Phenytoin therapy is a well recognized cause of gingival hyperplasia, a condition characterized by increased gingival collagen synthesis, and may also cause acromegalic-like facial features. Based on these clinical findings suggestive of anabolic actions, we sought to test the hypothesis that phenytoin acts on normal bone cells to induce osteogenic effects. To test the direct actions of phenytoin on human bone cells, we measured the dose responses to phenytoin for [H-3]thymidine incorporation, cell number, alkaline phosphatase specific activity, and collagen synthesis in human hip bone-derived cells. Phenytoin significantly and reproducibly increased [H-3]thymidine incorporation, cell number, alkaline phosphatase specific activity, and collagen synthesis in a biphasic manner with optimal stimulatory doses between 5-10 mu mol/L. Thus, micromolar concentrations of phenytoin can act directly on human bone cells to stimulate osteoblast proliferation and differentiation. We next sought to test the hypothesis that phenytoin stimulates bone formation in humans in vivo. Accordingly, three serum biochemical markers of bone formation, i.e. osteocalcin, skeletal alkaline phosphatase, and procollagen C-terminal extension peptide, were measured in 39 male epileptic patients, 20-60 yr of age, with an average duration of phenytoin therapy of 10.5 +/- 1.62 yr (mean +/- SEM). In this group of patients, the mean serum phenytoin level was 9.56 +/- 0.90 mg/L (mean +/- SEM; equivalent to 34.9 +/- 3.3 mu mol/L). Thirty apparently healthy male subjects of similar age and taking no medication were included as controls. Serum calcium, 25-hydroxyvitamin D-3, and PTH levels in the phenytoin-treated patients were not significantly different from those in the age-matched controls and were within the clinical laboratory normal range of our hospitals, indicating that the patients did not develop hypocalcemia, vitamin D deficiency, or secondary hyperparathyroidism. Serum levels of osteocalcin, skeletal alkaline phosphatase, and procollagen peptide in the phenytoin-treated patients were significantly increased compared to those in the age-matched subjects; in each case these biochemical markers were significantly correlated with the serum phenytoin level, but not with the dose or duration of phenytoin treatment. These findings are consistent with the interpretation that phenytoin increases the bone formation rate in humans in vivo. In summary, we have shown that phenytoin 1) in vitro acts directly on normal human bone cells at micromolar concentrations to stimulate [H-3]thymidine incorporation, cell number, alkaline phosphatase specific activity, and collagen synthesis (i.e, proliferation and differentiation); and 2) in vivo increases the serum level of bone formation markers (i.e, osteocalcin, skeletal alkaline phosphatase activity, and procollagen peptide) in human patients in vivo. In conclusion, our in vitro and in vivo findings together demonstrate that phenytoin increases markers of osteogenesis for the human species.
