Femoral Neck Fracture Nonunion

Presentation. The patient returns months after fixation with persistent groin or hip pain and painful weight-bearing, a limp, and a limb that is short and externally rotated. On radiographs, look for a persistent fracture line, back-out, cut-out or loosening of the hardware, and varus collapse.

When to call it. A femoral neck fracture that has not united by six months is a nonunion, but the decision rarely waits that long. Progressive varus collapse, screw back-out or cut-out, and a fracture line that is widening rather than fading at three to four months are declaring failure early, and in a young patient there is nothing to gain by waiting for the calendar.

Imaging. Radiographs, with or without CT, confirm the nonunion and show the hardware and alignment. Radiographs alone over-call it, because the screws obscure the line and an apparently persistent lucency may be healing. CT is the test that settles whether the neck has united, and it also shows the bone stock, the position of the hardware and any posterior comminution that will matter for planning.

Exclude infection before you reconstruct anything. A nonunion around retained metal is a septic nonunion until shown otherwise, and this is the single omission that turns a valgus osteotomy into a disaster. Check CRP and ESR, and ask specifically about wound problems or prolonged discharge after the index operation. If there is any doubt, aspirate the hip or take intra-operative samples for culture (multiple, off separate instruments) and histology before committing to a reconstruction.

Infection changes the plan completely, towards debridement, hardware removal and staged treatment, and it is far better found before a blade plate is inserted than after.

Screen the patient's biology. An atrophic nonunion deserves the standard metabolic work-up, the same principles set out in Nonunion Management:

• Vitamin D, calcium, phosphate, parathyroid hormone and thyroid function
• Smoking cessation
• Nutritional assessment
• A review of medications (NSAIDs, steroids)

It is unusual for a metabolic abnormality alone to explain a femoral neck nonunion, where mechanics and vascularity dominate. But correcting one is cheap, and failing to look is indefensible in a nonunion that has already defeated one operation.

AVN travels with nonunion. Avascular necrosis of the femoral head commonly coexists with, or develops after, femoral neck nonunion, since the same vascular insult underlies both. Whether the head is viable is the pivotal step: a viable head can be salvaged with osteotomy, whereas a dead or collapsing head mandates arthroplasty. AVN is also the main reason a valgus osteotomy may ultimately fail and require later replacement.

Reading the head. Confirm viability before committing to a head-preserving osteotomy.

• Plain radiographs show established AVN. Subchondral sclerosis, the crescent sign (a subchondral lucency) and head collapse or flattening all mark a failing, non-viable head.
• MRI is the most sensitive test for marrow viability and pre-collapse AVN (the marrow-signal and double-line changes). Beware metal artefact from the retained fixation, which may need metal-artefact-reduction sequences (MARS), or removal of the hardware first, before the head can be read. The disease itself is in Avascular Necrosis of the Hip.
• Bone scan or SPECT assesses head perfusion, reduced uptake meaning avascular, and is useful where MRI is degraded by metal (technique in Bone Scan Interpretation).
• The intra-operative drill test is the real-time check: drill the femoral head and look for punctate bleeding. A bleeding head is viable and can be salvaged by osteotomy; a dry, non-bleeding head is avascular and should be replaced.

Two questions decide it. Treatment turns on the patient's age and the viability of the femoral head, weighed alongside their demand and comorbidities.

Young patient, viable head: preserve it. The head-preserving operation of choice is a valgus intertrochanteric osteotomy, which converts the shear across the vertical nonunion into compression (planning and technique below). Bone grafting, for example a non-vascularised or vascularised fibular graft or a muscle-pedicle graft, is often added for biology, and an atrophic or gap nonunion needs it.

What the series show. Published series report union in most or all cases at around 12-13 weeks with good function. Read those numbers against their cohorts: the two series cited below contain 11 and 21 patients, there are no comparative trials, and "all united" out of eleven hips is a promising signal rather than a dependable rate. AVN remains a risk regardless.

Elderly patient, dead head or failed osteotomy: replace it. For the elderly, a non-viable (AVN or collapsed) head, established post-traumatic arthritis or a failed osteotomy, total hip arthroplasty is the reliable definitive treatment: one operation, immediate weight-bearing and predictable pain relief. It is not a primary hip replacement, for the reasons below.

The principle. A closing-wedge valgus osteotomy at the intertrochanteric level tilts the head into valgus, so the previously vertical fracture line becomes more horizontal and is loaded in compression, the Pauwels principle.

View larger

How much valgus. The goal sets the correction: bring the steep fracture line broadly into the compression range, roughly below 30 degrees from horizontal. The amount of valgus is the wedge angle needed to achieve that, templated pre-operatively (the classification itself is in Neck of Femur Fracture).

The construct. A laterally-based, medial closing wedge is taken at the intertrochanteric level and held by a fixed-angle device, an angled blade plate or equivalent, which sets and maintains the correction. Bone graft is added for biology.

The trade-offs to counsel. A medial closing-wedge valgus osteotomy shortens the limb and shifts the proximal femur. That can medialise the shaft or shorten the abductor lever arm, and leave a leg-length discrepancy and a Trendelenburg-type gait.

Modern variants. The subtrochanteric angulation/lateral-translation valgus (SALVA) osteotomy is designed to restore the abductor lever arm and correct the leg-length discrepancy, and it consolidates reliably in young adults. It also keeps the proximal femur favourable, so a later conversion to arthroplasty stays technically straightforward, which makes it useful for neglected or non-united neck fractures.

Arthroplasty is the commoner operation in this situation, and it is consistently underestimated. A total hip replacement after failed femoral neck fixation is a reconstructive procedure with higher complication rates than a primary, and the exam question is usually about exactly why.

The hardware has to come out. It may be broken, buried or stripped, so have extraction sets available, and accept that removal can take longer than the arthroplasty.

Screw holes are stress risers. Empty tracks in the lateral cortex and neck concentrate stress, and a fracture through one during broaching or in early weight-bearing is a recognised event. Consider a stem that bypasses the most distal hole by two cortical diameters, and be deliberate about it rather than hoping.

Exposure is harder. Scarring from the previous approach, capsular contracture, heterotopic bone and a shortened, adducted limb with contracted abductors all make dislocating the hip and restoring length more difficult than in a primary.

Length and offset. Restoring them is a genuine challenge when the neck has resorbed or collapsed into varus. Getting it wrong leaves the abductors weak and the patient with a limp and a Trendelenburg gait.

Dislocation and infection. Both are more frequent than after primary arthroplasty, because previous surgery, altered soft tissues and longer operative time all contribute. Implant choice should account for stability, and the patient should be counselled with these figures rather than primary-hip figures.

Total hip replacement, not hemiarthroplasty. THR is generally preferred here. These patients are usually younger and more active than the typical acute hip-fracture patient, and the acetabular cartilage has often already been damaged by the collapsed head or protruding metal. Hemiarthroplasty has a limited role in low-demand patients.

Guidance and Evidence Frameworks

No society publishes a dedicated clinical practice guideline for femoral neck nonunion; the salvage decision (valgus osteotomy versus arthroplasty) is extrapolated from the small series on this page and from general nonunion principles. The nearest anchors:

• Index-fixation guidance (AAOS hip-fracture CPG, NICE CG124 and equivalents) addresses the primary fracture - stable fixation, timely surgery, reduction quality - which is where nonunion is prevented; the FAITH trial is the pivotal fixation RCT.
• AO / trauma-association teaching supplies the mechanical principle (convert shear to compression) and the Pauwels framework that drives osteotomy planning.
• Head-preserving techniques (valgus osteotomy, vascularized fibular graft) remain centre-of-expertise operations; no guideline mandates them, and availability varies widely.

Registry Perspective

Arthroplasty registries (NJR, AJRR, AOANJRR and the Scandinavian registers) capture the endpoint of this pathway - conversion to total hip replacement after failed fixation - but do not record osteotomies or grafting, so there is no registry-level union-rate or survivorship data for head-preserving salvage. The evidence base is consequently confined to single-centre series, exactly as the cards above show.

Global Practice Variation

• High-resource settings: MRI/MARS viability assessment, blade-plate valgus osteotomy and microvascular fibular grafting are available at tertiary centres; arthroplasty is the default when the head is non-viable.
• Limited-resource settings: neglected femoral neck fractures presenting late are common, and the valgus osteotomy - requiring no implant beyond an angled plate - is often the only head-preserving option; where arthroplasty is unaffordable, osteotomy may be offered even in older patients, accepting less predictable results.
• The universal principle is unchanged: establish head viability first, preserve the viable young head, replace the dead one - and the earlier the nonunion is recognised, the more options remain open.

The femoral neck is uniquely prone to nonunion, for biological and for mechanical reasons.

Biology. The neck is intracapsular. Its blood supply is precarious and largely retrograde, running in the retinacular vessels from the medial femoral circumflex artery, and the fracture itself can disrupt it. It is bathed in synovial fluid, and it forms essentially no periosteal callus, so healing must occur by direct (intramembranous) union, which is unforgiving of any instability or gap.

Mechanics. A vertical fracture line, a high Pauwels angle (Pauwels III), is loaded in shear. Shear displaces the fragments rather than compressing them, and so defeats fixation.

Risk factors. They follow from both: displacement (Garden III-IV), a vertical or Pauwels III pattern, poor reduction, posterior comminution, and patient factors such as smoking.