Key points

Introduction

The discovery of bone morphogenetic proteins (BMP) by Urist in 1965 was met with great optimism. The finding of an osteoinductive compound created the potential for manufacturing a growth factor that would assist with bone formation and healing. Over the years, numerous studies (both animal and human) showed the efficacy of BMP in enhancing bone growth. Moreover, many of these studies have found ancillary benefits of BMP, such as decreasing infection rates and time to wound healing. In 2001, the US Food and Drug Administration (FDA) gave marketing clearance for rhBMP-7 (OP-1; Stryker [Stryker Corporation, Kalamazoo, Michigan]) to be used for recalcitrant long bone nonunions. Subsequently, in 2004, rhBMP-2 (INFUSE; Medtronic [Medtronic, Fridley, Minnesota]) was approved for treatment of open tibia shaft fractures. Unfortunately, recent reports of complications have overshadowed these early promising results. Increased wound drainage, excessive bone growth, neuropathy, and even carcinogenesis have been presented as complications after use of BMP. Additionally, concern over lack of mechanical strength and the high cost associated with BMP have been cited as shortcomings. Many of the reports on complications occurred after use around the spine; however, there are also reports of complications associated with fracture care.

Approximately 10 years ago BMP was seen as a miraculous adjuvant to assist with bone growth. However, in the face of an increasing number of complications and a lack of understanding its long-term effects, it is unclear what role BMP has in the current treatment of orthopedic trauma patients. This article to reviews the current recommendations, trends, and associated complications of BMP use in fracture care.

Introduction

The discovery of bone morphogenetic proteins (BMP) by Urist in 1965 was met with great optimism. The finding of an osteoinductive compound created the potential for manufacturing a growth factor that would assist with bone formation and healing. Over the years, numerous studies (both animal and human) showed the efficacy of BMP in enhancing bone growth. Moreover, many of these studies have found ancillary benefits of BMP, such as decreasing infection rates and time to wound healing. In 2001, the US Food and Drug Administration (FDA) gave marketing clearance for rhBMP-7 (OP-1; Stryker [Stryker Corporation, Kalamazoo, Michigan]) to be used for recalcitrant long bone nonunions. Subsequently, in 2004, rhBMP-2 (INFUSE; Medtronic [Medtronic, Fridley, Minnesota]) was approved for treatment of open tibia shaft fractures. Unfortunately, recent reports of complications have overshadowed these early promising results. Increased wound drainage, excessive bone growth, neuropathy, and even carcinogenesis have been presented as complications after use of BMP. Additionally, concern over lack of mechanical strength and the high cost associated with BMP have been cited as shortcomings. Many of the reports on complications occurred after use around the spine; however, there are also reports of complications associated with fracture care.

Approximately 10 years ago BMP was seen as a miraculous adjuvant to assist with bone growth. However, in the face of an increasing number of complications and a lack of understanding its long-term effects, it is unclear what role BMP has in the current treatment of orthopedic trauma patients. This article to reviews the current recommendations, trends, and associated complications of BMP use in fracture care.

History and mechanism of bone morphogenetic protein

BMPs are a part of the transforming growth factor-β superfamily that is responsible for tissue repair and embryogenesis. Twenty different BMPs have been discovered with many of them appearing to function in different ways. When acting together, these growth factors are able to provide a signal that causes mesenchymal stem cells to differentiate into osteoblasts, also known as osteoinduction . Specific proteins (BMP-2, -6, and -9) work early in the differentiation process, whereas most of the others, including BMP-7, help stimulate the final transition of preosteoblast to osteoblasts.

Urist is credited as being the founder of these growth factors after he implanted decalcified bone within rodent muscle and noticed subsequent bone growth. He defined this phenomenon as osteoinduction . In 1988, Johnson and colleagues reported the first clinical outcomes of purified human BMP used to treat femoral nonunions. Eleven of the 12 femoral nonunions treated healed at an average of 4.7 months. Johnson performed additional clinical trials that continued to show promising results. However, it became apparent early on that isolating large quantities of BMP from cadaveric bone was difficult and not a viable source for mass scale production. Moreover, the specific dose of BMP required for efficacy was unknown, which led to use of recombinant gene technology to create specific BMPs that show evidence of osteoinduction alone. Both rhBMP-2 (INFUSE) and rhBMP-7 (OP-1) are proteins that are now manufactured in mass quantities via recombinant technology. These proteins are currently the most widely used and most studied of the BMP family. Further studies followed that found that low doses of BMP resulted in minimal bone formation; however, higher doses of BMP can result in excessive bone formation and even bone resorption secondary to osteoclast activation. Termaat and colleagues explained that “The dose of BMP needed for clinical efficacy must overcome a threshold, and the dose-response curve becomes steeper as one progresses from rodent to nonhuman primate.” The specific reason for this is still unclear. At this time, the current recommended doses of BMP-2 and BMP-7 are more than 1000 times greater than those of native concentrations.

Treatment of acute fractures

The treatment of open fractures is associated with high complications rates and poor functional outcomes. Loss of soft tissue and bone in these injuries may lead to delayed healing and nonunion. Much of the clinical research with BMP use in acute extremity injuries involves open fractures, specifically, open tibia fractures.

In 2002, the BMP-2 Evaluation in Surgery for Tibial Trauma (BESTT) trial was performed to determine the safety and efficacy of rhBMP-2 in the treatment of open tibia shaft fractures fixed with both reamed and unreamed intramedullary nails (IMN). Four hundred fifty patients with open tibia shaft fractures were randomly divided into 3 groups:

• 1.
  

  The standard of care group (IMN plus soft tissue management)

  
  
• 2.
  

  The standard of care with 0.75 mg/mL of rhBMP-2

  
  
• 3.
  

  The standard of care with 1.50 mg/mL of rhBMP-2

The authors found that the group with 1.50 mg/mL of rhBMP-2 had significantly fewer reoperations, infections, wound complications, and hardware failures. Additionally, the group with the 1.50 mg/mL dose had faster healing times. At 1 year follow-up, the adverse events in the BMP group were similar to what was seen in a normal trauma setting. The authors stated that rhBMP-2 is a “novel adjunct” and advantageous when compared with the standard of care when treating long bone fractures. However, other investigators noted the disproportionate amount of patients in the control group that received unreamed IMN when compared with the study group. The effect of reaming when treating tibia fractures with IMN has been well studied. The Study to Prospectively Evaluate Reamed Intramedullary Nails in Patients with Tibial Fractures (SPRINT) trial found that the reaming is a strong confounding variable to consider in the BESTT trial results.

In 2006, Swiontkowski and colleagues combined the data from the BESST trial with data from another prospective, randomized trial using the same methods. Patients from 2 subgroups were analyzed:

• 1.
  

  131 patients with Gustilo-Anderson type 3A or 3B fractures

  
  
• 2.
  

  113 patients treated with reamed IMN

This analysis found significant improvements in secondary procedures and infections in those from the first subgroup receiving rhBMP-2 at a concentration of 1.50 mg/mL. The second subgroup showed no difference between those that received rhBMP-2 and those that did not, but a trend toward improved outcomes in the treatment group was noted by the authors ( P >.30). The study concluded that rhBMP-2 significantly reduces the frequency of bone grafting and secondary procedures when treating Gustilo-Anderson type 3 open tibia fractures. However, the authors admit that the study was not originally designed for a subgroup analysis and that the data alone “should be viewed with caution.”

McKee and colleagues evaluated the role of rhBMP-7 (OP1) in the treatment of open tibia fractures treated with IMN. One hundred twenty-four patients were randomly assigned to a control group (simple closure) and compared with those who had simple closure with the addition of 3.4 mg of rhBMP-7 at the fracture site. Patients in the rhBMP-7 group showed significant improvements in rates of reoperation, overall pain, and functional scores. No adverse events were associated with rhBMP-7 use. These data was presented at the Orthopaedic Trauma Association’s annual meeting in 2013 but have not been published to date.

Most recently, Aro and colleagues performed a randomized trial of 277 patients to evaluate the role of rhBMP-2 in the treatment of open tibia shaft fractures fixed with only reamed IMN. In contrast to the BESST study, no significant difference was found between the groups. A notable difference between the 2 studies was the higher proportion of Gustilo-Anderson type 3 injuries in the BESTT study (43% vs 32%). Additionally, the BESST study had a lower percentage (22%) of Gustilo-Anderson type 1 injuries than the study by Aro and colleagues (29%). This discrepancy could explain the different outcomes presented, as Gustilo-Anderson type 3 injuries have been associated with higher reoperation rates. A trend toward increased deep infection rates was seen in the rhBMP-2 group in the study by Aro and colleagues, but this did not reach significance ( P = .0645). This finding differs from those of other studies that suggest that BMP may assist with decreasing deep infection rates. The reason for this finding remains unclear. The authors suggest that it may be associated with the surgical technique, as many of the infections in the treatment group occurred in Gustilo-Anderson type 1 injuries. The collagen delivery device for the BMP could have potentially placed pressure on these small wounds leading to necrosis and possible infection. Notably, the authors state that although the treatment had a higher number of infections, the overall fracture healing rate was not affected.

A recent meta-analysis of randomized trials was published discussing BMP use in open tibia fractures. Four studies that specifically analyzed BMP use in open tibia fractures were evaluated. The authors concluded that there is some evidence for improved union rates in patients treated with BMP versus standard of care; however, “no clear answers” were established in the study. Final recommendations included the need for well-designed randomized, controlled trials. Box 1 summarizes the data for BMP use in open tibia fractures.

In addition to open tibia fractures, some investigators also evaluated the effects of BMP on acute closed tibial shaft fractures. A double-blinded, randomized, controlled trial of 369 patients with closed tibia shaft fractures was performed. Patients were randomly assigned to the standard of care (reamed, locked IMN), the standard of care plus 1 mg/mL of rhBMP-2 in an injectable calcium phosphate matrix (CPM), the standard of care plus 2 mg/mL of rhBMP-2 with CPM, and the standard of care with CPM only. The main outcome measures were time to union and time to return of normal function. The study was terminated early after 180 patients underwent interim analysis at 6 months. No difference in healing was seen between the groups in terms of healing or return to function; however, a significantly higher number of severe complications, edema, and deep venous thrombosis was seen in the treatment groups when compared with the standard of care. Of note, the rates of infection and compartment syndrome were comparable among the groups and with previously published data.

The authors concluded that time to union and return to function were not significantly reduced by adding BMP-2/CPM when treating patients with closed tibial shaft fractures with reamed IMN. The possibility of an injectable adjuvant that can assist with fracture healing is a desirable concept. Unfortunately, at this time, there is limited clinical evidence to suggest benefit in its current form.

Nonunion

Autologous bone grafting has traditionally been used to assist with healing when treating nonunions because it provides not only structural support but also osteoinductive and osteogenic factors to assist with healing. Additional benefits include limited immunogenicity and possibly decreased monetary cost, albeit with the expense of possible donor site morbidity. Although autograft has shown benefits in the treatment of bony deficits, it has come under scrutiny for increasing the risk of complications and donor site pain. Consequently, the idea of an implantable osteoinductive growth factor that can avoid donor site morbidity while acting as an adjuvant to osteoconductive fillers has become increasingly popular. This has led to numerous studies looking at BMP and its efficacy when treating nonunions and fractures with bone loss.

Fracture nonunion is often a multifactorial problem that can be difficult to treat. The etiology of many nonunions, particularly oligotrophic or atrophic variants, can be the result of a compromised healing response that prevents bony union. Additionally, many nonunions can result from significant bone loss at the time of injury. BMPs have been studied in each of these scenarios.

Friedlaender and colleagues studied the efficacy of rhBMP-7 (OP1) in the treatment of tibial nonunions. One hundred twenty-two patients with tibia nonunions were treated with reamed IMN and randomly assigned to autograft versus OP1 implantation at the nonunion site. Similar rates of healing were seen between the 2 groups. Additionally, a trend toward lower infection rates was seen in the rhBMP-7 group. Twenty percent of the autograft cohort had continued pain at the donor site. The authors concluded that rhBMP-7 is a safe and reasonable alternative to autologous bone grafting when treating tibial nonunions without the risk of donor site pain.

An additional multicenter study by Kanakaris and colleagues evaluated 68 patients with aseptic tibial nonunions treated with 3.5 mg of rhBMP-7 mixed with 1 g of collagen carrier. Surgeons were permitted to use autograft (36%) if deemed necessary to assist as a “graft expander.” They achieved a 90% union rate and had no complications or adverse events associated with BMP application.

Similar studies continue to show promising results with long bone nonunions. Unfortunately, when comparing these studies, there is variability in study design and inclusion criteria. Additionally, the types of nonunions treated, along with the specific definitions for nonunion, is inconsistent, making these studies difficult to compare.

A meta-analysis presented in 2014 stated that overall union rates were similar between BMP-treated and non–BMP-treated patients with nonunions. However, when looking specifically at rhBMP-7 use in tibial nonunions, the only FDA-approved indication, there was a significantly higher union rate in the BMP group. Additionally, a decrease in secondary interventions was seen in nonunions treated with BMP when compared with control. However, the analysis showed higher levels of significance in lower-level studies when compared with level 1 studies. This finding likely indicates confounding factors introduced during the lower-level studies, further blurring the findings.