BMI and gender increase risk of sacral fractures after multilevel instrumented spinal fusion compared with bone mineral density and pelvic parameters

Abstract

Background ContextSacral fractures are a rare but potentially devastating complication. Long-fusion constructs, including the sacrum, that do not extend to the pelvis may result in sacral fractures. Besides established risk factors including gender, age, and number of levels fused, body mass index (BMI), pelvic parameters, and bone mineral density (BMD) have also been proposed as potential risk factors for postoperative sacral fractures. The literature supporting this, however, is limited. PurposeThe aim of the present study was to assess whether preoperative pelvic parameters, BMI, or BMD of patients with sacral fracture are different compared with age, gender, and fusion level-matched non-fracture controls. Study Design/SettingThis is a case-control study. Patient SamplePatients undergoing posterior instrumented fusion at a single academic institution between 2002 and 2016 were included in the study. Outcome MeasuresThe outcome measure was occurrence of a postoperative sacral fracture. MethodsPatients with sacral fractures after posterior instrumented spinal fusion, including the sacrum, were retrospectively identified and matched 2:1 with non-fracture controls based on gender, age, and number of levels fused. Patients with concurrent spinopelvic fixation or missing preoperative computed tomography (CT) imaging were excluded. Preoperative sagittal balance was assessed using lateral radiographs. Quantitative computed tomography (QCT) assessment included standard measurements at L1/L2 and additional experimental measurements of the S1 body and sacral ala. ResultsTwenty-one patients with sacral fracture were matched to non-fracture controls. The majority of the patients with sacral fracture was female (76.2%) and of advanced age (mean 66.4 years). Fracture and control groups were well matched with respect to gender, age, and number of levels fused. Standard measurements at L1/L2 showed no significant difference in BMD between the fracture and the control groups (109.9 mg/cm3 vs. 116.4 mg/cm3, p=.414). Similarly, there was no significant BMD differences between the groups using the experimental measurements of the S1 body (183.6 mg/cm3 vs. 176.2 mg/cm3, p=.567) and the sacral ala (8.9 mg/cm3 vs. 4.8 mg/cm3, p=.616). Mean preoperative pelvic incidence-lumbar lordosis mismatch and pelvic tilt were not significantly different between the groups. Univariate conditional logistic regression analysis revealed that the odds of experiencing a sacral fracture was approximately six times higher for obese patients compared with normal or underweight patients. After controlling for BMI in multivariate conditional logistic regression models, BMD was still not significantly associated with the odds of experiencing sacral fractures. ConclusionsTo our knowledge, this is the first study to assess the association of preoperative BMD measured by QCT, pelvic parameters, and BMI with postoperative sacral fractures in a large patient cohort. Interestingly, our data do not show any difference in preoperative pelvic parameters and BMD between the groups. This is in line with previous reports that indicate only a few patients with sacral fracture after fusion surgery have clear evidence of osteoporosis. Bone mineral density as a measure of bone quantity, rather than bone quality, may not be as important in these fractures as previously thought. Obesity, however, was associated with higher odds of experiencing postoperative sacral fractures. The present study thereby challenges the widespread concept that obesity is a protective factor against fractures in the elderly. In summary, our results suggest that BMI and gender, more than pelvic parameters and BMD, are risk factors for postoperative sacral fractures.