(CSEMP070) TREATMENT OF SUBCLINICAL HYPERTHYROIDISM AND INCIDENT ATRIAL FIBRILLATION

Abstract:

Background: Treating overt hyperthyroidism can prevent and leads to resolution of atrial fibrillation (AF). While subclinical hyperthyroidism (SH) is also associated with AF, it is unknown if treating SH can prevent AF.

Objective: We aimed to assess the association between treatment of SH and incident AF.

Methodology: In a single-centre retrospective cohort study, patients ≥ 18 years with biochemical SH were identified from the Regional Laboratory database 2000–2021. Those with prior AF, thyroid or pituitary disease, or pregnancy were excluded. Patients treated for SH (medications, radioactive iodine, or surgery) were compared to untreated patients. The primary outcome was incident AF. The secondary outcomes were ECG p wave duration and echocardiographic ventricular and ascending aorta size. More rigorous pharmacoepidemiologic analyses to account for immortal time bias will follow.

Results: 360 of 5567 patients with SH met inclusion criteria (131 (37%) treated, 229 (64%) untreated).

5 treated and 15 untreated SH patients developed AF (3.8 % and 6.6%, respectively, p=0.26). In the treated group, we excluded time from SH diagnosis to SH treatment: 224.49 person-years (py) (median 0.56 years; IQR 0.19, 2.01) from analysis. Follow-up time in the treated group from SH treatment start was 496.36 py (median 2.85 years; IQR 1.26, 5.31), and in the untreated group from SH diagnosis was 920.44 py (median 3.12 years; IQR 0.82, 6.18). Incidence rate of AF was 1.0 %/year in the treated and 1.6 %/year for the untreated group (IRR 0.62, 95%CI 0.17-1.79, p=0.62). Mean p-wave duration, ventricular size and ascending aorta size in the treated vs untreated groups were 88.97 vs 84.95 ms (p=0.15), 4.18 vs 4.31 cm (p=0.81), and 3.36 and 3.21 cm (p=0.24), respectively.

In a sensitivity analysis, patients diagnosed with AF < 30 days after starting treatment for SH were included in the untreated group. In these treated and untreated groups, 3 versus 17 patients developed AF (2.3% versus 7.4% respectively, p=0.04), and incidence was 0.6%/year in the treated and 1.8%/year in the untreated group (IRR 0.33, 95%CI 0.06-1.14, p=0.06).

Discussion: Although there was an overall trend towards less AF in the treated group, there were no significant differences in AF incidence, or ECG or echocardiographic characteristics following SH treatment. Our small retrospective study does not support that treating SH directly prevents AF. Future larger prospective studies are required to further assess this question.