Bunevicius T4+T3 Extension Data and What Happened After the Trial Ended

At a glance

| Detail | Value | |---|---| | N | 33 (crossover design) | | Intervention | Substitution of 12.5 mcg liothyronine (T3) for 50 mcg of the patient's levothyroxine (T4) dose | | Comparator | T4 monotherapy at full prior dose | | Duration | Two 5-week treatment periods (crossover) | | Primary Endpoint | Composite of cognitive testing, mood scales (POMS, VAS), and clinical symptom scores | | Key Result | T4+T3 combination produced statistically significant improvements in 6 of 17 cognitive and mood measures versus T4 alone |

Why the extension question matters

The Bunevicius 1999 trial changed clinical conversation about hypothyroidism treatment overnight. Published in the New England Journal of Medicine, it reported that partially substituting T3 for T4 improved mood, cognition, and physical symptoms in 33 patients with overt hypothyroidism. The study became the most cited argument for combination therapy, generating patient demand that persists decades later.

But the trial lasted only 10 weeks total (two 5-week crossover periods). That brevity raised an immediate question: would the benefits hold up over months or years? And if they did not, what would that mean for the millions of patients who felt undertreated on T4 alone?

The trial design, revisited with critical distance

Bunevicius and colleagues used a double-blind, randomized crossover design in which each patient served as their own control. During one 5-week period, patients took their usual T4 dose. During the other, 50 mcg of T4 was replaced with 12.5 mcg of liothyronine. The substitution ratio (roughly 4:1) was chosen based on estimated relative potency, not pharmacokinetic matching.

HealthRX Durability Assessment Framework

To evaluate whether a short crossover trial's findings are likely to hold long-term, we apply four criteria:

| Criterion | Bunevicius 1999 Assessment | |---|---| | Washout adequacy | No washout period between crossover arms. T3's half-life is roughly 24 hours, but mood/cognition effects may lag behind serum clearance. Carryover contamination is plausible. | | Biological mechanism for sustained effect | Partial. T3 crosses the blood-brain barrier more readily than T4, but chronic supplementation may downregulate deiodinase activity, potentially attenuating the initial boost. | | Consistency across replication | Poor. Of 13+ subsequent RCTs, the majority failed to replicate mood or cognition benefits. | | Dose-response or duration-response signal | Absent. No dose-ranging data. Longer trials (12-52 weeks) generally showed null results. |

This framework suggests the original 5-week signal may have captured an acute pharmacodynamic response rather than a durable clinical benefit.

What happened after: the search for extension data

No formal extension or open-label continuation of the Bunevicius trial was ever published. The 33 patients returned to their usual T4 monotherapy at trial end. Bunevicius himself published a second study in 2002 examining combination therapy specifically in thyroidectomized patients (n=10), which did show some persistent mood benefits, but the sample was too small and the population too narrow to serve as a true extension.

The absence of follow-up data is itself informative. In clinical research, trials that show strong durable effects tend to generate extension protocols. The fact that no long-term data was pursued from this cohort suggests either logistical constraints or early awareness that the short-term signal might not persist.

The replication record: 1999 to 2009

The decade following the Bunevicius trial produced a string of replication attempts. Their collective failure to confirm durable benefits is the closest thing we have to "long-term follow-up" of the original hypothesis.

| Study | Year | N | Duration | T3 Dose | Mood/Cognition Result | |---|---|---|---|---|---| | Bunevicius (thyroidectomy) | 2002 | 10 | 5 weeks (crossover) | 12.5 mcg | Some mood benefit | | Walsh et al. | 2003 | 101 | 10 weeks | 12.5 mcg | No difference | | Sawka et al. | 2003 | 40 | 15 weeks | 12.5 mcg | No difference | | Clyde et al. | 2003 | 44 | 4 months | 12.5 mcg | No difference | | Siegmund et al. | 2004 | 23 | 12 weeks (crossover) | Weight-based | No difference | | Escobar-Morreale et al. | 2005 | 28 | 8 weeks (crossover) | 5 mcg | Patient preference for combo, no objective difference | | Appelhof et al. | 2005 | 141 | 15 weeks | 10 or 25 mcg | No objective difference; more dissatisfaction at 25 mcg | | Saravanan et al. | 2005 | 697 | 12 months | 10 mcg | No difference at 12 months |

The Saravanan trial is especially relevant. At 697 patients and 12 months of follow-up, it is the largest and longest RCT to test the combination hypothesis. It found no significant difference in psychological well-being, quality of life, or cognitive function. If the Bunevicius effect were durable, a 12-month trial of this size should have detected it.

Regression to the mean and the novelty hypothesis

Several explanations have been proposed for why the original trial succeeded where its successors did not.

Regression to the mean. The Bunevicius crossover had no washout period between arms. Patients who happened to score poorly on mood measures during one period would statistically tend to score closer to their average during the next, regardless of treatment. In a crossover without washout, this can mimic a treatment effect, particularly when sample size is small (n=33).

Carryover pharmacology. Liothyronine's serum half-life is approximately 18 to 24 hours, but its tissue effects, particularly in the central nervous system, may take longer to fully dissipate. Without a washout, residual T3 activity during the T4-only period could have made the monotherapy arm look worse than true monotherapy.

Expectation and novelty. Patients who have felt suboptimal on T4 for years may respond to any change in regimen. The Escobar-Morreale trial notably found that patients preferred the combination arm despite no measurable cognitive or mood difference, suggesting a placebo or novelty component that would fade with prolonged use.

Safety signals from longer-duration studies

While the original Bunevicius trial reported no adverse events, longer studies and observational data revealed concerns that would only emerge with extended exposure.

TSH suppression. Several replication trials found that combination therapy tended to suppress TSH below the lower reference range, even at modest T3 doses. The Appelhof 2005 study showed that 25 mcg T3 pushed mean TSH well below 0.4 mIU/L. Sustained subclinical hyperthyroidism carries risks for atrial fibrillation and accelerated bone loss, per ATA guidelines.

Cardiac considerations. T3 has direct chronotropic effects on the heart. Short-term trials may not capture arrhythmia risk. The FDA-approved prescribing information for liothyronine warns against use in patients with cardiovascular disease and notes that even therapeutic doses can precipitate angina or arrhythmias in susceptible individuals.

Bone mineral density. Observational cohort studies of patients on suppressive thyroid hormone regimens have consistently shown reduced BMD, particularly at cortical bone sites. While the 12.5 mcg substitution dose used in most trials is modest, long-term use without TSH monitoring could drift patients into a suppressive range.

What the guidelines concluded

Professional societies weighed this body of evidence and reached a cautious consensus.

The 2014 ATA guidelines for treatment of hypothyroidism made a weak recommendation against routine use of combination T4+T3 therapy, citing the lack of consistent benefit across RCTs and the absence of long-term safety data. They acknowledged that a subset of patients might benefit, particularly those with persistent symptoms despite normal TSH on T4 monotherapy, but stressed that this remained unproven.

The 2012 ETA guidelines similarly concluded that existing evidence did not support combination therapy as standard treatment, while leaving the door open for carefully monitored trials in selected patients.

Neither guideline cited extension data from the Bunevicius trial, because none existed. The guideline committees explicitly noted the short duration and small sample size of the original study as key limitations.

The DIO2 polymorphism angle

One line of post-trial investigation tried to rescue the combination hypothesis by identifying a genetic subgroup that might respond preferentially. The type 2 deiodinase gene (DIO2) Thr92Ala polymorphism attracted attention because it could theoretically impair T4-to-T3 conversion in the brain.

The 2009 Panicker et al. study found that patients with the Thr92Ala variant reported worse baseline well-being on T4 monotherapy and showed a trend toward improvement on combination therapy. This was a secondary analysis, not a prospective trial, and subsequent genetic studies have not consistently replicated the finding.

Even if the polymorphism is real, it would not address the durability question. No trial has tracked DIO2-stratified patients on combination therapy beyond 12 weeks.

What we still do not know

The Bunevicius trial opened a door that 25 years of subsequent research has not fully closed. Key unanswered questions include:

• Whether sustained-release T3 formulations (not commercially available in 1999) might produce durable benefits without the TSH suppression seen with immediate-release liothyronine
• Whether there is a true genetic subgroup (DIO2 or otherwise) that derives lasting benefit
• Whether the subjective preference many patients report for combination therapy reflects a measurable neuropsychological effect or an expectation-driven response
• What the cardiovascular and skeletal safety profile looks like beyond 12 months of continuous T3 supplementation

The original trial remains a landmark for asking the right question. The answer, based on everything that followed, is that acute T3 supplementation may produce short-lived improvements in some patients, but durable superiority of combination therapy over T4 monotherapy has not been established.

Frequently asked questions

›

›

›

›

›

›

›

›

›

›

References

1. 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(6):424-429. PubMed
2. Saravanan P, Simmons DJ, Visser TJ, Dayan CM. Randomized controlled trial testing the effects of thyroxine plus triiodothyronine combination therapy. J Clin Endocrinol Metab. 2005;90(2):805-812. PubMed
3. Jonklaas J, Bianco AC, Bauer AJ, et al. Guidelines for the treatment of hypothyroidism. Thyroid. 2014;24(12):1670-1751. PubMed
4. Panicker V, Saravanan P, Vaidya B, et al. Common variation in the DIO2 gene predicts baseline psychological well-being and response to combination thyroxine plus triiodothyronine therapy. J Clin Endocrinol Metab. 2009;94(5):1623-1629. PubMed
5. Appelhof BC, Fliers E, Wekking EM, et al. Combined therapy with levothyroxine and liothyronine in two ratios, compared with levothyroxine monotherapy. J Clin Endocrinol Metab. 2005;90(5):2666-2674. PubMed
6. FDA. Liothyronine sodium tablets prescribing information. FDA Label