An updated Cochrane review (abstract , review [Abstract]) included 26 studies with a total of 3 388 patients. Four trials examined the effect of duration of therapy on relapse rates of Graves' hyperthyroidism. In one trial using the titration regimen, longer duration therapy (18 months) had significantly fewer relapses (37% vs 58%) than six month therapy (OR 0.42, 95% CI 0.18 to 0.96). In one quasi-randomised trial using the block-replace regimen, there was no significant difference between the six and 12 month (relapses rates 41% versus 35%) regimens. Two other studies used even longer durations of therapy; extending the duration of therapy to over 18 months was not associated with improved relapse rates (Peto OR 0.75, 95% CI 0.39 to 1.43; 2 studies, n=186).
Twelve trials examined the effect of block-replace versus titration block-regimens. The relapse rates were similar in both groups (51% in the block-replace group vs. 54% in the titration block-group; OR 0.86, 95% CI 0.68 to1.08). Participants reporting rashes (10% versus 5%) and withdrawing due to side effects (16% versus 9%) were significantly higher in the block-replace group.
Three studies considered the addition of thyroxine with continued low dose antithyroid therapy after initial therapy with antithyroid drugs. There was significant heterogeneity between the studies and the difference between the two groups was not significant (OR 0.58, 95% CI 0.05 to 6.21).
Four studies considered the addition of thyroxine alone after initial therapy with antithyroid drugs. There was no significant difference in the relapse rates between the groups after 12 months follow-up with relapse rates being 31% (88/282) with thyroxine and 29% (82/284) with placebo (OR 1.15, 95% CI 0.79 to 1.67).
Two studies considered the addition of immunosuppressive agents. The results which were in favour of the interventions would need to be validated in other populations.
Comment: The quality of evidence is downgraded by study quality (inadequate or unclear allocation concealment).
The following decision support rules contain links to this evidence summary: