Thyroid Hormone Replacement Therapy in Primary Hypothyroidism: A Randomized Trial Comparing L-Thyroxine Plus Liothyronine With L- Thyroxine Alone
Background: Substituting part of the dose of L-thyroxine with small but supraphysiologic doses of liothyronine in hypothyroid patients has yielded conflicting results.
Objective: To evaluate combinations of L-thyroxine plus liothyronine in hypothyroid patients that match the proportions present in normal secretions of the human thyroid gland.
Design: Randomized, double-blind, crossover trial.
Setting: Academic research hospital.
Participants: 28 women with overt primary hypothyroidism.
Intervention: Crossover trial comparing treatment with L- thyroxine, 100 g/d (standard treatment), versus treatment with L- thyroxine, 75 g/d, plus liothyronine, 5 g/d (combination treatment), for 8-week periods. All patients also received L-thyroxine, 87.5 g/ d, plus liothyronine, 7.5 g/d (add-on combination treatment), for a final 8-week add-on period.
Measurements: Primary outcomes included serum thyroid hormone levels, results of quality-of-life and psychometric tests, and patients’ preference. Multiple biological thyroid hormone end points were studied as secondary outcomes.
Results: Compared with standard treatment, combination treatment led to lower free thyroxine levels (decrease, 3.9 pmol/L [95% CI, 2.5 to 5.3 pmol/L]), slightly higher serum levels of thyroid- stimulating hormone (increase, 0.62 mU/L [CI, 0.01 to 1.23 mU/L]), and unchanged free triiodothyronine levels. No improvement was observed in the other primary and secondary end points after combination treatment, with the exception of the Digit Span Test, in which the mean backward score and the mean total score increased slightly (0.6 digit [CI, 0.1 to 1.0 digit] and 0.8 digit [CI, 0.2 to 1.4 digits], respectively). The add-on combination treatment resulted in overreplacement. Levels of thyroid-stimulating hormone decreased by 0.85 mU/L (CI, 0.27 to 1.43 mU/L) and serum free triiodothyronine levels increased by 0.8 pmol/L (CI, 0.1 to 1.5 pmol/ L) compared with standard treatment; 10 patients had levels of thyroid-stimulating hormone that were below the normal range. Twelve patients preferred combination treatment, 6 patients preferred the add-on combination treatment, 2 patients preferred standard treatment, and 6 patients had no preference (P = 0.015).
Limitations: Treatment with L-thyroxine, 87.5 g/d, plus liothyronine, 7.5 g/d, was an add-on regimen and was not randomized.
Conclusions: Physiologic combinations of L-thyroxine plus Mothyronine do not offer any objective advantage over L-thyroxine alone, yet patients prefer combination treatment.
Ann Intern Med. 2005;142:412-424. www.annals.org
The recommended treatment for hypothyroidism is oral L-thyroxine sodium. This treatment is administered with the aim of restoring clinical euthyroidism and well-being and maintaining normal serum levels of thyroidstimulating hormone (TSH) (1). However, triiodothyroninc is the most active thyroid hormone because its affinity for the nuclear thyroid hormone receptor is 10- to 20-fold that of thyroxine (2, 3). The current practice of using L-thyroxine alone as replacement therapy for hypothyroidism assumes that peripheral conversion of thyroxine into triiodothyronine is able to restore normal triiodothyronine concentrations in target tissues. However, no experimental data support this assumption. On the contrary, we found that infusion of thyroxine alone into thyroidectomized rats was not able to restore euthyroidism (4); this was possible only by infusing combinations of thyroxine and triiodothyronine in proportions similar to those secreted by the rat thyroid gland (5).
An early study in hypothyroid patients compared treatment with the usual daily L-thyroxine dose, consisting of two or three 100-g tablets, versus the same number of tablets, each containing 80 g of L-thyroxine and 20 g of liothyronine (6). Although patients were probably over-treated throughout the study, adverse events were more frequent during treatment with the L-thyroxine-liothyronine combination (6), probably because of the excessive amount of liothyronine administered. More recently, several studies have evaluated combined levothyroxine-liothyronine treatment using a “triiodothyronine substitution” approach, in which a small yet supraphysiologic amount of liothyronine, ranging from 10 g/d to 15 g/ d, was substituted for 50 g of the total L-thyroxine dose (7-10). Bunevicius and colleagues (7, 8) reported that triiodothyronine substitution resulted in marked improvements in several items of the Profile of Mood States (POMS) and in a few indexes of psychometric function and quality of life, especially in athyreotic patients with thyroid cancer (8).
Recent studies, however, have failed to confirm any beneficial effects of triiodothyronine substitution on quality of life (9, 10). Even in selected hypothyroid patients presenting with depressive symptoms despite normalization of serum TSH concentrations, substitution of 50% of the L-thyroxine dose with a variable amount of liothyronine did not improve self-rated mood or well-being (11). However, 2 of these studies had a parallel-group design (10, 11), and in the study that included a crossover design, serum TSH levels were higher after triiodothyronine substitution than after treatment with L-thyroxine alone (9), suggesting undertreatment with the former agent (12).
These recent studies using triiodothyronine substitution had limitations. The ratios of L-thyroxine to liothyronine in the individual combinations administered to patients varied widely and differed from the physiologic ratio of thyroxine to triiodothyronine and from the absolute quantities of both hormones secreted by the human thyroid gland (5, 13). Moreover, these studies included men and women whose hypothyroidism varied in duration and severity. Some had subclinical hypothyroidism and might have retained some residual secretion of thyroxine and triiodothyronine. Prestudy L-thyroxine requirements also differed among patients. Therefore, the lack of beneficial effects of the combinations over L-thyroxine alone might depend not on the treatment itself but on these confounding factors. To avoid the limitations of previous studies, we examined L- thyroxine-liothyronine combinations with fixed ratios based on human thyroid secretion (14) in women with overt hypothyroidism at diagnosis.
METHODS
Study Objective and Participants
We performed this study to evaluate whether combinations of L- thyroxine plus liothyronine matching the proportion present in the normal secretion of the human thyroid gland offer demonstrable advantages over standard treatment with L-thyroxine alone in patients with overt primary hypothyroidism. We included women who were between 18 and 70 years of age; who presented with overt primary hypothyroidism, as defined by increased TSH levels and decreased serum free thyroxine levels at diagnosis; and who had maintained normal serum TSH levels while receiving treatment with a stable L-thyroxine dosage of 100 g/d for at least the previous year. Exclusion criteria were mental illness; affective disorders or use of psychotropic drugs; cardiovascular, renal, or hepatic disease; or osteoporosis. For the external euthyroid control group, we selected healthy women who were similar to the patients in terms of age, postmenopausal status, ethnicity, social and cultural background, academic degrees, and current professional occupation. Patients and controls were not taking any drug known to influence thyroid function, other than L-thyroxine in the former group. The Ethics Committee of the Hospital Ramn y Cajal approved the study, and written informed consent was obtained from all patients and controls.
Interventions
We used a randomized, double-blind, crossover design to compare treatment with L-thyroxine, 100 g/d (standard treatment), and treatment with a combination of L-thyroxine, 75 g/d, plus liothyronine, 5 g/d (combination treatment), in periods of 8 weeks each (Figure 1). The latter combination was selected after we considered the 14-to-1 proportion of thyroxine to triiodothyronine in human thyroid secretion (14) and the intestinal absorption of both hormones. Because our previous experience in rats indicated that it is difficult to guess the correct L-thyroxine-liothyronine combination, and because we feared that the combination we chose might result in undertreatment, all patients were also given a daily combination containing 87.5 g of L-thyroxine plus 7.5 g of liothyronine (add-on combination treatment) for a final 8-week add- on period (Figure 1).
On a daily basis, patients took one 50-g tablet of L-thyroxine (Euthyrox, Merck KgaA, Darmstad, Germany) from a red-tagged opaque box and another tablet from a yellow-tagged opaque box. The content of the yellow-tagged box differed during every 8-week period and contained 50-g L-thyroxine tablets (Levothroid, Aventis, Madrid, Spain) for standard treatment; Novothyral Mite tablets (Merck), containing 25 g of L-thyroxine plus 5 g of liothyronine, for combination treatment; or half-tablets of Novothyral 75 (Merck), containing 37.5 g of L-thyroxine plus 7.5 g of liothyronine, forthe add-on combination treatment.
Figure 1. Flow of patients through the study.
Outcomes
Primary outcomes included serum thyroid hormone levels, POMS, the Digit Symbol Substitution Test, the Digit Span Test, the Visual Scanning Test, and patients’ preference. Other tests of quality of life and psychometric function and multiple biological thyroid hormone end points were studied as secondary outcomes. We performed the same procedures in patients at 4 different times: when they were receiving their usual L-thyroxine treatment before the start of the study and at the end of each treatment period. Patients were evaluated while receiving usual treatment to avoid a placebo effect related to the fact of being evaluated; these data were not included in the final analysis. Controls were evaluated only once.
Patients and controls reported to the hospital after fasting overnight for 12 hours. Patients took the study medication with water immediately after waking up. Body mass index, office blood pressure, waist-to-hip ratio, body composition (estimated by bioelectrical impedance [BF-300, Omron Healthcare Inc., Bannockburn, Illinois]), and body temperature were measured. Immediately afterward, a broad evaluation of biological thyroid hormone end points was performed. The sequence of this evaluation was the same for every participant and lasted for no more than 3 hours.
Cardiovascular Evaluation
A single investigator performed echocardiography in all participants as described previously (15). The following measurements were recorded: left atrial diameter, systolic and diastolic left ventricular diameter, posterior wall and septum thickness, shortening fraction, left ventricular ejection fraction, left ventricular mass index, early-phase and maximal late mitral valve flows and their ratio, isovolumic relaxation time, deceleration time, aortic velocity time integral, heart rate, and cardiac output. Systemic vascular resistance was calculated as the mean arterial blood pressure divided by cardiac output. Patients were asked to record their heart rate using an oscillometric ambulatory blood pressure monitor (A&D Company Ltd., Tokyo, Japan) for 24 hours to exclude liothyronine-induced tachycardia.
Hormonal and Biochemical Measurements
Blood samples were obtained 2 hours after ingestion of study medication. Serum levels of free thyroxine, free triiodothyronine, TSH, and sex hormone-binding globulin were determined by using immunochemiluminescent assays (Immulite 2000, Diagnostic Products Corp., Los Angeles, California). The normal ranges were 0.4 to 4.0 mU/L for serum TSH, 10.3 to 24.5 pmol/L for free thyroxine, and 2.8 to 6.5 pmol/L for free triiodothyronine, as reported by the Central Laboratory of Hospital Ramn y Cajal.
An automated analyzer was used to perform a complete serum biochemistry analysis, including renal and liver function tests and lipid profiles, and to measure 24-hour urinary creatinine, calcium, and phosphate excretion (Aeroset, Abbott Laboratories, Abbot Park, Illinois). Levels of urinary hydroxyproline, pyridolines, and deoxypyridolines were determined by using chromatography (Bio-Rad Laboratories, Hercules, California). The mean coefficients of variation were below 10% for all of these assays.
Central and Peripheral Nervous System Evaluation
Immediately after blood sampling, participants had a small breakfast and an independent evaluator performed a battery of cognitive, affective, and quality-of-life tests, including the Zulewski score (16), POMS (17), Nottingham Health Profile and Short Form-36 Health Survey (18-22), Digit Symbol Substitution Test (23, 24), Digit Span Test (23), Visual Scanning Test (25, 26), and 16 Visual Analogue Mental Scales (VAMS) (sadness, confusion, fear, depression, irritability, tension, anger, tiredness, agitation, feeling cold, blurred vision, nausea, sleep, lightheadedness, drowsiness, and feeling hot, as previously described in detail [27]). Central and peripheral nervous system performance was objectively estimated by measuring the ankle teflex relaxation time (Cardioline, Elettronica Trentina, Cavareno, Italy) and obtaining visual and brainstem auditory evoked potentials (28, 29). At the end of the study, patients were asked which treatment they preferred.
Sample Size
We based our a priori power analysis on the beneficial effects of triiodothyronine substitution previously reported by Bunevicius and colleagues (7, 8). Calculations were done for a repeated-measures design, considering the autocorrelation of each test and setting α at 0.05, using PASS2000 software (Number Cruncher Statistical Systems, Kaysville, Utah). Our sample size provided 80% power to detect differences between the mean values after treatment of 2.1 points in the fatigue-inertia score, 3.6 points in the depression- dejection score, and 2.7 points in the anger-hostility score of the POMS; 0.5 digit in the total score of the Digit Span Test; and 1 point in the “items found” score of the Visual Scanning Test.
Randomization, Implementation, and Blinding
One investigator, who had no contact with the patients throughout the study, designed the protocol, obtained the computer-generated sequence of randomization, assigned participants to their groups, and prepared the boxes with the tablets. The physicians recruiting the patients, administering the interventions, and evaluating the outcomes were blinded to the design of the study and to the medication contained in the boxes. Blinding was maintained throughout the study. The study pills were counted to assess adherence only after all of the patients had completed the study. Adherence was 98%, 99%, and 98% for standard treatment, combination treatment, and add-on combination treatment, respectively.
Statistical Analysis
Data are presented as means with 95% CIs, unless otherwise stated. Logarithm or square root transformations were applied to ensure normality whenever possible. For normally distributed outcomes, the comparisons of standard treatment and combination treatment were evaluated with corrected t-tests by using the crossover analysis tools of NCSS 2004 software (Number Cruncher Statistical Systems). For outcomes that were not normally distributed after transformations, we compared treatments with Hodges-Lehman estimates of point median differences and exact 95% CIs, using StatXact software (Cytel Software Corp., Cambridge, Massachusetts) (30, 31).
The add-on combination treatment was compared with standard treatment by using paired t-tests or Wilcoxon signed-rank tests, as appropriate. Comparisons of the different treatments used in hypothyroid patients with controls were performed by using 1-way analysis of variance followed by the Dunnet test or by Kruskal- Wallis analysis of variance followed by Mann-Whitney tests, as appropriate. Discontinuous variables were analyzed by using the chi- square test. These analyses were performed by using SPSS for Macintosh, version 10 (SPSS, Inc., Chicago, Illinois). P values less than 0.05 were considered statistically significant.
Role of the Funding Sources
The funding sources (Instituto de Salud Carlos III, Ministerio de Sanidad y Consumo and Consejera de Educacin, Comunidad de Madrid, Spain, and Merck KgaA, Darmstad, Germany) had no role in the study design; the collection, analysis, and interpretation of data; the writing of the report; or the decision to submit the paper for publication.
RESULTS
Twenty-eight hypothyroid patients (mean age [SD], 48 11 years; mean body mass index [SD], 25.9 7.1 kg/m^sup 2^) were recruited from October 2000 to January 2003. Twenty healthy women (mean age [SD], 46 13 years; mean body mass index [SD], 25.4 5.7 kg/m^sup 2^) served as the external euthyroid control group. Causes of overt hypothyroidism were chronic lymphocytic thyroiditis in 23 patients and thyroid ablation for Graves disease or toxic multinodular goiter in 5 patients.
Two patients did not complete the study (Figure 1). One was lost to follow-up while receiving standard treatment. The other was removed from the study because of asymptomatic thyrotoxicosis after combination treatment, related to the ingestion of a homeopathic preparation containing desiccated thyroid extract. Her thyroid hormone levels returned to normal after the homeopathic compound was withdrawn, but her data were not used for analysis. The results from the 26 patients who completed the 3 treatment regimens are included in the final analysis.
Crossover Study Comparing Standard Treatment (L-Thyroxine, 100 g/ d) with Combination Treatment (L-Thyroxine, 75 g/d, plus Liothyronine, 5 g/d)
The results of the primary outcomes are presented in Figure 2 and Table 1. Compared with standard treatment, serum levels of free thyroxine decreased by 3.9 pmol/L (95% CI, 2.5 to 5.3 pmol/L) after combination treatment, whereas serum TSH levels increased slightly by 0.62 mU/L (CI, 0.01 to 1.23 mU/L) and levels of free triiodothyronine remained unchanged (Table 1). No improvement after combination treatment was observed in any scale of the POMS, on the Digit Symbol Substitution Test, or on the Visual Scanning Test. However, combination treatment slightly improved the backward and total scores of the Digit Span Test (0.6 digit [CI, 0.1 to 1.0 digit] and 0.8 digit [CI, 0.2 to 1.4 digit], respectively) (Table 1). At the end of the study, 12 patients preferred combination treatment, 6 preferred the add-on combination treatment, 2 preferred standard treatment, and 6 had no preference (chi-square value, 10.46; P = 0.015).
The results of the most important secondary outcomes are presented in Table 1. The remainder can be found in the Appendix Table (available at www.annals.org). As occurred with other tests of quality of life and psychometric function, no statistically significant differences between standard and combination treatment were observed in any scale of the Nottingham Health Profile or the Short Form-36 Hea\lth Survey or in any VAMS.
Compared with standard treatment, combination treatment led to an increase of 0.1 C (CI, 0.1 to 0.2 C) in body temperature, an increase of 0.4 mmol/L (16 mg/dL) (CI, 0.2 to 0.7 mmol/L [8 to 27 mg/ dL) in total cholesterol level, an increase of 0.4 mmol/L (16 mg/ dL) (CI, 0.1 to 0.6 mmol/L [4 to 23 mg/dL]) in low-density lipoprotein cholesterol level, and an increase of 1.6 mol (CI, 0.1 to 3.4 mol) in the ratio of urinary deoxypyridolines to creatinine (Table 1). Cardiac output showed a small decrease of 0.2 L/min (CI, 0.01 to 0.4 L/min), related to a subtle decrease in heart rate of 3.0 beats/min (CI, 0.4 to 5.7 beats/min) (Appendix Table, available at www.annals.org). No statistically significant differences between standard and combination treatment were observed in other secondary outcomes, including clinical, biochemical, and hormone measurements; ankle reflex relaxation time; evoked potentials; or echocardiographic measurements (Appendix Table, available at www.annals.org).
Add-On Combination Treatment (L-Thyroxine, 87.5 g/d, plus Liothyronine, 7.5 g/d)
Regarding primary outcomes, the add-on combination treatment resulted in overreplacement. Levels of TSH decreased by 0.85 mU/L (CI, 0.27 to 1.43 mU/L), and levels of free triiodothyronine increased by 0.8 pmol/L (CI, 0.1 to 1.5 pmol/L) compared with standard treatment (Table 2). In 10 women, TSH levels were below the lower limit of the normal range but were not suppressed. No statistically significant differences were observed for the add-on combination treatment compared with standard treatment in any scale of the POMS or the Digit Span Test. On the contrary, when compared with standard treatment, patients receiving the add-on combination treatment performed better on the copies score of the Digit Symbol Substitution Test (increase, 8.5 points [CI, 3.0 to 14.0 points]) and on the time needed to complete the Visual Scanning Test (decrease, 10.0 seconds [CI, 0.9 to 26.0 seconds]) (Table 2).
With respect to secondary outcomes, no statistically significant differences were observed in any scale of the Nottingham Health Profile and the Short Form-36 Health Survey or in any VAMS compared with standard treatment (Appendix Table, available at www.annals.org). Moreover, the number of patients who scored abnormal values on any category of the quality-of-life tests according to population standards did not differ depending on the treatment applied.
Compared with standard treatment, body temperature in patients receiving the add-on combination treatment increased by 0.2 C (CI, 0.1 to 0.3 C), and the Zulewski score decreased by 1.0 point (CI, 0.1 to 1.5 points) (Table 2). Serum creatinine concentration decreased by 3.2 mol/L (0.04 mg/dL) (CI, 0.4 to 5.9 mol/L [0.01 to 0.07 mg/dL]), whereas urinary bone-remodeling markers increased after the add-on combination treatment compared with standard treatment. The hydroxyproline-creatinine ratio increased by 0.5 mmol (CI, 0.09 to 1.23 mmol), the pyridoxines-creatinine ratio increased by 10.1 mol (CI, 2.5 to 17.7 mol), and the pyridolines-creatinine ratio increased by 2.8 mol (CI, 1.5 to 4.0 mol), suggesting that the add-on combination regimen led to overtreatment (Table 2, Appendix Table [available at www.annals.org]). There were also statistically significant differences in the isovolumic relaxation time and the left III latency of brainstem auditory evoked potentials (Appendix Table, available at www.annals.org). However, considering the large number of variables tested in our study, and considering that we did not correct for multiple testing, it is possible that some of these differences might have resulted from chance instead of being actually present in the sample. Finally, no statistically significant differences were found in other clinical or biochemical variables or in the ankle reflex relaxation time (Table 2, Appendix Table [available at www.annals.org]).
Comparison with External Euthyroid Controls
The results of the primary outcomes and the most important secondaiy outcomes obtained from the control group are presented in Table 3 and in the Appendix Table (available at www.annals.org). Standard treatment resulted in serum thyroid hormone concentrations similar to those of controls, with the exception of increased free thyroxine levels, whereas levels of TSH, free thyroxine, and free triiodothyronine did not differ from those of controls after combination treatment. On the contrary, serum TSH levels were decreased compared with controls after the add-on combination regimen, further suggesting overtreatment with this combination. No statistically significant differences between controls and patients were observed in any scale of the POMS, the Digit Symbol Substitution Test, and the Digit Span Test, regardless of the treatment applied. Patients performed worse than controls in the time score of the Visual Scanning Test after standard and combination treatment, but not after add-on combination treatment.
Figure 2. Evolution of primary outcomes during the study depending on the sequence of treatment.
Table 1. Primary and Selected Secondary Outcomes in the 26 Hypothyroid Patients Who Completed the Crossover Study*
Table 2. Primary and Selected Secondary Outcomes in the 26 Hypothyroid Patients Who Completed the Crossover Study: Data from Add-On Combination Treatment*
Table 3. Primary and Selected Secondary Outcomes in the Control Group of 20 Euthyroid Healthy Women*
Of note, when compared with euthyroid controls, patients presented with worse scores on the general health item of Short Form- 36 Health Survey; on the total, emotional, energy, and social scores of the Nottingham Health Profile; and on the depression, nausea, blurred vision, and drowsiness VAMS, regardless of the treatment applied. Also compared with euthyroid controls, the isovolumic relaxation time and some brainstem auditory evoked potentials were worse in patients after standard and combination treatment, but not after add-on combination treatment. Nevertheless, all of the individual values for these evoked potentials were within the normal ranges. No other differences were observed between patients and controls in any other primary or secondary outcomes.
Sequence and Period Effects
Statistically significant sequence effects were found for free triiodothyronine levels, low-density lipoprotein cholesterol levels, and body temperature. We explored this by adjusting these results by baseline values and reanalyzing the data. No statistically significant sequence effect persisted once we adjusted for baseline differences, so we concluded that carryover effects were unlikely. However, for completeness, we present results for these outcomes both overall and by sequence (Table 1). Of all of the variables studied here, an unexplained yet statistically significant period effect was found only for right P-100 amplitude in visual evoked potentials.
Adverse Effects
No adverse effects were reported with any of the treatments. Moreover, the proportion of increased heart rate readings (>120 beats/min in daytime or >100 beats/min at nighttime) in the patients who received ambulatory monitoring was similar (chi-square value, 1.95; P > 0.2) after standard treatment (13 of 420 readings), after combination treatment (13 of 376 readings), and after add-on combination treatment (7 of 377 readings).
DISCUSSION
Interpretation
According to our present results, treatment of primary hypothyroidism with L-thyroxine-liothyronine combinations matching the proportions present in human thyroid secretions does not offer clear advantages over standard treatment with L-thyroxiiie alone. None of the quality-of-life and mood indexes studied improved with L- thyroxineliothyronine combinations compared with standard treatment. In a previous study, we compared the same quality-of-life and mood indexes in patients with differentiated thyroid cancer in 2 instances, the first while they were experiencing mild or subclinical hyperthyroidism due to TSH-suppressive therapy and the second several days after withdrawal of L-thyroxine, when serum levels of free thyroxine and free triiodothyronine were within the normal range (27). We found very few changes in quality of life or mood despite more marked changes in thyroid hormone levels than those observed in the present study, suggesting that these indexes are relatively insensitive to thyroid hormones. Moreover, the rat cerebral cortex has an excellent homeostasis in triiodothyronine concentrations when thyroxine alone is infused (4), but this homeostasis is lost when triiodothyronine alone is infused (32). Therefore, it is unlikely that supplementation of L-thyroxine with liothyronine in humans improves indexes dependent on an organ that is itself virtually independent of the changes in circulating thyroid hormones when L-thyroxine alone is used for replacement therapy. However, it is perplexing that despite the lack of objective advantages, most patients preferred combined treatment. It is possible that combined L-thyroxine-liothyronine treatment offers subtle improvements in well-being that may not be detected by the relatively insensitive merhods used here to study quality of life.
Combinations of L-thyroxine and liothyronine did not clearly improve biological thyroid hormone end points pertaining to other organs and systems. We observed a slight improvement in the Digit Span Test with combination treatment compared with standard treatment, but this improvement was compensated by an undesirable increase in total and low-density lipoprotein cholesterol levels (together with the small increase in TSH levels and the small decrease in heart rate, this finding might suggest underreplacement in some organs, such as the pituitary gland, the heart, and the liver) and in excretion of urinary deoxypyridolines (which, on the contrary, might suggest overreplacement in bone). Some i\ndexes of cognitive function, brainstem auditory evoked potentials, and diastolic performance indexes were slightly better after the add-on combination treatment than after standard treatment. However, these results might be related to the fact that patients were actually overtreated with the add-on combination regimen, given that TSH levels were below the normal range in several patients and bone turnover was clearly increased. Therefore, clinicians should consider the possibility of undesirable effects when adding even small doses of liothyronine to L-thyroxine (33).
The comparison with the external euthyroid control group also provided interesting results. Perhaps the most important is that several items on the quality-of-life tests and several VAMS were worse in patients throughout the study when compared with euthyroid controls, independent of the treatment received. As previous studies showed, for hypothyroid patients receiving L-thyroxine alone (34), awareness of disease may be responsible for this difference (35, 36).
The finding that patients presented with higher serum levels of free thyroxine when receiving standard treatment than euthyroid controls who had similar TSH values is in conceptual agreement with previous results in hypothyroid humans (37) and thyroidectomized rats (4). This further suggests that supraphysiologic thyroxine concentrations are needed to ensure euthyroidism in hypothyroid patients when L-thyroxine is used alone as replacement therapy. Our results demonstrate that this may be avoided by using L-thyroxine- liothyronine combinations, given that serum concentrations of TSH, free thyroxine, and free triiodothyronine in patients receiving combination treatment were not different compared with controls. However, combination treatment did not offer other substantial advantages over treatment with L-thyroxine alone.
Generalizability
Because we included patients with overt primary hypothyroidism, in whom residual thyroid secretion is unlikely, our results may be extrapolated to most hypothyroid patients provided that thyroid hormone is given for replacement therapy, not with the aim of suppressing endogenous TSH secretion. This is important because the improvement in certain indexes of mood and cognitive performance with triiodothyronine substitution reported by Bunevicius and colleagues (8) was restricted to athyreotic patients with thyroid cancer who received supraphysiologic thyroid hormone doses to suppress TSH secretion; our present study did not include such patients. We administered liothyronine once daily, and although patients were evaluated during the peak postabsorption period (38) (making it unlikely that the short half-life of oral liothyronine influenced the lack of beneficial effects of L-thyroxine- liothyronine combinations), it is possible that slow-releasing forms of liothyronine might prove useful for combined L-thyroxine- liothyronine treatment in the future (39).
Overall Evidence
Of the 6 studies to date that have evaluated L-thyroxine- liothyronine treatment for hypothyroidism (Table 4), only the study by Bunevicius and colleagues (7, 8) found advantages of combined treatment over treatment with L-thyroxine alone. However, these beneficial effects appeared to be restricted to athyreotic patients with thyroid cancer who received TSH-suppressive treatment.
Our present study not only confirms the lack of benefits of L- thyroxine-liothyronine combinations but almost definitely solves some questions that had not been addressed in previous studies (36). First, instead of substituting liothyronine for part of the L- thyroxine dose, we used 2 fixed combinations of L-thyroxine- liothyronine, in proportions based on normal human thyroid secretion. In most studies, with the exception of that by Siegmund and associates (33), the amount of liothyronine used was variable and excessive compared with that of L-thyroxine. Second, to avoid patient heterogeneity, we included only women who had been treated with the same L-thyroxine dose for at least 1 year before entering the study, allowing us to ensure that all patients received exactly the same doses of L-thyroxine and liothyronine at each period of treatment. Third, because patients with subclinical hypothyroidism were excluded, all of our patients had long-term overt primary hypothyroidism, making it unlikely that residual thyroid function interfered with our results. Fourth, we did not restrict outcomes to quality of life and psychometric function but used a broad evaluation of thyroid hormone biological end points covering most organs and systems. This allowed us to detect undesirable effects of adding small doses of liothyronine to L-thyroxine, which may have been missed in all previous studies. Fifth, of importance, only our study includes an external control group of healthy euthyroid women. This allowed us to conclude that several outcomes were affected in hypothyroid patients regardless of whether L-thyroxine or L- thyroxine-liothyronine combinations were used as replacement therapy, a potential confounding factor in previous studies.
Table 4. Summary of Studies Evaluating L-Thyroxine-Liothyronine Combinations for the Treatment of Hypothyroidism*
Conclusions
In summary, treatment with L-thyroxine-liothyronine combinations that replicate the ratio of thyroxine to triiodothyronine in human thyroid secretion does not offer any objective advantage over treatment with L-thyroxine alone, although patients prefer combination treatment. However, the addition of even minimally excessive liothyronine doses to L-thyroxine may have undesirable effects. Until clear advantages of L-thyroxine -liothyronine combinations are demonstrated, L-thyroxine alone should remain the drug of choice for hypothyroid replacement therapy.
See also:
Editors’ Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . 413
Summary for Patients . . . . . . . . . . . . . . . . . . . . . . 1-55
Web-Only
Appendix Table
Conversion of figures and tables into slides
Context
Studies comparing the outcomes of treatment of hypothyroidism with L-thyroxine alone versus with a combination of L-thyroxine and liothyronine have had conflicting results. Some experts believe the conflicting results occurred because some studies used supraphysiologic doses of liothyronine.
Contribution
This crossover trial compared L-thyroxine with a combination of physiologic doses of L-thyroxine and liothyronine and found no objective advantages of combination therapy.
Implications
Treatment of hypothyroidism with L-thyroxine alone is sufficient.
-The Editors
References
1. Utiger RD. Hypothyroidism. In: De Groot LJ, ed. Endocrinology. 3rd ed. Philadelphia: WB Saundes; 1995:752-68.
2. Oppenheimer JH, Schwartz HL, Mariash CN, Kinlaw WB, Wong NC, Freake HC. Advances in our understanding of thyroid hormone action at the cellular level. Endocr Rev. 1987;8:288-308. [PMID: 3308445]
3. Morreale de Escobar G. Conversion of T4 to T3 and its relation to hormonal activity. In: Reinwein D, ed. Regulation of Thyroid Function, vol 1. Stuttgart: F.K. Schattaner Verlag; 1976:109-29.
4. Escobar-Morreale HF, Obregon MJ, Escobar del Rey F, Morreale de Escobar G. Replacement therapy for hypothyroidism with thyroxine alone does not ensure euthyroidism in all tissues, as studied in thyroidectomized rats. J Clin Invest. 1995;6:2828-38. [PMID: 8675653]
5. Escobar-Morreale HF, Escobar del Rey F, Obregon MJ, de Escobar G Escobar. Only the combined treatment with thyroxine and triiodothyronine ensures euthyroidism in all tissues of the thyroidectomized rat. Endocrinology. 1996;137:2490-502. [PMID: 8641203]
6. Smith RN, Taylor SA, Massey JC. Controlled clinical trial of combined triiodothyronine and thyroxine in the treatment of hypothyroidism. Br Med J. 1970;4:145-8. [PMID: 4097650]
7. Bunevicius R, Kazanavicius G, Zalinkevicius R, Prange AJ Jr. Effects of thyroxine as compared with thyroxine plus triiodothyronine in patients with hypothyroidism. N Engl J Med. 1999;340:424-9. [PMID: 9971866]
8. Bunevicius R, Prange AJ. Mental improvement after replacement therapy with thyroxine plus triiodothyronine: relationship to cause of hypothyroidism. Int J Neuropsychopharmacol. 2000;3:167-174. [PMID: 11343593]
9. Walsh JP, Shiels L, Lim EM, Bhagat CI, Ward LC, Stuckey BG, et al. Combined thyroxine/liothyronine treatment does not improve well- being, quality of life, or cognitive function compared to thyroxine alone: a randomized controlled trial in patients with primary hypothyroidism. J Clin Endocrinol Metab. 2003;88:4543-50. [PMID: 14557419]
10. Clyde PW, Harari AE, Getka EJ, Shakir KM. Combined levothyroxine plus liothyronine compared with levothyroxine alone in primary hypothyroidism: a randomized controlled trial. JAMA. 2003;290:2952-8. [PMID: 14665656]
11. Sawka AM, Gerstein HC, Marriott MJ, MacQueen GM, Joffe RT. Does a combination regimen of thyroxine (T4) and 3,5,3′- triiodothyronine improve depressive symptoms better than T4 alone in patients with hypothyroidism? Results of a double-blind, randomized, controlled trial. J Clin Endocrinol Metab. 2003;88:4551-5. [PMID: 14557420]
12. Kaplan MM, Same DH, Schneider AB. In search of the impossible dream? Thyroid hormone replacement therapy that treats all symptoms in all hypothyroid patients [Editorial]. J Clin Endocrinol Metab. 2003;88:4540-2. [PMID: 14557418]
13. Toft AD. Thyroid hormone replacement-one hormone or two? [Editorial]. N Engl J Med. 1999;340:469-70. [PMID: 9971874)
14. Pilo A, lervasi G, Vitek F, Ferdeghini M, Cazzuola F, Bianchi R. Thyroidal and peripheral production of 3,5,3'-triiodothyronine in humans by multicompartmental analysis. Am J Physiol. 1990;258:E715- 26. [PMID: 2333963]
15. Botella-Carretero JI, Gomez-Bueno M, Barrios V, Caballero C, Garcia-Robles R, Sancho J, et al. Chronic thyrotropin-suppressive therapy with levothyroxine and short-term overt hypothyroidism after thyroxine withdrawal are associated wi\th undesirable cardiovascular effects in patients with differentiated thyroid carcinoma. Endocr Relat Cancer. 2004;11:345-56. [PMID: 15163309]
16. Zulewski H, Muller B, Exer P, Miserez AR, Staub JJ. Estimation of tissue hypothyroidism by a new clinical score: evaluation of patients with various grades of hypothyroidism and controls. J Clin Endocrinol Metab. 1997;82:771-6. [PMID: 9062480]
17. McNair D, Lorr M, Droppleman L. POMS Manual Profile of Mood States. San Diego, CA: Educational and Industrial Testing Service; 1992.
18. Alonso J, Anto JM, Moreno C. Spanish version of the Nottingham Health Profile: translation and preliminary validity. Am J Public Health. 1990;80:704-8. [PMID: 2343954]
19. Alonso J, Prieto L, Anto JM. The Spanish version of the Nottingham Health Profile: a review of adaptation and instrument characteristics. Qual Life Res. 1994;3:385-93. [PMID: 7866356]
20. Alonso J, Regidor E, Barrio G, Prieto L, Rodriguez C, de la Fuente L. [Population reference values of the Spanish version of the Health Questionnaire SF-36]. Med Clin (Barc). 1998;111:410-6. [PMID: 9834913]
21. Bullinger M, Alonso J, Apolone G, Leplege A, Sullivan M, Wood- Dauphinee S, et al. Translating health status questionnaires and evaluating their quality: the IQOLA Project approach. International Quality of Life Assessment. J Clin Epidemiol. 1998;51:913-23. [PMID: 9817108]
22. Garratt AM, Ruta DA, Abdalla MI, Buckingham JK, Russell IT. The SF36 health survey questionnaire: an outcome measure suitable for routine use within the NHS? BMJ. 1993;306:1440-4. [PMID: 8518640]
23. Wechsler D. WAIS-R Manual: Wechsler Adult Intelligence Scale- Revised. New York: Harcourt Brace Jovanovich; 1981.
24. Kaplan E, Fein D, Morris R, Delis D. WAIS-R NI as a Neuropsychological Instrument. San Antonio, TX: Psychological Corp.; 1991.
25. Weintraub S, Mesulam M. Mental state assessment of young and elderly adults in behavioral neurology. In: Mesulam MM, ed. Principles of Behavioral Neurology. Philadelphia: FA Davis; 1985:71- 122.
26. Wechsler D. Escala de Inteligencia de Wechsler para Adultos- III. Madrid: TEA Ediciones, S.A.; 1999.
27. Botella-Carretero JI, Galn JM, Caballero C, Sancho J, Escobar- Morreale HF. Quality of life and psychometric functionality in patients with differentiated thyroid carcinoma. Endocr Relat Cancer. 2003;10:601-10. [PMID: 14713270]
28. Celesia GG, Brigeli MG. Recommended standards for pattern electroretinograms and visual evoked potentials. The International Federation of Clinical Neurophysiology. Electroencephalogr Clin Neurophysiol Suppl. 1999;52:53-67. [PMID: 10590976]
29. Pratt H, Aminorf M, Nuwer MR, Starr A. Short-latency auditory evoked potentials. The International Federation of Clinical Neurophysiology. Electroencephalogr Clin Neurophysiol Suppl. 1999;52:69-77. [PMID: 10590977]
30. Tudor G, Koch GG. Review of nonparametric methods for the analysis of crossover studies. Stat Methods Med Res. 1994;3:345-81. [PMID: 7889227]
31. Fleiss JL. The Crossover Study. The Design and Analysis of Clinical Experiments. New York: J Wiley; 1986:263-90.
32. Escobar-Morreale HF, Obregon MJ, Escobar del Rey F, Morreale de Escobar G. Tissue-specific patterns of changes in 3,53′-triiodo- L-thyronine concentrations in thyroidectomized rats infused with increasing doses of the hormone. Which are the regulatory mechanisms? Biochimie. 1999;81:453-62. [PMID: 10403175]
33. Siegmund W, Spieker K, Weike AI, Giessmann T, Modess C, Dabers T, et al. Replacement therapy with levothyroxine plus triiodothyronine (bioavailable molar ratio 14 : 1) is not superior to thyroxine alone to improve well-being and cognitive performance in hypothyroidism. CHn Endocrinol (Oxf). 2004;60:750-7. [PMID: 15163340]
34. Saravanan P, Chau WF, Roberts N, Vedbara K, Greenwood R, Dayan CM. Psychological well-being in patients on ‘adequate’ doses of 1-thyroxine: results of a large, controlled community-based questionnaire study. Clin Endocrinol (Oxf). 2002;57:577-85. [PMTD: 12390330]
35. Ladenson PW. Psychological wellbeing in patients. Clin Endocrinol (Oxf). 2002;57:575-6. [PMID: 12390329]
36. Cooper DS. Combined T4 and T3 therapy-back to the drawing board [Editorial]. JAMA. 2003;290:3002-4. [PMID: 14665664]
37. Fish LH, Schwartz HL, Cavanaugh J, Steffes MW, Bantle JP, Oppenheimer JH. Replacement dose, metabolism, and bioavailability of levothyroxine in the treatment of hypothyroidism. Role of triiodothyronine in pituitary feedback in humans. N Engl J Med. 1987;316:764-70. [PMID: 3821822]
38. LeBoff MS, Kaplan MM, Silva JE, Larsen PR. Bioavailability of thyroid hormones from oral replacement preparations. Metabolism. 1982;31:900-5. [PMID: 7121261]
39. Hennemann G, Docter R, Visser TJ, Postema PT, Krenning EP. Thyroxine plus low-dose, slow-release triiodothyronine replacement in hyporhyroidism: proof of principle. Thyroid. 2004;14:271-5. [PMID: 151423601
40. Bunevicius R, Jakubonien N, Jurkevicius R, Cernicat J, Lasas L, Prange AJ Jr. Thyroxine vs thyroxine plus triiodothyronine in treatment of hypothyroidism after thyroidectomy for Graves' disease. Endocrine. 2002;18:129-33. [PMID: 12374459]
Hector F. Escobar-Morreale, MD, PhD; Jos I. Botella-Carretero, MD; Manuel Gmez-Bueno, MD; Jos M. Galn, MD; Vivencio Barrios, MD, PhD; and Jos Sancho, MD, PhD
From Hospital Ramn y Cajal, Madrid, Spain.
Note: An earlier version of this paper was presented at the Endocrine Society’s 85th Annual Meeting in Philadelphia, Pennsylvania, 19-22 June 2003.
Acknowledgments: The authors thank Genoveva Gonzlez for excellent technical help; Robertas Bunevicius and Carlos Peir for invaluable advice regarding the tests of quality of life, mood, and psychometric functionality; and Ana Tabuenca for help with power analysis.
Grant Support: By Merck KgaA, Darmstad, Germany. Dr. BotellaCarretero is supported by fellowships from the Consejera de Educacin, Comunidad de Madrid (01/0430/01) and from the Fondo de Investigacin Sanitaria (01/F072), Instituto de Salud Carlos III, Ministerio de Sanidad y Consume, Spain.
Potential Financial Conflicts of Interest: Grants received: H.F. Escobar-Morreale (Merck KgaA, Darmstad, Germany).
Requests for Single Reprints: Hector F. Escobar-Morreale, MD, PhD, Department of Endocrinology, Hospital Raml, Carretera de Colmenar Km 9’1, E-28034 Madrid, Spain; e-mail, hescobarm.hrc@salud. mdrid.org.
Current author addresses and author contributions are available at www.annals.org.
Copyright American College of Physicians Mar 15, 2005