neck nonunion. Eliminating graft harvest simplifies surgery and avoids donor-site harm. Future prospective comparative studies are warranted.