Object: The authors compared fusion rates in transforaminal lumbar interbody fusion (TLIFs) when using either autograft or bone morphogenetic protein (BMP) placed in the interbody space.
Methods: Between September 2002 and December 2003, the authors performed 44 TLIF operations. Follow-up data were available for 40 patients. Of the 40 procedures, 19 involved cages filled with iliac crest autograft (Group 1) and 21 involved cages filled with a medium kit of recombinant human (rh) BMP-2 (Group 2). In all Group 2 patients, one BMP sponge was placed anterior to the cage and another was placed within the cage. In 12 of the Group 2 patients, iliac crest autograft was placed posterior to the BMP-filled cage (Group 2A). In the remaining nine Group 2 patients, only local autograft was placed posterior to the BMP-filled cage (Group 2B). Assessment of fusion was performed using dynamic radiography at 3-month intervals. Outcomes were assessed using the Prolo Scale, and iliac crest donor site pain was measured using a Visual Analog Scale (VAS). The mean follow-up period was 9 months (range 3-18 months). In Group 1 patients, one pseudarthrosis was detected. In Group 2 patients, dynamic radiography demonstrated solid fusion in all patients except one in Group 2B. Fifty-eight percent of patients in whom iliac crest autograft was used complained of donor site pain 6 months after surgery (5 of 10 points on the VAS). Symptomatic foraminal bone formation was not observed in any Group 2 patient.
Conclusions: The use of rhBMP-2 is safe in TLIFs when the sponges are placed away from the dura mater, and BMP promotes a more rapid fusion than iliac crest autograft alone. The use of rhBMP-2 in combination with local autograft is an excellent option for promoting solid fusion with TLIF, and it eliminates the possibility of iliac donor site pain.