The proximal humerus represents the third most common fracture site in patients older than 65 years, following only the hip and distal radius. In addition, proximal humerus fractures carry a more distinct age-related risk than other fractures of the upper limb.1,2 Proximal humerus fractures currently represent 5% of all fractures,3 and it is estimated that there will be a 3-fold increase in the incidence of these fractures during the next 30 years as the population ages.4,5 Although many of these fractures are nondisplaced or minimally displaced and can be treated conservatively, up to 20% will require operative treatment.6 Surgical options consist of open reduction and internal fixation, hemiarthroplasty, reverse total shoulder arthroplasty, and minimally invasive techniques such as percutaneous pinning, screw osteosynthesis, or intramedullary nailing.7
In the early 2000s, fixed-angle plates were developed. Proximal humeral locking plates with unicortical metaphyseal locking screws have been used for nearly a decade for the treatment of fractures of the proximal humerus. In biomechanical analyses, the stability of proximal humeral locking plates has shown superiority to that of conventional plating, intramedullary fixation, and blade plating.8–10 Benefits include lower rates of loss of fixation and angular deformities. Complication rates, however, range from 20% to 40%, and a reoperation rate of 25% has been reported in recent literature.11 The implant-related (impingement, screw loosening, and screw perforation of the humeral head) complication rate has been estimated to be between 6% and 30%, with a screw cutout rate of 11%.2 Südkamp et al12 reported a 34% (52 of 155) complication rate, with 40% (25 of 62) of those complications related to incorrect surgical technique.12 Specifically, they found that of 155 patients who underwent fixation with proximal humeral locking plates, 21 (13.5%) had primary screw perforation.
There remains a paucity of data that surgeons can use to determine the optimum number of locking screws. Although there is a theoretical increased risk of loss of fixation with fewer screws, as described above, there is also a risk of humeral head penetration associated with each additional screw placed.12,13 Therefore, it becomes desirable to place as few screws as possible without sacrificing strength of fixation. The purpose of this study was to determine the cyclic compressive displacement and failure characteristics of an osteotomy gap model of proximal humeri fixed with either 5 or 7 metaphyseal locking screws. The authors hypothesized that (1) 5 locking screws are potentially equal to 7 screws in biomechanical strength and stability of the construct, as shown by displacement across the osteotomy gap as well as stiffness and ultimate load on failure testing; and (2) there is no difference in the rate of screw cutout in the humeral head between 5 and 7 locking screws. If the stability and failure rates of 5 and 7 screws are similar, fewer screws may be used and rates of intraoperative complications could potentially be decreased.
Materials and Methods
For biomechanical testing, 10 matched pairs (20 total) of adult embalmed cadaveric humeri were used. Characteristics of the donors are listed in the Table. The ages of the donors ranged from 66 to 96 years (mean, 84.6 years). Prior to the start of the study, plain anteroposterior radiographs were obtained for all specimens to ensure that there were no large bony irregularities or noticeable differences in density within matched pairs. Proximal humeral locking plates (AxSOS; Stryker, Mahwah, New Jersey) were placed at the location of best fit, approximately 2 mm lateral to the bicipital groove and 10 mm distal to the apex of the greater tuberosity, according to the manufacturer's surgical technique manual. For diaphyseal fixation, three 3.5-mm diaphyseal nonlocking screws were placed in standard bicortical fashion in each specimen. In 1 humerus from each pair, five 4.0-mm monoaxial locking screws of appropriate length were placed in the humeral head, including an inferomedial calcar screw. In the other specimen from each pair, this procedure was repeated with seven 4.0-mm monoaxial locking screws placed in the humeral head (Figure 1). Pairs were randomized by drawing numbers to give half of the pairs 5 screws in the left humerus (7 in the right) and the other half 5 screws in the right humerus (7 in the left).
Age at Death and Sex of Paired Cadaveric Humeri
Placement of plates with segmental defect. Note the position of the 5 locking screws, including the calcar screw (A). A specimen containing 7 locking screws (B). A specimen mounted on testing apparatus. Differential variable reluctance transducer sensors are seen attached (C).
Experimentally standardized fractures were simulated using a protocol previously described by Lescheid et al.14 A 1-cm horizontal osteotomy gap defect was created to reproducibly create fractures simulating loss of medial cortical contact (Figure 1). The distal diaphyses of the humeri were cut 28 cm from the tip of the articular surface, removing the condyles. This allowed for potting in commercially available resin cement to a standard depth of 14 cm.
One differential variable reluctance transducer (DVRT) was placed along the medial cortex and another along the posterior cortex of the osteotomy sites in a reproducible location (Figure 1). These recorded gap displacement in 2 planes. The testing protocol described by Koval et al15 was modified so that a load was applied axially with 0° of abduction, as in the study by Schumer et al,16 to mimic the resultant glenohumeral force vectors predicted to occur with abduction.17 Stresses were applied for 1000 cycles at 1 Hz between 50 N and 250 N using a servohydraulic materials testing machine (model 812; MTS Systems Corporation, Eden Prairie, Minnesota). Data were recorded from the DVRTs and from the vertical displacement sensor on the servohydraulic materials testing machine.
Specimens with 5 initial metaphyseal locking screws were tested sequentially for 1000 cycles each with 5, 6, and then 7 screws in each test phase for a total of 3000 cycles. Specimens with 7 initial metaphyseal screws were tested sequentially with 7, 6, and then 5 screws in each test phase. This allowed for comparison of cyclic displacement with screw number within the same specimen while also counterbalancing for any differences in plastic deformation or loosening of the implants within the pairs of specimens. Data from the posterior and medial DVRTs as well as the servohydraulic materials testing machine were recorded using LabVIEW data acquisition software (National Instruments, Austin, Texas). Cyclic displacement for each DVRT was recorded as the accumulated displacement from the first to last cycle during each test phase. After cyclic loading was completed, the specimens were loaded to failure at 0.1 mm/s. Load displacement curves were generated and recorded. Failure was defined as discontinuity in the load-displacement curve or medial contact across the osteotomy site. The linear stiffness was measured as regression of the load-displacement data over the load limits of 25% to 75% of the ultimate load.
Statistical analyses of the cyclic displacement data and failure properties were performed using a paired Student's t test. If the test for normality failed, a Wilcoxon signed rank test was used. Statistical significance was set at P<.05. A pretest power analysis determined that a sample size of 8 specimens per group would be required to detect a 30% change in ultimate load assuming a standard deviation of 25% of the mean for a power of 0.80 and significance level of .05.
Following the completion of mechanical testing, specimens were examined with the screws and the implant still in place to evaluate the method of failure and any cutout or screw penetration. The articular surface was removed with a reciprocating saw to visualize screw displacement or screw cutout tracks in the metaphysis.
Nine of the 10 pairs were included in the analysis. In 1 pair, a specimen failed during the cyclic testing at lower loads; this pair was excluded because the failure was thought to be due to an existing diaphyseal crack. Owing to the study design, every specimen contained 6 screws during the second test phase of the 3 cyclic loading test phases and could not be compared in a pairwise fashion to samples from other test phases because of decrease in displacement from settling noted between each test phase. Therefore, displacement data for 5 and 7 screws were compared separately within the first and third test phase. In the first test phase, there was no difference in mean cyclic displacement between 5 and 7 screws for either medial DVRT (1.09 mm vs 1.12 mm, P=.834) or posterior DVRT (0.45 mm vs 0.42 mm, P=.791) placement. Similarly, in the third test phase, there was no difference in mean cyclic displacement between 5 and 7 screws for either medial DVRT (0.91 mm vs 0.72 mm, P=.098 using Wilcoxon signed rank test) or posterior DVRT (0.27 mm vs 0.21 mm, P=.293) placement. Differences in displacement are shown in Figure 2. In addition, there was no difference in cyclic displacement between 5 and 7 locking screws when evaluated by paired comparison within a specimen for either medial or posterior DVRT displacement.
The mean compressive displacement of posterior and medial differential variable reluctance transducers (DVRTs) was not significantly different between 5 and 7 screws in either the first phase or the last phase of testing.
On testing to failure, 7 screws, as compared with 5 screws, showed similar stiffness (336 N/mm vs 292 N/mm, P=.176), force at ultimate load (745 N vs 662 N, P=.309), and displacement at ultimate load (5.90 mm vs 4.36 mm, P=.080). There were no statistically significant differences noted between the 5- and 7-screw specimens. A Wilcoxon signed rank test of energy at ultimate load showed significantly (P=.039) greater energy in humeri containing 7 screws (median, 3535 N*mm) than in humeri containing 5 screws (median, 2026 N*mm). Within pairs, there was inconsistency of which sample exhibited greater load to failure (Figure 3). In 4 of 9 pairs, the sample within the pair that failed at a higher load contained 5 screws. Graphs of ultimate load, stiffness, displacement, and energy to ultimate load are shown in Figure 4. Importantly, all samples failed at the diaphyseal fixation, so any interpretation of difference related to metaphyseal locking screw number should be undertaken cautiously. No screw cutout was visualized prior to or after articular surface removal.
Differences in loads at failure within pairs were inconsistent and were, on average, similar between specimens containing 5 and 7 screws.
There was no significant difference in load to failure (A), stiffness (B), or displacement to ultimate load (C). Energy to ultimate load was significantly different with screw number (P<.05; Wilcoxon signed rank test) (D).
Proximal humeral locking plates continue to be a useful treatment option but are not without challenges. The rate of implant-related complications remains high and is related to both intraoperative complications and postoperative failure of constructs. Specifically, high rates of intra-articular perforation by locking screws have been described in multiple series.12,13 A multicenter analysis by Brunner et al13 showed primary screw perforation to be the single most frequent complication in proximal humeral locking plate placement, occurring in 14% (22 of 158) of cases. Although Erhardt et al18 reported that 5 screws (including an inferomedial screw) created the strongest construct in their biomechanical study, they did not explore the use of more than 5 screws. The current study adds to the knowledge gained from Erhardt et al18 by showing that additional screws above 5 do not increase strength of fixation or decrease implant failure, including screw perforation and cutout. The authors observed that 5 screws with an inferomedial screw were stronger than the diaphyseal fixation with 3 bicortical screws and that all samples failed in the diaphysis prior to metaphyseal cutout.
Using univariate and multivariate logistic regression modeling, Jung et al19 found that osteoporosis, initial varus displacement, medial comminution, and insufficient medial support were the primary independent risk factors for construct failure. Of those factors, only medial support is a modifiable risk factor. Additionally, several biomechanical studies have shown improved strength of fixation with inferomedial screw support or medial cortical support with calcium phosphate cement or bone blocks.18–24 Erhardt et al18 found that a construct that included an inferomedial screw increased strength of fixation before cutout in a cadaveric model. Large clinical studies are lacking, but at least 1 small study24 suggests that restoration of the calcar does significantly reduce complications. The current authors used inferomedial screws in all samples, which likely contributed to the stability of the constructs. The data suggest that much of the stability may come from this calcar restoration and that greater than 5 total screws does add significant strength, provided 1 is an inferomedial screw.
In any biomechanical study, investigators must make decisions regarding testing setup. Further, these studies are limited by imperfect representation of complex anatomic motion and inconsistent and unpredictable patient actions seen in clinical practice. This study was limited by unidirectional force application and a single position of the specimens during testing, a problem common to many biomechanical studies. Poppen and Walker17 demonstrated force vectors on the proximal humerus during elevation in the scapular plane and that, at 30° or more of elevation, the resultant vectors of the deltoid, rotator cuff, and other stabilizers remain close to parallel with the shaft of the humerus. The current model used compression parallel to the shaft, which mirrors the force vectors shown in Poppen and Walker's data.
Additional limitations of this study included the use of embalmed cadaver specimens that were frozen and thawed, which could have had deleterious effects on their biomechanical properties. Also, in clinical practice, the major failure mechanisms are screw cutout within the humeral head or shaft leading to varus collapse. It was anticipated that with the authors' loading model, screw cutout in the humeral head and varus collapse would occur; however, only loss of diaphyseal screw fixation with varus collapse was observed. One possible explanation is the use of cortical nonlocking screws for diaphyseal fixation in this model, as opposed to the use of locking screws for the entire construct. An alter-native explanation is that because of the absence of soft tissues allowing direct visualization of the whole bone while plating as well as a relatively simple fracture model, the authors were able to obtain better fixation than is clinically realistic in patients with complex fractures. The benefit of this study, in contrast to some previous cadaveric models or synthetic bone samples, is that the authors were able to use specimens from elderly individuals, a population especially at risk for failure of fixation.
This study of cadaveric humeri from elderly patients showed no statistically significant differences in cyclic displacement or strength of fixation between 5 and 7 proximal humeral locking screws in the AxSOS locking plate system. The authors believe this is an important finding that should be explored with additional biomechanical and clinical trials. On the basis of these results, the authors would recommend using no more than 5 metaphyseal locking screws, including 1 inferomedial calcar screw, for fixation of proximal humeral fractures.
- Baron JA, Barret JA, Karagas MR. The epidemiology of peripheral fractures. Bone. 1996; 18(3)(suppl):209S–213S. doi:10.1016/8756-3282(95)00504-8 [CrossRef]
- Thanasas C, Kontakis G, Angoules A, Limb D, Giannoudis P. Treatment of proximal humerus fractures with locking plates: a systematic review. J Shoulder Elbow Surg. 2009; 18(6):837–844. doi:10.1016/j.jse.2009.06.004 [CrossRef]
- Court-Brown CM, Garg A, McQueen MM. The epidemiology of proximal humeral fractures. Acta Orthop Scand. 2001; 72(4):365–371. doi:10.1080/000164701753542023 [CrossRef]
- Aaron D, Shatsky J, Paredes JC, Jiang C, Parsons BO, Flatow EL. Proximal humeral fractures: internal fixation. J Bone Joint Surg Am. 2012; 94(24):2280–2288.
- Shulman BS, Egol KA. Open reduction internal fixation for proximal humeral fractures: indications, techniques, and pitfalls. Bull Hosp Jt Dis (2013). 2013; 71(suppl 2):S54–S59.
- Robinson CM, Amin AK, Godley KC, Murray IR, White TO. Modern perspectives of open reduction and plate fixation of proximal humerus fractures. J Orthop Trauma. 2011; 25(10):618–629. doi:10.1097/BOT.0b013e31821c0a2f [CrossRef]
- Min W, Davidovitch RI, Tejwani NC. Three and four-part proximal humeral fractures: evolution to operative care. Bull NYU Hosp Jt Dis. 2012; 70(1):25–34.
- Chudik SC, Weinhold P, Dahners LE. Fixed-angle plate fixation in simulated fractures of the proximal humerus: a biomechanical study of a new device. J Shoulder Elbow Surg. 2003; 12(6):578–588. doi:10.1016/S1058-2746(03)00217-9 [CrossRef]
- Edwards SL, Wilson NA, Zhang LQ, Flores S, Merk BR. Two-part surgical neck fractures of the proximal part of the humerus: a biomechanical evaluation of two fixation techniques. J Bone Joint Surg Am. 2006; 88(10):2258–2264.
- Siffri PC, Peindl RD, Coley ER, Norton J, Connor PM, Kellam JF. Biomechanical analysis of blade plate versus locking plate fixation for a proximal humerus fracture: comparison using cadaveric and synthetic humeri. J Orthop Trauma. 2006; 20(8):547–554. doi:10.1097/01.bot.0000244997.52751.58 [CrossRef]
- Jost B, Spross C, Grehn H, Gerber C. Locking plate fixation of fractures of the proximal humerus: analysis of complications, revision strategies and outcome. J Shoulder Elbow Surg. 2013; 22(4):542–549. doi:10.1016/j.jse.2012.06.008 [CrossRef]
- Südkamp N, Bayer J, Hepp P, et al. Open reduction and internal fixation of proximal humeral fractures with the use of the locking proximal humerus plate: results of a prospective, multicenter, observational study. J Bone Joint Surg Am. 2009; 91(6):1320–1328. doi:10.2106/JBJS.H.00006 [CrossRef]
- Brunner F, Sommer C, Bahrs C, et al. Open reduction and internal fixation of proximal humerus fractures using a proximal humeral locked plate: a prospective multicenter analysis. J Orthop Trauma. 2009; 23(3):163–172. doi:10.1097/BOT.0b013e3181920e5b [CrossRef]
- Lescheid J, Zdero R, Shah S, Kuzyk PR, Schemitsch EH. The biomechanics of locked plating for repairing proximal humerus fractures with or without medial cortical support. J Trauma. 2010; 69(5):1235–1242. doi:10.1097/TA.0b013e3181beed96 [CrossRef]
- Koval KJ, Blair B, Takei R, Kummer FJ, Zuckerman JD. Surgical neck fractures of the proximal humerus: a laboratory evaluation of ten fixation techniques. J Trauma. 1996; 40(5):778–783. doi:10.1097/00005373-199605000-00017 [CrossRef]
- Schumer RA, Muckley KL, Markert RJ, et al. Biomechanical comparison of a proximal humeral locking plate using two methods of head fixation. J Shoulder Elbow Surg. 2010; 19(4):495–501. doi:10.1016/j.jse.2009.11.003 [CrossRef]
- Poppen NK, Walker PS. Forces at the glenohumeral joint in abduction. Clin Orthop Relat Res. 1978; 135:165–170.
- Erhardt JB, Stoffel K, Kampshoff J, Badur N, Yates P, Kuster MS. The position and number of screws influence screw perforation of the humeral head in modern locking plates: a cadaver study. J Orthop Trauma. 2012; 26(10):e188–e192. doi:10.1097/BOT.0b013e31823db922 [CrossRef]
- Jung SW, Shim SB, Kim HM, Lee JH, Lim HS. Factors that influence reduction loss in proximal humerus fracture surgery. J Orthop Trauma. 2015; 29(6):276–282. doi:10.1097/BOT.0000000000000252 [CrossRef]
- Gardner MJ, Weil Y, Barker JU, Kelly BT, Helfet DL, Lorich DG. The importance of medial support in locked plating of proximal humerus fractures. J Orthop Trauma. 2007; 21(3):185–191. doi:10.1097/BOT.0b013e3180333094 [CrossRef]
- Ponce BA, Thompson KJ, Raghava P, et al. The role of medial comminution and calcar restoration in varus collapse of proximal humeral fractures treated with locking plates. J Bone Joint Surg Am. 2013; 95(16):e113(1–7). doi:10.2106/JBJS.K.00202 [CrossRef]
- Osterhoff G, Baumgartner D, Favre P, et al. Medial support by fibula bone graft in angular stable plate fixation of proximal humeral fractures: an in vitro study with synthetic bone. J Shoulder Elbow Surg. 2011; 20(5):740–746. doi:10.1016/j.jse.2010.10.040 [CrossRef]
- Gradl G, Knobe M, Stoffel M, Prescher A, Dirrichs T, Pape HC. Biomechanical evaluation of locking plate fixation of proximal humeral fractures augmented with calcium phosphate cement. J Orthop Trauma. 2013; 27(7):399–404. doi:10.1097/BOT.0b013e318278c595 [CrossRef]
- Jung WB, Moon ES, Kim SK, Kovacevic D, Kim MS. Does medial support decrease major complications of unstable proximal humerus fractures treated with locking plate?BMC Musculoskelet Disord. 2013; 14:102. doi:10.1186/1471-2474-14-102 [CrossRef]
Age at Death and Sex of Paired Cadaveric Humeri
|Specimen No.||Age, y||Sex|
|Overall||Mean: 84.6||60% female|