Levothyroxine Absorption Test for Diagnosis of Levothyroxine Pseudomalabsorption: A Six-Year Experience and Proposal of a Cutoff Point

Levothyroxine (LT4) pseudomalabsorption corresponds to a factitious disorder. It should be suspected in subjects with persistently elevated thyroid stimulating hormone (TSH) despite doses ≥1.9 µg/kg per day of LT4. In this scenario, LT4 absorption test (LT4AT) could be performed, this is a non-invasive method to distinguish levothyroxine malabsorption from pseudomalabsorption however, more clinical studies are needed to determine a standard protocol with appropriate cutoffs. The major aim of this study is to estimate the optimal cut point of the free T4 (fT4) delta to determine LT4 pseudomalabsorption.

Methods:

Single-center, cross-sectional study. Patients included were adults 18 years or older with refractory primary hypothyroidism despite high LT4 doses, in whom a 1000 µg of levothyroxine test was performed to diagnosis LT4 pseudomalabsorption. Exclusion criteria were pregnant women and cases with incomplete clinical records. A normal test was considered a percentage of absorption ≥60%. We use the Mann-Whitney test to compare the basal fT4 with the two and four-hours measurements and the two and four-hours delta. Finally, we calculate an optimal threshold value for the four-hours fT4 delta using the Youden Index estimated by the nonparametric Kernel method.

Results:

Data from 2015 to 2021 of patients who underwent a 1000 µg of levothyroxine test was retrieved (n=76). Most of them were female, n=69 (90.8%), with a median age of 39 years (IQR 16.5). Chronic autoimmune thyroiditis was the most common cause of hypothyroidism, n=48 (63.2%). The percentage of absorption at 2 hours was 98.2% (IQR 59.8) and at 4 hours 117% (IQR 52.6), fT4 concentration increase significantly after the levothyroxine loading dose. The median increase at two hours was 0.8 ng/dL (95%CI 0.75 to 0.95), and after four hours, was 1.0 ng/dL (95%CI 0.92 to 1.07). The difference between the two and four-hours delta was also significant, with a median difference of 0.12 ng/dL (95%CI 0.05 to 0.19). For the fT4 four-hours delta, we found an optimal cutoff point of 0.65 with a 90% sensitivity and specificity, a 0.98 area under the curve (AUC), a positive predictive value of 1, and a negative predictive value of 0.3. In correlation with percentage of absorption ≥60%.

Discussion/Conclusion:

Our data suggest that most patients who complete a LT4AT demonstrate normal intestinal absorption of LT4 (percentage of absorption ≥60%). The single four hours post load LT4 value could be an optimal marker for calculating the percentage of absorption. The increase of fT4 ≥0.65 ng/dL at four hours respect to the baseline value is suggestive of pseudomalabsorption. The adoption of this simplification may facilitate and amplify the use of the LT4AT, and reduce costs by measuring only fT4 at four hours without needed to calculate the percentage of absorption.