Page 395 - ILAE_Lectures_2015

48 fractures for every 10,000 women treated with enzyme inducers for one year, including 10
hip fractures, and four additional hip fractures in every 10,000 men32. Thus it does appear that
at least the older AEDs, including all enzyme inducers and valproate, are themselves associated
with a modest increase in risk of fractures, and evidence is accumulating to support cause and
effect. There is insufficient data to draw conclusions with respect to any of the newer AEDs,
though there is some evidence to suggest that enzyme inducers carry a higher risk

Biochemical markers
Prior to the last decade or so, nearly all published data had been in the form of case reports,
cross-sectional, or retrospective studies, and thus subject to potential biases. Nonetheless there
were some consistent findings.

Enzyme inducers, such as carbamazepine (CBZ), phenytoin (PHT), topiramate (TOP) and
phenobarbitone (PB), would be expected to increase hepatic vitamin D catabolism, increasing
the risk of osteomalacia. This may be exacerbated by additional effects on sex hormones2.
Phenytoin is thought also to impair directly gastrointestinal calcium absorption33. There are
now many studies (only the most recent of which are cited here as examples), including in
ambulatory children and adults, consistently demonstrating significantly increased bone
alkaline phosphatase34-36 (particularly with phenytoin), reduced 25-hydroxy vitamin D3
levels34,37,38, reduced serum calcium34,36, and mildly elevated serum PTH38 in patients on
enzyme-inducing AEDs compared with matched controls. Lamotrigine looks to have few
effects36. Conflicting data has been reported, with oxcarbazepine35,39,40. Other markers of bone
turnover also appear to be consistently elevated in patients on AEDs, both enzyme-
inducing15,41,42 and non-inducing, such as valproate (VPA)43,44. However, not all studies are
consistent in terms of specific markers or individual drugs, and in terms of detail there are
many conflicting results.

A number of prospective studies have also now been reported, as summarised in Table 115,45-
47. Many of these studies have also evaluated vitamin D status, and demonstrated that increased
bone turnover appears to be independent of the presence of hypovitaminosis D. It is thought
some AEDs, including CBZ and VPA, may have direct effects on osteoclast/osteoblast
activity48. Of note, perhaps unsurprisingly4940, both clinical50,51 and rat52-54 studies also suggest
that the ketogenic diet may also have a negative effect on bone health, with effects on calcium
and phosphorus.

Taken together, AEDs, particularly though not exclusively enzyme inducers, do appear to have
effects on biochemical markers of bone metabolism, offering a number of biologically
plausible mechanisms that might underlie increased fracture risk. Hypovitaminosis D is
considered an independent risk factor for fracture, and may have other consequences including
muscle weakness and increased liability to falls56, and thus should reasonably be considered a
‘warning’ of clinically significant metabolic bone disease. It is also worryingly prevalent
(4080%) in epilepsy populations, both in the developed57 and the developing world58, and
may develop very quickly (within months) of starting AEDs. However, other markers of
increased bone turnover in themselves are not consistently associated with reduced BMD either
in adults, or children43. ‘AED bone disease’ should probably not be considered synonymous
with osteoporosis, which is supported by histomorphometric data, albeit limited, illustrating
increased bone remodelling59 and not necessarily decreased cortical bone mass. Thus, whilst
illustrating that there are changes to bone metabolism, the clinical significance of many of the
biochemical findings, with the exception of hypovitaminosis D, is currently uncertain, and
requires further study.