Retrospective analysis of 848 fractures of the thoracolumbar spine

Abstract

Klaus Fischer, MD, Bochum, Germany; J.C. Ward, MD, St. Gallen, Switzerland; E.J. Muller, MD, Bochum, Germany; F. Magerl, MD, St. Gallen, Switzerland; G. Muhr, MD, Bochum, GermanyPurpose: To gain epidemiological data for the fracture classification according to Magerl et al. [1], concerning fracture type and according neurologic deficit, fracture distribution, sex, age and fracture level.Methods: Evaluation of all patients treated between 1988 and 1995 with traumatic fractures of the thoracolumbar spine. Eight hundred forty-eight cases could be classified and were evaluated with chart reviews, X-ray dossiers and computed tomography scans. Data analysis and statistical evaluation was done with the SPSS program 9.0 (SPSS, Inc.).Results: This study included 533 male (average age, 38.1 years; range, 11 to 93 years) and 315 female (average age, 47.4 years; range, 7 to 92 years) patients, with an average age of 41.5 years (range, 11 to 93 years). In almost half of the patients (49.3%) a fall was the cause of injury, in 19.4% it was a sports injury and in 28.4% a traffic injury; 2.8% were classified as others. Subdividing the falls, 125 fall injuries were during walking and 293 falls were from significant height, including 20 falls in suicidal intention. In all patients only closed spinal injuries occurred, 70% were one-level spinal injuries, in 39.5% concomitant injuries were seen and 10.6% were classified as polytrauma patients. A total of 580 fractures were classified as type A injuries (compression type injuries), 105 type B injuries (flexion-distraction or hyperextension injuries) and 163 type C injuries (rotational injuries). More than two thirds of the fractures were located at the thoracolumbar junction, and a second peak was seen at the levels of T6–T7. Divided up into the different subgroups, there were 342 type A1, 20 type A2 and 218 type A3 injuries seen. The according neurological deficit increases significantly from A1 to A3 (A1, 0.8%; A2, 10%; A3, 19.7%). The type B injuries, subdivided into 64 B1 and 41 B2 injuries, showed also a significantly increased neurological deficit, compared with the type A injuries (B1, 26.6%; B2, 24.4%). However, no B3 injury (hyperextension injuries) was seen. The type C injuries were subdivided in 98 type C1, 56 C2 and 9 C3 injuries. A neurological deficit was seen in 42.9% for C1 injuries, 26.8% for C2 injuries and 66.7% for C3 injuries. Responding to fracture level and fracture type, the burst split fractures (type A321) were seen only between T12 and L5, with a significant correlation with the levels of L1/L3, type A13 fracture was significantly correlated with T8, type A313 was significantly correlated with L1, type B121 was significantly correlated with L1, type C132 was significantly correlated with L3 and type C133 was significantly correlated with L1/L3. Correlating fracture level, fracture type and sex, only a correlation for the female patients with type A1 injuries with the level of T6 could be found. Type A injuries (eg, A1) were significantly related to the age groups over 60 years, and type B and C injuries showed a higher correlation to the younger population.Conclusion: Only a part of the evaluated data could be shown for the abstract. To the authors' knowledge, there are no studies comparable to this one. Although the number of spinal injuries collected is high, only a limited statistical evaluation is possible because of the high number of different subtypes of the classification of Magerl et al. [1]. Nevertheless, a significant correlation, according to the injury mechanism, between the neurological deficit and the different fracture types, as well as age-related and fracture type–related correlations, could be found.