Bexarotene-induced hypothyroidism and dyslipidemia; a nation-wide study

Abstract

Central hypothyroidism and dyslipidemia are well-known adverse events (AEs) of bexarotene therapy. Although hypothyroidism is known to cause dyslipidemia, no study has examined the association between hypothyroidism and dyslipidemia in patients undergoing bexarotene therapy. The aim of this study is to examine this association. A retrospective observational study was performed among 294 patients who initiated bexarotene therapy in Japan (nation-wide postmarketing complete surveillance). Jonckheere-Terpstra (one sided) test was performed to evaluate the effect of the bexarotene dose on lipid metabolisms, and regression analyses were performed to evaluate associations of bexarotene dose, free thyroxine (FT4), body mass index (BMI), and lipid metabolisms. Most patients developed hypothyroidism. Two-third of patients showed FT4 values below the lower limit at 1 week. Triglycerides (TG) increased in a bexarotene dose–dependent manner, and grade ≥3 AEs on hypertriglyceridemia was observed in 39% of the patients. Additionally, one-third of grade ≥3 AEs on hypertriglyceridemia occurred within 1 week. The delta_FT4 (difference in FT4 from baseline) negatively correlated with TG increase at 1 week (p = 0.012) but not with low density lipoprotein cholesterol (LDL-C) increase at any week. Bexarotene-induced hypothyroidism is almost inevitable and occurred quickly. Bexarotene-induced hypertriglyceridemia showed positive bexarotene dose dependency and negative delta_FT4 dependency. Prophylactic and appropriate thyroid hormone compensation therapy and starting bexarotene at low doses with subsequent titration while managing dyslipidemia may have a beneficial effect for the successful continuation of bexarotene therapy without severe endocrine and metabolic AEs.

Introduction

Bexarotene has been available for cutaneous T-cell lymphoma (CTCL) since 1999 in the United States, based on the recommendations by the National Comprehensive Cancer Network [1] and the European Organisation for Research and Treatment of Cancer [2]. Central hypothyroidism is a frequent adverse event (AE) in bexarotene-treated patients, and prophylaxis with thyroid hormone compensation therapy is recommended at the time of bexarotene initiation by the UK Consensus Statement [3]. Importantly, we have previously reported that central hypothyroidism occurs very frequently in Japan, where bexarotene has been available for CTCL since January 2016 [4].

Dyslipidemia with elevated levels of triglycerides (TG) and cholesterol is another important AE in bexarotene-treated patients. The UK Consensus Statement [3] recommends starting treatment with fenofibrate or rosuvastatin 1 week before the initiation of bexarotene therapy. The association between thyroid dysfunction and dyslipidemia is well established, and it is a priori assumed that appropriate management of hypothyroidism leads to appropriate management of dyslipidemia. However, no study to date has examined whether this approach is applicable in bexarotene-treated patients because prophylactic therapy for thyroid dysfunction and dyslipidemia would complicate the interpretation of results in such studies. Prophylactic thyroid hormone administration at the start of bexarotene therapy has not been approved in Japan; therefore, data from Japanese patients on bexarotene therapy who are not receiving prophylactic therapy for hypothyroidism and dyslipidemia provide an opportunity to determine the association of bexarotene therapy with hypothyroidism and dyslipidemia. In this first study evaluating the association between thyroid dysfunction and dyslipidemia in bexarotene-treated patients, we retrospectively investigated the current status of endocrine and metabolic AEs associated with bexarotene use in Japan using real-world data.

Materials and methods

Data source

The data from an anonymized postmarketing surveillance of all the patients who initiated bexarotene treatment against CTCL in Japan between June 23, 2016 and June 30, 2018 were provided by Minophagen Pharmaceutical Co., Ltd., under a research contract.

Ethics

This study was conducted in accordance with the World Medical Association Declaration of Helsinki. The study was approved by the Research Ethics Committee (ethical reviews of clinical research, epidemiology research and diagnostics) of the University of Tokyo Hospital [review number: 2020204NI-(3)] and was conducted using an optout form of notification, as this study used previously collected and anonymized data.

Subjects

As shown in Fig. 1, a total of 294 cases were identified in the nation-wide postmarketing complete surveillance of bexarotene therapy in Japan and 291 patients were treated with bexarotene (full analysis set). For the analysis of thyroid function, 8 patients taking levothyroxine at baseline and 77 patients who were initiated levothyroxine without any measurements of thyroid hormones after the initiation of bexarotene were excluded. Additionally, 40 patients without data on baseline thyroid hormone levels, 38 patients whose baseline thyroid-stimulating hormone (TSH) or free thyroxine (FT4) levels were not within the reference range, and 3 patients whose thyroid hormones were not measured within the observation period after the initiation of bexarotene therapy were excluded. As a result, 125 patients were included in the thyroid analysis set.

Fig. 1

Study flowchart

Between June 2016 and June 2018, 294 patients were registered in the nation-wide postmarketing complete surveillance of bexarotene therapy in Japan and 291 patients were treated with bexarotene for CTCL (full analysis set). Of the 291 patients, 125 patients were included in the thyroid analysis set. Among thyroid analysis set, 120 patients developed hypothyroidism and 107 patients received thyroid hormone compensation therapy. In the remaining 5 patients, one patient received bexarotene for only 1 day and one patient received bexarotene for 9 days.

Standardization of thyroid function tests

The levels of TSH and FT4 could not be directly compared because multiple assays were used across multiple institutions over a 2-year period. Therefore, the measured TSH and FT4 levels were converted to standardized TSH and FT4, respectively, as follows: standardized hormone level = (measured hormone level – lower limit of reference range)/(upper limit of reference range – lower limit of reference range).

Statistical analyses

Nonnormally distributed patient background parameters, such as age, weight, and body mass index (BMI), were presented as medians with 25th and 75th percentile values. Laboratory parameters were presented as means ± standard deviations. Jonckheere-Terpstra (one sided) tests were performed to evaluate the effect of the bexarotene dose on TSH, FT4, TG, and low-density lipoprotein cholesterol (LDL-C). Multiple regression analysis was performed to evaluate the effects of the bexarotene dose, FT4, and BMI on TG and LDL-C. Single regression analysis was performed to evaluate the effect of TG increase on LDL-C levels using data of patients without lipid-lowering agents (LLAs).

All analyses were performed using the statistical program R version 4.1.2 (the R Foundation of Statistical Computing; www.r-project.org). For all analyses, a p value of <0.05 was considered to indicate statistical significance and a p value of <0.10 was considered to indicate trend.

Results

Patient background characteristics

As shown in Table 1, the cohort included 176 male and 115 female patients, with a median age of 67 years. The median BMI was 21.9 kg/m², and 64 of the 291 patients (22%) had a BMI of over 25 kg/m². The median starting dose of bexarotene was 279.6 [184.5–298.6] mg/m², which was close to the standard starting dose of 300 mg/m².

Table 1

Characteristics of the patients

parameters	Mean or median
Age [years old]	Median 67 [55–76]
Sex	Male:Female 176:115
Initial dose of bexarotene [mg/m2]	Median 279.6 [184.5–298.6]
BMI [kg/m2]	Median 21.9 [19.6–24.6]
T.Chol [mg/dL]	Mean 190.4 (±37.3)
HDL-C [mg/dL]	Mean 56.1 (±17.6)
TG [mg/dL]	Mean 131.0 (±66.9)
LDL-C [mg/dL]	Mean 110.9 (±28.8)
Standardized TSH	Mean 0.453 (±1.002)
Standardized FT4	Mean 0.373 (±0.290)
Standardized FT3	Mean 0.296 (±0.273)

Endocrine and metabolic AEs

The majority of patients (250 of 291; 86%) were reported to develop thyroid dysfunction as an AE, and 9 patients (3%) had grade 3 thyroid dysfunction. According to the laboratory tests, 244 patients (84%) had hypothyroidism based on FT4 levels below the lower reference limit, 251 patients (86%) were taking levothyroxine, and 30 patients (10%) had levothyroxine without a decrease in FT4 levels below the lower reference limit.

In the overall cohort, 280 of the 291 patients (96%) had grade ≥1 hypertriglyceridemia, including 113 patients (39%) with grade ≥3 hypertriglyceridemia and 24 patients (8%) with grade 4 hypertriglyceridemia. Specifically, in the first week, 36 patients (12%) experienced grade ≥3 adverse events (AEs) related to hypertriglyceridemia, while in the second week, the percentage increased to 66 patients (23%). In the third week, it rose further to 81 patients (28%), and in the fourth week, it reached 92 patients (32%). The total cholesterol level was above the upper reference limit (grade ≥1) in 220 patients (76%), which contained 4 patients with grade ≥3 hypercholesterolemia. Other endocrine AEs included central adrenal insufficiency in four patients, including one patient with grade 4 adrenal crisis.

Effect of bexarotene therapy on thyroid function

In the thyroid analysis set of 125 patients, 120 patients (96%) developed hypothyroidism and 107 patients (86%) were treated with levothyroxine (Fig. 1). Among the 85 patients whose FT4 levels were measured at 1 week, 56 exhibited FT4 levels below the reference range. Of the five patients who did not develop hypothyroidism, one patient received bexarotene for only 1 day and one patient received bexarotene for 9 days. In the remaining three patients without hypothyroidism, the FT4 levels remained within the reference range but were at or near the lower limit.

In the full analysis set, among 181 patients, 163 (90%) had TSH levels below the reference range at 1 week. Furthermore, in the Jonckheere-Terpstra tests, when dividing bexarotene doses into three groups: less than 150 mg/m²/day, 150 mg/m²/day or more but less than 250 mg/m²/day, and 250 mg/m²/day or more, a dose-dependent decrease in the 1 week to baseline TSH ratio and a difference in standardized FT4 from baseline at 1 week (p = 0.022 and p < 0.001, respectively; Fig. 2A and 2B). The decrease in the 1 week to baseline TSH ratio was similar to that described in the previous study [5].

Fig. 2

Bexarotene-induced hypothyroidism and its dose dependency

(A) A trend test of bexarotene dose-dependent 1 week to baseline TSH ratio was performed using the Jonckheere-Terpstra (one-sided) test. (B) A trend test of bexarotene dose-dependent standardized FT4 decrease at 1 week was performed using the Jonckheere-Terpstra (one-sided) test.

The trend in achievement of the recommended FT4 levels for bexarotene-treated patients by the UK Consensus Statement, the upper third of the reference range, in the thyroid analysis set is shown in Supplementary Fig. 1. Even though the proportion of patients receiving thyroid hormone compensation therapy and the dose of levothyroxine increased with the progression of bexarotene therapy (Supplementary Table 1), the recommended FT4 levels were rarely achieved throughout this study period (Supplementary Fig. 1).

Effects of bexarotene dose, bexarotene-induced hypothyroidism, and BMI on dyslipidemia

Before starting bexarotene therapy, 67 of the 291 patients (23%) were taking LLAs. The proportion of patients taking LLAs increased with the progression of bexarotene therapy. Statins were the most commonly used LLAs before bexarotene initiation, whereas fibrates were the most commonly used LLAs after 4 weeks of bexarotene therapy. The detailed information on the dosage of bexarotene, thyroid hormone replacement therapy, and the usage of LLAs are documented in Supplementary Table 1.

Jonckheere-Terpstra tests on the effect of the bexarotene dose revealed significance in the difference in TG from baseline (delta_TG) at 4 and 12 weeks (p = 0.0016 and p < 0.001, respectively; Fig. 3A and 3B). In addition, a multiple regression analysis also showed a correlation between the delta_TG and the bexarotene dose, the delta_FT4, and BMI at 1 week (p = 0.004, p = 0.005, and p = 0.016 respectively; R² = 0.17; Table 2). Jonckheere-Terpstra tests on the effect of the bexarotene dose showed significance in the difference in LDL-C from baseline (delta_LDL) not at 4 weeks but at 12 weeks (p = 0.13 and p < 0.001, respectively; Fig. 3C and 3D). Multiple regression analysis using the bexarotene dose, delta_FT4, and BMI as explanatory variables revealed that the delta_LDL did not correlate with the explanatory variables in the patients whose LLAs were not changed from baseline (Table 2).

Fig. 3

Changes in lipid metabolisms after the initiation of bexarotene therapy

Comparison of lipid metabolisms among three groups divided by the dose of bexarotene was performed using Jonckheere-Terpstra (one sided) tests. (A) The difference in triglyceride levels at 4 weeks of bexarotene therapy from baseline levels (delta_TG) according to the bexarotene dose. (B) The delta_TG at 12 weeks of bexarotene therapy according to the bexarotene dose. (C) The difference in low-density lipoprotein cholesterol levels at 4 weeks of bexarotene therapy from baseline levels (delta_LDL) according to the bexarotene dose. (D) The delta_LDL at 12 weeks of bexarotene therapy according to the bexarotene dose.

Table 2

Multiple regression analysis of factors influencing lipid metabolism

Dependent variable	β coefficient1	Standard error	p value	VIF
Delta_TG2 at 1 week (F value 11.1, p < 0.001, Adjusted R2 0.17)
Dose of bexarotene	0.59	0.20	0.004	1.10
Delta_FT43	–142	46	0.005	1.08
BMI	8.6	3.5	0.016	1.02
Delta_TG at 4 weeks (F value 4.07, p = 0.008, Adjusted R2 0.064)
Dose of bexarotene	0.47	0.21	0.029	1.03
Delta_FT4	–2.3	40	0.96	1.03
BMI	12.5	4.5	0.007	1.01
Delta_TG at 12 weeks (F value 8.37, p < 0.001, Adjusted R2 0.15)
Dose of bexarotene	0.48	0.13	<0.001	1.02
Delta_FT4	–17	24	0.49	1.01
BMI	9.5	2.7	<0.001	1.02
Delta_LDL4 at 1 week (F value 0.25, p = 0.86, Adjusted R2 –0.02)
Dose of bexarotene	0.02	0.05	0.74	1.10
Delta_FT4	6.1	10.7	0.56	1.07
BMI	–0.47	0.82	0.57	1.06
Delta_LDL at 4 weeks (F value 0.79, p = 0.50, Adjusted R2 –0.006)
Dose of bexarotene	–0.08	0.05	0.13	1.07
Delta_FT4	–4.9	13	0.71	1.07
BMI	0.14	1.1	0.90	1.01
Delta_LDL at 12 weeks (F value 1.47, p = 0.23, Adjusted R2 0.02)
Dose of bexarotene	0.07	0.05	0.16	1.003
Delta_FT4	–11	9.5	0.24	1.02
BMI	0.69	0.94	0.46	1.02

¹ β coefficient: coefficient of determination

² Delta_TG: Difference between TG values at baseline and at each week.

³ Delta_FT4: Difference between standardized FT4 values at baseline and at each week.

⁴ Delta_LDL: Difference between LDL values at baseline and at each week.

In general, thyroid hormones correlate more strongly with LDL-C than with TG [6-8]. Next, we suspect that bexarotene-induced hypertriglyceridemia might suppress the rise in LDL-C accompanied by bexarotene therapy and decrease in FT4. We conducted an analysis on the relationship between TG and LDL-C by extracting only patients not taking LLAs, as many treatments for dyslipidemia affect both TG and LDL-C. Indeed, simple regression analyses limited to the patients without LLAs revealed that the delta_LDL exhibited a significant negative correlation with the delta_TG at 4 weeks (p = 0.016 and R² = 0.20; Table 3) and a trend toward a negative correlation with the delta_TG at 12 weeks (p = 0.054 and R² = 0.23; Table 3).

Table 3

Association between TG and LDL-C with simple regression analysis in patients without taking lipid lowering agents

Objective variable	Weeks	Number of patients	delta_TG5 coefficient	R2	p value
Delta_LDL6	4	28	–0.20	0.20	0.016
Delta_LDL6	12	17	–0.11	0.23	0.054

⁵ Delta_TG: Difference between TG levels at baseline and at each week.

⁶ Delta_LDL: Difference between LDL levels at baseline and at each week.

Discussion

The UK Consensus Statement [3] recommends the prophylactic administration of levothyroxine, with the upper third of the reference range as the target FT4 level. Studies have also reported that bexarotene enhances peripheral thyroid hormone metabolism [9, 10] and that the required amount of levothyroxine is expected to be higher than that used in primary hypothyroidism. In this study, the mean and median levothyroxine dose increased over time of bexarotene therapy; however, up to a maximum of one patient per week achieved the target FT4 levels in the thyroid analysis set (Supplementary Fig. 1 and Supplementary Table 1). The observed insufficient management of hypothyroidism was expected to have AEs on lipid metabolism. Furthermore, only a small number of patients had FT4 levels above the upper reference limit during the study and no patient had increased FT3 levels above the upper reference limit. In other words, although the FT4 levels were slightly elevated in some patients in this study, thyrotoxicosis was not an issue in any of the patients and there was no cause for hesitation regarding prophylactic levothyroxine compensation therapy.

In this study, most patients revealed the FT4 levels below the reference limit and hypothyroidism seemed to be almost inevitable, as reported previously. In addition, similar to the results of a Japanese clinical trial [4], the FT4 levels were below the lower reference limit in 56 of the 86 patients whose thyroid hormones were measured at 1 week of bexarotene therapy, suggesting that thyroid hormone compensation is justified with the coadministration of bexarotene. The early onset of hypothyroidism after bexarotene initiation may be due in part to the bexarotene-induced increase in thyroid hormone degradation in peripheral tissues [9, 10], in addition to the suppression of TSH secretion [5, 10].

Dyslipidemia is the second most common AE in bexarotene-treated patients, and hypertriglyceridemia occurs in about 80% of patients [11]. The management of dyslipidemia is important for the continuation of bexarotene therapy.

Hypothyroidism is well known to affect lipid metabolism, and LDL-C and TG levels have been reported to decrease with levothyroxine compensation therapy for overt hypothyroidism [12]. Therefore, we hypothesized that bexarotene-induced hypothyroidism might be a factor further exacerbating dyslipidemia and investigated the correlation of the bexarotene dose and FT4 levels with LDL-C and TG levels. As shown in Fig. 4, bexarotene and bexarotene-induced hypothyroidism affect dyslipidemia in various ways.

Fig. 4

Potential mechanism of bexarotene-induced hypothyroidism and dyslipidemia

Solid boxes indicate parameters shown to be involved in bexarotene-induced hypothyroidism and dyslipidemia in this study, and dashed boxes indicate parameters known to be involved in hypothyroidism-induced dyslipidemia but not examined in this study.

Bexarotene induces suppression in TSH levels, resulting from repressing TSHβ subunit expression through a specific binding site, RXRE [22]. And reduced TSH levels result in hypothyroidism. Hypothyroidism induces increased LDL levels through decrease in HMG-COA reductase expression, and induces increased TG levels through reduction in activity of lipoprotein (LPL) [23, 24]. Bexarotene activates RXR/LXR and RXR/PPAR and their activation can induce dyslipidemia through various factors, such as sterol regulatory element-binding protein 1c (SREBP1c), carbohydrate response element-binding protein (ChREBP), stearoyl-CoA desaturase (SCD1), and fatty acid synthase (FAS) [22, 25]. Thus, bexarotene and bexarotene-induced hypothyroidism may exacerbate both TG increase and LDL increase. Elevated TG levels are associated with increased small dense LDL-C (sdLDL) levels via TG-rich LDL and hepatic lipase activity [26, 27], which might mask LDL-C increase at least in part.

RXRE, retinoid X receptor response element; RXR/LXR, retinoid X receptor/liver X receptor heterodimer; RXR/PPAR, retinoid X receptor/peroxisome proliferator-activated receptors; SREBP1c, sterol regulatory element-binding protein 1c; ChREBP, carbohydrate response element-binding protein; SCD1, stearoyl-CoA desaturase; FAS, fatty acid synthase; CETP, cholesteryl ester transfer protein; LPL, lipoprotein lipase; HL, hepatic lipase

First, we found that the bexarotene dose significantly correlated with elevated TG levels (Fig. 3A, 3B, and Table 2) and that the effect of FT4 on TG was limited to 1 week after bexarotene initiation (Table 2). Since TG and LDL-C levels were greatly affected by LLAs and a large proportion of patients have started taking LLAs during bexarotene therapy, we could not sufficiently explain the effect of bexarotene dose and FT4 in TG after 4 weeks because of no significant values in p or low values in the R². The frequency of bexarotene-induced hypertriglyceridemia in this study was consistent with previous CTCL trials of patients treated with the standard bexarotene dose of 300 mg/m², which reported hypertriglyceridemia in 80% of the patients [13, 14]. A study evaluating low-dose bexarotene (0.5–3.0 mg/kg) as a treatment option for psoriasis reported hypertriglyceridemia in 56% of the patients [15]. Starting bexarotene at a high dose might increase TG in a bexarotene dose-dependent manner and cause bexarotene-induced decrease in FT4 levels, which also had bexarotene dose dependency.

Second, although the initial incentive of this study was based on the assumption that the increases in TG and LDL-C levels associated with the decrease in FT4 levels would have an additional effect on the bexarotene-induced increase in TG and LDL-C levels (Fig. 4), the effect of FT4 was limited to the early period of the study. LDL-C levels have already been shown to increase because of a lack of thyroid hormone action [6-8]; therefore, we investigated the cause of the observed lack of the effect. LDL-C levels have been reported to decline in patients with severe hypertriglyceridemia [16], and the increase in TG levels might have suppressed an increase in LDL-C levels [17-20]. As LLAs used for dyslipidemia are known to affect both LDL-C and TG levels, we analyzed the correlation between the delta_TG and the delta_LDL in patients not taking LLAs and found that the delta_LDL exhibited a negative correlation with the delta_TG (Table 3). Thus, excluding the effect of LLAs, the increase in LDL-C levels was more modest in patients with higher TG levels.

Third, grade ≥3 AEs on hypertriglyceridemia was reported in 39% of patients, which is higher compared to 7 out of 32 patients (22%) reported from a facility in France [21]. In this report from France, preventive thyroid hormone replacement therapy was administered in 29 out of 32 patients. Additionally, one-third of grade ≥3 AEs on hypertriglyceridemia occurred within 1 week and only few patients on bexarotene therapy have achieved the target FT4 range in Japan. Therefore, it is desirable to initiate early thyroid hormone replacement therapy to reliably achieve the target FT4 range, ensuring the continuation of bexarotene therapy without severe AEs.

This study has several limitations. This was a retrospective observational study, and the frequency and intervals of follow-up were quite variable. Furthermore, thyroid hormones were not measured in many patients who were followed in outpatient visits. Therefore, the rates of hypothyroidism and dyslipidemia might have been underestimated. Second, the thyroid hormone levels were standardized to facilitate the comparison of the thyroid hormone status of patients treated in different institutions. The standardization might have been insufficient to account for the differences in assay methods. Nevertheless, the data source in this study included multiple institutions in Japan and the study results are valuable in understanding the real-world relationship among bexarotene use and endocrine and metabolic AEs.

Conclusion

This real-world study examining the effect of bexarotene therapy on thyroid dysfunction and dyslipidemia in a multicenter study in Japan indicated that hypothyroidism after bexarotene initiation was almost inevitable and occurred quickly. Bexarotene-induced increases in TG showed positive correlation with bexarotene dose and negative correlation with delta_FT4. The apparent lack of FT4 effect on LDL-C might be caused by LDL-C suppression accompanied by increased TG. Based on the features of bexarotene induced hypothyroidism and dyslipidemia, prophylactic and appropriate thyroid hormone compensation therapy should be started along with bexarotene at low doses with subsequent titration while managing dyslipidemia as this might have a beneficial effect for the successful continuation of bexarotene therapy without severe endocrine and metabolic AEs.

Acknowledgments

This study has been conducted by the collaboration of Minophagen Pharmaceutical Co., Ltd. We appreciate all the clinician who provide medical care of the patients and participate in the postmarketing surveillance in Japan.

The authors would like to thank Enago (www.enago.jp) for the English language review.

Funding

Noriko Makita has received funds regarding the investigation of bexarotene-induced thyroid dysfunction from Minophagen Pharmaceutical Co., Ltd., the manufacturer of bexarotene. This research was also supported by a JSPS KAKENHI (19K08996 to NM).

Author contributions

N.M. designed the study. K.M. performed the clinical study, and K.M., J.S. and N.M. analyzed the data. K.N., N.M. and T.I. wrote the manuscript. A.M., H.S., H.B., T.Fuji, Y.H., T.S., C.T., E.K., I.M, H.N., R.A., K.F., C.N., E.N., K.Y., T.Funa., M.A., T.M., M.S. and T.H. followed patients. All authors discussed the data, and helped to revise the manuscript.

Conflict of interest statement

This study has been conducted by the collaboration between Minophagen Pharmaceutical Co., Ltd. and the Japanese Bexarotene Study Group. All data were provided by Minophagen Pharmaceutical Co., Ltd.

N.M. is an editorial board member of Endocrine Journal.

N.M., A.M., T.Fuji, T.S., C.T., E.K., I.M., K.F., C.N., T.Funa, M.S.,T.M. and T.H. have received speaker fees outside of this study. N.M., A.M., T.S. and T.H. have received consultancy fees, and N.M., A.M., T.S., C.N. and T.H. have received research grants from Minophagen Pharmaceutical Co., Ltd.

Data availability

All the data that support the findings of this study was provided by Minophagen Pharmaceutical Co., Ltd. under a research contract. The data are not publicly available due to the contract.

Supplementary Table 1

Thyroid hormone compensation therapy during bexarotene therapy

Weeks	0	1	4	12	24	52
Full analysis set
Number of patients	291	283	236	198	158	121
Median dose of bexarotene	N/A	280 [185–298]	214 [161–287]	193 [129–275]	188 [133–249]	187 [129–257]
Taking L-T41	8	57	171	166	141	111
Mean dose of L-T4 (mcg)	92.2	53.6	53.7	71.1	79.5	95.0
Statins	43	58	70	63	59	41
Fibrates	17	54	102	96	76	46
Omega-3 fatty acids	3	12	45	48	39	27
Ezetimibe	5	7	18	14	21	19
Tocopherol	2	4	16	24	18	12
Thyroid analysis set
Number of patients	125	124	107	90	71	58
Median dose of bexarotene	N/A	285 [224–298]	220 [180–285]	191 [122–285]	169 [119–258]	170 [118–216]
Taking L-T4	0	1	70	73	62	54
Mean dose of L-T4 (mcg)	N/A	25	50.2	76.0	79.7	85.2
Statins	25	30	31	28	27	16
Fibrates	4	19	47	44	33	18
Omega-3 fatty acids	0	3	23	23	18	11
Ezetimibe	4	5	14	8	13	10
Tocopherol	0	1	8	12	6	3

¹ L-T4: levothyroxine

Supplementary Fig. 1

Changes in thyroid hormone managements over time

Changes in the FT4 levels in the thyroid analysis set. Solid areas in the bar graph indicate the number of patients with sufficient thyroid hormone levels (FT4 > 2^nd tertile of the reference range). Dotted areas in the bar graph indicate the number of patients with insufficient thyroid hormone levels (FT4 ≤ 2^nd tertile value of the reference range).

References

• 1   Mehta-Shah  N,  Horwitz  SM,  Ansell  S,  Ai  WZ,  Barnes  J, et al. (2020) NCCN guidelines insights: primary cutaneous lymphomas, version 2.2020. J Natl Compr Canc Netw 18: 522–536.
• 2   Trautinger  F,  Eder  J,  Assaf  C,  Bagot  M,  Cozzio  A, et al. (2017) European organisation for research and treatment of cancer consensus recommendations for the treatment of mycosis fungoides/Sezary syndrome—Update 2017. Eur J Cancer 77: 57–74.
• 3   Scarisbrick  JJ,  Morris  S,  Azurdia  R,  Illidge  T,  Parry  E, et al. (2013) U.K. consensus statement on safe clinical prescribing of bexarotene for patients with cutaneous T-cell lymphoma. Br J Dermatol 168: 192–200.
• 4   Makita  N,  Manaka  K,  Sato  J,  Mitani  K,  Nangaku  M, et al. (2019) Bexarotene-induced hypothyroidism: characteristics and therapeutic strategies. Clin Endocrinol (Oxf) 91: 195–200.
• 5   Sherman  SI,  Gopal  J,  Haugen  BR,  Chiu  AC,  Whaley  K, et al. (1999) Central hypothyroidism associated with retinoid X receptor-selective ligands. N Engl J Med 340: 1075–1079.
• 6   Pearce  EN (2004) Hypothyroidism and dyslipidemia: modern concepts and approaches. Curr Cardiol Rep 6: 451–456.
• 7   Rizos  CV,  Elisaf  MS,  Liberopoulos  EN (2011) Effects of thyroid dysfunction on lipid profile. Open Cardiovasc Med J 5: 76–84.
• 8   Liu  XL,  He  S,  Zhang  SF,  Wang  J,  Sun  XF, et al. (2014) Alteration of lipid profile in subclinical hypothyroidism: a meta-analysis. Med Sci Monit 20: 1432–1441.
• 9   Smit  JW,  Stokkel  MP,  Pereira  AM,  Romijn  JA,  Visser  TJ (2007) Bexarotene-induced hypothyroidism: bexarotene stimulates the peripheral metabolism of thyroid hormones. J Clin Endocrinol Metab 92: 2496–2499.
• 10   Sherman  SI (2003) Etiology, diagnosis, and treatment recommendations for central hypothyroidism associated with bexarotene therapy for cutaneous T-cell lymphoma. Clin Lymphoma 3: 249–252.
• 11   Quereux  G,  Saint-Jean  M,  Peuvrel  L,  Brocard  A,  Knol  AC, et al. (2013) Bexarotene in cutaneous T-cell lymphoma: third retrospective study of long-term cohort and review of the literature. Expert Opin Pharmacother 14: 1711–1721.
• 12   Kotwal  A,  Cortes  T,  Genere  N,  Hamidi  O,  Jasim  S, et al. (2020) Treatment of thyroid dysfunction and serum lipids: a systematic review and meta-analysis. J Clin Endocrinol Metab 105: 3683–3694.
• 13   Duvic  M,  Hymes  K,  Heald  P,  Breneman  D,  Martin  AG, et al. (2001) Bexarotene is effective and safe for treatment of refractory advanced-stage cutaneous T-cell lymphoma: multinational phase II–III trial results. J Clin Oncol 19: 2456–2471.
• 14   Duvic  M,  Martin  AG,  Kim  Y,  Olsen  E,  Wood  GS, et al. (2001) Phase 2 and 3 clinical trial of oral bexarotene (Targretin capsules) for the treatment of refractory or persistent early-stage cutaneous T-cell lymphoma. Arch Dermatol 137: 581–593.
• 15   Smit  JV,  Franssen  ME,  de Jong  EM,  Lambert  J,  Roseeuw  DI, et al. (2004) A phase II multicenter clinical trial of systemic bexarotene in psoriasis. J Am Acad Dermatol 51: 249–256.
• 16   Packard  CJ,  Boren  J,  Taskinen  MR (2020) Causes and consequences of hypertriglyceridemia. Front Endocrinol (Lausanne) 11: 252.
• 17   Otvos  JD,  Collins  D,  Freedman  DS,  Shalaurova  I,  Schaefer  EJ, et al. (2006) Low-density lipoprotein and high-density lipoprotein particle subclasses predict coronary events and are favorably changed by gemfibrozil therapy in the veterans affairs high-density lipoprotein intervention trial. Circulation 113: 1556–1563.
• 18   Rosenson  RS,  Wolff  DA,  Huskin  AL,  Helenowski  IB,  Rademaker  AW (2007) Fenofibrate therapy ameliorates fasting and postprandial lipoproteinemia, oxidative stress, and the inflammatory response in subjects with hypertriglyceridemia and the metabolic syndrome. Diabetes Care 30: 1945–1951.
• 19   Rosenson  RS (2008) Fenofibrate: treatment of hyperlipidemia and beyond. Expert Rev Cardiovasc Ther 6: 1319–1330.
• 20   Rosenson  RS,  Underberg  JA (2013) Systematic review: evaluating the effect of lipid-lowering therapy on lipoprotein and lipid values. Cardiovasc Drugs Ther 27: 465–479.
• 21   Quéreux  G,  Saint-Jean  M,  Peuvrel  L,  Brocard  A,  Knol  AC, et al. (2013) Bexarotene in cutaneous T-cell lymphoma: third retrospective study of long-term cohort and review of the literature. Expert Opin Pharmacother 14: 1711–1721.
• 22   Graeppi-Dulac  J,  Vlaeminck-Guillem  V,  Perier-Muzet  M,  Dalle  S,  Orgiazzi  J (2014) Endocrine side-effects of anti-cancer drugs: the impact of retinoids on the thyroid axis. Eur J Endocrinol 170: R253–R262.
• 23   Sinha  RA,  Singh  BK,  Yen  PM (2018) Direct effects of thyroid hormones on hepatic lipid metabolism. Nat Rev Endocrinol 14: 259–269.
• 24   Liu  H,  Peng  D (2022) Update on dyslipidemia in hypothyroidism: the mechanism of dyslipidemia in hypothyroidism. Endocr Connect 11: e210002.
• 25   Lalloyer  F,  Pedersen  TA,  Gross  B,  Lestavel  S,  Yous  S, et al. (2009) Rexinoid bexarotene modulates triglyceride but not cholesterol metabolism via gene-specific permissivity of the RXR/LXR heterodimer in the liver. Arterioscler Thromb Vasc Biol 29: 1488–1495.
• 26   Sniderman  AD,  Rosenbloom  M (2005) If apoB is so good, why isn’t everybody measuring it? One reason why we need The Netherlands Journal of Medicine! Neth J Med 63: 232–235.
• 27   Hayashi  T,  Koba  S,  Ito  Y,  Hirano  T (2017) Method for estimating high sdLDL-C by measuring triglyceride and apolipoprotein B levels. Lipids Health Dis 16: 21.