TRH stimulation test in patients with repeatedly elevated TSH and normal FT4

Subclinical hypothyroidism (SCH) is defined by elevated thyroid-stimulating hormone (TSH) with normal free thyroxine (FT4). We aimed to evaluate the thyrotropin-releasing hormone (TRH) stimulation test in patients with repeatedly elevated TSH (up to 10 mIU/l) and normal FT4, but without apparent thyroid disease. Women with TSH > 4.5 and ≤ 10 mIU/l (in two measurements) and normal FT4 were selected. Women with a known non-thyroid cause of TSH elevation, those treated with anti-thyroid drugs, amiodarone, lithium, and those with a history of thyroidectomy, neck radiotherapy and 131 I treatment were excluded. Seventy women had negative antithyroperoxidase antibodies. Ultrasonography revealed a eutopic thyroid, usual echogenicity, and a volume ≤ 15 ml, and they underwent the TRH stimulation test during initial evaluation. After stimulation with TRH, TSH > 30 mIU/l was observed in 38 women (expected response), while 32 women had TSH < 20 mIU/l (inadequate response). Age, basal TSH or thyroid volume did not differ between both groups, but FT4 concentrations were significantly lower in the first group. Follow-up was available for 66/70 women. Seven women developed a need for levothyroxine, all of them in the group with an adequate response to TRH [7/36 (19.4%) versus 0/30]. The results suggest that some cases of TSH elevation (even persistent) do not represent the early stage of thyroid insufficiency.

introduction Subclinical hypothyroidism (SCH) is traditionally defined as serum thyroid-stimulating hormone (TSH) elevation in the presence of circulating thyroid hormone concentrations within the reference range (1)(2)(3)(4) . It is known that not all cases of hyperthyrotropinemia are related to thyroid insufficiency. Some conditions that can elevate serum TSH are easily recognized: untreated adrenal insufficiency, chronic renal failure, antidopaminergic drugs, and recovery from severe acute disease. Additionally, slight TSH elevation is seen in older adults (> 70 years) when the normal range obtained for adults is used. Obese individuals may also exhibit slightly elevated TSH. Elevation of TSH can, likewise, be the result of a predominance of TSH isoforms with low biological activity or analytical interferences from macro TSH and human anti-mouse antibodies. Finally, even within the population reference interval, the concentrations of free thyroxine (T4) above and below which TSH secretion is inhibited and stimulated, respectively, show wide individual variation and are genetically determined.
Palabras clave: hipotiroidismo subclínico; hormona liberadora de tirotropina; pruebas de función de la tiroides. and who are potential candidates for L-T4 therapy have in fact hyperthyrotropinemia unrelated to thyroid insufficiency and may be treated erroneously (5) . We previously showed that this applies to a significant number of patients (6) . Even after confirmation of the result and exclusion of known non-thyroid causes of TSH elevation, clinical history, TPOAb measurement and US did not disclose thyroid disease in approximately 20% of women with elevated TSH ≤ 10 mIU/l and normal free T4 (6) . However, currently no recommendation exists regarding additional tests in patients without apparent thyroid disease designed to distinguish individuals without thyroid dysfunction and with elevated TSH due to other reasons, such as those mentioned above, from individuals who actually have initial thyroid dysfunction (1)(2)(3)(4) . Individuals with TSH ≤ 10 mIU/l deserve greater attention because they account for approximately 80% of SCH cases (1,3,4) .
In patients with hyperthyrotropinemia due to thyroid insufficiency, a TSH hyper-response upon stimulation with thyrotropin-releasing hormone (TRH) is expected (7)(8)(9) . Indeed, this response can be seen early even when TSH is still within the normal range (10) . We evaluated the TRH stimulation test in women with repeatedly elevated TSH (up to 10 mIU/l) without a known cause of hyperthyrotropinemia and with normal free T4 concentrations, who could be diagnosed with SCH (1)(2)(3)(4) , but who did not exhibit apparent thyroid disease (based on clinical history, TPOAb, and US).
The aim of this study was to evaluate the usefulness of this test as a tool to distinguish patients without thyroid dysfunction and with elevated TSH due to other reasons from those who actually have initial thyroid dysfunction.

methods
Women not treated with L-T4 and with TSH > 4.5 and ≤ 10 mIU/l were initially selected. Women treated with antithyroid drugs, amiodarone, lithium and alpha-interferon and those with a history of thyroidectomy, neck radiotherapy or 131 I treatment for hyperthyroidism, or with a known non-thyroid cause of TSH elevation (untreated adrenal insufficiency, chronic renal failure, antidopaminergic drugs, recovery from severe acute disease) were excluded. The remaining patients were reevaluated by measurement of TSH and free T4 after 8-12 weeks. Finally, among patients with persistently elevated TSH ranging from 4.5 to 10 mIU/l and normal free T4, 70 had negative TPOAb and US revealed a eutopic thyroid, usual echogenicity, and a volume ≤ 15 ml (group of interest) (6) . All of these women were ≤ 70 years and had a body mass index < 35 kg/m 2 (6) .
TSH, TPOAb and free T4 were measured with a chemiluminescent assay (Immulite 2000, Diagnostic Products Corporation, Los Angeles, CA), with reference values of 0.4-4 mIU/l, up to 35 kIU/l and 10.3-23 pmol/l, respectively. TRH stimulation tests were performed in the fasting state (8-10 hours). All patients received 200 mg TRH intravenously as a bolus injection. Blood samples were obtained at baseline, and 30 and 60 minutes after TRH stimulation for TSH analysis (7,10,11) .
Statistical analysis was performed using Fisher's exact test. A p-value of < 0.05 was considered significant.

results
After stimulation with TRH, TSH > 30 mIU/l (31 to 52 mIU/l) was observed in 38 women. This was considered an expected response (7,10,11) . In contrast, 32 women exhibited peak TSH < 20 mIU/l (8 to 17 mIU/l), which was considered an inadequate response (7,10,11) . Age (p = 0.85), basal TSH (p = 0.2) or thyroid volume on US (p = 0.3) did not differ between the two groups. Although within the normal range in all women, free T4 concentrations were significantly lower in those with an expected TSH response (p = 0.01).
Twenty women reported adverse reactions during the TRH test: nausea (n = 8), bad taste in the mouth (n = 8), urinary urgency (n = 4), and hot flashes (n = 5). All symptoms occurred immediately after the infusion and only lasted a few minutes.
Follow-up was available for 66 women (6) . After a follow-up of seven to 14 years (median 10 years), only seven women exhibited progression of SCH (TSH > 10 mIU/l accompanied by reduced free T4 or two TSH values > 10 mIU/l within an interval of eight weeks) (6) . Progression of SCH was only observed in women with an expected TSH response on initial evaluation [7/36 (19

discussion
We believe that the inadequate response to TRH stimulation itself already indicates hyperthyrotropinemia unrelated to primary thyroid insufficiency (7)(8)(9) . This hypothesis was supported by the difference in free T4 concentrations and progression to hypothyroidism found between women with an "expected" response versus those with "inadequate" response.
When the study began, we considered the TRH stimulation test to be the best parameter for identifying cases of hyperthyrotropinemia not secondary to thyroid dysfunction. Although the response of serum TSH to TRH shows individual variation, the results of the present study confirm that the traditional peak TSH cut-off of 30 mIU/l (7,10,11) identifies individuals with true thyroid dysfunction. Some years after the start of the study, polyethylene glycol (PEG) precipitation became the focusof the literature (12,13) . It is reasonable to currently begin the investigation for hyperthyrotropinemia without apparent cause by PEG precipitation, a simple, noninvasive and low-cost method (12,13) . However, this method does not detect all causes of elevated TSH not resulting from initial thyroid dysfunction. Indeed, we recently performed PEG precipitation following the protocol of Hattori et al. (2016) (12) in 12 women with an inadequate response to TRH on initial evaluation and persistently elevated TSH, and precipitation was normal (< 70%) in nine of them (data not shown).
Regardless of the screening method, the study calls attention to the possibility that, in the absence of underlying thyroid disease, a considerable number of cases with repeatedly elevated TSH and normal free T4 concentrations, even without a known nonthyroid cause of this alteration, do not represent the early stage of thyroid insufficiency but rather hyperthyrotropinemia of another nature. We therefore agree that much care should be taken when diagnosing SCH in the absence of apparent thyroid disease (5,6) . Although we did not study this group, this concern certainly also applies to patients with TSH > 10 mIU/l (12,13) . In the present study, age, basal TSH and thyroid volume were not useful for identifying these cases, reinforcing the need for specific investigation. If not for all, this investigation should at least be performed for patients who are candidates for L-T4 therapy, so that they are not treated erroneously.