<h3>BACKGROUND CONTEXT</h3> Clinical outcomes for patients with intervertebral disc injury or disease could be enhanced by a new generation of regenerative matrices that create tissue environments to control angiogenesis, chondrogenesis and bone formation. Regenerative matrices are a new class of controlled release biomaterials that use surface chemistry, materials architecture, and drug delivery kinetics to create optimised conditions for cell interactions that result in high quality tissue formation. A novel biodegradable polymer matrix that releases bone morphogenetic protein 2 (BMP-2) over a period of one month with a low Cmax and sustained presentation of active protein is investigated in this study. <h3>PURPOSE</h3> BMP-2 release kinetics and protein activity from a new regenerative matrix was quantified to develop an implant with 1-month release of the drug within a macroporous matrix. A rabbit posterolateral (PLF) vertebral fusion model was performed to assess bone formation and fusion outcomes as a function of dose and source of BMP-2 (E coli and CHO cells). <h3>STUDY DESIGN/SETTING</h3> The study was designed to investigate the effect of various biomaterial design factors and manufacturing methods on the rate of release of BMP-2, the activity of BMP-2 and the extent of in vivo bone formation. <h3>PATIENT SAMPLE</h3> This study did not involve patients. <h3>OUTCOME MEASURES</h3> Outcome measures were the mechanical properties of the regenerative matrix, the architecture and porosity of the matrix, BMP-2 release kinetics, BMP-2 activity as a function of release time and bone formation. <h3>METHODS</h3> BMP-2 activity was tested using two in vitro cell-based assays. BMP-2 solutions and regenerative matrix extracts were tested for their ability to induce alkaline phosphatase (ALP) activity in W-20-17 stromal cells using a modified version of the ASTM standard protocol for in vitro biological activity of recombinant human BMP-2 (F2131-02). Cells were exposed to serial dilutions of BMP-2 for 24 h, assayed for ALP activity and EC50s calculated allowing the relative potency to NIBSC WHO BMP-2 standard to be calculated. BMP-2 release was assessed with an in vitro assay of ALP activity in the mouse muscle myoblast cell line C2C12. Formulations were suspended above a C2C12 monolayer in an insert with a porous (8 µm) PET membrane. Following 3-4 days of exposure, the inserts containing LDGraft were transferred to fresh cells for a further 3-4 days. Cells exposed to LDGraft in this manner were assayed for ALP activity (normalised to protein content) once the inserts had been transferred to fresh cells. Fusion rate and new bone formation were compared to INFUSE in a single-level lumbar (L4-L5) posterolateral (PLF) vertebral fusion model in the New Zealand White rabbit based on standard protocols (Boden et al, 1995, ASTM F3207). Pre-formed implants (3 cc) of INFUSE (rhBMP-2 and absorbable collagen sponge) were implanted on decorticated dorsal surfaces of the transverse processes of L4 and L5 adjacent to the laminae. Fusion was assessed by radiograph at 6 weeks and by manual palpation, radiography and microCT at 12 weeks. <h3>RESULTS</h3> BMP-2 activity was retained over one month of release with the addition of protectants in the formulation and precise control of manufacturing processes. The matrix had a porosity of >70% with a high proportion of macropores (>200 micron diameter). Low doses of E. coli-produced BMP-2 induced fusion in the PLF model. <h3>CONCLUSIONS</h3> A one-month releasing BMP-2 formulation is described. This new regenerative matrix had internal architecture designed to host angiogenesis. Manufacturing and design challenges can be overcome to retain high biologic activity throughout the release period. <h3>FDA DEVICE/DRUG STATUS</h3> BMP-2, not approved for this indication.