Letter to the Editor: Hennessey Response to Fitzgerald et al. (DOI: 10.1089/thy.2019.0535)

Dear Editor:

Fitzgerald et al. have provided us with a unique and extensive review of the medical literature regarding the correlation of circulating thyroid hormone levels, thyrotropin (TSH) and thyroid-related outcomes (1). We are reassured that TSH abnormalities, which direct us to consider over- or underactivity of the thyroid gland, are indeed the first step in clinical decision making. Logically, outcomes correlate better with the tissue level and circulating thyroid hormone levels we use to classify thyroid dysfunction into subclinical and overt.

The authors extrapolate their findings, suggesting that TSH levels should not be preferred over thyroid hormone levels in monitoring thyroid treatments (1). I agree in cases of thyrotoxicosis, where TSH levels remain suppressed as therapeutic progress is made, and recovery is measured by assessment of free thyroxine (fT4) and triiodothyronine (T3). However, I disagree in those with hypothyroidism. The authors point out, the use of a controlling hormone to diagnose and differentiate thyroid disease is less than optimal (1). I certainly agree. In contrast, in a patient on replacement therapy for primary hypothyroidism, TSH is no longer the controlling hormone, rather TSH reflects the impact of activated thyroid hormone at a tissue level.

There are few if any other parameters so readily available for our assessment of thyroid hormone therapy. Secondly the type of thyroid hormone replacement dictates which circulating thyroid hormone levels might best indicate adequacy of thyroid hormone replacement. For example, levothyroxine (LT4) treatments result in higher circulating fT4 levels than a normalized TSH would seem to justify (2,3). This potentially euthyroid hyperthyroxinemia may result from blood sampling at the peak period after oral absorption or because of unique characteristics of LT4 metabolism ingested exogenously. These observations could result in less than optimal dosing should the author's favored fT4 level be substituted for TSH. In those on LT4 monotherapy, many have demonstrated normalization of circulating T3 levels (2,3) while others have found lower levels of T3 (4) in presumably euthyroid LT4 treated subjects. Another issue arises with liothyronine, either alone or in combination with LT4 or in the form of thyroid hormone extract. Postabsorptive peaks of T3 in individuals with normal TSH may simulate over replacement (4) and preabsorptive nadirs in T3 may also trigger dose adjustments that might not be in the patient's best interest. Thyroid hormone extract use results in fairly low circulating fT4 levels while TSH is normal (4) again potentially resulting in dose adjustment that may not be of benefit to the patient.

TSH gives us an integrated answer as to the adequacy of replacement therapy and sustained normal TSH levels are associated with mortality comparable with background populations (4), it would seem that maintaining a more simplified approach, as we currently do is in the best interest in our patients and avoids much of the confusion that continues to be associated with thyroid function testing in nonendocrine circles.