Page 396 - ILAE_Lectures_2015

Table 1. Prospective studies assessing bone health markers in relation to antiepileptic drug

exposure.

Ref Population Drugs (n) 1st and 2nd time Controlled for BMD, site Main biochemical

point on drug confounders* findings

15 Children and CBZ (60) 0 and 2y Exercise, Vit D ND ↑ turnover
adolescents,
range 6-19y

38 Adult men, mean Any AED (81; Variable and + Smoking, ↓ femur  all
1.8%/yr
45y, range 25 – most CBZ, mean 19m alcohol, diet,

54y PHT, or VPA) (range 12-29) exercise, other

drugs

46 Adults, mean VPA (50) Mean 6.7 +/- 4y Alcohol, ↓ lumbar ND

28.9 +/- 5y Control (60) and +6m smoking, coffee, and femur

Mean 30.4 +/- diet, exercise  duration
5.6y

45 Children, mean CBZ or VPA 0 and >1y Diet, exercise ND  most, ↓ Vit D
season
7.4 +/- 3.3y (51)

Control (80)  all

55 Children, mean VPA (15), 0 and 2y BMI  lumbar  all

7.8 +/- 3.7y, CBZ(11),

range 3-15.5 PB(4)

47 Adults, range 18 - CBZ (10) 0 and 6m BMI, diet, ↓  most ↓ Vit D
50y VPA (15) exercise calcaneus  incl Vit D, ↑ ctn
LTG (8)  incl Vit D, ↑ ctn




52 Orchidectomised LEV (8) 0 and 12 weeks - ↓ femur  most, ↓ OPG, ↑
CTX1
adult rats Control (8)

 all

49 Adults, mean LEV (61) 0 and 14.1+/- BMI, diet, ↑ lumbar  all
31.0 +/- 13.1y 3.4m exercise  other

53 Adult mice PHT (6) 0 and 4m ↓ lumbar  most, ↓AlkP ↓ HxP

VPA (6) ↓ lumbar  most, ↓AlkP ↓ HxP
LEV (6)   all

40 Adults, mean OXC (41) 0 and 11.6 +/- BMI, diet, ↓ lumbar  most, ↓ Ca & AlkP

28.2 +/- 8.4y 6m exercise

*All controlled for gender, body mass index and age

y = years; m = months; n = number; CBZ = carbamazepine; LEV = levetiracetam; LTG = lamotrigine;

PB = phenobarbital; VPA = Valproate; ND = not done; Vit = vitamin;  no significant change; AlkP =

alkaline phosphatase; BMD = bone mineral density; ctn = calcitonin; CTX1 = cross-linking telopeptide

of type I collagen; HxP = hydroxyproline; OPG = osteoprotegerin

BMD
Given the limitations of biochemical markers, BMD remains the gold standard in terms of
assessing fracture risk, monitoring disease and treatment effects in metabolic bone disease, and
several studies, mostly cross-sectional or retrospective, have now reported on BMD in epilepsy
patients. Many of the studies claiming a significant reduction in BMD with both enzyme-
inducing AEDs60,61 and VPA44,62 use non-validated methods/sites and are thus difficult to
interpret. However those that use DEXA scanning at appropriate sites (spine, hip), and take
adequate care to control for confounders, mostly support that AED use is independently
associated with reduced BMD, at least in adults on older AEDs. One study found a significant
association only for PHT41, but most show reduced BMD in adults on any of the older AEDs
(PHT, PB, CBZ, VPA)37,38,63 in whom up to 59% are classified as osteopenic and 23% as
osteoporotic by WHO definitions. However the studies are inconsistent as to the size of any
AED effect and whether or not this reduction correlates with either duration of AED therapy