Anterior versus Posterior Treatment of Thoracolumbar Burst Fractures without Neurological Deficit: A Prospective Randomized Study

Methods: From May 1995 to Mar 2001, a consecutive series of subjects meeting the criteria were included. All had acute isolated burst fractures of the thoracolumbar junction (T10 - L2) without neurological deficit. Those willing to participate were randomized to receive either an anterior fusion with instrumentation or a posterior fusion with instrumentation. No specific degree of canal compromise or kyphosis was a means of exclusion, nor were laminar fractures. However, posterior ligamentous disruption (exam, radiography) or facet dislocation was an exclusion. Upon admission, all subjects received standard radiography including 2D CT. Radiographs were repeated at 2, 4, 6, 12, and 24 months. The CT scan was also repeated at 24 months. Measured were kyphosis, degree of canal compromise, instrumentation issues, and fusion rates. Hospital stay, cost, operating time, blood loss, complications, return to work, medication use were measured. All patients completed visual analog pain scales and a Roland and Morris back disability questionnaire on admission. At 24 months they were repeated along with the Owestry and SF- 36 questionnaires.

Results: 39 subjects met the criteria for inclusion and were randomized. The average follow-up was 42 months (range 24-108). Average age for both groups was 43. Twenty received a posterior spine fusion and nineteen an anterior approach. Motor vehicle accidents (16) and falls (15) were the most common etiologies. L1 (24) was the vertebra most commonly fractured. For those treated with a posterior approach, the average kyphosis on admission was 12º; on discharge 5º; and at follow-up 12º. The average canal compromise on admission was 39% and at followup 20%. The average pain scale (0-10) was 6.3 on admission, and 3.4 at follow-up. The Roland and Morris (0-25) was 1 at the time of the accident, and 9 at follow-up. At final follow-up the Oswestry was 22. Of those treated with anterior surgery, the admission kyphosis averaged 13.7º; 2.6º (P<0 .05) at discharge, and 3.7º follow-up. The average canal compromise was 40% admission 18% pain scale on 5.8 2.3 (p="0.008)" R-M 1 6(n.s.) Owestry 16. There were 19 complications in 16 patients treated posteriorly; 3 minor anteriorly.

Discussion: Both anterior and posterior surgery for thoracolumbar burst fractures offer acceptable return to work rates and patient satisfaction.

Conclusions: Although patient outcomes are similar, anterior fusion and instrumentation for thoracolumbar burst fractures present fewer complications, retain sagittal alignment, and exhibit a trend towards less pain when compared with posterior surgery.