Efficacy of thyroid hormone suppression for benign thyroid nodules: meta-analysis of randomized trials.

OBJECTIVE To determine the efficacy of thyroid hormone suppressive therapy (THST) to decrease benign thyroid nodule volume. DESIGN Meta-analysis. METHODS Systematic search using electronic databases (PubMed, Medline, Cochrane Library) through August 2004, paper review, and contacting experts and drug manufacturers. Only randomized controlled studies of THST vs no treatment or placebo, for reduction of benign thyroid nodule volume, were included. Exclusion criteria were: <6-month treatment, lack of ultrasound volume measurement, and region of endemic goiter. Primary outcome was clinically relevant nodule volume reduction (>50%), with a random effects model (RevMan4.2). RESULTS Nine randomized trials were included (609 subjects). Subjects were 88% more likely to experience >50% nodule volume reduction with THST than placebo or no treatment (relative risk = 1.88; 95% CI = 1.18-3.01; P = 0.008). However, 8 subjects must be subjected to the risk of cardiac and skeletal side effects from THST, for one to benefit from therapy (number needed to treat = 8, risk difference = 0.13; 95% CI = 0.06-0.19; P = 0.0003). Sensitivity analysis reveals that 15 null studies would have to have been missed to reverse statistical significance (fail-safe N = 15). Review of the only study with long-term treatment (5 years) suggests no significant difference in nodule volume reduction between THST and placebo. Further, studies with follow-up after THST withdrawal demonstrate rapid increase in thyroid nodule and goiter volumes. CONCLUSION THST appears more likely than placebo or no treatment to significantly reduce benign thyroid nodule volume, but long-term treatment may be less effective and regrowth is likely following cessation of therapy. Given the risks of THST, routine use is not recommended for benign nodules.