Does Vertebral End Plate Morphology Affect Cage Subsidence After Transforaminal Lumbar Interbody Fusion?

OBJECTIVE: To investigate the effect of end plate morphology on cage subsidence and to compare the surgical outcomes among patients with different vertebral end plate morphologies. METHODS: We reviewed a series of consecutive patients from January 2009 to January 2016 who had undergone monosegment L4/5 transforaminal lumbar interbody fusion (TLIF) with a follow-up >2 years. The enrolled patients were divided into 3 groups based on the preoperative vertebral end plate morphology on T1-weighed sagittal magnetic resonance scans: concave group (C group), flat group (F group), and irregular group (Ir group). Lumbar lordosis (LL), segmental lordosis (SL), and disc height (DH) were measured on the plain image at each follow-up, and three-dimensional computed tomography (3D-CT) was obtained at 1 year follow-up to evaluate the cage subsidence and solid fusion. RESULTS: A total of 145 consecutive patients (41 males and 104 females) were included in this study, with a mean follow-up of 33.8 +/- 12.3 months. The age was significantly older in the Ir group than in the C group or F group (P < 0.05). Cage subsidence was detected in 23 patients (15.9%) at 1 year follow-up through 3D-CT. The incidence of cage subsidence was significantly higher in the Ir group than in the F group or C group (P < 0.05). Patients in the Ir group had significant loss of DH, SL, and LL at the latest follow-up, compared with those in the C group and F group (P < 0.05). Patients with cage subsidence had a significantly older age (P < 0.05). The presence of cage subsidence was associated with end plate morphology as shown by logistic regression analysis (P < 0.05). Before surgery, Oswestry Disability Index and visual analog scale back pain scores were significantly higher in the Ir group than in the C and F groups (P < 0.05). After surgery and until the latest follow-up, each group experienced significant improvement in contrast to preoperative scores regardless of end plate morphology (P < 0.05). CONCLUSIONS: Morphology of the end plate plays an important role in the development of cage subsidence after TLIF surgery. Fused segments with irregular end plates are prone to cage subsidence. Although cage subsidence does not affect short-term clinical outcomes, measures should be taken to prevent cage subsidence-related loss of SL and total LL.