Clinical manifestations

Presentation and course

Gluten-related disorders affect the nervous system in various ways. Ataxia (often concurrent with cerebellar degeneration) is, by far, the most common manifestation. Other manifestations include peripheral neuropathy, ranging from ganglionopathy to length-dependent axonal sensorimotor neuropathy to small fiber neuropathy; headaches, often with abnormal white matter on MR imaging; myoclonic ataxia; epilepsy, sometimes with occipital calcifications; and a number of psychiatric ailments (anxiety, depression, and psychosis).

Gluten ataxia

In the absence of an alternative etiology for the ataxia, gluten ataxia was originally defined as cerebellar ataxia with positive anti-gliadin antibodies (30; 32; 31). In a series of 1500 patients with progressive ataxia evaluated over a period of more than 20 years at a specialist ataxia center (Sheffield, UK), gluten ataxia had a prevalence of 20% amongst all ataxias and a prevalence of more than 40% amongst idiopathic sporadic ataxias (39). A systematic review of all studies reporting the prevalence of anti-gliadin antibodies in idiopathic ataxias confirmed an association between these antibodies and ataxia across different geographic regions, although there was variability in the prevalence (52). Such variability may relate to geographic variations in the prevalence of gluten sensitivity or the serological assay used.

Gluten ataxia usually presents with pure cerebellar ataxia that primarily affects gait, occasionally in combination with a peripheral neuropathy (32) or, rarely, in combination with myoclonus (70). Gluten ataxia is usually of insidious onset; however, it can also be rapidly progressive, mimicking paraneoplastic cerebellar degeneration (60). Gaze-evoked nystagmus is common. As with most immune-mediated ataxias, there is a predilection for vermian involvement, and, therefore, all patients have gait ataxia. Less than 10% of patients with gluten ataxia have any gastrointestinal symptoms; however, about 40% of the patients have enteropathy on duodenal biopsy.

Depending on the duration of the ataxia, patients with gluten ataxia usually have evidence of cerebellar atrophy on MR imaging, with a particular predilection for the cerebellar vermis. MR spectroscopy of the vermis is abnormal in all patients with gluten ataxia (low N-acetyl aspartate/creatine-NAA/Cr area ratio). MR spectroscopy is abnormal, even in patients with gluten ataxia without cerebellar atrophy (39).

The management of patients with gluten ataxia is based on the input of a specialist dietician advising on a strict gluten-free diet with regular follow-up to ensure that the antibodies are completely eliminated, which usually takes 6 to 12 months.

The response to treatment depends on the duration of the ataxia prior to diagnosis and the strictness of the gluten-free diet. A systematic case-controlled study of 43 patients presenting with ataxia due to gluten sensitivity, with or without an enteropathy, showed that patients who adhered to a strict gluten-free diet (having serological evidence of elimination of anti-gliadin antibodies while being on the diet) had significant improvement on ataxia test scores and on the subjective global clinical impression scale compared to the control group, which comprised of patients who decided not to go on a gluten-free diet (28). The two groups did not differ with regards to age, gender, and severity of ataxia. Interestingly, the improvement was apparent irrespective of the presence of enteropathy.

In a large study of 117 patients with gluten ataxia, a gluten-free diet was found to have a positive effect on MR spectroscopy of the cerebellum (36). In particular, the NAA/Cr area ratio from the cerebellar vermis increased in 62 out of 63 (98%) patients on a strict gluten-free diet (as indicated by the elimination of anti-gliadin antibodies), in 9 out of 35 (26%) patients on a gluten-free diet but with positive antibodies, and in only 1 out of 19 (5%) patients not on a gluten-free diet. Demonstration of an increased NAA/Cr ratio on repeat scanning following a strict gluten-free diet supported previous findings of clinical improvement of the ataxia. The presence of enteropathy was not a prerequisite for such improvement; indeed, there were no differences in the response between patients with or without enteropathy.

Most patients with gluten ataxia respond well to a gluten-free diet. The absence of such a response raises three possibilities. The commonest explanation would be ongoing exposure to gluten due to partial adherence to the gluten-free diet. This can be inferred by the ongoing positive serology in such patients and the presence of gluten immunopeptides in the urine of these patients when tested randomly. Rarely, patients may require additional treatment with immunosuppressive medication because diet alone may not be able to fully suppress the autoimmune process that results in cerebellar damage. Such patients are often negative for the relevant antibodies, and a review by a dietitian suggests very strict adherence to the gluten-free diet. These patients may have refractory celiac disease, which, by definition, is no longer responsive to a gluten-free diet, and they may benefit from the use of mycophenolate as an additional treatment for their ataxia. The diagnosis of refractory celiac disease can be made with repeat duodenal biopsies that show specific histopathological abnormalities. Finally, the absence of response may be due to the fact that the diagnosis is wrong, and the patient has an alternative or additional cause of ataxia (eg, genetic).

Myoclonic ataxia, in which the myoclonus is of cortical origin, can be seen in refractory celiac disease, where a gluten-free diet alone is not sufficient to suppress the immune process. In the largest published series of myoclonic ataxia secondary to refractory celiac disease, five patients were treated with mycophenolate and one with rituximab and IVIG (70). Although the ataxia and enteropathy improved, the myoclonus remained the most disabling feature of the illness. Two of the patients have been treated with cladribine, which has resulted in significant neurologic improvement.

Gluten neuropathy

Gluten neuropathy is defined as an otherwise idiopathic sporadic neuropathy with serological evidence of sensitivity to gluten (34). The most common type is symmetric sensorimotor axonal (large fiber), length-dependent peripheral neuropathy (about 75% of cases), followed by sensory ganglionopathy, an asymmetric form of pure sensory neuropathy in which the pathology is within the dorsal root ganglia (40).

Small fiber neuropathy, which is characteristically painful, can also be seen in the context of gluten sensitivity. The majority of patients with pure small fiber neuropathy related to gluten sensitivity report a burning sensation in a length-dependent manner (mainly soles and palms). However, a small proportion of patients present with a non-length-dependent pattern of symptoms, suggesting that the predominant pathology lies in the dorsal root ganglia (08; 80). Small fiber involvement may also cause autonomic dysfunction (20). Though the exact prevalence of gluten-related pure small fiber neuropathy has yet to be confirmed, more than half of the patients with gluten neuropathy report pain, suggesting that small fibers are often involved (78).

Gluten neuropathy is a slowly progressive condition. About 25% of patients have evidence of enteropathy on biopsy (celiac disease), but the presence or absence of an enteropathy does not influence the positive effect of a strict gluten-free diet. In a systematic, controlled study on the effect of a gluten-free diet in a large series of patients with gluten neuropathy of the sensorimotor axonal type, a clear clinical and neurophysiological improvement was demonstrated in the patients adhering to a strict gluten-free diet after 12 months, with serological elimination of the anti-gliadin antibodies (34). In particular, there was a significant increase in the sural sensory nerve action potential, the predefined primary endpoint, as well as subjective improvement of the neuropathic symptoms. Subgroup analysis showed that there is less capacity for recovery when the neuropathy is severe. In a large series of patients with gluten sensitivity and sensory ganglionopathy, it was shown that strict adherence to a gluten-free diet may result in stabilization of the neuropathy irrespective of the presence of enteropathy (34).

Gluten neuropathy can cause a significant burden on the overall quality of life of patients with gluten neuropathy. In a case-controlled study, patients had worse scores in physical functioning, role limitations due to physical health, energy or fatigue, and general health subdomains of the SF-36 questionnaire compared to controls (79). Interestingly, after adjusting for age, gender, and disease severity, being on a strict gluten-free diet correlated with better SF-36 scores in the pain domain and in the overall health change domain, suggesting that along with clinical and neurophysiological improvement, a strict gluten-free diet can lead to better overall quality of life.

The beneficial effect of IVIG as an adjuvant option to a gluten-free diet for the treatment of neuropathic pain associated with gluten-related small fiber neuropathy (two patients) and gluten neuropathy (one patient) with concomitant gluten ataxia has been suggested (71). Similarly, another study reported clinical improvement following IVIG and gluten-free diet implementation in an isolated case of a patient with a combination of gluten neuropathy and gluten ataxia (02). However, there are no reports or controlled trials on the use of IVIG as a monotherapy in gluten neuropathy. Therefore, the therapeutic potential of IVIG in gluten neuropathy in patients who are not adhering to a gluten-free diet remains questionable.

Gluten encephalopathy

The term “gluten encephalopathy” was introduced in 2001 to describe a combination of frequent, often intractable headaches, cognitive complaints (sometimes patients describe these as “foggy brain”), and white matter abnormalities on brain MR imaging (35). Despite the fact that the white matter abnormalities do not resolve following a gluten-free diet, such a diet stops the progression of MRI changes, and both associated cognitive difficulties and headaches usually resolve.

The headaches take the form of chronic migraines and are often unresponsive to the usual acute and prophylactic medications. Headaches affect both adults and children with celiac disease and gluten sensitivity approximately two times more frequently than controls (53; 18; 59; 77). Although there is no correlation between the number of years on a gluten-free diet and migraine severity, Lionetti and colleagues showed that a gluten-free diet has a beneficial effect: a reduction in headache frequency in the majority of patients (53). Approximately 25% of patients manage to become headache free.

In a population-based retrospective cohort study conducted in Sweden, the authors reported that among 28,638 patients with celiac disease and 143,126 controls, headache-related visits occurred in 4.7% and 2.9% of each group, respectively, suggesting a hazard ratio of 1.7 (95% CI 1.6–1.8; p < 0.0001) (49).

The white matter changes often seen in the context of gluten encephalopathy may be responsible for the cognitive deficits that can be found in patients with celiac disease (14). A study confirmed cognitive deficits in patients with celiac disease, which appear to be established on diagnosis. These deficits stabilize assuming patients adhere to a strict gluten-free diet. A possible association between progressive cognitive impairment and celiac disease has been suggested in small case series (56). Swedish epidemiological data of patients with celiac disease show that these patients are not at an increased risk for dementia overall; however, subgroup analysis indicates an increased risk of vascular dementia (48). The white matter abnormalities encountered in gluten-sensitive patients are identical to what gets labeled as “vascular” disease, which is really what defines vascular dementia in the context of cognitive deficits.

Therapeutic options for patients with advanced dementia are not available. Even a gluten-free diet at this advanced stage of neural damage is unlikely to have a beneficial effect; however, it may arrest progression. In newly diagnosed celiac disease, cognitive performance improves with adherence to a gluten-free diet parallel to mucosal healing (14). Therefore, a gluten-free diet should be introduced as soon as possible in patients with gluten sensitivity or celiac disease as, even in the absence of mild cognitive impairment, it has a potentially protective effect.

An important issue regarding the neurology of celiac disease is the presence of neurologic deficits in patients with “classic” celiac disease, who are often treated purely by gastroenterologists and without any referrals to neurology. A study confirmed the presence of a cognitive deficit (in reaction time) and white matter tract injury as evidenced via diffusion tensor imaging (13). The study used third-party population data from a National UK Biobank, thereby ensuring that participants with a diagnosis of celiac disease (who were otherwise healthy) were represented. Further, as the data collection had been performed by another study team, there was no ascertainment or recruitment bias.

A clinical study of 100 consecutive patients with newly diagnosed celiac disease presenting to gastroenterologists showed that gait instability was reported in 24%, persisting sensory symptoms in 12%, and frequent headaches in 42% (27). On examination, 29% had gait ataxia, 11% had nystagmus, and 10% had distal sensory loss. Sixty percent of these patients had abnormal brain imaging, either in the form of reduced NAA/Cr from the cerebellum or excessive white matter changes. The study concluded that neurologic deficits are already present at the time of diagnosis of celiac disease. However, this cohort of patients has a good prognosis given that the diagnosis of celiac disease was made early, and treatment was instigated immediately, thus, protecting against future neurologic deterioration. Indeed, patients with celiac disease who present solely with neurologic manifestations are, on average, 10 years older than the ones who present with classic celiac disease.

Miscellaneous manifestations

Epileptic seizures presenting in the context of gluten sensitivity include patients with and without overt brain pathology who may or may not respond to antiepileptic drugs. This spectrum includes a range of interesting pathological entities, including epilepsy and cerebral calcifications, hippocampal sclerosis, and temporal lobe epilepsy as well as patients who display no pathological clues as to the cause of the epilepsy (03; 21; 51; 44).

More than half of patients with epilepsy and gluten sensitivity or celiac disease respond positively to a gluten-free diet (defined as either decreased frequency of seizures, cessation of seizures, or reduction or cessation of antiseizure medications following the introduction of a gluten-free diet). The response to a gluten-free diet could reflect the resolution or reduction of neurologic insult caused by gluten ingestion. When specifically considering patients with cerebral calcifications, a syndrome that refers to patients with focal, medically refractory epilepsy and parieto-occipital calcifications, response to antiseizure medication alone appears to be poor, with the majority (73%) being unresponsive to treatment. Response to a gluten-free diet appears to be more effective, with 53% of patients on a gluten-free diet demonstrating a good response. Patients with cerebral calcifications are often children and have primarily been reported in Italy, Argentina, and Spain. The mean age of epilepsy onset is 6 years, with the focus being primarily occipital. The most effective therapeutic intervention is the gluten-free diet. Interestingly, three prospective cohort studies appear to have demonstrated an inverse relationship between the effectiveness of a gluten-free diet and the duration of epilepsy prior to the gluten-free diet, perhaps due to permanent neurologic damage.

Myopathy cases in both adults and children with celiac disease have been described as isolated case and series reports over the last 4 decades (45; 26). The myopathy was the first manifestation of celiac disease in the majority, if not all, of the cases, and a gluten-free diet was proven to be effective in resolving the patients’ symptoms. It has been speculated that such acquired myopathies are related to a deficiency in fat-soluble vitamin D or E due to malabsorption secondary to celiac disease, which is reversed following a gluten-free diet. However, there have also been cases of inflammatory myopathy (myositis with or without inclusion bodies) associated with celiac disease; therefore, an immune-mediated mechanism to account for the myopathy has also been suggested (26).

Hadjivassiliou and colleagues presented the largest series to date of patients who presented with a clinical picture of myopathy and in whom the diagnostic work-up led to a diagnosis of gluten sensitivity (26). The mean age at onset of the myopathic symptoms was 54 years. Inflammatory myopathy was the most common finding on neuropathological examination. Not all patients had enteropathy, suggesting that gluten sensitivity without enteropathy can also be linked to myopathy. Although some patients were on immunosuppressive treatment (ie, azathioprine, methotrexate, prednisolone), some showed clinical improvement of the myopathy with just a gluten-free diet. The improvement was also associated with a reduction or normalization of the serum creatine kinase level. Myopathy, however, is a rare manifestation of gluten sensitivity.

Finally, some cases of isolated chorea have been linked to gluten sensitivity and celiac disease. Such cases also respond to a strict gluten-free diet, often with complete and permanent resolution of the chorea; however, re-exposure to gluten can cause recurrence of the chorea (62; 50).

Prognosis and complications

The prognosis of neurologic complications is very much dependent on the strictness of the gluten-free diet. Ataxia and peripheral neuropathy have been reported to stabilize with adherence to a gluten-free diet, whereas some studies have shown development or progression despite a gluten-free diet. Studies reporting progression are usually case reports or small series. Adherence to the gluten-free diet is assumed rather than demonstrated by negative serology. The systematic studies mentioned above demonstrated that patients on a gluten-free diet with elimination of anti-gliadin antibodies improve. Those not on the diet progress. Those on a gluten-free diet but with positive serology also follow a progressive course, but at a slower rate than the group not on the diet. Strict adherence to a gluten-free diet is not an easy task. This is reflected by the fact that up to 50% of patients with neurologic manifestations still have serological positivity (lower titers) when tested for anti-gliadin antibodies.

A small number of patients may have refractory celiac disease. This is certainly the case in patients with myoclonic ataxia. Refractory celiac disease, by definition, does not respond to a gluten-free diet. Immunosuppression is essential to achieve stabilization, but even then, these patients are at risk of developing enteropathy-associated lymphoma.

Meta-analysis has shown that patients with celiac disease are at a slightly increased risk of mortality and a 2- to 4-fold increased risk of non-Hodgkin lymphoma, particularly T-cell non-Hodgkin lymphoma (enteropathy-associated), dependent on the level of mucosal healing (73). An increased risk of small intestinal adenocarcinoma has also been reported, although the absolute risk of this rare cancer is low (19).

Clinical vignette

Case 1. A 45-year-old female teacher presented with gait instability and falls. She had difficulty with her work as a teacher because of an element of slurring of speech, particularly towards the end of lessons.

Neurologic examination showed mild dysarthria and gait ataxia. She was able to walk without a walking aid. MRI showed mild atrophy of the cerebellum with reduced MR spectroscopy of the vermis, less so in the right hemisphere. Extensive investigations as to the cause of the ataxia (including whole genome sequencing) were negative excepting elevated anti-gliadin antibodies. Gastroscopy and duodenal biopsy showed no evidence of enteropathy.

She started a gluten-free diet, but the anti-gliadin antibody level remained positive on repeated outpatient appointments. She was referred for further dietetic review and was seen several times. Two years later, she noticed some mild progression of her ataxia and had to give up her teaching job. At this stage, the MR spectroscopy showed a reduction of the NAA/Cr ratio when compared to the baseline scan. Her anti-gliadin antibodies remained elevated despite what she perceived as being on a strict gluten-free diet.

Further review by the dietitian identified cross contamination issues, both at home and when she was eating out. After addressing these issues, the antibodies became negative for the first time a year later. This was associated with a substantial increase in the NAA/Cr ratio from the cerebellar vermis. The patient still has a mild degree of cerebellar atrophy, which has not progressed further, and feels stable.

Case 2. A 51-year-old Italian woman reported anemia since her mid-20s, which lead to the diagnosis of celiac disease at age 32, based on elevated gliadin (IgG and IgA) and endomysial antibodies and duodenal biopsy showing subtotal villous atrophy. She immediately began following a gluten-free diet.

About 5 years later, despite adherence to the diet, she developed progressive dysarthria and ataxia. On examination, she had dysmetria, impaired rapid alternating movements, decreased pinprick perception in her feet, ankle areflexia and a severely ataxic, wide-based gait. A modified International Cooperative Ataxia Rating Scale (ICARS) score was 31/96 (without Archimedes spiral drawing).

Increased IgA antibodies to transglutaminase and glutamic acid decarboxylase were noted. Cerebral MRI showed superior vermian atrophy and EMG/nerve conduction studies showed a mild sensorimotor polyneuropathy. Prior median somatosensory evoked potentials showed a delay between the lower brainstem and cortex.

She was treated with IVIG, 2 grams/kg initially with 0.5 grams/kg given 2 weeks later. Within 1 month, she reported substantial improvements in her speech and gait. On examination, the modified ICARS score was 3/96.

She developed a rash, so IVIG was discontinued, and she received a single dose of methylprednisolone; however, 5 weeks later, she deteriorated and her ICARS score, calculated 8 weeks after her last infusion, had worsened to 17/96.

A different IVIG formulation was given, and 3 weeks later her ICARS score had improved to 3/96. She was initially stable with maintenance dosing (0.5 grams/kg/month) (66). IVIG was later again discontinued due to a rash and periodic plasmapheresis was initiated. This was discontinued after symptom worsening and poor tolerance. Her ICARS score was 22/100 when IVIG was resumed (age 44 years).

There has been some mild progression and decline over time so that she now requires a rolling walker when outdoors, and her ICARS score was 30/100 when last tested.

This case is illustrative of ataxia in a case of biopsy-confirmed celiac disease developing after initiation of a gluten-free diet. However, at the time she developed the ataxia, there was serological positivity for transglutaminase antibodies, suggesting ongoing exposure to gluten. Despite subsequent presumed strict adherence and IVIG therapy, there has been some disease progression.

Both cases demonstrate the importance of close monitoring of patients, both by a clinician and a dietitian, with regular serological testing for the relevant antibodies.

Biological basis

Etiology and pathogenesis

Postmortem data from patients with gluten ataxia demonstrate patchy loss of Purkinje cells throughout the cerebellar cortex, an end stage nonspecific finding in many cerebellar disorders. However, findings supporting an immune-mediated pathogenesis include diffuse infiltration, mainly of T-lymphocytes within the cerebellar white matter, as well as marked perivascular cuffing with inflammatory cells (32). The peripheral nervous system also shows sparse lymphocytic infiltrates, with perivascular cuffing observed in sural nerve biopsies of patients with gluten neuropathy and in dorsal root ganglia in patients with sensory neuronopathy (34). Patients with celiac disease produce an immune response to gluten involving both innate and adaptive immunity (43). Antibodies to gliadin are part of this response, and their systemic levels appear to mirror the immune reaction triggered by gluten in the intestine. There is cross-reactivity of these antibodies with antigenic epitopes on Purkinje cells as demonstrated by the fact that serum from patients with gluten ataxia and from patients with celiac disease recognize Purkinje cells of both human and rat origin (25; 07). This reactivity can also be seen using polyclonal anti-gliadin antibodies and can be eliminated by absorption with crude gliadin. However, when using sera from patients with gluten ataxia, there is evidence of additional antibodies targeting Purkinje cell epitopes as elimination of anti-gliadin antibodies alone is not sufficient to abolish such reactivity (25). There is evidence that these additional antibodies are one or more transglutaminase isoenzymes (transglutaminase 2, transglutaminase 6) (25).

Transglutaminase 2 belongs to a family of enzymes that crosslink or modify proteins. Gluten proteins (from wheat, barley, and rye), the immunological trigger of celiac disease, are glutamine-rich donor proteins amenable to deamidation. Deamidation of gluten peptides enhances binding with disease-relevant human leukocyte antigens and thereby enhances presentation, leading to the development of gluten-specific Th1-like CD4+ T cells (43). Thus, activation of transglutaminase 2 and deamidation of gluten peptides appears to be central to disease development. In genetically predisposed individuals, this is at the center of the chronic inflammatory reaction manifesting with enteropathy in the context of celiac disease. Apart from the gluten-specific T-cell response, one of the hallmarks of gluten-related disorders is an IgA autoantibody response to transglutaminase 2. Events leading to the formation of autoantibodies against transglutaminase 2, transglutaminase 3, or transglutaminase 6 are less clear. Intestinal deposits of IgA antibodies targeting transglutaminase 2 are present at all stages of celiac disease, including early (when there is no overt enteropathy) and late stages; this includes refractory celiac disease. It is important to keep in mind that B cells have roles beyond antibody production, including highly effective antigen presentation for T-cell responses. Therefore, B cells may drive clonal expansion of gluten-specific T cells, which, in turn, may support the development of B cells specific to transglutaminases. This potentially puts B cells at center stage in the pathogenesis of gluten-related disorders.

Questions also remain as to the contribution of these autoantibodies to organ-specific deficits. Anti-transglutaminase 2 antibodies have been shown to be deposited in the small bowel mucosa of patients with gluten-related disorders and may contribute to the development of the enteropathy (47). Such deposits have been found in extraintestinal tissues, such as muscle and liver, and are more relevant around brain vessels in the context of gluten ataxia (38). This finding suggests that such autoantibodies could play a role in the pathogenesis of the whole spectrum of manifestations seen in gluten-related disorders. Variations in the specificity of antibodies produced in individual patients could explain the wide spectrum of manifestations. Although transglutaminase 2 has been shown to be the autoantigen in celiac disease, the epidermal transglutaminase, transglutaminase 3, has been shown to be the autoantigen in dermatitis herpetiformis (17; 69). Antibodies against transglutaminase 6, primarily a brain-expressed transglutaminase, have been shown to be present in patients with gluten ataxia (24). Like anti–transglutaminase 2, the production of these anti–transglutaminase 3 and anti–transglutaminase 6 antibodies in patients with dermatitis herpetiformis and gluten ataxia, respectively, is gluten-dependent, which substantiates the link to a gluten-specific T-cell population. In gluten ataxia and dermatitis herpetiforme, IgA deposits of transglutaminase 6 and transglutaminase 3, respectively, seem to accumulate in the periphery of blood vessels at sites where, in health, the respective proteins are absent. It is likely that inflammation around vessels is at the heart of gluten ataxia. Indeed, perivascular cuffing with lymphocytes is a common finding in brain tissue from patients with gluten ataxia but is also seen in the peripheral nerves and muscle in patients with gluten neuropathy or myopathy (63). In most patients who are sera reactive with more than one transglutaminase isoenzyme, distinct antibody populations are responsible for such reactivity rather than this being a result of cross-reactivity with different transglutaminase isozymes. This makes shared epitopes less likely to be the cause of B-cell development to other transglutaminases and points to the possibility that transglutaminase isozymes other than transglutaminase 2 can be the primary antigen in extraintestinal manifestations.

IgA deposition in brain vessels and the pathological finding of perivascular cuffing with inflammatory cells may indicate that vasculature-centered inflammation may compromise the blood-brain barrier, allowing exposure of the CNS to pathogenic antibodies, and, therefore, triggering nervous system involvement.

It is also possible that additional factors other than the autoantibodies themselves play a role. These may affect either vascular permeability, blood-brain barrier integrity, or antigen availability. An unrelated infection or other brain insult (eg, stroke or head injury) that causes local inflammation may, in the presence of circulation-derived autoantibodies, bring about pathogenic immune complexes at the blood-brain barrier (55).

One may argue that the development and deposition of antibodies is an epiphenomenon rather than being pathogenic. One method of demonstrating the pathological effect of an antibody is the passive transfer of the disease through antibody injection into a naive animal. A common problem in such studies is the ability to demonstrate whether it is these specific antibodies or other circulating autoantibodies that cause neuronal damage. Using a mouse model, it has been shown that serum from patients with gluten ataxia, as well as clonal monovalent anti-transglutaminase immunoglobulins derived using phage display, cause ataxia when injected intraventricularly in mice (07). Therefore, these data provide evidence that anti-transglutaminase immunoglobulins (derived from patients) compromise neuronal function in selected areas of the brain once exposed to the CNS.

Epidemiology

Celiac disease occurs in about 1% of the worldwide population, although most people are undiagnosed. In the United States, the biopsy-verified prevalence is 0.7% to 0.8% and appears more common in white people than African Americans or Hispanics. Estimated prevalence is higher in studies based on serology screening (transglutaminase and endomysial antibodies) rather than those requiring confirmation through intestinal biopsy. There appears to be an increased prevalence of both diagnosed and undiagnosed disease over time.

DQ2 and DQ8 haplotypes are necessary for celiac disease, and their absence essentially excludes the diagnosis. The lifetime risk in first-degree relatives is estimated to be 4% to 17%. First-degree relatives who are homozygous for HLA DQ2 are at high risk for developing celiac disease, with 26% developing the disease during childhood.

The prevalence of non-celiac gluten sensitivity has been reported to be 6 times that of celiac disease and affects women more than men (33; 68). There is more diagnostic uncertainty in the group of patients that present to gastroenterology, although the majority of them are positive for anti-gliadin antibodies. The lack of a more specific biomarker for the gastroenterology group and the potential for overlap with other conditions, such as irritable bowel syndrome, makes estimation of the prevalence difficult. A consensus committee has established guidelines for the diagnosis of nonceliac gluten sensitivity. For patients with neurologic manifestations, gluten ataxia has been shown to account for up to 20% of all ataxias (39).

The prevalence of neurologic complications in patients with established celiac disease has been estimated to range from 10% to 22% (09). A prospective study demonstrated that up to 67% of patients with newly diagnosed celiac disease have neurologic symptoms or signs and that 46% have abnormal MR spectroscopy of the cerebellum (27). Nineteen percent have excessive white matter changes compared to 3% of the healthy population.

Prevention

Although strict adherence to a gluten-free diet is the cornerstone of therapy, existing data indicate that this is not always adequate because patients are not always strict with their diet. Inadvertent exposure to gluten is common, particularly when eating out. This is the most likely reason neurologic manifestations develop or worsen, despite strict adherence. Furthermore, the beneficial effect of a gluten-free diet takes up to a year to manifest, assuming that the patient is very strict with their diet. It would be of interest to investigate the introduction of immunomodulatory medication after the diagnosis to establish whether such medication in combination with a gluten-free diet, at least for the first year, may improve outcome.

Promising results of two phase 1 randomized controlled studies of Nexvax2®, a vaccine directed against gluten-specific CD4 T cells, lends support for further investigation of this therapeutic strategy (22).

Differential diagnosis

Associated or underlying disorders

Symptoms of celiac disease may overlap with other gastrointestinal conditions, such as irritable bowel syndrome, food intolerances (eg, lactose, fructose), microscopic colitis, or small intestinal bacterial overgrowth.

Careful history, examination, and diagnostic studies are necessary to exclude other causes of ataxia and neuropathy, such as genetic disorders (eg, spinocerebellar ataxia), toxicity (eg, alcohol-induced cerebellar ataxia), and nutritional deficiencies (eg, myeloneuropathy due to vitamin E, B12, copper or folate deficiency) (57).

Diagnostic workup

Serologic testing, particularly in patients with an autoimmune diathesis and first-degree relatives with celiac disease, should be pursued (64). The presence of IgA transglutaminase or endomysial antibodies are specific for enteropathy in celiac disease and are often not detected in patients with non-celiac gluten sensitivity. For patients with IgA deficiency, IgGs for transglutaminase, endomysial antibodies, and deamidated gliadin peptides (DGP) may be tested (64). HLA typing for DQ2 or DQ8 can also be pursued. Celiac disease is confirmed by duodenal or jejunal biopsy showing characteristic features, such as villous atrophy, crypt hyperplasia, and increased intraepithelial lymphocytosis.

In the past, gluten sensitivity was diagnosed by the detection of IgG or IgA gliadin antibodies (AGA) (67; 64). However, a study found that anti-gliadin antibodies IgG is detected in only 56.4% of gluten sensitivity cases and IgA in 7.7% (75). Therefore, resolution of symptoms on gluten-free diet and their reappearance in response to a gluten challenge are additional requirements for the diagnosis of no-celiac gluten sensitivity (Carroccio et al 2012; 74). Other possible diagnostic alternatives must be ruled out, even in those patients who fulfill diagnostic criteria of celiac disease or gluten sensitivity.

The preceding information is relevant to patients with gastrointestinal presentations. Neurology cases will, by definition, have positive serology for anti-gliadin antibodies. In addition, those with enteropathy will be positive for transglutaminase 2 and endomysial antibodies. Emerging evidence suggests that transglutaminase 6 antibodies may prove to be more specific and more sensitive to the neurologic manifestations of gluten sensitivity.

Management

Gluten-free diet. Adherence to a gluten-free diet has been reported to be efficacious in the treatment of various gluten-related neurologic disorders, including ataxia, neuropathy, progressive myoclonic ataxia, epilepsy, myopathy, stiff-person syndrome, anxiety, depression, attention deficit hyperactivity disorder, schizophrenia, and autism (15; 01; 46; 28; 37; 26; 23; 61; 42; 65; 76; 10; 04).

Immunomodulatory agents. There are case reports that support the efficacy of IVIG in ataxia and peripheral neuropathy associated with gluten-related disorder (11; 66; 12). Additionally, patients with progressive myoclonic ataxia refractory to gluten-free diet achieved clinical stability on mycophenolate, and patients with myopathy have had favorable clinical responses to various immunomodulatory agents, including azathioprine, prednisolone, and methotrexate (29). Azathioprine, mycophenolate, plasmapheresis, and steroids have shown little promise in treating gluten-related disorder neuropathy (12).

Surgery. Patients with celiac disease, epilepsy, and cerebral calcifications (CEC) have a poor clinical response to gluten-free diet (21; 72). Patients often undergo temporal lobectomy or calcification resection to achieve better seizure control (58).

Outcomes

In a small cohort study, gluten ataxia improved with gluten-free diet (28). A small prospective cohort study showed that patients who followed a gluten-free diet for 1 year had improvement in neuropathy-associated symptoms and sural sensory action potentials, as compared to those who did not follow a restricted diet (37). Elimination of anti-gliadin antibodies in the serum confirms strict adherence to a gluten-free diet. In stiff-person syndrome, gluten-free diet led to a reduction in anti-GAD antibodies (23). Due to the rarity of neurologic disorders secondary to celiac disease, randomized controlled studies are lacking, and the evidence supporting gluten-free diet efficacy is largely anecdotal with various case studies showing mixed results.

Outcomes for immunomodulatory agents and surgical interventions are solely case based.