The typical anterolateral approach (including the minimally invasive anterolateral approach) is the working horse for the surgical treatment of proximal humerus fractures with locking fixation[13, 14]. However, despite the biomechanical superiority and clinical availability of this treatment approach, several studies have documented reduction loss and varus deformity in proximal humeral fractures treated with locking plates. A common pattern of failure is the increased varus diaplacement of the humeral head, leading to secondary screw penetration or plate breakage[13, 15–17].
Many studies have considered that medial support in proximal humerus fracture surgery is crucial to avoid reduction loss and varus displacement. Jung et al. revealed that insufficient medial support was an independent risk factor for reduction loss in proximal humerus fractures (p < 0.01). Maier et al.  also found that short or absent metaphyseal head extension and disruption of medial periosteal hinge reliably predict posttraumatic humeral head ischemia. Hertel et al. identified that the reliable predictors of ischemia were short (< 8 mm) posteromedial metaphyseal extension of the head fragment and disruption of the medial hinge with displacement of > 2 mm. Krappinger et al. pointed out that local bone mineral density, restoration of the medial column, nonanatomic reduction, and age were significant predictors of fixation failure (P < 0.01). Osteosynthesis of osteoporotic fractures with nonreconstructable comminution of the medial column is prone to failure.
To resolve the issues associated with medial reduction loss and varus deformity, two points were presented (1) using calcar screws to provide angular stability, facilitating the maintenance of humeral head reduction; and (2) recovering the continuity of the medial cortex to provide cortical support for the humeral head. The available literature has reported that the use of medial support screws, screw-tip augmentation with bone cement and bone grafts are currently the most frequently used and assessed tools to increase the stability of the locking plate complex[23–25]. However, their functional outcomes are unpredictable and inconsistent[26, 27].
Additionally, the use of the double-plating technique has been reported in previous studies. To achieve medial stability, Wanner et al. suggested a technique with the additional use of one-third of tubular plates positioned ventrally and at a right angle to the lateral adjusted standard plate. However, this technique resulted in a decrease in biomechanical stability compared to fixation with locking plate systems. The ventrally inserted plate may damage the blood supply of the arcuate artery, which is a branch of the anterior humeral circumflex artery. There are few reports in the literature describing the recovery of medial continuity using the medial plate directly. Park SG et al. showed that the use of additional medial buttress plate fixation in proximal humeral fractures with unstable medial column restoration after lateral locking compression plate fixation provided strong support for the medial column. However, an additional medial buttress T-plate was placed through only a typical anterolateral approach. Therefore, extensive surgical exposure and the use of large metal buttress T-plate implants may place the axillary and musculocutaneous nerves at risk. In this study, we used an anatomical locking plate to fix the medial column, which has two advantages. First, compared with those of anatomical plates, the locking mechanism and interference with the mechanical stability of the plate may be changed by plate bending during the operation. Furthermore, no intraoperative plate bending will reduce the operation time.
In this study, the combined application of minimally invasive lateral locking plate fixation and medial locking plate fixation was performed, which could provide a stable dual-column buttress for the treatment of proximal humerus fractures with unstable medial columns or osteoporosis. A potential problem of the double-plating technique seems to be the medial plate, which might damage the blood supply of the humeral head by compressing the branch of the circumflex artery. No humeral head necrosis was observed in this study, which suggests that blood supply is sufficient for survival of the humeral head. The PCHA and ACHA have been proven to be crucial for the blood supply of the humeral head. To avoid posttraumatic humeral head ischemia, the PCHA and ACAH were protected carefully during the operation in this study. According to a previous study and our anatomical study, there are no communicative branches between the PCHA and ACHA. In addition, beneath the humeral head, the distance between the PCHA and ACAH is approximately 25–30 mm. Therefore, both the PCHA and ACAH can be avoided by using this interval plane, and there is enough space to place the plate at the medial column. Moreover, a recent case report by Gerber et al. underlined the clinical relevance of intraosseous blood supply by anastomoses of the deep brachial artery because avascular necrosis of the humeral head was absent after posttraumatic rupture of the anterior and posterior humeral circumflex arteries. It should be emphasized that the surgeon must take care to avoid injury to the GHL. In addition, based on our measurement and experience, the exposed fore-and-aft region should not exceed 3 cm when the medial capsule is opened longitudinally to avoid injury to the ACHA and PCHA. However, we should admit that the medial approach remains difficult to learn because of the relatively complex anatomical structure and spatial relationships.
According to our observations, it is not complicated to place the medial column plate, and there is limited risk of injury to the nerve and vessels while fixing the plate and screws. Appropriate plating was needed for fixation of the medial column because of its relatively complex anatomic structure. The proximal portion of the plate should be plate between PCHA and ACHA, and its width should be limited to less than 25 mm to avoid those two vessels. In our cases, an anatomical locking plate was designed and used, and 3 locking screws were inserted into the humeral head with 2 locking screws inserted into the humeral shaft.
In this clinical trial, we describe not a completely new treatment strategy but a new surgical technique and the use of a new implant, to restore medial cortical support in proximal humeral fractures. The Constant score was used for the clinical evaluation of the results in this study. In this regard, we observed a mean Constant score of 82.8 24 months after the operation. Thus, the clinical results from our patient population appear to be satisfactory, and we consider our results relevant despite the small number of included patients.
This study has several limitations. First, the sample size was small, so it may not have been sufficient for accurately assessing the efficacy of this technique. Second, the follow-up period was short, and the biomechanics of this medial locking plating have not been studied. Third, there was no control group to permit a head-to-head comparison with other surgical techniques. Because of these limitations, more work needs to be done in terms of clinical and biomechanical studies.