B. E. W. Brownlie

Lithium associated thyrotoxicosis: a report of 14 cases, with statistical analysis of incidence

OBJECTIVE Lithium is known to cause goitre and hypothyroidism, and has been associated less commonly with hyperthyroidism. We report a series of 14 patients with lithium associated thyrotoxicosis (LiAT), and have used epidemiological data to assess the association between long‐term lithium treatment and the development of thyrotoxicosis.

The epidemiology of thyrotoxicosis in New Zealand: incidence and geographical distribution in north Canterbury, 1983-1985.

In a 3‐year (1983‐1985) epidemiological study of thyrotoxicosis in North Canterbury, New Zealand, the annual incidence was 25.8 per 100 000 (female 40.7, male 10.5). Thyroid scintiscanning showed that 64% had diffuse hyperplasia (DH), 27% toxic multinodular goitre (TMG), 7% toxic uninodular goitre (TUG), and 2% zero uptake. The calculated annual incidence of toxic diffuse goitre (DH) was 15 per 100 000, and for toxic nodular goitre (TMG and TUG combined) was 8 per 100 000. The age‐related incidence for toxic diffuse goitre peaked in middle life whereas toxic nodular goitre showed an increasing incidence with age. There was no significant seasonal variation or rural/urban difference in incidence. Analysis of geocoded addresses did not identify areas of high incidence. The variable duration of symptoms prior to diagnosis limits the search for possible environmental trigger factors. North Canterbury was an endemic goitre area prior to the introduction of iodized salt 50 years ago, and the incidence of toxic nodular goitre is likely to fall in future.

LITHIUM AS AN ADJUNCT TO RADIOIODINE THERAPY FOR THYROTOXICOSIS

16 patients with diffuse thyroid hyperplasia were given lithium carbonate (400 mg daily) for 1 week before and 1 week after a standardised 5 mCi therapy dose of 131I. A comparable control group of 16 patients were treated with 5 mCi of 131I without lithium therapy. The % retention of the therapy dose was measured in all patients at 7 days (168-hour 131I uptake). In the lithium-treated group the 24-hour 131I uptake showed no significant change after the first week of lithium therapy. The mean 48-hour protein-bound 131I, however, fell considerably from 1-21 to 0.55%/dose/1. The mean 24-168 hour % thyroidal 131I uptake drop was significantly less in the lithium group. These results show that low-dosage lithium therapy increases the retention of a standard-therapy dose of 131I. Lithium promises to be a useful adjunct to 131I therapy in patients with a rapid thyroidal iodine turnover and particularly in young patients where the total body-radiation dose must be kept to a minimum.

Subclinical thyrotoxicosis in an outpatient population – predictors of outcome

Objective  Individuals with endogenous subclinical thyrotoxicosis (SCT) may subsequently require treatment for overt disease. We aimed to evaluate the frequency of progression to hyperthyroidism and factors influencing this outcome.

Lithium associated thyrotoxicosis

Lithium therapy for affective disorder is goitrogenic and may cause iatrogenic hypothyroidism. Despite lithium’s mild antithyroid action, thyrotoxicosis may occur during a long-term lithium treatment and we have previously reported 14 patients with lithium associated thyrotoxicosis (LiAT) with statistical analysis of incidence. A more recent report from North America has emphasised that thyrotoxicosis occurring during lithium therapy may be transient and due to painless thyroiditis. In this communication we report our recent experience in the management of 23 patients with LiAT – the 23 patients were identified from our thyroid clinic database over a 12-year period (1995–2006). The cohort included 20 females and three males with a median age of 41 year (range 19–75 year). None of the females were postpartum. These 23 patients made up 1Æ4% of thyrotoxic patients referred to our regional thyroid clinic over the 12-year period. For comparison, during this same period 4Æ8% of thyrotoxic patients had amiodarone-associated thyrotoxicosis. The LiAT patients were all currently receiving lithium maintenance therapy, and the median duration of lithium therapy was 6 years (range 0Æ6–25 year). The only patient receiving lithium for <1 year presented with subacute thyroiditis – with neck pain, mild hyperthyroidism and elevated inflammatory markers. Three patients had a past history of goitre – two had been investigated at our clinic some years previously when euthyroid for diffuse goitre during lithium therapy, and one patient had a long-standing multinodular goitre. Three patients had a first degree relative with thyrotoxicosis – two mothers had received radioiodine treatment, and a sister thyroidectomy. This is in contrast to our previous LiAT series in which none of the 14 patients had a family history of thyrotoxicosis. All 23 LiAT patients had elevated thyroid hormone levels, and suppressed TSH levels prior to outpatient referral. Patients were assessed clinically and had repeat thyroid function tests (total serum thyroxine, free thyroxine index, total serum T3, sensitive TSH), thyroid antibodies, and Tc pertechnetate thyroid scintiscans. The scan findings were: 10 patients with diffuse hyperplasia, two with toxic multinodular goitres, and 11 with inadequate thyroid visualisation due to very low Tc uptake – including nine patients with painless thyroiditis, the patient with subacute thyroiditis, and one patient with diffuse goitre and strong family history of Graves’ disease who later admitted to taking a dietary supplement containing kelp. Carbimazole treatment was initiated in nine of the 10 patients with diffuse hyperplasia, and the untreated patient was an asymptomatic young female with T3 toxicosis and large diffuse goitre. The two patients with toxic multinodular goitre also received carbimazole treatment. No antithyroid medication was given to the 11 patients with low Tc uptake scans (nine with painless thyroiditis, the subacute thyroiditis patient, and the kelp-ingesting patient) and thyroid hormone levels spontaneously returned into the normal range over a 6–10 week period. During follow up, six of the 10 patients with diffuse hyperplasia have required definitive treatment for persistent thyrotoxicosis – the two youngest patients by near-total thyroidectomy, and four received radioiodine therapy. One of the two patients with multinodular goitres has received radioiodine. None of the low Tc uptake patients have required definitive therapy, but five of the nine with painless thyroiditis, and the subacute thyroiditis patient are currently receiving thyroxine replacement. The patient who ingested kelp is currently euthyroid. Our recent experience confirms that a significant proportion of patients with LiAT have transient thyrotoxicosis, with 9/23 (39%) having investigations consistent with painless thyroiditis. This is a higher proportion than in our previous report where two of 14 LiAT patients had painless thyroiditis. The increase in painless thyroiditis cases in recent year may in part be due to the increased use of sensitive TSH measurements which results in the diagnosis of additional patients with mild hyperthyroidism. The higher incidence of painless thyroiditis during lithium therapy reported from North America was documented in a retrospective review of patients with very low thyroidal I radioiodine uptake results. The pathogenesis of LiAT remains uncertain but autoimmunity may play an important role in many patients. Antithyroid peroxidase antibodies were positive in five of the ten patients with diffuse hyperplasia, and two of the nine patients with painless thyroiditis. TSH receptor antibody results were available for recent diffuse hyperplasia patients and were positive in seven of nine patients. The thyroid histology for our two diffuse hyperplasia patients treated by thyroidectomy showed diffuse lymphocytic infiltration consistent with Graves’ disease. None of our painless thyroiditis patients had a thyroid biopsy, and the few published reports have shown granulomatous thyroiditis, lymphocytic thyroiditis or non-specific thyroiditis. Lithium is concentrated by the thyroid, and inhibits thyroid hormone synthesis and release, and this may cause expansion of the intrathyroidal iodine pool which could precipitate thyrotoxicosis in patients with a genetic predisposition to Graves’ disease or nodular thyroids. How lithium predisposes to painless thyroiditis remains unknown but a possible direct toxic effect of lithium on the thyroid has been suggested. Most of our LiAT patients were diagnosed following routine follow up thyroid function testing during lithium treatment, but patients with Graves’ disease usually present with more obvious clinical features, and one of our younger patients had atrial fibrillation. Closer surveillance of thyroid function has been recommended for lithium-treated patients with thyroid antibodies, goitre, or family history of thyroid disease. If elevated thyroid hormone levels or suppressed TSH levels are detected the investigation should be repeated, and if persistently abnormal referral for endocrine evaluation is indicated. We have found routine thyroid Clinical Endocrinology (2011) 75, 402–407

Results of lithium-131I treatment of thyrotoxicosis

The role of lithium as an adjunct to 131I therapy of thyrotoxicosis has been assessed. Seventeen lithium-131I treated patients and 16 control — 131I treated patients have been followed for almost three years. Five lithium — 131I and 3 control — 131l patients have become hypothyroid with the lithium treated patients developing earlier thyroid failure. Ten lithium and 10 control patients remain clinically and biochemically euthyroid. Two lithium — 131I and 2 control — 131I patients remain on low dose antithyroid drugs.

Thyroid vascularity and trapping function: Analysis of very early thyroidal technetium “uptake”

Abstract A method is presented for analysing flow curves recorded over the neck during the first three minutes following the injection of 99 m Tc-pertechnetate, using a gamma camera fitted with a pinhole collimator interfaced to a minicomputer. The analysis involves the identification of two distinct components in the flow curve, namely “thyroid vascularity” (derived from the height of the initial rise in the curve) and “uptake slope,” which is a sensitive measure of true thyroidal pertechnetate trapping function. Clinical results on 157 patients revealed that whereas the uptake slope correlates well with 10-min uptake and other established thyroid function tests, the gland vascularity correlates rather poorly. Consequently, it is shown that in certain types of thyroid physiology misleading results can be obtained from conventional “early uptake” measurements, when a high vascularity occurs together with a low uptake slope, or vice versa. Amongst possible further applications of this method, it should be particularly valuable in elucidating the effect of long term antithyroid drugs upon thyroid physiology.