Long-term follow-up after vertebroplasty - A mean 10-years follow-up control study

Abstract

Objectives: To evaluate the clinical 10 year outcome of patients treated with percutaneous vertebroplasty for vertebral compression fractures and to determine the incidence of new fractures in this time interval, as well as the mortality of the patients who underwent this procedure.

Methods: All patients undergoing vertebroplasty for vertebral compression fractures between May 2007 until July 2008 were prospectively followed up at 10 years postoperatively. Patients were assessed for radiologic outcome and self-reported outcome parameters (PROs). Gathered parameters remained unmodified to the initial ones analyzing QoL improvement (EQ5D 3L and NASS score) and pain alleviation (VAS, NRS). Mortality was defined as an additional endpoint. Exclusion criteria include additional instrumentation, use of additional devices such as kyphoplasty balloons/stentoplasty, cognitive impairment, insufficient radiological documentation or absent re-consent.

Results: Of 280 patients who underwent vertebroplasty, 49 (17.5%) were available for re-assessment with a mean follow-up of 10.5 years (9.9-11.1). Thirty patients (10.7%) were assessed clinically and radiologically, 16 (5.7%) in written form and three (1.1%) by phone only. A total of 186 (66.4%) died during the follow up period. Out of the remaining 45 patients, 27 patients declined participation, eight couldn't participate due to cognitive impairment, four had insufficient radiologic documentation. Six patients were lost to follow-up. At 10 years, patients reported a consistently improved quality of life (EQ-5D; p < 0.01) and global satisfaction. Vertebroplasty demonstrated a substantial and enduring effect on alleviating back pain over 10 years (p < 0.001). 26 (53%) patients experienced a new fracture since the initial procedure.

Conclusion: A decade following vertebroplasty, patients continue to demonstrate a quality of life and pain level comparable to short and medium-term assessments, with a significant difference from baseline measurements. More than half (53%) of the patients participating at last follow-up experienced new fractures during this interim period. The cohort as a whole has been impacted by an elevated mortality rate over the time period.

Keywords: Osteoporosis; Thoracolumbar; Trauma; Vertebroplasty.

Fig. 1

Flow-chart.

Fig. 2

Note: Only fractures diagnosed/treated in Inselspital Bern were collected after 2 years assessment for patients who did not participate on 10 yrs extension. Note: For patients who occurred several fractures at different days during FU, only the first event was counted. Note: Patients who didn't have a new fracture confirmed are censored at the time of the last visit, which can be either 10 yrs visit or the last visit of the original study.

Fig. 3

For patients with a known date of death, this date was used for the analysis. Patients who died between 2 and 10 years, for whom only the year of death was available, were imputed with the 30th June of the corresponding year. Patients who deliberately did not consent to the 10 year visit were marked as censored as at 31 December 2017. Patients lost to FU were censored at the time of their last study visit.

Fig. 4

EQ5D.

Fig. 5

NRS back pain.

Fig. 6

NASS lumbar pain.