Radiological Outcomes, Complications, and the Influence of Risk Factors in PHILOS Repair of Three- and Four-Part Proximal Humerus Fractures with and Without Femoral Head Allograft: Insights from a Cohort of 116 Patients

TL;DR

This study examines the outcomes of PHILOS repair for proximal humerus fractures with and without allograft augmentation in 116 patients, finding that allograft improves stability but does not prevent complications from poor surgical technique.

Contribution

The study provides insights into the influence of allograft augmentation and risk factors on radiological outcomes and complications in proximal humerus fracture repair.

Findings

Allograft augmentation increases surgical time but does not prevent complications like avascular necrosis or screw cut-out.

Greater tubercle comminution and calcar loss are significant risk factors for head collapse and screw cut-out.

Medial hinge restoration and allograft use are associated with reduced redislocation of the greater tubercle.

Abstract

Background: Complications after proximal humerus osteosynthesis are not uncommon. The aim of this study was to compare the outcomes of osteosynthesis using PHILOS with fresh-frozen femoral head allograft augmentation and without it, and to assess the influence of risk factors and their impact on the occurrence of postoperative complications. Methods: This retrospective study evaluates the radiological outcomes and complications of treating proximal humerus fractures (Neer III–IV) in 116 patients over 50 years of age treated between 2017 and 2021. Results: Osteosynthesis without allograft was performed in 84 patients and with allograft in 32 patients. In total, 42 patients (36%) had a three-part fracture and 74 (64%) had a four-part fracture. The Deltoid Tuberosity Index was comparable between the groups (1.59 ± 0.25 vs. 1.50 ± 0.26; p = 0.802). The average duration of surgery was 101.3 ± 21.3 min with allograft and 86.0 ± 31.9 min without allograft (p = 0.004). AVN was verified in four patients (3.5%), head collapse in nine (8%), cut-out in six (5%), reoperation in eight (7%), infection in three (2.5%), and pseudoarthrosis in one (1%) case. Conclusions: An allograft augmentation improves construct stability, but cannot compensate for inadequate surgical technique. None of the risk factors significantly influenced the development of AVN and pseudoarthrosis. The greater tubercle comminution (p = 0.005), calcar loss (p = 0.020, p = 0.112), allograft augmentation (p < 0.001), and medial hinge restoration (p = 0.012, p = 0.002) were significant risk factors associated with HC and screw cut-out, respectively. The greater tubercle redislocation was influenced by its comminution, calcar loss, and the use of allograft augmentation. HFZ and DTI had no significant impact on surgery results or complications.

Figures