Levothyroxine dosage and the limitations of current bioequivalence standards
James V Hennessey About the author
Correspondence Rhode Island Hospital—Brown University, Hallett Center for Diabetes and Endocrinology, 1 Hoppin Street, Providence, RI 02903, USA
Email james_hennessey@brown.edu
Precision in the dosing of levothyroxine (LT4) products is guided by sensitive TSH measurements, and is essential to optimize therapeutic outcomes and assure patient safety. FDA regulation of LT4 formulations aims to provide consistent drug content and bioavailability of the individual products approved for distribution in the US.1, 2 What is unclear at this point, however, is whether true generic interchangeability among approved products has been achieved.
Once absorbed, orally dosed LT4 exerts physiologic effects on metabolism, and a dynamic feedback effect on the hypothalamic–pituitary axis.3 In healthy individuals, the resultant modulation of the production and clearance of endogenous T4 probably affects total circulating hormone levels. As LT4 is converted to T3 for therapeutic effect, measuring the concentration of the active ingredient (T3) at the site of action—as is the intent of bioequivalence determinations—becomes impractical. In this situation, the quantification of total T4 levels in the blood serves as the FDA-designated pharmacokinetic measure of choice.4
Precise and sensitive determination of therapeutic equivalence is important when comparing drugs acknowledged as having narrow therapeutic ratios.4 In a study of hypothyroid patients on optimal replacement therapy (mean dose 108 
g/day)—determined by their TSH response to TSH-releasing hormone—participants received a series of doses differing by up to 50 g daily.5 When patients ingested 25 g less than the optimal dose (a 0.23-fold difference), seven of nine patients (89%) were classified as hypothyroid. Conversely, when participants received 25 g more than optimal (also a 0.23-fold difference), 11 of 20 (55%) of patients were classified as thyrotoxic.5 The 25 g increment might be too large to reflect common clinical experience, however, as dosage adjustments of 
12% are frequently made to achieve therapeutic goals.
The stated purpose of FDA bioequivalence evaluations is to demonstrate therapeutic consistency between preparations,4 which should ensure that substitution of products found to be therapeutically equivalent can occur safely, and without the need for dose adjustment or follow-up therapeutic monitoring beyond that of the usual annual or semiannual interval. The choice of methods to document bioequivalence has recently been the subject of debate between clinicians and regulators.3 From a regulatory standpoint, the preferred method for documenting bioequivalence is the pharmacokinetic measurement of T4 levels.4 The justification for this position rests on the understanding that measurement of the active ingredient at the site of action is not possible, and that some relationship exists between the efficacy and safety of the product and the concentration of the active moiety in the systemic circulation.4
A previous attempt to demonstrate bioequivalence uncovered discrepancies between the pharmacokinetic measures generated by clinical doses of LT4 and the results of steady state TSH determinations.3 In standardized pharmacokinetic studies, ingested LT4 is followed into the circulation, and levels of T4 in the blood are measured for 48 h.3 Bioavailability is calculated from pharmacokinetic parameters, including maximum T4 concentration (Cmax) and the area under the T4 curve observed over the 48 h period (AUC0–48). Assessments are made for a name brand (reference) drug and the candidate generic formulation. Healthy volunteers ingest 600 g doses of each product in a randomized cross-over manner, with 35-day wash out periods in between.4 Products are considered bioequivalent if the observed differences between the Cmax values and the AUC0–48 values are insignificant. Statistically, the 90% CI of these differences must fall within 80–125% of the reference drug's value. This method should accommodate a 5% difference in each limit for a 10% total error;6 it is considered unlikely that generic products with bioavailability that differs by more than 10% could meet the CI requirements, and virtually impossible if the differences approach 20%.6 When initially introduced, correction for the endogenous T4 pool was not required.7, 8 Subsequently, the sensitivity of the original method to detect clinically relevant differences in absorbed LT4 was assessed, found to be flawed,8 and a baseline correction method to account for endogenous T4 was adopted in 2003.7 As a result, confidence has been established that differences in bioavailability of 
25% would be evident.8 Clinicians have questioned, however, whether differences of <25% are reliably detected.
To allay these concerns, the FDA stated that differences in tablet content or bioavailability among therapeutically equivalent products should never exceed 9%, and the actual differences were more likely to average 
3.5%, as reported for previously approved generic products.3 Clinicians agree that differences approaching 9% are clinically significant. Additionally, in the US, most clinicians recognize that differences of 9% exceed the increments of many currently available FDA-approved LT4 doses. A difference of this degree would essentially render a prescription for 137 g daily the equivalent of dispensing either 125 g or 150 g.
There are now four generic formulations and three reference preparations for LT4 approved. Although the 90% CIs fall within the accepted limits, the arithmetic mean differences of the AUC0–48 reveal clinically significant variations. For example, the Sandoz generic LT4 product is 12.5% more bioavailable than Synthroid® (Abbott Laboratories, Abbott Park, IL), but is 2.3% less bioavailable than Levoxyl® (King Pharmaceuticals, Bristol, TN).9 Synthroid® is 9% less bioavailable than the generic product from Mylan Laboratories, and 3% less bioavailable than LT4 Lannett (the generic version of Unithroid®, Jerome Stevens Pharmaceuticals, Bohemia, NY).9 Data on the relative bioavailability of the generic LT4 product from Genpharm are not yet available.
These results have led the Endocrine Society, the American Association of Clinical Endocrinologists, and the American Thyroid Association into discussions with the FDA. The clinicians expressed concerns about the potential consequences of underdosage or overdosage of LT4, should products that differ as much as those outlined above be substituted for one another, and discussed alternative approaches to bioequivalence assessment.3 The suggested TSH standard was rejected by the FDA in May 2005, but the societies hope for further discussion of this concept.
In conclusion, the clinical community has welcomed the assurance of high quality LT4 products through the FDA's new-drug application process.1 These regulations require that the unique formulation of each product remains unaltered, to assure consistent bioavailability. Each step of this standardization process has moved clinicians closer to the goal of enhanced patient care. The endocrine societies listed above have raised concerns that the sensitivity of the pharmacokinetic assessment methods for LT4 bioequivalence could fall short of assuring that true interchangeability of products has been achieved. In the interim, the three societies have issued a joint statement that recommends maintaining patients on a consistent source of LT4 by designating the brand name of the product given.10 Once generic substitution has occurred, refills with any of four formulations (three of which are rated as noninterchangeable with one another) are likely to occur. The societies suggest that patients should be educated to verify the LT4 brand that they are prescribed. Patients are asked to note the tablet shape, color, and inscriptions to avoid unannounced substitution. The societies urge that additional serum TSH levels be re-evaluated after a 6–8 week equilibration period after any change in LT4 source, to assess the need for retitration to maintain goal TSH values.10
References
- Hennessey JV (2003) Levothyroxine a new drug? Since when? How could that be? Thyroid 13: 279–282 | Article | PubMed |
- US Department of Health and Human Services, FDA, and Center for Drug Evaluation and Research (2002) Guidance for industry: bioavailability and bioequivalence studies for orally administered drug products—general considerations [http://www.fda.gov/cder/guidance/4964dft.pdf] (accessed 19 July 2006)
- FDA (2005) Itinery for the joint public meeting on equivalence of levothyroxine sodium products [http://www.fda.gov/ohrms/dockets/dockets/05n0137/05n-0137-lst0001.pdf] (accessed 19 July 2006)
- FDA (2005) Code of federal regulations, title 21, vol 5, subchapter D, part 320 [http://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfcfr/CFRSearch.cfm?CFRPart=320&show] (accessed 19 July 2006)
- Carr D et al. (1988) Fine adjustment of thyroxine replacement dosage: comparison of the thyrotropin releasing hormone test using a sensitive thyrotropin assay with measurement of free thyroid hormones and clinical assessment. Clin Endocrinol 28: 325–333
- Henderson JD and Esham RH (2001) Generic drug substitution: issues for problematic drugs. South Med J 94: 16–21 | PubMed | ChemPort |
- FDA (2003) Advisory committee for pharmaceutical science 12–13 March 2003 [http://www.fda.gov/ohrms/dockets/ac/03/briefing/3926B1_01_H-FDA-Additional%20Information.doc] (accessed 19 July 2006)
- Blakesley VA et al. (2004) Are bioequivalalence studies of levothyroxine sodium formulations in euthyroid volunteers reliable? Thyroid 14: 191–200 | Article | PubMed |
- FDA (2006) Drugs@FDA [http://www.accessdata.fda.gov/scripts/cder/drugsatfda/index.cfm] (accessed 19 July 2006)
- American Association of Clinical Endocrinologists, the Endocrine Society and the American Thyroid Association (2005) Joint position statement on the use and interchangeability of thyroxine products [http://www.thyroid.org/professionals/advocacy/04_12_08_thyroxine.html] (accessed 19 July 2006)
Competing interests
Research support from Novartis and Abbott Laboratories; Speaker honoraria from Abbott Laboratories and Aventis.
Contact the journal about this article
Subject areas under which this article appears: Thyroid gland
