We describe five patients with a KCNJ11 mutation developing tooth discoloration 1–55 months after transfer from insulin onto glibenclamide. The severity of this novel side effect varied from easily removable tooth staining to nonreversible discoloration and loss of enamel.
Tooth discoloration has not previously been described despite widespread use of glibenclamide in adults. There are many possible explanations for this. First, our patients are much younger than patients with type 2 diabetes, and tooth discoloration is more noticeable in white deciduous than in the permanent teeth, which tend to be darker. Second, the doses used in children are usually higher than the maximum doses used in adults (3
). However, there seems to be no clear relationship between glibenclamide dose and the development of tooth discoloration within our cohort, since patient 2 was on a low dose (0.1 mg · kg−1
), and no tooth discoloration was noted in a further 62 patients with KCNJ11
diabetes who were successfully managed on similar doses of sulfonylureas. Third, and most likely, the teeth may have been exposed to high local concentrations of glibenclamide because of tablets being chewed or taken in solution. In keeping with this, most evidence indicates that the cause of tooth staining is the precipitation of ingested chromogens onto dental surface (5
). However, the possible pathogenic mechanism for the more severe effect on enamel seen in patient 2 remains unclear. Many other pediatric liquid medicaments have an erosive effect on the primary enamel surface (6
). In addition to this local effect, it may relate to a decrease in blood flow to the teeth, since glibenclamide, a nonselective sulfonylurea, reduces blood flow to the dental pulp by 70% (7
) by acting on vascular KATP
channels (composed of Kir6.1 and SUR2B) (8
). It might also be possible that loss of enamel is unrelated to sulfonylurea therapy, since it was present in deciduous teeth but not in permanent teeth.
Clinicians should be aware of this novel side effect of glibenclamide therapy in patients with neonatal diabetes resulting from a KCNJ11 mutation. While the cause is uncertain, patients should probably be advised not to chew tablets. Although the effect seems to have mainly a cosmetic consequence and should thus not preclude such treatment, this previously unreported association warrants further investigation.