Benefits and Harms of Levothyroxine/L-Triiodothyronine Versus Levothyroxine Monotherapy for Adult Patients with Hypothyroidism: Systematic Review and Meta-Analysis

Abstract

Background:

Combined therapy with levothyroxine (LT4)/L-triiodothyronine (LT3) has garnered attention among clinicians and patients as a potential treatment alternative to LT4 monotherapy. The objective of this study was to compare the benefits and harms of LT4/LT3 combined therapy and LT4 monotherapy for patients with hypothyroidism.

Methods:

A systematic search in MEDLINE, Scopus, EMBASE, Web of Science, and Cochrane Central Register of Controlled Trials was performed by a librarian from inception date until September 2020. Randomized clinical trials and quasiexperimental studies comparing combined therapy (LT4/LT3) versus monotherapy (LT4) for adult patients with hypothyroidism were considered for inclusion. Independent data extraction was performed by paired reviewers. A meta-analysis comparing standardized mean differences of the effect of each therapy was performed on clinical outcomes and patient preferences. Proportions of adverse events and reactions were assessed narratively.

Results:

A total of 1398 references were retrieved, from which 18 fulfilled the inclusion criteria. Results supported by evidence at low-to-moderate certainty evidence did not display a difference in treatment effect between therapies on clinical status, quality of life, psychological distress, depressive symptoms, and fatigue; all measured with standardized questionnaires. Furthermore, meta-analysis of patient preferences revealed higher proportions of choice for combined therapy (43%) when compared with monotherapy (23%) or having no preference (30%). When evaluating treatment adverse events or adverse reactions, similar proportions were observed between treatment groups; meta-analysis was not possible.

Conclusions:

The available evidence at low-to-moderate certainty demonstrates that there is no difference in clinical outcomes between LT4/LT3 combined therapy and LT4 monotherapy for treating hypothyroidism in adults, except for a higher proportion of patient preferring combined therapy. Adverse events and reactions appear to be similar across both groups, however, this observation is only narrative. These results could inform shared decision-making conversations between patients with hypothyroidism and their clinicians.

PROSPERO Registration ID:

CRD42020202658.

Introduction

Before 1970, the standard of care for the treatment of hypothyroidism consisted of desiccated thyroid extract, which contained thyroid hormones: thyroxine (T4) and triiodothyronine (T3) (1). Concerns regarding its variability and high rate of overdose-related symptoms led to adopting levothyroxine (LT4) monotherapy as a safer alternative (2). To date, major professional organizations and guidelines have overwhelmingly recommended LT4 monotherapy as standard of treatment for hypothyroidism (3 –5).

LT4 monotherapy has been proven to be effective both in normalizing thyrotropin (TSH) and T4 (free T4) levels (3). Still, a significant number of patients complain of hypothyroidism-related symptoms and/or have persistent dissatisfaction with treatment (6,7). Fatigue, decrease in health-related quality of life (QoL), and cognitive and psychological impairment are among the main complaints (2,8 –11). Among proposed treatment alternatives, LT4/L-triiodothyronine (LT3) combined therapy has garnered much attention among patients and clinicians (5,12 –15). From 2000 to 2009, several clinical trials and a meta-analysis compared LT4/LT3 combined with LT4 monotherapy without evidence of superiority of any treatment modality (2,5,16,17 –20). Consequently, by 2012, the European Thyroid Association (ETA) guidelines stated that LT4/LT3 combined therapy could be used as an experimental therapy in some LT4-treated patients with persistent hypothyroidism-related symptoms despite normalized thyroid hormone levels (21).

Although clinical evidence does not support the routine use of LT4/LT3 combination therapy, patient preferences and evidence of adequate restoration of thyroid function lead to controversies and practice variation. In fact, a 2017 survey of members of the American Thyroid Association (ATA) showed that approximately one-third would consider alternative treatment options such as combined therapy for their patients depending upon the circumstances (22). The last meta-analysis that compared clinical outcomes between these treatment modalities was published more than a decade ago; additional clinical trials are now available for analysis (23 –28). Patients and clinicians would benefit from a comprehensive evaluation of the most updated clinical evidence. To further clarify the value of LT4/LT3 therapy, the primary aim of this systematic review and meta-analysis was to assess the benefits and harms of LT4/LT3 combined therapy when compared with LT4 monotherapy for treatment of adult patients with hypothyroidism.

Methods

Protocol registration and guidelines

This review adhered to the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) statement. Before initiating, a review protocol was submitted and accepted by the International Prospective Register for Systematic Reviews (PROSPERO) with the following ID: CRD42020202658.

Eligibility criteria

Studies that were considered for inclusion were randomized clinical trials (RCTs) of parallel or crossover designs and quasiexperimental studies (an interventional study design in which the independent variable is manipulated, where participants are not randomly assigned to conditions or orders of conditions) comparing the efficacy of LT4/LT3 synthetic combined therapy or desiccated thyroid extracts versus LT4 monotherapy on adult patients with a diagnosis of either primary or central hypothyroidism, independent of intervention time or dose. There were no language restrictions.

Outcomes

We evaluated the following outcomes measured quantitatively and by previously validated questionnaires: treatment effect on clinical status (measured by any clinical scoring or grading system for hypothyroidism; e.g., Billewicz score, Zulewski's clinical score, Thyroid Symptoms Questionnaire), QoL (e.g., Hypothyroid Health-Related Quality-of-Life Questionnaire), psychological distress (measured by the global score of scales aimed at giving a general measure of psychological distress; e.g., General Health Questionnaire, Symptoms Checklist, Profile of Mood States), depressive symptoms (e.g., Beck Depression Inventory), and fatigue (e.g., Piper Fatigue Scale). Other outcomes that were searched for were anxiety symptoms, insomnia, somatic symptoms, social dysfunction, all-cause mortality, proportion of participants achieving suppressed TSH, and cognitive parameters.

We also evaluated patient preferences for each treatment modality and adverse treatment events (medical occurrence temporally associated with the use of a medicinal product, but not necessarily causally related) or reactions (a response to a drug, which is noxious and unintended, and which occurs at doses normally used for the prophylaxis, diagnosis, or therapy of disease, or for the modification of physiological function) as the proportion of patients preferring any treatment or experiencing any adverse event/reaction.

Information sources and search strategy

An experienced librarian performed the search strategy with input from the study investigators. The following scientific databases were assessed in our search strategy: PubMed, Scopus, EMBASE, Web of Science, and COCHRANE Central database. Searches were made from each database inception date until September 2020 with no language restrictions. The search strategy consisted of a series of Medical Subject Heading terms and keywords that relate to the population (hypothyroidism), intervention (synthetic LT4/LT3 combined therapy, desiccated thyroid extracts), and comparison (LT4 monotherapy) of interest. Previous systematic and narrative reviews on the subject were screened for any missing references as well as other listed references in the studies that were included. Gray literature was addressed via the Open System for Information on Grey Literature (OpenSIGLE) database (opensigle.inist.fr) and the National Technical Information Service (NTIS) database (www.ntis.gov/).

Selection process

Independent reviewers worked at screening each reference in a duplicate manner during the selection process to ensure eligibility. Screening process consisted of the following two phases, a title/abstract and full-text screening phase, respectively. During the first phase, any study in which there was a decision conflict between reviewers was included and reviewed in the full-text phase where any conflict was resolved by either consensus or the intervention of a third reviewer. Before each phase started, a pilot test was performed between reviewers to ensure adequate inter-rater agreement (defined as a kappa index equal as or higher than 0.7). The selection process was performed on Distiller SR Software.

Data collection process

Independent members of the research team worked to collect information from each included study in a duplicate manner. Any conflict during data collection was resolved by either consensus or the intervention of a third reviewer. The following data were collected for baseline description: study characteristics (author, year, and country where study was performed), population characteristics (age, gender, etiology of hypothyroidism, and years of diagnosis), biochemical and clinical parameters (baseline and follow-up weight, body mass index, total/free T3, total/free T4, TSH, and cholesterol levels), presence of the Thr92Ala-type II iodothyronine deiodinase (DIO2)-polymorphism in study population, intervention characteristics (LT4 dose, LT4/LT3 dose with T4:T3 ratio, frequency and time of administration), and type of combined therapy (synthetic or desiccated thyroid extract). Furthermore, the following data were collected for outcome analysis: baseline and follow-up scores for validated scales reporting clinical status, QoL, psychological distress, depressive symptoms, fatigue, anxiety symptoms, insomnia, somatic symptoms, social dysfunction, all-cause mortality, proportion of participants achieving suppressed TSH, and cognitive parameters. In addition, we assessed patient preferences and adverse events and reactions from each intervention.

Risk of bias in individual studies

Working independently and in duplicate, reviewers appraised the risk of bias of the RCTs using Cochrane's RoB 2.0 tool: a revised tool for assessing risk of bias in randomized trials. Risk of bias was assessed for the primary outcome of each study. Studies were rated as “high,” “some concerns,” or “low” risk of bias for each domain with an overall assessment also obtained for the study. For quasiexperimental studies, risk of bias was assessed using the Risk of Bias in Non-Randomized Studies -of Interventions (ROBINS-I).

Data synthesis and analysis

Data regarding population characteristics, weight, body mass index and biochemical parameters (thyroid function measures and cholesterol levels), patient preferences, and adverse events/reactions are narratively presented in the results. A meta-analysis was performed to assess the effect of combined therapy versus monotherapy on scales evaluating clinical status, QoL, psychological distress, depressive symptoms, and fatigue. Before analysis, mean differences between follow-up and baseline measures were obtained for each treatment group. Furthermore, standard deviations (SDs) for these mean differences were also calculated by either obtaining the standard error from the difference between the effect of each treatment and later transforming it into an SD or by calculating one from the baseline and follow-up SD for each treatment group by using a prespecified formula (SD = square root [((SD pretreatment)2 + (SD post-treatment)2) − (2R × SD pretreatment × SD post-treatment)]).

When the SD for the baseline or follow-up measure was not available for any given intervention in a study, it was imputed from either an available value in the study or obtained from another included study using the same scale for evaluation. When a study presented results from a determined scale divided by domains rather than reporting an overall score, a sum or mean of these domains was obtained depending on if the nature of the scale allowed for a calculation of an overall score for analysis.

To perform the meta-analysis for each quantitative outcome, mean differences and SD from the baseline and follow-up assessments were needed for each intervention, all of which were obtained directly from the studies using the abovementioned methods. This information was introduced into the Review Manager Software version 5.3 (The Nordic Cochrane Centre, The Cochrane Collaboration, 2014) for meta-analysis. Since all outcomes were measured quantitatively and scales measuring each outcome were heterogeneous, standardized mean differences (SMD) were obtained for the analysis as this measure of effect size is appropriate for studies that report efficacy and is recommended by the Cochrane Collaboration, considering an effect size ≥0.5 clinically significant (29).

Heterogeneity across studies was evaluated with the I ² and Cochran's Q test. According to the obtained heterogeneity, meta-analysis was performed with a mixed-effect model via the inverse-variance or Mantel–Haenszel method when heterogeneity was lower than 50%. When heterogeneity was higher than 50%, a random-effects model via the DerSimonian and Laird method was used (30). Subgroup and sensitivity analyses were planned for the following variables: etiology of hypothyroidism, presence of Thr92Ala-DIO2 polymorphism, type of combined therapy (synthetic or desiccated thyroid extract), LT3 dosing, duration of therapy, study design (crossover vs. parallel), and risk of bias (low vs. some concerns/high).

A meta-analysis to obtain proportions of patients preferring combined therapy, monotherapy, or having no preference was also performed by using a binomial-normal model for meta-analysis of proportions via a generalized linear mixed model. Meta-analysis of proportions was performed in R statistical software version 4.0.2.

Confidence in cumulative evidence

The Grading of Recommendations, Assessment, Development and Evaluations (GRADE) approach was used to assess the quality of the body of evidence for each evaluated outcome.

Results

Study selection process

A total of 1399 references were retrieved by the search strategy and from additional sources. At the end of the selection process, 18 studies fulfilled the inclusion criteria for this review and 11 were used in the quantitative synthesis (meta-analysis) (Fig. 1). Reasons for excluding studies from meta-analysis included not evaluating outcomes of interest and the inability to obtain mean change scores due to missing information in studies (e.g., absence of baseline scores for a specific scale). All studies included in the meta-analysis were performed on participants with overt hypothyroidism.

FIG. 1.

Study selection flow diagram.

Study characteristics and biochemical parameters

Eighteen trials comprising a total of 1563 participants were included in this review (12,14,23 –28,31–40). Ten studies had a crossover design and eight were parallel trials (including one quasiexperimental study). Duration of treatment across studies ranged from 5 weeks to 12 months. In all studies but one, participants were mostly female. Administered LT4/LT3 and LT4 dose as well as other additional information regarding general study characteristics can be found in Table 1. We excluded the 1999 study by Bunevicius et al. (41) as patients were included in their 2000 report, which is included in this review (32). None of the included studies provided information regarding the proportion of participants carrying the Thr92Ala-DIO2 polymorphism. Information regarding baseline and follow-up values of thyroid (TSH, total and free T4, T3) and metabolic measures (weight, body mass index, total cholesterol, low-density lipoprotein, high-density lipoprotein, and triglycerides) for each treatment (monotherapy and combined therapy) is available in Supplementary Tables S1 and S2.

Table 1.

Study Design, Demographics, and Intervention Characteristics

Author, year	Design	Duration	N	Age (years)	Female (n, %)	Type of hypothyroidism	Time since diagnosis (years)	T4:T3 ratio	Combined therapy (LT4/LT3) dose	LT4/LT3 dosing frequency	Monotherapy (LT4) dose
Appelhof, 2005+ (36)	Parallel	15 weeks	141/a130	48.37 (9.54)	120 (85.1)	Overt	ND	10:1 5:1	25 μg of usual LT4 dose replaced by LT3 in a 10:1 manner 25 μg of usual LT4 dose replaced by LT3 in a 5:1 manner	Twice daily	Usual dose
Bunevicius, 2000 (32)	Crossover	5 weeks	26	46.1 (10.13)	26 (100)	Overt	ND	ND	50 μg of usual LT4 dose replaced by 12.5 μg of LT3	Once daily	Usual dose
Bunevicius, 2002 (12)	Crossover	5 weeks	10	36.25 (11.33)	10 (100)	Overt	ND	5:1	50 μg of usual LT4 dose replaced by 10 μg of LT3	Once daily	Usual dose
Clyde, 2003 (34)	Parallel	16 weeks	46/a44	44.15 (10.46)	36 (81.81)	Overt	ND	3:1	50 μg of usual LT4 dose replaced by 7.5 μg of LT3	Twice daily	Usual dose
Escobar-Morreale, 2005 (31)	Crossover	8 weeks	28	48 (11)	28 (100)	Overt	ND	15:1	75 μg of LT4 plus 5 μg of LT3	Once daily	100 μg
Fadeyev, 2005 (24)	Parallel	24 weeks	58	39.59 (9.14)	58 (100)	Overt	ND	ND	25–50 μg of usual LT4 dose replaced by 12 μg of LT3	Once daily	Usual dose
Fadeyev, 2010 (28)	Parallel	24 weeks	36	39.96 (9.68)	36 (100)	Overt	ND	ND	25 μg of LT4 replaced by 12.5 μg of LT3	ND	Usual dose
Hoang, 2013 (14)	Crossover	16 weeks	78/a70	46.22 (32.32)	53 (75.71)	Overt	ND	ND	Desiccated thyroid extract (dose not specified)	Once daily	—
Kaminski, 2016 (26)	Crossover	8 weeks	32	42.6 (13.3)	30 (94)	Overt	10.75 (8)	5:1	75 μg of LT4 plus 15 μg of LT3	Once daily	125 or 150 μg
Krysiak, 2018 (27)	Quasirandomized	24 weeks	37	30.54 (5.93)	37 (100)	Overt	2.11 (0.49)	5:1	Half of usual LT4 dose replaced by LT3 in a 5:1 manner	ND	Usual dose
Nygaard, 2009 (23)	Crossover	12 weeks	68/a59	47.04 (12.61)	55 (93.2)	Overt	ND	2.5:1	50 μg of LT4 plus 20 μg of LT3	Once daily	100 μg
Rodriguez, 2005 (38)	Crossover	6 weeks	27	47.5 (12.9)	25 (83)	Overt	ND	5:1	50 μg of usual LT4 dose replaced by 10 μg of LT3	Once daily	Usual dose
Saravanan, 2005 (35)	Parallel	12 months	697	57.34 (11.04)	584 (83.78)	ND	ND	5:1	50 μg of usual LT4 dose replaced by 10 μg of LT3	Once daily	Usual dose
Sawka, 2003 (39)	Parallel	15 weeks	40	47.25 (11.07)	36 (90)	Overt	9.2 (7.05)	2:1	Half of usual LT4 dose plus 12.5 μg of LT3	Twice daily	Usual dose
Siegmund, 2004 (40)	Crossover	12 weeks	26/a23	ND	21 (84)	Overt	ND	14:1	95% of the dose corresponded to LT4 and 5% to LT3	ND	Usual dose
Slawik, 2007 (37)	Crossover	5 weeks	32/a29	51 (10.77)	8 (27.5)	Central	ND	10:1	1.44 μg/kg of LT4 plus 0.16 μg/kg of LT3	ND	1.6 μg/kg
Walsh, 2003 (33)	Crossover	10 weeks	110/a101	47.7 (11.7)	101 (92)	Overt	8 (8.3)	5:1	50 μg of usual LT4 dose replaced by 10 μg of LT3	Once daily	Usual dose
Valizadeh, 2009 (25)	Parallel	16 weeks	71/a60	38.6 (11.37)	48 (80)	Overt	ND	4:1	50 μg of usual LT4 dose replaced by 12.5 μg of LT3	Twice daily	Usual dose

This study included two combined therapy groups (10:1 and 5:1 LT4/LT3 concentration) that were combined for the purpose of analysis; a: number of participants included in analysis; usual dose is dependent on each study definition; in the “N” column, the number of randomized participants is reported in every study and analyzed (a) participants reported when available.

LT3, L-triiodothyronine; LT4, levothyroxine; ND, no data.

Risk of bias

The overall risk of bias of the body of evidence was deemed as low to moderate, mostly driven by inconsistencies in the randomization domain. Among 17 RCTs, risk of bias was deemed to be low in nine. However, some concerns were observed in five, and three were deemed to present a high risk of bias. In the parallel RCTs, risk of bias was affected due to inconsistencies within the randomization domain and in the crossover trials due to deviations from the intended interventions. Additional information regarding evaluated domains during the risk-of-bias assessment of RCTs is available in Figure 2. For the quasiexperimental study, overall assessment was rated as critical risk of bias, due to concerns for bias regarding confounding, deviations from intended intervention, classification of the outcome, and selection of the reported result.

FIG. 2.

Risk-of-bias assessment.

Clinical outcomes

General results

A meta-analysis to compare the effect of each therapy was performed for the following outcomes: clinical status, QoL, psychological distress, depression, and fatigue, all of which were evaluated with clinical questionnaires evaluating the corresponding outcome. No differences were observed between combined therapy and monotherapy (Fig. 3). Used scales for the evaluation of each outcome across studies are listed in Supplementary Table S3, the following are some examples: Zulewski clinical score (Clinical Status), Billewicz score (Clinical Status), Short Form Health Survey (Quality of Life), General Health Questionnaire (Psychological Distress), Symptoms Checklist (Psychological Distress), Beck Depression Inventory (Depressive Symptoms), Center For Epidemiological Studies Depression Scale (Depressive Symptoms), and Piper Fatigue Scale (Fatigue). We were not able to obtain data for the following variables: anxiety symptoms, insomnia symptoms, somatic symptoms, social dysfunction, cognitive parameters, all-cause mortality, and proportion of participants with suppressed TSH after treatment.

FIG. 3.

Forest plot comparing the efficacy of LT3/LT4 combined therapy versus LT4 monotherapy on (A) clinical status, (B) quality of life, (C) psychological distress, (D) depressive symptoms, and (E) fatigue. LT3, L-triiodothyronine; LT4, levothyroxine.

Subgroup analyses

Performance of subgroup analysis was only possible for the following variables: duration of treatment, LT3 dose, and LT3 dosing frequency per day. The subgroups were defined according to the values available in the studies. Other planned subgroup analyses were not feasible due to lack of information.

Subgroup analysis according to duration of treatment (≤12 weeks vs. >12 weeks) was possible for change in clinical status, psychological distress, and depressive symptoms scores. Regarding clinical scores, monotherapy was favored when intervention time was ≤12 weeks (SMD, 0.15; 95% confidence interval [CI 0.02–0.28]; p, 0.02; I ², 0%). On the contrary, a tendency favoring combined therapy was observed when intervention time was >12 weeks, with no statistically significant differences between groups (SMD, −0.23 [CI −0.52 to 0.06]; p, 0.12; I ², 0%). Furthermore, no differences based on intervention time were observed for psychological distress and depressive symptoms (Supplementary Figs. S1–S3).

Regarding LT3 dose, subgroup analysis was possible for changes in depression and psychological distress scores, comparing studies that provided doses of 10 μg or less versus studies that provided more than 10 μg as part of the LT3 intervention. A tendency favoring combined therapy was observed when doses of LT3 were higher for both changes in depressive symptoms (SMD, −0.21 [CI −0.41 to −0.01]; p, 0.04; I ², 0%) and psychological distress (SMD, −0.2 [CI −0.42 to 0.01]; p, 0.06; I ², 0%), only observing a statistically significant difference on the former. When evaluating studies providing lower doses of LT3, no tendency favoring any group was observed (Supplementary Figs. S4 and S5).

Subgroup analysis regarding LT4/LT3 dosing frequency was possible for depressive symptoms and psychological distress outcomes, where studies were subgrouped according to if the administration regimen was given once or twice daily. No differences in treatment effects were observed after performing this subdivision (Supplementary Figs. S6 and S7).

Sensitivity analysis

Sensitivity analysis according to study design (parallel and crossover) and study quality (high quality and low/moderate quality) was possible for the evaluation of treatment effect on clinical status, depressive symptoms, and general psychological scores. In none of these analyses was a difference between groups observed. (Supplementary Figs. S8–S13).

Patient preferences

Of 10 included crossover trials, 6 evaluated patient preferences after study completion. Participants were asked to choose between preference for synthetic combined therapy, preference for monotherapy, and no preference at all. Pooled preference for combined therapy was 43% ([CI 34–52%]; I ² = 48%), compared with 23% of preference for monotherapy ([CI 14–35%]; I ² = 77%). Furthermore, 30% ([CI 21–41%]; I ² = 66%) of participants reported no preference for any treatment (Fig. 4).

FIG. 4.

Forest plot of patient preferences across crossover trials. (A) Preference for combined therapy, (B) preference for monotherapy, and (C) no preference.

Adverse events and reactions related to combined therapy

Overall, incidence of these with combined therapy did not appear to differ when compared with monotherapy. However, a statistical comparison was not performed due to the heterogeneous form of reporting across studies, and judgment was based on each study conclusions.

A total of 13 studies including 829 patients randomized to combined therapy reported reasons for treatment discontinuation, where a total of 60 (7.23%) participants withdrew due to adverse events or reactions. From these, 18 were due to possibly hyperthyroidism-related symptoms (tachycardia, nervousness, heat intolerance, weight loss, headaches, tremors, feelings of fainting, nightmares, and sleeping difficulty), 32 due to possibly hypothyroidism-related symptoms (fatigue, poor performance, depressive symptoms, weight gain, increasing tiredness, sluggish), and 8 to nonspecified symptoms. In one study, treatment was discontinued in one subject who experienced atrial fibrillation with arrhythmia associated with suppressed TSH, and in another one, a subject discontinued therapy due to digestive symptoms. Furthermore, among 793 subjects randomized to monotherapy, 50 (6.3%) participants withdrew due adverse events or reactions (Supplementary Table S4).

A total of 8 studies comprising 390 patients randomized to combined therapy reported adverse reactions among participants who completed the intervention, where a total of 18 were observed. Moreover, among 391 subjects randomized to monotherapy, 29 adverse reactions were observed. In both groups, these included palpitations, excessive sweating, psychological instability, agitation, fatigue, among other nonspecified ones (Supplementary Table S4).

Assessment of the certainty of evidence

The certainty of evidence across outcomes was rated as low to moderate according to the GRADE approach. Concerns on the specific outcome regarding risk of bias were identified for change in clinical and fatigue scores as well as patient preference assessment. Inconsistency due to significant heterogeneity was observed for the change in QoL outcome. Indirectness was observed due to differences in intervention and time differences in outcomes across all outcomes (Supplementary Table S5).

Discussion

Previous evidence has described that even when LT4 normalizes TSH levels, euthyroidism may not be achieved in all tissues, and combination therapy with LT4/LT3 could help achieve this goal (42). As such, it can be suggested that LT4 monotherapy may not be the ideal treatment for all patients with hypothyroidism, yet, it is the most used as most patients will be benefited and be satisfied (8). However, about 10% of patients with hypothyroidism, on monotherapy and have normal TSH levels, will complain of persistent disease-related symptomatology. Although the reason for their lingering symptoms is not known, posited theories include the following: chronic nature of hypothyroidism, associated autoimmune diseases, thyroid autoimmunity, as well as inadequacy of LT4 dose, or regimen use for thyroid hormone replacement (8).

One important hypothesis for unresponsiveness to monotherapy in this subset of patients could be related to the presence of a single-nucleotide polymorphism DIO2 Thr92Ala, which has been shown to reduce the activity of the DIO2 enzyme, which converts T4 (LT4) into its active form, triiodothyronine (T3). The presence of this polymorphism has been linked to psychological impairment among patients on monotherapy, better response to combined therapy, and higher preferences for it (13,15). As such, in the face of a significant subset of patients who do not achieve clinical euthyroidism even when by laboratory parameters it may be apparent, alternative therapies for the disease have been explored lately, such as LT4/LT3 combined therapy (8).

In this comprehensive systematic review and meta-analysis aiming to compare the benefits and harms of LT4/LT3 combined therapy versus LT4 monotherapy in the treatment of hypothyroidism, we found no difference in clinical scores, QoL, psychological distress, depression, and fatigue. As part of our subgroup analysis, we only observed a statistically significant effect favoring LT4/LT3 therapy on depressive symptoms among studies where the dose of LT3 was higher than 10 μg. However, this statistical difference may not be important on a clinical level (29). Furthermore, we observed that the incidence of adverse events/reactions among participants on combined therapy did not tend to differ substantially from those on monotherapy, yet no statistical comparisons were performed.

Regarding patient preferences, combined therapy was more frequently preferred when compared with monotherapy or having no preference. Our results support previous meta-analysis on the topic, concluding that based on the available evidence, there is no difference in clinical outcomes (e.g., clinical status, QoL, psychological distress, depressive symptoms, fatigue); however, one interesting finding of this review was the higher observed preference for combined therapy (43%) when compared with monotherapy (23%). Moreover, this is the first review to summarize any kind of observed treatment adverse events or reactions from treatment across clinical trials of combined therapy. It is also the first systematic review to evaluate clinical outcomes, patient preferences, and adverse events/reactions as a whole.

Although clinical evidence does not demonstrate a clear benefit stemming from LT4/LT3 combined therapy for hypothyroidism when compared with LT4 monotherapy, crossover trials have demonstrated that a significant number of participants would still prefer this form of treatment when compared with monotherapy, which is supported by our meta-analysis (12,16,23,38). Furthermore, a recent meta-analysis evaluating treatment preferences for combined therapy also confirmed these results, where authors reported that almost half of the participants tend to prefer this treatment modality (43). Similarly to combined therapy, patients also prefer desiccated thyroid extract over LT4 monotherapy citing negative past experiences with synthetic LT4, as well as clinical improvement in symptoms as their reasons; however, this report is based on online posts rather than controlled studies (44).

The previously mentioned raises some concerns on whether previous clinical trials have been adequately designed to assess the question of interest, which is treatment efficacy on clinical outcomes. In addition, our review allows evaluation of the methods used on previous studies to evaluate the treatment efficacy (LT4 vs. LT4/LT3 treatment) and help us plan future studies, in other words, what we can learn from the included population, combined therapy regimens, evaluation of predictors, and outcome assessment. For example most, if not all clinical trials, have not specifically selected hypothyroid patients dissatisfied on monotherapy (45). These studies rather had a more general approach for selecting participants with hypothyroidism. Moreover, little emphasis has been placed on factors such as residual thyroid function, presence of thyroid peroxidase antibodies, Thr92AlaD2 polymorphisms, among others, in the process of patient recruitment. Selecting an appropriate patient population for combined therapy in the setting of an adequately powered study would be a crucial step in the performance of new clinical trials on the topic (45,46).

Some aspects that could have hindered results in studies included in this meta-analysis are the use of once-daily LT4/LT3 preparations (45,47,48), heterogeneous T4:T3 dose ratios, use of generic patient-reported outcome measures (PROMs), or inadequately calculated sample size among others. To improve and homogenize the design of future studies, new evidence has suggested the use of twice-daily LT4/LT3 regimens, disease-specific PROMs for evaluating response to treatment (specifically focusing on thyroid-related QoL), optimal trial duration, and use of parallel designs rather than crossover studies, among other several considerations (45,46). A recent consensus document based on a joint conference held by the ATA, British Thyroid Association (BTA), and ETA agreed on that there was equipoise for new clinical trials of combined therapy in hypothyroidism, taking into consideration new areas of opportunity designed in this document for future redesigned clinical trials that would inform future treatment recommendations (45).

This systematic review and meta-analysis provide the most contemporary and robust evidence regarding the use of combined therapy (LT4/LT3) versus monotherapy (LT4) for the treatment of hypothyroidism. Some limitations, however, must be specified. First, key aspects related to the design of the included trials could bias results, as it has been previously described. Second, half of the included studies were rated as having some concerns or high risk in the bias evaluation. Third, some studies were excluded from meta-analysis due to missing information.

Finally, some key factors that could have influenced our results have a relatively short duration of intervention, where more than half of the included studies had a duration of 12 weeks or less; a sample size smaller than 100 in 13 of the 18 included studies, which could have underpowered any statistically significant difference between groups; heterogeneity in outcome assessment, since it was not uncommon for different scales evaluating the same outcome to be reported across studies; and differently used T4:T3 dose ratios across studies being specially relevant as they had a direct effect on both outcome efficacy and treatment adverse events or reactions. In the process of determining the true efficacy of combined therapy, future research should focus on establishing a standardized T4:T3 dose ratio as part of the treatment regimen and providing trials with an optimal sample size.

We have to place certain emphasis on the fact that more than half of the included studies only had a follow-up of ≤12 weeks as it may be insufficient. The consensus document released by the ATA/BTA/ETA states that future trials of combined therapy should last at least 12 months to capture relevant information regarding efficacy and safety, allow therapy adjustments and gather data on potential toxicity. Strengths of this study include a strict systematic review process that incorporates a sensitive search strategy, meticulous evaluation of the risk of bias across studies, and assessment of the certainty of evidence using the GRADE approach.

Our results highlight the importance of performing new clinical trials that take into consideration knowledge gained from the current body of literature in terms of patient selection, combined therapy regimen, and outcome assessment. In fact, the consensus statement supports the need for new trials that build up on current knowledge and provide significant insight into the role of combined therapy in the treatment of hypothyroidism (45).

Conclusion

The body of evidence at low-to-moderate certainty demonstrates that there is no difference in clinical and surrogate outcomes between LT4/LT3 combined therapy and LT4 monotherapy for the treatment of adult patients with hypothyroidism with the exception that patients preferred the use of combined therapy. Adverse events and reactions appear to be similar across groups, however, this observation is only narrative. These results could provide support for clinicians and patients with hypothyroidism in the shared decision-making process for initiating combined therapy. Future studies designed to address current study design limitations should help clinicians and patients have more confidence in the decision-making process.

Footnotes

Authors' Contributions

Drs. Rodríguez-Gutiérrez, Millán-Alanis, and González-González contributed to the conceiving of the research idea and developed the first draft of the research protocol. Drs. Singh Ospina, Brito, Maraka, and González-Velazquez revised and commented on the first version of the research protocol. All authors revised and approved the final version of the research protocol. Drs. Millán-Alanís, González-González, Flores-Rodríguez, Moreno-Peña, and González-Velázquez screened all the corresponding abstracts and full-text documents. Drs. Millán-Alanís, Flores-Rodríguez, and Moreno-Peña extracted the corresponding information from the included studies and appraised the risk of bias. Dr. Millán-Alanís performed the statistical analysis. Drs. Millán-Alanís, González-González, and Rodríguez-Gutiérrez worked on the first draft of the article. Drs. Singh Ospina, Brito, Maraka, and González-Velazquez revised and commented on the first version of the article. All authors contributed and approved the final version of the article.

Disclaimer

The content is solely the responsibility of the authors.

Author Disclosure Statement

No competing financial interests exist.

Funding Information

Dr. Maraka receives support by the Arkansas Biosciences Institute, the major research component of the Arkansas Tobacco Settlement Proceeds Act of 2000. Dr. Singh Ospina was supported by the National Cancer Institute of the National Institutes of Health under Award No. K08CA248972.

Supplementary Material

Supplementary Figure S1

Supplementary Figure S2

Supplementary Figure S3

Supplementary Figure S4

Supplementary Figure S5

Supplementary Figure S6

Supplementary Figure S7

Supplementary Figure S8

Supplementary Figure S9

Supplementary Figure S10

Supplementary Figure S11

Supplementary Figure S12

Supplementary Figure S13

Supplementary Table S1

Supplementary Table S2

Supplementary Table S3

Supplementary Table S4

Supplementary Table S5

References

Kaufman

, Gross

, Kennedy DL (1991) Thyroid hormone use: trends in the United

States from

. 1960. through 1988. Thyroid, 1:285–291.

McAninch

, Bianco AC (2016) The history and future of treatment of hypothyroidism. Ann Intern Med 164:50.

Jonklaas

, Bianco

, Bauer

, Burman

, Cappola

, Celi

, et al. (2014) Guidelines for the treatment of hypothyroidism: prepared by the American Thyroid Association Task force on thyroid hormone replacement. Thyroid 24:1670–1751.

Azizi F, Amouzegar A, Mehran L, Abdi H (2020) LT4 and slow release T3 combination: optimum therapy for hypothyroidism? Int J Endocrinol Metab 18:e100870.

McAninch

, Bianco AC (2019) The swinging pendulum in treatment for hypothyroidism: from (and toward?) combination therapy. Front Endocrinol (Lausanne) 10:446.

Saravanan

, Chau

, Roberts

, Vedhara

, Greenwood

, Dayan

(2002) Psychological well-being in patients on “adequate” doses of L-thyroxine: results of a large, controlled community-based questionnaire study. Clin Endocrinol (Oxf) 57:577–585.

Peterson

, Cappola

, Castro

, Dayan

, Farwell

, Hennessey

, et al. (2018) An online survey of hypothyroid patients demonstrates prominent dissatisfaction. Thyroid 28:707–721.

Wiersinga WM (2019) T4 + T3 combination therapy: any progress? Endocrine 66:70–78.

Wekking

, Appelhof

, Fliers

, Schene

, Huyser

, Tijssen

JGP

, et al. (2005) Cognitive functioning and well-being in euthyroid patients on thyroxine replacement therapy for primary hypothyroidism. Eur J Endocrinol 153:747–753.

10.

Panicker

, Evans

, Bjøro

, Åsvold

, Dayan

, Bjerkeset O (2009) A paradoxical difference in relationship between anxiety, depression and thyroid function in subjects on and not on T4: findings from the HUNT study. Clin Endocrinol (Oxf) 71:574–580.

11.

Peterson SJ, McAninch EA, Bianco AC (2016) Is a normal TSH synonymous with “euthyroidism” in levothyroxine monotherapy? J Clin Endocrinol Metab 101:4964–4973.

12.

Bunevicius

, Jakubonien

, Jurkevicius

, Cernicat

, Laas

, Prange AJ (2002) Thyroxine vs thyroxine plus triiodothyronine in treatment of hypothyroidism after thyroidectomy for

Graves' disease

. Endocrine 18:129–133.

13.

Panicker

, Saravanan

, Vaidya

, Evans

, Hattersley

, Frayling

, et al. (2009) Common variation in the DIO2 gene predicts baseline psychological well-being and response to combination thyroxine plus triiodothyronine therapy in hypothyroid patients. J Clin Endocrinol Metab 94:1623–1629.

14.

Hoang

, Olsen

, Mai

, Clyde

, Shakir MKM (2013) Desiccated thyroid extract compared with levothyroxine in the treatment of hypothyroidism: a randomized, double-blind, crossover study. J Clin Endocrinol Metab 98:1982–1990.

15.

Carlé

, Faber

, Steffensen

, Laurberg

, Nygaard

(2017), Hypothyroid patients encoding combined MCT10 and DI, O2 gene polymorphisms may prefer L-T3 + L-T4 combination treatment—data using a blind, randomized, clinical study. Eur Thyroid J 6:143–151.

16.

Escobar-Morreale

, Botella-Carretero

, Escobar Del Rey

, Morreale De Escobar G (2005) Review: treatment of hypothyroidism with combinations of levothyroxine plus liothyronine. J Clin Endocrinol Metab 90:4946–4954.

17.

Grozinsky-Glasberg

, Fraser

, Nahshoni

, Weizman

, Leibovici L (2006) Thyroxine-triiodothyronine combination therapy versus thyroxine monotherapy for clinical hypothyroidism: meta-analysis of randomized controlled trials. J Clin Endocrinol Metab 91:2592–2599.

18.

Joffe

, Brimacombe

, Levitt

, Stagnaro-Green A (2007) Treatment of clinical hypothyroidism with thyroxine and triiodothyronine: a literature review and metaanalysis. Psychosomatics 48:379–384.

19.

Chao

, Jiawei

, Xia

, Guoming

, Yangang

, Xufu

, et al. (2009) Thyroxine alone or thyroxine plus triiodothyronine replacement therapy for hypothyroidism. Nucl Med Commun 30:586–593.

20.

Medici

, la Cour

, Michaelsson

, Faber

, Nygaard B (2017) Neither baseline nor changes in serum triiodothyronine during levothyroxine/liothyronine combination therapy predict a positive response to this treatment modality in hypothyroid patients with persistent symptoms. Eur Thyroid J 6:89–93.

21.

Wiersinga

, Duntas

, Fadeyev

, Nygaard

, Vanderpump MPJ (2012) 2012. ETA guidelines: the use of L-T4 + L-T3 in the treatment of hypothyroidism. Eur Thyroid J, 1:55–71.

22.

Jonklaas

, Tefera

, Shara N (2019) Prescribing therapy for hypothyroidism: influence of physician characteristics. Thyroid 29:44–52.

23.

Nygaard

, Jensen

, Kvetny

, Jarløv

, Faber

. (2009) Effect of combination therapy with thyroxine (T4) and 3,5,3′-triiodothyronine versus T4 monotherapy in patients with hypothyroidism, a double-blind, randomised cross-over study. Eur J Endocrinol 161:895–902.

24.

Fadeyev

, Morgunova

, Sytch

, Melnichenko

. (2005) TSH and thyroid hormones concentrations in patients with hypothyroidism receiving replacement therapy with L-thyroxine alone or in combination withL-triiodothyronine. Hormones (Athens) 4:101–107.

25.

Valizadeh

, Seyyed-Majidi

, Hajibeigloo

, Momtazi

, Musavinasab

, Hayatbakhsh

, (2009) Efficacy of combined levothyroxine and liothyronine as compared with levothyroxine monotherapy in primary hypothyroidism: a randomized controlled trial. Endocr Res 34:80–89.

26.

Kaminski

, Miasaki

, Paz-Filho

, Graf

, de Carvalho GA (2016) Treatment of hypothyroidism with levothyroxine plus liothyronine: a randomized, double-blind, crossover study. Arch Endocrinol Metab 60:562–572.

27.

Krysiak

, Szkróbka

, Okopień B (2018) Sexual function and depressive symptoms in young women with hypothyroidism receiving levothyroxine/liothyronine combination therapy: a pilot study. Curr Med Res Opin 34:1579–1586.

28.

Fadeyev

, Morgunova

, Melnichenko

, Dedov

. (2010) Combined therapy with L-thyroxine and L-tiiodothyronine compared to L-thyroxine alone in the treatment of primary hypothyroidism. Hormones 9:245–252.

29.

Faraone SV (2008) Interpreting estimates of treatment effects: implications for managed care. P T 33:700–711.

30.

Higgins, Julian

, et al., eds. Cochrane handbook for systematic reviews of interventions. John Wiley & Sons, 2019.

31.

Escobar-Morreale

, Botella-Carretero

, Gómez-Bueno

, Galán

, Barrios

, Sancho

. (2005) Thyroid hormone replacement therapy in primary hypothyroidism: a randomized trial comparing L-thyroxine plus liothyronine with L-thyroxine alone. Ann Intern Med. 142.

32.

Bunevicius

, Prange AJ (2000) Mental improvement after replacement therapy with thyroxine plus triiodothyronine: relationship to cause of hypothyroidism. Int J Neuropsychopharmacol 3:167–174.

33.

Walsh

, Shiels

, Lim

, Bhagat

, Ward

, Stuckey

BGA

, et al. (2003) 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 88:4543–4550.

34.

Clyde

, Harari

, Getka

, Mohamed Shakir KM (2003) Combined levothyroxine plus liothyronine compared with levothyroxine alone in primary hypothyroidism: a randomized controlled trial. JAMA 290:2952–2958.

35.

Saravanan

, Simmons

, Greenwood

, Peters

, Dayan CM (2005) Partial substitution of thyroxine (T4) with tri-iodothyronine in patients on T4 replacement therapy: results of a large community-based randomized controlled trial. J Clin Endocrinol Metab 90:805–812.

36.

Appelhof

, Fliers

, Wekking

, Schene

, Huyser

, Tijssen

JGP

, et al. (2005) Combined therapy with levothyroxine and liothyronine in two ratios, compared with levothyroxine monotherapy in primary hypothyroidism: a double-blind, randomized, controlled clinical trial. J Clin Endocrinol Metab 90:2666–2674.

37.

Slawik

, Klawitter

, Meiser

, Schories

, Zwermann

, Borm

, et al. (2007) Thyroid hormone replacement for central hypothyroidism: a randomized controlled trial comparing two doses of thyroxine (T4) with a combination of T4 and triiodothyronine. J Clin Endocrinol Metab 92:4115–4122.

38.

Rodriguez

, Lavis

, Meininger

, Kapadia

, Stafford LF (2005) Substitution of liothyronine at a 1:5 ratio for a portion of levothyroxine: effect on fatigue, symptoms of depression, and working memory versus treatment with levothyroxine alone. Endocr Pract 11:223–233.

39.

Sawka

, Gerstein

, Marriott

, MacQueen

, Joffe

. (2003) 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, 88:4551–4555.

40.

Siegmund

, Spieker

, Weike

, Giessmann

, Modess

, Dabers

, et al. (2004) 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. Clin Endocrinol (Oxf) 60:750–757.

41.

Bunevicius

, Kazanavicius

, Zalinkevicius

, Prange

Jr . (1999) Effects of thyroxine as compared with thyroxine plus triiodothyronine in patients with hypothyroidism. N Engl J Med 340:424–429.

42.

Escobar-Morreale

H F

, Del-Rey

, Obregon

, de Escobar

(1996). Only the combined treatment in all tissues the thyroidectomized by passing this aqueous phase through. Endocrinology, 137:2490–2502.

43.

Akirov

, Fazelzad

, Ezzat

, Thabane

, Sawka AM (2019) A systematic review and meta-analysis of patient preferences for combination thyroid hormone treatment for hypothyroidism. Front Endocrinol (Lausanne) 10:477.

44.

Toloza

FJK

, Suarez

NRE

, El Kawkgi

, Golembiewski

, Ponce

, Yao

, et al. (2020) Patient experiences and perceptions associated with the use of desiccated thyroid extract. Medicina 56:161.

45.

Jonklaas

, Bianco

, Cappola

, Celi

, Fliers

, Heuer

, et al. (2020) Evidence-based use of levothyroxine/liothyronine combinations in treating hypothyroidism: A consensus document. Thyroid 31:156–182.

46.

Hoermann

, Midgley

JEM

, Larisch

, Dietrich

. (2018) Lessons from randomised clinical trials for triiodothyronine treatment of hypothyroidism: have they achieved their objectives? J Thyroid Res 2018:3239197.

47.

Jonklaas

, Burman KD (2016) Daily administration of short-acting liothyronine is associated with significant triiodothyronine excursions and fails to alter thyroid-responsive parameters. Thyroid 26:770–778.

48.

Van Tassell

, Wohlford

, Linderman

, Smith

, Yavuz

, Pucino

, et al. (2019) Pharmacokinetics of L-triiodothyronine in patients undergoing thyroid hormone therapy withdrawal. Thyroid 29:1371–1379.