215 - Stand-alone LLIF procedure: in which patient it can be used without ca...

General Session: MIS-4

Presented by: L. Pimenta - View Audio/Video Presentation (Members Only)


L. Marchi(1), R. Amaral(1), F. Fortti(1), L. Oliveira(1), E. Coutinho(1), R. Jensen(1), L. Pimenta(1,2)

(1) Instituto de Patologia da Coluna, Sao Paulo, Brazil
(2) USCD, Neurosurgery, San Diego, CA, United States


The lateral lumbar interbody fusion (LLIF) has evolved into an effective and less invasive treatment option. A stand-alone construction with the maintenance of the ALL has been shown to be sufficient to provide stabilization. Although, cage subsidence is a potential important complication. Some reports have demonstrated that wider cage footprint diminish this event, but no work have studied patient and surgery factors in the appearance of subsidence. The objective of this work was to identify factors correlated with the lack of cage subsidence in stand-alone procedures. Single center retrospective review of prospective collected data between 2008-2015. Inclusion criteria: single level stand-alone lumbar LLIF. Exclusion criteria: cages with 18mm of antero-posterior dimension; any previous lumbar arthrodesis/arthroplasty surgery; any kind of supplementation (posterior/ lateral/ anterior). 3-month lateral orthostatic radiographs and CT, when necessary, were reviewed. Subsidence grade evaluated in a 4-point scale(0-III) as previously described. Preoperative diagnosis and images reviewed in order to identify the occurrence of spondylolisthesis and scoliosis. ALL rupture, cage rotation and endplate damage documented in the intraoperative data. A risk score was designed based in factors with significant correlation with subsidence grade. Student's t-test, Fischer exact test and ANOVA used with an alpha of 0.05. Ninety-four patients were initially enrolled, 14 were excluded due lack of images, and 80 were eligible for the analysis. Average age of 58.7 y/o (25-84, range), mean bone mass index (BMI) of 27.9(19-50, range), 37 were females(47%). The surgeries involved L4L5 in 84%(66) of the total cases (L1L2, 1 case; L2L3, 5 cases; L3L4, 6 cases; L5VT, 1 case). Twenty-five cases(31%) had spondylolisthesis, 12 cases(15%) had scoliosis, and other ones either DDD, instability and/or stenosis. Two cases(3%) experienced unwanted intraop ALL insufficiency/rupture, and two cases (3%) some extent of endplate damage or cage rotation. Subsidence analysis in 3-month images revealed the following

Results: grade 0= 58 cases(73%); grade I= 10 cases(13%); grade II= 8 cases(10%); grade III= 4 cases(5%). The following risk factors were correlated with high-grade subsidence (grade II and III) when compared with low-grade subsidence group: spondylolisthesis (92% vs 21%; p< 0.001); scoliosis (33% vs 12%; p=0.002); female gender (75% vs 42%; p=0.033); older patients (average 57.3 vs 67.7y/o; p=0.003). Cage rotation/endplate damage showed borderline significance (1 vs 2 cases; p=0.06). Some other factor did not show statistical correlation: ALL rupture (p=0.289; body mass index(p=0.459); cage AP dimension(p=0.495). In order to score the risk factors found, the cases evaluated scored in an unweighted fashion adding one unit as they presented the factors that correlated with grade II/III subsidence. Minimum score=0 and maximum=4. Accordingly, patients presented higher risk scores as they had greater subsidence(p< 0.001). The results of using the score higher or equal to 2 as a predictor of high grade subsidence was 92% of sensitivity and 72% of specificity. It was possible to correlate the degree of subsidence in stand-alone LLIF procedures with demographic (age and gender) and pathology data (spondylolisthesis and scoliosis). With a score based on risk factors, for any particular score< 2, the probability that it is true negative (negative predictive value) is 98%. Further studies are needed to submit this score for validation for the preoperative planning.