Abstracts

Rationale: Non-enzyme inducing antiepileptic drugs (NEIAEDs) have been associated with changes in bone quality in an animal study and significant bone loss in older men after one year of treatment. We aimed to evaluate bone mineral density (BMD) and indices of structural integrity or bone quality using dual energy X-ray absorptiometry (DXA) and high-resolution peripheral quantitative computed tomography (HR-pQCT) in older men with epilepsy treated with NEIAEDs. Methods: Older men (aged 50 - 70) with epilepsy treated with NEIAEDs (gabapentin, lamotrigine, levetiracetam, pregabalin alone or in combination) for at least 6 months were enrolled (n=15). A control group was also recruited (n=8). Men with known bone disease, medical conditions known to affect bone (liver disease, renal disease, endocrinopathy) or taking medications known to affect bone (glucocorticosteroids, anticoagulants, methotrexate) were excluded. Areal BMD (aBMD) was measured by DXA (Hologic QDR 4500) and expressed as absolute BMD (g/cm2) and T scores at the lumbar spine, total hip, femoral neck, distal 1/3 and ultradistal radius. Volumetric BMD (vBMD), trabecular and cortical microarchitecture were measured at the nondominant distal radius and tibia by HR-pQCT using the SCANCO Xtreme CT (resolution, ~82 microns). Measurements obtained included total area, total density, cortical thickness, and trabecular density, number, thickness, and separation. Results: Eight men with epilepsy were on lamotrigine, 2 on levetiracetam, 2 on gabapentin, and 5 were on a combination of NEIAEDs. Men on NEIAEDs and control men did not differ by average age (58.0 ± 3.7 vs 57.8 ± 4.8 years), proportion of Caucasians, height, weight, or BMI. Areal BMD by DXA and T scores, unadjusted or adjusted for age and weight, did not differ between groups at any bone site. In contrast, age and weight adjusted total vBMD by HR-pQCT was 11.7% lower at the radius in NEIAED treated men than controls (316 + 55 vs 358 + 55 g/cm3, respectively; p = 0.03). The deficit was attributable to reduced cortical thickness (by 15%; 0.829 + 0.167 vs 0.978 + 0.205 mm; p = 0.07) as trabecular parameters did not differ. Findings at the distal tibia were similar to those at the radius: lower total vBMD and cortical thickness (both, p=0.07) with no difference in trabecular measurements. Conclusions: In summary, although areal BMD by DXA could not distinguish between men with epilepsy treated with NEIAEDs and normal controls, HR-pQCT revealed significantly lower total volumetric BMD at the radius and a trend toward lower volumetric BMD at the tibia in the cases. The deficits appeared to be related to decreased cortical thickness rather than to abnormal trabecular microarchitecture. These results suggest that cortical bone may be adversely affected in men with epilepsy treated with NEIAEDs and raise concern regarding increased risk of fracture with prolonged NEIAED exposure. Supported by NIAMS 1 K23 AR053113