The Roof Technique in the Treatment of Bone Tumors of the Spine. A Ten Years’ Experience

Introduction: Besides appropriate oncological principles for resection, spine tumors require complex reconstructive techniques in a setting of biologic compromise. Fusion outcomes in this cohort are influenced by several variables. Most previous studies focused more on the oncology related prognosis and only briefly reported on fusion rates and instrumentation failure. The aim of our study was to evaluate the efficacy of using bone graft for fusion across long posterior column defects after en bloc resection. Material and Methods: This retrospective observational study analyzed 30 patients with primary spinal tumors who underwent posterior column reconstruction after en bloc resections. In all cases, the resected anterior column was reconstructed with 3D-printed custom-made prosthesis, titanium mesh cages (TMC) or Polyetheretherketone/Carbon fiber (PEEK/CF) Vertebral Body Replacement (VBR). In regions where the posterior elements were completely resected, a structural fresh-frozen allograft or autologous bone graft was placed to span the defect and cover spinal cord. Fashioned bone graft was prepared with saddle cuts on its proximal and distal end to sit on the surface of the extremities of the spinous processes anchored to posterior instrumentation with sublaminar bands. Posterior fusion graft assessment was performed using CT scans with multiplanar reconstruction (MPRs) at 6 months, 12 months, and at the last follow-up. Fusion status was recorded according to four grades: complete fusion (cortical union of the structural allograft with central trabecular continuity), partial fusion (cortical union with partial trabecular incorporation), unipolar pseudarthrosis and bipolar pseudarthrosis (central trabecular discontinuity with superior or inferior cortical non-union, or both, along with a complete lack of central trabecular continuity, respectively). Results: Complete posterior graft fusion was achieved in 11 patients (36.67%, p=0.32) at six months of follow-up while 16 patients (53.33%, p=0.25) achieved partial fusion at the first follow-up. 28 patients achieved grade I graft fusion at the final follow-up (mean 56.38 ± 23.40 months, p=0.18). A rod fracture, without evidence of graft dislodgment, occurred only in one patient who underwent revision surgery at 20 months of follow-up. 3 (3%) patients experienced wound infections treated with DAIR without affecting the final degree of fusion. The last clinical and radiographic follow-up with a mean of months (±) showed no evidence of graft dislodgment or increased axial pain. In 4 patients there was evidence of local recurrence (LR) within 12 ± 3.4 months of followup. In one case a revision surgery for decompression was required at 11 months of follow-up. Conclusion: This innovative technique of fashioning the graft posteriorly allows for both early and long-term stability with solid fusion. Additionally, in case of revision surgery for local recurrence, it avoids direct exposure of the spinal cord.