*2.1.1 rhBMP-2 pre-clinical*

*Clinical Implementation of Bone Regeneration and Maintenance*

benefits in their clinical use.

this data to make clinical use decisions.

**2.1 Bone morphogenetic protein: BMP**

genetically engineered version of the natural protein.

manufacturer determines a product meets all of these requirements and follows the appropriate regulations, the manufacturer can place the product on the market by simply notifying the FDA of the intent to do so. There is no premarket review by FDA for safety or effectiveness of such products. Therefore, there are no requirements for preclinical or clinical data. Since most HCT/P products have little to no peer-reviewed human clinical data, the surgeons must extrapolate the likely

Synthetic bone grafts and demineralized bone matrices (DBMs) fall under Class II under Section 510(k) of the Federal Food, Drug and Cosmetic Act. Section 510(k) describes a regulatory process for the clearance of products that have been demonstrated to the FDA's satisfaction to be "substantially equivalent" in safety and effectiveness to another lawfully marketed device when used for the same purpose. Market clearance requires the filing and review of a 510(k) application and a subsequent FDA review. This review is generally based on a single animal study and bench-top testing comparing the subject device with a predicate. Once again, since most 510(k) cleared products have little to no peer-reviewed human clinical data,

Drug-device combination bone grafts are Class III and require an investigational

In 1965, Dr. Marshall R. Urist showed that demineralized bone matrix (DBM) could induce bone formation when implanted under the skin or intramuscular [1]. Urist pioneered the theory that a substance naturally present in bone was responsible for the osteoinductive bone healing activity of DBM (now Class II). He named this substance, the bone inductive principle, bone morphogenetic protein (BMP) and initiated an extensive and difficult search for these active protein molecules [2]. More than two decades later, advances in molecular biology by Dr. John Wozney's research team at the Genetic Institute described the first isolated extraction and recombinant form of BMP-2 in 1988 [3, 4]. BMP-2 and BMP-7 have been shown to have the most bone forming potential out of the 15 BMPs identified and studied to date [5]. The recombinant protein available commercially today is a synthetic,

BMPs are potent bone forming agents in bone regeneration and bone repair activity and are members of the TGF-Beta superfamily of cytokines. BMPs drive mesenchymal stem cells (MSCs) into osteoblastic lineage. These active protein dimeric molecules are osteoinductive and generally require a collagen sponge or ceramic carrier to enhance their handling characteristics. BMPs initiate endochondral bone formation, presumably by stimulating local MSCs and augmenting bone collagen synthesis. The BMP-2 ligand acts as a rigid clamp connecting type I and

device exemption (IDE) clinical trial followed by a premarket approval (PMA) application. The IDE study requirements include strict oversight from the FDA from statistical and protocol design through clinical follow-up, data integrity and analysis. These IDE studies must be prospective, controlled, blinded and statistically-powered prior to the onset. Moreover, the FDA-required outcomes for approval must be stipulated in the clinical design protocol. The review of both the IDE and PMA filings for drug-device combination products involve both the device (CDRH) and drug (CDER) branches of the FDA. The result of this rigorous process is the highest quality Level I human clinical data available. Surgeons may rely upon

surgeons must extrapolate the likely benefits in their clinical use.

**2. Infuse bone graft: bone morphogenetic protein (BMP)**

**50**

The next direction of rhBMP-2 research was to attempt to define proper dosing of the potent protein and to determine if this dosage would be specific to site or to carrier/scaffold. Formative work in this inquiry was performed in two different models by Sandhu et al. [7, 8].

The investigators first attempted to characterize the dose-response relationship of rhBMP-2 in a canine intertransverse spine fusion model. They compared increasing logarithmic rhBMP-2 doses (58, 115, 230, 460, and 920 mg) on a porous polylactic acid polymer carrier. Successful fusions postoperatively at 3 months were shown throughout this dosing range. A prior study done by this same research team demonstrated superiority of a 2300 mg rhBMP-2 dose to autologous iliac crest bone graft (ICBG) using this same technique. Quality differences above a threshold dose were not reflected in the mechanical, radiographic, or histologic features in the canine intertransverse spine fusion model from a 40-fold variation in rhBMP-2 dose [7].

Learning from the canine work, the investigators continued their work in an ovine lumbar interbody fusion model in conjunction with a cylindrical fenestrated titanium interbody fusion cage (INTER FIX, Medtronic Sofomor Danek, Inc., Memphis, TN). The cage was filled with rhBMP-2-collagen or ICBG (control). The sheep all appeared radiographically fused at 6 months. However, the histologic evaluation revealed that 33% (2/6) of the control group were fused as compared with 100% (6/6) of the rhBMP-2 group (*P* < 0.001). The scar involving the control group was 16-fold more than that seen with the rhBMP-2 group (*P* < 0.01) [8].

A rhesus monkey nonhuman primate model with rhBMP-2 on a collagen carrier within a titanium cage (Sofomor Danek, Memphis, TN) was the subsequent evolutionary step from the ovine model [9]. As optimal dosing for rhesus monkeys had not been previously established, three concentrations of rhBMP-2 [0.00 mg/ mL sham (buffer only), 0.75 mg/mL low dose, or 1.50 mg/mL high dose) were tested. The results demonstrated that both the investigational rhBMP-2 groups achieved arthrodesis at 6 months histologically as compared to the sham group. As before with the ovine model, the blinded nonhuman primate model radiographic assessment was suboptimal but the sagittal CT assessment was consistent with the histology. The higher dose rhBMP-2 (1.5 mg/mL) caused faster and denser bone formation; this study established the dose used in the upcoming US IDE trial.

Fusion environments differ. A cage environment places the contained collagen carrier under protected direct compression forces between large vascularized opposing bony vertebral endplate surfaces. Posterolateral fusion presents a difficult environment with limited surface area and an intertransverse process fusion gap under distraction forces. Moreover, the surrounding muscle envelope applies mechanical compression on the graft material and may contribute to a pseudarthrosis or an hourglass configured fusion.

A standard compression resistant carrier with concomitant dosing concentration for rhBMP-2 was deemed necessary. Boden et al. studied a ceramic carrier [60% hydroxyapatite and 40% tricalcium phosphate (TCP)] in a nonhuman primate laminectomy model. Concentrations of rhBMP-2 (0, 6, 9, or 12 mg) were compared to ICBG. Fusion occurred with each rhBMP-2 carrier including the 0 mg/rhBMP-2 ceramic carrier alone group [10]. No significant overgrowth occurred involving the

thecal sac, as bone growth induction was confined to ceramic carrier. The ceramic carrier was judged to be satisfactory for posterolateral application.

## *2.1.2 rhBMP-2 clinical trials*

The "elimination" of the gold standard, autologous iliac crest bone graft (ICBG) harvest, in lumbar fusion was scientifically proposed by Burkus et al. in 2002 [11]. Anterior lumbar interbody fusion (ALIF) utilizing a combination of rhBMP-2 (1.50 mg/mL concentration) on an absorbable collagen sponge (ACS) carrier-filled tapered titanium fusion cage was shown to have an equivalent (~90%) radiographic (X-ray and CT) fusion rate to that of ICBG. This 2 year, multicenter, prospective, randomized, nonblinded human trial design compared an investigational Class III drug/device combination product (Infuse®, Medtronic Sofamor Danek, Memphis, Tennessee) on a Type-1 ACS (143 patients) to a control ICBG (136 patients); all 273 patients received the same tapered titanium fusion cage (LT-cage, Medtronic Sofomor Danek, Memphis, Tennessee). Osseous fusion rate was confirmed in 94.5% of the investigational group versus 88.7% of the control group at the two-year follow up. Outcomes of particular surgical interest including operative time (1.6 h), estimated blood loss (109.8 mL), adverse events from iliac crest harvest (0%), reported bone graft site discomfort (0%), and bone graft site appearance complaint (0%) were all less in the investigational group as compared to the control group at 2.0 h, 153.1 mL, 5.9, 32, and 16%, respectively. FDA approval was granted on July 2, 2002 for the rhBMP-2/ACS combination product in conjunction with the tapered titanium fusion cage (Infuse®/LT-Cage or Medtronic Sofamor Danek, Memphis, Tennessee) in the treatment of lumbar degenerative disc disease.

A traditional posterolateral spine fusion application (PLF) was the next step in the evaluation of rhBMP-2 on a ceramic granule carrier (60% hydroxyapatite and 40% TCP) in humans as a forged extrapolation of the seminal nonhuman primate work by Boden et al. [12]. Randomization of 25 patients (whose spondylolisthesis was ≤ Grade 1) into one of three groups was performed; five patients (control group) received autograft PLF with pedicle screw instrumentation, 11 patients received rhBMP-2 PLF with pedicle screw instrumentation, and 9 patients received rhBMP-2 PLF in situ only. A 20 mg rhBMP-2 dose was evenly divided in a bilateral, posterolateral application in those patients receiving rhBMP-2 on a ceramic granule carrier. Load bearing through the hardware until osseous fusion ensued was revelatory in the Oswestry scores. Oswestry scores demonstrated significant improvement in the rhBMP-2 PLF in situ only group at 6 weeks, rhBMP-2 PLF with pedicle screw instrumentation group at 3 months, and control, autograft PLF with pedicle screw instrumentation group, at 6 months. The fusion rate with the combined rhBMP-2 PLF groups (in situ only or with pedicle screw instrumentation) was 100% (20/20) and with the control, autograft PLF with pedicle screw instrumentation, however, was 40% (2/5). The radiographic fusion rate for the combined rhBMP-2 PLF groups was statistically significantly higher than for the control (*P* = 0.004).

A trauma application was then explored in a prospective, controlled, randomized multicenter clinical trial evaluation of patients with open tibial fractures. All 450 patients received intermedullary nail stabilization. Patients were randomized equally (*n* = 150), dividing them among one of three treatments: a standard of care (control group) or alternatively two different concentrations of rh-BMP [(0.75 mg/mL, total dose 6 mg) or (1.5 mg/mL, total dose 12 mg) respectively)] on ACS carrier. The control group standard of care was defined, for purposes of this study, as routine soft tissue management. The specific key measure outcome in the study was defined by the proportion of patients for whom secondary intervention was required due to delayed union or nonunion within the index postoperative year.

**53**

*Class III Spine Grafts*

*DOI: http://dx.doi.org/10.5772/intechopen.87706*

The 1.50 mg/mL group demonstrated a 44% reduction in the risk of failure requiring secondary intervention because of delayed union. The 1.50 mg/mL rhBMP-2 group had both significantly accelerated fracture healing and wound healing, higher osseous union rates, significantly fewer secondary interventions, less hardware failure, and less infections (Gustilo-Anderson type III associated injuries). Govender et al. further concluded that the 1.50 mg/mL rhBMP-2 concentration treatment was significantly superior care to the control, standard of care [13]. FDA approval for Infuse® (rhBMP-2 and ACS) in conjunction with an intermedullary nail for acute, open tibial fracture treatment was issued on April 30, 2004.

Between 2005 and 2009, three journal articles were published on the results of the FDA approval studies on 2-stage maxillary sinus floor augmentation [14–16]. Boyne et al. evaluated two concentrations of rhBMP-2/ACS at 0.75 and 1.50 mg/mL versus bone graft control; this pilot study was the first randomized controlled trial (RCT) demonstrating safe de novo bone induction by a recombinant human protein, rhBMP-2. Core biopsies retrieved after subsequent dental implant restoration confirmed normal bone formation in all groups; the proportion of dental implants that remained functionally loaded at 36 months was 62, 67, and 76% in the control group, 0.75 mg/mL rhBMP-2/ACS group, and 1.5 mg/mL rhBMP-2/ACS group, respectively. Triplett et al. performed a pivotal, multicenter, prospective, randomized, parallel evaluation of two treatments for a 2-stage maxillary sinus floor augmentation comparing a 1.50 mg/mL rhBMP-2/ACS group with an autograft control group; this study demonstrated no rhBMP-2 related adverse events at 6 months after dental restoration and similarly effective functional loading performance in both groups. Fiorellini et al. performed a randomized, masked, placebo-controlled, multicenter clinical trial evaluating de novo bone formation for dental implant restoration following tooth extraction using 0.75 mg/mL rhBMP-2/ACS, 1.5 mg/mL rhBMP-2/ACS, placebo control (ACS alone), or no treatment control. The 1.50 mg/ mL rhBMP-2/ACS group demonstrated significantly greater (twice as great) bone augmentation compared to both controls (*P* ≤ 0.05). Furthermore, bone density and histology disclosed no difference between newly induced and native bone. A compression resistant matrix consisting of bovine collagen and Betatricalcium phosphate-hydroxyapatite in conjunction with rhBMP-2 was next compared to ICBG (control). A prospective, randomized, multi-center trial comparing the clinical and radiographic outcomes of an investigational optimized rhBMP-2 formulation to ICBG in one level instrumented traditional PLF in 463 patients with symptomatic degenerative disc disease (DDD) with spondylolisthesis ≤ Grade 1 [17]. Osseous fusion rate was radiographically (X-ray and CT) confirmed in 96% of the investigational group versus 89% of the control group at a 2 year follow up (*p* = 0.014). Outcomes of particular surgical interest including operative time (2.5 h), estimated blood loss (343.1 mL), reported donor site morbidity (0%), failures because of nonunion (six patients), and number requiring secondary surgeries (20 patients) were all significantly less in the investigational group as compared to the control group at 2.9 h, 448.6 mL, 60%, 18 patients (*p* = 0.011), and 36 patients (*p* = 0.015), respectively. The investigators concluded that clinical outcomes were similar between groups; they further concluded that morbidity was eliminated with the use of the optimized 2 mg/mL rhBMP-2 concentration in the compression resistant matrix. Eight patients (3.3%) with cancer (basal cell carcinoma, lung, lymphoma, ovarian, pancreatic, prostate, squamous cell carcinoma, and vocal cord) were reported in the optimized rhBMP-2 matrix group as compared to two patients (0.9%) with cancer (colon and lymphoma) in the control group. The fourfold increase in cancer in the optimized rhBMP-2 matrix group was not reported as possible device-related adverse events, as the cancer types were heterogeneous and statistically nonsignificant (*p* = 0.107). A nonapproval letter was received by

#### *Class III Spine Grafts DOI: http://dx.doi.org/10.5772/intechopen.87706*

*Clinical Implementation of Bone Regeneration and Maintenance*

*2.1.2 rhBMP-2 clinical trials*

carrier was judged to be satisfactory for posterolateral application.

Tennessee) in the treatment of lumbar degenerative disc disease.

was statistically significantly higher than for the control (*P* = 0.004).

A trauma application was then explored in a prospective, controlled, randomized multicenter clinical trial evaluation of patients with open tibial fractures. All 450 patients received intermedullary nail stabilization. Patients were randomized equally (*n* = 150), dividing them among one of three treatments: a standard of care (control group) or alternatively two different concentrations of rh-BMP [(0.75 mg/mL, total dose 6 mg) or (1.5 mg/mL, total dose 12 mg) respectively)] on ACS carrier. The control group standard of care was defined, for purposes of this study, as routine soft tissue management. The specific key measure outcome in the study was defined by the proportion of patients for whom secondary intervention was required due to delayed union or nonunion within the index postoperative year.

A traditional posterolateral spine fusion application (PLF) was the next step in the evaluation of rhBMP-2 on a ceramic granule carrier (60% hydroxyapatite and 40% TCP) in humans as a forged extrapolation of the seminal nonhuman primate work by Boden et al. [12]. Randomization of 25 patients (whose spondylolisthesis was ≤ Grade 1) into one of three groups was performed; five patients (control group) received autograft PLF with pedicle screw instrumentation, 11 patients received rhBMP-2 PLF with pedicle screw instrumentation, and 9 patients received rhBMP-2 PLF in situ only. A 20 mg rhBMP-2 dose was evenly divided in a bilateral, posterolateral application in those patients receiving rhBMP-2 on a ceramic granule carrier. Load bearing through the hardware until osseous fusion ensued was revelatory in the Oswestry scores. Oswestry scores demonstrated significant improvement in the rhBMP-2 PLF in situ only group at 6 weeks, rhBMP-2 PLF with pedicle screw instrumentation group at 3 months, and control, autograft PLF with pedicle screw instrumentation group, at 6 months. The fusion rate with the combined rhBMP-2 PLF groups (in situ only or with pedicle screw instrumentation) was 100% (20/20) and with the control, autograft PLF with pedicle screw instrumentation, however, was 40% (2/5). The radiographic fusion rate for the combined rhBMP-2 PLF groups

thecal sac, as bone growth induction was confined to ceramic carrier. The ceramic

The "elimination" of the gold standard, autologous iliac crest bone graft (ICBG) harvest, in lumbar fusion was scientifically proposed by Burkus et al. in 2002 [11]. Anterior lumbar interbody fusion (ALIF) utilizing a combination of rhBMP-2 (1.50 mg/mL concentration) on an absorbable collagen sponge (ACS) carrier-filled tapered titanium fusion cage was shown to have an equivalent (~90%) radiographic (X-ray and CT) fusion rate to that of ICBG. This 2 year, multicenter, prospective, randomized, nonblinded human trial design compared an investigational Class III drug/device combination product (Infuse®, Medtronic Sofamor Danek, Memphis, Tennessee) on a Type-1 ACS (143 patients) to a control ICBG (136 patients); all 273 patients received the same tapered titanium fusion cage (LT-cage, Medtronic Sofomor Danek, Memphis, Tennessee). Osseous fusion rate was confirmed in 94.5% of the investigational group versus 88.7% of the control group at the two-year follow up. Outcomes of particular surgical interest including operative time (1.6 h), estimated blood loss (109.8 mL), adverse events from iliac crest harvest (0%), reported bone graft site discomfort (0%), and bone graft site appearance complaint (0%) were all less in the investigational group as compared to the control group at 2.0 h, 153.1 mL, 5.9, 32, and 16%, respectively. FDA approval was granted on July 2, 2002 for the rhBMP-2/ACS combination product in conjunction with the tapered titanium fusion cage (Infuse®/LT-Cage or Medtronic Sofamor Danek, Memphis,

**52**

The 1.50 mg/mL group demonstrated a 44% reduction in the risk of failure requiring secondary intervention because of delayed union. The 1.50 mg/mL rhBMP-2 group had both significantly accelerated fracture healing and wound healing, higher osseous union rates, significantly fewer secondary interventions, less hardware failure, and less infections (Gustilo-Anderson type III associated injuries). Govender et al. further concluded that the 1.50 mg/mL rhBMP-2 concentration treatment was significantly superior care to the control, standard of care [13]. FDA approval for Infuse® (rhBMP-2 and ACS) in conjunction with an intermedullary nail for acute, open tibial fracture treatment was issued on April 30, 2004.

Between 2005 and 2009, three journal articles were published on the results of the FDA approval studies on 2-stage maxillary sinus floor augmentation [14–16]. Boyne et al. evaluated two concentrations of rhBMP-2/ACS at 0.75 and 1.50 mg/mL versus bone graft control; this pilot study was the first randomized controlled trial (RCT) demonstrating safe de novo bone induction by a recombinant human protein, rhBMP-2. Core biopsies retrieved after subsequent dental implant restoration confirmed normal bone formation in all groups; the proportion of dental implants that remained functionally loaded at 36 months was 62, 67, and 76% in the control group, 0.75 mg/mL rhBMP-2/ACS group, and 1.5 mg/mL rhBMP-2/ACS group, respectively. Triplett et al. performed a pivotal, multicenter, prospective, randomized, parallel evaluation of two treatments for a 2-stage maxillary sinus floor augmentation comparing a 1.50 mg/mL rhBMP-2/ACS group with an autograft control group; this study demonstrated no rhBMP-2 related adverse events at 6 months after dental restoration and similarly effective functional loading performance in both groups. Fiorellini et al. performed a randomized, masked, placebo-controlled, multicenter clinical trial evaluating de novo bone formation for dental implant restoration following tooth extraction using 0.75 mg/mL rhBMP-2/ACS, 1.5 mg/mL rhBMP-2/ACS, placebo control (ACS alone), or no treatment control. The 1.50 mg/ mL rhBMP-2/ACS group demonstrated significantly greater (twice as great) bone augmentation compared to both controls (*P* ≤ 0.05). Furthermore, bone density and histology disclosed no difference between newly induced and native bone.

A compression resistant matrix consisting of bovine collagen and Betatricalcium phosphate-hydroxyapatite in conjunction with rhBMP-2 was next compared to ICBG (control). A prospective, randomized, multi-center trial comparing the clinical and radiographic outcomes of an investigational optimized rhBMP-2 formulation to ICBG in one level instrumented traditional PLF in 463 patients with symptomatic degenerative disc disease (DDD) with spondylolisthesis ≤ Grade 1 [17]. Osseous fusion rate was radiographically (X-ray and CT) confirmed in 96% of the investigational group versus 89% of the control group at a 2 year follow up (*p* = 0.014). Outcomes of particular surgical interest including operative time (2.5 h), estimated blood loss (343.1 mL), reported donor site morbidity (0%), failures because of nonunion (six patients), and number requiring secondary surgeries (20 patients) were all significantly less in the investigational group as compared to the control group at 2.9 h, 448.6 mL, 60%, 18 patients (*p* = 0.011), and 36 patients (*p* = 0.015), respectively. The investigators concluded that clinical outcomes were similar between groups; they further concluded that morbidity was eliminated with the use of the optimized 2 mg/mL rhBMP-2 concentration in the compression resistant matrix. Eight patients (3.3%) with cancer (basal cell carcinoma, lung, lymphoma, ovarian, pancreatic, prostate, squamous cell carcinoma, and vocal cord) were reported in the optimized rhBMP-2 matrix group as compared to two patients (0.9%) with cancer (colon and lymphoma) in the control group. The fourfold increase in cancer in the optimized rhBMP-2 matrix group was not reported as possible device-related adverse events, as the cancer types were heterogeneous and statistically nonsignificant (*p* = 0.107). A nonapproval letter was received by

Medtronic on March 9, 2011 regarding the optimized AMPLIFY rhBMP-2 Matrix. The FDA nonapproval stemmed from the fourfold increased cancer risks in the investigational group and was linked to the high dose rhBMP-2 form of AMPLIFY versus the prior approved low dose forms.

Amidst the high-profile controversy, Yale University Open Data Access (YODA) retrieved Medtronic's safety and efficacy data on file in toto. Contract funding support of both the research and preparation of the work was provided by Medtronic to Yale. The Centre for Reviews and Dissemination (CRD) was then commissioned by the YODA initiative in an unprecedented effort by industry to facilitate unbiased review of the relevant benefits and harms of rhBMP-2 as used specifically in spinal fusion surgery; CRD has no direct financial conflict with Medtronic. Two successive publications in the Annals of Internal Medicine were issued in 2013 regarding the findings of the YODA initiative; the dissimilitude between the two publications were the extraction methods and the different studies included [18, 19]. Simmons et al. found rhBMP-2 had increased fusion rates versus ICBG, 12% higher (CI, 2–23%); Fu et al. found similar overall lumbar fusion rates between rhBMP-2 and ICBG Simmons et al. found nonsignificant increased cancer risk after rhBMP-2 (relative risk, 1.98 [CI 0.86–4.54]); Fu et al. found rhBMP-2 at 24 months had increased cancer risk (risk ratio, 3.45 [95% CI, 1.98–6.00]). Fu et al. also found rhBMP-2 to have associated increased risk for wound complications and dysphagia in off-label use in anterior cervical spine surgery, and nonsignificant increased risk for retrograde ejaculation and urogenital problems after on-label ALIF.

The data synthesis from the YODA initiative and more recent publications report mixed findings with regards to rhBMP-2 usage complications and cancer incidences after rhBMP-2 [20–23]. This same data synthesis suggests that an informed public might have benefited from earlier disclosure and blinded outcome assessment in retrospect.
