Fixing Cho Type IIC Distal Clavicle Fractures with Hook Plates Leads to a High Incidence of Subacromial Osteolysis: A Retrospective Study and Literature Review

Abstract

Background

This retrospective study investigates the complications, particularly subacromial osteolysis (SAO), associated with hook plate (HP) fixation, in the treatment of unstable distal clavicle fractures characterized by complete coracoclavicular (CC) ligament rupture. The decision-making process for employing HP in fractures of this nature, such as Neer types IIB and V and Cho classification IIC, involves considerations of distal fragment size and displacement. While HP offers advantages in clinical practice, it is not without complications, with SAO being a notable concern. Factors such as non-anatomic hook tip placement and fracture classification may influence the risk of SAO.

Methods

The study comprises a retrospective analysis of unstable distal clavicle fractures treated with HP at our institution from 2019 to 2022. Exclusions include non-displaced fractures, those treated with other locking plates, and pathologic fractures. A total of 91 patients with displaced distal clavicle fractures underwent open reduction and internal fixation with HP. Cho classification was employed to differentiate cases with CC ligament rupture. Patient demographics, classifications, postoperative radiographs, distal fragment size, plate position, timing of implant removal, and complications, including SAO, were recorded.

Results

Among the 91 patients, 32 were classified as Cho IIB, 43 as Cho IIC, and 16 as Cho IID. Ninety-one percent exhibited solid union before implant removal. The prevalence of SAO was 43.8%, 76.7%, and 62.5% in Cho IIB, IIC, and IID, respectively. Univariate analysis revealed a significant difference only in Cho classification (p = 0.014). Binary logistic regression identified Cho classification type IIC as the sole risk factor for SAO (p = 0.021; odds ratio, 4.48; 95% confidence interval, 1.56–12.87).

Conclusions

Cho type IIC fractures, characterized by CC ligament deficiency causing horizontal instability, demonstrated the highest SAO rate. In contrast, Neer type IIB fractures retained the trapezoid ligament, and Neer type V fractures had intact CC ligaments, resulting in lower SAO rates. Biomechanically, combining HPs with CC ligament reconstruction provided better structural stability than using HPs alone in treating Cho type IIC fractures.

Clinical decisions regarding the treatment of distal clavicle fractures are often based on the size of the distal fragment and whether the fracture is displaced. The decision of whether a patient needs surgical treatment and the type of implant to be used depend on these factors. According to the Neer classification, type I fractures are nondisplaced and can be treated non-surgically. Type II and type V fractures, which are displaced, have a nonunion rate of 30%–40% if not treated surgically, making surgical treatment advisable.¹,²,³⁾ For fractures with large distal fragment size, such as Neer type IIA fractures, a pre-contoured locking compression plate is most commonly used.⁴⁾ For fractures with smaller distal fragment size, such as some Neer type IIB and type V fractures, where there is insufficient bone mass for screw purchase, a hook plate (HP) can provide stability and favorable union rates and therefore is currently the most popular method.¹,²,⁵⁾ Other methods for these kinds of fractures including Kirschner wires,⁶⁾ Knowles pins,⁷,⁸⁾ tension band fixation,⁹,¹⁰⁾ and CC fixation with Mersilene tapes⁵,¹¹⁾ or screws¹²,¹³⁾ have also been described.

However, the literature suggests that using an HP for treatment can lead to a complication rate of 44%,²⁾ with subacromial osteolysis (SAO) being the most common complication.⁵,¹⁴⁾ The severity of SAO has different clinical implications. Mild cases may not cause noticeable symptoms and can be managed by removing the HP after fracture site union. Severe cases can result in postoperative shoulder pain, acromion perforation, or acromion fracture, leading to chronic shoulder pain and poor long-term functional outcomes, even after HP removal.

Currently, there is literature indicating that the placement of the hook tip closer to the posterior margin of the acromion,¹⁵,¹⁶⁾ the use of a hook depth design that leads to overreduction,¹⁵,¹⁷⁾ or a larger hook angle¹⁸⁾ can increase the risk of postoperative SAO in patients with unstable distal clavicle fractures. This study aims to investigate whether different fracture classifications treated with HPs result in distinct complication rates, particularly focusing on the incidence of SAO. Additionally, potential risk factors contributing to an increased rate of complications following HP application will be explored.

In 2018, Cho et al.⁴⁾ proposed treatment recommendations based on different fracture classifications, introducing a novel classification (Cho type IIC) that significantly departs from the established Neer classifications. While Cho et al.⁴⁾ recommended specific treatment modalities for each type, a definitive rationale for the treatment of Cho type IIC fractures remains unclear. For the new Cho classification type IIC, which differs from Neer type IIB as it involves complete rupture of both trapezius and conoid ligaments within the coracoclavicular (CC) ligament, the use of an HP is not recommended. Instead, trans-acromial intramedullary fixation and CC fixation are suggested. Cho et al.⁴⁾ did not provide a definitive reason for this treatment recommendation.

Despite existing recommendations, there is a noticeable gap in the literature regarding the comparative outcomes of HP treatment across different fracture classifications, including Neer types IIB and V and the newly introduced Cho type IIC. Our study seeks to fill this gap by examining whether these classifications exhibit distinct complication rates and by identifying additional risk factors that may influence postoperative outcomes.

METHODS

This study was approved by the Research Ethics Committee (REC) of Kaohsiung Veterans General Hospital (REC No. 210630-1). This study was exempted from obtaining informed consent from patients.

Patients

We retrospectively enrolled patients from January 2015 to December 2022 with displaced unstable distal clavicle fractures (Neer classification type IIB and type V or Cho’s classification types IIB, IIC, and IID) who underwent open reduction and internal fixation by using a HP manufactured by a single company (DePuy Synthes) in our hospital. Patients with any concurrent shoulder injury on the affected side, a history of shoulder trauma or surgery, inadequate radiographic follow-up, or age under 18 years were excluded from enrollment. We collected patients’ data until they were re-admitted to our hospital for HP removal surgery. Patient demographics, injury classification (Cho’s classification⁴⁾ and Neer classification⁶⁾) distal fragment size, HP position parameters, HP removal timing, and complications were recorded. All patients were classified into a group with SAO and a group without SAO.

Parameter Measurement

Plain radiographs included an anteroposterior and oblique view of the bilateral shoulders of each patient and were taken and reviewed. The distal clavicle fragment size, the distance between the upper border of the coracoid process and the lower cortex of the clavicle, known as the coracoclavicular distance (CCD),¹⁹⁾ the depth of hook body, the depth of hook tip, and the hook coverage ratio were measured on anteroposterior views as described by Xu et al.²⁰⁾ and Tamiya et al.¹⁶⁾ The CCD was expressed as a percentage compared to the contralateral-side shoulder. The anteroposterior position of the hook tip related to acromion was measured on an oblique view as described by Tamiya et al.¹⁶⁾ (Fig. 1).

Fig. 1. (A) The distal clavicle fragment size (a) and the preoperative coracoclavicular distance (CCD, b) were measured on the shoulder anteroposterior (AP) view. The contralateral side CCD was also measured with the same method on the contralateral side shoulder AP view. (B) The anteroposterior positioning of the hook tip in relation to the acromion was illustrated by the ratio of d to c. A larger ratio indicates a more posterior placement of the hook plate. (C) The hook coverage ratio, denoted as f to e, illustrates the extent of lateral placement of the hook plate relative to the acromion. A higher ratio signifies a more outward positioning of the hook plate (HP). The postoperative CCD (g) was measured. (D) The image obtained after HP removal showed presence of subacromial osteolysis, which is shown in the circle.

Statistical Analysis

The means and standard deviations of continuous variables were calculated. To evaluate and compare continuous variables between the group with SAO and the group without SAO, the paired Student t-test was performed. Pearson’s chi-square test was used to assess categorical variables, which were expressed as numbers and percentages. Risk factors for SAO were identified by binary logistic regression and expressed as odds ratio (OR) and 95% confidence interval (CI). A p < 0.05 was deemed statistically significant. All statistical analyses were done using SPSS Advanced Statistics 27.0 (IBM Corp.).

RESULTS

Patient Characteristics

A total of 91 patients were included in this study. The overall union rate was 91.2%, complication rate was 24.2%, and incidence of SAO was 62.6%. Two typical cases are illustrated in Figs. 2 and 3. There was a significant difference in complication rate among the Neer types IIB and V and Cho’s type IIC (p < 0.001). The details of each number of complications for each classification are described in Table 1. Fifty-seven patients had SAO and 44 patients did not. Patient demographics of each group are shown in Table 2. Thirty-one men and 26 women were included in the group with SAO. Sixteen men and 18 women were included in the group without SAO. There was no significant difference in the sex ratio between these 2 groups (p = 0.499). The age of patients with SAO was 45.75 ± 16.03 years and for those without SAO, it was 45.06 ± 15.96 years (p = 0.845). The body mass index in the SAO group was 23.85 ± 3.98 kg/m², and in the non-SAO group, it was 22.91 ± 2.95 kg/m² (p = 0.202). The timing of implant removal was 6.19 ± 2.33 months in the group with SAO and 5.70 ± 2.00 months in the group without SAO, and there was no significant difference between these 2 groups (p = 0.316).

Fig. 2. Patient 1: a 66-year-old man presented with a Cho type IIC distal clavicle fracture. He was treated using a hook plate. The postoperative plain film revealed satisfactory reduction; however, during follow-up, there was progression of subacromial osteolysis, culminating in eventual acromion cut-through.

Fig. 3. Patient 2: A 44-year-old man presented with a Cho type IIC distal clavicle fracture. He underwent treatment with a hook plate. The postoperative plain film indicated successful reduction; however, progressive subacromial osteolysis with upper migration of the hook into the acromion was identified. Acromion perforation accompanied by peri-implant fracture was noted at postoperative 5 months of follow-up. Eventually, the plate had to be removed due to the development of severe pain.

Table 1. Complications in Hook Plate Fixation of Distal Clavicle Fractures.

	Variable	Cho type IIB (n=32)	Cho type IIC (n=43)	Cho type IID (n=16)	p-value
Total complications		25	60	15	< 0.001
	Malunion	1	3	1
	Nonunion	3	4	0
	Loss reduction	0	4	0
	Impingement	1	4	0
	Rotator cuff tear	0	0	0
	Acromion perforation	2	2	1
	Distal clavicle osteolysis	4	7	3
	Clavicle midshaft peri-implant fracture	0	2	0
	Subacromial osteolysis	14	33	10
	Hook escape	0	1	0

Table 2. Demographic Characteristics of Patients.

Risk factor			SAO group (n=57)	Non-SAO group (n=34)	p-value
Sex (male : female)			31 : 26	16 : 18	0.499
Age (yr)			45.75 ± 16.03	45.06 ± 15.96	-
BMI (kg/m2)			23.85 ± 3.98	22.91 ± 2.95	0.202
Fracture factor
	Cho classification				0.014
		Type IIB	14	18
		Type IIC	33	10
		Type IID	10	6
	Distal fragment size (cm)		1.58 ± 0.84	1.62 ± 0.54	0.799
		< 1	15	5
		1–1.5	18	10
		1.5–2	8	10
		> 2	16	9
	Preoperative CCD (%)		216.23 ± 99.60	208.15 ± 77.85	0.687
	Postoperative CCD (%)		81.06 ± 34.35	74.87 ± 48.22	0.478
Reduction quality					0.523
	Under-reduction (> 120%)		10	3
	Neutral (80%–100%)		13	8
	Over-reduction (< 80%)		34	23
Plate removal timing (mo)			6.19 ± 2.33	5.70 ± 2.00	0.316
Anatomical factor
	Acromial shape				0.490
		Flat	22	10
		Curved	6	4
		Hooked	7	2
		Reversed curved	22	18
Hook plate position factor
	DBA (cm)		0.65 ± 0.33	0.59 ± 0.30	0.402
	DHT (cm)		1.71 ± 0.37	1.72 ± 0.33	0.948
	Hook coverage (%)		57.77 ± 16.83	60.65 ± 17.60	0.439
	Anterior-posterior position (%)		63.73 ± 13.74	66.78 ± 15.58	0.333

Values are presented as mean ± standard deviation.

SAO: subacromial osteolysis, BMI: body mass index, CCD: coracoclavicular distance, DBA: the distance between the hook body and the acromion inferior cortex, DHT: the depth of hook tip, which was the shortest distance from the hook tip to the acromial articular surface.²⁰⁾

Potential Influential Fracture Factors, Anatomical and HP Position Factors

In terms of Cho’s classification, the group with SAO included 14 IIB, 33 IIC, and 10 IID patients. The group without SAO included 18 IIB, 10 IIC, and 6 IID patients (Table 2). There was a significant difference between these 2 groups (p = 0.014). In terms of distal fragment size (SAO vs. no SAO: 1.58 ± 0.84 cm vs. 1.62 ± 0.54 cm, p = 0.799), both preoperative CCD and postoperative CCD (SAO vs. no SAO: preoperative CCD 216.23% ± 99.60% vs. 208.15% ± 77.85%, p = 0.687; postoperative CCD 81.06% ± 34.35% vs. 74.87% ± 48.22%, p = 0.478), the timing of HP removal (SAO vs. no SAO: 6.19 ± 2.33 months vs. 5.70 ± 2.00 months; p = 0.316), and reduction quality (p = 0.523), no significant difference was presented between these 2 groups.

The ratio of each acromion in each type of fracture had no significant difference in the occurrence of SAO. All the depth of the hook body (SAO vs. no SAO: 0.65 ± 0.33 cm vs. 0.59 ± 0.30 cm, p = 0.402), the depth of the hook tip (SAO vs. no SAO: 1.71 ± 0.37 cm vs. 1.72 ± 0.33 cm, p = 0.948), the percentage of hook coverage (SAO vs. no SAO: 57.77% ± 16.83% vs. 60.65% ± 17.60%, p = 0.439), and the anteroposterior position of hook tip (SAO vs. no SAO: 63.73% ± 13.74% vs. 66.78% ± 15.58%, p = 0.333) were not significantly different between these 2 groups.

Risk Factor for SAO

We enrolled Cho classification based on significant differences found in univariate analysis, and reduction quality and anteroposterior position of hook tip based on previous studies using binary logistics regression analysis.¹⁵,¹⁶,¹⁷⁾ Patients with Cho type IIC fractures had a significantly higher risk of SAO compared to those with Neer type IIB (OR, 4.48; 95% CI, 1.56–12.87). Neer type V had no significant difference in OR for causing SAO (OR, 2.26; 95% CI, 0.64–7.92) compared to Neer type IIB. P values of both reduction quality and anteroposterior position of hook tip were not significantly different for causing SAO (Table 3).

Table 3. Binary Logistic Regression Model for Predictors of Subacromial Osteolysis.

Variable		OR	95% CI	p-value
Cho classification				0.021
	Type IIB	Reference	Reference
	Type IIC	4.48	1.56–12.87
	Type IID	2.26	0.64–7.92
Reduction quality				0.635
	Neutral reduction (90%–100%)	Reference	Reference
	Under-reduction (> 110%)	1.16	0.218–6.132
	Over-reduction (< 90%)	0.67	0.218–2.033
	Anteroposterior position of thehook tip related to acromion (%)	0.98	0.953–1.017	0.338

OR: odds ratio, CI: confidence interval.

DISCUSSION

Neer classification is currently the most commonly used classification for distal clavicle fractures. The fracture stability is based on whether the fracture displacement exceeds 100%, the position of the fracture relative to the CC ligament, and whether the CC ligament is injured in this classification.⁶⁾ Fracture with an injured CC ligament is considered unstable and therefore needs to be treated surgically due to higher than 30% nonunion rates if left untreated nonoperatively.¹⁴⁾ In a clinical context, most unstable distal clavicle fractures are typically categorized as Neer types IIB and V. However, there are instances where the fracture pattern is so unstable that surgeons encounter difficulty classifying it specifically as Neer type IIB or V. Moreover, the low reliability in both interobserver and intraobserver assessments of this classification system often leaves surgeons unable to definitively classify the fracture type. This lack of clarity can lead to confusion in determining the most suitable treatment. In 2018, Cho et al.⁴⁾ introduced a new classification with high intraobserver and interobserver reliability including a new type of fracture for unstable distal clavicle fractures, type IIC. In this type of fracture, both the conoid ligament and trapezoid ligament in the CC ligament are completely ruptured, without any residual trapezoid ligament, which is different from Neer IIB.⁴⁾ The purpose of this classification is to differentiate between a complete CC ligament tear and a partial CC ligament tear, as complete tears can cause complete loss of stability in the proximal clavicle stump in both horizontal and vertical directions. In our opinion, in treating this type of fracture, surgeons need to consider therapies that enhance stability more than dealing with traditional Neer type IIB or type V.

For unstable clavicle fractures, diverse surgical methods, such as CC stabilization, HP fixation, intramedullary fixation, interfragmentary fixation, and K-wire plus tension band wire, have shown promising outcomes, with union rates exceeding 90% and no significant differences among the methods.²,⁵,²¹⁾ The HP is widely favored due to its ability to provide the most rigid vertical stability by hooking onto the acromion, compensating for the compromised inferior pulling strength of the CC ligament. Nevertheless, the treatment algorithm outlined by Cho et al.⁴⁾ stipulates that for Cho type IIC fractures, the recommended modalities are CC stabilization or intramedullary fixation, as opposed to the utilization of a HP—a departure from the treatment advocated for Neer type IIB and V fractures. In our opinion, Cho’s recommendation is likely based on the extreme instability of fractures involving a complete tear of the CC ligament. Using a HP alone may carry a high risk of complications. Unfortunately, current clinical research has not provided us with information on the clinical outcomes of using a HP for this specific type of fracture. High complication rates exceeding 40% after utilizing HPs were reported in treating unstable distal clavicle fractures based on Neer classification in several studies.²,⁵,²¹⁾

Compared to CC stabilization, hook plating has approximately 4 times higher complication rates.²¹⁾ Common complications include SAO, acromion fracture, distal clavicle osteolysis peri-implant fracture, hook escape, rotator cuff tear, and subacromial impingement. Among them, SAO has an incidence of 50% to 100%.¹⁵,²²,²³⁾ In our study, we found that using HPs for treating Cho type IIC fractures had significantly higher complication rates, especially acromion-related complications, as compared to that for treating Neer types IIB and V (Figs. 2 and 3). Furthermore, our study identified Cho classification type IIC as the sole risk factor for SAO (up to 4.48 times) using binary logistic regression. It is worth noting that 39.4% exhibited either an erosion depth exceeding 5 mm or concomitant acromion perforation or an acromion fracture in patients with Cho type IIC fractures who developed SAO.

For distal clavicle fractures treated with HP fixation, Shimpuku et al.¹⁵⁾ and Tamiya¹⁶⁾ found that posterior hook tip placement was the risk factor correlated with the acromion erosion depth. Besides, posterior hook tip placement should be avoided to prevent hook escape. Over-reduction is also found to be the risk factor. Shimpuku et al.¹⁵⁾ revealed that 95% of patients with SAO were over-reduced. Another biomechanical study¹⁷⁾ found that shorter hook depth can cause over-reduction in an excessive pressure exerted on the acromion by the hook tip. Our results showed that both posterior placement of hook tip and shorter hook depth did not significantly increase ORs for causing SAO. The only risk factor for SAO in this study was the distal clavicle fracture with a completely ruptured CC ligament (Cho type IIC). The main function of the CC ligament is to provide vertical stability of the shoulder and the acromioclavicular (AC) joint, which plays a big role in providing horizontal stability of the clavicle. Once the CC ligament is totally ruptured, not only the vertical stability but also the horizontal stability of the clavicle will be affected. In this kind of injury, a HP can provide vertical stability but not horizontal stability. This makes the hook section of the plate produce excessive horizontal micromotion over the bottom of the acromion, eventually leading to SAO.

Rockwood type V AC joint dislocation shared common injury pattern characteristics with Cho type IIC distal clavicle fracture in terms of total CC ligament rupture and similar deforming force. Previous literature on Rockwood type V AC joint dislocation manifested that adding CC ligament reconstruction in treatment can reduce the incidence of SAO by 70%.²⁴,²⁵⁾ In our opinion, restoring the CC ligament is important to prevent proceeding SAO in Cho type IIC. In Neer type IIB or V fractures, the CC ligament is retained; therefore, rigid fixation without CC ligament reconstruction may provide satisfactory outcomes with fewer complications.

There are some limitations in our study. The HPs in our study were all manufactured by a single company. Therefore, it was impossible for us to evaluate whether the difference in hook angle would influence the incidence of SAO and the rate of other complications.¹⁸⁾ The radiographic set is affected by individual position, which may be altered by radiographic shoulder position despite efforts to standardize the measurement of hook position-related values. Furthermore, the functional outcomes of the shoulder after implant removal need to be assessed in the future.

Compared to other types of unstable distal clavicle fractures, distal clavicle fractures with total CC ligament ruptures like Cho type IIC have as high as 66.2% higher SAO after fixation with HPs alone. Adding CC ligament reconstruction to restore not only the vertical but also horizontal stability is recommended to decrease the overall complications rate, especially in terms of SAO.