The range of allograft products for spinal fusion has been extended with the development of cellular bone matrices (CBMs). Most of these combine demineralized bone with viable cancellous bone prepared in a manner that retains cells with differentiation potential. The purpose of this study was to compare commercially-available human CBMs in the athymic rat model of posterolateral spinal fusion. The products compared were Trinity ELITE® (TEL, OrthoFix), ViviGen (VIV, DePuy Synthes), Cellentra (CEL, Zimmer Biomet), Osteocel® Pro (OCP, NuVasive), Bio4 (BIO, Stryker) and map3 (MAP, RTI Surgical). Bone from the ilia of syngeneic rats was used as a control to approximate the human gold standard.

All implants were stored, thawed, and prepared per manufacturer's instructions and all implantations occurred within the manufacturer's time allowance for use after preparation. In total, fifteen 9–10 week old male rats were implanted per implant type, with three different lots of each implant used per five rats to account for lot-to-lot variability. Under anesthesia, a posterior midline longitudinal skin and subcutaneous incision was made, followed by bilateral longitudinal paraspinal myofascial incisions to expose the transverse processes at the L4–5 level. Implants (0.3 cc of allograft or freshly harvested syngeneic iliac bone graft) were placed bilaterally. Surgeons were blinded as to CBM implant type. Incisions were closed with sutures and in vivo microCT scans performed within 48 hours of surgery. A second microCT scan was taken at euthanasia, six weeks after surgery, and the lumbar spines harvested. Fusion was evaluated by manual palpation by three independent, blinded reviewers. MicroCT analysis was performed by an independent CRO (ImageIQ, Cleveland OH). Anonymity of implant type was rigorously kept to avoid bias.

By manual palpation, 5/15 (33%) spines of the syngeneic bone group were fused at 6 weeks. The TEL (8/15, 53%) and CEL (11/15, 73%) groups were not significantly different from each other but were from all other CBM groups. Only 2/15 (13%) of VIV-implanted spines fused and none (0/15, 0%) of the OCP, BIO and MAP CBMs produced stable fusion. The mineralized cancellous bone component of the allografts confounded radiographic analysis but microCT analysis indicated bone volume increased over six weeks for all groups except the syngeneic bone (−4.3%). TEL (+65%) and CEL (+73%) were not different from each other but were significantly increased over all other groups (VIV 29%, OCP 37%, BIO 19%, and MAP 45%, respectively).

CBMs have distinct formulations and are likely processed differently. The claimed live cell and stem cell contents differ between products. Additionally, map3 has cells added at the time of surgery, whereas the other CBMs are processed to retain matrix-adherent cells. Given the wide range of formulations, differences in performance were not surprising, and Trinity ELITE and Cellentra did significantly better than other implants at both forming new bone and achieving fusion. The other CBMs did not have greater bone formation than the control and were very poor at forming a solid fusion. These findings suggest more careful consideration of these allograft products is needed at the clinical level.