Allan-Herndon-Dudley syndrome (AHDS) is a rare disease characterized as an X-linked mental retardation syndrome that is associated with neuromuscular involvement and thyroid function abnormality. The first case of AHDS was reported by Allan et al. [1] in 1944, but its association with solute carrier family 16 member 2 (SLC16A2) mutation was first described in 2004 [2,3].

SLC16A2 encodes monocarboxylate transporter 8 (MCT8), a transmembrane transporter of the active form of thyroid hormone, triiodothyronine (T₃) [4]. Thyroid hormones are important for neurodevelopment and normal growth. Because MCT8 is highly expressed in the brain-blood barrier and neuronal cells of the developing brain [5], MCT8 defects result in severe psychomotor deficits including central hypotonia, generalized spasticity, dystonia, and moderate-to-severe intellectual disability [2,3]. Impaired myelination on brain magnetic resonance imaging (MRI) has been reported in the majority of affected patients, but the heterogeneity of MRI findings needs further evaluation [6,7].

The thyroid profile of patients with AHDS shows elevated T₃ levels, low-to-normal free tetraiodothyronine (T₄) levels, and normal-to-high thyroid-stimulating hormone (TSH) serum levels [2,3]. Due to the elevated serum levels of T₃, other organs, such as the heart, muscles, and liver, which are less dependent than the brain on the MCT8 protein for transporting thyroid hormones into the cell, are affected by peripheral thyrotoxicosis, the symptoms of which include muscle wasting, low body weight, and tachycardia [8,9].

The phenotypic spectrum of AHDS is highly varied, including patients who achieve independent walking with mild intellectual disability to those with severe encephalopathy [10,11]. The phenotype-genotype correlation is not yet understood fully, but residual protein activity is suggested to be associated with clinical severity [12].

Herein, we present nine patients who were diagnosed with genetically confirmed AHDS. Their clinical, laboratory, radiologic, and genetic characteristics were described in detail to broaden the scope of known phenotypic characteristics and to expand our understanding of the disease.

In this study, we conducted a detailed investigation of the clinical, genotypic, and laboratory characteristics of nine male patients with genetically confirmed AHDS.

After the causative gene SLC16A2 was revealed, the pathophysiology of AHDS has been understood as resistance to thyroid hormones in the brain due to the deficiency of the thyroid hormone transporter, MCT8, which plays a key role in transporting thyroid hormones across the blood-brain barrier and into neuronal and glial cells in developing brains [13]. The MCT family, which is a member of the major facilitator superfamily (MFS), has a shared structural fold with 12 transmembrane domains (TMDs) and cytosolic N- and C-termini [14]. This transmembrane protein is thought to control the cellular influx and efflux of thyroid hormones and regulate intracellular hormone levels [4,15]. An expression analysis of MCT8 revealed that it is present in cerebral neurons, glial cells, and the cellular structures that compose the blood-brain barrier and blood-cerebrospinal fluid barrier in the fetal brain [5,16]. In addition to the brain, MCT8 is also abundant in the thyroid glands, kidneys, and liver [4].

A distinctive thyroid hormone profile is a hallmark of AHDS [2,3]. Typically, the T₃ level is elevated, the free T₄ level is normal to low, and the TSH level is normal to mildly elevated. Considering the diseases resulting from defects of other thyroid hormone transporters, such as OATP1C1 deficiency, which shows a normal thyroid hormone profile [17], abnormal serum thyroid hormone levels not only provide a clue for diagnosis, but also show the association of MCT8 and the hypothalamus-pituitary-thyroid (HPT) axis. Studies on the relevance of MCT8 to the HPT axis in Mct8-knock-out mice have shown that the hypothalamus and pituitary gland seem to poorly sense serum thyroid hormone concentrations, which suggests that TSH level is not a reliable indicator of the thyroid hormone status of the brain and body [18,19]. Although an elevated T₃ level was thought to be a fingerprint of AHDS, normal T₃ levels were reported in some cases [10,11]. In our cohort, patient 5, who had an elevated T₃ level of 190 ng/dL (reference range, 64 to 152) at the age of 9 months showed normal serum T₃ levels of 144.5 and 115.7 ng/dL at the ages of 4 and 5 years, respectively, with decreased free T₄ levels, but the ratio of T₃ to free T₄ was still elevated. If AHDS is clinically suspected, it is recommended to test the patients’ thyroid hormone status repeatedly and keep in mind that the T₃ level can vary. An increased T₃ to free T₄ ratio may be a more reliable parameter for clinical suspicion of AHDS.

The types of mutations in the SLC16A2 gene vary considerably, from single-nucleotide variants to large deletions. In our cohort, we reported three missense mutation, one nonsense mutation, one frameshift mutation, one splicing-site mutation, and two exon deletions. Several functional studies have shown that genetic mutations of SLC16A2 cause a variable degree of functional defects in MCT8, including impairments of homodimerization, localization to the plasma membrane, and transport function [20,21]. Two previously reported missense mutations were found in patients 4 and 5, who had c.439G>A [p.G147R] and c.449C>T [p.A150V] mutations, respectively. The c.439G>A [p.G147R] mutation changes the glycine located in the second TMD to arginine. In functional studies [20], it impaired the thyroid hormone transporter activity without affecting localization to the plasma membrane. The clinical manifestation of two patients with the p.G147R mutation from the same family reported previously [18] was similar to patient 4, who had a severe phenotype with spastic paraplegia and absence of verbal output, except for seizures, which were not observed as of the age of 4 years (at his last visit). The c.449C>T [p.A150V] mutation in patient 5 changed alanine to valine in the second TMD. This has been reported previously, and all cases had the same severe phenotype as patient 5 [19,22,23]. In the functional analysis [19,21], this mutation was reported as not affecting dimer assembly [21], but impairing expression in the plasma membrane [19]. The novel 1396G>C [p.A466P] missense mutation from patients 8 and 9 changed alanine to proline in the 11th TMD and caused a relatively mild phenotype, as the patients gained the ability to walk and had normal brain MRI findings. A functional study was not performed for these patients. In our cohort, exon deletion was found in patients 1 and 7, who showed exon 1 and exon 5 deletion, respectively; sole exon 5 deletion has not been reported previously. Both patients had a severe phenotype, with spastic paraplegia and no verbal output. Patient 3 had a c.1026+1G>A spicing site mutation, which was previously reported in one case [24]. In that case, the clinical presentation was similar, showing a severe phenotype, but the thyroid hormone profile was different because the previously reported case showed lower free T₄ levels than patient 3, who had normal free T₄ levels [24]. The novel frameshift mutation c.80delG [p.G27Vfs*57] and nonsense mutation c.58C>T [p.Q20X] on exon 1 were observed in patients 2 and 6, respectively, who had severe phenotypes.

Axial hypotonia and spasticity were the most common features in our cohort, which corresponds well with previously reported cases [10,11]. Among several extrapyramidal phenotypes, exaggerated startles or myoclonus was a notable phenotype in this cohort, as observed in 55% (5/9) of the patients. Dystonia, which is also a commonly reported manifestation of AHDS, was also widely observed in this study.

Along with the increasing number of reported patients and genetic mutations of AHDS over the last decade, some patient groups have shown milder phenotypes than others, with mild intellectual disability or achieving ambulation [10,18,25-29]. In this cohort, patients 8 and 9, who had the same missense 1396G>C [p.A466P] mutation at the end of the 11th TMD, showed a distinctively milder phenotype compared to the other patients. Their age of onset was later than that of the others and they achieved ambulation around 5 years of age. Although both had the same variants and could be categorized as having a mild phenotype, they had phenotypic differences. At the initial neurologic examination at the age of 12 months, patient 9 showed more classical clinical features, such as hypotonia, spasticity, and hyperreflexia, whereas patient 8 showed none of these features except for mild hypotonia. During the follow-up period, these clinical features became evident in patient 8. In this patient with a mild phenotype, the initial neurologic examination was uncertain; thus, thoughtful follow-up is crucial.

To our knowledge, cataplexy, which was a notable shared phenotype of patients 8 and 9, has not previously been reported in patients with AHDS. As one of the features of narcolepsy, cataplexy is thought to be related to the deficiency of orexin, a neurotransmitter secreted by neurons in the lateral hypothalamus [30], which plays a role in the sleep-wakefulness cycle. There is a topographical similarity in that both orexin-producing neurons and MCT8-expressing cells exist in the hypothalamus [31]. Although the underlying mechanism is unclear, several studies have shown evidence supporting the association between the HPT axis and the orexin system. In an animal study, treatment with a thyrotropin-releasing hormone analogue improved cataplexy [32] and a low level of TSH was reported in patients with orexin-deficient narcolepsy [33]. The finding of cataplexy in patients with AHDS could also support the suggested link between the two systems; thus, further research is needed.

Several phenotypes in patients with AHDS, including poor weight gain, decreased muscle bulk, and tachycardia as reported previously, were thought to be caused by peripheral thyrotoxicosis. Poor weight gain was also a common phenotype in this cohort. Skeletal muscle is one of the major target organs for thyroid hormones, which regulate homeostasis. The regeneration of skeletal muscle and excessive T₃ levels in patients with AHDS changes the expression pattern of numerous proteins toward a catabolic status, which results in muscle wasting [34]. Elevation of lactic acid was observed in about half of the patients, and one-third of them experienced metabolic acidosis. Along with lactic acid, the increased levels of AST, ALT, CK, and ammonia were thought to reflect the catabolic status of skeletal muscle [8]. These laboratory abnormalities can sometimes lead to the misdiagnosis of mitochondrial or muscle disease; for example, previous reports of brain MRI findings in patients with AHDS [6] showed that delayed myelination or hypomyelination was a major accompanying feature with or without cerebral atrophy. In some previous research, normalization of myelination in the successive follow-up of brain MRI was reported [6,7]; however, this was not observed in our case. The observation of severe hypomyelination on brain imaging can lead to a misdiagnosis of other neurodegenerative diseases that are categorized into leukodystrophy, such as lysosomal storage disease or mitochondrial disease. Normal brain MRI findings were also observed in two patients with milder phenotypes, which makes AHDS difficult to diagnose.

Due to the specific pathophysiology and significant burden of AHDS, various efforts have been made to obtain an effective treatment [35]. Recently, a second-phase clinical trial of treatment with triiodothyroacetic acid, which is a thyroid hormone analogue, showed clinical improvement of peripheral thyrotoxicosis and the possibility of improving patients’ neurologic status if treated at a young age [36]. Therefore, an early diagnosis is becoming more important.

Herein, we report nine patients who were genetically confirmed with AHDS with mutations of SLC16A2, which encodes MCT8, a thyroid hormone transporter. Phenotypic and genetic variability is a notable feature of the disease, which makes an early diagnosis difficult. Understanding its broad clinical features, which reflect the opposite patterns of thyroid hormone status in the brain and periphery, with a variable serum thyroid profile and MRI findings will help to stimulate clinical suspicion and guide early confirmatory genetic tests. Importantly, a thyroid function test including T₃ levels should be considered when encountering male patients with severe developmental delay and spasticity.