Autograft is still regarded as the gold standard in bone-grafting, despite its well-appreciated limitations. The iliac crest has been the primary source of autologous bone, but other sites, such as the tibial metaphyses and the calcaneus, have also been commonly used. The techniques of autograft harvesting have recently evolved to include reamer-irrigator-aspirator (RIA) harvesting. RIA-harvested autograft can be used as stand-alone graft material or, for enhanced structural integrity, in combination with firmly packed cancellous graft. Both cancellous and RIA-harvested grafts have the distinction of being osteogenic, osteoinductive, and osteoconductive, all in their physiologic ratio. However, a need for additional surgical exposure, limited availability, and potential risks associated with graft harvesting have prompted a search for suitable autograft alternatives.
The advent of recombinant human bone morphogenetic proteins (rhBMPs) generated optimism that the pursuit for an optimal alternative to autogenous bone graft had, at last, successfully ended. Bone morphogenetic proteins (BMPs) proved to be an extremely potent autograft substitute as evidenced by numerous preclinical and clinical studies. Although cleared by the U.S. Food and Drug Administration (FDA) for the highly selected orthopaedic indications, BMPs (rhBMP-2, rhBMP-7) have been extensively used off-label by many clinicians. However, the prevalent and serious adverse events associated with the on and off-label use of rhBMPs call for a revisiting of its clinical safety and efficacy1. The quest for an optimal bone graft alternative is, again, back on track.
Recombinant human platelet-derived growth factor BB (rhPDGF-BB) has been introduced as another potent biologic graft substitute to stimulate cell proliferation, recruitment, and revascularization. Platelet-derived growth factor (PDGF) natively occurs as a dimeric polypeptide with A, B, C, or D chains, and their arrangements lead to five PDGF isoforms, of which PDGF-BB homodimer is the most potent2. Many human cells naturally produce PDGF, but its major source is activated platelets. PDGF physiologically increases DNA synthesis, cell replication, collagen synthesis, and cross-linking as determined by numerous in vitro and in vivo studies. Its stimulatory effects on bone healing are likely due to the recruitment and increased proliferation of mesenchymal stem cells and enhanced angiogenesis.
However, the PDGF release and increased activity have been implicated in the etiology of atherosclerosis as well as malignancies (gliomas, sarcomas, breast cancer)3. Indeed, PDGF exhibits extremely potent mitogenic, chemotactic, and angiogenic activity, especially on fibroblasts and smooth muscle, endothelial, and glial cells, and a dysregulation of autocrine and paracrine PDGF pathways can lead to malignant transformation, tumor therapeutic resistance, and increased risk of metastasis.
Clinically, rhPDGF-BB as becaplermin (Regranex Gel, Healthpoint Biotherapeutics) is approved by the FDA as a topical agent delivered in a hydrogel for the treatment of refractory leg and foot cutaneous wounds, especially diabetic neuropathic ulcers. Another rhPDGF-BB formulation, GEM 21S (Osteohealth), has also gained FDA approval for clinical use. GEM 21S combines rhPDGF-BB and beta-tricalcium phosphate (β-TCP) for the treatment of periodontal defects and a few other dental surgical procedures. Although the initial clinical experience has favored the efficacies of both Regranex Gel and GEM 21S, the safety has been criticized mainly because of a suspected increased risk of cancer mortality among patients who received three or more Regranex Gel treatments.
DiGiovanni et al. report a prospective, multicenter, 2:1 randomized, controlled, blinded trial determining if non-inferiority exists in clinical efficacy and safety between rhPDGF-BB plus β-TCP (the same as in GEM 21S) composite graft substitute (Augment Bone Graft, BioMimetic Therapeutics) and autogenous bone graft for hindfoot or ankle arthrodesis. The study involved thirty-seven clinical sites across the United States and Canada, included 397 patients requiring bone-grafting, and constituted a pivotal trial seeking market approval for Augment. Within a 10% non-inferiority margin, the authors demonstrated significant non-inferiority in the clinical efficacy for rhPDGF-BB/β-TCP compared with autograft. For the primary outcome of the fusion rate determined by computed tomography at twenty-four weeks, the rhPDGF-BB/β-TCP group was non-inferior to the autograft group. In the full-complement analysis (p = 0.038), the fusion rate was 61.2% for patients in the rhPDGF-BB/β-TCP group and 62.0% for patients in the autograft group. In the all-joints analysis (p < 0.001), the fusion rate was 66.5% in the rhPDGF-BB/β-TCP group and 62.6% in the autograft group. Furthermore, non-inferiority was also established for the secondary outcomes (radiographic, clinical, functional, and quality of life) in fourteen of the sixteen end-point assessments at twenty-four weeks and in fifteen of the sixteen end-point assessments at fifty-two weeks. The treatment success rate at twelve months was 86.2% for patients in the rhPDGF-BB/β-TCP group compared with 87.6% for patients in the autograft group as determined by the analysis of the full complement and 88.3% in the rhPDGF-BB/β-TCP group compared with 87.2% in the autograft group as determined by the all-joints analysis. The application of rhPDGF-BB in the study was considered safe as determined by the lack of unusual serious adverse events and the absence of anti-PDGF-BB antibodies. There were five reported malignancy-related serious adverse events, and the overall cancer incidence was 1.1% in the rhPDGF-BB/β-TCP group compared with 1.4% in the autograft group.
The article by DiGiovanni et al. is a very well-designed, well-executed, and well-described Level-I study establishing non-inferiority between rhPDGF-BB/β-TCP and autograft for ankle and hindfoot fusion in both primary and secondary study outcomes. Unlike superiority studies designed to identify a treatment with a better outcome and to determine the extent of that improvement, this non-inferiority study demonstrates that rhPDGF-BB/β-TCP is at least as effective as autograft. Although advantages may exist for rhPDGF-BB/β-TCP over autograft, such as no need for graft harvesting and a more favorable cost-to-benefit ratio, the study with such a design would not be able to detect them.
The FDA recognizes non-inferiority (not worse) and equivalence (not worse and not better) as standard clinical trial methods for obtaining clearance and/or approval for marketing of new devices or drugs. Conventional superiority trials, although very prevalent and convincing in comparing clinical efficacies, may not always be feasible. This is especially the case when the current treatment standard produces good results, so that it becomes extremely difficult to improve its primary outcome, or the extent of that improvement is expected to be marginal. The merits of non-inferiority trials exist if there is potential for improvement or advantages in the secondary outcome measures between the treatment options4. Non-inferiority trials are more complex to design, to conduct, and to interpret than superiority trials. The critical aspects include determining the sample size, selecting the control to establish inferiority, and setting up the non-inferiority margin5. Therefore, assessing the efficacy of rhPDGF-BB/β-TCP is very suitable for an inferiority trial, and all of the critical parameters associated with that trial design are well planned and selected.
In recent years, many clinicians have considered allograft rather than autograft the new gold standard for bone-grafting. This transition has evolved because the cancellous allograft can now be used in conjunction with demineralized bone matrix as an osteoconductive-osteoinductive alternative bone graft option6. An attractive next step in assessing the clinical efficacy of rhPDGF-BB/β-TCP as a bone graft alternative would therefore be to establish its non-inferiority compared with allograft (e.g., cancellous croutons combined with demineralized bone matrix) as both of these autograft contenders do not have any issues associated with harvesting. The future will show if the results of the study by DiGiovanni et al. and the subsequent FDA approval of rhPDGF-BB/β-TCP (Augment) for clinical use as a biologically potent bone graft alternative will result in a change on the throne of bone-grafting.