Intramedullary cement augmentation in locking plate fixation for Neer four-part proximal humeral fractures: a technique for preventing early mechanical failure

BACKGROUND: Locking plate fixation for Neer four-part proximal humeral fractures is frequently complicated by varus collapse and screw cut-out, leading to mechanical failure, particularly in osteoporotic bone. This study evaluates the clinical and radiographic outcomes of a modified technique involving intramedullary polymethylmethacrylate (PMMA) bone cement augmentation applied during its thread-forming phase to create an internal support structure. METHODS: In this retrospective cohort study, 66 consecutive patients with Neer four-part proximal humeral fractures treated between January 2023 and June 2024 were enrolled and divided into two groups: 31 patients underwent locking plate ORIF with intramedullary PMMA cement augmentation (augmentation group), and 35 patients received conventional locking plate fixation without cement (control group). Patient demographics, surgical details, and perioperative data were recorded. Primary outcomes were radiographic: fracture union time, and changes in the neck-shaft angle (NSA) and humeral head height (HHH). Secondary outcomes included the Constant-Murley Score (CMS), Visual Analog Scale (VAS) for pain, and all complications. RESULTS: The two groups were well matched demographically (mean age 77.9 ± 3.6 years; P > 0.05). All fractures achieved union, with significantly faster healing in the augmentation group (14.42 ± 1.45 vs. 15.10 ± 1.12 weeks; P = 0.0245). Radiographic stability was markedly superior in the augmentation group: mean loss of NSA was 1.4° ± 1.2° versus 2.8° ± 1.9° (P = 0.001), and mean loss of HHH was 0.6 ± 0.5 mm versus 1.8 ± 1.1 mm (P < 0.001). Critically, no patient (0%) in the augmentation group experienced significant loss of reduction (defined as > 5° change in NSA or > 3 mm loss of HHH), compared to three patients (8.57%) in the control group (P = 0.095). Early postoperative pain was significantly reduced in the augmentation group on day 1 (VAS: 4.35 ± 1.27 vs. 5.20 ± 1.39; P = 0.0065) and day 3 (3.29 ± 0.45 vs. 3.63 ± 0.64; P = 0.0044). Functional recovery was significantly better at 1 month (CMS: 51.61 ± 6.16 vs. 48.54 ± 7.06; P = 0.0447) and 3 months (66.10 ± 10.32 vs. 62.09 ± 8.56; P = 0.0467). The complication rate was low; notably, radiographic avascular necrosis occurred in only one patient (3.23%) in the augmentation group versus five patients (14.29%) in the control group (P = 0.122). No cases of screw cut-out, implant failure, or reoperation were recorded. CONCLUSION: Intramedullary augmentation with thread-forming phase PMMA cement is a highly effective adjunct to locking plate fixation for Neer four-part proximal humeral fractures. In this cohort, no patient in the augmentation group experienced significant loss of reduction, suggesting that the technique may reduce the risk of early mechanical failure, provide superior radiographic stability, reduce early postoperative pain, accelerate early functional recovery, and demonstrate a clinically meaningful trend toward lower AVN incidence. It represents a reliable joint-preserving option for managing these complex fractures, particularly in patients where maintaining the native glenohumeral joint is a priority.