Abstract
Background: Distraction osteogenesis creates a challenging bone-healing environment with protracted demand for cells of the osteoblast lineage. Platelet-derived growth factor-BB (PDGF-BB) is an osteoblast mitogen and chemotaxin that has been shown to accelerate and/or enhance bone-healing in several preclinical studies. The purpose of the present study was to determine whether recombinant human platelet-derived growth factor-BB (rhPDGF-BB) would have a similar effect on regenerate healing after distraction osteogenesis.
Methods: Unilateral 7-mm mid-diaphyseal femoral lengthening procedures were performed in eighty-three male Sprague-Dawley rats that were separated into five experimental groups. During the distraction period (Days 7 to 28), each animal received a weekly 50-µL injection of either sodium acetate buffer, bovine collagen dissolved in sodium acetate buffer, or one of three concentrations of rhPDGF-BB (100, 300, or 1000 µg/mL) into the distraction site. Animals from each group were killed on Days 35, 42, 49, 56, and 63. Healing was assessed with biweekly serial radiographs, micro-computed tomography of the explanted bones, and histologic analysis.
Results: rhPDGF-BB treatment significantly increased new-bone formation at the midconsolidation time points (Days 42, 49, and 56) as well as the union rate. On Day 49 regenerate bone volume was significantly greater in each of the three rhPDGF-BB-treated groups than in the controls (p < 0.05, p = 0.0002, and p < 0.05 for the 100, 300, and 1000 µg/mL rhPDGF-BB groups, respectively), whereas on Day 42 regenerate bone volume was significantly greater in the 300 and 1000 µg/mL rhPDGF-BB groups than in the controls (p = 0.0002 and p < 0.05, respectively) and on Day 56 regenerate bone volume was significantly greater in the 100 and 300 µg/mL rhPDGF-BB groups than in the controls (p < 0.05 and p < 0.0001, respectively). The overall union rate was 40.4% (nineteen of forty-seven) in the rhPDGF-BB-treated animals, compared with 4.5% (one of twenty-two) in the controls (p = 0.01). The radiographic and histologic results were consistent with new-bone formation as quantified by micro-computed tomography, although they were less definitive.
Conclusions: The administration of exogenous rhPDGF-BB into the distraction site during diaphyseal distraction enhanced bone-healing in a rat model of distraction osteogenesis as evidenced by both increased regenerate new-bone formation and a higher union rate.
Clinical Relevance: The ability of rhPDGF-BB to enhance healing in this model suggests that it may be able to shorten treatment time and to decrease the nonunion rate in the challenging healing environment created during distraction osteogenesis.
Distraction osteogenesis is a technique that is used to stimulate and prolong bone regeneration through the incremental distraction of immature fracture callus, providing a means of bridging what would otherwise be a large osseous defect. It is used to treat a variety of challenging musculoskeletal conditions, such as congenital and acquired limb-length discrepancy, segmental bone loss, deformity and fracture nonunion, and chronic medical conditions such as osteomyelitis1-3.
Despite its efficacy, a major disadvantage of distraction osteogenesis is the protracted nature of the treatment process. For example, the average time to complete an uncomplicated 5-cm femoral lengthening in a twenty to twenty-nine-year-old patient is seven months (including latency, distraction, and consolidation)4. During this time, the patient must contend with the inconvenience of a bulky external fixator and chronic pain as well as the risks of infection, delayed consolidation, and nonunion5. Accordingly, several groups have explored strategies to reduce treatment time by hastening the maturation of the regenerate with use of biophysical stimulation6-9 or biological intervention10.
As the cellular and molecular events in fracture-healing have become increasingly well-characterized, interest has grown in the use of growth factors to augment normal11-16 and compromised17-19 bone repair. Recombinant versions of two members of the transforming growth factor-beta (TGF-ß) superfamily of growth factors, bone morphogenetic protein-2 (rhBMP-2) and osteogenic protein-1 (rhOP-1, also known as rhBMP-7), have received approval from the United States Food and Drug Administration for limited use in trauma, spine, and maxillofacial (rhBMP-2 only) applications20. Additionally, recombinant platelet-derived growth factor-BB homodimer (rhPDGF-BB) has been approved for the treatment of periodontal bone loss20. While there have been a few studies on the use of growth factors to accelerate ossification of the regenerate after distraction osteogenesis10,21,22, to date none of the attempts have moved beyond preclinical testing in animal models.
The platelet-derived growth factors (PDGFs) are potent mesenchymal cell chemotaxins and mitogens that are released from degranulating platelets in the fracture hematoma and also are synthesized by inflammatory and reparative cells23. PDGF was originally identified as an essential component for the culture of serum-dependent cells24. Now considered to be a family of five heterodimeric and homodimeric proteins (PDGF-AB, PDGF-AA, PDGF-BB, PDGF-CC, and PDGF-DD), PDGFs act by binding with the tyrosine kinase receptors PDGF-aa, PDGF-ßß, and PDGF-aß25.
The response to PDGF depends on the isoform of PDGF delivered, the type of target cell, and the specific cell-surface receptor expressed on the target cell25. At a wound site, PDGF attracts neutrophils and macrophages and stimulates macrophages to release additional growth factors that are important for wound-healing26. PDGF receptors have been found on all of the connective-tissue cells associated with bone-healing, including fibroblasts, vascular smooth-muscle cells, osteoblasts, and chondrocytes26,27. PDGF induces fibroblast proliferation, is mitogenic28 and chemotactic29,30 for osteoblasts, and is proliferative for chondrocytes31.
Preclinical animal studies have demonstrated that PDGF-BB has a stimulatory influence on bone formation. In perhaps the most comprehensive study to date, systemic administration of rhPDGF-BB to ovariectomized rats for six weeks resulted in increased bone density and strength throughout the skeleton32. Other studies have suggested that rhPDGF-BB can accelerate bone-healing. rhPDGF-BB released from chitosan/tricalcium phosphate sponges has been shown to induce the filling of critical-sized rat calvarial defects33, whereas rhPDGF-BB, delivered in collagen or ß-TCP-collagen matrices, has been shown to stimulate long-bone-healing in rabbits34 and geriatric osteoporotic rats35. rhPDGF-BB has also been shown to increase the restoration of bone lost as a result of advanced periodontal disease in humans36. To our knowledge, however, there have been no studies of the effects of rhPDGF on healing during distraction osteogenesis, a particularly challenging environment that might benefit from the prolonged recruitment of bone-forming cells.
The present study was performed to evaluate the effect of exogenous rhPDGF-BB on new-bone formation during distraction osteogenesis. Our hypotheses were that regenerate new-bone formation would be increased at distraction sites treated with rhPDGF-BB as compared with sites treated with collagen or acetate buffer alone and that new-bone formation would increase with increasing rhPDGF-BB doses.
Following approval by our Institutional Animal Care and Use Committee, eighty-three six-month-old male Sprague-Dawley rats (mean weight, 425 ± 35 g) were separated into five experimental groups, including two negative control groups and three rhPDGF-BB-treated groups. The negative control groups included seventeen animals treated with acetate buffer and sixteen animals treated with bovine collagen (dissolved in acetate buffer). The rhPDGF-BB-treated groups included sixteen animals treated with 100 µg/mL rhPDGF-BB, seventeen animals treated with 300 µg/mL rhPDGF-BB, and seventeen animals treated with 1000 µg/mL rhPDGF-BB. In all cases, the rhPDGF-BB was delivered in bovine collagen dissolved in acetate buffer.
Our experimental design called for fifteen animals in each group, with tissues from three animals per group being analyzed at each of five time points. Additional animals were included to provide replacements for animals that might be excluded perioperatively because of surgical or anesthetic complications. The three doses of rhPDGF-BB (100, 300, and 1000 µg/mL) were selected because similar doses had been used with success in a previous rat fracture model35.
The rhPDGF-BB was delivered in a viscous solution of Type-I bovine collagen dissolved in 20-mM sodium acetate buffer (pH 6.0). Collagen was selected as a carrier because it is biocompatible, it is easily injected through a narrow-bone syringe, and it functions as a scaffold for cell ingrowth. The collagen was provided as a lyophilized brick (Kensey Nash, Exton, Pennsylvania) that was hydrated with the rhPDGF-BB/acetate buffer solution or acetate buffer alone, and the resulting mixture was loaded into a 1.0-mL syringe for injection.
A custom-made four-pin monolateral fixator was applied to the right femur of each animal, followed by a periosteal-sparing mid-diaphyseal osteotomy. The wounds were closed in layers, and the animals were returned to their cages and were allowed unrestricted weight-bearing. A single dose of cefazolin (20 mg/kg, administered intramuscularly) was used preoperatively for antibiotic prophylaxis, and buprenorphine (0.3 mg/kg, administered subcutaneously two times per day on postoperative Day 1, and 0.2 mg/kg, administered subcutaneously two times per day on postoperative Day 2) was used for postoperative pain control. Low-dose oral antibiotics were given in the drinking water to reduce the risk of pin-track infections (cephalexin [0.23 mg/mL] for four weeks, alternating with sulfamethoxazole [0.83 mg/mL]/trimethoprim [0.17 mg/mL] for four weeks). Distraction was initiated on postoperative Day 7 and involved two 0.17-mm lengthenings per day for twenty-one days, for a total lengthening of 7 mm6,37-39.
On postoperative Days 7, 14, 21, and 28, 50 µL of the rhPDGF-BB or control solution was injected directly into the distraction gaps (yielding 5, 15, and 50 µg of rhPDGF-BB per injection for the 100, 300, and 1000-µg/mL treatment groups, respectively). Healing was followed with use of biweekly plain radiographs (Fig. 1, A and B) made with a high-resolution cabinet x-ray system (MX-20; Faxitron X-Ray, Lincolnshire, Illinois). At the conclusion of the study, all of the radiographs from a given time point (e.g., Day 42) were ranked from least to most healed by two independent reviewers (M.G.E. and D.C.M.) who were blinded to treatment.
A minimum of three animals from each group were scheduled to be killed on Days 35, 42, 49, 56, and 63. After the animals were killed, the femora were removed en bloc and were placed in 10% formalin. High-resolution three-dimensional images (isometric voxel size, 16 µm) of a 16.5-mm region of each femoral diaphysis were generated with a desktop micro-computed tomography system (µCT 40; SCANCO Medical AG, Brüttisellen, Switzerland). The 16.5-mm scan region was selected to include the 7-mm distraction site and ~4.75 mm of the adjacent proximal and distal fragments. Once acquired, the original gray-value images were segmented with use of a low-pass noise-reducing filter (s = 1, support = 1.0) and fixed threshold to extract only mineralized tissue (bone), and volume renderings were generated for visualization (Fig. 1, C). Regenerate new-bone formation (BV) and bone volume fraction (BV/TV, where TV represents the total tissue volume) were then calculated on the basis of a 6.4-mm (400-slice) segment of the thresholded image centered in the distraction gap (Fig. 1, D). Union was assessed by inspection of the original volume renderings for clear evidence of bone bridging.
After scanning, all of the bones were demineralized and embedded in paraffin and 6-µm midsagittal sections were cut and stained with hematoxylin and eosin, Mason trichrome, and safranin-O/fast green. Healing of the regenerate was graded by two blinded reviewers (M.G.E. and D.C.M.) using a 6-point grading scale: 0 (no new bone), 1 (new bone only at the cut bone ends), 2 (new bone bridging one-third of the distraction gap or less), 3 (new bone bridging more than one-third to two-thirds of the distraction gap), 4 (new bone bridging the entire gap but not united), or 5 (clearly united and remodeling).
Statistical Analysis
All data were analyzed with use of SAS Software (Version 9.1.3; SAS Institute, Cary, North Carolina). Mixed linear models were fit with use of residual estimation of maximum likelihood (REML, PROC MIXED). Each model included effects for group, time, and the interaction of group by time. Orthogonal contrasts were used to compare the treatment groups at each time point (our a priori hypotheses) and to test for trends in new-bone formation as a function of rhPDGF-BB dose (e.g., 0, 100, 300, and 1000 µg/mL). Separate models were used for the comparison of new-bone formation (BV), bone volume fraction (BV/TV), and histologic grade. Post hoc comparisons were adjusted with use of the Holm test (sequential Bonferroni) with alpha maintained at 0.05. Prior to analysis, the data on new-bone formation (BV) were logarithmically transformed to reduce positive skewness (p = 0.28 for the Shapiro-Wilk test of deviation from normality). Group means and standard error limits from the mixed models were used for graphical presentation. Intraclass correlation coefficients were calculated for the x-ray ranking and histologic grading data to assess inter-rater reliability40. There was good inter-rater reliability for the histologic grading (fixed set intraclass correlation coefficient = 0.86), so for these data the means of the raters' gradings were analyzed. The reliability of the radiographic rankings was poorer (intraclass correlation coefficient = 0.559), so analysis of these rankings included a term for rater and its interaction with group. Analysis of the radiographic rankings was limited to Days 28 and 42 because of sample-size limitations on Days 56 and 63. Time was not included as a factor as the scale of the rankings was influenced by the numbers of rankings (sample size).
Source of Funding
Funding for this study was provided by BioMimetic Therapeutics. The funds were used to pay for all aspects of study, including the animals, housing, supplies, and salaries. The bovine collagen used in this study was provided by Kensey Nash Corporation.
Seventy-two animals survived the study and were available for analysis (n = 15 for the 100-µg/mL group, n = 16 for the 300-µg/mL group, n = 16 for the 1000-µg/mL group, n = 14 for the buffer control group, and n = 11 for the collagen control group). Six animals were excluded because of intraoperative surgical complications. In addition, five of the control animals (including two in the buffer control group and three in the collagen control group) that were operated on early in the study died as a result of gastrointestinal problems, which were attributed to the postoperative analgesic and antibiotics (Table I). Consequently, the buprenorphine dose was reduced (to 0.03 mg/kg, administered twice per day for two days) and the prophylactic antibiotics in the drinking water were eliminated for the remaining animals (including six animals in the collagen control group and all of the animals in the rhPDGF-BB-treated groups). Three additional samples from the collagen control group were unavailable for micro-computed tomographic analysis because they were processed for histologic analysis before being scanned (Table I).
The serial radiographs revealed new-bone formation in each experimental group at every time point. Typically, periosteal and regenerate new bone was visible by the time distraction was completed (Day 28), with progressively more new bone forming centripetally from the gap ends throughout consolidation (Fig. 2). Although there was low concordance in the rankings of our two radiographic reviewers (intraclass correlation coefficient = 0.55), our mixed linear model demonstrated that the reviewers did not rank the treatment groups differently (p = 0.9965). This made differences difficult to detect (because of increased variability and therefore reduced power), but it also provided a measure of confidence that any detected differences were real. Despite the decreased power, we found significant differences on Days 28 and 42. On Day 28, the healing in the acetate buffer control group was ranked significantly lower than that in all other treatment groups, whereas the healing in the 300-µg/mL group was ranked significantly higher than that in the 1000-µg/mL group and the collagen (and buffer) controls (p < 0.05 for all). On Day 42, the 300-µg/mL group ranked significantly higher than the buffer control group (p < 0.05). None of the other comparisons reached the level of significance.
Analysis of the bone volume and bone volume fraction data revealed significant differences among the experimental groups (p < 0.0001 for both) and evaluation time points (p < 0.0001 for both) as well as significant interactions between experimental group and evaluation time point (p = 0.0395 and 0.0022, respectively). New-bone formation was lowest in the buffer and collagen control groups; however, attrition was an issue in the collagen control group (as noted above) (Table I). Therefore, the buffer and collagen groups were equally weighted and were compared as a combined control with the rhPDGF-BB treatment groups using complex orthogonal constructs. This technique minimized the influence of chance differences between the buffer and collagen controls rather than weighting the buffer controls more heavily, which would have occurred if the animals from the two groups had been simply combined. The rationale for combining the groups was that they were both negative controls and were expected to perform similarly. However, they were combined only after they were specifically compared at each time point (pairwise) and overall (simple effect). These comparisons revealed no significant differences in bone volume or bone volume fraction between the buffer and collagen controls.
In general, bone volume (BV) and bone volume fraction (BV/TV) were greater in the rhPDGF-BB-treated animals than in the combined buffer and collagen controls, and bone volume tended to increase with increasing rhPDGF-BB dose (Figs. 3 and 4). Bone volume in the 100-µg/mL rhPDGF-BB group was significantly greater than that in the combined controls on Days 49 and 56 (p < 0.05 for both), and bone volume in the 1000-µg/mL rhPDGF-BB group was significantly greater than in the controls on Days 42 and 49 (p < 0.05 for both). Bone volume in the 300-µg/mL rhPDGF-BB group was significantly greater than in the controls on Days 42, 49, and 56 (p = 0.0002, p = 0.0002, and p < 0.0001, respectively). Similarly, bone volume fraction in the 100-µg/mL rhPDGF-BB group was significantly greater than that in the combined controls on Day 49 (p = 0.0009), bone volume fraction in the 1000-µg/mL rhPDGF-BB group was significantly greater than that in the controls on Days 42 and 49 (p = 0.0019 and p < 0.0001, respectively), and bone volume fraction in the 300 µg/mL rhPDGF-BB group was significantly greater than that in the controls on Days 42, 49, and 56 (p = 0.0007, p < 0.0001, and p < 0.0001, respectively). There were no significant treatment-related differences in bone volume or bone volume fraction on Days 35 or 63, nor were there any significant differences between rhPDGF-BB doses at any time point. However, the trend analysis did reveal significant positive relationships between both bone volume and bone volume fraction and rhPDGF-BB dose on Days 42, 49, and 56 (p < 0.001 for both).
Inspection of the reconstructed micro-computed tomography images suggested that there was a general increase in the rate of bone bridging in the rhPDGF-BB-treated animals (Table II), but with the small number of samples available in each group, none of the comparisons according to treatment or time reached the level of significance. However, when the buffer and collagen controls were combined and the effects of time and rhPDGF-BB dose were ignored, the overall union rate for the rhPDGF-BB-treated animals at the time of death was significantly greater than that for the combined controls (40.43% compared with 4.55%; p = 0.0127, chi-square test).
Histologically, healing on Day 35 was similar in all groups and was typified by the presence of periosteal new bone, which merged with regenerate bone emanating from the cut bone ends and extended into the distraction gap. Small areas of bone bridged the distraction gap in a few specimens, whereas in most specimens the gap was filled with centrally located regions of longitudinally oriented, dense, highly cellular fibrous connective tissue separating regions of intramembranous woven bone. In the majority of sections, there were also small foci (~5% of the regenerate area) of safranin-O-stained cartilage tissue. Large populations of osteoblasts covered the bone surfaces at the fibrous tissue-bone interface. Subjectively, however, there were no gross differences among the five treatment groups.
At the midconsolidation time points (defined as Days 42, 49, and 56), the amount of new bone in the distraction gap appeared to increase similarly in all groups, whereas the amount of fibrous tissue in the gap was much lower in the rhPDGF-BB-treated animals than it was in the buffer or collagen controls (Figs. 5-A and 5-B). The one exception was the animals in Group 5 (1000 µg/mL) on Day 56, which also retained a fair amount of fibrous tissue. By Day 49, the regenerate in the rhPDGF-BB-treated animals emulated a nearly united fracture, with a narrow region of cartilage at the center of the distraction gap separating the newly formed bone on the proximal and distal fragments (Fig. 6). In contrast, the cut bone ends in the buffer and collagen controls were separated by several millimeters of fibrous tissue. By Day 63, the amount of bone and fibrous tissue in the control animals was similar to that in the rhPDGF-BB-treated animals, although there was still some cartilage at the midpoint of the distraction gap in fifteen of the sixteen bones analyzed at that time point. We did not notice any gross differences in cellularity—or shifts in the relative amounts of different cell types—between the rhPDGF-BB-treated and control animals.
In general, the blinded histologic grading data were consistent with the micro-computed tomography data in that new-bone formation increased with time and the average grade for the rhPDGF-BB-treated animals was higher than that for the buffer and collagen controls (Table III). Analysis of the blinded histologic grading data revealed significant main effects associated with experimental group (p < 0.0001) and evaluation time point (p < 0.005), although the interaction between group and time did not quite reach the level of significance (p = 0.052). The histologic grading of the collagen and buffer controls was not significantly different on Days 42, 49, 56, and 63. Therefore, for these time points, the collagen and buffer control groups were equally weighted and combined as a single control. On Day 35, when the data for the two control groups were not combined, the histologic appearance of the slides from the 300-µg/mL rhPDGF-BB and collagen control groups was graded higher than that of the slides from the 1000-µg/mL rhPDGF-BB group (p < 0.05 for both). At the other time points, the only significant differences were that the slides from the 300-µg/mL rhPDGF-BB group were graded higher than those of the combined controls on Days 42 and 56 (p < 0.05 for both).
Distraction osteogenesis creates an environment of protracted demand for new bone-forming cells. In contrast to the healing of uncomplicated fractures, which is largely completed within six weeks, active new-bone formation must continue for months during distraction osteogenesis. During that time, bone-forming cells must be recruited and expanded to meet the need for bone regeneration at the distraction site. Our rationale for treatment with rhPDGF-BB was that sustained recruitment by repeated stimulation of inflammatory cells and osteoblasts would hasten and/or increase new-bone formation, ultimately leading to more rapid union. We selected PDGF because it is mitogenic and chemotactic for cells in the osteoblastic lineage41-43, and we selected PDGF-BB in particular because it has been used with success in several dental44 and orthopaedic applications32-35. We chose to deliver it during distraction because distraction is a period of increased demand for reparative cells and, we reasoned, because the populations of receptive cells would be highest (as PDGF is normally released early in the wound-healing process).
Our results are generally consistent with the two previous studies of rhPDGF-BB augmentation of fracture-healing in long bones. In the first, by Nash et al., tibial osteotomies in seven rabbits were treated with an intramedullary injection of 80 µg of rhPDGF-BB delivered in 0.15 mL of bovine collagen, and seven control animals were treated with 0.15 mL of bovine collagen alone34. At four weeks, the authors found more robust callus formation (on radiographs), increased callus stiffness and strength, and more mature callus formation in the rhPDGF-BB-treated animals as compared with the collagen-treated controls. More recently, Hollinger et al. reported positive results after treating tibial osteotomy sites in geriatric osteoporotic rats with rhPDGF-BB in an injectable matrix of ß-tricalcium phosphate and bovine collagen35. Specifically, the authors reported that the torsional strength of rhPDGF-BB-treated tibiae was indistinguishable from the strength of the contralateral controls at five weeks, whereas the non-rhPDGF-BB-treated tibiae remained weaker. Our results are also consistent with the study by Lee et al., which showed that rhPDGF-BB increased new-bone formation in a rat calvarial defect model33.
We observed the largest treatment-related increases in bone volume and bone volume fraction on postoperative Days 42, 49, and 56 in our model. This finding was somewhat surprising, given that our injections were performed on Days 7, 14, 21, and 28. We anticipated a more rapid effect, given that most (40% to 80%) of the PDGF-BB delivered in collagen matrices is burst-released33,45 and is cleared from plasma with a half-life of less than two minutes24. In retrospect, however, the delay is not unreasonable given that PDGF-BB is mitogenic, angiogenic, and chemotactic. It would take time for the attracted new cells to migrate into and proliferate at the distraction site, and in some cases, to differentiate into mature osteoblasts. It is also possible that there may have been an anabolic effect due to a chronic low-level release of the remaining rhPDGF-BB as the collagen matrix was progressively degraded. Interestingly, Hollinger et al.35 also reported a response only at the later time point in their fracture-healing study, with no effect at three weeks postoperatively, followed by discernable differences at five weeks.
In addition to finding increased midconsolidation new-bone formation (on Days 42, 44, and 56) in association with rhPDGF-BB administration, our histologic findings also suggest that the regenerate matured more quickly in the treated animals. By Day 49, the distraction gap in the rhPDGF-BB-treated animals was largely filled with newly formed and reorganizing woven bone, with only a thin strip of cartilage separating the proximal and distal fragments. In contrast, in both the buffer and collagen control groups at Day 49, there was still a substantial amount of fibrous tissue in the distraction zone. As healing and maturation of the regenerate is a cell-mediated process, we believe that the difference was due to sustained recruitment of the cells associated with bone repair through rhPDGF-BB injection. Our periodic injections repeatedly reintroduced a potent wound-healing chemokine and mitogen that has a direct stimulatory effect at numerous points along the wound-healing cascade46. We believe that this stimulation offset the challenging, prolonged wound-healing environment seen with distraction osteogenesis. Our examination of the histologic slides did not reveal any obvious cellular differences between the controls and the rhPDGF-BB-treated animals. However, the present study was not originally designed to generate these data. Additional work will be required to clarify how rhPDGF-BB upregulates new-bone formation during distraction osteogenesis.
Our secondary objective was to assess the impact of increasing doses of rhPDGF-BB on the timing and extent of regenerate new-bone formation. Our data indicate that in the range of doses evaluated in the present study (100 to 1000 µg/mL), new-bone formation increases with increasing rhPDGF-BB dose. We found that mean new-bone formation (BV) in the gap increased with increasing rhPDGF-BB concentration on Days 42 and 49, and for two of the three concentrations (100 and 300 µg/mL) on Day 56. Although none of the comparisons between the rhPDGF-BB dosages reached the level of significance because of small individual sample sizes, our trend analysis revealed that new-bone formation was indeed dose-dependent. We cannot explain the sharp decrease in new-bone formation on Day 56 in the 1000-µg/mL group, especially since bone volume increased again on Day 63. At this point, we suspect that the data from this group and time point are outliers as they are inconsistent with both the Day 42 and Day 63 results for the same treatment group as well as the ranking of the three rhPDGF-BB treatment groups on Days 42 and 49. It is possible that it simply reflects missed dosing, although that is unlikely as all of the animals in each dose group (for example, 1000 µg/mL) were injected on the same days with PDGF-BB/collagen from the same batches.
Our experiment was designed to simultaneously generate data on PDGF-BB treatment as well as on dose and postoperative time point. Despite the use of eighty-three animals, the initial size of each of our individual experimental groups was only three or four animals, which made the study vulnerable to dropout. This was mostly an issue in our collagen control group, in which five animals were lost perioperatively and there were no micro-computed tomography data for three specimens. However, by collapsing the buffer and collagen groups into a single combined control we were able to make comparisons with the rhPDGF-BB groups that minimized the risk of biasing to one control group or the other. We justified combining the groups on the grounds that they were both a priori negative controls, and we did so only after we confirmed that their means were similar (and not significantly different) at every time point.
As we interpret our data, we must acknowledge the possibility that our results were influenced by reduction of the analgesic dose and elimination of the oral antibiotic for the animals in the rhPDGF-BB groups once a problem had been detected in the control groups. However, we think it is unlikely that they were, for several reasons. First, the analgesic was only administered two days postoperatively for all animals, so there is little chance that it would have affected bone formation on Days 28 through 63. In addition, the animals in the control groups that were affected were excluded from analysis. Finally, the average weights of the animals at the time of death were similar in each group.
Our findings add to the small but growing literature on the positive effect of PDGF-BB in bone-healing applications33-36. Our findings provide encouraging evidence that there may be therapeutic potential for rhPDGF-BB in the challenging environment created by distraction osteogenesis. 
Aldegheri R. Distraction osteogenesis for lengthening of the tibia in patients who have limb-length discrepancy or short stature. J Bone Joint Surg Am.1999;81:624-34.81624Â
1999Â
[PubMed] Â
Maffuli N, Fixsen JA. Distraction osteogenesis in congenital limb length discrepancy: a review. J R Coll Surg Edinb.1996;41:258-64.41258Â
1996Â
Â
Perry CR. Bone repair techniques, bone graft, and bone graft substitutes. Clin Orthop Relat Res.1999;360:71-86.36071Â
1999Â
[CrossRef] Â
Fischgrund J, Paley D, Suter C. Variables affecting time to bone healing during limb lengthening. Clin Orthop Relat Res.1994;301:31-7.30131Â
1994Â
Â
Paley D. Problems, obstacles, and complications of limb lengthening by the Ilizarov technique. Clin Orthop Relat Res.1990;250:81-104.25081Â
1990Â
Â
Eberson CP, Hogan KA, Moore DC, Ehrlich MG. Effect of low-intensity ultrasound stimulation on consolidation of the regenerate zone in a rat model of distraction osteogenesis. J Pediatr Orthop.2003;23:46-51.2346Â
2003Â
[CrossRef] Â
Eyres KS, Saleh M, Kanis JA. Effect of pulsed electromagnetic fields on bone formation and bone loss during limb lengthening. Bone.1996;18:505-9.18505Â
1996Â
[CrossRef] Â
Pepper JR, Herbert MA, Anderson JR, Bobechko WP. Effect of capacitive coupled electrical stimulation on regenerate bone. J Orthop Res.1996;14:296-302.14296Â
1996Â
[CrossRef] Â
van Roermund PM, ter Haar Romeny BM, Hoekstra A, Schoonderwoert GJ, Brandt CJ, van der Steen SP, Roelofs JM, Scholten F, Visser WJ, Renooij W. Bone growth and remodeling after distraction epiphysiolysis of the proximal tibia of the rabbit. Effect of electromagnetic stimulation. Clin Orthop Relat Res.1991;266:304-12.266304Â
1991Â
Â
Mizumoto Y, Moseley T, Drews M, Cooper VN 3rd, Reddi AH. Acceleration of regenerate ossification during distraction osteogenesis with recombinant human bone morphogenetic protein-7. J Bone Joint Surg Am.2003;85 Suppl 3:124-30.85124Â
2003Â
Â
Bostrom MP, Saleh KJ, Einhorn TA. Osteoinductive growth factors in preclinical fracture and long bone defects models. Orthop Clin North Am.1999;30:647-58.30647Â
1999Â
[CrossRef] Â
Einhorn TA. Enhancement of fracture-healing. J Bone Joint Surg Am.1995;77:940-56.77940Â
1995Â
Â
Einhorn TA, Trippel SB. Growth factor treatment of fractures. Instr Course Lect.1997;46:483-6.46483Â
1997Â
Â
Gamradt SC, Lieberman JR. Genetic modification of stem cells to enhance bone repair. Ann Biomed Eng.2004;32:136-47.32136Â
2004Â
[CrossRef] Â
Lieberman JR, Daluiski A, Einhorn TA. The role of growth factors in the repair of bone. Biology and clinical applications. J Bone Joint Surg Am.2002;84:1032-44.841032Â
2002Â
Â
Lind M. Growth factor stimulation of bone healing. Effects on osteoblasts, osteomies, and implants fixation. Acta Orthop Scand Suppl.1998;283:2-37.2832Â
1998Â
Â
Friedlaender GE, Perry CR, Cole JD, Cook SD, Cierny G, Muschler GF, Zych GA, Calhoun JH, LaForte AJ, Yin S. Osteogenic protein-1 (bone morphogenetic protein-7) in the treatment of tibial nonunions. J Bone Joint Surg Am.2001;83 Suppl 1(Pt 2):S151-8.83S151Â
2001Â
Â
Johnson EE, Urist MR, Finerman GA. Repair of segmental defects of the tibia with cancellous bone grafts augmented with human bone morphogenetic protein. A preliminary report. Clin Orthop Relat Res.1988;236:249-57.236249Â
1988Â
Â
Lee FY, Sinicropi SM, Lee FS, Vitale MG, Roye DP Jr, Choi IH. Treatment of congenital pseudarthrosis of the tibia with recombinant human bone morphogenetic protein-7 (rhBMP-7). A report of five cases. J Bone Joint Surg Am.2006;88:627-33.88627Â
2006Â
[CrossRef] Â
Center for Devices and Radiological Health, US Food and Drug Administration. Humanitarian use devices and new device approval. . 2008. Accessed 2009 June 17.Â
2008Â
www.fda.gov Â
Hamdy RC, Amako M, Beckman L, Kawaguchi M, Rauch F, Lauzier D, Steffen T. Effects of osteogenic protein-1 on distraction osteogenesis in rabbits. Bone.2003;33:248-55.33248Â
2003Â
[CrossRef] Â
Sciadini MF, Dawson JM, Banit D, Juliao SF, Johnson KD, Lennington WJ, Schwartz HS. Growth factor modulation of distraction osteogenesis in a segmental defect model. Clin Orthop Relat Res.2000;381:266-77.381266Â
2000Â
[CrossRef] Â
Andrew JG, Hoyland JA, Freemont AJ, Marsh DR. Platelet-derived growth factor expression in normally healing human fractures. Bone.1995;16:455-60.16455Â
1995Â
Â
Ross R. Platelet-derived growth factor. Lancet.1989;1:1179-82.11179Â
1989Â
Â
Alvarez RH, Kantarjian HM, Cortes JE. Biology of platelet-derived growth factor and its involvement in disease. Mayo Clin Proc.2006;81:1241-57.811241Â
2006Â
[CrossRef] Â
Heldin CH, Westermark B. Mechanism of action and in vivo role of platelet-derived growth factor. Physiol Rev.1999;79:1283-316.791283Â
1999Â
Â
Fujii H, Kitazawa R, Maeda S, Mizuno K, Kitazawa S. Expression of platelet-derived growth factor proteins and their receptor alpha and beta mRNAs during fracture healing in the normal mouse. Histochem Cell Biol.1999;112:131-8.112131Â
1999Â
[CrossRef] Â
Gruber R, Karreth F, Frommlet F, Fischer MB, Watzek G. Platelets are mitogenic for periosteum-derived cells. J Orthop Res.2003;21:941-8.21941Â
2003Â
[CrossRef] Â
Fiedler J, Etzel N, Brenner RE. To go or not to go: migration of human mesenchymal progenitor cells stimulated by isoforms of PDGF. J Cell Biochem.2004;93:990-8.93990Â
2004Â
[CrossRef] Â
Lind M, Deleuran B, Thestrup-Pedersen K, Søballe K, Eriksen EF, Bünger C. Chemotaxis of human osteoblasts. Effects of osteotropic growth factors. APMIS.1995;103:140-6.103140Â
1995Â
Â
Guerne PA, Sublet A, Lotz M. Growth factor responsiveness of human articular chondrocytes: distinct profiles in primary chondrocytes, subcultured chondrocytes, and fibroblasts. J Cell Physiol.1994;158:476-84.158476Â
1994Â
[CrossRef] Â
Mitlak BH, Finkelman RD, Hill EL, Li J, Martin B, Smith T, D'Andrea M, Antoniades HN, Lynch SE. The effect of systemically administered PDGF-BB on the rodent skeleton. J Bone Miner Res.1996;11:238-47.11238Â
1996Â
[CrossRef] Â
Lee YM, Park YJ, Lee SJ, Ku Y, Han SB, Klokkevold PR, Chung CP. The bone regenerative effect of platelet-derived growth factor-BB delivered with a chitosan/tricalcium phosphate sponge carrier. J Periodontol.2000;71:418-24.71418Â
2000Â
[CrossRef] Â
Nash TJ, Howlett CR, Martin C, Steele J, Johnson KA, Hicklin DJ. Effect of platelet-derived growth factor on tibial osteotomies in rabbits. Bone.1994;15:203-8.15203Â
1994Â
[CrossRef] Â
Hollinger JO, Onikepe AO, MacKrell J, Einhorn T, Bradica G, Lynch S, Hart CE. Accelerated fracture healing in the geriatric, osteoporotic rat with recombinant human platelet-derived growth factor-BB and an injectable beta-tricalcium phosphate/collagen matrix. J Orthop Res.2008;26:83-90.2683Â
2008Â
[CrossRef] Â
Nevins M, Giannoble WV, McGuire MK, Kao RT, Mellonig JT, Hinrichs JE, McAllister BS, Murphy KS, McClain PK, Nevins ML, Paquette DW, han TJ, Reddy MS, Lavin PT, Genco RJ, Lynch S. Platelet-derived growth factor stimulates bone fill and rate of attachment level gain: results of a large multicenter randomized controlled trial. J Periodontol.2005;76:2205-15.762205Â
2005Â
[CrossRef] Â
Jazrawi LM, Majeska RJ, Klein ML, Kagel E, Stromberg L, Einhorn TA. Bone and cartilage formation in an experimental model of distraction osteogenesis. J Orthop Trauma.1998;12:111-6.12111Â
1998Â
[CrossRef] Â
Moore DC, Leblanc CW, Müller R, Crisco JJ 3rd, Ehrlich MG. Physiologic weight-bearing increases new vessel formation during distraction osteogenesis: a micro-tomographic imaging study. J Orthop Res.2003;21:489-96.21489Â
2003Â
[CrossRef] Â
Pacicca DM, Moore DC, Ehrlich MG. Physiologic weight-bearing and consolidation of new bone in a rat model of distraction osteogenesis. J Pediatr Orthop.2002;22:652-9.22652Â
2002Â
[CrossRef] Â
Canalis E, McCarthy TL, Centrella M. Effects of platelet-derived growth factor on bone formation in vitro. J Cell Physiol.1989;140:530-7.140530Â
1989Â
[CrossRef] Â
Hock JM, Canalis E. Platelet-derived growth factor enhances bone cell replication, but not differentiated function of osteoblasts. Endocrinology.1994;134:1423-8.1341423Â
1994Â
[CrossRef] Â
Piché JE, Graves DT. Study of the growth factor requirements of human bone-derived cells: a comparison with human fibroblasts. Bone.1989;10:131-8.10131Â
1989Â
[CrossRef] Â
Lynch SE, Wisner-Lynch L, Nevins M, Nevins ML. A new era in periodontal and periimplant regeneration: use of growth-factor enhanced matrices incorporating rhPDGF. Compend Contin Educ Dent.2006;27:672-8.27672Â
2006Â
Â
Hart CE, BioMimetic Therapeutics. Personal communication; 2007.Â
2007Â
Â
Hollinger JO, Hart CE, Hirsch SN, Lynch S, Friedlaender GE. Recombinant human platelet-derived growth factor: biology and clinical applications. J Bone Joint Surg Am.2008;90 Suppl 1:48-54.9048Â
2008Â
[CrossRef] Â