Medial column instability has been identified as a key factor contributing to the increased failure rates observed after locking plate (LP) fixation for unstable proximal humerus fractures (PHFs) ^{[1, 2, 3]}. Medial column instability, primarily characterized by calcar comminution, is particularly severe in patients with osteoporosis ^[4]. Over the past decade, numerous augmentation strategies have been introduced to enhance medial column stabilization and improve outcomes in LP fixation including calcar screws ^[5], allogenic or autologous bone grafting ^{[6, 7]}, and cement augmentation ^[8]. Recent studies have reported promising biomechanical ^{[9, 10]} and clinical outcomes ^{[11, 12]} with the use of LP fixation augmented by an additional plate for medial column stabilization.

The current article provides an overview of the available clinical literature for dual plate fixation of unstable PHFs with medial column instability. To our knowledge, this is the first review article addressing the current topic.

For the arrangement of this narrative non-systematic review, an exploratory search in the MEDLINE (via PubMed) database using the keywords combinations: “proximal humerus fracture” and “dual plate”; “proximal humerus fracture” and “double plate”; “proximal humerus fracture” and “additional plate” was conducted. The search was originally performed in November 2024 to include the most recent literature. The selected studies were limited to English only. The results of the search were critically evaluated and clinical studies were included in a detailed review. Reference lists from the articles retrieved were further examined to identify any additional studies of interest. Inclusion criteria for the study were: clinical studies with proximal humerus fractures due to trauma; patients older than 18 years of age; more than 6 patients included in the study; at least 12 months follow-up; studies with fractures having medial column instability; patients treated only with proximal humerus locking plate (PHLP) and additional plate; Exclusion criteria considered: biomechanical (experimental) studies; studies on pathological fractures; patients younger than 18 years of age; studies with less than 6 included patients; studies with patient follow-up of less than 12 months; studies with fractures without having medial column instability; patients treated with proximal humerus locking plate (PHLP), additional plate and other augmentation, such as structural allograft or cement;

The initial search in Pubmed yielded 24 studies. 15 studies did not meet the inclusion criteria: 8 were biomechanical experiments; 1 study did not use PHLP; 3 studies had less than six patients; one study was available only in Mandarin; one study used PHLP, additional plate and structural allograft; One study had a follow-up period of just 6 months. As a result only 9 studies satisfied the inclusion criteria and were subject to further analysis. Two studies were performed prospectively ^[11,13,] while the other seven studies were conducted retrospectively.

There were 165 patients with PHFs in the nine included studies. The average age was 57 years (range 18–84 years). The average follow-up was 21.8 months (range 12–52 months). By fracture type, there were 32 (19.4%) two-part fractures, 64 (38.8%) three-part fractures, 61 (37%) four-part fractures and 8 (4.8%) fracture-dislocations.

Various plates were used as additional fixation to the PHLP: three studies employed a 1/3 tubular plate for supplementary fixation ^[14–16] with one study also using either a 2.4/2.7 mm reconstruction plate or a 1/3 tubular plate ^[15]. One study utilized a Variable Angle Locking Compression Plate (Distal Radius System, Synthes, Switzerland) ^[13], another study reported the use of a non-locking 3-hole T-plate (Zimmer or Synthes) ^[17], and a 2.7 mm T-shaped locking plate was employed in one study ^[18]. One study used an anatomical medial locking plate (Waston, China) ^[19], and a 2.7 mm micro-locking plate (Synthes Inc., Paoli, PA, USA) was used in one study ^[12]. Finally, a locking three-hole T-plate (Zimmer, Inc Warsaw, IN or Synthes) plate was described in one study ^[20].

Eight authors used the delto-pectoral approach and one author used anterolateral (MIPO) approach with a “novel” medial approach ^[19].

In all 9 studies, rehabilitation with passive and/or active-assisted range of motion began as soon as patients were comfortable and pain-free.

Absolute Constant-Murley Score (CMS) was used for functional evaluation in 8 out of 9 studies. Relative CMS calculated by using reference values of the respective age and gender group described by Constant and Murley was used in 1 study ^[14]. The average CMS at the final follow-up for the patients who had undergone dual plate fixation for PHFs was 79.9 (71.5–90.4).

There were a total of 12 (7.3%) reported clinically relevant complications and 15 (9.1%) re-interventions.

The results and complications for each separate study are detailed in Table 1.

The medial calcar support consists of two main components: the length of the posteromedial metaphyseal extension and the integrity of the medial hinge ^[21].

According to Gardner et al., positioning a locking plate along the lateral cortex of the proximal humerus creates a mechanical construct that acts as a tension band. When the rotator cuff activates and exerts varus-deforming forces on the humeral head, these forces can be redirected into medial compression forces. This process effectively reduces the load on the implant, promoting a load-sharing mechanism between the implant and the bone. However, the mechanical stability of the construct depends on the integrity and alignment of the medial cortex, which must be capable of effectively transmitting the load ^{[22, 23]}. Beyond its mechanical function, the medial calcar plays a crucial role in bone biology by supporting blood flow to the humeral head through the vessels of the posteromedial hinge ^{[22, 24]}.

In an experimental study by Ponce et al., medial comminution was found to reduce the load to failure by 48% compared to cadaveric specimens with an intact medial cortex ^[25]. In a clinical study Osterhoff compares the functional outcomes between patients with PHFs with and without metaphyseal comminution treated with LP. Patients with medial cortex comminution exhibit lower absolute CMS (CSabs < 65) compared to the average for both groups (72.4) ^[26].

The simplest approach of achieving medial support is by medializing the diaphysis and laterally impacting the humeral head, or simply fixing it in varus ^[23]. Although biomechanically justified, these techniques could theoretically lead to complications, pain, and impaired shoulder function ^[25]. The other simple method involves placing locking screws in the inferomedial quadrant (calcar screws) of the humeral head ^[23]. However, Gardner reports a 29% complication rate, despite the use of calcar screws ^[23].

Supplemental fibular allograft augmentation for medial column insufficiency has been shown to improve both radiological and clinical outcomes while reducing the rate of complications ^{[27, 28]}. However, fibular allograft carries potential risks, including infection, disease transmission, and the possibility of fracture during insertion. Despite its benefits, its high cost and availability constraints may make it less accessible in all healthcare settings ^[29–31].

Cement augmentation for PHFs, especially in elderly patients, can provide enhanced stability, but it does carry several risks. Potential complications include cement leakage, which can lead to adjacent joint or soft tissue damage, and the risk of thermal injury due to the high temperatures generated during cement setting, which may cause necrosis of the bone tissue. Additionally, while cement augmentation may improve fixation in osteoporotic bone, its use remains controversial, with concerns about the long-term effectiveness and the risk of fractures or delayed healing in some cases ^{[8, 32]}.

To address the limitations of previously mentioned techniques, several authors have proposed using an additional plate to augment medial column insufficiency ^[12–20] (Fig. 1). Warner et al. introduced double-plate fixation for unstable proximal PHFs, utilizing a 1/3 tubular plate placed laterally and another positioned ventrally at a 90° angle to the first. This configuration demonstrated high internal stability, enabling early shoulder mobilization in all patients treated with this method ^[33]. However, subsequent biomechanical study comparing the use of two 1/3 tubular plates with a proximal humeral locking plate (PHLP) found that the PHLP offers superior mechanical stability, providing significantly greater stiffness and exhibited less irreversible deformation ^[34]. Further biomechanical studies comparing PHLP alone to PHLP combined with a second plate have demonstrated that the additional medial column support provided by the second plate enhances stability in the proximal humerus ^{[9, 10, 35–38]}.

The majority of authors who use an additional plate to the PHLP for treating unstable PHFs with medial column instability report good clinical results with little or no complications (Table 1). As shown in Table 1, the positioning of plates varies between studies, with each placement requiring additional soft tissue dissection regardless of its location ^{[14, 18]}. A ventrally positioned plate carries the risk of compromising the blood flow through the anterior circumflex humeral artery (ACHA). Similarly, a medially positioned plate may endanger the blood supply to the humeral head by compressing branches of the anterior or posterior circumflex humeral arteries, both of which are vital for maintaining its vascular integrity ^{[19, 39]}. Anyway, the rate of avascular necrosis (AVN) in the studies of dual plating is relatively low. The highest incidence of AVN was reported in the study by Warnhoff et al., where the complication occurred in 2 (8%) out of 25 patients treated with a PHLP combined with an additional anterior one-third tubular plate ^[14]. The authors attributed this outcome primarily to the destruction of the medial calcar and the more complex fracture morphology observed in all cases ^[14]. Zhang et al. reported one case of AVN among 15 patients treated with an additional 2.7 mm locking T-plate positioned ventrally at the lesser tuberosity ^[18]. The authors highlighted a potential advantage of the locking mechanism of the anteromedial plate, which minimizes pressure on the periosteum, thereby preserving the remaining blood supply to the humeral head. This contrasts with the one-third tubular plate, which may exert greater pressure and potentially compromise vascular integrity ^[18]. In the study by Liu et al., which included the largest patient cohort, the authors reported that no complete AHCA was found beneath the lesser tuberosity, before plate fixation in the 37 patients. According to the authors, microplate fixation beneath the lesser tuberosity offers a distinct advantage in preventing excessive medial placement, which could otherwise further compromise the blood supply to the comminuted bone fragments of the medial calcar and the posterior circumflex humeral artery (PCHA). As a final result none of the patients had humeral head necrosis at the last follow-up visit. In contrast, a study by Wang et al., which utilized a medial plate in eight patients, reported no complications, including AVN^[19]. This clinical study was preceded by a biomechanical experiment using the same anatomically designed plate and an anatomical study for a newly developed medial approach ^[35]. The authors observed that there are no communicative branches between the PCHA and the ACHA. Additionally, beneath the humeral head, the distance between the PCHA and ACHA was approximately 25–30 mm. This anatomical gap allows for the use of an interval plane to avoid both arteries while providing sufficient space to position the plate along the medial column ^[19]. Seok et al. utilized a locking three-hole T-plate, contoured to the medial cortex, to buttress the medial metaphyseal head extension in 26 PHFs with medial comminution and varus deformity in patients with osteoporosis. To minimize the risk of ACHA injury during bone exposure, the plate was placed on the latissimus dorsi tendon. The authors reported that the plate‘s position did not significantly affect postoperative function or range of motion, with a mean CMS of 71.5. Furthermore, no cases of iatrogenic nerve injury, vascular injury, or AVN were observed. In comparison, functional outcomes were worse and complication rates were higher in similar fractures and patient characteristics treated with PHLP alone ^[15].

As a conclusion from the studies in the present review, the anterior position of the second plate could possibly be more dangerous for the humeral head supply.

Another potential drawback of dual plate fixation for PHFs is the possible restriction in range of motion, primarily due to mechanical impingement and, secondly, muscle violation, particularly with anterior or posterior plate placement. In the study by Choi et al., the authors dissected the subscapularis or supraspinatus muscles to position the VA-LCP distal radius plate anteriorly or posteriorly, respectively, as a buttress to prevent anterior-posterior angulation of the humeral head. The muscle tendons were reattached to the plate after its fixation. However, the functional results were excellent with CMS of 90.4 points ^[13]. Zhang et al. stripped the attachment of pectoralis major for 2–3 cm and repaired it by suturing to the plate or by drilling the humeral cortical bone after fracture reconstruction. If necessary, the subscapularis tendon was also partially released from the lesser tuberosity. The CMS was good – 79.8 ^[18].

One advantage of the anterior plate, compared to the medial and posterior plates, is that it is easier to place and can serve as a temporary fixation method before the placement of the PHLP^{[12, 16]}.

The present study has some limitations. First, it was limited by its narrative design, which may have introduced selection bias. Second, the number of included studies was relatively small. Third, the majority of the studies included in the final analysis were retrospective case series, and further research with higher levels of evidence is needed to confirm these findings. However, this study also has several strengths. One strength is the strict inclusion criteria used, which helped minimize potential bias. Another strength is the large total number of patients, which is adequate for a meaningful analysis of the results. Lastly, the mean follow-up period in the included studies is sufficient to objectively assess functional outcomes and detect any subsequent complications.

Adding a second plate to PHLP fixation for unstable PHFs with medial column insufficiency has shown promising clinical outcomes with a relatively low complication rate, despite the additional soft tissue dissection required. The anterior placement of the secondary plate is less technically demanding and can serve as a temporary reposition tool. However, anterior plating has been associated with a higher risk of AVN compared to medial plating. The variability in techniques and implants used for dual plating complicates drawing definitive conclusions. Moreover, the current evidence is limited, and additional studies with a higher level of evidence are needed to support the efficacy and routine application of the dual plating technique for unstable PHFs.