Background/Purpose: :  

To describe the pattern and frequency of incident vertebral fractures (VF) in glucocorticoid (GC)-treated children and to determine the clinical factors at baseline that are associated with incident VF.

Methods:

In children with leukemia, rheumatic disorders and nephrotic syndrome, VF were assessed prospectively each year following GC initiation for 3 years, according to the Genant semi-quantitative method. An incident VF was defined as a new fracture in a previously normal vertebral body or worsening of an existing fracture. The 3-year cumulative VF incidence rate was then calculated. Multivariable Poisson regression was used to examine associations between the 3-year total number of incident VF and clinical factors at baseline, including diagnosis, age, gender, pubertal stage, height and body mass index (BMI) Z-scores, calcium and vitamin D intake, physical activity, back pain, and lumbar spine (LS) bone mass density (BMD) Z-score. In addition, prevalent VF (at baseline) as well as cumulative GC exposure and the number of days in receipt of GC until the baseline visit were also assessed for their relationship with 3-year, incident VF.

Results:

404 children were enrolled at a median age of 6.2 years, range 1-17; 50% boys; 188 (46%) had leukemia, 136 (34%) rheumatic conditions, and 80 (20%) nephrotic syndrome. The baseline study visit occurred at a median of 18 days following steroid initiation (inter-quartile range 11-24 days). Forty-four (11%) children had VF at baseline, while 134 incident VF were detected in 55 children over the 3 years.  Overall, 17% of children (95% CI 13-22) had at least one incident VF over the 3 years. The proportions of children with incident VF were as follows: Leukemia: 24% (95% CI 16-32); rheumatic disorders: 13% (95% CI 6-19); nephrotic syndrome 9% (95% CI 1-17). The annual proportion of children with incident VF peaked at 12 months and declined thereafter (p=0.04). Among those with incident VF, 24/55 children (44%) had 1 or more moderate or severe fracture. Most of the VF were new fractures in previously normal vertebral bodies (86%), compared to worsening of existing VF. In Poisson multivariable modeling assessing baseline clinical factors, the following were associated with higher, 3-year VF incident rates: the presence of VF at baseline (incidence Rate Ratio (RR) 6.3, 95% CI 3.2-12.4), female gender (RR 1.8; 95% CI 1.0-3.3), pre-pubertal status (RR 2.1; 95% CI 0.8-5.4), and lower BMD Z-scores (RR 1.4; 95% CI 1.1-1.7). Underlying diagnosis and back pain at baseline were highly correlated with these significant factors (co-linear). Other factors at baseline were not significantly associated with incident VF.

Conclusion:

Within 3 years of steroid initiation, 17% of children had incident VF. VF incidence peaked at 12 months, and almost half of the VF were moderate or severe. Of the factors measured at baseline, prevalent VF were most strongly associated with incident VF over the ensuing 3 years. Female gender, pre-pubertal status and low BMD at baseline were also associated with 3-year incident VF (for every 1 standard deviation reduction in spine BMD Z-score, there was a 40% increased incident VF risk).

Funded by CIHR FRN 64285

---

« Back to 2013 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/vertebral-fractures-in-the-3-year-period-following-steroid-initiation-among-children-with-chronic-illnesses/