LETTER TO EDITOR Lack of imatinib-induced thyroid dysfunction in a cohort of non-thyroidectomized patients Jose ´ Miguel Dora, Murilo Anderson Leie, Bruno Netto, Laura Maria Fogliatto 1 , Lucia Silla 1 , Felipe Torres 2 and Ana Luiza Maia Endocrine Division, Thyroid Section, Hospital de Clı ´nicas de Porto Alegre, Universidade Federal do Rio Grande do Sul, Porto Alegre, Brazil, 1 Hematology Division and 2 Radiology Division, Hospital de Clı ´nicas de Porto Alegre, Porto Alegre, Brazil (Correspondence should be addressed to A L Maia who is now at Servic ¸o de Endocrinologia, Hospital de Clı ´nicas de Porto Alegre, Ramiro Barcelos 2350, 90035-003 Porto Alegre, RS, Brazil; Email: almaia@ufrgs.br) To the Editor Understanding the role of tyrosine kinase (TK) proteins in the pathogenesis of various tumors triggered the development of drugs that specifically block TK actions (1). Imatinib, one of the drugs in that class, has favorably altered the natural history of chronic myeloid leukemia (CML) and gastrointestinal stromal tumors (1). However, given the short experience with these drugs, the full spectrum of adverse effects remains unknown. Cardiotoxicity (2) and bone metabolism alterations (3) have been reported in association with TK inhibitor therapy. Recently, high rates of hypothy- roidism have been described in patients receiving sunitinib (4, 5). In addition, adjustment of thyroid hormone replacement in thyroidectomized patients under sunitinib or imatinib therapy, with the need of up to 350% increase in levothyroxine dose, has also been reported (6). Due to similar mechanisms, it is possible that imatinib is also associated with thyroid dysfunction in non-thyroidectomized patients. Because of the widespread use of this drug, we sought to evaluate the impact of imatinib therapy on thyroid function. Patients with CML under imatinib therapy, followed at the Hematological Division of Hospital de Clinicas de Porto Alegre (Porto Alegre, Brazil), between March and October 2007, were eligible for the study. Patients who used drugs with potential interference in thyroid function tests 6 months prior to the study entry or with known previous thyroid dysfunction were excluded. All measurements were performed using ECLIA (Roche). Interassay coefficients of variation were as follows: thyrotrophin (TSH), 1.6%; thyroxine (T 4 ), 3.5%; (free T 4 ), 3.0%; tri-iodothyronine (T 3 ), 3.4%; Tg, 1.9%; and anti-TPO, 7.1%. Fifty-four patients underwent thyroid ultrasound (US) for thyroid volume estimation. Eleven patients underwent a 24-h radio- iodine uptake (RAIU). A total of 70 patients were eligible to enter in the study. Two patients were excluded because of a previous diagnosis of hypothyroidism. Thus, 68 patients were included. Table 1 shows the clinical and laboratory characteristics of study participants. All study subjects displayed levels of T 4 , FT 4 , and T 3 in the normal range. Serum TSH was in the normal range of 63/68 (92.6%) and slightly elevated in 5 patients (range 5.08–12.55 mU/l). The serum TSH levels before and after imatinib therapy, available for a subgroup of ten patients, were similar (2.39 (1.82–3.05) vs 2.71 (1.71–3.25) mU/l, PZ0.64). Furthermore, there was no correlation between serum TSH levels and dose (rZK0.043, PZ0.73), duration of therapy (rZK0.084, PZ0.50), or cumulative dose of imatinib (rZK0.105, PZ0.39). Because of a previous study report supporting thyroiditis in patients under sunitinib therapy (4, 7), Tg and anti-TPO were also measured. Serum Tg was at normal levels and only one patient displayed positivity for anti-TPO. Thyroid volume and RAIU were in the normal range. The power calculation of our study was the following: for correlations, considering an a error of 0.05 and a b error of 0.20, estimations of r values of 0.3, 0.4, and 0.5 resulted in calculated required samples of 84, 46, and 29 patients respectively (NCSS Statistical & Power Analysis Software 2007, Kaysville, UT USA). Despite previous positive reports, here we found no influence of imatinib on thyroid function. The lack of correlation between TSH levels with dose, duration, or cumulative dose of imatinib therapy suggests that this drug has no adverse effect on thyroid function. In addition, the prevalence of subclinical hypothyroidism in this sample was w10.0%, similar to that reported for our population (8). Imatinib exerts its effects through inhibition of multiple TKs, including Bcr-Abl, platelet-derived growth factor receptors a and b, c-Fms, and c-kit (1). The increased demand for levothyroxine induced by imatinib in patients under levothyroxine replacement (6) might indicate increased peripheral metabolism of thyroid hormones. As suggested by others (6), induction of hepatic conjugation with glucoronates and sulfates would be a possible explanatory mechanism. Another potential explanation could be increased hormonal deiodination to reverse T 3 through enhanced deiodinase type 3 (D3) activity. Because both metabolic pathways would imply an increase in thyroid hormone output to overcome the increased peripheral clearance, it would be expected to find borderline-low values of T 4 and T 3 and borderline-high values of TSH in non-thyroidectomized European Journal of Endocrinology (2008) 158 771–772 ISSN 0804-4643 q 2008 Society of the European Journal of Endocrinology DOI: 10.1530/EJE-08-0006 Online version via www.eje-online.org