Long-term differences in clinical prognosis between crossed- and parallel-cannulated screw fixation in vertical femoral neck fractures of non-geriatric patients.

BACKGROUND AND PURPOSE: Vertical femoral neck fractures (VFNFs) are one of the most difficult fractures to stabilize, with high non-union (17%), avascular necrosis (AVN, 21%), and femoral neck shortening (FNS, 29.0%) rates. The objective of this investigation was to directly compare the long-term clinical complication rates of VFNFs repaired by crossed (Alpha fixation) or parallel screws in non-geriatric patients.

PATIENTS AND METHODS: We conducted a retrospective comparative study of VFNFs in patients (<60 years) between January 2014 and December 2017, with at least 2 years of follow-up. VFNFs were fixed with either three parallel screws (G-TRI) or augmented with a crossed screw (G-ALP). Confounding variables included age, gender, initial displacement, ISS (Injury Severity Score), general comorbidities, combined fractures, Pauwels angle, reduction quality. Complications, including non-union, AVN of the femoral head and FNS were compared as outcome indicators. Risk factors associated with these variables were further analysed using multivariate analysis.

RESULTS: A total of 157 patients (97 G-TRI; 60 G-ALP) met inclusion criteria. G-ALP had a significantly lower rate of FNS (8.3% vs. 28.9%, p = 0.039) than that of G-TRI. non-union (3.3% vs. 11.3%), AVN (21.7% vs. 25.8%), and reoperation rates (21.7% vs. 23.7%) were lower in G-ALP than G-TRI but was not statistically significant. Multivariate analyses showed significant relationships between NU and unacceptable reduction quality (OR=7.610; 1.823-31.770, adjusted-p = 0.015); between AVN and initial displacement (Garden III and IV) (OR=7.885; 1.739-35.744, adjusted-p = 0.021); and between FNS and screw configuration (OR=5.713; 1.839-17.743, adjusted-p = 0.009).

CONCLUSION: For the treatment of VFNFs, satisfactory reduction still remains the key surgical goal that prevents NU, while the incidence of AVN strongly depends on the initial displacement at the time of injury. Crossed screws were associated with a markedly lower FNS rate than parallel screws, which promote further randomised controlled trials to establish a guideline for optimal fixation selection in VFNFs.