Could direct transdeltoid approach to severely displaced proximal humerus fracture be advantageous for a better reduction?

BACKGROUND: Surgical treatment of proximal humeral fractures (PHF) is a challenge for orthopaedic surgeons. Despite the wide application of open reduction and internal fixation with locking plates, the optimal surgical approach of PHF is still debated. This study aims to evaluate the radiological outcomes, defined as anatomical restoration of the greater tuberosity and humeral head-shaft angle, of the deltopectoral (DPA) and the lateral transdeltoid (LTA) approaches in three- and four-part PHF, treated with locking plate.
MATERIALS AND METHODS: This retrospective series review identifies 74 PHF surgically treated between January 2012 and December 2019. Patients were divided into two groups according to the surgical approach (DPA vs LTA). Demographic data, duration of surgery, radiological pre- and post-surgery parameters (greater tuberosity displacement and humeral head-shaft angle) were collected. The association between the surgical approach and the quality of fractures reduction was assessed.
RESULTS: The use of LTA approach correlates with a better reduction of greater tuberosity displacements compare to DPA (63% in DPA vs 100% LTA). No significant association was found with the humeral head-shaft angle (restored in 89% of the patients in DPA and 86% in LTA group), and surgical times (range 40 - 210 minutes ± DS 33,56 for the DPA; range 45 - 170 minutes ± 29,60 for LTA).
CONCLUSIONS: The results of this radiological study suggest that PHF with significant displacement of the grater tuberosity could benefit from the adoption of a lateral transdeltoid approach for the ORIF procedure. Further studies are needed to confirm these findings.

Introduction

Proximal humeral fracture (PHF) accounts for 5% of all fractures and, after distal radius and hip fractures represents the most common fracture of the extremities (1).

They are usually related to low-energy trauma occurring in elderly patients, particularly women affected by post-menopausal osteoporosis (2-5). The incidence of this kind of fracture is expected to rise since the geriatric population is growing continuously (6).

Despite various kinds of treatment strategies available, the management of complex PHF remains demanding.

Conservative treatment with short-term immobilization in bandage has been a well experienced treatment option in stable and simple fractures, but it has been recently demonstrated to be a good option for complex fractures in low-demanding elder patients (7).

The main surgical choices are open reduction with internal fixation (ORIF), closed reduction and internal fixation, minimally invasive percutaneous plating osteosynthesis (MIPPO), closed reduction and external fixation and arthroplasty (8). Each approach may be appropriate based on the specific fracture's and patient's "personality" (9,10).

ORIF technique with locking plate has been developed and refined in the last decade, and according to the literature, it leads to a good clinical and functional outcome for even the most complex fractures such as 3- 4 parts pattern according to Neer classification (9,11).

Another controversial topic for surgery is the choice of surgical approach.

The deltopectoral approach (DPA) remains widely used because of its excellent exposition of the anterior structure with limited concern about injuring the axillary nerve. Furthermore is more convenient for a potential intraoperative conversion to arthroplasty (12). However, this approach could not extensively expose the lateral-posterior aspect of the proximal humerus, provides a disadvantageous lever arm for the screws (due to the anterior-lateral plate location), and involves extensive soft tissue dissection and muscle retraction which may increase the risk of avascular necrosis (13).

The direct lateral transdeltoid approach (LTA), instead, is less invasive, provides a more advantageous lever arm for the screws, and permits a direct lateral view of the humeral greater tuberosity. On the other hand, this approach could lead to axillary nerve injury.

The goal of a good reduction of any PHF is to restore the humeral head-shaft angle and the correct position of the greater tuberosity in order to achieve a better clinical outcome (14-18).

The aim of this study was to retrospectively evaluate the radiological outcomes comparing direct lateral transdeltoid and deltopectoral approaches in three-and four-part PHF according to Neer classification, treated by ORIF technique with locking plate (11).

Ethical Approval: Patient data was retrospectively analysed and did not change patient care. Ethical Committee approval was therefore deemed unnecessary.

Materials and methods

All participants provided written informed consent to participate in this study. This study was con-ducted under the principles of the Declaration of Helsinki. We retrospectively selected from hospital's records 257 PHF surgically treated between January 2012 and December 2019. Inclusion criteria were three-part fracture with surgical neck and greater tuberosity involvement or four-part fracture with involvement of surgical neck, greater and lesser tuber-osity according to the Neer classification, treated by open reduction and internal fixation using a locking compression plate (Philos, Synthes, Oberdorf, Switzerland), the availability of preoperative X-ray and CT scan, and postoperative radiographs, age of 18 years or older. Exclusion criteria were inability to acquire the planned imaging, minimal displacement PHF or two-part fracture, type 5 and fracture-dislocation of the humeral head according to the Neer classification, pediatric fractures (physeal injuries), osteosynthesis with intramedullary nail, percutaneous pinning, external fixation, MIPPO and shoulder arthroplasty replacement (Table 1).

Table 1.

Study inclusion and exclusion criteria

INCLUSION CRITERIA	EXCLUSION CRITERIA
- Three-part fracture with surgical neck and greater tuberosity involvement according to the Neer classification- Four-part fracture with involvement of surgical neck, greater and lesser tuberosity according to the Neer classification	- Minimal displacement PHF Two-part fracture according to the Neer classification- Type 5 and fracture-dislocation of the humeral head according to the Neer classification
- ORIF using locking compression plate	- Osteosynthesis with intramedullary nail, Percutaneous pinning, Ex-fix, MIPPO- Shoulder arthroplasty replacement
- Pre- and postoperative radiographs- Preoperative CT scan	Inability to acquire:- Pre- and postoperative X-ray- Preoperative CT scan
	- Pediatric fractures

According to inclusion criteria we recruited 74 patients (Fig. 1).

Figure 1.

Patient selection algorithm. The blue circles show the progressive patients selection flow. The red boxes show the number and the criteria of the excluded patients.

For each patient were recorded: demographic data, duration of surgery, side of the fracture and surgical approach.

These patients were divided into 2 groups based on the surgical approach: deltopectoral (group A-DPA) or direct lateral transdeltoid approach (group B-LTA).

Two Authors independently classified the fractures' pattern by preoperative CT scans from hospital's records and evaluated plane radiographs preoperative-ly and at the last follow-up, measuring the greater tu-berosity fracture displacement (> or < 5mm), the humeral head-shaft angle (anatomical range 120°-145°) (14-19).

Statistical analysis was performed using R v3.6.3 (https://www.r-project.org).

The association between surgical approach and duration of surgery was detected using the Wilcox-on-Mann-Whitney test.

The association between surgical approach and surgeon's chance to reduce the fracture into normal ranges was detected using the Fisher test for dichoto-mous nominal variables.

Surgical approaches and Technique

Four senior surgeons, who were trained in both surgical approaches, operated both groups using the beach-chair position on an OPT 100 table with modular helmet headrest (Opt surgicalsystems ® Calliano (TN), Italy). The mobile C-arm with image intensifier was positioned at the head of the patient, on the ho-molateral side. Fluoroscopy was carried out in antero-posterior and, when possible, axillary view to define fragment configuration, position, and size.

DPA (Fig. 2 A-G) consists of a straight skin incision starting from the coracoid process, following the delto-pectoral sulcus. An incision of the subcutaneous tissue is made until the fascial plane is reached. Identified and protected the cephalic vein, an incision of the fascia is made following the superficial internervous plane consisting of the deltoid muscle laterally (axillary nerve) and the pectoralis major muscle medially (medial and lateral pectoral nerve).

Figure 2.

Representative intra-operative images from DPA group. (A) Surface marking of bony landmarks; (B) skin incision following the delto-pectoral sulcus and careful hemostasis; (C-D) incision of the fascia and isolation of the cephalic vein; (E) identification the internervous plane between deltoid muscle (axillary nerve) laterally and pectoralis major muscle (medial and lateral pectoral nerves) medially; (F) exposure the tendon of the subscapularis muscle; (G) capsule incision and intra-articular access.

The deltoid muscle is separated from the pecto-ralis muscle and the cephalic vein is mobilized either medially or laterally as needed. The clavipectoral fascia is opened, allowing the identification of the lateral margin of the conjoined tendon. Below this structure runs the subscapularis muscle that covers the joint capsule. It is necessary to extrarotate the limb to bring the circumflex nerve posteriorly, which normally crosses postero-anteriorly. If the fracture does not involve the small tuberosity resulting in its separation from the humeral head, the tendon of the subscapularis muscle should be incised at about 2 cm from its insertion to preserve vascularization, exposing the joint capsule and the fracture site.

LTA (Fig. 3 A-E) consists of a straight skin incision about 5 cm anterolateral to the deltoid, starting from the lateral margin of the acromion along the humeral diaphyseal axis. The fascia covering the deltoid muscle is incised and the osteotendinous plane is reached through the deltoid muscle. It is important not to extend the incision below 5 cm from the lateral margin of the acromion because of the risk of damaging the circumflex nerve, which leaves the posterior wall of the axilla by crossing the quadrangular space of Velpeau. This nerve surrounds the humerus and penetrates deeply into the deltoid.

Figure 3.

Representative intra-operative images from LTA group. (A) Surface marking of bony landmarks; (B) straight skin incision; (CD) incision of the fascia to access to the osteotendinous plane; (E) exposure of the proximal humerus through the deltoid muscle fibers.

Results

In our series, DPA approach was predominant 52/74 (72%) vs 22/74 (28%). Group A was composed of 31 females and 21 males, the mean age was 57 ± SD 9,77 (range 39-77), while group B was composed of 16 females and 6 males, and the mean age was 60 ± SD 15,33 (range 20-80) (Table 2).

Table 2.

Demographic data related to patients included in the study

Surgical approach	Gender	Side of fracture	Neer 3	Neer 4
Deltopectoral approach	M 21 / F 31	L 31 / R 21	39 (75%)	13 (25%)
Lateral trandeltoid approach	M 6 / F 16	L 9 / R 13	19 (86,4%)	3 (13,6%)

The average radiological follow-up period was 12 months (range 10 - 15, ± SD 1.15).

The radiological outcomes are summarized in Table 3.

Table 3.

Radiological outcomes and duration of surgery for both surgical approaches.

Surgical approach	Preoperative greater tuberosity fracture displacement > 5mm	Postoperative greater tuberosity fracture displacement > 5mm	Preoperative humeral head-shaft angle malalignment (<120° or >145°)	Postoperative humeral head-shaft angle malalignment (<120° or >145°)	Duration of surgery (Mean, ± SD)
Deltopectoral approach	38	14 (37 %)	44	5 (11%)	84,8, ± 33,5
Lateral trandeltoid approach	16	0 (0%)	14	2 (14%)	80,6, ± 29,6
	p < 0,05		> 0,05		p > 0,05

In group A-DPA there were 38 PHF with preop-erative displacement of the greater tuberosity > 5mm. The reduction was obtained in 24 patients (63 %) while in 14 (37 %) this was not achieved (Fig. 4 A-F).

Figure 4.

Clinical case 1: Male, 61 years old. Proximal humeral fractures Neer 4 treated by ORIF using deltopectoral approach (DPA). (A) Preoperative anteroposterior X-Rays (XR); (B-D) 3D CT-scan Anterior/Lateral/Posterior views; (E-F) Postoperative XR AP and LL views. The use of this surgical approach succeeded in the humeral head-shaft angle restoration, but not the in greater tuberosity reduction (> 5 mm of residual displacement).

In group B-LTA there were 16 patients with preoperative greater tuberosity displacement > 5mm. The greater tuberosity anatomy was restored in all of cases (100 %). The humeral head-shaft angle malalignment (< 120° or >145°) was restored in 12 patients out of 14 (86 %). (Fig. 5 A-F).

Figure 5.

Clinical case 2: Female, 67 years old. Proximal humeral fractures Neer 4 treated by ORIF using direct lateral transdeltoid approach (LTA). (A) Preoperative AP view X-Rays (XR); (B-D) 3D CT-scan Anterior/Lateral/Posterior view; (E-F) postoperative AP and LL X-Ray. The restoration of the humeral head-shaft angle was obtained in combination with a good reduction of the humeral greater tuberosity.

The Fisher test was used to investigate the possible relationship between the surgical approach and the quality of the reduction in displaced fractures. The LTA significantly correlated with a better reduction of the greater tuberosity displacement (p < 0.05). No correlation was found between the surgical approach and restoration of the humeral head-shaft angle (p > 0.05). The mean time of surgery was 84,85 ± 33,56 minutes for the DPA (range 40 - 210) and 80,59 ± 29,60 minutes for LTA (range 45 - 170) (p > 0.05).

Discussion

PHF treatment remains a challenge in orthopaedic surgery (20) and there is no consensus upon the best treatment option (21-24). The decision-making process, in addition to the fracture's pattern, is also influenced by factors related to the surgeon and the patient. Most patients with PHF are elderly with many comor-bidities and limited expectations. Operative treatment is then rarely indicate in patients older than 80 years (25). Conservative management continues to be the best option for the majority of patients but approximately in 20% of them surgery is required (25). When the fracture is severely displaced, surgery is recommended but the final outcome is correlated to multiple factors (24,26). Among patients younger than 65 years old, anatomic reduction with ORIF is crucial (26).

In case of surgery, the choice of the best surgical approach is crucial (27). DPA is mostly performed by surgeons and described in literature. However, this approach requires an extensive soft-tissue dissection including a partial release of the deltoid insertion and its over retraction. These actions on the muscle can lead to functional deficits of the muscle. Furthermore, soft tissue release increases the risk of avascular necrosis of the humeral head (13).

LTA allows a direct plating zone of the lateral proximal humerus, requiring less deltoid retraction and soft tissue dissection. It does not affect the blood supply of the humeral head, namely the anterolateral branch of the anterior humeral circumflex artery. This approach, however, has a higher risk of axillary nerve damages and paralysis of the anterior portion of the deltoid muscle (28). The axillary nerve lies anterior to the subscapu-laris, wraps around the surgical neck of the humerus, and passes through the quadrangular space to innervate the teres minor and deltoid muscles. The location of the axillary nerve is around 6.32 cm (range, 5.20-7.6 cm) distal from the anterolateral aspect of the acromion and the course of this nerve must be kept in mind during the surgical approach to avoid neural injuries (29).

Defining which is the better surgical approach is still an open debate (24). In 2013 Buecking et al. published a prospective randomized trial of 120 patients that compared these two approaches using three main parameters (pain, clinical Constant score, complications) and did not detect any significant differences (30). A recent systematic review and meta-analysis by Xie and Zhang suggested that LTA had less humeral head necrosis and shorter surgical time (27). The authors also concluded that both approaches have similar results in functional outcomes, number of complications, and time of hospitalization.

Although in literature many studies about PHF can be found, just few of them take into account defined radiological parameters and none of them unequivocally answers which is the most appropriate surgical approach for each specific type of fracture and each specific patient (27). In our study, X-rays and CT-scans of patient before PHF and immediate X-ray post-surgery were detected, thus comparing the two approaches by two radiological parameters strictly related to the clinical outcome: the humeral greater tuberosity displacement and the head-shaft angle. In fact, the greater tuberosity acts as a fulcrum over which 3 of the 4 muscles of the rotator cuff act on; therefore its good reduction is crucial for a good functional recovery of the shoulder (18,31-33). Posterosuperior displacement of the greater tuberosity of more than 5 mm can result in malunion and impingement of the shoulder due to an altered rotator cuff insertion site influencing the motion in the glenohumeral joint (34).

The physiological humeral head-shaft angle is normally 135° with an interindividual variability of 10-15°; this angle measures the proximal humeral displacement on the coronal plane being measured on true anterior-posterior (AP) radiographs. The goal of a good reduction of any proximal humeral fracture is to restore this anatomic-functional axis (35-37).

No statistically significant differences regarding the restoration of the humeral head-shaft angle were found between the two surgical approaches. This is because the reduction of the fracture at the anatomical or surgical neck level may be obtained through both surgical windows. Concerning the displacement of the humeral greater tuberosity instead, the result could be better through LTA, since this approach allows a direct exposure of this structure, permits to isolate it and to pull it parallelly to its axis obtaining an anatomical reduction and conferring a better tightness of the screws as their force vector is perpendicular to the fracture line.

Furtherly this surgical approach does not significantly affect the length of surgery.

Regarding early postoperative neurological complication, no damages of the circumflex nerve were reported by hospital's records in the LTA group, being the nerve isolated and protected during the surgical procedure.

The main limitations of this study are the small sample size, the uneven distribution of patients through the two groups, a higher percentage of PHF type Neer 4 in the DPA group compared with LTA group, and the lack of clinical evaluation at follow up.

In the future it would be interesting to perform a clinical assessment of these patients to evaluate if the two surgical approaches have an impact on the shoulder function and range of motion.

Conclusion

Direct lateral deltoid splitting approach (LTA) compared to deltopectoral approach (DPA) has significant advantages in the reduction of the grater tuber-osity displacement.

No difference between the two approaches in the restoration of the physiological humeral head-shaft angle was found.

Lastly, the length of the surgery was similar in both groups, being not significantly affected by the surgical approach.