INTRODUCTION:

Lumbar interbody fusion using bone substitutes is a common procedure. The use of silicated calcium phosphate bone substitute (Actifuse^®) is associated with fusion rates in the order of 80% in previous studies. The aim of this study was to review the rates and characteristics of the interbody arthrodesis in patients undergoing lumbar interbody fusion surgery and compare traditional pedicle screw and cortical bone trajectory techniques.

METHODS:

A consecutive series of patients undergoing primary single level posterior interbody fusion for degenerative lumbar conditions were treated by a single surgeon and prospectively studied. The choice between techniques was based on surgical pathology. Fusion was assessed at 12 months by an independent radiologist using fine cut computed tomography. The radiological parameters assessed were implant-related (screw loosening, screw breakage, cage subsidence and lucency around the cage) and bone-related (bone growth through the cage, bone growth outside the cage, new end-plate sclerosis and new end-plate cyst formation). Fusion was assessed as a non-union, a locked pseudarthrosis or a solid union depending on the presence of bridging bone through the cage. A functional fusion was considered to be a locked pseudarthrosis or solid union. Data was analysed using Fisher exact and unpaired t-tests.

RESULTS:

Between July 2015 and January 2018, 121 patients met the inclusion criteria. The 12-month CT scan was available for 106 patients. Of these, the cortical technique was used in 42 and the traditional in 64. Overall, 91/106 patients (85.8%) achieved functional fusion, with 83/91 (91.2%) of these having a solid union and 8/91 (8.8%) a locked pseudarthrosis. Non-union rates between the cortical screw and pedicle screw groups (14.2% vs 14.1%) were similar. There were more solid unions in the cortical group (46/64 (71.8%) vs 37/42 (88.1%)) but this did not reach statistical significance (p=0.056). For the rest of the analysis we considered the two groups together and assessed the characteristics of the non-unions, of which there were 15/106 (14.2%). The nonunions when compared with the functional fusions, showed a statistically significant association with cysts (93.3 vs 7.7%), sclerosis (86.7 vs 12%), screw loosening or breakage (60 vs 2.2%), lucency (66.7 vs 32.9%) and subsidence (93.3 vs 6.6%). All values were highly significant (p<0.01).

DISCUSSION:

We demonstrated a satisfactory rate of fusion in both groups of 85.8%, which is comparable to previously published studies. We found no statistical difference in functional fusion rate between pedicle and cortical screw techniques. We identified radiological predictors of non-union on CT scan, the most significant of which was the presence of new end-plate cyst formation and sclerosis. These may represent a consequence of the osteostimulatory nature of the bone graft substitute. Accepted classification systems of interbody fusion success have not included these criteria. We propose that end-plate cyst formation and sclerosis should be considered as additional parameters in assessment of fusion, especially where bone substitutes are used.