Treatment of proximal humerus fractures in geriatric patients - Can pathological DEXA results help to guide the indication for allograft augmentation?

• Main Authors: Sascha Halvachizadeh, Till Berk, Thomas Rauer, Christian Hierholzer, Roman Pfeifer, Hans-Christoph Pape, Florin Allemann
• Format: Article
• Language: English
• Published: Public Library of Science (PLoS) 2020-01-01
• Series: PLoS ONE
• Online Access: https://doi.org/10.1371/journal.pone.0230789

INTRODUCTION:Reconstruction of proximal humerus fracture continues to represent a challenge, especially in severe osteopenia. However, there still is a lack of consensus and clear indication on use of allograft augmentation. Therefore, this study aims to investigate outcome after osteosynthesis with and without allograft augmentation. It focuses on bone density results obtained by DEXA as potential examination that might help decision-making. METHODS:This study included patients aged 65 years and older that were treated at one Level 1 trauma center between 2007 and 2018. Inclusion criteria: Proximal humerus fracture treated with or without allograft, conclusive data-sets. Exclusion criteria: prior surgical treatment of the proximal humerus, open fracture with bone loss, neurological damage. Patients were stratified according to the use of allograft augmentation in two groups: Group NA (no allograft augmented PHILOS) and Group A (PHILOS with allograft augmentation). Comorbidity was assessed using the Charlson Comorbidity Index (CCI). Fractures were graded according to the classification by Neer. Radiographic union was analyzed at 6 weeks, 12 weeks, and at year follow up. Complications include surgical site infection, implant failure, humeral head necrosis, or delayed union. Allograft was used in cases of 1inch/3cm3 bone-loss or an egg-shell situation, where the patient refused arthroplasty. RESULTS:This study included 167 patients, with 143 (85%) in the Group NA, and 24 (15%) in the Group A. There were no significant differences in age, gender, injury distribution, and distribution of Neer classification or CCI. Patients in Group A had significantly lower T-scores preoperatively (-2.87 ± 1.08 versus -0.9 ± 2.12, p = 0.003). No difference occurred in any of the complications. At one-year follow-up, the range of motion was comparable in both groups. CONCLUSION:In patients with allograft augmentation and severe osteopenia, similar clinical and radiological results were obtained when compared with patients with better preoperative bone density scores (T-scores, DEXA). In view of a lack of guidelines indicating the indication for the use of allograft, this difference may be worth further study.