Clinical and radiographic outcome of open reduction and internal fixation of 3- and 4-part proximal humerus fractures augmented with impaction bone grafting and a magnesium-based bone filler

Abstract

Background

Open reduction and internal fixation (ORIF) of proximal humerus fractures (PHFs) is a challenging operation with high rates of loss of reduction, screw cut out, avascular necrosis, and subsequent unplanned reoperation. Augmentation of the repair with synthetic bone fillers and other alternatives has shown promise in decreasing adverse outcomes. Our aim is to report radiographic and clinical outcome of impaction grafting with cancellous allograft chips and injection of magnesium-based bone filler for ORIF augmentation of PHFs.

Methods

All patients that underwent ORIF for a 3- or 4-part PHF with the standardized protocol with a minimum of 6 months radiographic follow-up by a single surgeon (VE) were included. Radiographs were taken at standardized time points up to 6 months, followed by a final radiographic follow-up, to define radiographic healing or failure. Patient-reported outcome measures were collected at final follow-up. Patient characteristics, complications, reoperations and radiographic measures of reduction quality were also recorded.

Results

17 patients were identified with a mean 34.1 months of radiographic follow-up. Median Penn Shoulder Score was 89 (interquartile range 19), American Shoulder and Elbow Surgeons score was 92 (25), Veterans RAND 12-Item Health Survey Mental Component Score was 53.9 (11.9), Veterans RAND 12-Item Health Survey Physical Component Score was 51.6 (12.8), and Single Assessment Numerical Assessment was 85 (18.5). 14 patients (82.3%) had routine radiographic healing. Two patients (11.8%) developed avascular necrosis with screw cutout. Two patients (11.8%) had reoperation including one hardware removal and one conversion to reverse total shoulder arthroplasty for a subsequent rotator cuff tear. No signs of glenohumeral arthritis were present in 12/17 patients (70.6%), and no signs of cuff tear arthropathy were noted in 13/17 patients (76%). Seven out of nine (78%) patients who worked prior to injury returned to work at a mean of 14.7 weeks postoperatively.

Conclusion

Augmentation of 3- and 4-part proximal humeral fractures with a standardized protocol utilizing cancellous chips and a synthetic magnesium-based bone filler results in a high rate of maintenance of fracture reduction, radiographic healing, and satisfactory patient outcomes. Further comparative data is needed to evaluate the efficacy of this technique compared to alternative methods of augmentation.

Proximal humerus fractures (PHFs) account for 4%-5% of all fractures, with an increasing incidence over the past two decades.² Although the majority of PHFs are treated nonoperatively, open reduction and internal fixation (ORIF) has been shown to improve outcomes when compared to nonoperative management in certain patient populations and fracture patterns.^19,28,32,36 Although there is no consensus for surgical indications in the literature, ORIF is commonly chosen in younger patients with displaced PHF involving tuberosities or patients in whom nonoperative treatment had less than ideal outcome such as humeral head split with large intraarticular step-off .^{9,13,22,37,43} Complications after ORIF of PHF have been reported to occur in 25%-40% of patients, including stiffness, nonunion, malunion, loss of fixation, intra-articular screw penetration, screw cut out, avascular necrosis (AVN), subacromial impingement of the plate, and post-traumatic arthritis.^15,34,40 The advent of locked proximal humeral plates and improved surgical techniques have reduced the frequency of major complications in the past two decades, but the rate of short and long terms complications remains high despite these efforts.

Lack of stability in 3- and 4-part PHFs as well as poor bone quality due to advanced age and osteoporosis makes ORIF of these fractures challenging. In addition to locking plate technology, various forms of biologic and synthetic augmentations including cancellous chips, fibular strut allografts, and various bone fillers have been proposed in the last two decades to mitigate the high risk of failure in these fractures.^{4,12,14,26,27,44} Polymethylmethracrylate (PMMA) has been shown biomechanically to augment fixation strength but has not clinically been shown to be superior.¹¹ Fibula struts are a popular option; however, they can lead to significant difficulty if revision to arthroplasty is needed.³⁹ The graft may be incorporated into host bone up to 80% of the time and they are challenging to remove, requiring excavation of the graft and causing significant humeral bone loss.²³ Multiple case series have shown that calcium phosphate and calcium sulfate can be used to fill a metaphyseal void with acceptable clinical and radiographic outcomes.^16,33 However, both calcium phosphate and calcium sulfate cement have the potential to create inflammatory host reactions or complications related to accelerated resorption of the cement over time.²⁴

The goal of PHF augmentation is to fill bone voids created by fracture comminution to increase time zero fixation and prevent early fracture collapse without interfering with bone healing and potential future surgeries. The authors have developed a protocol to augment displaced 3- and 4-part PHFs with impaction grafting with cancellous allograft and injection of a magnesium-based bone filler (Osteocrete; Bone Solutions, Colleyville, TX, USA). Magnesium is an inert material and there is less chance of inflammatory reaction by the host.²⁰ Also, the reported residence time for magnesium-based bone filler is more than 18 months.⁶ The present study seeks to evaluate clinical and radiographic outcomes after utilizing this protocol, as well as a description of the technique developed to maximize the benefit of this augmentation strategy.

Materials and methods

Institutional review board approval was obtained by the Cleveland Clinic Foundation Institutional Review Board (Protocol 22-261). All patients who underwent ORIF of 3- and 4-part PHF from 2019 to 2021 by the senior author (VE) were included. Patients were excluded if their fracture extended to diaphyseal region in the humeral shaft, had prior open surgery in the same shoulder or had less than 6 months radiographic follow-up. Patients underwent X-rays at routine follow-up durations. Patients were excluded if they did not have at least 6 months of radiographic follow-up. All eligible patients were contacted at a minimum of 24 months postoperatively to obtain final radiographs and patient-reported outcome measures (PROMs). The primary outcome measure was radiographic evidence of fracture healing. The secondary and tertiary outcomes were PROMs and complications included AVN with or without humeral head collapse, screw cut out, loss of reduction, nonunion, development of posttraumatic arthritis, and reoperation.

Surgical technique

All cases were performed in the beach chair position through a deltopectoral approach. The fracture fragments were initially identified and the tuberosities were tagged, and the fracture was anatomically reduced and provisionally pinned using three to four 2.0 mm Kirschner wires. An example of a complex fracture dislocation is shown in Figure 1, A and B. Tuberosity repair was performed with multiple horizontal sutures placed at the rotator cuff tendon-bone junction, and they were passed through the peripheral holes in the proximal humeral locking plate. Separate reduction sutures were used to control the position of the tuberosities. Intraoperative fluoroscopy was used in two orthogonal planes to confirm anatomic reduction. Through a lateral cortical window commonly created by the fracture lines, 15-30 cc of cancellous bone chips were impacted into the central portion of the humeral head until the metaphyseal region was filled by allograft (Fig. 1, C). A precontoured proximal humerus plate (PHILOS; DePuy Synthes, Raynham, MA, USA) with locking screws was then provisionally placed on the anterolateral aspect of the proximal humerus, and nonlocking 3.5 mm screws were initially placed in the shaft holes of the plate to assist with indirect reduction of the fracture. Tuberosity sutures were then tied to the plate and the humeral head was filled with 4.0 mm cannulated locking screws after they were measured by sounding technique in which a K-wire was used to drill until subcortical bone was felt, measured, and a screw 4mm shorter than the measurement was placed. The maximum number of screws possible was placed in the humeral head as dictated by the plate and the size of the patient’s humeral head, and a minimum of 1 calcar screw (CS) was placed in all cases. A 3.0 mm hole was drilled through the lateral aspect of the greater tuberosity through a proximal locking screw hole in the plate. The Osteocrete (Bone Solutions, Colleyville, TX, USA) was mixed on the back table until a pasty consistency was achieved after approximately 5-7 minutes that allows diffusion of the bone filler through the impacted chips without extravasation (Fig. 1 D-F). The syringe was inserted thorough the lateral hole and the bone filler was injected under live fluoroscopy and pressurized into the metaphyseal region and the humeral head. Care was taken to ensure that the bone filler was spreading within the head/neck area and there was no extravasation into the joint. A final postoperative radiograph is shown in Figure 1, G.

Figure 1.

(A and B). Radiographs showing a 3-part proximal humerus fracture dislocation. (C) Intraoperative photograph showing cancellous chip insertion through cortical window using a funnel. (D-F) Intraoperative photographs demonstrating how Osteocrete is mixed using the included instrumentation, and inserted using a pressurizer and long narrow tip. (G) Immediate postoperative radiograph in another patient demonstrating reduced fracture with plate fixation, impaction of cancellous graft, and osteocrete.

Radiographic measurements

A preoperative anteroposterior (AP) and Grashey radiographs were used to classify the fractures based on Neer and Arbeitsgemeinschaft für Osteosynthesefragen/Orthopaedic Trauma Association classifications, as well as the medial cortical ratio. Medial cortical ratio is defined by dividing the thickness of the medial cortex by the width of the humeral shaft measured 20 mm below the medial calcar (Fig. 2, A).²¹ The initial postoperative AP and Grashey radiographs were utilized to assess quality of fracture reduction and tuberosity repair based on previously published parameters, including the neck-shaft angle (NSA), humeral head displacement (HHD), cranialization of the greater tuberosity, and integrity of the medial calcar, ³⁵ presence of a CS,²⁷ and overall anatomic/acceptable/malreduction.³⁸ (Fig. 2, B-D, Table I). The number of screws in the shaft and the number of screws in the head were also recorded. Radiographic outcome after ORIF was assessed on the final AP and Grashey radiographs based and healing was defined as bony union of all fracture fragments on both views. The final grade of osteoarthritis and rotator cuff tear arthropathy was recorded based on the Samilson-Prieto and Hamada classifications using consensus reads by two observers (Fig. 2, D).^10,35 AVN of the humeral head was determined based on the presence or absence of humeral head collapse.

Figure 2.

(A) Preoperative radiographs demonstrating measurement of medial cortical ratio. (B) First postoperative radiographs demonstrating measurement of head-shaft displacement. (C) First postoperative radiographs demonstrating residual greater tuberosity cranialization. (D) Measurement of neck-shaft angle on final X-rays, which demonstrate Hamada Stage 3 and Samilson-Prieto Grade 2 arthritis.

Table I.

Anatomic, acceptable, and malreduction criteria (Adapted from Schnetzke et al³⁸).

Metric	Quality of reduction
	Anatomical	Acceptable	Malreduced
Head-shaft displacement (mm)	0	≤5	>5
Head-shaft angulation (deg)	120°-150°	110° to <120°	<110° or >150°
Greater tuberosity cranialization (mm)	0	≤5	>5

Patient reported outcome measures

Patients were contacted for an in-person postoperative visit beyond two years postoperatively. Patients’ pain level, satisfaction, range of motion and reoperations were recorded. Standardized PROMs were also collected including Veterans RAND 12-Item Health Survey (VR-12), the Single Assessment Numeric Evaluation, American Shoulder and Elbow Surgeons score (ASES), Penn Shoulder Score and Kerlan-Jobe Orthopedic Clinic Shoulder and Elbow Score (if the patient participated in athletic activities). Return to work and sport were also collected.^1,7,17,45,46

Statistical analysis

Demographic and PROM data was reported using descriptive statistics. Median and interquartile ranges were used for to present the distribution of continuous variables.

Results

20 patients with 3- or 4- part PHFs underwent ORIF with the described technique during the study period. Demographic data for this cohort are listed in Table II. Three patients were lost to follow-up within 6 months (two were deceased, and one did not come back after 3 month follow-up); therefore, we included 17 patients in our radiographic analysis (mean follow-up 34.1 months). Median age was 62 years (interquartile range [IQR] 11) with 5 males (29.4%) and 12 females (70.6%) patients. Two of the remaining 17 patients passed away prior to being contacted for later in-person follow-up. Mean follow-up duration for this later time point with PROMs (n = 15) was 34 months (13-57). In the final cohort, 13 patients (76.5%) had 3-part fractures, whereas 4 (23.5%) had 4-part fractures.

Table II.

Demographic data of cohort undergoing open reduction and internal fixation with standardized augmentation for proximal humerus fractures (n = 17).

Variables	No (%)	Median (IQR)
Age		62 (11)
Gender (% female)	12 (70.6)
CCI		3 (4)
BMI		26.6 (7.8)
Fracture classification
3-part	13 (76.5)
4-part	4 (23.5)
Time to surgery (weeks)		1.6 (1.9)
Smoker (% yes)	2 (11.8)

IQR, interquartile range; BMI, body mass index; CCI, Charlson comorbidity index.

Reduction quality

Early postoperative reduction parameters on the first postoperative AP and orthogonal radiograph (1-4 weeks postoperatively) are listed in Table III. All patients had a CS as part of their fixation constructs. The median number of shaft screws was 3 (2-8), whereas the median number of head screws was 8 (7-10). With respect to reduction parameters, mean HHD was 3.2 mm (0-7.8 mm), mean greater tuberosity cranialization was 0.5 mm (0-4.4 mm), the medial hinge was intact in 11/17 cases (64.7%), and the mean NSA was 130.3° (117.4°-137.9°). Based on previously defined criteria,³⁸ 13/17 (76.5%) of fractures had an anatomical or acceptable reduction, whereas 4/17 (23.5%) were malreduced. All malreductions were based on HHD >5mm.

Table III.

Radiographic parameters before and after open reduction and internal fixation with standardized augmentation for proximal humerus fractures (n = 17).

	No (%)	Median (IQR)
Preoperative
Neer classification
3-part	70%
4-part	30%
AO classification
11B1.1	60%
11B1.2	6%
11C3.1	30%
11C3.3	6%
Medial cortical ratio	0.17
Dislocation (%)	6%
Early postoperative (within 6 weeks)
Calcar screw present (%)	100%
Head screws (n, range)		8.2 (6-10)
Shaft screws (n, range)		3 (2-8)
Humeral head displacement (mm)		3.2 (SD 3)
GT cranialization (mm)		0.5 (SD 1.3)
Medial hinge intact (%)	65%
Head-shaft angle (deg)		130.3 (SD 6.1)
Presence of osteocrete in glenohumeral joint (%)	0%
Latest follow-up (mean 34.1 mo)
Tuberosity healing (%)	94%
Tuberosity resorption (%)	18%
Head-shaft healing (%)	100%
Head-shaft angle (deg)		133.3 (SD 6.6)
Humeral head AVN (%)	12%
Screw cutout (%)	12%
Samilson-Prieto osteoarthritis classification (%)
Grade 0	70%
Grade 1	12%
Grade 2	18%
Hamada classification
Stage 1	76%
Stage 2	18%
Stage 3	6%

IQR, interquartile range; GT, greater tuberosity; SD, standard deviation; AVN, avascular necrosis; AO, Arbeitsgemeinschaft für Osteosynthesefragen.

Percentages are reported as percentages of the final cohort. Numerical values are reported as mean and standard deviation.

Radiographic outcomes

Final radiographic analysis (mean 34.1 months) showed a high rate of healing and low rate of complications in this series. Complete radiographic healing was observed in 14/17 patients (82%) at latest follow-up. All patients had radiographically confirmed healing of the humeral head to the shaft. One patient had subsequent rotator cuff failure 9 months after ORIF and underwent reverse total shoulder arthroplasty 11 months later.

Complications were uncommon (2/17 patients) and included tuberosity migration and resorption and late AVN of the humeral head. Two patients (12%) developed AVN with subsequent humeral head collapse and intraarticular screw penetration. Of these two patients, one was doing well until 2 years postoperatively at which point she began undergoing treatment for advanced cancer and last radiograph showed AVN with collapse. She underwent a percutaneous hardware removal as a definitive treatment. The other patient did not follow-up after their 4-week visit; however, she followed up at 3 years postoperatively and noted to have developed pain and dysfunction 18 months after surgery. She opted to proceed with nonoperative management of her AVN. Tuberosity nonunion occurred in one patient and two additional patients experienced tuberosity resorption despite initial evidence of healing. In each case, radiographs demonstrated healed fractures at 3 months, and tuberosity resorption was noted at later follow-up between 2 and 3 years. The Samilson-Prieto arthritis grade was zero in 12 patients (70%), one in 2 patients (12%), and two in 3 patients (18%). The Hamada grade was one in 13 patients (76%), two in 3 patients (18%), and three in 1 patient (6%).

Patient reported outcomes

PROMs were available for 15/20 (75%) of patients at latest follow-up. The overall median Penn score was 89 (IQR = 19), the ASES score was 92 (IQR = 25) and the Single Assessment Numeric Evaluation score was 85 (IQR = 18.5). The median VR-12 mental component summary was 53.9 (IQR = 11.9), whereas the VR-12 PCS was 51.6 (IQR = 12.8). Seven out of nine (78%) patients who worked prior to injury returned to work at a mean of 14.7 weeks postoperatively. Three patients returned to sporting activities, with a mean Kerlan-Jobe Orthopedic Clinic score of 70.3.

Discussion

The present study demonstrates that when electing to treat 3- and 4-part PHFs with ORIF using a precontoured proximal humeral locking plate, a standardized augmentation protocol with impaction grafting using cancellous bone allograft and a magnesium-based bone filler results in a high rate of radiographic healing, good short term clinical outcomes, and a low complication rate. We did not observe any loss of reduction or failures within the first 6 months postoperatively. At latest radiographic follow-up, 14/17 (82%) of patients had complete union of their fractures with no AVN or hardware-related complications. In addition, across various PROMs evaluating pain, function, and satisfaction patients did well postoperatively.

A systematic review of 3- and 4-part PHFs treated surgically identified 2939 patients who underwent ORIF.⁸ They found that ORIF had higher ASES scores than arthroplasty and reported that the mean ASES score was 80.7. A systematic review evaluating fixation with and without fibular strut grafts found four comparative studies that reported ASES scores; the mean score was 84.3 in the fibular strut group and 78.7 in the locking plate-only group.⁴ Although the sample size of the present study was smaller than these studies, the mean ASES scores were higher than in previously reported data.

Previous review articles and multicenter studies have described high complication and reoperation rates after PHF locking plate fixation, though most do not separately evaluate 3- and 4-part fractures. In a systematic review of 2-, 3-, and 4-part fractures (n = 514) across 12 studies with minimum 18 months follow-up, Sproul et al found an overall complication rate of 33% excluding varus malunion, with an AVN rate of 10% and a screw penetration rate of 8%.⁴⁰ A prospective, multicenter study evaluating 155 patients with 12 months follow-up duration including all fracture types found a complication rate of 34% with a reoperation rate of 19%.⁴² Another study reviewed all PHFs treated at a major institution over 27 years and found the revision rate in 3- and 4-part fractures to be greater than 2-part fractures.³⁴ A randomized trial comparing ORIF to nonoperative management of 3-part PHFs found improved outcomes in the ORIF group; however, the reoperation rate was 30% despite a high-quality reduction in 86% of patients.²⁸

The major complication in our cohort was late AVN that was observed in 2 patients (12%). Of these two patients, one was doing well clinically and radiographically at one year postoperatively prior to beginning treatment with chemotherapy and multiple courses of steroids and developing AVN three years postoperatively. The other did well clinically for 18 months and followed up at almost 3 years postoperatively, where she was found to have AVN and collapse with intraarticular screw penetration and opted to proceed with nonoperative treatment.

Several studies have sought to evaluate surgical techniques and methods to improve outcomes after proximal humerus ORIF. Several factors related to the fixation construct have been shown to be significant with respect to improving outcomes, including suture augmentation, number and length of screws in the humeral head and, most importantly, support of the medial calcar with CSs in the inferior 25% of the head.^25,29, 30, 31^,41 Quality of reduction has been shown to be a critical factor, and various metrics have been defined that correlate with improved healing rates, including restoration of the medial calcar, anatomic reduction of the head and tuberosities, and restoration of the appropriate NSA. Reduction metrics in the present study were excellent, with mean displacement of 3.2 mm (standard deviation [SD] 3.0), mean GT cranialization of 0.5 mm (SD 1.3), and mean head-shaft angle of 130.3° (SD 6.1).

In addition to these metrics, given the challenges in obtaining and maintaining a good reduction in these fractures, there has been increasing interest in augmentation with allograft, autograft, or synthetic materials, especially in the setting of poor bone quality. Various methods of augmentation have been studied, including femoral head allograft, fibula strut grafts, cancellous allograft chips, resorbable bone fillers, and bone cement (PMMA) with or without fenestrated screws.^29,41

A comprehensive review of the published comparative clinical data (eight studies) on fibular strut allografts found that fibula strut grafts may improve rates of healing, although these studies are all relatively small, heterogenous cohorts.²⁷ Another review evaluating clinical and biomechanical studies found improvements in load to failure, construct stiffness, and fracture displacement as well as better clinical outcomes, particularly in patients with four-part fractures.³ However, a randomized controlled trial with 80 patients with medial calcar comminution that underwent ORIF with or without fibular strut graft augmentation found no difference in any clinical or radiographic outcomes at 12 months postoperatively.⁴⁴ It is also well-established that revision arthroplasty of PHFs that previously had fibular strut augmentation can present significant challenges.³⁹ In this setting, as much as 80% of grafts may be fully integrated into host bone resulting in difficulty of humeral reconstruction during revision surgery and in some cases significant bone loss when the strut graft needs to be removed for shoulder arthroplasty.²³

The use of PMMA has been shown in several biomechanical studies to increase the stability and efficacy of proximal humeral fixation constructs.³ However, a recent randomized-controlled trial showed no difference comparing PMMA augmentation to standard locked plating was terminated prior to complete enrollment due to the minimal differences seen between groups.¹¹ A systematic review and meta-analysis on the subject found that there was no difference in revision rate, AVN, or Constant score, but the rate of implant failure was lower in the PMMA augmentation group.¹⁸

There is limited clinical data evaluating the effects of augmentation with bone fillers after proximal humerus ORIF. A review by Biermann et al evaluated the biomechanical studies in the literature investigating bone fillers such as cancellous allograft and bone cement.³ They found that constructs with synthetic bone fillers had decreased interfragmentary strain, increased maximum load to failure, and increased axial stiffness. A retrospective review compared patients that received no augmentation, augmentation with cancellous chips only, and augmentation with calcium phosphate cement.⁵ They found that augmentation with calcium phosphate alone reduced rates of humeral head settling and decreased screw penetration postoperatively when compared to the other groups.

We have developed a protocol that combines anatomic reduction techniques and augmentation using impaction grafting of cancellous bone chips with injection of a synthetic magnesium-based bone filler (Osteocrete; Bone Solutions, Colleyville, TX, USA) to help reduce postoperative loss of reduction. The present study is the first to document clinical performance of this technique and demonstrates there is a high success rate with only two patients demonstrating late collapse and AVN. This has become the preferred technique for the senior author (VE) and is utilized in a standardized fashion for all 3- and 4-part fractures treated operatively with ORIF.

There are several limitations to the present study. This is a case series, and as such there is no comparison group, which limits the ability to draw conclusion about the relative benefits of the current technique. In addition, the present study has a relatively small sample size with three patients lost to follow-up, and this may have influenced results. Preoperative computed tomography scan was not utilized for categorization of fractures. Despite these limitations, there are several strengths of the study. The mean follow-up duration is over 24 months, allowing ample time for fracture healing. In addition, we performed a robust clinic and radiographic surveillance of patients including fracture healing and progression of arthritis in this cohort. We only included 3- and 4-part PHFs that are considered difficult fracture patterns for ORIF, which demonstrates the validity of this standardized fracture augmentation technique.

Conclusion

Augmentation of 3- and 4-part proximal humeral fractures with a standardized protocol utilizing impaction grafting with cancellous chips and a synthetic magnesium-based bone filler in combination with proximal humeral locking plate fixation results in a high rate of maintenance of reduction and satisfactory patient outcomes. Further comparative data in larger series is needed to evaluate the efficacy of this technique compared to alternative methods of augmentation.

Disclaimers:

Funding: No funding was disclosed by the authors.

Conflicts of interest: Jason C. Ho, MD reports Consultant and advisory board member for Shoulder Innovations; Paid Speaker for Enovis/DJO; Eligible for Royalties from Aevumed; Board Member: Ohio Shoulder Elbow Society, Phoenix Shoulder Society. Eric Ricchetti, MD reports AAOS: Board or committee member; American Board of Orthopaedic Surgery, Inc.: Board or committee member; American Shoulder and Elbow Surgeons: Board or committee member; Enovis: IP royalties, Paid consultant, Paid presenter or speaker; Journal of Shoulder and Elbow Surgery: Editorial or governing board. Vahid Entezari, MD reports Stryker designer and consultant; Depuy Synthes consultant; Orthokinetic Track LLC founder and medical advisor; Tigon medical LLC designer and consultant; Ateron LL consultant and SAB member; ASES research and patient centered initiative (PCI) committee member; OREF and Enovis research grant recipient. The other authors, their immediate families, 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.