Risk of Death and Adverse Effects in Patients on Liothyronine

Suhani Bahl, Peter N Taylor, Lakdasa D Premawardhana, Mike Stedman, Adrian Heald, Colin M Dayan, Onyebuchi E Okosieme
The Journal of Clinical Endocrinology & Metabolism, Volume 110, Issue 11, November 2025, Pages 3278–3288
https://doi.org/10.1210/clinem/dgaf449

Abstract

Context

Although some patients with hypothyroidism prefer combination therapy with liothyronine (LT3) and levothyroxine (LT4), the safety of LT3 remains unresolved.

Objective

We undertook a multisource systematic review and meta-analysis of LT3 safety.

Data Sources

We searched PubMed for articles relating to death, adverse events (AEs), and cardiovascular outcomes in LT3 users. We also searched AEs data in the UK Yellow Card scheme and US Food and Drug Administration Adverse Reporting System (FAERS).

Data Extraction

Data was extracted independently by 2 reviewers. Out of 1814 articles identified, 52 studies were selected, comprising 21 randomized controlled trials (RCTs), 4 cohort studies, and 27 case reports. Meta-analyses were conducted for adverse outcomes in RCTs and cohort studies of combination vs monotherapy.

Data Synthesis

LT3-related AEs were only reported with unregulated LT3 use or pharmacy compounding errors. LT3 and LT4 showed similar adverse severity profiles in the Yellow Card scheme. Disproportionality analysis in the FAERS database showed no increased LT3 safety signals. A meta-analysis of RCTs (n = 2128) showed a similar AEs risk for combination vs monotherapy [relative risk (RR) 1.22, 95% confidence interval (CI) 0.66–2.25]. A cohort study meta-analysis (LT3 vs LT4-only users, n = 630,254) showed no increased risk of atrial fibrillation (RR 1.10, 95% CI 0.74–1.63), heart failure (RR 1.54, 95% CI 0.95–2.47), or strokes (RR 0.86, 95% CI 0.11–6.75), but reduced mortality risk was observed for LT3 (RR 0.70, 95% CI 0.62–0.78).

Conclusion

Our findings are reassuring that regulated LT3 use is not associated with the risk of death or serious AEs. More studies are needed to supplement existing data.

You may also like...