The Levothyroxine Absorption Test: A Four-Year Experience (2015

Abstract

Background:

Levothyroxine (LT4) is the mainstay of therapy for hypothyroidism. Yet, despite physician efforts at dose titration, some patients remain hypothyroid on LT4 doses in excess of weight-based calculations, a condition known as refractory hypothyroidism. The LT4 absorption test (LT4AT) has been proposed to have utility in these patients by enabling distinction of LT4 malabsorption from pseudomalabsorption, a condition of intentional nonadherence. Given its rare use in clinical practice, we reviewed our institution's experience with the LT4AT to assess its impact on management of refractory hypothyroidism.

Methods:

We reviewed the charts of 16 patients diagnosed with refractory hypothyroidism and who had completed the LT4AT between January 2015 to January 2019. The primary aim was to determine the utility of this test in distinguishing LT4 malabsorption from pseudomalabsorption. Secondary aims were to determine whether the results of this test impacted physicians' management decisions, as well as to report on clinical outcomes at follow-up. Our LT4AT is a six-hour test wherein patients receive a weight-based dose of LT4 followed by serial measurements of total thyroxine (TT4) and thyrotropin (TSH). Percentage absorption is calculated using the following formula, with normal absorption being ≥60%:

Results:

Percentage absorption was calculated in 13 of 16 patients due to lack of TT4 data for 3 patients. Absorption was impaired in one patient (% absorbed = 0), who had known causes of malabsorption. The remaining 12 patients had normal absorption by hour 4 of the test (% absorption 60–158) in conjunction with upward TT4 trends. Clinical follow-up ranged from 1 to 32 months (median 6.5 months), with 11 patients having follow-up data. Six of these had normal or suppressed TSH values at most recent follow-up, and four had improved but persistent TSH elevations. The one said patient with malabsorption improved with intravenous LT4.

Conclusions:

The LT4AT can provide valuable information for distinguishing malabsorption from pseudomalabsorption. Our findings support the combined use of calculated percentage absorptions with TT4 trends for at least a four-hour time frame when making determinations regarding absorption.

Introduction

Levothyroxine (LT4) is one of the most commonly prescribed medications in the United States, with >13 million patients utilizing it for management of hypothyroidism (1). Absorption occurs primarily in the small intestine within the first three hours of ingestion. Normally, only 60–80% of the administered dose of the medication is absorbed (2). Initial LT4 dosing is traditionally performed using a weight-based calculation, with most patients requiring between 1.6 and 1.8 mcg/kg to achieve thyrotropin (TSH) values in the reference range (2 –4). While most patients do well on this therapy, there is a small subset of individuals who have persistent hypothyroidism despite relatively large LT4 doses, a phenomenon known as refractory primary hypothyroidism. This is defined as a TSH level above the upper limit of the reference range, despite a daily LT4 dose of ≥1.9 mcg/kg (5,6). In these cases, a stepwise approach for evaluating underlying causes should be implemented.

A diagnostic challenge occurs when patients exhibit refractory hypothyroidism despite reported adherence to large doses of LT4 in the absence of known conditions or medications that impair LT4 absorption. In this scenario, alternative testing can be employed to evaluate for LT4 malabsorption. The LT4 absorption test (LT4AT) is a noninvasive and safe method for potentially distinguishing LT4 malabsorption from intentional nonadherence, or pseudomalabsorption (7,8). It is estimated that this test is utilized in <1% of patients evaluated for hypothyroidism at Mayo Clinic.

Since this test is rarely utilized and follows a nonstandardized protocol across institutions, we reviewed our institution's experience with use of the LT4AT in an effort to better define its role in the management of refractory hypothyroidism.

Materials and Methods

Using the Advanced Cohort Explorer, a clinical data repository at Mayo Clinic, we conducted a search of all patients from January 2015 to January 2019 who had undergone LT4 absorption testing through the Endocrine Testing Center (ETC) at Mayo Clinic-Rochester, in the Department of Endocrinology. Patients were identified based on chart searches using the term “levothyroxine absorption test.” Inclusion criteria were age ≥18 years at the time of testing; evidence of primary hypothyroidism; and completion of the LT4AT based on our institution-specific protocol, which allows for calculation of percentage LT4 absorption. Patients who had undergone the LT4AT outside of our ETC were excluded. This study was approved by the Mayo Clinic Institutional Review Board.

Individual chart reviews were performed to document pertinent workups for malabsorption. These included referrals to gastroenterology, serum and endoscopic evaluations for celiac disease, vitamin deficiencies, and medications known to impair LT4 absorption. Underlying comorbidities such as decompensated heart failure, nephrotic syndrome, cirrhosis, pancreatic insufficiency, gastroparesis, Helicobacter pylori infection, gastritis, pernicious anemia, short bowel syndrome, or gastric bypass surgery were also documented. Timing of daily LT4 administration and adherence to therapy were recorded when available.

Before the test, patients were instructed to fast overnight and to hold their dose of LT4 the day of the test. They were also advised not to take any medications known to affect LT4 absorption before or for four hours after the test. The LT4 dosing and protocol used at our institution were derived from previous studies and are outlined hereunder (9 –11):

Standard LT4 dose:

Age 18–65 years, body mass index (BMI) <40 kg/m²: 1000 mcg

Age 18–65 years, BMI ≥40 kg/m²: 1500 mcg

Age 65 years and older: 600 mcg

Nurse-driven protocol for LT4AT

After the patient checks into the outpatient ETC, a medication reconciliation is performed. A review of the patient's allergies is performed to specifically assess for an allergy to LT4 preparations. A peripheral intravenous catheter is then inserted, vitals are recorded, and a baseline blood sample is collected for TSH and total thyroxine (TT4) measurements. The appropriate LT4 dose, as determined by the mentioned parameters, is then administered. Patients are directly observed for the entirety of the test.

Blood specimens are collected for six hours. A TT4 measurement (reference range 4.5–11.7 mcg/dL) is collected at baseline, 1, 2, 3, 4, and 6 hours. TSH (reference range 0.3–4.2 mIU/L) is measured at baseline and at six hours. After the last sample is collected, vital signs are obtained, symptoms are reassessed, and the patient is discharged if deemed clinically stable. For each patient, we calculate the percentage LT4 absorption using the following formula (9,12): $\begin{array}{l} % Absorbed = [[Increment TT4 (mcg/dL \times 10/ \\ total administered LT4 (mcg)]] \times Vd (L) \times 100 . \end{array}$

Increment TT4: peak [TT4] − baseline [TT4].

Vd (volume of distribution) = 0.442 × BMI.

Results

Between January 2015 and January 2019, we identified 16 patients who had undergone the LT4AT for refractory primary hypothyroidism. Fifteen patients (94%) were female with ages ranging from 19 to 60 years. Average weight and BMI were 91.1 kg (range 45.3–138.7 kg, median 90.2 kg) and 32.5 kg/m² (range 16.3–49.5 kg/m², median 31.4 kg/m²), respectively. Causes for hypothyroidism included postsurgical (n = 6, 37.5%), Hashimoto's thyroiditis (n = 7, 43.8%), and unspecified primary hypothyroidism (n = 3, 18.8%). Fourteen patients (87.5%) underwent evaluation for causes of malabsorption, including testing for serum tissue transglutaminase IgA antibodies (n = 14), endoscopy with duodenal biopsies (n = 5), and stool studies (n = 1). One patient had a urine protein analysis to assess for nephrotic syndrome. Pertinent comorbidities included gastroparesis (n = 1, 6.7%), autoimmune gastritis based on an endoscopic biopsy (n = 1, 6.7%), cardiomyopathy (n = 3, 20%), and prior gastric bypass surgery (n = 1, 6.25%). One patient (I) had documented “gastric surgery,” but further details regarding the surgical procedure performed were unknown, therefore, no determinations regarding absorptive capacity could be deduced from the clinical history. Three patients (20%) were on a proton pump inhibitor, two (13.3%) were on ferrous sulfate, and one (6.7%) was on calcium carbonate. One patient was taking both calcium carbonate and ferrous sulfate at the same time as LT4. Patients who were taking interfering medications were advised on proper administration and timing before LT4AT, yet remained poorly controlled leading up to the test. Table 1 displays the baseline characteristics of the cohort.

Table 1.

Baseline Characteristics

Patient	Age (years)	Weight (kg)	BMI (kg/m²)	Free T4 (ng/dL)^a	Baseline TSH (mIU/L)	Cause of hypothyroidism	Malabsorption evaluation (Y/N)	Comorbidities	Interfering medications
A	44	68.1	22.1	0.6	120.1	Postsurgical	Y	None	None
B	44	72.8	26.7	0.5	261	Hashimoto's	Y	None	None
C	21	104.5	33.2	0.8	102.4	Hashimoto's	Y	None	Ferrous sulfate, Topamax
D	29	72	29.6	0.8	138.3	Postsurgical	Y	None	None
E	31	138.7	49.5	—	9	Hashimoto's	Y	None	None
F	39	80.3	27.9	<0.3	85.8	Hashimoto's	Y	Type 1 DM	None
F	39	80.3	27.9	<0.3	85.8	Hashimoto's	Y	Gastroparesis atrophic gastritis dilated cardiomyopathy	None
G	23	110.1	39.6	1.3	7.7	Unspecified	Y	None	None
H	37	135.5	45.2	0.5	17.6	Postsurgical	Y	Cardiomyopathy	Omeprazole
I	60	101.9	41.1	<0.3	141.9	Hashimoto's	N	Gastric surgery, cardiomyopathy	Omeprazole
J	27	58.8	23.0	<0.3	194.9	Postsurgical	Y	None	None
K	19	81.5	30.3	0.8	223.4	Postsurgical	Y	None	None
L	50	118.6	40.8	<0.3	215.1	Unspecified	Y	None	None
M	56	98.8	34.7	<0.3	93	Unspecified	N	Gastric bypass	Calcium carbonate
N	27	45.3	16.3	<0.3	401.9	Postsurgical	Y	None	None
O	34	70.3	26.8	0.4	113.6	Hashimoto's	Y	None	Prilosec
P^b	43	99.8	32.5	<0.3	281.1	Hashimoto's	Y	None	Pantoprazole, magnesium oxide, rosuvastatin
Average (range)	36.5 (19–60)	91.1 (45.3–138.7)	32.4 (16.34–49.49)	(<0.3–1.3)	162.3 (7.7–401.9)

Normal reference range 0.9–1.7 ng/dL.

Patient P was the only male in the cohort.

BMI, body mass index; DM, diabetes mellitus; T4, thyroxine; TSH, thyrotropin.

LT4AT results

Percentage LT4 absorption was calculated for 13 of 16 patients in the cohort due to lack of TT4 measurements for the other 3 patients (free thyroxine [T4] drawn for unclear reasons). Normal absorption was considered to be ≥60%. Several patients had calculated percentage LT4 absorptions >100%, suggesting a tendency toward overestimation using the described formula. There were no adverse events reported during or after the tests.

All patients included in the analysis demonstrated ≥60% LT4 absorption with the exception of Patient F (Table 2). TT4 trends among all but Patient F demonstrated a rise toward peak levels (Fig. 1a). Nine patients reached peak TT4 levels between three and four hours. One of these patients achieved ≥60% absorption by hour 1 with a TT4 increment of 5 mcg/dL; 5 patients did so by hour 2 with TT4 increments ranging from 5.5 to 7.6 mcg/dL; 2 patients did so by hour 3 with TT4 increments of 5.9 and 11 mcg/dL; and 1 patient did so by hour 4 with a TT4 increment of 5 mcg/dL. Patients B, C, and D continued to show a rise in TT4 by hour 6; however, they all achieved ≥60% absorption before that time. Patient B did so by hour 4 with a corresponding TT4 increment of 6.9 mcg/dL. Patients C and D did so by hour 2 with TT4 increments of 5.2 and 6.1 mcg/dL, respectively (Table 3).

FIG. 1.

(a) Total thyroxine trends during LT4AT. (b) TSH trends during LT4AT. LT4AT, levothyroxine absorption test; TSH, thyrotropin.

Table 2.

Levothyroxine Absorption Calculations

Patient	Absorption (%)
A	67
B	83
C	104
D	158
E	94
F	0
G	105
H	N/A
I	127
J	89
K	67
L	109
M	N/A
N	79
O	N/A
P	148

N/A, not available.

Table 3.

Total Thyroxine Increment and Corresponding Percentage Absorption

ID	One hour		Two hours		Three hours		Four hours		Six hours
ID	TT4 increment	Absorption (%)	TT4 increment	Absorption (%)	TT4 increment	Absorption (%)	TT4 increment	Absorption (%)	TT4 increment	Absorption (%)
A	2.9	28	6.5	63	NA	NA	6.9	67	5.8	57
B	0.2	2	1.1	13	3.7	44	6.4	76	7	83
C	0.8	12	5.2	76	5.6	82	6	88	7.1	104
D	1.3	17	6.1	80	10.4	136	10.8	141	12.1	158
E	0.8	11	1.7	23	5.9	81	6.9	94	5.1	70
F	0	0	0	0	0	0	0	0	0	0
G	0.3	5	5.5	96	4.9	86	6	105	4	70
I	5	91	6.9	125	7	127	6.5	118	5.8	105
J	1.2	12	6.8	69	8.8	89	8.8	89	7.5	76
K	−0.1	−1	0.2	3	2	27	5	67	4.6	62
L	3.2	38	7.1	85	9.1	109	8.9	107	6.7	81
N	2.5	18	8	58	11	79	10.3	74	9.6	69
P	3.5	50	7.6	109	10.3	148	8.6	124	7.9	113

The increment TT4 corresponding to ≥60% absorption varied between individuals. While all individuals with the exception of Patient F achieved an increment of at least 5 mcg/dL by hour 4, our small sample size limited our ability to propose a minimum TT4 increment suggestive of adequate absorption. The pattern of TT4 rise toward a peak seemed to better correlate with normal absorption. Normal absorption could be detected by hour 4 of the LT4AT. The only true nonabsorber in the cohort exhibited a flat slope in TT4 during the test.

TSH levels declined from baseline to hour 6 in all but Patients D, L, and F (Table 4 and Fig. 1b). Patient D had a TSH of 212 mIU/L one day before the LT4AT, and a baseline TSH on the day of the test of 138.3 mIU/L. By six hours, the TSH was 149 mIU/L. However, as shown in Figure 1a, TT4 trended upward, consistent with adequate absorption during the test. In addition, calculated absorption for this patient was >60%. After direct observation of oral LT4 administration for one week, this patient's TSH decreased to 15.4 mIU/L.

Table 4.

Thyrotropin Trends During Levothyroxine Absorption Test

Patient ID	Baseline TSH (mIU/L)	Six-hour TSH (mIU/L)
A	120.1	79.1
B	261	197
C	102.4	55.6
D	138.3	149
E	9	7.2
F	85.8	97
G	7.7	3.8
H	17.6	14
I	141.9	108.5
J	194.9	172.1
K	223.4	113.4
L	215.1	219.3
M	93	74.9
N	401.9	331
O	113.6	42.2
P	281.1	231.3

Patient L had a slight increase in TSH from 215.1 to 219.3 mIU/L during the test, but her TT4 trend was also upward and percentage absorption >60%, indicating normal absorption. Patient F had a slight rise in TSH from baseline (85.8 mIU/L) to hour 6 (97 mIU/L). Her TT4 slope remained flat, and percentage absorption was 0%, suggesting no absorption. Thus, the TSH trends seemed less reliable in determining absorptive capacity when compared with either percentage absorption or TT4 trends.

Clinical follow-up

Clinical follow-up periods ranged from 1 to 32 months (median 6.5 months). Only 11 patients had laboratory tests and clinical follow-up documented after completion of the LT4AT. Ten of these patients demonstrated normal LT4 absorption. Six patients with normal absorption (B, L, O, P, M, and D) achieved TSH suppression or normalization during follow-up (Table 5). Patients B, L, O, and P achieved TSH normalization (range 0.3–2.4 mIU/L) within one year after the LT4AT. Patients D and M achieved TSH suppression by the time of follow-up (TSH 0.2 mIU/L and <0.1 mIU/L, respectively). All but one of these patients remained on LT4 doses above their estimated doses based on weight, suggesting perhaps some continued issues with adherence.

Table 5.

Patients with Thyrotropin ≤ Normal Reference Range After Levothyroxine Absorption Test

Patient	Dose prior LT4AT mcg/day (mcg/kg)	Dose immediately after LT4AT	Lowest dose at which TSH normalized mcg/day (mcg/kg)	Baseline TSH (mIU/L)	TSH at follow-up (mIU/L)
B	125 (1.7)	150 (2.1)	200 (2.7)	261	2.4
L	250 (2.1)	400 (3.4)	200 (1.7)	215.1	0.4
O	225 (3.2)	150 (2.3)	175 (2.7)	113.6	1.3
P^a	400 (4.0)	300 (3.0)	300 (3.0)	281.1	0.3
M	325 (3.3)	275 (2.8)	275 (2.8)	93	<0.1
D	125 (1.7)	200 (2.8)	200 (2.8)	138.3	0.2

Dose reduced at most recent follow-up due to nearly suppressed TSH value on 300 mcg daily.

LT4AT, levothyroxine absorption test.

Four of the 10 patients demonstrating normal absorption on the LT4AT had improvements in TSH values at follow-up, but had not achieved complete normalization (Table 6). Three achieved improvement on LT4 doses lower than those before the LT4AT. This suggested that improvement of hypothyroidism was possible at lower doses, and perhaps normalization would have been observed with longer follow-up and continued dose titrations.

Table 6.

Patients with Thyrotropin > Normal Reference Range After Levothyroxine Absorption Test

Patient	Dose prior LT4AT mcg/day (mcg/kg)	Dose immediately after LT4AT	Dose at latest follow-up mcg/day (mcg/kg)	Baseline TSH (mIU/L)	TSH at follow-up (mIU/L)
A	150 (2.2)	150 (2.2)	150 (2.2)	120.1	4.6
I	500 (4.9)	500 (4.9)	125 (1.3)	141.9	11
J	225 (3.8)	400 (6.8)	175 (2.7)	194.9	21
N	343 (7.6)	150 (3.3)	150 (3.3)	401.9	133

Patient F, who had a calculated absorption of 0% on the LT4AT with no change in TT4, had a history of type 1 diabetes, atrophic autoimmune gastritis, and gastroparesis. She was managed with a combination consisting of i.v. LT4 200 mcg twice weekly and oral LT4 200 mcg daily until approximately three weeks before the LT4AT. At that time, i.v. LT4 was discontinued and she continued on daily oral LT4 therapy. Upon completion of the test, she had local follow-up and was restarted on i.v. LT4. She presented to Mayo Clinic two years later for repeat evaluation, at which time her TSH was 0.3 mIU/L, but on repeat testing two weeks later, it had increased to 65.3 mIU/L. She had been switched to oral LT4, Tirosint^® 600 mcg daily, in the month before this laboratory test. The patient subsequently developed a rise in her TSH to 141.3 mIU/L, and Tirosint was increased to 900 mcg daily. Ten months later, the TSH was <0.01 mIU/L on Tirosint 500 mcg daily, thus prompting further dose reduction.

Immediately after completion of the test, changes in LT4 doses were documented. Sixty three percent of patients had immediate dose increases, 25% had dose reductions, and 12.5% had no change in dose (Table 7). Documentation of physician discussions with patients regarding test results is available in all cases. Few physicians opted to lower the LT4 dose despite normal test results, suggesting that the results themselves had less influence on physician decisions but served to more pointedly guide physician–patient discussions.

Table 7.

Levothyroxine Dose Adjustments Immediately After Levothyroxine Absorption Test

Patient	LT4 formulation	Weight (kg)	Weight-based estimated LT4 daily requirement mcg (mcg)	Average daily LT4 dose patient was taking before LT4AT mcg	LT4 daily dose prescribed after LT4AT mcg	% LT4 dose change
Patient	LT4 formulation	Weight (kg)	Weight-based estimated LT4 daily requirement mcg (mcg)	(mcg/kg)^a	(mcg/kg)	% LT4 dose change
A	Levoxyl^®	68.1	108	150 (2.2)	150 (2.2)	0
B	Synthroid^®	72.8	116	125 (1.7)	150 (2.0)	20
C	i.m. LT4	104.5	167	86 (0.82)	225 (2.1)	161
D	Generic	72	115	125 (1.7)	200 (2.8)	60
E	i.v. LT4	138.7	222	171 (1.2)	200 (1.4)	17
F	i.v. LT4 + Synthroid	80.3	128	371 (4.6)	471 (5.9)	27
G	i.v. LT4	110.1	176	96 (0.87)	250 (2.3)	160
H	i.v. LT4 + Generic PO	135.5	216	439 (3.3)	600 (4.4)	37
I	Generic	101.9	163	500 (4.9)	500 (4.9)	0
J	Synthroid	58.8	94	225 (3.8)	400 (6.8)	78
K	Synthroid	81.5	130	300 (3.7)	350 (4.3)	16
L	Generic	118.6	190	250 (2.1)	400 (3.4)	60
M	Levothroid^®	98.8	158	325 (3.3)	300 (3.0)	−7
N	Synthroid	45.3	72	342 (7.6)	150 (3.3)	−50
O	Synthroid	70.3	112	225 (3.2)	150 (2.3)	−33
P	i.m. LT4 + Tirosint^®	99.8	160	414 (4.1)	300 (3.0)	−28

LT4 doses expressed as oral daily dose equivalents. For patients not taking LT4 daily, or for those taking i.v. formulations, doses were averaged for seven days and calculated to obtain daily oral dose equivalents.

Discussion

Our results suggest that most patients who complete an LT4AT demonstrate normal intestinal absorption. Normal absorption was characterized by calculated absorptions ≥60% and corresponding rises in TT4 toward peak levels. Based on our data, the duration of the LT4AT could be reasonably reduced to four hours while still providing optimal information regarding absorptive capacity.

There is a paucity of data regarding the use of nonisotopic oral LT4AT in clinical practice, with much of the literature being limited to case reports and case series. There is also significant heterogeneity between the various protocols described (3,7,8,13 –23). To date, there has been only one other report published using the LT4AT protocol employed by our institution (24). Our protocol takes into account percentage absorption based on TT4 increment, dose of LT4 administered, and volume of distribution (Vd) based on BMI. Twelve patients in the cohort received 1000 mcg of LT4, and four patients received higher doses according to their BMI. Even among those who received 1000 mcg, there was vast variation in percentage absorptions with a tendency toward overestimation of percentages (>100%) using the parameters in our formula. This is a significant limitation and suggests that gross overestimations may have been made within the cohort. Although the cutoff of ≥60% was determined historically using isotopic absorption tests, there are clear limitations in using this derived formula to make accurate determinations regarding percentage absorption in nonisotopic tests. Thus, we argue that interpretation of the calculated percentage absorption be done so in the context of the TT4 trends observed during the test. Increments in TT4, and similarly in free T4, have been positively correlated with LT4 absorption, supporting their utility as surrogates for determining normal absorption (9,19). We were unable to ascertain specific TT4 increment cutoffs that would reliably correspond to normal absorption given the interindividual variation and small sample size.

Unique to our study is the observation of TSH trends during the test. While six hours is a limited timeframe to detect the full effects of LT4 on thyrotrophic production of TSH, Goichot et al. highlight the potential utility of a six-hour TSH measurement, noting a 39% reduction in TSH within two hours of oral T4 ingestion (22). A postulated mechanism for this, as described by Spencer et al., is the rapid inhibition of release of preformed TSH from the anterior pituitary, which is then followed by the more delayed inhibition of TSH synthesis by thyrotrophs (25). In our study, TSH levels did not reliably decline by six hours in all patients. Given our findings and recognizing the kinetics of TSH release and formation, it is likely that TSH measurements can be excluded from the LT4AT, although it may provide an additional piece of information.

Of the 16 patients included in this study, only one demonstrated abnormal absorption as evidenced by 0% calculated absorption and no changes in TT4. Subsequent medical investigations for this patient revealed why the absorption was indeed impaired: a gastric emptying study confirmed gastroparesis by demonstrating only 5% emptying of a technetium-labeled meal at 1 hour (normal 7–27%), 26% at 2 hours (normal 31–67%), and 48% at 4 hours (normal 81–100%). Small bowel transit by the standard six-hour measurement of accumulation was 6%, suggesting marked delay (normal mean accumulation in small intestine expected to be ≥44% by six-hours). Since adequate absorption of most common oral LT4 formulations requires dissolution under acidic gastric conditions, conditions such as atrophic gastritis can certainly impair dissolution (2,26). Gastroparesis can also adversely affect LT4 absorption by delaying exposure of the drug to the intestinal mucosa and allowing admixture of the LT4 preparation with gastric content (2). Tirosint is a gel-cap formulation of LT4 that is dissolved in glycerin, thus bypassing the need for gastric dissolution (20). Assuming that no other conditions known to impair intestinal absorption or transit are present, this is a reasonable consideration for patients who remain refractory to alternative less expensive formulations. When gastroparesis is present, consideration of a parenteral formulation should be considered.

Other conditions known to affect intestinal absorption of LT4 are celiac disease, lactose intolerance, pancreatic insufficiency, short-bowel syndrome, cirrhosis, and cardiomyopathy (3,7,14). A comprehensive evaluation for contributory comorbidities was undertaken for the vast majority of our patients. Unless clinically indicated, extensive evaluations for patients with refractory hypothyroidism may be avoided with use of the LT4AT given it is noninvasive, simple to perform, and can be completed within a short time frame.

Perhaps equally imperative in the evaluation of altered LT4 absorption is the discussion of medication adherence. In the majority of cases, adherence can be easily assessed, and modifications to the plan can be negotiated between the clinician and patient. However, when adherence is endorsed by the patient despite a negative malabsorption workup, this can lead to frustration for both patients and clinicians. It is in these cases wherein the LT4AT may be of the most utility (14). It provides the clinician with an additional tool of objectivity that can serve to more definitively guide management discussions, and can provide important insight for patients.

Our review does not establish a causal relationship between long-term improvement in hypothyroidism and completion of the LT4AT, but it does suggest that the results of this test can aide in facilitating more pointed and open discussions between clinicians and patients regarding the disease and options for long-term management. Strategies for enhancing compliance can include direct observation of medication administration, either through family or clinical support (16). Use of programmed cell phone reminders or a seven-day pillbox may also serve as reminders to patients. In some cases, administration of a weekly dose under direct observation can also be helpful, although this practice should be exercised with caution in patients with underlying atrial arrhythmias or coronary artery disease (14).

Our study is inherently limited by its retrospective design. Although this review constitutes the largest cohort published to date on this topic, the sample size remains small, reflecting the rare use of this test in clinical practice, and the rarity of clinical presentations necessitating this test. In addition, follow-up data were lacking in a third of our cohort, and follow-up periods were highly variable. Three patients who completed the six-hour test had free T4 measurements rather than TT4, impairing our ability to calculate percentage absorptions for these subjects.

Finally, a potential criticism of our protocol is that the doses of LT4 administered are higher than what would be needed to demonstrate adequate absorption on the basis of TT4 changes. However, the proposed amounts (600 mcg, 1000 mcg, and 1500 mcg) correspond with the average expected total weekly amounts of LT4 needed for most individuals in the cohort according to weight-based calculations. One could consider individualizing the protocol further such that the LT4 dose corresponds precisely with weight-based total weekly calculations.

Another limitation of the test is that it does not provide information regarding the optimal dose needed to achieve euthyroidism, nor can it be credited for altering physician judgments regarding dose adjustments. In our cohort, dose increases were made in 62.5% of patients immediately after LT4AT, a decision that one can reasonably expect of clinicians caring for patients with refractory hypothyroidism. Thus, the LT4AT results may better serve to guide open discussions regarding adherence rather than to influence decision-making regarding dose adjustments.

Conclusions

Using the protocol described in this study, we were able to distinguish patients with true malabsorption from those with pseudomalabsorption. Our findings support the combined use of calculated percentage absorptions with TT4 trends over at least a four-hour time frame when making determinations regarding absorption. Although performance of the LT4AT may not alter ultimate decisions regarding dose adjustments, the clinical utility of this test perhaps lies more in its ability to provide objective information that can guide open discussions between clinicians and patients regarding strategies for medication adherence. As reflected in the clinical follow-up of patients in this cohort, many achieved normalization or improvement of TSH over time, but most did so by still requiring relatively large doses of LT4. This perhaps reflects continued issues of nonadherence in the setting of chronic disease and emphasizes the importance of close follow-up for patients with refractory hypothyroidism.

Footnotes

Acknowledgments

We would like to acknowledge and thank the Mayo Clinic ETC staff: Russell Ward, Elisha Vicente, Corey Schweitzer, David Schick, Heather Johnson, and Annamarie Erlbacher.

Author Disclosure Statement

No competing financial interests exist.

Funding Information

No funding was received for this article.

References

US Food and Drug Administration. Levothyroxine Sodium Product Information. Available at https://www.fda.gov/Drugs/DrugSafety/PostmarketDrugSafetyInformationforPatientsandProviders/ucm161257.htm (accessed April 25, 2019 ).

Skelin

, Lucijanic

, Amidzic Klaric

, Resic

, Bakula

, Liberati-Cizmek

, Gharib

, Rahelic

. 2017. Factors affecting gastrointestinal absorption of levothyroxine: a review. Clin Ther, 39:378–403.

Rdzak

, Whitman

, Inzucchi

. 2018. Levothyroxine pseudo-malabsorption: testing and treatment in the outpatient setting. Ther Adv Endocrinol Metab, 9:217–222.

Morris

. 2009. How do you approach the problem of TSH elevation in a patient on high-dose thyroid hormone replacement?. Clin Endocrinol (Oxf), 70:671–673.

Centanni

, Benvenga

, Sachmechi

. 2017. Diagnosis and management of treatment-refractory hypothyroidism: an expert consensus report. J Endocrinol Invest, 40:1289–1301.

Ramadhan

, Tamilia

. 2012. Treatment-refractory hypothyroidism. CMAJ, 184:205–209.

Ain

, Refetoff

, Fein

, Weintraub

. 1991. Pseudomalabsorption of levothyroxine. JAMA, 266:2118–2120.

Livadariu

, Valdes-Socin

, Burlacu

, Vulpoi

, Daly

, Beckers

. 2007. Pseudomalabsorption of thyroid hormones: case report and review of the literature. Ann Endocrinol (Paris), 68:460–463.

Greenstadt

, Nicoloff

, Spencer

, Kaptein

. 1981. Development of a non-isotopic method for determination of oral T4 absorption. Clin Res, 29:A83.

10.

Sherman

, Malecha

. 1995. Absorption and malabsorption of levothyroxine sodium. Am J Ther, 2:814–818.

11.

Surks

, Schadlow

, Stock

, Oppenheimer

. 1973. Determination of iodothyronine absorption and conversion of L-thyroxine (T 4) to L-triiodothyronine (T 3) using turnover rate techniques. J Clin Invest, 52:805–811.

12.

Sherman

, Tielens

, Ladenson

. 1994. Sucralfate causes malabsorption of L-thyroxine. Am J Med, 96:531–535.

13.

Srinivas

, Oyibo

. 2010. Levothyroxine pseudomalabsorption and thyroxine absorption testing with use of high-dose levothyroxine: case report and discussion. Endocr Pract, 16:1012–1015.

14.

Walker

, Shillo

, Ibbotson

, Vincent

, Karavitaki

, Weetman

, Wass

, Allahabadia

. 2013. A thyroxine absorption test followed by weekly thyroxine administration: a method to assess non-adherence to treatment. Eur J Endocrinol, 168:913–917.

15.

Lips

, van Reisen

, Voigt

, Venekamp

. 2004. Diagnosis and treatment of levothyroxine pseudomalabsorption. Neth J Med, 62:114–118.

16.

Eledrisi

, Szymajda

, Alshanti

, Urban

. 2001. Noncompliance with medical treatment: pseudomalabsorption of levothyroxine. South Med J, 94:833–836.

17.

Balla

, Jhingan

, Rubin

. 2015. Rapid Levothyroxine Absorption Testing: A Case Series of Nonadherent Patients. Int J Endocrinol Metab, 13:e31051.

18.

Ogawa

, Otsuka

, Mimura

, Ueno

, Hashimoto

, Kishida

, Ogura

, Makino

. 2000. Pseudomalabsorption of levothyroxine: a case report. Endocr J, 47:45–50.

19.

Sun

, Pantalone

, Faiman

, Gupta

, Olansky

, Hatipoglu

. 2014. The clinical utility of free thyroxine in oral levothyroxine absorption testing. Endocr Pract, 20:925–929.

20.

Vita

, Benvenga

. 2014. Tablet levothyroxine (L-T4) malabsorption induced by proton pump inhibitor; a problem that was solved by switching to L-T4 in soft gel capsule. Endocr Pract, 20:e38–41.

21.

Van Wilder

, Bravenboer

, Herremans

, Vanderbruggen

, Velkeniers

. 2017. Pseudomalabsorption of levothyroxine: a challenge for the endocrinologist in the treatment of hypothyroidism. Eur Thyroid J, 6:52–56.

22.

Goichot

, Vinzio

, Luca

, Sirlin

, Sapin

, Schlienger

. 2007. In vivo evidence for a direct ultra-fast negative feedback of thyroxine on TSH secretion in humans: a case of L-thyroxine pseudomalabsorption. Clin Endocrinol (Oxf), 67:952–953.

23.

Benvenga

, Bartolone

, Squadrito

, Lo Giudice

, Trimarchi

. 1995. Delayed intestinal absorption of levothyroxine. Thyroid, 5:249–253.

24.

Simsir

, Soyaltin

, Ozgen

. 2019. Levothyroxine absorption test results in patients with TSH elevation resistant to treatment. Endocrine, 64:118–121.

25.

Spencer

, LoPresti

, Nicoloff

, Dlott

, Schwarzbein

. 1995. Multiphasic thyrotropin responses to thyroid hormone administration in man. J Clin Endocrinol Metab, 80:854–859.

26.