Operative treatment of proximal and/or middle one-third humeral shaft fractures with a helical plate is a safe procedure with a good to excellent shoulder function at one-year follow-up. Contrary to conventional surgical techniques, we did not observe any iatrogenic radial nerve palsies while maintaining similar functional outcome measures and also obtaining excellent healing rates [3].
Although most humeral shaft fractures can be treated conservatively, conventional plate osteosynthesis is the gold standard in operative treatment of displaced proximal and/or middle one-third humeral shaft fractures. As a result, anatomic reduction could be attempted so angular deformity can be corrected in order to maximize cortical contact. However, radial nerve palsy is a serious complication following conventional plate osteosynthesis [12]. A helical plate could combine the benefits of anatomic reduction while also avoiding radial nerve palsy and deltoid insertion. This is obtained by the design of the plate, covering both the lateral side of the proximal third of the humerus, avoiding the long head of the biceps, and the anterior side of the middle/distal third of the humerus, avoiding the radial nerve and deltoid insertion [8]. Klepps S. et al noted that a release of more than one fifth of the anterior deltoid insertion could compromise the anterior deltoid [9].
Three nerves are at risk in open reduction internal fixation of humeral shaft fractures, in particular the radial, axillary and musculocutaneous nerves. A radial nerve palsy can be caused by trauma (i.e. fracture) or is due to surgery. Clinically, it will present itself as a loss of sensation of the dorsal hand, as well as loss of active extension of wrist and fingers at the metacarpophalangeal joints [13]. Artico M. et al. outlined the surgical anatomy of the radial nerve and showed it has a consistent distance of 121 (± 13) mm between the lateral humeral epicondyle to the lateral point of crossing the posterior aspect of the humerus [14]. Belayneh et al. showed no statistical differences in recovery time between nonoperative and iatrogenic radial nerve palsies [15]. Mean time to recovery of a complete palsy was 25.2 weeks, and surgical intervention did not lead to faster recovery [15]. Although management of iatrogenic radial nerve palsy is mainly conservative, sometimes late exploration is necessary if there is no spontaneous recovery at three to six months [13]. A systematic review by Shao et al. showed a full recovery rate of 88,1%, with a mean time to recovery of 6.1 months (range 3.4–12 mo.) [2].
Although there is a high overall recovery rate, the rehabilitation of patients is delayed with an average of six months and incomplete recovery can necessitate tendon transfers [16]. These arguments show that a radial nerve palsy is a high impact complication. One which could be completely avoided with the use of a helical plate due to its design, a benefit consistently confirmed in all case reports available [4, 17–19]. Recently, Da Silva et al. reported a 10-year retrospective study of 62 patients where no radial nerve damage was reported in the helical plate group [17]. This contrasts to conventional open reduction internal fixation (ORIF) where there are approximately 8,3 % iatrogenic radial nerve palsies [6]. However, this percentage can depend on the surgical approach used. Claessen et al. reported an iatrogenic radial nerve palsy of 22% when using a lateral approach, 4% with a anterolateral approach and 11% with a posterior approach [7]. However, a posterior approach is not suitable for fractures of proximal and combined proximal/middle humeral shaft fractures [20]. A posterior approach is well favored for distal or combined middle/distal humeral shaft fractures, due to the crossing of the radial nerve on the anterolateral aspect of the distal humerus between the brachialis and brachioradialis [20]. When using a helical plate for proximal and/or middle humeral shaft fractures, the high-risk lateral approach is not indicated anymore.
The axillary nerve, on the other hand, elongates depending on plate-bone distance. Dauwe et al. demonstrated on 42 fresh frozen cadaveric humeri that a helical plate significantly lowers plate-bone distance. This could imply less risk of nerve damage due to lower axillary nerve elongation [21].
Lastly, Gardner et al. described that the musculocutaneous nerve is most at risk when using a helical plate, due to its location on the anterior side of the middle/distal humerus. However, a ‘safe zone’ can be created due to a predictable and consistent anatomic location of the musculocutaneous nerve (99% CI: 12,2–18,8 cm distal from the greater tuberosity) [22].
In our study, all sixteen humeral fractures consolidated clinically and radiographically within three months. A consistency in healing rates can be found in other small case reports. Combined, all reported humeral shaft fractures in literature healed when treated with a helical plate [4, 18, 19, 23, 24]. Moreover, Yang et al. recorded a 100% healing rate in ten comminuted fractures of the proximal and/or middle one-third of the humerus [23]. In traditional operative management nonunion rates range from 0–9% [5]. A systematic review of Beeres et al. showed a nonunion rate of 8,5% in patients treated with ORIF [6]. These excellent healing rates of a helical plate could be related to biomechanical advantages. Krishna et al. described that a helical plate had a better gap closure in oblique fractures, reduced stress shielding, absorbed tensile stress caused by torsion and had a higher screw-holding power due to different orientation of screws [25].
However, in self-molded plates it is known that excessive deformation during contouring will damage the locking mechanism and has an impact on the fatigue properties of the plate [25]. Controlled manufacturing of helical plates could resolve this. In this study the pre-contoured A.L.P.S proximal humeral plating system (Zimmer Biomet®) was used and we are the first to report clinical results.
We reassessed twelve patients, after written informed consent, with a minimum of one-year follow-up. Functional outcome measures were taken and calculated by the same Orthopaedic Surgery resident. All measurements were made within one month to minimize intra-observer variability. Furthermore, it is known that the CMS has a high degree of reproducibility with a low intra-observer error of 3% [26]. Normalized CMS were comparable with those reported in other case reports (80 vs. 77, 88) [18, 24]. The systematic review of van de Wall et al. included two randomized controlled trials, for those treated operatively a mean DASH score of 15 was reported [3]. Brunner et al. used a PHILOS plate in 15 humeral shaft fractures with a median CMS of 74 (56–100) and a median DASH score 34 (24–48) [27]. These results are comparable with our mean DASH score (22 ± 19) and our mean normalized CMS (80 ± 19).
It is, however, important to note some limitations. First of all, a retrospective single center study design was used with a small study population. Consequently, only descriptive statistical analysis was made in between groups.
Secondly, this study consists of inhomogeneous patient characteristics and fracture types as can be seen in Table 1 and Table 2. Despite these differences, all fractures healed within three months and no radial nerve palsies were detected. On the other hand, these inhomogeneous patient characteristics can account for the large standard deviations in our functional outcome measures. Finally, functional outcome measures were taken at different follow-up times, however, a minimum of one year was respected.