Long-term data shows that non-operative treatment may be better than operative treatment for patients with stable thoracolumbar burst fractures


The results of a study, presented at the annual meeting of the North American Spine Society (NASS, 23–27 October, Dallas, USA), indicated that after 15-20 years of follow-up, non-operative treatment for stable thoracolumbar burst fractures is associated with reduced pain and improved function compared with operative treatment.

Lead author and study presenter Kirkham Wood, Massachusetts General Hospital, Boston, USA, said: “Burst fractures of the thoracolumbar spine, even those without instability or neurologic deficit, can lead to significant disability and the treatment of these fractures has evolved over the years since they were first described more than 100 years ago. But, whether to operate or not has remained a little bit uncertain.”

He added that, in 2003, he and co-authors of the study found that operative treatment of patients with stable thoracolumbar burst fractures was not associated with any advantages compared with non-operative treatment after four and a half years of follow-up (Wood et al. J Bone Joint Surg Am 2003; 85:773–81). However, Wood explained, other investigators—again after around four and half years of follow-up—have found that operative treatment was associated with improved functional outcomes and was more cost-effective compared with non-operative treatment.

Therefore, Wood and co-authors revisited their original group to review the long-term effect of operative treatment versus non-operative treatment. The patient sample was 47 consecutive patients, with stable burst thoracolumbar fractures, who had been prospectively randomised to receive operative or non-operative treatment in the original study. Wood explained: “At an average of 17 years (ranging from 15 to 20 years) of follow-up, patients were visited again. They were asked to fill out patient-related outcome scores and underwent radiographic evaluation.” He added that of the original sample, 19 operative patients (13 male; six female) and 18 non-operative patients (15 male; three female) were available for the long-term evaluation.

According to Wood, non-operative patients “did seem to drift into a bit more kyphosis” compared with operative patients but this difference was not significant (in the four-year study, there was no significant difference in average kyphosis between the groups). He said: “After two decades of follow-up, we did see adjacent segment degeneration in approximately three quarters of those patients—whether they were treated surgically or not. But, what was significantly different was that distally, in the lower aspect of the lumbar spine, the frequency of degeneration was much higher in those who were treated with an operation as opposed to those who were treated with a cast or a brace (77% vs. 33%, respectively; p=0.02).” He added that, after the long-term follow-up, the Visual Analogue Scores of the non-operative patients were nearly half that of those treated operatively (1.7 vs. 3.9, respectively; p=0.14) but this finding did not “quite reach statistical significance”.

The non-operative group, however, did have significantly better Roland and Morris Disability scale scores than the operative group (23.5 vs. 17, respectively; p=0.003) and also had significantly better Oswestry Disability Index scores (4.9 vs. 17.2, respectively; p=0.002). Wood stated: “All indices of the SF-36 favoured the non–operative group, although only social (p=0.02), pain (p=0.01), and general health (p=0.01) reached actual statistical significance.” Furthermore, three quarters of non-operative patients were still working compared with just under half of operative patients.

Wood concluded: “With this first study that has significant follow-up, we still cannot see any real advantage of operative treatment for those stable, neurologically intact, burst fractures. In fact, it may be that non-operative treatment—at long-term follow-up—is significantly advantageous.”