Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
ORIGINAL ARTICLE E n d o c r i n e C a r e Impact of Lithium on Efficacy of Radioactive Iodine Therapy for Graves’ Disease: A Cohort Study on Cure Rate, Time to Cure, and Frequency of Increased Serum Thyroxine After Antithyroid Drug Withdrawal Fausto Bogazzi, Clara Giovannetti, Rezene Fessehatsion, Maria Laura Tanda, Alberto Campomori, Emanuele Compri, Giuseppe Rossi, Claudia Ceccarelli, Paolo Vitti, Aldo Pinchera, Luigi Bartalena, and Enio Martino Department of Endocrinology and Metabolism (F.B., C.G., R.F., A.C., C.C., P.V., A.P., E.M.), University of Pisa, and Epidemiology and Biostatistics Unit (G.R.), Institute of Clinical Physiology, National Research Council, 56124 Pisa, Italy; and Department of Clinical Medicine (M.L.T., E.C., L.B.), University of Insubria, 21100 Varese, Italy Context: Radioactive iodine (RAI) is a common therapy for hyperthyroidism due to Graves’ disease. A small but significant proportion of patients have recurrence of hyperthyroidism after RAI therapy. Lithium might increase RAI effectiveness by increasing RAI retention in the thyroid. However, whether lithium favorably affects the long-term outcome of RAI therapy is still a matter of argument. Objective: The objective of the study was to compare the efficacy of RAI given with or without concomitant lithium treatment. Design: This was a retrospective cohort study. Setting: The study was conducted at a tertiary university center. Patients: Six hundred fifty-one patients with newly diagnosed Graves’ disease participated in the study. Intervention: Two hundred ninety-eight patients were treated with RAI plus lithium (900 mg/d for 12 d) and 353 with RAI alone. Main Outcome Measures: Proportion of cured patients and time to achieve cure of hyperthyroidism during 1 yr of follow-up was measured. Results: Patients treated with RAI plus lithium had a higher cure rate (91.0%) than those treated with RAI alone (85.0%, P ⫽ 0.030). In addition, patients treated with RAI plus lithium were cured more rapidly (median 60 d) than those treated with RAI alone (median 90 d, P ⫽ 0.000). Treatment with lithium prevented the serum free T4 increase after methimazole withdrawal and RAI therapy. Side effects after RAI therapy occurred in a subset of patients and were mild, transient, and without differences in the two groups. Conclusions: RAI combined with lithium is safe and more effective than RAI alone in the cure of hyperthyroidism due to Graves’ disease. (J Clin Endocrinol Metab 95: 201–208, 2010) ISSN Print 0021-972X ISSN Online 1945-7197 Printed in U.S.A. Copyright © 2010 by The Endocrine Society doi: 10.1210/jc.2009-1655 Received August 3, 2009. Accepted October 20, 2009. First Published Online November 11, 2009 Abbreviations: BSA, Body surface area; CI, confidence interval; FT3, free T3; FT4, free T4; OR, odds ratio; RAI, radioactive iodine; RAIU, radioiodine uptake; TgAb, thyroglobulin antibody; TPOAb, thyroperoxidase antibody; TRAb, TSH receptor antibody; UIE, urinary iodine excretion. J Clin Endocrinol Metab, January 2010, 95(1):201–208 jcem.endojournals.org 201 202 Bogazzi et al. Lithium Increases RAI Efficacy adioactive iodine (RAI) is widely used for hyperthyroidism due to Graves’ disease either as a first-line treatment or when hyperthyroidism relapses after a course of antithyroid drug treatment (1– 4). The target of RAI therapy (cure of hyperthyroidism) is a stable restoration of euthyroidism or permanent hypothyroidism. To this purpose, several regimens have been proposed, including, for example, selection of patients and different schedules based on variable RAI doses (5). Pretreatment with thionamides may reduce RAI efficacy (6 –12), as pointed out in a recent metaanalysis, irrespective of the type of drug and discontinuation protocol (13). Studies reported variable percentage of failure after RAI therapy, depending on several variables, including 131-I dose (14, 15), although a recent metaanalysis did not show difference in the outcome of calculated or fixed doses (16). Lithium accumulates in the thyroid (17) and increases (radio)iodine retention without affecting thyroidal radioiodine uptake (RAIU) (18). The addition of lithium to RAI has been associated with an increase in the radiation dose delivered to the thyroid (19). Whether this results in a higher cure rate of hyperthyroidism is still a matter of argument. Conflicting results have been reported on this issue (20, 21). Furthermore, the potential toxicity of lithium raised some concern about its use in combination with RAI for hyperthyroidism (22). The aims of this study were to evaluate, in a very large series, whether lithium offers an additional value to RAI in terms of cure of hyperthyroidism and whether this treatment, at the doses used, is safe. R Subjects and Methods Subjects The study included 651 patients with Graves’ disease [143 men (22%), 508 women (78%); mean (⫾SD) age 44 ⫾ 13 yr, range 18 –77 yr] referred to the Department of Endocrinology and Metabolism, at University of Pisa, Italy, from January 2004 to June 2007. Inclusion criteria were the following: recent-onset (ⱕ6 months); untreated (or treated with antithyroid drugs for less than 15 d, see Treatment) Graves’ disease; aged 18 yr or older; mild or absent Graves’ ophthalmopathy. Moderate to severe Graves’ ophthalmopathy, previous treatment of hyperthyroidism with radioiodine or partial thyroidectomy, and contraindications to glucocorticoids (peptic ulcer, gastritis, active infectious disease, and previous tuberculosis infection) or lithium treatment (kidney disease) were exclusion criteria. The ethical board of our department approved the study. Study design The present study was a retrospective cohort study in nature because data were obtained from medical records of the patients, who were, indeed, prospectively followed up with a standard follow-up. In addition, although patients were not randomly J Clin Endocrinol Metab, January 2010, 95(1):201–208 allocated to the treatment regimen, the propensity to the treatment was evaluated with an ad hoc score (propensity score). This score is the conditional probability of an experimental unit (i.e. patient) being assigned to a particular treatment in a study given a set of known covariates. Propensity scores are used to reduce selection bias by equating groups based on these covariates, thus adjusting for confounding factors. To control for a possible bias owing to nonrandomized assignment of patients to the type of treatment (RAI or RAI plus lithium), we performed a multivariate analysis using both the propensity score and the covariates more linked to the cure rate. Evaluation, study period, and definition of cure Baseline evaluation included evaluation of thyroid function, 3- and 24-h RAIU determination, thyroid echography, ophthalmological examination, white cells count, differential count, hematocrit, platelet count, blood urea nitrogen, creatinine, serum electrolytes, urinalysis, and electrocardiogram. Serum free T4 (FT4), free T3 (FT3), and TSH concentrations were measured on the day of methimazole withdrawal (T-5), at T-3, T0 (the day of RAI administration), T⫹1, T⫹3, T⫹5, T⫹7, T⫹14, and T⫹30 and then every month for the entire follow-up period. Thyroid echography was performed at baseline and at the end of the 1-yr follow-up period. Patients were considered cured when they developed permanent hypothyroidism or were stably euthyroid. All patients were followed for at least 1 yr. Stable euthyroidism was defined as serum FT4, FT3, and TSH within the normal range and confirmed during the following 12-month period; thus, a patient, who achieve euthyroidism 6 months after RAI therapy was considered cured if he or she remained euthyroid during the following 12 months. Hypothyroidism was defined as low serum thyroid hormone and increased TSH concentrations; after the first evidence, hypothyroidism was confirmed in the following 2 months, meaning that a patient who developed hypothyroidism 12 months after RAI was considered having permanent hypothyroidism if it was confirmed during the following 2 months. Hypothyroidism or persistent hyperthyroidism after RAI treatment was corrected by levothyroxine or methimazole administration, as appropriate. A second RAI dose was given to patients with persistent hyperthyroidism at the end of the follow-up period. Follow-up lasted 1 yr after RAI therapy. Treatment All patients were treated with methimazole for 3– 6 months to restore euthyroidism before RAI therapy. Seventeen patients (2.6%) were on antithyroid drug at referral, whereas in the remaining 97.4%, methimazole was initiated after the referral to our department. Methimazole was withdrawn 5 d before RAI therapy. Lithium (900 mg/d for 12 d) was started the day of methimazole withdrawn (i.e. 5 d before RAI therapy) and continued for 7 d after RAI. This regimen was shorter (12 vs. 19 d) than that of a previous pilot study (19) because the effect of lithium on serum thyroid hormone concentrations occurred 3–5 d after RAI therapy (19). All patients received a short course of oral prednisone (starting dose, 0.5 mg/kg 䡠 d) beginning the seventh day after RAI therapy (the day of lithium discontinuation) to avoid RAI-associated progression of Graves’ ophthalmopathy (23, 24); prednisone was gradually tapered (0.1– 0.3 mg/kg every 7–15 d) and withdrawn in 2 months. Radioiodine dose was 7 MBq/g (260 mCi/g) estimated thyroid tissue, corrected for the 24-h RAIU value as previously reported (19). J Clin Endocrinol Metab, January 2010, 95(1):201–208 Assays Serum FT4, FT3 (Vitros Immunodiagnostics, The Broadway, Amersham, Bucks, UK), TSH (Immulite 2000, third generation TSH; Diagnostic Products Corp., Los Angeles, CA), TRAb (TSH receptor antibody; TRAK human; Brahms, Hennigsdorf, Germany), TgAb (thyroglobulin antibody; AIA-Pack TgAb; Tosoh, Tokyo, Japan), and TPOAb (thyroperoxidase antibody; AIA-Pack TPOAb; Tosoh) levels were assayed by commercial kits. Normal values in our laboratories are as follows: FT4, 7–17 pg/ml (9.0 –22.0 pmol/liter); FT3, 2.7– 4.5 pg/ml (4.2–7.0 pmol/ liter); TSH, 0.4 –3.4 mU/liter; TRAb, less than 1 U/liter; TgAb, less than 1 U/ml; TPOAb, less than 1 U/ml. Urinary iodine excretion (UIE) Morning random urinary samples were collected for iodine measurements using an autoanalyzer apparatus (Technicon, Rome, Italy). Median UIE in our areas is 110 g/liter. Serum lithium concentration Serum lithium concentration was measured by a colorimetric assay (Cobas c; Roche Diagnostics GmbH, Mannheim, Germany); therapeutic levels for psychiatric disorders range from 0.6 to 1.2 mEq/liter. Thyroid ultrasonography Thyroid volume was measured by ultrasonography and calculated by the ellipsoid model (width ⫻ length ⫻ thickness ⫻ 0.52 for each lobe) as previously described (25). Thyroid volume was normalized by body surface area (BSA), using the following formula [BSA (square meters) ⫽ 公 height (centimeters) ⫻ weight (kilograms)/3600) (26); normal values in our areas are 3.5–13 ml/m2 (27). Thyroidal RAIU Thyroidal RAIU was measured 3 and 24 h after the administration of a tracer dose (50 Ci) of 131I, starting the day before RAI therapy. The normal 3- and 24-h RAIU values in our area are 10 –20 and 30 – 45%, respectively. Side effects after radioactive iodine therapy Side effects after RAI therapy were evaluated through a questionnaire filled by each patient at the 1-month control and reflected symptoms occurred during the previous 30 d. The questionnaire was based on previous studies (19, 20, 28) and included the following symptoms: neck pain or tenderness (suggesting actinic thyroiditis); swelling and tenderness of the salivary glands (suggesting sialoadenitis); dry mouth; nausea and upset stomach; loss of appetite (gastrointestinal symptoms); taste changes; headache; tremors; and symptoms suggesting lithium toxicity (see Table). Statistical analysis Baseline values (T-5) were expressed as mean ⫾ SD for quantitative variables and percentage for qualitative variables, respectively. The baseline characteristics of the two groups were compared by nonparametric Mann-Whitney test or ANOVA for quantitative variables, as appropriate, or by the Fisher’s exact test for qualitative variables, respectively. Differences of cure rate between the two groups at 1 yr was evaluated by logistic regression using univariate and multivariate analysis. In the multivariate analysis, treatment effect was adjusted for the covari- jcem.endojournals.org 203 ates significantly associated with the cure rate in the univariate analysis. In addition, a propensity score to the treatment, defined as the conditional probability of being treated given the covariates, has been estimated by logistic regression, keeping into consideration the following variables: age, serum FT4 concentrations and serum FT3 concentrations at T-5, TRAb levels, and normalized thyroid volume. The propensity score was then used in the multiple logistic regression to obtain an adjusted treatment effect. The nonhealing curves, in the two groups, were analyzed using survival curves at 12 months by the Kaplan-Meier method. The two curves were compared by the log-rank test. Comparison of the cure time between the two groups was performed by the Cox proportional hazard regression model. The time trend of serum FT4 and FT3 concentrations in the two treatment groups was compared by ANOVA with repeated measures, with a trial factor (time) and a grouping factor (treatment). When a significant treatment-time interaction was present, simple main effects were evaluated. Multiple comparisons among times were performed by a modified Dunnett’s test. The minimal value of serum lithium concentration to obtain the lithium effect on the cure rate was identified using the receiver-operating characteristic curve. The Youden’s index was used to obtain the cutoff value. A relationship between serum creatinine and lithium concentration was evaluated by linear regression analysis. Difference in normalized thyroid volume at the end of the follow-up period in the two treatment groups was compared by ANOVA using the normalized thyroid volume at baseline as covariate. Difference in the incidence of symptoms after RAI therapy in the two treatment groups was evaluated by the Fisher’s exact test. Results Among the 651 patients submitted to RAI, 353 were treated with RAI alone and 298 with RAI combined with lithium. Four patients, treated with RAI and lithium, were lost to follow-up. Baseline clinical and biochemical findings of the two groups of patients are shown in Table 1; the two treatment groups did not differ for their main clinical and biochemical features. However, patients treated with RAI plus lithium were slightly thinner, had a slightly larger goiter, and received a slightly smaller dose of radioiodine (Table 1); the two latter parameters were slightly favorable to the group of patients treated with RAI alone. Cure rate The proportion of cured patients was significantly higher [91.0% (267 of 294)] in the RAI plus lithium group than the RAI alone group [85.0% (300 of 353)] (P ⫽ 0.030; Fig. 1). When the difference in the cure rate in the two groups was adjusted for covariates, which were significantly associated with this event at the univariate analysis (TRAb, P ⫽ 0.033, and normalized thyroid volume, P ⫽ 0.000, Table 2), the effect of lithium was even en- 204 Bogazzi et al. Lithium Increases RAI Efficacy J Clin Endocrinol Metab, January 2010, 95(1):201–208 TABLE 1. Clinical and biochemical features of the study groups at baseline Patients (male/female), n Mean age 关(yr) (range)兴 BMI (kg/m2) Body surface (m2) Smokers (%) Onset of hyperthyroidism (months) Presence of ophtalmopathy (%) Mean thyroid volume (ml) Mean normalized thyroid volume (ml/m2) Mean serum FT4 at T-5 (pmol/liter) Mean serum FT3 at T-5 (pmol/liter) Hyperthyroid patients at the time of RAI (%) Positive TRAb (%) TRAb (U/liter) Positive TgAb (%) Positive TPOAb (%) UIE (g/liter) 3-h RAIU value (%) 24-h RAIU value (%) 131-I administered dose (MBq) Mean serum lithium concentration (mEq/liter) Mean methimazole dose (mg/d) Mean time of methimazole treatment (months) RAI group 353 (84/269) 44.9 ⫾ 12.7 (18 –71) 25 ⫾ 4.5 1.8 ⫾ 0.2 39.7 3.5 ⫾ 1.3 54.8 21.4 ⫾ 11.9 12.0 ⫾ 6.3 12.1 ⫾ 6.5 4.1 ⫾ 3.0 17.1 87.0 12.3 ⫾ 27.2 60.3 79.3 30.4 ⫾ 39.0 42.6 ⫾ 22.1 63.3 ⫾ 18.4 531.3 ⫾ 65.9 N/A 12.4 ⫾ 3.7 4.3 ⫾ 0.3 RAI ⴙ Lithium group 298 (48/240) 43.1 ⫾ 13.3 (18 –77) 24.0 ⫾ 3.6 1.7 ⫾ 0.2 43.4 3.3 ⫾ 1.2 43.6 24.2 ⫾ 12.9 14.0 ⫾ 7.0 12.6 ⫾ 7.2 4.7 ⫾ 3.2 20.0 89.9 12.1 ⫾ 18.1 64.7 84.7 27.5 ⫾ 35.7 48.1 ⫾ 22.6 68.9 ⫾ 17.9 520.0 ⫾ 70.5 0.56 ⫾ 0.23 13.1 ⫾ 2.9 4.1 ⫾ 0.4 P 0.214 0.070 0.002 0.014 0.419 0.543 0.182 0.005 0.000 0.444 0.337 0.404 0.303 0.911 0.702 0.450 0.357 0.002 0.000 0.028 N/A 0.579 0.656 Mean thyroid volume was measured by ultrasonography and calculated by the ellipsoid model and normalized by BSA; normal values in our areas are 3.5–13 ml/m2. Normal values in our laboratories are as follows: FT4, 7–17 pg/ml (9.0 –22.0 pmol/liter); FT3, 2.7– 4.5 pg/ml (4.2–7 pmol/liter); TRAb, less than 1 U/liter; TgAb, less than 1 U/ml; TPOAb, less than 1 U/ml; to convert FT4 and FT3 to picograms per milliliter, multiply by 1.29 and 1.54, respectively. To convert megabequerels to millicuries, divide by 37. The normal 3- and 24-h RAIU values in our area are 10 –20 and 30 – 45%, respectively. BMI, Body mass index. hanced [P ⫽ 0.002, odds ratio (OR) 2.618, 95% confidence interval (CI) 1.444 – 4.749]. The difference in the proportion of cured patients between the two treatment groups was further analyzed adjusting for the propensity FIG. 1. Cumulative hazard of cured patients estimated by the KaplanMeier method. The two curves were compared by the log-rank test (P ⫽ 0.000). Difference in the cure rate at the end of the follow-up period was assessed by Fisher’s exact test (P ⫽ 0.030). The outcome of patients treated with RAI and lithium is indicated by solid line and that of patients treated with RAI alone by dotted line. The cumulative hazard of cure is related to the cumulative proportion of patients not cured. The insert figure shows the percentage of patients not cured in the two groups. score to treatment and for TRAb values; variables significantly affecting the propensity score were age (P ⫽ 0.005) and normalized thyroid volume (P ⫽ 0.007). Even when adjusted for the propensity score and TRAb values, the adjunctive effect of lithium on the proportion of cured patients was highly significant (P ⫽ 0.002, OR 2.459, 95% CI 1.383– 4.374). Cure time The median cure time was 60 d in the RAI plus lithium group and 90 d in the RAI-alone group (P ⫽ 0.000). Patients treated with radioiodine plus lithium had a more rapid control of hyperthyroidism and a higher probability of controlling hyperthyroidism during the first months after radioiodine therapy than those receiving RAI alone (Figure 1). The favorable effect of lithium on the median cure time [P ⫽ 0.000, hazard ratio 1.512, 95% CI 1.220 –1.875] was present even when adjusted for the covariates, which were significantly associated to the median cure time at the Cox univariate analysis (TRAb, P ⫽ 0.007; normalized thyroid volume, P ⫽ 0.000; FT3, P ⫽ 0.046). Effect of serum lithium concentrations on cure rate Mean serum lithium concentration was 0.56 ⫾ 0.23 mEq/liter; only four patients (1.3%) had a serum lithium concentration greater than 1 mEq/liter, and among them J Clin Endocrinol Metab, January 2010, 95(1):201–208 jcem.endojournals.org 205 TABLE 2. Factors associated with the outcome of RAI therapy Age Male sex BMI Body surface Mean thyroid volume Mean normalized thyroid volume Smoke (yes/no) Number of cigarettes per day Lithium treatment Serum FT4 at T-5 Serum FT3 at T-5 Serum FT4 at T 0 Serum FT3 at T 0 Serum TSH at T 0 TRAb UIE OR 0.994 1.308 0.982 0.415 0.934 0.888 95% CI 0.977–1.012 0.764 –2.239 0.929 –1.038 0.131–1.316 0.918 – 0.952 0.859 – 0.919 P 0.524 0.327 0.518 0.135 0.000 0.000 0.840 0.982 0.485–1.457 0.960 –1.006 0.535 0.141 1.747 0.996 0.942 0.951 0.850 1.082 0.992 0.999 1.068 –2.857 0.960 –1.034 0.881–1.008 0.922– 0.980 0.797– 0.906 0.919 –1.272 0.984 – 0.999 0.972–1.006 0.026 0.846 0.085 0.001 0.000 0.344 0.033 0.197 Factors associated with the probability of cure were evaluated using univariate analysis by logistic regression. Treatment with lithium, lower thyroid volume, lower TRAb concentrations, lower thyroid hormone concentrations, and higher TSH concentrations at time 0 were associated with a higher probability of cure. However, only lithium, TRAb, and thyroid volume maintained a predictive value in the multivariate regression analysis (see Results). Serum FT4 and FT3 concentrations were measured at T-5 (the day of methimazole withdrawal and lithium initiation) and at T-0 (the day of radioiodine therapy). BMI, Body mass index; OR, odds ratio of probability of cure. only one patient (0.3%) reached a 1.5 mEq/liter serum lithium concentration. Using the receiver-operating characteristic curve, we identified a cutoff value of 0.45 mEq/ liter (sensitivity 0.70 and specificity 0.56). When patients treated with RAI plus lithium were divided into two levels based on lithium cutoff value, the probability of cure was higher when serum lithium concentration was 0.45 mEq/ liter or greater (OR 2.97, 95% CI 1.04 – 8.81, P ⫽ 0.042); the proportion of cured patients was 93.3% when serum lithium concentration was 0.45 mEq/liter or greater and 82.3% when serum lithium concentration was less than 0.45 mEq/liter. However, serum lithium concentrations 0.7 mEq/liter or greater were not associated with a further increase in the cure rate when compared with a serum lithium concentration less than 0.45 mEq/liter. Finally, serum lithium concentrations were related, although weakly, with serum creatinine levels (r ⫽ 0.21, P ⫽ 0.014), which were within the normal range in all patients. Time trend of serum thyroid hormone concentrations Mean serum FT4 and FT3 concentrations, which did not differ in the two groups at baseline (T-5) (P ⫽ 0.444 and P ⫽ 0.337, respectively, Table 1), significantly changed during the follow-up period in each group (P ⬍ 0.0001; Fig. 2). FIG. 2. Outcome of serum thyroid hormone concentrations. The shaded area shows the normal range for serum FT4 and FT3 levels. Outcome of mean serum FT4 and FT3 concentrations (⫾SE) in patients treated with RAI and lithium are shown by white circles and that in patients treated with RAI alone by black circles. Difference in the outcome of serum FT4 concentrations between the RAI plus lithium group and the RAI alone group were different at T⫹3 and T⫹5 (P ⫽ 0.0139 and P ⫽ 0.0373, respectively); on the contrary, mean serum FT3 concentrations, although lower in patients treated with RAI plus lithium, were not significantly different in the two treatment groups. Serum FT4 concentrations increased after methimazole withdrawal and RAI therapy, reaching a peak between 3 and 5 d after radioiodine therapy in the RAI group (P ⬍ 0.0001 for both times). On the contrary, serum FT4 concentrations remained in the normal range, although slightly increasing after methimazole withdrawal and RAI therapy in patients treated with RAI plus lithium (P ⫽ 0.138 at T⫹3 and P ⫽ 0.050 at T⫹5, Fig. 2). Mean serum FT4 concentrations in the RAI group were significantly higher in the RAI group than in the RAI plus lithium group at T⫹3 and T⫹5 (P ⫽ 0.0139 and P ⫽ 0.0373, respectively). Subsequently, serum FT4 levels declined reaching the normal range between 14 and 30 d. Serum FT3 concentrations had a similar trend in the two groups, increasing shortly after methimazole withdrawal and reaching a peak the day of RAI administration (P ⫽ 0.0003 in both groups). Mean serum FT3 concentrations were higher in the RAI group than the RAI plus lithium group, although not reaching a significant difference, at T⫹3 and T⫹5 (P ⫽ 0.090 and P ⫽ 0.182, respectively; Fig. 2). 206 Bogazzi et al. Lithium Increases RAI Efficacy J Clin Endocrinol Metab, January 2010, 95(1):201–208 Outcome of Graves’ ophthalmopathy Graves’ ophthalmopathy was absent or mild in all subjects evaluated by a complete ophthalmological evaluation; no patient in the present cohort of subjects in either group had a worsening of eye disease after RAI. 1.5 mEq/liter) had no symptoms heralding a toxic effect of lithium (Table 3). No patient reported symptoms related to glucocorticoids excess. Discussion Outcome of thyroid volume As expected, normalized thyroid volume significantly decreased in both groups after RAI therapy; mean normalized thyroid volume reduced from 12.0 ⫾ 6.3 to 5.0 ⫾ 3.9 ml in the RAI group (P ⬍ 0.0001) and from 14.0 ⫾ 7.0 to 4.8 ⫾ 4.5 ml in the RAI plus lithium group (P ⬍ 0.0001). At the end of the follow-up period, normalized thyroid volume did not differ in the two groups (P ⫽ 0.762). Symptoms occurring after RAI therapy Symptoms occurring after RAI therapy were evaluated by a questionnaire administrated to each patient 1 month after radioiodine therapy. Overall, side effects occurred in few patients; no patient in either group had major side effects, which were, on the contrary, mild and transient. Twenty-five percent of patients in both groups had mild symptoms suggesting actinic thyroiditis, which lasted 2–3 d. Fifteen percent of patients in both groups had referred gastrointestinal symptoms; finally, the incidence of symptoms did not differ in the two groups of patients (Table 3). No toxic effects of lithium were reported; moreover, the four patients with serum lithium concentration greater than 1 mEq/liter (including the patient with serum lithium concentration of TABLE 3. Incidence of symptoms after RAI therapy Actinic thyroiditis (%) Pain in the neck region Neck swelling Sialoadenitis (%) Nausea (%) Vomiting (%) Diarrhea (%) Somnolence (%) Dizziness (%) Tremor (%) Asthenia (%) Altered taste (%) Thirst (%) Polyuria (%) Weakness (%) Seizures (%) Anorexya (%) Mental confusion (%) RAI group RAI ⴙ lithium group P 26.8 4.8 5.3 12.0 2.4 7.3 36.6 9.8 12.1 48.8 17.1 19.5 29.3 22.0 0.0 0.0 0.0 25.4 5.9 4.9 19.4 2.5 7.5 28.4 7.5 7.4 32.8 20.9 14.9 23.9 19.4 0.0 0.0 0.0 0.891 0.586 0.678 0.429 1.000 1.000 0.399 0.728 0.500 0.109 0.803 0.599 0.652 0.808 1.000 1.000 1.000 Symptoms were obtained through a questionnaire filled by each patient at the 1-month visit. Incidence of side effects is expressed as percentage of the total. Hyperthyroidism due to Graves’ disease can be treated with thionamides, thyroidectomy, or RAI (29). Thionamides are associated with a high relapse rate of hyperthyroidism (30); thyroidectomy may be rarely (if the surgeon is skilled) complicated by laryngeal nerve palsy or hypoparathyroidism (31). RAI treatment is widely used because it is easy to be administered, widely available, and effective in most patients (4). However, hyperthyroidism recurs or persists in 15–18% of patients after RAI. Failure of RAI treatment may be influenced by several factors, including thionamide treatment, TRAb titers, large goiter, and high RAIU values (4, 13–15); on the contrary, other studies did not find an association between TRAb levels and RAI failure (32). Lithium, owing to its effect on (radio)iodine retention in the thyroid, is a good candidate for increasing RAI effects. However, conflicting results have been reported so far (19 –22, 33, 34). The addition of lithium to RAI was associated with an increased cure rate in the short term (33) but not in the long run (34). Two pilot studies from our group suggested that lithium could indeed increase the effect of RAI in Graves’ disease (19, 20). In neither study, however, a statistical difference was achieved, either because patients with large goiters, more suitable for surgery, were included (20) or the number of enrolled patients was not high enough to allow detection of differences in the proportion of cured patients (19). The results of the present, large study clearly show that the addition of lithium is beneficial for Graves’ disease patients treated with RAI. In fact, a higher proportion of patients were cured compared with those treated with RAI alone. Lithium treatment resulted in an additional 6% cure rate, leading to an overall cure after 1 yr greater than 90%, the highest reported so far. Our results differ from those reported by Bal et al. (21), showing no different outcome in hyperthyroid patients treated with RAI plus lithium and those treated with RAI alone. It is worth mentioning that in the study by Bal et al. (21), the cure rate was low in all patients (averaging 65%), and the study groups were not homogeneous because patients with both Graves’ disease and toxic multinodular goiter were included. Patients enrolled in our study were homogeneous (only Graves’ disease), making assessment of lithium adjuvant treatment more appropriate. Proportion of cured patients in the RAI-alone group was superimposable to that reported in other studies (14, 15). J Clin Endocrinol Metab, January 2010, 95(1):201–208 Another important result of our study is that the addition of lithium adjuvant therapy allows a prompter control of hyperthyroidism after RAI and lithium, thus confirming the results of our previous pilot study (20). In fact, 50% of patients were cured in 2 months if given lithium, 3 months if treated with RAI alone. This difference may appear marginal, but a prompt control of hyperthyroidism is indeed important in some categories of patients, such as the elderly or those with underlying heart problems, who need a rapid restoration of euthyroidism (or the occurrence of permanent hypothyroidism). Serum FT4 concentrations increased after methimazole withdrawal in patients treated with RAI alone; a further increase occurred after RAI reaching peaking 3–5 d after treatment. Conversely, serum thyroid hormone increase was blunted in patients treated with RAI and lithium, leaving thyroid hormone levels well within the normal range, in keeping with the inhibitory effect of lithium on the release of thyroid hormone release in cell culture (35). Again, this effect is not negligible in elderly patients and those with preexisting cardiac problems. Serum FT3 concentrations increased after methimazole withdrawal mainly in the RAI group, without a statistical difference with the RAI plus lithium group. Resumption of thionamides after RAI therapy may be an alternate tool to avoid exacerbation of thyrotoxicosis (36). The increased cure rate observed in the RAI plus lithium group was associated with serum lithium concentrations 0.45 mEq/liter or greater. This lithium level is below the therapeutic concentrations of lithium for psychiatric disorders (37). Serum lithium concentrations of 0.7 mEq/liter or greater were not associated with a higher cure rate. In our opinion, this is an important finding because it suggests that lithium increases RAI efficacy at blood concentrations far from those considered to be risky for the occurrence of side effects (22, 37). In keeping with this assumption, the proportion of patients who developed symptoms after RAI therapy was similar in the two groups, indicating that lithium, under the conditions used in this study, is safe. Finally, the lithium regimen used in the present study is cheap (about 4 U.S. dollars) and may avoid a further exposure to radiometabolic therapy in 6% of Graves’ patients and reduce environmental pollution. In conclusion, the results of this study, carried out in a large number of patients with Graves’ disease, suggest that a short course of lithium is safe and beneficial for patients treated with RAI, increasing the cure rate, shortening the cure time and preventing serum thyroid hormone surge. jcem.endojournals.org 207 Acknowledgments Address all correspondence and requests for reprints to: Fausto Bogazzi, Department of Endocrinology and Metabolism, University of Pisa, Ospedale Cisanello, Via Paradisa, 2, 56124 Pisa, Italy. E-mail: [email protected]; or [email protected]. This work was partially supported by grants from the University of Pisa (Fondi d’Ateneo) (to F.B. and E.M.). Disclosure Summary: F.B., C.G., R.F., M.L.T., A.C., E.C., G.R., C.C., P.V., A.P., L.B., and E.M. have nothing to declare. References 1. Kendall-Taylor P, Keir MJ, Ross WM 1984 Ablative radioiodine therapy for hyperthyroidism: long-term follow-up study. Br Med J 289:361–363 2. Peters H, Fischer C, Bogner U, Reiners C, Schleusener H 1997 Treatment of Graves’ hyperthyroidism with radioiodine: results of a prospective randomized study. Thyroid 7:247–251 3. Wartofsky L 1997 Radioiodine therapy for Graves’ disease: case selection and restrictions recommended to patients in North America. Thyroid 7:213–216 4. Weetman AP 2007 Radioiodine treatment for benign thyroid diseases. Clin Endocrinol (Oxf) 66:757–764 5. Allahabadia A, Daykin J, Sheppard MC, Gough SC, Franklyn JA 2001 Radioiodine treatment of hyperthyroidism-prognostic factors for outcome. J Clin Endocrinol Metab 86:3611–3617 6. Marcocci C, Gianchecchi D, Masini I, Golia F, Ceccarelli C, Bracci E, Fenzi GF, Pinchera A 1990 A reappraisal of the role of methimazole and other factors on the efficacy and outcome of radioiodine therapy of Graves’ hyperthyroidism. J Endocrinol Invest 13:513– 520 7. Andrade VA, Gross JL, Maia AL 2001 The effect of methimazole pretreatment on the efficacy of radioactive iodine therapy in Graves’ hyperthyroidism: one-year follow-up of a prospective randomized study. J Clin Endocrinol Metab 86:3488 –3493 8. Burch HB, Solomon BL, Cooper DS, Ferguson P, Walpert N, Howard R 2001 The effect of antithyroid drug pretreatment on acute changes in thyroid hormone levels after 131I ablation for Graves’ disease. J Clin Endocrinol Metab 86:3016 –3021 9. Bonnema SJ, Bennedbaek FN, Veje A, Marving J, Hegedüs L 2006 Continuous methimazole therapy and its effect on the cure rate of hyperthyroidism using radioactive iodine: an evaluation by a randomized trial. J Clin Endocrinol Metab 91:2946 –2951 10. Bogazzi F, Martino E, Bartalena L 2003 Antithyroid drug treatment prior to radioiodine therapy for Graves’ disease: yes or no? J Endocrinol Invest 26:174 –176 11. Sabri O, Zimny M, Schulz G, Schreckenberger M, Reinartz P, Willmes K, Buell U 1999 Success rate of radioiodine therapy in Graves’ disease: the influence of thyrostatic medication. J Clin Endocrinol Metab 84:1229 –1233 12. Imseis RE, Vanmiddlesworth L, Massie JD, Bush AJ, Vanmiddlesworth NR 1998 Pretreatment with propylthiouracil but not methimazole reduces the therapeutic efficacy of iodine-131 in hyperthyroidism. J Clin Endocrinol Metab 83:685– 687 13. Walter MA, Briel M, Christ-Crain M, Bonnema SJ, Connell J, Cooper DS, Bucher HC, Müller-Brand J, Müller B 2007 Effects of antithyroid drugs on radioiodine treatment: systematic review and meta-analysis of randomised controlled trials. Br Med J 334:514 –520 14. Boelaert K, Syed AA, Manji N, Sheppard MC, Holder RL, Gough SC, Franklyn JA 2009 Prediction of cure and risk of hypothyroidism in patients receiving 131I for hyperthyroidism. Clin Endocrinol (Oxf) 70:129 –138 15. Erem C, Kandemir N, Hacihasanoglu A, Ersöz HO, Ukinc K, Kocak M 2004 Radioiodine treatment of hyperthyroidism: prognostic factors affecting outcome. Endocrine 25:55– 60 208 Bogazzi et al. Lithium Increases RAI Efficacy 16. de Rooij A, Vandenbroucke J, Smit J, Stokkel M, Dekkers O 2009 Clinical outcomes after estimated versus calculated activity of radioiodine for the treatment of hyperthyroidism: systematic review and meta-analysis Eur J Endocrinol 161:771–777 17. Temple R, Berman M, Robbins J, Wolff J 1972 The use of lithium in the treatment of thyrotoxicosis. J Clin Invest 51:2746 –2756 18. Robbins J 1984 Perturbations of iodine metabolism by lithium. Math Biosci 72:337–347 19. Bogazzi F, Bartalena L, Campomori A, Brogioni S, Traino C, De Martino F, Rossi G, Lippi F, Pinchera A, Martino E 2002 Treatment with lithium prevents serum thyroid hormone increase after thionamides withdrawal and radioiodine therapy in patients with Graves’ disease. J Clin Endocrinol Metab 87:4490 – 4495 20. Bogazzi F, Bartalena L, Brogioni S, Scarcello G, Burelli A, Campomori A, Manetti L, Rossi G, Pinchera A, Martino E 1999 Comparison of radioiodine with radioiodine plus lithium in the treatment of Graves’ hyperthyroidism. J Clin Endocrinol Metab 84:499 –503 21. Bal CS, Kumar A, Pandey RM 2002 A randomized controlled trial to evaluate the adjuvant effect of lithium on radioiodine treatment of hyperthyroidism. Thyroid 12:399 – 405 22. Dunkelmann S, Künstner H, Nabavi E, Eberlein U, Groth P, Schümichen C 2006 Lithium as an adjunct to radioiodine therapy in Graves’ disease for prolonging the intrathyroidal effective half-life of radioiodine. Useful or not? Nuklearmedizin 45:213–218; quiz N51–N52 23. Bartalena L, Marcocci C, Bogazzi F, Panicucci M, Lepri A, Pinchera A 1989 Use of corticosteroids to prevent progression of Graves’ ophthalmopathy after radioiodine therapy for hyperthyroidism. N Engl J Med 321:1349 –1352 24. Bartalena L, Tanda ML 2009 Clinical practice. Graves’ ophthalmopathy. N Engl J Med 360:994 –1001 25. Massaro F, Vera L, Schiavo M, Lagasio C, Caputo M, Bagnasco M, Minuto F, Giusti M 2007 Ultrasonography thyroid volume estimation in hyperthyroid patients treated with individual radioiodine dose. J Endocrinol Invest 30:318 –322 26. Mosteller RD 1987 Simplified calculation of body-surface area. N Engl J Med 317:1098 27. Bogazzi F, Bartalena L, Tomisti L, Rossi G, Tanda ML, Dell’Unto J Clin Endocrinol Metab, January 2010, 95(1):201–208 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. E, Aghini-Lombardi F, Martino E 2007 Glucocorticoid response in amiodarone-induced thyrotoxicosis resulting from destructive thyroiditis is predicted by thyroid volume and serum free thyroid hormone concentrations. J Clin Endocrinol Metab 92:556 –562 Kita T, Yokoyama K, Higuchi T, Kinuya S, Taki J, Nakajima K, Michigishi T, Tonami N 2004 Multifactorial analysis on the shortterm side effects occurring within 96 hours after radioiodine-131 therapy for differentiated thyroid carcinoma. Ann Nucl Med 18: 345–349 Brent GA 2008 Clinical practice. Graves’ disease. N Engl J Med 358:2594 –2605 Kharlip J, Cooper DS 2009 Recent developments in hyperthyroidism. Lancet 373:1930 –1932 Schüssler-Fiorenza CM, Bruns CM, Chen H 2006 The surgical management of Graves’ disease. J Surg Res 133:207–214 Sabri O, Zimny M, Schreckenberger M, Reinartz P, Ostwald E, Buell U 1999 Radioiodine therapy in Graves’ disease patients with large diffuse goiters treated with or without carbimazole at the time of radioiodine therapy. Thyroid 9:1181–1188 Turner JG, Brownlie BE, Rogers TG 1976 Lithium as an adjunct to radioiodine therapy for thyrotoxicosis. Lancet 1:614 – 615 Brownlie BE, Turner JG, Oveden BM, Rogers TG 1979 Results of lithium-131I treatment of thyrotoxicosis. J Endocrinol Invest 2:303– 304 Mori M, Tajima K, Oda Y, Matsui I, Mashita K, Tarui S 1989 Inhibitory effect of lithium on the release of thyroid hormones from thyrotropin-stimulated mouse thyroids in a perifusion system. Endocrinology 124:1365–1369 Bonnema SJ, Bennedbaek FN, Gram J, Veje A, Marving J, Hegedus L 2003 Resumption of methimazole after 131I therapy of hyperthyroid diseases: effect on thyroid function and volume evaluated by a randomized clinical trial. Eur J Endocrinol 149:485– 492 Baldessarini RJ, Tarazi FI 2005 Pharmacotherapy of psychosis and mania. In: Brunton L, Lazo J, Parker K, eds. Goodman, Gilman’s, the pharmacological basis of therapeutics, 11e, section III. Drugs acting on the central nervous system. Chap 18; 314 –316. New York; McGraw-Hill