Several growth and differentiation factors have shown potential as therapeutic agents to support periodontal wound healing/regeneration, although optimal dosage, release kinetics, and suitable delivery systems are still unknown. Experimental variables, including delivery systems, dose, and the common use of poorly characterized preclinical models, make it difficult to discern the genuine efficacy of each of these factors. Only a few growth and differentiation factors have reached clinical evaluation. It appears that well-defined discriminating preclinical models followed by well-designed clinical trials are needed to further investigate the true potential of these and other candidate factors. Thus, current research is focused on finding relevant growth and differentiation factors, optimal dosages, and the best approaches for delivery to develop clinically meaningful therapies in patient-centered settings.
The periodontium encompasses the alveolar bone, root cementum, periodontal ligament (PDL), and gingiva, which are the tissues surrounding and supporting the teeth. Periodontitis is an inflammatory disease characterized by destruction of the alveolar bone, root cementum, PDL, and gingiva as a response to insults elicited by microbial accumulations on tooth surfaces. Regeneration is defined as the reproduction or reconstitution of a lost or injured part, with form and function of lost structures restored. Periodontal regeneration includes regeneration of alveolar bone, cementum, PDL, and gingiva.
Periodontal defects may be morphologically characterized as suprabony or intrabony, as furcation or gingival recession defects, or their combinations. Defect configuration appears to be one important factor that may predict outcomes of periodontal reconstructive and regenerative procedures. The spatial distribution of vascular and cellular tissue elements circumscribing the defect plays an important role for the healing of any periodontal defect. These tissue elements, which have periodontal or alveolar origin, are dramatically reduced in two- and one-wall intrabony defects, class II and III furcation defects, and in supra-alveolar defects. Accordingly, the number of bone walls and adjoining periodontal tissues appears to be a critical determinant to treatment outcomes in clinical settings. This means that supra-alveolar periodontal defects have significantly reduced potential for regeneration.
Animal models are useful for observing the nature of wound healing following periodontal regenerative therapy, and for evaluating the efficacy and safety of biomaterials, devices, and biologic factors included in surgical protocols to support or induce periodontal regeneration before such protocols are applied clinically. Animal studies have been performed using rodent, feline, porcine, caprine, ovine, canine, and nonhuman primate models. Variations among species including, but not limited to, anatomy and dimensions of teeth and alveolar processes, amount and character of the gingiva, local physiologic environment, animal behavior, and healing rate make each animal model unique. None of the species listed provide an anatomic and physiologic environment equal to the human oral cavity. However, dogs, which are skeletally larger and easier to handle than many of the other animal models, are preferred for study of periodontal wound healing and regeneration because their physiology is reasonably close to that of humans, they have a clinically relevant size and tooth configuration, they are significantly easier to handle than most other animals during essential postoperative management, and they are the subjects of extensive published research experience. Many other species, including nonhuman primates, may not be as useful in oral wound healing studies because of difficulties in gaining access for required postsurgery follow-up and, especially in the case of nonhuman primates, because of their ability to perturb surgical sites.
Several animal models and defect types have been proposed in the literature for the study of periodontal wound healing and regeneration. Wikesjö and colleagues developed and characterized supracrestal periodontal defects into a critical-size defect model. The model comprises surgically created, circumferential, supra-alveolar periodontal defects, 5 to 6 mm in height (from the cementoenamel junction to the reduced alveolar crest) around the mandibular third and fourth premolar teeth in dogs. Through various experimental studies, this critical-size, supra-alveolar, periodontal defect model has been proven to be a discriminating tool, a “litmus test,” for the preclinical evaluation of candidate therapies for periodontal wound healing and regeneration, including therapies involving bone biomaterials, devices for guided tissue regeneration, growth and differentiation factors, and their combinations. Significant regeneration in this challenging model warrants clinical evaluation of the therapeutic concept; conversely, limited regeneration under optimal circumstances for wound healing does not warrant clinical follow-up.
This article reviews studies evaluating growth and differentiation factors considered candidate therapeutic agents for periodontal wound healing/regeneration. The article focuses on studies using relevant preclinical models and pivotal clinical trials where available. These studies support a rationale for the clinical evaluation and eventual use of a number of growth and differentiation factors to enhance or secure outcomes following regenerative procedures in periodontal defects.
Regeneration of periodontal structures constitutes a complex multifactor process regulated by interactions among cells, hormones, growth factors, and extracellular matrices. These interactions trigger a series of events leading to de novo tissue formation. This process is still incompletely understood. However, advances in molecular and cellular biology have contributed to an understanding of the potential role of growth factors in periodontal wound healing and regeneration and their use as therapeutic agents. Several matrix, growth, and differentiation factors have received attention because of their ability to actively regulate various functions of PDL cells. The effects of such factors as they apply to periodontal regeneration have been evaluated. Examples of growth factors associated with periodontal tissues and considered as candidate agents in support of periodontal wound healing or regeneration include platelet-derived growth factor (PDGF), insulinlike growth factors I and II (IGF-I and -II), acidic and basic fibroblast growth factors (aFGF and bFGF), and transforming growth factor β (TGF-β) ( Table 1 ).
|Factor (Preclinical or Clinical)||Model; Platform||Dose; Carrier; Healing Interval||Major Observation||References|
|PDGF (preclinical)||Fenestration; dog||10 μg/mL; topical application; 1, 3, 7d||Increased fibroblast proliferation|
|Class III furcation; dog||0.5 μg/mL; topical application; 5, 8, 11 wk||Favorable periodontal regeneration including bone fill|
|Chronic periodontitis; nonhuman primate||10 μg; methylcellulose gel; 4, 12 wk||Increased new attachment and bone fill|
|PDGF (clinical)||Phase III randomized controlled trial, 180 patients, ≥4-mm intrabony defects; clinical||0.3, 1.0 mg/mL; β-TCP; 3, 6 mo||0.3 mg/mL: improved attachment level at 3 mo; 1.0 mg/mL: not different from control|
|Case series, 8 patients, defects; clinical and biopsies||0.3, 1.0 mg/mL; β-TCP; 6 mo||Limited periodontal regeneration|
|IGF (preclinical)||Chronic periodontitis; nonhuman primate||10 μg; methylcellulose gel; 4, 12 wk||No periodontal regeneration|
|FGF (preclinical)||3-wall intrabony; dog||30, 40, 50 μg; fibrin gel; 6 wk||Dose-dependent periodontal regeneration|
|2–3 wall intrabony and class II furcation; nonhuman primate||30, 40, 50 μg; fibrin gel; 8 wk||Dose-dependent periodontal regeneration|
|Class II furcation; dog||30 μg; topical application; 6 wk||Increased PDL, bone formation|
|Class II furcation; nonhuman primate||0.1%, 0.4%; gelatin; 8 wk||Dose-dependent bone and cement regeneration|
|Class III furcation; dog||0.5, 1.0 mg; topical application; 90 d||Low dose: greater cement and bone formation|
|Reimplanted incisor; dog||0.1, 1,5 μg; collagen gel; 4, 8 wk||Enhanced cementum formation, PDL fiber|
|FGF (clinical)||Phase II randomized controlled trial, 74 patients, 2–3 wall intrabony defects; clinical||0.03%, 0.1%, 0.3%; hydroxypropyl cellulose; 9 mo||Enhanced alveolar bone height, PDL regeneration|
|TGF-β (preclinical)||Class II furcation; sheep||80 μg/mL; 25% pluronic F-127; 6 wk||TGF-β1 plus GTR enhanced bone formation over TGF-β1 alone|
|Supra-alveolar; dog||20 μg; calcium carbonate composite; 4 wk||Limited cementum and bone formation|
|Class II furcation; nonhuman primate||1.5, 2.5 μg; gelatinous, heterotopic induced ossicles, minced muscle tissue; 8 wk||Enhanced vascularity, substantial regeneration|
|RhPDGF-B/IGF-1 (preclinical)||Periodontitis defects; dog||1/1 μg; aqueous gel; 2, 5 wk||Enhanced bone and cementum formation|
|Chronic intrabony defects; nonhuman primate||10/10 μg; methylcellulose gel; 4, 12 wk||Greater periodontal regeneration|
|RhPDGF-B/IGF-1 (clinical)||Phase I/II randomized controlled trial, 38 patients, bilateral intrabony and furcation defects; clinical||50/50, 150/150 μg/mL; gel; 6, 9 mo||150/150 μg/mL rhPDGF-BB/IGF-1 dose increased bone fill|
Platelet-Derived Growth Factor
PDGF, a polypeptide growth factor, has potent stimulatory effects as a chemoattractant and mitogen for mesenchymal cells (including osteogenic cells), along with an ability to promote angiogenesis in wound healing. The PDGF family includes four isoforms: PDGF-A, PDGF-B, and recently discovered PDGF-C and PDGF-D. PDGF-A and -B are both present in gingival epithelium. PDGF-A may have a prominent role during early wound healing, while PDGF-B may regulate later events. PDGF-A and -B form homodimers (AA or BB) and a heterodimer (AB). In vitro, studies have demonstrated that all three forms enhance gingival and PDL fibroblast chemotaxis, proliferation, and protein synthesis, with PDGF-BB apparently being the most effective ligand. PDGF-BB application onto surface demineralized dentin has been shown to stimulate human PDL cell proliferation and increase cementoblast mitogenesis in vitro. Moreover, PDGF-BB stimulates human PDL cell proliferation and collagen synthesis in a time- and dose-dependent order reaching maximum effect at 24 hours at a dose of 10 ng/mL.
The effect of PDGF in a carrier or combined with guided tissue regeneration (GTR) has been evaluated in periodontal fenestration defects in dogs. Autoradiography showed significantly increased fibroblast proliferation following PDGF application compared with GTR or sham-surgery controls at 1 and 7 days postsurgery. In other studies using chronic class III furcation defects in dogs, a PDGF-BB/GTR/root surface demineralization protocol apparently produced favorable periodontal regeneration compared with the carrier control. Nonhuman primate periodontal defects implanted with PDGF-BB showed significantly greater new attachment formation and bone fill compared with vehicle control at 12 weeks postsurgery. In summary, in vitro studies suggest that PDGF-BB exerts several important effects on cells native to the periodontal environment; and in vivo preclinical studies suggest that PDGF-BB exerts stimulatory effects on periodontal wound healing and regeneration, thus motivating clinical follow-up.
A multicenter phase III randomized controlled clinical trial assessed the safety and efficacy of recombinant human PDGF-BB (rhPDGF-BB) in a β-tricalcium phosphate (β-TCP) carrier. One hundred-eighty subjects requiring surgical treatment of 4-mm or deeper intrabony periodontal defects were randomized to receive rhPDGF-BB at 0.3 or 1.0 mg/mL, or carrier control. Clinical and radiographic evaluations were performed presurgery, and at 3 and 6 months postsurgery. No serious adverse effects attributable to treatments were recorded. Significantly improved attachment levels were observed for sites receiving rhPDGF-BB at 0.3 mg/mL compared with control at 3 months; however, the mean effect was limited (ie, 3.8 vs 3.3 mm). There were no significant differences in attachment level gain at 6 months, attachment level gain averaging 3.8 versus 3.5 mm, respectively. Compared with control or rhPDGF-BB at 0.3 mg/mL, rhPDGF-BB at 1.0 mg/mL exhibited no remarkable or significant differences. A biopsy study including intrabony periodontal in eight patients treated with rhPDGF-BB at 0.3 or 1.0 mg/mL in the β-TCP carrier showed limited periodontal regeneration in 12 of 16 defects (range 0.3–1.6 mm) following a 6-month or greater healing interval. Bone formation never juxtaposed new cementum formation. A majority of the defect sites were filled with residual β-TCP, with bone formation never penetrating the β-TCP mass or contacting the particles. The unremarkable clinical and histologic improvements noted following rhPDGF-BB treatment in these studies raises questions about the relevance and utility of this treatment modality in patient-centered settings.
Insulinlike Growth Factors
IGF-1 and -2 play a critical role in stimulating organogenesis and growth during early stages of embryogenesis as well as in regulation of specific tissue and organ functions at later stages of development. IGF-1 affects cementoblast mitogenesis, phenotypic gene expression, and mineralization ; and stimulates bone formation, growth, and resorption. In vitro, IGF-1 enhances rat and human PDL and gingival fibroblast migration and proliferation in a dose and temporal order, but IGF-1 does not exhibit an apparent effect on type I collagen synthesis. These observations suggest that IGF-1 might play a role in periodontal wound healing and regeneration.
Preclinical studies using IGF-1 in a methylcellulose gel carrier, however, failed to show regeneration of the periodontal attachment following application into induced nonhuman primate chronic periodontal defects. Incremental increases in osteoblast numbers and bone formation compared with sham-surgery control were reported in canine periodontitis defects implanted with IGF-1. Collectively the results may be interpreted to suggest that IGF-1 has limited, if any, appreciable effects on periodontal wound healing or regeneration.
Fibroblast Growth Factor
FGFs exert a range of biologic effects on cells of endodermal, ectodermal, and mesodermal origin; are considered potent growth and differentiation regulators and angiogenic factors; and play important roles in development and wound healing. BFGF (also called FGF-2) found in bone matrix is a multifunctional factor that induces proliferation and morphogenesis in a wide range of cells and tissues, including the PDL. BFGF also appears to exert profound effects on bone growth and development, and enhances fracture healing. Moreover, bFGF inhibits alkaline phosphatase activity and PDL cell mineralized nodule formation in vitro.
Surgically created three-wall intrabony defects implanted with bFGF at various dosages exhibited significantly greater periodontal regeneration compared with carrier or sham-surgery controls at 6 weeks in dogs and at 8 weeks in nonhuman primates. The high dose generated an approximately twofold increase in cementum and bone formation regardless of the species used for testing. Epithelial down growth or root resorption/ankylosis were not observed. In a second study, topical application of rhbFGF (30 μg/site) in a gelatinous carrier into surgically created class II furcation defects in dogs induced increased PDL, cementum, and bone formation compared with control at 6 weeks. Dose-dependent periodontal regeneration was observed at 8 weeks in nonhuman primates using 0.1% or 0.4% rhbFGF with the gelatinous carrier; the high-dose group showed significant bone and cementum regeneration. The effect of bFGF (0.5 or 1.0 mg/site) combined with GTR on periodontal wound healing/regeneration was analyzed in surgically induced mandibular premolar class III furcation defects in dogs. Test sites received bFGF after root conditioning with tetracycline hydrochloride (HCl). Increased regeneration was observed in sites receiving bFGF compared with control at 90 days. Notably, the low-dose group exhibited greater cementum and bone formation compared with the high-dose group. Root resorption/ankylosis was not observed. In yet another study, cementum formation was enhanced following application of bFGF (0.1, 1, and 5 μg/site) in a collagen gel into dentinal defects in freshly extracted and then reimplanted mandibular incisors in dogs. Random PDL fibers attached to dentin were observed at 4 weeks. Newly synthesized dense fibers invading alveolar bone and cementum were observed at 8 weeks in the 1-μg bFGF group. In vitro and in vivo preclinical observations justified the need for clinical trials to determine the potential of bFGF to promote periodontal regeneration also in humans.
A randomized controlled phase II clinical trial enrolled 74 patients with two- or three-wall intrabony periodontal defects to evaluate the effect of rhbFGF in a hydroxypropylcellulose gel carrier on periodontal wound healing/regeneration. A significant increase in alveolar bone height was noted after 36 weeks, suggesting that rhbFGF stimulates regeneration of the PDL also in humans. Taken together, the studies suggest that bFGF may serve as a useful therapeutic adjunct to surgical procedures aimed at promoting periodontal wound healing and regeneration.
Transforming Growth Factor–β
TGF-β stimulates PDL cell extracellular matrix synthesis, mitogenesis, and proliferation. TGF-β receptors are up-regulated in regenerated PDL tissues, suggesting that TGF-β may also be capable of mediating periodontal regeneration.
TGF-β 1 at 80 μg/mL in a gelatinous carrier, 25% Pluronic F-127 (poloxamer 407), and TGF-β 1 in the gelatinous carrier combined with GTR were implanted into surgically created, mandibular premolar class II furcation defects in sheep. Significantly, enhanced bone formation was demonstrated for TGF-β 1 compared with carrier control at 6 weeks; the TGF-β 1 /GTR combination significantly enhanced bone formation over TGF-β 1 alone. Contrasting results were reported in studies using the canine critical-size, supra-alveolar periodontal defect model. Using a split-mouth design and a 4-week healing interval, contralateral defects in six animals received rhTGF-β 1 (20 μg/defect) in a calcium carbonate (CaCO 3 ) composite carrier versus carrier control, both combined with GTR. Defects in another six animals received rhTGF-β 1 versus carrier control without GTR, and still another six animals received carrier control combined with GTR versus GTR without additions. The histometric analysis showed limited, if any, cementum regeneration, without obvious differences between experimental groups, and bone formation generally limited to the apical aspect of the defects. Collectively, the results from these studies suggest that TGF-β 1 possesses a clinically insignificant, if any, potential to stimulate periodontal wound healing or regeneration. Other studies suggest that TGF-β 3 may enhance periodontal wound healing/regeneration. Surgically created class II furcation defects in nonhuman primates, implanted with TGF-β 3 in a Matrigel carrier (a gelatinous protein mixture), TGF-β 3 plus carrier plus heterotopic TGF-β 3 –induced ossicles, TGF-β 3 plus carrier plus minced muscle tissue, or carrier alone showed pronounced regeneration in TGF-β 3 –implanted sites compared with controls at 8 weeks. Striking vascularization in sites receiving TGF-β 3 and displaying multiple capillaries along the edge of the alveolar bone appeared to preside insertion of Sharpey fibers. Notably, substantial regeneration was observed in defects implanted with heterotopically TGF-β 3 –induced ossicles and with TGF-β 3 plus minced muscle tissue. A clinical follow-up of a TGF-β 3 –based therapy appears to be not yet available. Unfortunately, the use of ad hoc defect models without thorough characterization makes the results difficult to interpret and compare with those from established discriminating critical-size defect models, which is why these promising observations also need to be confirmed in such established models.
Growth Factor Combinations
Combinations of growth factors might be used to synergistically improve periodontal wound healing/regeneration. Most investigators have evaluated the effects of single factors only and might thus have overlooked potential large biologic responses comparable with those documented in the literature when growth factors used in combinations interact synergistically in vitro. For example, combinations of PDGF-BB, IGF-1, and TGF-β 1 stimulated PDL cell mitogenesis and adhesion. Interactions among IGF-1, PDGF-BB, TGF-β 1 , and bFGF were evaluated in other studies. Fetal bovine osteoblasts were assessed for surrogates of bone formation and metabolism/remodeling, including osteoblast mitogenesis, collagenous and noncollagenous protein synthesis, and alkaline phosphatase activity. Even though synergistic interactions between IGF-1 and the other factors relative to osteoblast mitogenic activity and protein synthesis were observed, IGF-1 failed to increase alkaline phosphatase activity when combined with TGF-β 1 , PDGF-BB, and bFGF.
An rhPDGF-B/IGF-1 construct surgically implanted into periodontitis defects in dogs significantly increased bone and cementum formation compared with control following a 2- and 5-week healing interval. In a parallel study using nonhuman primates, induced chronic intrabony defects were implanted with rhPDGF-B/IGF-I, rhPDGF-B, rhIGF-1, or carrier control. Significantly greater periodontal regeneration was observed in sites receiving the rhPDGF-B/IGF-1 combination compared with individual factors or the carrier control following 4 and 12 weeks.
The positive in vitro and in vivo preclinical evaluation of the rhPDGF-B/IGF-1 combination motivated a clinical evaluation. Thirty-eight patients with bilateral intrabony and furcation defects participated in a phase I/II clinical trial. Defect sites received surgical implantation of rhPDGF-BB/IGF-I (50 or 150 μg/mL each) in a gel carrier, and were compared with carrier control or sham surgery. Bone fill was evaluated using surgical reentry at 6 to 9 months postsurgery. Subjects receiving the rhPDGF-BB/IGF-I (50/50 μg/mL) combination showed similar bone fill in experimental and control sites, whereas subjects receiving the rhPDGF-BB/IGF-I (150/150 μg/mL) combination showed statistically significant increased bone fill corresponding to a mean of 2.1-mm vertical gain (42% fill) compared with 0.8 mm (19% fill) for the controls. No serious local or systemic adverse effects attributable to treatments were observed. It is noteworthy that despite these encouraging observations, further studies on the rhPDGF-BB/IGF-1 combination have not been reported. Perhaps growth factor combinations for periodontal wound healing/regeneration may never be developed for clinical use because of the substantial, complex, and costly evaluation needed to meet regulatory demands.
Bone morphogenetic proteins (BMPs) form a unique family within the TGF-β superfamily of proteins and have essential roles in regulation of bone formation, maintenance, and repair. While BMPs are frequently referred to as growth factors, it is more precise to regard them as differentiation factors because BMPs play important roles in cell migration, proliferation, differentiation, and apoptosis, and are involved in morphogenesis and organogenesis in such diverse tissues and organs as kidney, eye, nervous system, lung, teeth, skin, and heart. More than 20 BMPs have been identified, and several trials have evaluated rhBMPs for tissue engineering. BMP-2, -3 (osteogenin), -4, -6, -7 (also known as osteogenic protein-1 [OP-1]), -12 (also known as growth/differentiation factor-7 [GDF-7]), and -14 (also known as GDF-5, or cartilage derived morphogenetic protein-1 [CDMP-1]) have been evaluated for periodontal wound healing/regeneration ( Table 2 ).
|Factor (Preclinical or Clinical)||Model; Platform||Dose; Carrier; Healing Interval||Major Observation||References|
|BMP-2 (preclinical)||Supra-alveolar; dog||0.05–0.4 mg/mL; PLGA, DBM, PLA, ACS, CP cement, HY sponge; 8, 24 wk||Significant bone and cementum formation, no PDL, root resorption/ankylosis|
|Supra-alveolar; dog||0.4 mg/mL; gelatin/PLGA; 12 wk||Enhanced bone, cementum, PDL|
|3-wall intrabony; nonhuman primate||0.4 mg/mL; ACS, α-BSM; 16 wk||Enhanced periodontal regeneration|
|3-wall intrabony; dog||0.2 mg/mL; ACS; 8, 24 wk||Enhanced bone but not cementum formation|
|Supra-alveolar; dog||0.1 mg/mL; gelatin sponge, spacer membrane; 12 wk||Spacer eliminated root resorption/ankylosis, but reduced bone formation|
|BMP-3/osteogenin (preclinical)||Class II furcation; nonhuman primate||250 μg/site; type I collagen; 8 wk||Enhanced PDL and bone formation|
|BMP-3; osteogenin (clinical)||Case series, 16 patients, intrabony defects; clinical and biopsies||200 μg/site; type I collagen; 6 mo||Enhanced periodontal regeneration|
|BMP-6 (preclinical)||Fenestration/rat||0, 1, 3, 10 μg/site; type I collagen; 4 wk||3 μg: greatest bone and cementum formation|
|BMP-7/OP-1 (preclinical)||Class II furcation; nonhuman primate||0, 100, 500 μg/g; type I collagen; 8 wk||Enhanced cementogenesis and PDL|
|Class II furcation; nonhuman primate||0.5, 2.5 mg/g; type I collagen; 24 wk||Enhanced PDL and alveolar bone formation|
|Class III furcation; dog||0.75, 2.5, 7.5 mg/g; type I collage; 8 wk||Enhanced periodontal regeneration|
|Class II furcation; nonhuman primate; (rhBMP-2 vs rhOP-1)||100 μg/g; type I collagen; 8 wk||RhOP-1: enhanced cementogenesis; rhBMP-2: enhanced bone formation|
|BMP-12/GDF-7 (preclinical)||Supra-alveolar; dog (GDF-7 vs rhBMP-2)||GDF-7: 0.04, 0.1, 0.2 mg/mL; ACS; 8 wk. RhBMP-2: 0.2 mg/mL; ACS; 8 wk||GDF-7: PDL regeneration. RhBMP-2: no PDL formation|
|BMP-14/GDF-5 (preclinical)||1-wall intrabony; dog||20 μg/site; β-TCP; 8 wk||Enhanced bone and cementum formation, PDL|
|1-wall intrabony; dog||1, 20, 100 μg/site; ACS; 8 wk||Enhanced bone and cementum formation, PDL|
|Supra-alveolar; dog||500 μg/g β-TCP; PLGA; 8 wk||Enhanced cementum and bone formation, PDL|
|Dehiscence; dog||93 μg/site; PLGA; 2, 4, 6, 8 wk||Accelerated bone regeneration|
|1-wall intrabony; dog||RhGDF-5: 500 μg/g; β-TCP; 8 wk. RhPDGF: 0.3 mg/mL; β-TCP; 8 wk||RhGDF-5: enhanced bone and cementum formation over rhPDGF|
|BMP-14/GDF-5 (clinical)||Phase IIa randomized controlled trial, 20 patients, ≥4-mm intrabony defects; clinical and biopsies||500 μg rhGDF-5/g; β-TCP; 6 mo||Twice greater clinical attachment gain, favorable bone and periodontal regeneration|