Implant-specific rates of post-operative displacement and reoperation in isolated greater tuberosity fracture open reduction and internal fixation

Abstract

Background

Isolated fractures of the greater tuberosity (GT) comprise nearly 20% of the fractures of the proximal humerus. The Mutch classification categorizes these injuries by morphology into avulsion (Type I), split (Type 2), and depression (Type 3) patterns. While the efficacy of arthroscopic repair vs. open reduction and internal fixation is well-studied, there is less-convincing data on the comparative effectiveness of different open reduction and internal fixation constructs in GT fracture care.

Methods

A retrospective review analyzed patients with isolated GT fractures of the humerus from February 2008 to June 2024 from Kaiser Permanente Southern California. Patients were treated with constructs consisting of (1) plate osteosynthesis alone, (2) suture anchors alone, (3) cannulated or solid lag screws alone, (4) combined plate osteosynthesis and screws, or (5) combined plate osteosynthesis and suture anchors. Primary outcomes were the rate of post-operative displacement and reoperation for each construct. A secondary outcome was reason for reoperation.

Results

The study population included 164 patients with a mean age of 52 years (range 20-88 years). The distribution of injuries was 99 Mutch 1 fractures (60.4%), 44 Mutch 2 fractures (26.8%), and 21 Mutch 3 fractures (12.8). There was no significant difference in Mutch 1 fractures between constructs in post-operative displacement (P = .538) or reoperation rates (P = .915). There was no significant difference in Mutch 2 fractures in post-operative displacement between constructs (P = .368); however, the difference in rate of reoperation was significant (P = .034) on univariate analysis. There was no significant difference in Mutch 3 fractures between constructs in post-operative displacement (P = 1) and none required reoperation. There were no significant differences in post-operative displacement or reoperation on multivariate analysis.

Conclusion

Overall, we did not find a statistically significant difference in rate of post-operative displacement or reoperation between fixation type for Mutch 1, 2, and 3 fractures on multivariate analysis. Surgeons should use the technique they are most comfortable with to achieve an appropriate reduction to restore proper rotator cuff biomechanics.

Proximal humerus fractures represent almost half of the fractures of the humerus and nearly 5% of all fractures.^5,9 As a subset of proximal humerus fractures, isolated fractures of the greater tuberosity (GT) have a reported incidence of 20% within this group.¹³ Isolated GT fractures tend to result from higher energy mechanisms of injury compared to all other proximal humerus fractures, with up to 30% associated with glenohumeral dislocations.¹⁸

Historically, GT fractures are identified as parts in the Neer and AO fracture classification,¹⁵ which only takes into account fragment displacement and angulation. In 2014, Mutch et al described a novel classification based on fracture morphology, size, and orientation. The three described patterns are avulsion (Type I), split (Type 2), and depression (Type 3) (Fig. 1). Based on a series of 199 fractures, they reported superior interobserver and intraobserver reliability compared to the Neer/AO classification.¹⁴ Furthermore, they suggest this morphologic classification may shed light on specific mechanisms of injury, identify concomitant injuries to the glenohumeral joint, and help guide and plan surgical technique and treatment algorithms.¹⁹

Figure 1.

Mutch classification of greater tuberosity fractures. (1) Avulsion type, (2) split type, and (3) depression type.

Optimizing the treatment of GT fractures represents an important endeavor in orthopedic care, as the prevalence of this injury is expected to rise with an increasingly aging population. The economic implications of prolonged morbidity as a result of this injury are significant, as both direct in-hospital costs and indirect out-of-hospital costs would also be expected to rise.^1,6,20 Therefore, it is critical to understand the optimal way to treat these injuries to improve patient outcomes and mitigate the consequences of post-operative morbidity.

Historically, most isolated GT fractures were treated conservatively with immobilization alone. However, Platzer et al¹⁷ found superior outcomes in patients with operatively-treated fractures with greater than 5 mm of displacement compared to their non-operative counterparts. Consequently, open reduction and internal fixation (ORIF) became the mainstay of treatment for displaced fractures. However, open approaches to the shoulder are not benign and carry an increased risk of injury to the axillary nerve, post-operative shoulder stiffness, infection, and cosmetic dissatisfaction. Therefore, many authors have advocated for arthroscopic repair with a double-row suture anchor technique and have demonstrated non-inferior outcomes compared to ORIF in this setting.^2,8,11

While the efficacy of arthroscopic repair vs. ORIF is well-documented, there is a paucity of literature on the effectiveness of different fixation constructs within the ORIF approach to GT fracture care. This study seeks to compare rates of post-operative fracture displacement and all-cause reoperation rates between different fixation techniques of GT fractures within each Mutch classification sub-group.

Materials and methods

Following institutional review board approval, a retrospective review was conducted which included patients with isolated GT fractures of the humerus from February 2008 to June 2024. The study population was collected from the Research Data Warehouse of Kaiser Permanente Southern California from thirteen individual medical facilities. A computer algorithm was utilized to extract all patients from the registry with operatively treated fractures of the GT using International Classification of Diseases (ICD) codes specifying exact encounter diagnoses (ICD-9 821.03; ICD--10 S42.521-5) and Current Procedural Terminology (CPT) codes encompassing ORIF of the GT (CPT 23615, 23630) and Open Treatment of Shoulder Dislocation (CPT 23670). After curation of a list of 367 candidates, meticulous radiographic and chart analysis was conducted to filter out patients who met the exclusion criteria: (1) musculoskeletal polytrauma, (2) fractures involving other portions of the humeral head or neck, (3) no imaging available, (4) no post-operative follow-up, (5) arthroscopic repair, (6) brachial plexus palsy, and (7) periprosthetic fractures. After this process, we arrived at a final study population of 164 patients (Fig. 2).

Figure 2.

Study population flowchart.

Patients were treated with one of the following constructs at the discretion of the surgeon: (1) plate osteosynthesis only, (2) suture anchors only, (3) cannulated or solid lag screws only, (4) combined plate osteosynthesis and screws, or (5) combined plate osteosynthesis and suture anchors. All procedures were performed through either a standard deltopectoral or deltoid split approach. For the “combined plate osteosynthesis and screws” group, all constructs had independent lag screws placed outside of the plate. For the “combined plate osteosynthesis and suture anchors” group, constructs where suture was placed through the rotator cuff and then tied down through bone tunnels or through the plate were included in this cohort in addition to those with independent suture anchors placed into bone.

Demographic information was collected for each patient, including age, gender, race, body mass index, active smoking status, and history of diabetes mellitus through manual chart review. Clinical characteristics were also recorded, including date of injury, delay to surgery, duration of follow-up, presence or absence of injury dislocation event, fixation type, and surgical approach.

All fractures were assessed radiographically according to the Mutch classification. Primary outcome measures for the study were the rate of post-operative fracture displacement and the all-cause reoperation rate for all constructs within each Mutch sub-group. Post-operative fracture displacement was defined in this study as deviation of 5 mm or greater from its immediate post-operative reduction on anterior-posterior Radiographs at any point during post-operative follow-up. To reduce error from interobserver variations, all radiographs were reviewed and scored by the lead author. A secondary outcome measure was reason for reoperation. Indications for reoperation were acquired from operative reports. We hypothesized that plate osteosynthesis alone would display the highest rates of displacement among all three fracture types as this construct does not incorporate fixation through the rotator cuff and may lead to inadequate reduction and stability.

Statistical methods

The mean and standard deviation were determined for all variables in the analysis. Continuous variables were compared using analysis of variance, while categorical variables were compared using a chi-square test or in cases comparing variables including categories with 0 patients, Fisher exact test was used. Multivariable logistic regression was used to model the probability of post-operative displacement and re-operation; P < .05 was considered significant, and all statistical analysis was performed using R 4.5.1 (R Foundation, Vienna, Austria).

Results

Application of the chosen coding parameters yielded 367 patients for review from the Research Data Warehouse. Eighty-four patients were excluded due to polytrauma status, 58 due to no available imaging, 24 due to no post-operative follow-up, 20 due to arthroscopic repair, 8 due to concomitant brachial plexus palsy, 8 due to periprosthetic fracture, and 1 due to recurrent trauma yielding a final study population of 164 patients with isolated GT fractures. Cohort demographics and clinical characteristics are detailed in Table I.

Table 1.

Demographics and clinical characteristics.

Characteristics	N	Mean ± SD (range) or n (%)
Sex
Male	87	53.0%
Female	77	47.0%
Age (yr)	164	52 ± 17 (20-88)
Race
White	86	52.4%
Hispanic	57	34.8%
Asian	12	7.3%
Black	8	4.9%
Native Hawaiian or Other Pacific Islander	1	0.6%
Body mass index classification
Underweight	2	1.2%
Normal weight	26	15.9%
Overweight	56	34.1%
Obese	80	48.8%
Current smoker	24	14.6%
Diabetes	42	25.6%
History of injury dislocation	64	39.0%
Fracture classification
Mutch 1	99	60.4%
Mutch 2	44	26.8%
Mutch 3	21	12.8%
Fixation type
Plate only	22	13.4%
Anchor only	60	36.6%
Screws only	12	7.3%
Anchor and screw	31	18.9%
Anchor and plate	39	23.8%
Follow-up time (d)	164	184 ± 17 (24-801)
Delay to surgery (d)	164	12 ± 14 (1-124)

SD, standard deviation.

N mean ± SD (range) or n (%).

Mutch 1 fractures accounted for 60.4% of injuries in this study. The rate of injury dislocation in Mutch 1 fractures was 17%. Of the five constructs, screw alone fixation displayed the highest rate of post-operative displacement (40%), followed by plate alone (28.6%), anchor and plate (22.7%), anchor alone (17.8%), and anchor and screw (10%) constructs. The rate of reoperation was 12.1% among all constructs. On univariate analysis, there was no significant difference in post-operative displacement (P = .538) or reoperation (P = .915) between constructs. The most common indication for reoperation was stiffness causing persistent functional deficits, followed by fracture displacement with functional deficits, and rotator cuff tear (Table II).

Table II.

Rate of displacement and re-operation by fixation type.

Fixation type	Usage N (%)	Displacement N (%)	P value	Reoperation N (%)	P value
Mutch 1
Plate alone	7 (7.1%)	2 (28.6%)		1 (14.3%)
Anchor alone	45 (45.5%)	8 (17.8%)		4 (8.9%)
Screw alone	5 (5.1%)	2 (40.0%)		1 (20.0%)
Anchor and screw	20 (20.2%)	2 (10.0%)		3 (15.0%)
Anchor and plate	22 (22.2%)	5 (22.7%)		3 (13.6%)
Total	99 (100%)	19 (19.2%)	.538	12 (12.1%)	.915
Mutch 2
Plate alone	14 (31.8%)	2 (14.3%)		0 (0.0%)
Anchor alone	5 (11.4%)	1 (20.0%)		1 (20.0%)
Screw alone	5 (11.4%)	0 (0.0%)		2 (40.0%)
Anchor and screw	8 (18.2%)	0 (0.0%)		0 (0.0%)
Anchor and plate	12 (27.3%)	0 (0.0%)		1 (8.3%)
Total	44 (100%)	3 (6.8%)	.368	4 (9.1%)	.032
Mutch 3
Plate alone	1 (4.8%)	0 (0.0%)		0 (0.0%)
Anchor alone	10 (47.6%)	2 (20.0%)		0 (0.0%)
Screw alone	2 (9.5%)	0 (0.0%)		0 (0.0%)
Anchor and screw	3 (14.3%)	0 (0.0%)		0 (0.0%)
Anchor and plate	5 (23.8%)	1 (20.0%)		0 (0.0%)
Total	21 (100%)	3 (14.3%)	1	0 (0.0%)

Bold value indicates statistically significant P < .05.

Mutch 2 fractures accounted for 26.8% of injuries in this study. The rate of injury dislocation in Mutch 2 fractures was 43%. Of the five constructs, anchor alone fixation displayed the highest rate of post-operative displacement (20.0%), followed by plate alone fixation (14.3%). The screw alone, anchor and screw, and anchor and plate groups did not have any patients experience post-operative displacement. The rate of reoperation was 9.8% among all constructs. On univariate analysis, there was no significant difference in post-operative displacement between constructs (P = .368). However, there was found to be a significant difference in the rate of re-operation (P = .034) between constructs. The most common indication for reoperation was stiffness causing persistent functional deficits, followed by hardware irritation, and fracture displacement with functional deficits (Table II).

Mutch 3 fractures accounted for 12.8% of injuries in this study. The rate of injury dislocation in Mutch 2 fractures was 48%. Of the five constructs, anchor alone and anchor and plate fixation displayed the highest rate of post-operative displacement (20.0%). The screw alone, anchor and screw, and plate alone groups did not have any patients experience post-operative displacement. No patients sustaining a Mutch 3 fracture required reoperation in this study. On univariate analysis, there was no significant difference in post-operative displacement between constructs (P = 1). Additionally, there was no significant difference in the rate of reoperation between Mutch 1, 2, and 3 groups regardless of fixation type (P = .262) (Table II).

A Multivariate analysis was performed, which demonstrated no significant differences in post-operative displacement or reoperation when modeled by patient demographics, construct type, or fracture pattern (Table III).

Table III.

Multivariate model for post-operative displacement and reoperation.

Characteristics	Post-operative displacement		Reoperation
	Odds ratio	P value	Odds ratio	P value
Demographics
Age	1.01	.735	1.03	.11
Female	1	Reference	1	Reference
Male	1.75	.401	1.53	.427
BMI	1.01	.854
Social history
Non-smoker	1	Reference	1	Reference
Active smoker	4.07	.203	1.76	.484
Medical history
No diabetes	1	Reference
History of diabetes	2.12	.309
Fixation construct
Plate alone	1	Reference	1	Reference
Anchor alone	1.69	.66	0.7	.63
Screw alone	8.07	.111	0.891	.912
Anchor and screw	1.94	.6	0.255	.169
Anchor and plate	1.94	.582	0.676	.617
Fracture type
Mutch 1	1	Reference	1	Reference
Mutch 2	0.63	.501	0.283	.08
Mutch 3	0	.99	0.795	.746

BMI, body mass index.

Discussion

This study sought to identify differences in post-operative fracture displacement and rate of reoperation between common fixation methods in ORIF of isolated GT fractures. To our knowledge, this is the largest study examining these outcomes for isolated GT fractures covering a wide array of fixation strategies. Importantly, this study examines fixation strategies according to fracture morphology in the Mutch classification, which offers a more reproducible way to characterize fractures and may guide treatment decisions.

While isolated GT fractures are, indeed, an osseous deformity, we believe the associated injury to the rotator cuff is just as important. As such, soft tissue handling and fixation that restores proper rotator cuff biomechanics are paramount. For this reason, the plate alone construct was chosen as the reference class in our multivariate analysis model. This construct utilizes no fixation through the rotator cuff which may compromise shoulder biomechanics and lead to sub-optimal reduction. We hypothesized that this construct would perform worse compared to others that achieve fixation through the rotator cuff.

Mutch 1 fractures highlight the importance of understanding fracture biomechanics, as they most closely resemble true avulsion fractures of the rotator cuff footprint. In our study, the only two constructs without fixation through the soft tissue (plate alone and screws alone) demonstrated the highest level of post-operative displacement and the highest and third-highest rates of reoperation for Mutch 1 fractures. However, these results were not statistically significant. Since plate alone and screw alone fixation were not the predominant form of fixation selected for this fracture pattern, these groups may be underpowered to detect true statistical significance. These observations were somewhat mitigated in the Mutch 2 fracture pattern where plate alone fixation demonstrated lower levels of displacement than in Mutch 1 fractures, possibly due to a buttress effect of the plate on the larger bony shear fragment. Mutch 3 fractures accounted for less than 13% of injuries in this study, and only three patients experienced post-operative displacement without need for reoperation.

Several studies have compared outcomes in fixation strategies for ORIF of isolated GT fractures. Biomechanical studies suggest that screw fixation is inferior to both suture anchor and locking plate fixation, which is corroborated by the findings in our study of higher reoperation rate in both Mutch 1 and Mutch 2 fractures.^3,10,12,16 Yoon et al report on a series of patients with isolated GT fractures treated with screw and washer fixation. They found satisfactory union rates; however, many patients developed functional deficits due to post-operative stiffness.²² We report similar findings in our study, with 25% of patients in the screw fixation group requiring reoperation for lysis of adhesions due to stiffness.

The efficacy of locking plate fixation has also been evaluated in the setting of isolated GT fractures. Several studies report satisfactory radiographic and functional outcomes after application of locking plates through either a deltopectoral or minimally invasive deltoid split approach for unclassified GT fractures.^21,23 Specifically for split-type fractures, Hu et al⁷ report a low complication profile, rates of secondary displacement, and implant failure with locking plate fixation. Additionally, a biomechanical study found superior strength and stiffness of repair for locking plate fixation compared to suture anchor fixation for split type fractures.⁴ This finding is supported by our study, which found lower rates of displacement and reoperation for plate alone fixation compared to anchor alone constructs for split-type fractures, although not statistically significant.

Strengths of this study are that it compares all five of the most common fixation strategies for isolated GT fractures across each morphologic fracture type. Additionally, to our knowledge, this is the largest study comparing post-operative displacement and reoperation outcomes between fixation types. Additionally, data were collected from 13 hospitals across Southern California, encompassing diverse populations to increase the generalizability of our findings.

Limitations of this study are that it is a retrospective review, which is inherently subject to selection bias. Furthermore, since surgeries were performed by multiple surgeons, we were unable to control for differences in surgical experience, quality of reduction, rotator cuff management, or non-standardization of post-operative rehab protocols. Retrospective review of radiographs is subject to interpretation bias and misclassification. To mitigate this, all radiographs were reviewed and scored by the same author to negate interobserver variations. Follow-up duration was inconsistent, leading to comparisons between groups at different timepoints in recovery. Finally, indications for reoperation are subjective and at the discretion of the treating surgeon, potentially leading to varying levels of reoperation between institutions. Indications for reoperation in this study included stiffness causing functional limitations and fracture displacement leading to sub-acromial impingement or loss of motion.

Since GT fractures represent essentially rotator cuff-equivalent injuries, future studies should be planned to evaluate the relationship between post-operative displacement, reoperation, and functional outcomes with respect to the quantity fixation through the rotator cuff.

Conclusion

Overall, we did not identify a statistically significant difference in the rate of post-operative displacement or reoperation between fixation types for Mutch 1, 2, and 3 fractures on multivariate analysis. Surgeons should be aware of potentially higher rates of displacement and reoperation with plate alone and screw alone fixation in Mutch 1 fractures and anchor alone fixation in Mutch 2 fractures. However, they should use the technique they are most comfortable with to restore bony anatomy and proper rotator cuff biomechanics.

Disclaimers:

Funding: No funding was disclosed by the authors.

Conflicts of interest: The author, their immediate family, and any research foundation with which they are affiliated have not received any financial payments or other benefits from any commercial entity related to the subject of this article.