Paraspinal muscle CT measurements in predicting subsequent fractures after percutaneous kyphoplasty in elderly women

Subsequent fractures are a common complication after percutaneous kyphoplasty (PKP) in elderly women with osteoporotic vertebral compression fractures (OVCF). Although bone mineral density (BMD) is a known predictor, its limitations highlight the need for additional prognostic markers.

This study aimed to evaluate the predictive value of lumbar paraspinal muscle parameters measured by preoperative computed tomography (CT) for subsequent fractures after PKP, and to assess whether incorporating these muscle parameters improves risk stratification beyond BMD alone.

This prospective longitudinal study enrolled 185 elderly female patients (mean age 70.9 ± 8.6 years) with OVCF who underwent PKP. Preoperative quantitative CT (QCT) was used to measure muscle parameters-including density (Hounsfield units [HU]) and functional cross-sectional area (FCSA) of the extensor (erector spinae and multifidus) and flexor (psoas major) muscles at the L4 level-as well as volumetric BMD (vBMD) and vertebral HU. Incident subsequent fractures were defined according to the semiquantitative Genant classification during a mean follow-up of 4.7 years. Cox proportional hazards models, Kaplan-Meier survival analysis, and integrated discrimination improvement (IDI) were employed to assess predictive performance.

During follow-up, 56 patients (30.3%) developed incident fractures. In multivariable analysis adjusted for age, BMI, fracture history, and baseline vertebral fracture status, the extensor-to-flexor density ratio remained an independent protective factor for any incident fracture (hazard ratio [HR] = 0.29 per SD, 95% Confidence interval [CI]: 0.11-0.79), while vBMD showed the strongest protective effect for vertebral-specific fractures (HR = 0.60 per SD, 95% CI: 0.41-0.88). Incorporating muscle parameters into a baseline BMD model significantly improved predictive performance for both fracture endpoints (IDI = 0.067 for any fracture, p = .036; IDI = 0.09 for vertebral fracture, p = .028).

CT-based measurements of paraspinal muscles, particularly the extensor-to-flexor density ratio, are independent predictors of subsequent fractures after PKP in elderly women. Combining muscle parameters with vBMD enhances risk stratification, underscoring the clinical value of assessing the "muscle-bone unit" in postoperative management. These findings suggest that postoperative rehabilitation strategies should prioritize strengthening of the spinal extensor musculature to mitigate fracture risk.