Bone morphogenetic protein alone vs in combination with bone graft in management of non-union of tibia fractures

Abstract

BACKGROUND

Both bone morphogenetic protein (BMP) and bone graft have been shown to promote healing in fracture non-unions.

AIM

To compare the effectiveness of BMP alone vs BMP combined with bone graft in achieving radiological union in tibial non-union fractures over 3 months after intervention.

METHODS

A retrospective cohort study was conducted, collecting data from January to July 2023. A total of 126 participants with tibial non-unions were categorized into either a BMP group (n = 63) or a BMP with bone graft group (n = 63). The radiological union was assessed at 3 months post-surgery. χ² tests were applied to compare union rates. Stratified analyses were conducted to examine union outcomes across subgroups based on age, gender, injury mechanism, fracture type, diabetes status, and non-union duration.

RESULTS

Radiological union was achieved in 61.9% of the BMP-alone group compared to 22.2% in the BMP with bone graft group (P < 0.001). Subgroup analyses showed that BMP alone had higher union rates in younger patients and female participants, patients with open fractures, those with diabetes, and cases with non-union duration over three months. BMP alone achieved an 81.4% union rate in diabetic participants vs 46.7% in the BMP with bone graft group. In patients with non-union duration exceeding three months, BMP alone had a union rate of 78.6% compared to 12.9% in the combined treatment group.

CONCLUSION

BMP alone demonstrated superior radiological union outcomes in tibial non-unions compared to BMP combined with bone graft, particularly in patients with specific characteristics such as diabetes, longer non-union duration, and open fractures.

Key Words: Tibia; Non unions; Bone morphogenetic protein; Bone grafts; Fractures; Healing; Fractures

Core Tip: Non-union of tibial fractures remains a significant challenge in orthopaedic trauma practice, often associated with prolonged morbidity and a substantial economic burden on healthcare systems. Bone morphogenetic proteins (BMPs) have emerged as a promising biological agent in fracture healing; however, their use in clinical practice, either alone or in combination with bone grafting techniques, remains variable and sometimes controversial due to cost-effectiveness and clinical outcomes. Our study directly compares the outcomes of BMP use alone vs in combination with autologous bone grafts in managing tibial non-unions. This manuscript contributes provides valuable insights that can aid clinical decision-making, particularly in resource-limited settings where the optimal use of biologics must be balanced against surgical and economic factors.

INTRODUCTION

A non-union fracture leads to considerable health complications, extended hospital stays, and higher costs[1]. Most fractures usually heal within five months or 20 weeks, but a non-union fracture shows incomplete healing after six months[2]. The incidence of non-union or delayed fractures in the tibia varies from 2.5% to 46% based on the bone's location and the extent of injury to the blood vessels, soft tissue, and bone[3]. Treating non-union fractures usually requires biological repair and mechanical fixation, often with an autologous bone graft (ABG) to aid osteogenic growth[4].

Despite the availability of bone graft substitutes, ABG remains the preferred treatment due to its convenience and the high presence of growth factors and progenitor cells in patients[3-5]. Recombinant human bone morphogenetic proteins (rhBMPs), a type of bone matrix glycoprotein, help stimulate the migration, proliferation, and differentiation of stem cells into chondroblasts and osteoblasts[1,6]. With advancements in graft adjuvants and synthetic implants, combining ABG with rhBMPs has become a popular approach to enhancing fracture healing, with reported healing rates of up to 92% in persistent non-union cases[7]. A meta-analysis indicated that rhBMPs and ABG offer similar healing outcomes in tibia non-union cases[8]. Surgery for long bone non-union remains challenging due to the associated risks. Current bone healing approaches focus on applying growth factors like rhBMPs, regarded as powerful osteo-inductive agents. Clinical data suggest rhBMPs have a higher potential for promoting long bone non-union healing compared to standard treatments, including ABG[6]. The rationale of the study is to compare bone morphogenetic protein (BMP) alone vs in combination with bone graft in the management of patients with non-union of long bone fractures, as rhBMP-2 use continues to be a debate because literature provides scarce evidence regarding its effectiveness. Hence, this study aims to compare the effect of BMP alone vs in combination with bone graft in the management of patients with non-union of tibia fractures.

MATERIALS AND METHODS

Data source and data collection

This retrospective cohort study was conducted in the Department of Orthopaedic Surgery at Jinnah Postgraduate Medical Centre, Karachi, over a six-month period from September 2023 to March 2024.

The sample size of 126 participants, with 63 individuals per group, was calculated using the World Health Organization sample size calculator for a one-sided hypothesis test of two population proportions[9]. The calculation was based on an expected union rate of 61% in the BMP group and 39% in the BMP with bone graft group[10], with a significance level of 5% and a power of 80%. The sample included patients aged 18-70 years with a tibial aseptic non-union fracture. Patients who otherwise met the inclusion criteria but had complicating factors such as pathological fractures, active infection, or a history of malignancy were excluded. Sampling was performed using non-probability consecutive sampling.

Data collected included age, gender, mechanism of injury, fracture duration, fracture characteristics (open or closed), and history of diabetes. Participants were divided into two groups: Group A received BMP alone, while group B received BMP with bone graft. All patients underwent open reduction and internal fixation using a tibial plating system supplemented with either BMP alone or BMP combined with bone graft. All surgeries were performed by a surgeon with at least two years of orthopaedic experience, and patients were followed postoperatively according to departmental protocol. three months after surgery, bone union was assessed based on bridging bone on at least three cortices on radiographs in anteroposterior and lateral.

Statistical analysis

For data analysis, IBM SPSS Statistics for Windows, Version 26.0. Armonk, NY, United States: IBM Corp was used to calculate frequencies and percentages for categorical variables, including gender, injury mechanism, fracture characteristics, diabetes history, and radiological union. Age and duration of non-union were summarised using mean ± SD. Union outcomes between groups were compared using the χ² test, with P ≤ 0.05 considered statistically significant. To control for potential effect modifiers (age, gender, injury mechanism, fracture characteristics, diabetes history, and non-union duration), stratification was applied, followed by post-stratification χ² analysis. This study investigates the difference in effects of BMP alone vs BMP combined with bone graft in managing tibia non-union fractures.

RESULTS

A total of 126 patients were included in the study, with 63 each in BMP and BMP with bone graft groups. The participants had an average age of 42.92 years. Of the total, 24.6% were male and 75.4% were female. Road traffic accidents (RTAs) accounted for 30.2% of injuries, while falls made up 69.8%. Most fractures were closed (77.0%), while 23.0% were open fractures. Regarding diabetes, 46.0% of participants had diabetes mellitus, with the remaining 54.0% non-diabetic. Non-union fractures had an average duration of 3.21 months. In terms of distribution, 64.3% had a non-union duration of three months or less, while 35.7% exceeded three months. Radiological union was achieved in 42.1% of cases, while 57.9% showed no union (Table 1).

Table 1 Characteristics of the participants and their fracture (n = 126), mean ± SD/n (%).

Characteristics
Age (years)	42.92 ± 7.99
Gender
Male	31 (24.6)
Female	95 (75.4)
Mechanism of injury
RTA	30 (30.2)
Fall	88 (69.8)
Fracture characteristics
Open	97 (77.0)
Closed	29 (23.0)
Diabetes mellitus
Yes	58 (46.00)
No	68 (54.00)
Duration of non-union (months)	3.21 ± 0.86
Non-union duration distribution (months)
≤ 3	81 (64.30)
> 3	45 (35.70)
Radiological union
Yes	53 (42.10)
No	73 (57.90)

RTA: Road traffic accidents.

In Table 2, 61.9% achieved radiological union in the BMP group, while 38.1% did not, making for a significant difference (P < 0.001). Baseline demographic and clinical characteristics of the study population are summarized in Table 1, and stratified comparisons of treatment outcomes according to these baseline variables are presented in Table 3. The BMP with bone graft group showed lower success, with only 22.2% achieving radiological union and 77.8% not achieving it.

Table 2 Comparison of radiological union with respect to group (n = 126), n (%).

Group	Radiological union achieved (yes)	Radiological union not achieved (no)	Total	P value
BMP	39 (61.9)	24 (38.1)	63 (100)	< 0.001
BMP with bone graft group	14 (22.2)	49 (77.8)	63 (100)

BMP: Bone morphogenetic proteins.

Table 3 Stratified analysis of radiological union outcomes by participant characteristics and treatment group, n (%).

Factor	Group	Radiological union achieved (yes)	Radiological union not achieved (no)	Total	P value
Age ≤ 43 years	BMP	39 (62.9)	23 (37.1)	62 (100)	0.208
	BMP with bone graft	5 (41.7)	7 (58.3)	12 (100)
Age > 43 years	BMP	0 (0)	1 (100)	1 (100)	> 0.999
	BMP with bone graft	9 (17.6)	42 (82.4)	51 (100)
Gender (male)	BMP	9 (56.3)	7 (43.8)	16 (100)	0.066
	BMP with bone graft	3 (20.0)	12 (80.0)	15 (100)
Gender (female)	BMP	30 (63.8)	17 (36.2)	47 (100)	< 0.001
	BMP with bone graft	11 (22.9)	37 (77.1)	48 (100)
Mechanism (RTA)	BMP	8 (66.7)	4 (33.3)	12 (100)	0.004
	BMP with bone graft	5 (19.2)	21 (80.8)	26 (100)
Mechanism (fall)	BMP	31 (60.8)	20 (39.2)	51 (100)	0.001
	BMP with bone graft	9 (24.3)	28 (75.7)	37 (100)
Fracture (closed)	BMP	7 (35.0)	13 (65.0)	20 (100)	> 0.999
	BMP with bone graft	3 (33.3)	6 (66.7)	9 (100)
Fracture (open)	BMP	32 (74.4)	11 (25.6)	43 (100)	< 0.001
	BMP with bone graft	11 (20.4)	43 (79.6)	54 (100)
Diabetes	BMP	35 (81.4)	8 (18.6)	43 (100)	0.017
	BMP with bone graft	7 (46.7)	8 (53.3)	15 (100)
No diabetes	BMP	4 (20.0)	16 (80.0)	20 (100)	0.719
	BMP with bone graft	7 (14.6)	41 (85.4)	48 (100)
Non-union duration ≤ 3 months	BMP	28 (57.1)	21 (42.9)	49 (100)	0.025
	BMP with bone graft	10 (31.3)	22 (68.8)	32 (100)
Non-union duration > 3 months	BMP	11 (78.6)	3 (21.4)	14 (100)	< 0.001
	BMP with bone graft	4 (12.9)	27 (87.1)	31 (100)

BMP: Bone morphogenetic proteins; RTA: Road traffic accidents.

A χ² test was applied, with a significance level of 0.05. Stratification was done to see the effect of baseline characteristics on the radiological non-union, presented in Table 3. For participants aged ≤ 43, those treated with BMP alone had a radiological union rate of 62.9%, while this rate dropped to 41.7% in the BMP with bone graft group. Among participants over 43, radiological union was rare, with only the BMP with bone graft group showing any success (17.6%), albeit minimal. Gender also influenced outcomes. Females treated with BMP had a significantly higher union rate (63.8%) than those in the BMP with bone graft group (22.9%). In males, BMP was still more effective, though the difference was smaller.

Mechanism of injury impacted union rates: Those injured in RTAs and treated with BMP achieved union in 66.7% of cases, compared to only 19.2% in the BMP with bone graft group. Similarly, for those who had falls, BMP was more successful at achieving union (60.8%) than BMP with bone graft (24.3%). Moreover, open fractures responded better to BMP, with a union rate of 74.4%, while BMP with bone graft had much lower success (20.4%). Diabetes seemed to predict better union outcomes in the BMP group, achieving union in 81.4% of cases, compared to only 46.7% in the BMP with bone graft group. In those without diabetes, the union rates were notably lower for both groups. Non-union duration also played a role. For durations over three months, BMP yielded a high union rate of 78.6%, while BMP with bone graft had a low success rate of 12.9%. For shorter non-union durations, BMP still outperformed BMP with bone graft but with a narrower margin (57.1% vs 31.3%).

Overall, BMP alone generally led to better radiological union outcomes, especially in younger participants, females, those with open fractures, and those with diabetes or longer non-union durations.

DISCUSSION

Non-union in long bone fractures leads to higher risks of complications and poor long-term outcomes[11]. Fracture healing depends on several factors: The fracture type and location, surgical dissection extent, fracture stability, presence of infection, blood supply, quality of mechanical fixation, nutrition, and chronic health conditions[12]. While most fractures typically heal within five months or 20 weeks, non-union is defined as a lack of healing after six months[2]. To promote healing in cases of delayed union or non-union, local biological stimulation can be applied at the fracture site. Current bone healing approaches focus on applying growth factors like rhBMPs, regarded as powerful osteoinductive agents.

The results of this study emphasise the efficacy of BMP alone over BMP combined with bone graft in achieving radiological union in tibial non-union fractures. This outcome aligns with research indicating the potent osteo-inductive capabilities of BMPs, which can effectively stimulate osteoblast differentiation and bone formation. These findings suggest that additional grafting material may not always be necessary.

Contemporary studies have increasingly highlighted BMP’s potential as a solo agent in promoting bone healing, particularly due to its strong osteo-inductive properties. Multiple studies indicate that BMPs are effective in promoting non-union healing, and they also contribute to reducing infection and alleviating pain[1,13,14]. Peterson et al[14] highlighted the effectiveness of bone grafts in improving healing, particularly in cases with weakened bone structure, where autografts helped with osseous bridging and boosted callus formation. However, our results indicate that BMP alone might be sufficient for many tibia non-union cases, especially in patients with specific demographic and fracture profiles. This aligns with findings from Patil et al[15], who reported high union rates with less invasive autologous methods like bone marrow injections, which offer osteogenic support without the drawbacks of traditional grafting. Similarly, a meta-analysis also indicated that using BMPs, mesenchymal stem cells and platelet rich plasma in fracture treatment has proven effective and safe. Moreover, these methods help to speed up bone healing and lead to good postoperative functional outcomes[16].

When comparing BMP to ABG, Gupta et al[17] found that autografts alone achieved strong union rates, whereas adding synthetic materials such as beta-tri-calcium phosphate to grafts often led to poorer outcomes in tibia fractures. This supports our observation that BMP alone can outperform BMP combined with grafts, likely because it results in fewer immune responses and complications after surgery. Konda et al[18] also noted that while iliac crest bone grafts were effective for atrophic tibial non-unions due to their osteoinductive properties, BMP showed similar or even better results, avoiding the common complications of graft harvesting. A study found that using ABG alone was as effective and less expensive than combining it with recombinant BMP for treating aseptic non-unions in tibia, femur, or humerus[19,20]. Some suggest that the drive behind BMP treatment development may stem more from its commercial potential in the bone graft substitute market than from actual gains in safety and effectiveness[2,21].

Patient-specific factors also revealed unique insights, particularly in diabetic participants, where BMP alone resulted in an 81.4% union rate, consistent with newer literature showing BMP’s adaptability in patients with compromised healing capacity[22]. Some studies have questioned BMP’s efficacy in diabetic patients due to potential cellular response delays. Our findings suggest that BMP alone may indeed overcome such limitations, indicating its therapeutic potential across diverse patient profiles.

This study has some limitations. The follow-up period was relatively short, which limits insights into long-term outcomes and potential late complications. Our sample size, though statistically sufficient, was drawn from a single centre, which could affect how widely the findings apply. While BMP showed positive results in this study, variations in BMP formulation, dosage, and application methods may impact its effectiveness in other clinical settings. Larger, multicentre randomised controlled trials with longer follow-up and a focus on cost-effectiveness would provide a more detailed view of BMP’s role in treating tibial non-unions.

CONCLUSION

BMP alone demonstrated better radiological union outcomes compared with BMP combined with bone graft. BMP alone achieved higher radiological union rates than when combined with bone graft. Hence, it can be inferred that BMP’s osteoinductive properties might eliminate the need for additional graft material in certain cases. The study highlights the efficacy of BMP alone in treating tibial non-unions, especially in specific patient groups, such as younger individuals, females, those with diabetes, and cases involving open fractures. BMP alone could be a reliable treatment option, avoiding the complications and potential immune responses linked with grafting procedures. Clinicians should consider that BMP alone may represent a less invasive treatment option, with reduced operation time, surgical risks, and potential complications.