ABSTRACT
BACKGROUND: Proximal humerus fractures (PHF) are common and lead to post-traumatic humerus head necrosis (HHN) in 3-35% after ORIF with an internal locking plate. Few studies focus on this condition and risk factors remain a discussion topic. Hertel's criteria for initial head ischemia right after fracture (fracture complexity, medial hinge displacement and short metaphyseal head extension) have recently been correlated to HHN, but there is still a clear lack of evidence on the topic. Due to its anatomical similarities to the proximal femur, some authors argue that PHF may as well benefit from early surgery to avoid head necrosis. METHODS: In this 10-year retrospective study, we assessed 305 patients from a single center. All cases were treated with a PHILOS plate through a deltopectoral approach. The mean follow-up time was 467 days. The primary endpoint was HHN. RESULTS: HHN was diagnosed in 12 patients (4%), 10 of which were diagnosed within the first year and one case 4 years after surgery. A positive correlation (p < 0.04) was found between HHN and fracture type (both in AO and Neer's classification), initial neck-shaft-angle (NSA) and metaphyseal head extension (MHE). Medial hinge displacement (MHD) occurred in all HHN cases. Achieving perfect reduction (< 2 mm dislocation) was relevant to avoiding HHN (p = 0.035). Although HHN developed in 32% of the high risk cases (four-part fractures with a short MHE), it was completely avoided (0%) when perfect reduction was achieved. Time until surgery after admission was neither a protective nor a risk factor for HHN. CONCLUSION: We conclude that fracture complexity (four-part and C-fractures) as well as disruption of the medial hinge with a metaphyseal head extension smaller than 8 mm are relevant risk factors for humerus head necrosis. A combination of these criteria generated an high risk pattern with a 32% rate of HHN. Though often difficult to achieve, perfect reduction was a clear protective factor and reduced HHN to 0%. Perfect reduction may be key to inosculation and, therefore, salvage of the humerus head, especially in high risk cases. Surgery timing did not correlate with HHN. LEVEL OF EVIDENCE: Level 3, retrospective cohort study.
