Evaluation of Thyroid Hormone Replacement Dosing in Overweight and Obese Patients After a Thyroidectomy

Introduction

Surgical resection of the thyroid gland has inherent postoperative consequences for thyroid hormone replacement (THR). Selection or choice of postoperative dosing for THR has been typically extrapolated from studies assessing hormone replacement in primary hypothyroidism, while other studies have suggested that the dosing for post-thyroidectomy patients may not be the same (1 –7). In primary hypothyroidism that is most commonly secondary to Hashimoto's thyroiditis, there often is residual gland function that can complement a given exogenous dose of thyroid hormone and serve as a buffer to insufficient dosage. However, this “buffer” is lacking after total thyroidectomy, thereby rendering achievement of the appropriate dose in surgical patients more challenging. Overdosing can lead to metabolic abnormalities, cardiac dysfunction, bone loss, and fractures (8 –12). Underdosing can lead to weight gain, dyslipidemia, and decreased cognitive function (2,13,14).

A common formula to determine the initial dose of levothyroxine (LT4) for THR is to start the patient on 1.6 mcg/kg body weight in the postoperative period without correction for the patient's body mass index (BMI). This dose is subsequently titrated based on thyrotropin (TSH) values and symptomatology. Replacement dosage recommendations are often ambiguous in regard to whether the dose is based on measured body weight or estimated lean body weight (LBW). This distinction relates to the effect of obesity on dosage requirements. Over the past two decades, the proportion of the U.S. population that is classified as obese has been increasing (15). It has been shown in multiple studies that obese patients can develop thyrotoxicosis if dosing is based on the same calculations that were developed for the nonobese population (16 –23), yet no corrected dosing regimen has been uniformly recommended.

Our hypothesis was that patients with high BMI require a relatively lower dose of LT4 in relation to their body weight in kilograms, and this study aimed to calculate the proper dosing algorithm of LT4 for obese patients.

Methods

Following institutional review board approval, we created a database of patients who had undergone a total thyroidectomy at our tertiary referral center between January 2012 and December 2015. Patients with a surgical pathology of malignancy were excluded, as were pregnant or lactating patients. Patient demographics, preoperative laboratory results, surgical pathology, operative indications, and postoperative laboratory results were collected for analysis.

Patients were considered to be on a stable LT4 dose when two consecutive TSH values were within the goal range, 0.45–4.50 mIU/L while on the same dose of LT4. The time to achieve euthyroidism was defined as the time from their surgery to the first normal TSH value. Continuous variables were analyzed with a two-sample t-test or an ANOVA test, and categorical variables were analyzed with a chi-square test. Regression analysis was performed for calculation of LT4 dosing with a dependent variable to weight. Actual weight (AW) was the preoperative weight of the patient. Ideal body weight (IBW) was determined as the predicted weight of a patient based on the patient's height as derived by the Devine formula (24). Determination of LBW was based on the LBW 2005 formula (25). Adjusted body weight (ABW) was expressed as predicted body weight, which takes into account the IBW and AW to correct for the shortcomings of IBW and is more appropriately used for patients with an actual body weight 30% greater than the predicted IBW. ABW is calculated based as IBW +0.4 (AW − IBW) on the Ramanathan formula (26). Body surface area (BSA) is a common formula used for medication dose titrations (27). Data analysis was performed with SAS Software, Version 9.4 (Cary, NC).

Results

Charts of patients who underwent total thyroidectomy between July 2011 and December 2015 were reviewed. A total of 428 patients were identified, of whom 173 were excluded due to diagnosis of cancer on final pathology. Of the remaining 255 patients, 114 had follow-up data and achieved euthyroidism on LT4. Our cohort consisted of these 114 patients.

The mean age of the cohort was 55 years (range 28–77 years) with 84% females. Hashimoto's thyroiditis or LT4 use was documented in 19 patients before surgery. Surgical indications included symptomatic goiter (52%), suspicious nodule (45%), and Graves' disease (3%). The mean BMI of the group was 31, with 48% of the population having a BMI of 30 or greater. The number of patients with a BMI less than 25, 25–29, 30–34, 35–39, and greater than 40 were 26 (23%), 33 (29%), 23 (20%), 19 (17%), and 13 (11%), respectively. Patients with a higher BMI did require a higher absolute dose (mcg) of LT4 to achieve euthyroidism (p < 0.01), but the dose was significantly lower in relation to their weight (mcg/kg) (p < 0.01), as shown in Table 1. The LT4 doses required to achieve euthyroidism based on the previously mentioned BMI categories were 1.76, 1.47 , 1.42 , 1.27 , and 1.28 mcg/kg. Of the patients who did have a follow-up within the first three months, 25% of patients were hypothyroid, 40% were euthyroid, and 35% were thyrotoxic. The euthyroid state with the initial LT4 dosing for each BMI category is presented in Figure 1.

OPEN IN VIEWER

FIG. 1.

Thyroid function results with initial dose of levothyroxine.

A mean of 50 weeks elapsed after surgery to achieve euthyroidism, with no significant difference between the BMI categories (p = 0.58). Additionally, all the BMI groups showed weight loss until euthyroidism was achieved with no significant difference between the groups (p = 0.61). These results are summarized in Table 2.

Linear regression models were created using the patient's preoperative BMI, actual body weight, IBW, ABW, LBW, and BSA and are shown in Table 3.

Discussion

The number of thyroidectomies that are performed in the United States has been increasing over the past decade. Approximately 100,000 patients undergo a thyroidectomy on an annual basis in the United States (28). Currently, THR dosing is based on studies evaluating patients with autoimmune hypothyroidism, despite the fact that recent literature has shown that the dosing requirements for THR may not be equivalent for patients hypothyroid due to a thyroidectomy when compared with those who are hypothyroid due to an autoimmune disorder (4). Moreover, one area that has lacked specific recommendation is the need for dose adjustments in the obese. It has been well established that the rate of obesity has been increasing in the United States. In an estimate by the National Institutes of Health, over 23% of the U.S. population was noted to be overweight, over 37% was noted to be obese, and 7.7% were extremely obese, all reflecting significant increments from previous decades (15). Our concern was that past standard textbook or guideline recommendations for thyroxine replacement dosage may not have been reconciled with current population trends to obesity. Thus, this latter trend to increasing body weight since the development of original dosing regimens for LT4 may indicate that a better model for dose initiation is required for more accurate dosing to minimize the potential adverse effects of either over- or underdosage.

Despite these concerns, few studies have evaluated the impact of obesity on thyroid function and hormone replacement in an obese post-thyroidectomy group. A study of thyroid dosing for the obese by Singh showed that in patients with autoimmune hypothyroidism, a higher BMI required a lower dose per kilogram body weight and that IBW was a better predictor for appropriate dosage than was actual body weight (16). These results were reaffirmed in reports by Devdhar et al. (20), Baehr et al. (18), Glymph and Gosmanov (21), and Santini et al. (17).

The issue of optimal dosage replacement based on body weight was further assessed by Jin et al. in a retrospective analysis; LT4 dosing based on BMI, IBW, and BSA in 280 patients who had undergone total thyroidectomy or lobectomy showed that the best predictor of dosing was equivalent to 1.5 μg/kg (19). Unfortunately, they did not evaluate dosing at various BMI categories but rather analyzed the entire cohort in a regression analysis. Similarly, Ojomo et al. showed in another retrospective study that patients with a higher BMI require a lower LT4 dose per kilogram body weight to achieve euthyroidism (23). Their regression analysis was employed to create a formula for LT4 dosing based on BMI rather than actual kg body weight.

Our findings are consistent with these latter data indicating that obese patients require a higher absolute dose of levothyroxine but a lower relative dose of LT4 compared with the patient's body weight. Further breakdown analysis of our data into defined BMI categories to assess the dosing at various levels of obesity demonstrated that there was statistically significant different LT4 dosing for each BMI category. In evaluating which predictor could most accurately be used for the dose calculation, we noted that the regression lines for the patient's AW, IBW, ABW, or BSA all showed high correlations, indicating that any of those can be used to calculate the LT4 dose with near equal accuracy. The actual body weight can be used to calculate a postoperative LT4 dose, if the patient's BMI is taken into account and the proper coefficient is used for the dosing calculation. Yet for ABW, LBW, and BSA, there seems to be less variability in the calculated coefficients in the regression analysis between the different BMI categories, which implies that a single coefficient can be sufficient to dose the LT4 for any weight, and consequently, a breakdown into BMI categories may not be necessary. We believe that the most practical method for calculating the patient's initial LT4 dose is the simplest one. The two simplest methods would be either to calculate the patient's ABW and use a single dosing coefficient to calculate the LT4 dose regardless of the patients weight or use the patient's actual body weight and BMI (both commonly found information in most medical records) and calculate the LT4 dose by using the BMI appropriate dosing coefficient, as shown in Table 3.

Our study does have limitations, the major one being its retrospective nature. Additionally, our data were collected from electronic medical records, and compliance with dosing or proper administration timing could not be verified.

Furthermore, it is important to note that postoperative patients can continue to be either thyrotoxic or hypothyroid for as long as 50 weeks from the date of the operation while their LT4 dose is being titrated. Inaccuracies with regard to weight-based dosing of LT4 could be largely responsible for the prolongation of the titration period. Thus, we recommend using either the AW of the patients with adjustment of dosing based on the BMI or the ABW without regard to the patient's BMI.