Double trouble: a case of an ataxic young man with coeliac disease and cerebrotendinous xanthomatosis

Abstract

We present the case of a 29-year-old south Asian man born of consanguineous marriage, presenting with ataxia, peripheral neuropathy and cognitive impairment. An initial diagnosis of coeliac disease was thought to explain the pertinent clinical features; however, further investigation led to an additional diagnosis of the rare yet treatable autosomal recessive condition, cerebrotendinous xanthomatosis. With both conditions employing highly diverse and overlapping clinical phenotypes, this contributed to a delay in diagnosis. Our report highlights the importance of paying close attention to both the clinical phenotype and family history.

Background

Cerebrotendinous xanthomatosis (CTX) is a rare, autosomal recessive disorder of cholesterol and bile acid metabolism. With a heterogeneous clinical phenotype, symptoms of CTX can include tendon xanthomas, chronic diarrhoea, juvenile cataracts, cognitive impairment, neuropathy and ataxia.^1–3 Most inherited conditions causing ataxia and neuropathy merit no specific treatment; conversely, CTX is different. The earlier the diagnosis is made, the sooner oral chenodeoxycholic acid (CDCA) therapy can commence, ultimately improving the prognosis.⁴ However, CTX can be notoriously difficult to diagnose, in part due to its diverse presentation. Here, we present the first case of CTX with coeliac disease (CD), a disorder whose own diverse phenotype had the potential to disguise many of our patient’s other symptoms and contribute to a delay in diagnosis of this rare disorder.

Case presentation

A 29-year-old south Asian man was referred to the neurology clinic with a 6-month history of falls and proximal leg weakness. Most of the history was provided by his father. The patient was revealed to have longstanding diarrhoea. Normal developmental milestones in childhood were reached. He attended a mainstream school before being identified as having a moderate learning disability at age 15, with scores below the 1st centile in MISIC-III (Malin's Intelligence Scale for Indian Children) testing. He denied any sphincter or bulbar symptoms. He did not take any regular medications. He was one of four siblings born to healthy first-cousin parents, who were themselves part of a multiply consanguineous family (figure 1A). His siblings were said to be healthy. A paternal first cousin had an undiagnosed neurological condition characterised by difficulty walking and cognitive impairment, but as they were resident in Pakistan no further clinical details was available.

Figure 1.

Composite image demonstrating key features of our patient’s initial presentation. (A) Our patient’s family tree, revealing consanguinity across multiple generations (consanguineous marriage is depicted through linkage of family members with double horizontal lines). Please note that while one of the patient’s cousins was suspected to also have cerebrotendinous xanthomatosis, this was not able to be confirmed; they have hence not been included as an affected individual here. (B) Photographs of our patient demonstrating smoothly thickened Achilles tendons.

Physical examination revealed a thin gentleman with a broad-based ataxic gait. He was unable to tandem walk and had a mild bilateral postural tremor. He had pes cavus with high arches, hammer toes and Achilles tendon thickening (figure 1B), with noticeable wasting of the quadriceps muscles. Motor examination demonstrated symmetrical proximal leg and arm weakness. There was a loss of vibratory sensation and proprioception in a length-dependent manner, with retained pinprick sensation. Lower limb reflexes were brisk with upgoing plantars observed bilaterally. A positive Romberg’s test suggested a peripheral component to the ataxia. There was no dysarthria or nystagmus. Cranial nerve and ophthalmological examination were both unremarkable.

Investigations

On initial presentation, a basic blood panel was ordered. All results were normal except for an isolated hyperbilirubinaemia and low vitamin D and E levels, which were subsequently treated (table 1). Further testing revealed positive tissue transglutaminase and antiendomysial antibodies, with duodenal biopsy confirming a histological diagnosis of coeliac disease (CD).

Table 1.

Blood panel on initial presentation

Test	Result	Reference range
Haemaglobin	162 g/L	130–180 g/L
White cell count	7.1×109/L	4.0–11.0×109/L
Mean cell volume	82.0	80.0–100.0 fL
Platelet count	265	150–400×109/L
Erythrocyte sedimentation rate (ESR)	2	1–10 mm/hour
Serum ferritin	56	20–300 µg/L
Serum folate	5.5	3.1–20.0 µg/L
Serum vitamin B12	463	200–900 ng/L
C reactive protein	<1	0–10 mg/L
Total bilirubin	46	<20 µmol/L
Alanine transaminase (ALT)	19	<50 U/L
Aspartate transaminase (AST)	20	<40 U/L
Alkaline phosphatase (ALP)	95	30–130 U/L
Albumin	41	35–50 g/L
Sodium	139	133–146 mmol/L
Potassium	4.8	3.5–5.3 mmol/L
Urea	4.9	2.5–7.8 mmol/L
Creatinine	71	40–130 µmol/L
Creatine kinase	80	40–320 U/L
25-OH vitamin D	43	≥50 nmol/L
Vitamin E	8	15–45 µmol/L

Despite adherence to a gluten-free diet, the diarrhoea persisted and his ataxia worsened, prompting additional investigation. An MRI scan of the brain and spinal cord demonstrated scattered white matter lesions on T2-weighted imaging, with symmetrical hypoattenuation on T1 in the dentate nucleus of the cerebellum (figure 2A). There was no evidence of signal change in the cord. Genetic testing for spinocerebellar ataxia (SCA) 1, 2, 3, 6, 7 and Friedreich’s ataxia were negative.

Figure 2.

Summary of key investigation findings; (A) MRI brain T1 axial section demonstrating bilateral symmetrical hypoattenuation in the dentate nuclei of the cerebellum (white arrows); (B) electroencephalogram revealing a moderately severe encephalopathic picture with diffusely slow background rhythms and (C) neurophysiological studies of motor nerves showing mildly slowed velocities in keeping with an intermediate picture. NCS, nerve conduction studies; Amp, amplitude of the distal compound muscle action potential; DML, distal motor latency; F wave, F wave minimum latency; MCV, motor conduction velocity; NA, not applicable; ND, not done. Reference values for individual nerves given in brackets as per Glasgow Neurophysiology departmental guidelines.

An electroencephalogram (EEG) revealed a moderately severe encephalopathic picture with a diffusely slow background rhythm (figure 2B). While sensory nerve conduction studies were normal, lower limb motor studies demonstrated mild conduction slowing across both common peroneal nerves with delayed F waves across all four lower limb nerves studied (figure 2C). Electromyography (EMG) revealed mild neurogenic changes with either reduced or minimal recruitment. The reporting neurophysiologist wondered if these changes could represent a distal inherited motor neuropathy.

Consequently, the clinical phenotype could now be refined as ataxia, neuropathy and encephalopathy. The pes cavus suggested the neuropathy to be longstanding and hereditary. The patient was referred to the neuropathy clinic, where physical examination led to a presumptive diagnosis of CTX (figure 1B); this was confirmed on discovery of a raised cholestanol level of 98 µmol/L (reference range 3–16) and characteristic bile acid analysis. Sequencing of the CYP27A1 gene identified homozygosity for a missense variant, c.776A>G; p.(Lys259Arg). Carrier status was confirmed in the parents. The variant, which is found at low frequencies in population databases, has been reported in several cases in the literature.^{5 6} In silico splice prediction tools predict creation of an acceptor splice site. This, together with the diagnostic biochemical result, led to the National Health Service (NHS) genetics laboratory to report the variant as pathogenic according to the Association for Clinical Genomic Science (ACGS) Best Practice Guidelines for Variant Classification.⁷

Differential diagnosis

When faced with a neurological problem, the role of the clinician is to first localise the lesion. This young man demonstrated both upper (brisk reflexes, upgoing plantar responses) and lower (wasting, pes cavus) motor neuron signs. Features of a central ataxia (broad-based gait) and a peripheral sensory ataxia (dorsal column loss in the lower limbs, positive Romberg’s test) were also present. Multifocal pathology of the central nervous system such as multiple sclerosis was initially thought most likely; however, closer attention to the clinical phenotype pointed to dysfunction of the entire neural axis.

The combination of upper and lower motor neuron signs along with a gait ataxia, in the absence of other cerebellar features such as dysarthria, intention tremor and abnormal eye movements, could indicate spinal cord pathology; for example, subacute combined degeneration of the cord (SACD). SACD causes degeneration of the corticospinal and spinocerebellar tracts and the dorsal columns, resulting in gait and sensory ataxia with spasticity; this is also commonly associated with a peripheral neuropathy. Vitamin B12 or E deficiency, secondary to a malabsorption disorder such as CD, can lead to SACD.^8–11 However, there was no radiological evidence of SACD on MRI scan and no improvement with vitamin E supplementation.

CD could have explained many of the clinical features in our patient, with its diverse clinical phenotype encompassing ataxia and peripheral neuropathy.^{8 10} However, a gluten-free diet did not improve his diarrhoea. Genetic ataxic disorders such as Friedreich’s ataxia and SCA were subsequently considered. However, testing for the common SCAs was negative, while Friedreich’s was considered unlikely given the late onset of ataxia.

His neurophysiological studies yielded unusual results, with mildly slowed velocities and delayed F waves confined to the lower limb motor nerves only. The presence of pes cavus prompted consideration of a hereditary neuropathy such as Charcot–Marie–Tooth (CMT) disease. However, the neuropathy in our patient was part of a more complex clinical syndrome, rather than a pure neuropathy as seen in CMT. Ultimately, genetic studies for CMT-related genes were negative.

A key observation in this case was thickening of the Achilles tendons. This can be attributed to a number of disorders, including Achilles tendinosis, Achilles tendon xanthoma, rheumatoid arthritis, neoplasms and CTX.¹² Here, it prompted consideration of CTX.

Treatment

The patient started CDCA therapy, the current first-line treatment for CTX.⁴ CDCA provides negative feedback on the cholesterol pathway, thereby reducing any further accumulation of cholestanol and cholesterol.¹³ Additionally, a statin was prescribed with the aim of exerting further negative feedback on cholesterol metabolism.

Outcome and follow-up

One year into treatment, our patient’s Scale for the Assessment and Rating of Ataxia Score was 9.5/40 compared with 10/40 from the previous year, indicating stabilisation in his condition. His cholestanol levels decreased from 98 µmol/l to 13 µmol/L (normal range: 3–16); these will continue to be monitored on an annual basis. A follow-up EEG demonstrated a definite improvement in background rhythms, with subsequent neurophysiological tests evidencing modest improvement in the lower limb motor nerve conduction velocities.

Genetic testing of the patient’s relatives revealed his sister to have CTX; investigations are pending in other relatives.

Discussion

CTX is a rare autosomal recessive disorder of cholesterol and bile acid metabolism, resulting from a functional defect of the mitochondrial enzyme sterol 27-hydroxylase, as encoded by the CYP27A1 gene.^{1 2} Sterol 27-hydroxylase normally catalyses the hydroxylation reactions involved in cholesterol and bile acid synthesis; with CTX, the lack of sterol 27-hydroxylase results in the accumulation of cholesterol and bile acid precursors in the plasma and tissues.^{1 2}

With a heterogeneous phenotype, CTX has the potential to be masked by other disorders. In the case of our patient, there were several clinical features that could be explained by CD, thereby potentially endangering the discovery of the additional diagnosis of CTX. To the best of our knowledge, this is the first case of CTX with CD.

CTX presents with features such as diarrhoea (31%), cognitive impairment (57%), cataracts (75%), xanthomas (73%) and neurological symptoms (65%).^{2 3 14} Further delving into CTX’s neurological landscape, a recent literature review identified the breakdown of symptom prevalence as: ataxia (59%), cognitive decline (40%), gait difficulties (38%), sensory loss (21%), seizures (19%) and parkinsonism (10%).¹⁵ All of these neurological features have also been reported in CD.8 10 16 17

Like CTX, CD also employs a diverse clinical phenotype, and can be crudely categorised as either intestinal or extraintestinal based on its presentation.¹⁶ Intestinal disease can encompass diarrhoea, anorexia, bloating, abdominal pain, constipation, weight loss and malabsorption, while extraintestinal CD can range from reproductive issues to neurological problems.^{16 17} Ataxia and peripheral neuropathy are the two most common neurological manifestations in CD; notably, our patient presented with both.^{8 10} Furthermore, he lacked cataracts—the most ubiquitous symptom among CTX patients worldwide. While not conclusive, the clinical picture presented thus far was arguably deceptive and in favour of CD.

Although the presentation appeared to fit with CD, there were other factors to consider. Consanguinity is an important risk factor for autosomal recessive disease, particularly among south Asians where endogamy is social practice.¹⁸ The MRI findings of high signal changes in the dentate nuclei were also not in keeping with CD, being more suggestive of CTX.¹⁹ Furthermore, the presence of xanthomas is not synonymous with CD. The continued deterioration of the patient, despite adhering to a gluten-free diet, was an extra factor in searching for an additional diagnosis. Taking into account all the afore-mentioned, we can confidently attribute our patients’ pertinent clinical features to the diagnosis of CTX.

Through our case, we hope to convey the importance of thorough investigation of those presenting with neurological dysfunction. While our patient’s clinical features were generally consistent with a diagnosis of CD, delving deeper revealed an additional treatable condition which had implications for the wider family. By noting details such as the patient’s consanguineous origins, MRI findings and the presence of xanthomas, we were able to target subsequent investigations, facilitating treatment to ultimately improve prognosis.³

Learning points.

• The devil is in the detail: accurate description of the patient’s clinical phenotype and localising the lesion(s) is paramount in reaching a diagnosis.

• Do not underestimate the importance of an extended family history—here, it raised suspicion of a possible autosomal recessive disorder.

• There are two principles of relevance in neurology: Occam’s razor, with the simplest explanation often being correct, and Hickham’s dictum, whereby ‘a man can have as many diseases as he damn well pleases.’ Our patient’s case highlights the latter; the coexistence of another disease can mask the features of an alternative diagnosis, and atypical features should always prompt the clinician to consider a wider differential.