A well-consolidated fusion mass is fundamental to pain reduction and increased function, which are desired clinical outcomes after a spinal fusion procedure1. The use of bone morphogenetic proteins (BMPs) contributes to rapid and effective fusion2. However, consistent de novo bone formation is still considered a clinical challenge even when BMPs are used3,4. For instance, fusion consolidation is particularly difficult in patients who have had previous surgery at the same anatomical site; those who have multilevel disease, diabetes, metabolic bone disease, or nutritional deficiencies; and those who are elderly. Difficulty using biologics in these patients may be due to an inadequate cellular microenvironment at the implant site or fusion bed, a lack of viable cells, and/or deficient vascularization2,5. Other challenges associated with the use of BMPs in spine fusion include high costs2,6-8 and a dose-dependent risk of adverse events9,10.
Bone marrow has emerged as a potential adjuvant to improve the efficiency and effectiveness of BMP in spine fusion11-13. During surgery, bone marrow is readily available to augment biologic or other grafting materials in situ as autologous bone marrow aspirate (BMA). The stimulation of osteogenesis by bone marrow has been repeatedly demonstrated since 186914. BMA is typically combined with bone-grafting products to increase the likelihood of spinal fusion15. The osteoprogenitor components of BMA, mesenchymal stem cells (MSCs), are essential to the osteogenic activity of BMA16-19. Clinical20,21 and animal12,13,22-25 studies have demonstrated the efficacy of bone marrow-derived MSCs in improving fusion success. Accordingly, many studies have focused on the benefits of manipulated and/or expanded MSCs combined with BMP26-28; however, the time required and the technical and regulatory obstacles of preparing MSCs limit their clinical application16.
In addition to MSCs, freshly harvested and clinically readily available BMA includes osteoinductive factors that may optimize the environment for both MSCs and BMPs to promote fusion29. Muschler et al.30 observed that bone-grafting was most improved when cancellous bone matrix was enriched with osteoblastic progenitor-concentrated cells and bone marrow clot. The synergistic effects of recombinant human BMP-2 (rhBMP-2), fibroblast growth factors, and osteoinductive components in BMA on the osteogenic differentiation of MSCs were observed both in vitro31,32 and in spine fusion21. The results of those studies suggested that BMA may provide a more favorable environment for rhBMP-2-induced bone formation.
The present study was performed to test whether the transplantation of freshly harvested BMA with a low dose of rhBMP-2 optimizes the efficiency of rhBMP-2 for increased efficacy at a reduced dose. Rao et al.13 observed that the posterolateral fusion mass was increased by an efficient dose of rhBMP-2 augmented with processed BMA in a murine model. Improving rhBMP-2 performance through the use of BMA has the potential to lower costs, to decrease adverse effects due to dosing, and to improve the host environment for better outcomes, especially in compromised patients.
The specific objectives of this study were (1) to determine the rhBMP-2/absorbable collagen sponge (ACS) dose-dependent fusion rates following posterolateral spinal fusion procedures in Lewis female rats, finding a suboptimal dose at which only 50% of the rats had fusion, and (2) to evaluate the ability of freshly harvested, unmanipulated bone marrow to enhance posterolateral fusion with a suboptimal concentration of rhBMP-2 at which only 50% of the rats were expected to achieve posterolateral transverse process fusion (effective dose 50% [ED50]).
This study was conducted in three phases; a flowchart is presented in Figure 1.
Phase 1: Selection of Subeffective Dose and rhBMP-2 Dosing (in Vivo Assay Study)
Materials
Clinical grade rhBMP-2 (INFUSE; Medtronic Sofamor Danek, Memphis, Tennessee) in buffer was provided in freeze-dried form. It was reconstituted with sterile water to desired concentrations of 0.0015, 0.003, 0.006, 0.032, and 0.160 of rhBMP-2 mg/mL. Concentrations were validated by means of enzyme-linked immunosorbent assay (ELISA) (R&D Systems, Minneapolis, Minnesota).
An ACS (Helistat; Integra Life Sciences, Plainsboro, New Jersey) was used for delivery. Each ACS was cut into 0.5 × 1 × 0.5-cm sized pieces for all implant conditions.
Implant Conditions
In a separate series of dilution experiments, fifty-three mature female Lewis rats underwent an L4-L5 posterolateral transverse process spinal fusion (see Appendix) with implants containing decreasing concentrations of 0.16, 0.032, 0.006, 0.003, and 0.0015 mg/mL of rhBMP-2 per side (Table I). Rats were killed after eight weeks, and L4-L5 spinal segments were tested by means of manual palpation; radiographic imaging was performed.
Phase 2: Syngeneic Bone-Marrow Transplants and Subeffective Dose of rhBMP-2 (in Vivo Study)
Subjects
Fifty-two female adult (220 to 260-g) Lewis rats (Charles River Laboratories, Wilmington, Massachusetts) were utilized in Phase 2. Of these, twenty-four (n = 18 and n = 6) were hosts, four were controls, and twenty-four served as donors for fresh isogenic BMA transplants (autologous BMA in this rat subspecies may be transplanted without immune rejection of the graft or host disease).
rhBMP-2/ACS
The concentration of 0.006-mg/mL rhBMP-2 (total dose, 1.68 μg/side/220-g rat) was selected as the subeffective concentration as it was the concentration at which 33% to 50% of rats had fusion (ED33 to 50) in the rhBMP-2 dosing-in vivo assay study (described above) (Table I). rhBMP-2 was diluted to the desired concentration of 0.006 mg of rhBMP-2 per mL; at the time of surgery, rhBMP-2 was applied with use of ACS. Samples were maintained from each dilution and were validated with use of ELISA (R&D Systems).
Harvesting of Bone-Marrow Aspirate from Donor Syngeneic Lewis Rats
BMA was freshly harvested from the long bones (the femur and tibia) of separate donor Lewis rats (n = 24) while two groups of recipient rats (n = 18 and n = 6) simultaneously underwent survival surgery to receive the prepared implants (described below).
The donor Lewis rats were killed with an overdose of pentobarbital. The femur and tibia of the hindlimbs were harvested in a sterile fashion. A blunt cut was made on both ends of the long bone shaft, and articular and osseous prominences were removed. A sterile syringe (10 mL) (Tyco Healthcare Group, Mansfield, Massachusetts) with an 18-gauge needle (PrecisionGlide; Becton Dickinson, Franklin Lakes, New Jersey) was filled with air. The tip of the needle was placed within one end of the marrow cavity, and the other end was placed into a sterile receiving vial (1.5 mL) (Cryogenic Vial; Corning, Corning, New York). The air was forced through the cavity, emptying the fresh BMA into the vial; this process was repeated until the cavity was empty. BMA from one femur (i.e., the right femur) and one tibia (i.e., the right tibia) of a donor rat was implanted into the corresponding side of a recipient rat.
Implant Study Conditions
Implant conditions consisted of an isograft of fresh BMA from a Lewis donor rat (simulating an autologous transplant) mixed with 0.006-mg/mL rhBMP-2/two absorbable collagen sponges per side (BMA+rhBMP-2/two ACS) (n = 18). Control conditions consisted of 0.006-mg/mL rhBMP-2/two ACS per side (n = 12) and 0.006-mg/mL rhBMP-2/one ACS per side (n = 12) from the dosing study (Fig. 1) (Table I), BMA in one ACS per side (n = 6), and one ACS alone (n = 4).
Depending on the experimental implant condition, either one or two ACS (0.5 × 1 × 0.5 cm) were placed in the vials of BMA combined with rhBMP-2, BMA only, or rhBMP-2 only. The ACS was soaked until the solution was absorbed (approximately two minutes). For the implantation of BMA+rhBMP-2/ACS, the ACS was saturated to 50% volume with rhBMP-2 and 12% to 20% volume of BMA.
Surgical Technique
The posterolateral intertransverse process fusion surgical procedure is described briefly here; a detailed description of this procedure is provided in the Appendix33-56. With the rat under 2% to 2.5% inhaled isoflurane anesthesia, a posterior midline longitudinal incision was made at the lumbar spine. Dissection was performed to expose the transverse processes of L4 and L5. The transverse processes were decorticated with a high-speed burr until punctate bleeding, whereas the lamina and facet joints were left intact (not decorticated). No internal fixation was used. The surgical site was irrigated with antibiotic Ringer solution and was packed with gauze until the implants were prepared. The surgeon was blinded to the experimental assignment until after surgical preparation.
A 7 to 10-mm space exists between the transverse processes of L4 and L5. The loaded ACS implants were placed bilaterally in the paraspinal muscle bed, between and touching the transverse processes of L4 and L5 (see Appendix). The fascia and skin were closed with 3-0 absorbable suture and 3-0 nylon suture, respectively. The rats were housed in separate cages, were allowed to eat and drink ad libitum, and were managed with subcutaneous injections of buprenorphine (0.5 mg/kg) for the control of perioperative and postoperative pain. The condition of the rats was monitored daily. The rats were killed at eight weeks after surgery.
Manual Palpation Fusion (Main End Point)
Fusion was determined via manual palpation of segments by three independent observers; the association between manual palpation and biomechanical testing has been demonstrated in this model49. Any motion that was detected between the L4 and L5 segment, including the transverse process or vertebral bodies, was considered to indicate a fusion failure38-56. Each side was tested separately; the observation of no motion on both the right and left sides was considered to indicate a successful fusion.
Radiographic Evaluation
Radiographs were made monthly until the animals were killed, and high-resolution radiographs were made after the animals were killed. Radiographs were evaluated for increasing density between the transverse processes of L4 and L5; the continuity of newly formed radiodensity (bone) compared with immediate postoperative radiographs was noted. Radiographic data were utilized as supporting data because the main end point of this study was manual palpation.
Biomechanical Testing
Biomechanical testing was performed on spinal segments that were fused. Each specimen was cleaned of musculature, was potted in custom aluminum cups with use of a two-part epoxy resin on either end of the fusion mass, was mounted onto a servohydraulic actuator (MTS Bionix 370.02 with Spine Subsystem; MTS, Eden Prairie, Minnesota) equipped with a mini load cell (Mini45 Transducer; ATI, Apex, North Carolina), and was tested to failure at an angular deformation rate of 30°/minute while the corresponding applied moment and rotation were recorded. Strength was measured as the maximum torque to failure. Stiffness was calculated as the slope of the linear regression between the initial and maximum applied torque on failure.
Phase 3: Cell Counts of Bone-Marrow Aspirate from Lewis Rats
Subjects
A separate group of eighteen Lewis rats served as donors for the quantitative cell study in BMA from the femur and tibia.
Procedure for Bone-Marrow Aspirate Harvests
Lewis rats were killed with an overdose of pentobarbital, and BMA was harvested from the long bones (the femur and tibia) in the same manner as described above.
Procedure for Cell Counts
BMA from one femur and one tibia of each of eighteen female adult Lewis rats was mixed with heparin (20 units/mL) (Sigma-Aldrich, St. Louis, Missouri). Each sample was carefully laid over a 0.5-mL density gradient solution (Lymphocyte Separation Medium; Mediatech, Herndon, Virginia) in a centrifuge tube (Corning) and was centrifuged at 400 times gravity at room temperature for thirty minutes. The sediment containing erythrocytes was removed, and supernatant containing mononuclear cells was collected and counted with use of a hemacytometer (Hausser Scientific, Horsham, Pennsylvania).
Statistical Evaluation (Phases 1, 2, and 3)
All statistical analyses for Phases 1, 2, and 3 were performed with use of SAS statistical software (SAS Institute, Cary, North Carolina).
Evaluation of Fusion via Manual Palpation Data (Phases 1 and 2)
The Fisher exact test was applied to the variable of manual palpation (fused versus not fused) for comparison among groups. Because of the small sample size, confidence intervals were computed with use of the Exact method. The number of rats needed for each experimental condition was computed with alpha set at 0.05 and with 80% power (an 80% chance of detecting a difference). To test the effect of 0% versus 100% successful fusion by manual palpation, a minimum sample size of approximately four rats was required for each implant condition, yielding a right-sided p value of 0.01.
Biomechanical Data (Phase 2)
Average strength and stiffness data were compared between implant conditions of BMA + 0.006-mg/mL rhBMP-2/ACS and 0.006-mg/mL rhBMP-2/ACS with use of the Student t test. Power was determined in a post hoc manner.
Evaluation of Cell Counts (Phase 3)
Pearson correlation was used to compare the cell counts of BMA from one side of a rat with those from the other side. Analysis of variance (ANOVA) was used to determine if the quantities of BMA were significantly variable among rats (subject/between-rats effect).
Source of Funding
No outside funding was provided for this research. rhBMP-2 was provided gratis by Medtronic Sofamor Danek, Memphis, Tennessee.
Phase 1: Selection of Subeffective Dose and rhBMP-2 Dosing (in Vivo Assay Study)
rhBMP-2/ACS dose-dependent manual palpation fusion rates are presented in Table I. Diluting the concentration of rhBMP-2 from 0.16 mg/mL (fusion rate, six of six [100%; 95% confidence interval (CI), 54% to 100%]; ED100) to 0.032 mg/mL (fusion rates, two of two [100%; 95% CI, 16% to 100%] and nine of nine [100%; 95% CI, 66% to 100%]; ED100) to 0.006 mg/mL reliably yielded a fusion rate of 33% to 50% (four of twelve [33%; 95% CI, 10% to 65%] to six of twelve [50%; 96% CI, 21% to 79%]; ED33 to 50). Although the lowest concentration of 0.0015 mg/mL yielded no fusions (zero of six [0%; 95 CI, 0% to 46%]; ED0), the concentration of 0.006 mg/mL yielded an effective dose or concentration of rhBMP-2/ACS at which 33% to 50% of the rats demonstrated a response in terms of posterolateral fusion (ED33 to 50) by eight weeks; the latter concentration (0.006 mg/mL) was therefore selected as the target dose for investigation with BMA (as described below). Fusion occurred in four of twelve rats that received one ACS per side and in six of twelve rats that received two ACS per side (Table I). The comparison between the use of rhBMP-2 with one ACS per side as compared with two ACS per side was performed because two sponges per side were used when rhBMP-2 was mixed with BMA (experiments described below). There was no significant difference between the use of one and two ACS per side (p = 0.68, Fisher exact test). The concentration of 0.006 mg/mL was selected for further evaluation to test the synergistic effect when BMA is mixed with rhBMP-2.
Phase 2: Syngeneic Bone Marrow Aspirate Transplants and Subeffective Dose of rhBMP-2 (in Vivo Study)
Radiographic Fusion
In a dose-dependent manner, radiographic density was observed between the L4 and L5 transverse processes in sixteen of eighteen rats (89%; 95% CI, 65% to 99%) that received rhBMP-2/two ACS with BMA, in four of twelve rats (33%; 95% CI, 10% to 65%) that received rhBMP-2/one ACS without BMA, and in six of twelve rats (50%; 95% CI, 21% to 79%) that received rhBMP-2/two ACS without BMA (Fig. 2). Radiographic density between L4 and L5 transverse processes was observed in zero of six rats (0%; 95% CI, 0% to 46%) that received BMA/ACS (without rhBMP-2) and in zero of four rats (0%; 95% CI, 0% to 52%) that received ACS alone.
Manual Palpation Fusion
Fusion via manual palpation (no motion) was observed between the L4 and L5 segments at eight weeks in sixteen of eighteen rats (89%; 95% CI, 65% to 99%) that received rhBMP-2/ACS+BMA (Table II). Fusion was observed in four of twelve rats (33%, CI: 21% to 79%) and in six of twelve rats (50%, CI: 21% to 79%) that received 0.006-mg/mL rhBMP-2/ACS without BMA with use of one or two absorbable collagen sponges, respectively, in the dosing study (Fig. 3). Fusion was observed in zero of six rats that received BMA/ACS alone (without rhBMP-2) (0%; 95% CI, 0% to 46%) and in zero of four rats that received ACS alone (0%; 95% CI, 0% to 52%). Thus, the addition of BMA significantly increased the fusion rate to 89% (sixteen of eighteen) compared with the fusion rate of 33% (four of twelve) or 50% (six of twelve) for rats that received the same concentration of rhBMP-2, without fresh BMA, with use of one ACS (p < 0.01, Fisher exact test) or two ACS (p < 0.05, Fisher exact test).
Biomechanical Testing
Ten of the fused segments (including six from rats that received rhBMP-2/ACS+BMA and four from rats that received rhBMP-2/ACS only) were selected for biomechanical testing. There were no significant differences between the fusion masses from rats that received rhBMP-2/ACS+BMA and those that received rhBMP-2/ACS alone in terms of average strength (492.04 ± 92.73 compared with 497.60 ± 56.55 Nmm; p = 0.79) or stiffness (26.04 ± 3.07 compared with 27.96 ± 3.60 Nmm/deg, p = 0.41) (Fig. 4).
Phase 3: Cell Counts in Bone Marrow Aspirate from Independent Group of Lewis Rats
The volume of BMA per side ranged from 30 to 50 μL. An average of 1.23 × 106 bone marrow cells (range, 0.60 to 2.60 × 106 bone marrow cells), without red blood cells, were employed per side. There was no correlation in the number of cells in BMA between right or left sides of the rat (Pearson rho = −0.1). There was no significant rat subject effect (ANOVA).
The goal of the present study was to investigate the use of BMA to increase the biologic efficiency or osteogenic productivity of rhBMP-2 as measured by the fusion rate and biomechanical properties of the resultant osseous mass. To our knowledge, this is the first study to combine fresh, unmanipulated BMA with a subeffective dose of rhBMP-2. The direct application of BMA mixed with an experimentally determined subeffective dose of rhBMP-2 (ED33 to 50) significantly improved the fusion rate to 89%, compared with 33% to 50% without BMA, a twofold improvement. BMA alone in carrier was not sufficient to induce fusion. Also, the frequency of fusion induced by rhBMP-2 was not inhibited by the addition of BMA. These results suggest that the use of BMA together with a subeffective dose of rhBMP-2 increases osteogenesis and improves the frequency of spinal fusion success in rats.
BMA providing cells and other factors appears to increase the rhBMP-2-induced fusion rate and remodeling but not the biomechanical quality of fusion itself.
Rao et al.13, in a similar investigation, found that rhBMP-2 and BMA had a synergistic effect that improved fusion. The primary differences were the use of a high optimal dose of rhBMP-2 instead of an experimentally determined subeffective dose, serially passaged and centrifuged BMA, and a murine model. Even at this high optimal dose, Rao et al. reported a significantly greater fusion area, density, and volume in association with rhBMP-2 with BMA as compared with rhBMP-2 alone. The fusion mass was characterized by a thicker cortical perimeter with more trabeculae and active bone formation. Their quantitative and qualitative data support our results suggesting that the rate of bone formation and remodeling may be increased with BMA in combination with rhBMP-2.
The synergistic interaction between BMA and rhBMP-2 may be an effective mechanism for improving fusion success clinically, as supported by the results of this study. When rhBMP-2 is used in a clinical fusion application, osteoprogenitor cells are recruited from surrounding muscle, from soft tissues at the implant site or fusion muscle bed, or from bleeding bone after decortication57-59. Previous studies involving the use of bone marrow-derived and isolated MSCs have also demonstrated fusion enhancement11-13. The osteogenic action of rhBMP-2 is to differentiate MSCs in the immediate environment to osteoprogenitors as well as the differentiation of preosteoblasts to mature osteoblasts60. Clinically, bone marrow is a readily available source of MSCs. The yield of MSCs in BMA can be quite variable26-28. However, the direct concentration and application of MSCs to BMP may provide optimum synergy. BMA allows direct administration to augment the grafting construct after harvesting in the same patient undergoing a spinal fusion procedure.
The addition of autologous BMA slightly increases the number and types of undifferentiated cell targets in direct proximity to rhBMP-2 at the critical time of implantation, which may initiate and promote the local differentiation process prior to clearance or degradation of rhBMP-261-64.
Furthermore, the addition of autologous BMA also adds a variety of osteoinductive factors29 that have been suggested to further increase the osteogenicity of the fusion site, including endogenous BMPs and other cytokines, growth factors, and signaling regulators.
Adverse effects after the use of exogenous rhBMP-2 clinically are dose-dependent9,10 and are related to the supraphysiologic doses needed because it is administered only once during the surgical procedure for fusion. Higher doses are also often required for fusion performed in certain locations and difficult tissue environments. Furthermore, rhBMPs have a multitude of effects that are not entirely understood, including indirectly inducing bone resorption by osteoclast formation and activity65. The use of rhBMP-2 in spine fusion substantially increases costs2,6-8. Our ability to achieve an increased fusion rate with use of a lower dose of rhBMP-2 with BMA may ultimately make rhBMP-2 a more cost-effective treatment. The use of BMA may increase the fusion efficiency of rhBMP-2 in challenging patients (such as those undergoing multilevel fusion procedures or revision same-site surgery because of fusion failure) or in patients in whom the surrounding fusion site and muscle bed is compromised.
Our conclusions should be interpreted in light of several study limitations. Our study utilized Lewis rats for syngeneic transplant of bone marrow from an identical twin (isogenic transplant). The iliac crest is not a viable source of autogenous bone marrow harvest for survival fusion surgery in rats because of the small size and cortical nature of bone. Syngeneic transplant in the Lewis rat subspecies simulates autogenous/autologous bone marrow transplantation because Lewis rat subspecies are considered an identical gene match for marrow transplants; thus, transplants herein are modeled to be similar to surgical procedures that may be performed for spine fusion in patients. No host disease, adverse reactions, or signs of rejection were observed because of the syngeneic qualities of these rats, although this was not specifically studied. Samples from donor rats that were not implanted but were evaluated for cell number demonstrated considerable intra-rat variability (when comparing aspirate from the right and left sides of same rat) and inter-rat variability (when comparing aspirate from different rats). This variability would exist in the clinical situation as well.
Because of the use of freshly harvested and directly applied BMA, our study method did not allow us to quantitatively evaluate the composition of the actual BMA implanted. Therefore, we were unable to determine which specific components within the harvested BMA were responsible for the increased rate of fusion and interpreted our results on the basis of previous studies that have quantitatively assessed the composition of BMA26-29.
Last, during biomechanical testing, we assumed that the geometry and microstructure of the fusion masses were relatively homogeneous, allowing comparison of material properties. A more qualitative assessment of the fusion mass is necessary to validate this assumption. In addition, the testing was limited by the small number of specimens available for mechanical testing, giving us an estimated power of 0.80 to detect a difference of 25% in stiffness and 30% in strength between treatment groups (α = 0.05). Future studies with an expanded number of animals per treatment group are needed to strengthen the validity of these results. Nonetheless, our results are consistent with those of others who have found increased rates of fusion and increased size of the fusion mass without demonstrating significant changes in the intrinsic properties of the bone material once sufficient fusion has been achieved11,12.
In conclusion, BMA may provide a readily available, safe, and cost-effective means to increase the osteogenic potency and biologic efficiency of rhBMP-2. This study provides evidence that the use of BMA in conjunction with rhBMP-2 may allow for the reduction of the effective dose of rhBMP-2 or to expand its use in patients presenting with a surgically challenging tissue environment. The direct application of BMA avoids the potential risks associated with cultured MSCs or gene therapy and is already widely used clinically for related applications. The observed osteogenic potency and biologic efficiency of rhBMP-2 combined with BMA warrants further investigation to establish feasibility for clinical applications.
Disclosure: None of the authors received payments or services, either directly or indirectly (i.e., via his or her institution), from a third party in support of any aspect of this work. One or more of the authors, or his or her institution, has had a financial relationship, in the thirty-six months prior to submission of this work, with an entity in the biomedical arena that could be perceived to influence or have the potential to influence what is written in this work. Also, one or more of the authors has had another relationship, or has engaged in another activity, that could be perceived to influence or have the potential to influence what is written in this work. The complete Disclosures of Potential Conflicts of Interest submitted by authors are always provided with the online version of the article.