CHAPTER 114 NEUROLOGY OF GASTROENTEROLOGY AND HEPATOLOGY
The enteric nervous system within the gut wall, replete with its neuronal networks and numerous neurotransmitters, mirrors many aspects of the central nervous system (CNS) and intimately interfaces with the CNS via the autonomic nervous system. Although seldom recognized, the number of neurons in the enteric nervous system is actually comparable to the number of neurons in the spinal cord, leading some authorities to refer to the enteric nervous system as the “little brain.”1,2 The striking similarities and intricate interaction between the neurological and gastrointestinal systems have been highlighted by the emergence of “neurogastroenterology” as a field of growing research and clinical interest.3
The boulevard connecting the two disciplines is a two-way street. Gastrointestinal manifestations of some neurological diseases are well known; the sometimes prominent gastrointestinal manifestations of Parkinson’s disease serve as a ready example.4 However, the potential presence of neurological dysfunction in the setting of gastrointestinal disease is often overlooked. Although neurological dysfunction in the setting of hepatic failure is very well known to both neurologists and gastroenterologists, the potential for neurological dysfunction in other gastrointestinal disease processes often eludes recognition. This chapter highlights several of these disease processes and also addresses some neurological aspects of hepatic failure.
GASTROINTESTINAL DISEASE
Celiac Disease
Celiac disease, also known as nontropical sprue, celiac sprue, or gluten-sensitive enteropathy, has historically been conceptualized as a disorder of the small intestine that in its adult-onset form is characterized clinically by steatorrhea, intermittent diarrhea, abdominal bloating, flatulence, malabsorption, weight loss, and aphthous stomatitis. Pathological features include villous atrophy and crypt hyperplasia. The proximal small intestine, especially the jejunum, is most prominently involved. Exposure to gluten, the protein fraction of wheat, appears to trigger an autoimmune response that in turn produces the intestinal mucosal damage characteristic of celiac disease. There are two distinct proteins in gluten (glutenin and gliadin), and it is the production of immunoglobulin A antibodies to gliadin that has been traditionally held to be the vehicle for the mucosal damage, although it has also been suggested that these antibodies, although diagnostic markers of the disease, are not crucial to its pathogenesis.5 Other similar proteins, called prolamins, present in barley, rye, and oats, may induce a similar response.
Most reports have estimated the prevalence of celiac disease at approximately 1% of the white population; one study indicated that 1 per every 120 to 300 persons in the United States and Europe is affected.6 Celiac disease appears to be most prevalent in northwestern Europe. Two peaks of clinical appearance have been identified: the first during infancy and the second between the ages of 30 and 50. A genetic component is presumably present, inasmuch as approximately 90% of individuals with diagnosed celiac disease carry the human leukocyte antigen DQ2 haplotype.7,8 Possible susceptibility loci have also been identified on chromosomes 2 and 5,9 but specific genetic mutations have not been delineated.
Not all individuals with celiac disease manifest clinical symptoms. Serological tests and even intestinal biopsy may demonstrate the presence of the pathological process in completely asymptomatic individuals. Clinicians have also realized that celiac disease is not merely a gastrointestinal disease but is, in fact, a multisystemic disorder. Dermatitis herpetiformis is characterized primarily by a blistering skin rash, but patients with this disorder also demonstrate evidence of gluten-induced intestinal pathology, the presence of gluten-induced antibodies, and a response to dietary gluten restriction, which suggests that the two disorders may be part of a pathological spectrum.10 Osteoporosis and infertility have been reported as complications of celiac disease,6 as has psychiatric dysfunction.11 Individuals with celiac disease are also at increased risk for developing certain types of cancer, including lymphoma and carcinomas of the small intestine, esophagus, and pharynx.12 A particular variety of T cell lymphoma, now designated enteropathy-associated T cell lymphoma, has been specifically associated with celiac disease. Associations with other autoimmune diseases, such as autoimmune thyroid disease, type I diabetes mellitus, Sjögren’s syndrome, and primary biliary cirrhosis have also been described.13
Neurological dysfunction has also been identified in the context of celiac disease (Table 114-1). Neurological complications may, in fact, develop in 6% to 12% of persons with celiac disease.14–16 CNS disorders have undergone the most intense investigation, but peripheral nervous system dysfunction has also been recognized.
Ataxia
Ataxia is the neurological complication that has received the most interest and attention in the setting of celiac disease. In 1966, Cooke and Smith described 16 patients with celiac disease who had also developed progressive ataxia (along with other neurological abnormalities) that was unresponsive to gluten restriction.17 It was not until 1996, however, that the idea that ataxia might be a direct consequence of gluten sensitivity was proposed by Hadjivassiliou and colleagues when they published the first of a series of reports that delineated the presence of antigliadin antibodies in individuals with sporadic adult-onset ataxia of unknown etiology and subsequently created the term gluten ataxia for the condition.7,8,18,19 They later documented the presence of antigliadin antibodies (immunoglobulin G and/or immunoglobulin A) in 41% (54) of 132 individuals with sporadic idiopathic ataxia, in comparison with only 15% (5) of 33 persons with clinically probable multiple system atrophy, 14% (8) of 59 patients with familial ataxia, and 12% (149) of 1200 normal controls.19 In the same report, a second group of 44 patients with sporadic idiopathic ataxia from another clinic was also studied; antigliadin antibodies were present in 32% (14) of this group. Other investigators have also reported elevations of antigliadin antibodies in patients with sporadic idiopathic ataxia but with lower frequencies, ranging from 11% to 27%.14,20–22 This finding has not been universal, however. Some investigators have not encountered this elevation in their study populations.23,24
Gluten ataxia, as currently recognized by its proponents, has no particularly distinguishing clinical characteristics. Gait ataxia is, by definition, present in all individuals with gluten ataxia. Limb ataxia, dysarthria and ocular signs are present in the vast majority. Cerebellar atrophy is evident on magnetic resonance imaging (MRI) in 79% of patients, whereas white matter hyperintensities are evident in only 19%.19 Evidence of classical celiac disease was found in only 24% (12) of 51 individuals with gluten ataxia who underwent gastroscopy and duodenal biopsy. In another report by the same group of investigators, 26 of 43 patients placed on a gluten-free diet were able to maintain the diet for 1 year; all 26 experienced improvement in the ataxia, regardless of whether an enteropathy was present.25
Some investigators are not convinced that gluten ataxia exists as a distinct entity.5,23 They point to the nonspecificity of antigliadin antibodies, as reflected in the fact that antigliadin antibodies are present in significant numbers of normal controls. Moreover, antigliadin antibodies have been noted to be present in 44% (23) of a group of 52 patients with Huntington’s disease, which prompted speculation that antigliadin antibodies might simply be an epiphenomenon in certain neurodegenerative diseases and that gluten ataxia may not be a distinct clinical entity.26 The absence of any clearly defined pathophysiological mechanism that might account for the cerebellar dysfunction in affected individuals has been another criticism of the concept of gluten ataxia as a distinct clinical entity. However, a chronic, immune-mediated inflammatory process has been proposed to be responsible for the cerebellar damage,27 and autopsy examination in several affected individuals has demonstrated Purkinje cell loss and lymphocytic infiltration within the cerebellum and posterior columns of the spinal cord.8
Hadjivassiliou and colleagues (2003) suggested that screening for gluten sensitivity be undertaken in all individuals who present with adult-onset ataxia without any other obvious cause.19 There is, however, no universal agreement regarding the specific screening test that should be employed. Antireticulin, antiendomysial and antitissue transglutaminase antibodies may actually be identical,5 and their presence is quite specific for celiac disease. In contrast, antigliadin antibody, especially the immunoglobulin G isotype, is nonspecific and actually present in more than 10% of healthy blood donors. This diminishes the usefulness of antigliadin antibody testing as a screening test for celiac disease, but Hadjivassiliou and colleagues believed that immunoglobulin G antigliadin antibodies are the best diagnostic marker for gluten ataxia in that they mark the whole spectrum of gluten sensitivity and not just gluten enteropathy alone.19 It thus appears that research on gluten ataxia is currently unfinished.
Epilepsy
Some investigators have suggested that the prevalence of epilepsy is increased among individuals with celiac disease. Epilepsy has been reported to be present in 3.5% to 5.5% of individuals with celiac disease.28,29 Examining the converse, Cronin and associates studied a group of patients with epilepsy and found the frequency of celiac disease, diagnosed from the presence of endomysial antibody, to be 2.3%, in comparison with 0.4% in a control group.30 A specific syndrome of epilepsy, bilateral occipital calcification, and celiac disease has also been described.31 These reports have been largely from Italian investigators, but reports from other parts of the globe have also been published.32 The mechanism for such an association is obscure. It also is not entirely clear whether a gluten-free diet improves seizure control in patients with celiac disease and epilepsy, although some investigators have reported that it does.32 An association between celiac disease and epilepsy has not been found by all investigators.33,34 Moreover, in one study, the presence of celiac disease–associated antibodies (antigliadin, antiendomysial, and antitissue transglutaminase) did not differ between 968 patients with epilepsy and a reference group of 584 individuals.35
Migraine
An association between celiac disease and migraine has also been proposed. In a report by Gabrielli and colleagues, 4.4% (4) of 90 patients with migraine were found to have serological evidence of celiac disease; subsequent jejunal biopsy confirmed the presence of celiac disease in all four of these.36 Migraine severity improved in all individuals on a gluten-free diet. Abnormalities of regional cerebral blood flow, which were noted on single photon emission computed tomographic scanning in all patients with the combination of celiac disease and migraine, also improved in all individuals. Additional case reports have described resolution of migraine with treatment of celiac disease.37 Low serotonin levels have been proposed to be a common link between the two disorders.38
Learning Disabilities
Learning disabilities, such as attention deficit/hyperactivity disorder, have been shown to occur with increased frequency in the setting of celiac disease.27 The co-occurrence of dyslexia and celiac disease has also been reported.39
Peripheral Neuropathy and Myopathy
Both peripheral nerve involvement and myopathy have been described in celiac disease. In a retrospective chart review, Vaknin and associates found that peripheral neuropathy accounted for 17% of the neurological abnormalities present in a group of patients with celiac disease.16 Another group of investigators identified chronic axonal sensorimotor neuropathy in 23% (6) of 26 patients with celiac disease and abnormalities on neurophysiological testing in 31% (8).40 These findings were present despite the fact that all 26 individuals studied had been on gluten-free diets for a median of 3 years (range, 2 to 28 years). Sural nerve biopsy, when performed, has also demonstrated the presence of axonal injury.41
Inflammatory myopathy has also been described in the setting of celiac disease and in at least one instance has been associated with vitamin E deficiency; reversal of both clinical symptoms and muscle biopsy abnormalities was documented in that individual after treatment with vitamin E and institution of a gluten-free diet.42
Other Processes
A number of other neurological manifestations of celiac disease have also been reported but less extensively evaluated. Examples include myoclonic ataxia,43 progressive multifocal leukoencephalopathy,44 chorea,45 autonomic neuropathy,46 and neuromyelitis optica.47 The significance of these reported associations is uncertain. Lymphoma within the CNS, with the immunophenotype of enteropathy-associated T cell lymphoma, has also been reported in celiac disease.12
Inflammatory Bowel Disease
Two similar but distinct disease entities, ulcerative colitis and Crohn’s disease (regional enteritis, granulomatous colitis), are the most widely recognized members of a group of conditions collectively labeled inflammatory bowel disease (IBD). An autoimmune etiology, characterized by a dysregulated mucosal immune response to antigens normally present within the intestinal lumen, is suspected in both.48,49 During the latter half of the 20th century, a significant rise in the incidence of Crohn’s disease, but not ulcerative colitis, was noted, especially in North America and northern Europe.50 The explanation for this increased incidence, which has appeared to stabilize since about 1980, is uncertain, but suspicion has fallen on a variety of environmental factors. Genetic factors also appear to play a role in the generation of the inappropriate immune response. This has been most clearly identified in Crohn’s disease, in which the NOD2/CARD15 gene, which is involved in the immune detection of bacterial products, has been identified as a susceptibility gene for Crohn’s disease.51,52 Despite many similarities, the clinical features and pathological profiles of the two conditions also demonstrate decided differences (Tables 114-2, 114-3, and 114-4). Neurological dysfunction has been described in both.
TABLE 114-3 Peripheral Nervous System Dysfunction in Inflammatory Bowel Disease
TABLE 114-4 Gastrointestinal Features of Ulcerative Colitis
Ulcerative colitis is characterized clinically by urgent, bloody diarrhea (see Table 114-4). Its course is typically marked by exacerbations and remissions. The pathological hallmark of ulcerative colitis is diffuse inflammation of the mucosa and superficial submucosa of the colon, extending a variable distance proximally from the rectum but not beyond the colon.53
The primary clinical characteristics of Crohn’s disease, in contrast, consist of abdominal pain and nonbloody, less urgent diarrhea (see Table 114-2). Weight loss is also common. Scarring and stricture formation can lead to partial intestinal obstruction, and fistula formation is also frequent. As with ulcerative colitis, the clinical course of Crohn’s disease often entails exacerbations and remissions. Although Crohn’s disease may involve virtually all levels of the gastrointestinal tract, it demonstrates a distinct predilection for the distal small intestine and proximal colon. In contrast to ulcerative colitis, gastrointestinal tract involvement in Crohn’s disease is patchy or segmental and extends deeply, often transmurally. Noncaseating granulomas may form but are not invariably present.
Extraintestinal manifestations of both Crohn’s disease and ulcerative colitis are surprisingly common, with reported frequencies ranging from approximately 25% to more than 50% of affected individuals.54–57 Some of the extraintestinal manifestations, such as involvement of joints, skin, mouth, and eyes, seem to be correlated with the presence of active colonic inflammation and are observed in both Crohn’s disease and ulcerative colitis. Other clinical processes, such as gallstones and renal calculi, are correlated more directly with small intestinal involvement and are observed primarily in Crohn’s disease.58 Primary sclerosing cholangitis is a particularly important extraintestinal complication of ulcerative colitis that may progress to hepatic failure, independent of the severity of the ulcerative colitis itself.
In comparison with many of the systemic manifestations of IBD, neurological involvement (myopathy and myasthenia gravis) occurs less frequently in both ulcerative colitis and Crohn’s disease. Lossos and colleagues (1995) reported the presence of neurological involvement in 3% (19) of 638 persons they studied with either ulcerative colitis or Crohn’s disease.59 In their comprehensive review, they described four categories of neurological involvement in Crohn’s disease and ulcerative colitis: peripheral neuropathy, myopathy or myoneural junction dysfunction, cerebrovascular disease, and myelopathy. Other groups of investigators have also documented seizures and encephalopathy in patients with IBD. Treatment for both Crohn’s disease and ulcerative colitis involves potent medications; thus, a number of neurological complications of treatment for these diseases have also been reported. Although neurological impairment often becomes evident during periods of disease activity, it can also emerge when the disease process is quiescent.
Peripheral Nervous System Involvement in Inflammatory Bowel Disease
Peripheral Neuropathic Disease
The peripheral nervous system is the dominant site of neurological dysfunction in ulcerative colitis and Crohn’s disease, accounting for 31.5% of neurologically affected patients in the experience of Lossos and colleagues.59 A rather extensive array of peripheral neuropathic processes has received recognition (see Table 114-3). Acute inflammatory demyelinating neuropathy (Guillain-Barré syndrome), axonal sensorimotor neuropathy, mononeuropathy, brachial plexopathy, mononeuritis multiplex, multiple compressive neuropathies, and cranial neuropathies have all been described in the setting of Crohn’s disease and ulcerative colitis.59–62
Gondim and associates performed a retrospective study of 18 patients with Crohn’s disease and 15 with ulcerative colitis who had developed peripheral neuropathy with no other identifiable etiology.63 The neuropathy was demyelinating in nature in slightly fewer than 30% of the patients, small- or large-fiber axonal sensory in about 30%, and large-fiber axonal sensorimotor in approximately 40%. Small-fiber axonal sensory neuropathy was more frequently present in younger patients, whereas large-fiber axonal neuropathy tended to be present in older patients. Although both axonal and demyelinating neuropathies often responded to immunotherapy, the response was more consistent and robust in patients with demyelinating neuropathy.
Two specific and unusual processes consisting of cranial nerve involvement have been noted in individuals with IBD. Melkersson-Rosenthal syndrome is characterized by the clinical constellation of recurrent facial nerve palsy, intermittent orofacial swelling, and fissuring of the tongue (lingua plicata). Not only has this syndrome been noted in the presence of Crohn’s disease, but also its pathological hallmark—noncaseating granuloma formation—is observed in Crohn’s disease. This has led some to suggest that Crohn’s disease and Melkersson-Rosenthal syndrome are actually part of the same pathological spectrum.64 Acute sensorineural hearing loss has been described, primarily in persons with ulcerative colitis.65–68 The presumption of an autoimmune basis for the hearing loss is based in part on reported responses to steroid administration. Chronic, subclinical hearing loss, discovered on audiometric screening, has also been documented in ulcerative colitis.69
Muscle and Myoneural Junction Disease
Myopathy has been described in the setting of both ulcerative colitis and Crohn’s disease, but it is present more frequently in Crohn’s disease. It accounted for 16% of the cases of neurological dysfunction in the series compiled by Lossos and colleagues.59 Dermatomyositis, polymyositis, rimmed vacuole myopathy, and granulomatous myositis have all been noted in this setting. As with peripheral nerve involvement, an autoimmune basis is presumed to be present. Localized myositis involving the gastrocnemius muscles has also been reported in Crohn’s disease and is known as gastrocnemius myalgia syndrome.70 In approximately 50% of cases, the appearance of myopathic pathology seems to be correlated with disease activity in the bowel. Myasthenia gravis, another autoimmune disorder, has also been reported in the setting of both Crohn’s disease and ulcerative colitis, although reports of this association are quite sparse.
Central Nervous System Involvement in Inflammatory Bowel Disease
CNS involvement in inflammatory bowel disease can assume many guises. Cerebrum, brainstem, and spinal cord can all be involved (Table 114-5).
TABLE 114-5 Central Nervous System Involvement in Inflammatory Bowel Disease
Myelopathic and Motor Neuron Disease
Chronic, slowly progressive myelopathy is yet another neurological manifestation of inflammatory bowel disease, accounting for 26% of the patients with neurological involvement in the series of Lossos and colleagues.59 Most of these individuals were suffering from Crohn’s disease rather than ulcerative colitis. An inflammatory basis was suspected, and oligoclonal banding was noted in one patient. Other investigators have suggested a possible association between ulcerative colitis and multiple sclerosis,71,72 and it has been reported that the incidence of multiple sclerosis is threefold greater in persons with ulcerative colitis than in the general population.71 Transverse myelitis has been reported in an individual with ulcerative colitis who also was found to have anti–Jo-1 antibody (antisynthetase) syndrome.73 Whether cases such as these are causal or coincidental is uncertain. A more definite causal relationship can be drawn with the occurrence of spinal empyema in the setting of Crohn’s disease secondary to fistula formation.74
There has been one case report of the coexistence of ulcerative colitis with motor neuron disease, although the issue of cause versus coincidence cannot be escaped in this instance either.75
Cerebrovascular Disease
Vascular complications are well-documented extraintestinal manifestations of inflammatory bowel disease. In a massive undertaking, Talbot and coworkers reviewed the records of 7199 patients with either Crohn’s disease or ulcerative colitis and noted the presence of vascular complications in 1.3%.76 Deep venous thrombosis and pulmonary embolus were the most common sites of involvement, whereas cerebrovascular events accounted for only 9.8% of the total vascular complications (in 9 of 92 patients). Vascular complications occur more frequently in ulcerative colitis than in Crohn’s disease; the reason for this difference is uncertain.77 Hypercoagulability has been presumed to be responsible for the thromboembolic events, and elevations of factors V and VIII and fibrinogen levels, along with decreased antithrombin III levels, have been noted.78 Anticardiolipin antibodies, thrombopoietin, and homocysteine levels have also been shown to be elevated in individuals with ulcerative colitis and Crohn’s disease, but the elevations are not correlated with increased risk of thromboembolic events.79–81 In fact, no single or specific abnormality has been identified as the hypercoagulable culprit, and no consistent abnormality of the routinely recognized coagulopathy susceptibility factors has been discovered.82
A variety of cerebrovascular events has been reported in ulcerative colitis and Crohn’s disease. Both large artery and lacunar infarcts have been described.83,84 Cerebral vasculitis has been identified in both Crohn’s disease and ulcerative colitis, which led to the proposal that an autoimmune basis is responsible for the cerebrovascular events. However, responses to both immunosuppressive therapy (corticosteroids and azathioprine) and anticoagulation have been reported, which suggests that a single explanation is improbable and that both hypercoagulable and autoimmune processes may have roles in different individuals.85–87 Both dural and cortical venous sinus thrombosis have been the subjects of numerous case reports in individuals with inflammatory bowel disease, predominantly but not exclusively in ulcerative colitis.77,88 Such events have occurred both during active exacerbations and during periods of disease quiescence. As with other cerebrovascular events, the mechanism of the coagulopathy is not entirely clear.
Seizures and Encephalopathy
Seizures are an infrequently reported neurological complication of inflammatory bowel disease. They may occur as a complication of the surgical management of these diseases, presumably precipitated by factors such as fluid overload, electrolyte imbalance, hypoxia, and steroid administration or withdrawal.89 Seizures have also been reported as a complication of cyclosporine treatment in an individual with Crohn’s disease.90 Status epilepticus, with long-term sequelae, has been described in the setting of ulcerative colitis; the genesis for this has not been clearly delineated.91
Diffuse encephalopathy with altered consciousness may also develop in individuals with ulcerative colitis or Crohn’s disease. This may occur in the context of cerebral vasculitis85 but also as a consequence of nutritional deficiencies. Both Wernicke’s encephalopathy and possible selenium-induced encephalopathy have been described in individuals with Crohn’s disease receiving total parenteral nutrition.92,93 Wernicke’s encephalopathy has also been reported in an individual with clinically inactive Crohn’s disease.94
