Endocrine Abstracts

Research in recent years has shown that biochemically euthyroid levels of thyrotropin (TSH) and serum free thyroxine (FT4) as determined by laboratories may not necessarily equate to clinical well-being of patients. The defined normal range of TSH and FT4 are wide and may differ from patients’ optimal range of TSH and FT4. We have previously described a deterministic mathematical model that can predict each patient’s unique homeostatic set point, which gives the optimal TSH and FT4 values of each patient. This model is built on a parameterised inverse exponential relationship between TSH and FT4. In this retrospective study of 142 patients who had undergone thyroidectomy, we used the model to compute the predicted TSH and FT4 values based on thyroid function tests obtained post-thyroidectomy. The predicted values were then compared with the average TSH and FT4 obtained pre-thyroidectomy, when the patients were clinically euthyroid. Bland-Altman analysis of the differences between the predicted values and the average pre-thyroidectomy values shows that there is a mean difference of 0.15 mU/l, 95% CI (−0.03, 0.34), between the predicted TSH and average pre-thyroidectomy TSH and a mean difference of 2.87 pmol/l, 95% CI (2.55, 3.24), between the predicted FT4 and average pre-thyroidectomy FT4. Despite minor differences between the predicted values by the model and the average pre-thyroidectomy values of the patients, our mathematical model is able to narrow the range at which a patient can be considered as being euthyroid. Therefore, the personalised euthyroid targets derived from the model is able to provide clinicians with more guidance in managing patients who remain symptomatic despite having normal TSH or FT4 values.