Article Text


Neurology of Whipple's disease
  1. Queen Elizabeth Neuroscience Centre, Edgbaston
  2. Birmingham B15 2TH, UK
  3. email Milne.Anderson{at}

    Statistics from

    In 1907, George Hoyt Whipple, then an instructor in pathology at Johns Hopkins University, published a case report of a 36 year old physician who had been domiciled in Constantinople.1 He developed recurring arthropathy, weight loss, and steatorrhoea, became worse, and died. At necropsy the intestine and mesenteric lymph glands were infiltrated by mononuclear and polynuclear giant cells and deposits of fat and fatty acids. Whipple described the large “foamy” mononuclear cells in the intestinal mucosa which were later shown to contain periodic acid Schiff (PAS) positive material. The aetiology of the condition was not established and he wrote “Gland tissue treated by the Levaditi method shows great numbers of a peculiar rod-shaped organism (?) which does not stain by the aniline dyes . . .” Whether this is the aetiological factor in this disease cannot be determined from this case”. A prescient observation. He suggested intestinal lipodystrophy as a name for the disease.

    General features

    Whipple's disease is rare—since its first description less than 1000 cases have been reported in the literature which is certainly an underestimate of all cases. Men are affected much more often than women (over 80%) and the mean age of onset approaches 50 although the age range extends from childhood to senility. Cases which have been reported have originated from Europe and North America in the main. Whether this is a true reflection of its aetiology or relates to technical difficulties in reaching a diagnosis is not clear. The natural habitat of the organism is not known, nor are the mechanisms by which infection takes place. The number of cases described in reports is too low to determine whether there is a racial susceptibility. The most common clinical presentation is a malabsorption syndrome with diarrhoea, abdominal pain, weight loss, generalised wasting, variable fever, and lymphadenopathy. Many cases are associated with longstanding and relapsing and remitting arthropathies. The onset is insidious and the symptoms and signs are commonly atypical. The CNS, lungs, heart, eyes, and skin may be involved and the disease may first manifest in these organs. Some present with a pyrexia of unknown origin, lymphadenopathy, and a sarcoidosis-like picture.2 Perhaps as many as 15% of patients do not have gastrointestinal symptoms throughout their illness3 and jejunal biopsy may be normal.4 It is therefore not surprising that in many of these atypical cases the diagnosis may not be made for some time and unfortunately, it may be missed until it is demonstrated at necropsy.

    Tropheryma whippelii

    The aetiology and pathogenesis of Whipple's disease have remained elusive until recently although a bacterial cause has been postulated since the original description. Attempts to culture the organism over the years have been negatory, yet accumulated circumstantial evidence has led to the identification of a bacillus with unique characteristics as the cause. It has been consistently recognised in affected tissue using electron microscopy,5-7 lying free or degraded to varying degrees within macrophages. Patients with Whipple's disease who are treated with antibiotics improve and improvement is accompanied by disappearance of bacilli: recurrence of disease is accompanied by re-emergence of bacteria.8 9 The morphology of the organism has been described.10 It is a weakly gram positive rod shaped bacillus which is not acid fast. It is 1–2 μm in length and has a thick wall, the inner layer of which stains with PAS dyes and it is this characteristic which accounts for the brightly staining PAS positive macrophages which are seen in biopsy material and which contain bacillary debris. It has resisted attempts to be cultured. In 1997 reproduction of the organism in tissue culture was reported by Schoeden et al 11who used interleukin-4 to deactivate macrophages. The difficulties experienced in isolating the organism stimulated the application of molecular genetic techniques to the search and led to the identification of a single 16S rRNA gene sequence from tissue derived from small bowel biopsy of patients with Whipple's disease.12 13 Using the polymerase chain reaction (PCR), positive results have been obtained from other tissues including the heart,14 vitreous fluid,15 peripheral blood cells,16 17 and pleural effusion cells18 It has proved particularly difficult to identify sequences in brain tissue19 yet a recent report gave a high yield for PCR on CSF.20 Analysis of the 16S rRNA gene sequence found a phylogenetic association with the Actinomycetesand Whipple's bacillus is thought to be a novel actinomycete. It has been given the name Tropheryma whippelii.Whether the same bacterium causes all forms of Whipple's disease and its multisystem manifestations remains to be determined.


    Because it has been so difficult to isolate the organism associated with Whipple's disease, little is known of its occurrence in nature, the method of its transmission to humans, and the mechanisms of production of disease once infection takes place. Humans are the only known host for the disease. There is no evidence for person to person transmission and it does not seem to occur in clusters. Maiwaldet al 21 have recently detected DNA specific for Whipple's bacillus in sewage water in Germany and they argue for an environmental source of infection. Because of the gastrointestinal location of the disease it is inferred that infection takes place by ingestion of the organism. From there it probably disseminates through the body via lymphatics and bloodstream and spreads infection to other organs. The brain is a favoured site but the mechanisms by which the organism breaches the blood-brain barrier are not known, neither is the role of host immunity. Whether the clinical manifestations of the disease result from direct bacterial invasion, or from a bacterially provoked inflammatory response, is not clear.


    Precedence in describing changes in the CNS in Whipple's disease is given to Sieracki22 and Sieracki et al 23 who were closely followed by Lampertet al 24 and Badenochet al.25 Since then numerous descriptions of pathological changes in the CNS have appeared26-28 and in 1977 Romanul et al 29 published details of a case in which the disease seemed to be confined to the CNS. It is likely that cases of Whipple's disease with neurological involvement had been described before but it was not recognised in these cases that the underlying disease was Whipple's. Also, the significance of neurological symptoms in established cases was not recognised. The gross pathological features of Whipple's disease in the CNS are generalised cerebral atrophy and small chalky nodules or granulomas up to 2 mm in diameter scattered diffusely in grey matter of the cerebral and cerebellar cortex and in subependymal grey matter around the ventricles and the aqueduct. The changes are focal and one area of the brain may be normal whereas an adjacent area may show florid abnormality. Microscopically, those granulomas have been shown to contain strongly positive PAS staining macrophages surrounded by large reactive astrocytes. With more widespread disease PAS positive cells infiltrate white matter and may burst through into the subarachnoid space and be associated with death of neurons, formation of vacuoles, and demyelination. Debris of bacilli may be found in the PAS positive material. Microinfarcts have been described, perhaps caused by emboli: vegetations on the heart valves have been found in a substantial proportion of cases at necropsy.30 In the areas of the brain with the heaviest PAS positive staining, bacilli and bacillary debris may be seen. Electron microscopic appearances were described in 196927 31 and corresponded to that of the bacilli seen in intestinal Whipple's disease. The electron microscopic appearance of the Whipple's bacillus is considered to be quite specific.32


    Neurological aspects

    Because of the rarity of Whipple's disease it is difficult to know just how often the CNS is affected. Reviews report figures which vary between 5% to just over 40%.2 8 9 33 34 Perhaps in as many as 5% of all cases of Whipple's disease, the presentation is neurological and the disease remains confined to the nervous system for the most part of the evolution of the disease.29 35-41 In several of these cases there has been no evidence of clinical involvement of the gastrointestinal system which makes diagnosis particularly problematical. Irrespective of disease elsewhere in the body, the neurological manifestations of Whipple's disease are now sufficiently well described to enable us to recognise some patterns which point to the diagnosis. Dementia, disturbances of ocular movement, abnormal involuntary movements, particularly myoclonus, and deranged function of the hypothalamus are most often found. Epilepsy, focal cerebral signs, ataxia, and meningitic features may also be present. It is not usual to get involvement of the spinal cord or of muscle or peripheral nerve although myelopathy has been described.42 Headache is a very common symptom.

    There is nothing remarkable about the mental changes which accompany Whipple's disease (table). They constitute the most often encountered neurological abnormalities in just over half of all cases,2 34 43 and coexist with other CNS syndromes. Depression, cognitive decline, confusion, and behavioural and personality change and memory loss may progress insidiously and in the beginning may be dismissed as changes of aging or even of the onset of Alzheimer's disease. Disturbances of ocular movement are the next most common finding. Almost always the ophthalmoplegia is of supranuclear gaze palsy type with involvement of vertical rather than horizontal movement. Pure oculomotor palsies are rare. Internuclear ophthalmoplegia may occur.

    Movement disorders are the next most common and may involve ocular movement. Oculomasticatory myorhythmia and oculofacial-skeletal myorhythmia occur and are said to be pathognomonic for Whipple's disease—Louis et al 34 claim that these findings have not been documented in cases other than Whipple's disease and consider them to be diagnostic findings. Oculomasticatory myorhythmia is characterised by pendular convergent divergent oscillations of the eyes, synchronous with involuntary rhythmic contraction of the muscles of mastication at a rate of approximately one per second.43 44 Oculofacial-skeletal myorhythmia is a more widespread expression of the same disorder with synchronous movements at the same rate of muscles of the extremities and face.33 45 Myoclonus is well recognised and has been described affecting limbs alone, and the facial muscles.46-48 Ataxia has been recognised in some cases. The triad of dementia, ophthalmoplegia, and myoclonus occurs in about 10% of cases and is highly suggestive of Whipple's disease.

    Hypothalamic derangement causing sleep disruption and excess, polydipsia, and hyperphagia has been described.39 49 50It will be recalled that the pathology in Whipple's disease of the CNS may be patchy throughout the cerebral hemispheres and that microembolism may take place. Consequently signs and symptoms focal to the area of the affected brain may be found, hence dysphasia, seizures, cortical blindness, cranial nerve disturbances including trigeminal neuralgia, brainstem, and cerebellar signs have been recorded.

    The eye is also a site of Whipple's disease.15 51 52Uveitis, retinitis, vitritis, keratitis, optic neuritis, and papilloedema may be found. The clinical picture may easily mimic sarcoidosis and forms of cerebral vasculitis. According to Louiset al 34 80% of the patients in their review had clinical evidence of systemic disease and this encompassed migratory arthralgias and polyarthritis, unexplained weight loss, chronic diarrhoea, abdominal pain, steatorrhoea, abdominal distension, pyrexia of unknown origin, lymphadenopathy, nights sweats, malaise, and uveitis. Such a constellation of symptomatology invites a wide differential diagnosis. The same authors proposed diagnostic guidelines for establishing definite, and possible, Whipple's disease of the CNS. It is important to remember that the progression of Whipple's disease is naturally slow and insidious although rapid deterioration may take place. After a lingering prelude of arthralgia and fatigue, weight loss, fever and gastrointestinal complaints follow to merge into cachexia, malnutrition, and death, the whole evolution taking years.


    Routine laboratory studies are commonly abnormal in a non-specific way. The erythrocye sedimentation rate may be raised, anaemia and hypoalbumenaemia from steatorrhoea may occur. Liver function tests may be deranged. Routine examination of CSF also discloses non-specific results—protein may be raised, there may be a pleocytosis, or the results may be normal. PAS positive staining cells have been found in centrifuged deposits.53 If the disease is suspected then duodenal biopsy provides the answer in most cases by demonstrating PAS positive foamy macrophages throughout the lamina propria32and the characteristic morphology of the bacillus may be evident on electron microscopy. However, not all cases have small bowel infiltration. The application of these techniques to tissue obtained from sites clinically affected by disease may be helpful but brain biopsy has not been very successful, probably because of the patchy nature of cerebral infiltration. Recently diagnostic material was obtained from stereotactic biopsy of a lesion in the putamen.54

    The application of a PCR assay against Tropheryma whippelii has transformed the ease of diagnosis. As reported above, positive results have been obtained from several tissues12-18 and from CSF20 and PCR is more sensitive and specific than other techniques.55 PCR has been shown to be positive in biopsy normal specimens of duodenum.15 It is now the diagnostic method of choice although its limitations have not yet been defined.

    Brain imaging techniques and electroencephalography show abnormalities which are not diagnostic. The EEG has shown slow wave activity.56 CT and MR images have been normal, have shown atrophic changes, mass lesions with contrast enhancement, white matter non-space occupying high signal areas, ring enhancing lesions, and hydrocephalus.37 38 40 41 43 57-66 On clinical grounds and after imaging, it can be appreciated that the differential diagnosis of Whipple's disease in the CNS encompasses a large slice of neurology. Several forms of encephalopathy and some of the more indolent encephalitides, demyelination, and the range of CNS vasculitides have many features in common with neurological Whipple's disease. Granulomatous disease including sarcoid and more chronic CNS infections such as tuberculosis, particularly if atypical bacteria are implicated, and AIDS is associated can cause diagnostic confusion. The early stages of cerebral degeneration such as Alzheimer's disease, and if abnormal movements are present, Creutzfeld-Jacob disease can be similar. Therefore it is necessary for the clinician to retain a high index of suspicion that Whipple's disease may be the cause of these various syndromes.


    Antibiotics have been used to treat Whipple's disease as they have become available. Results have been variable. It is difficult to reverse established neurological defects and there is a tendency for the disease to relapse once antibiotics have been withdrawn and indeed, in some cases, deterioration has taken place while treatment has continued. Neurological complications are likely to be the cause of relapse in the first place. When relapse occurs it is more difficult to treat. Furthermore, neurological relapse often takes place some time after treatment has been stopped.8 9 67 Experience suggests that relapse may occur more often in those patients treated with single drug regimes and with antibiotics which do not penetrate the blood-brain barrier. The combination of trimethoprim-sulfamethoxazole (cotrimoxazole) has been reported to be effective68-70 but some treatment failures have occurred with this regime.47 63 67 Not surprisingly, when there is so little concentrated experience of the condition in any one centre, no consensus of opinion has been reached regarding the optimal drug regime for the treatment of Whipple's disease. Most agree that initial treatment with a combination of parenteral penicillin and streptomycin for at least 14 days is appropriate, thereafter cotrimoxazole orally 3 times a day for at least one and probably for 2 years. Others would recommend following on after penicillin/streptomycin with third generation cephalosporins, and this is my favoured option—ceftriaxone parenterally for at least a month to be followed by 2 years of oral cefixime. Chloramphenicol has been used successfully. Experience with rifamicin and the macrolides is too small to allow conclusions to be drawn. There does not seem to be any place for the use of steroids.71 Symptomatic treatment is given for concurrent complications. Anticonvulsant drugs are given for seizures: valproate, clonezepam and piracetam have been tried for myoclonus and abnormal movements.

    It is important that the patient is closely monitored throughout the duration of treatment. Serial brain imaging can demonstrate improvement or deterioration in existing lesions and detect the appearance of new ones. In the past it was usual to monitor progress by serial small bowel biopsy and to gauge success by demonstrating resolution of the histological changes and eradication of the organism from the specimens. Unfortunately this did not always guarantee a successful outcome—recurrence would take place in at least a third of cases.55 Also histological changes do not always correlate with clinical improvement or with molecular biological test results.72 PCR is now recognised to be the best tool for monitoring progress55 and in cases of neurological Whipple's disease it is prudent to check the CSF as well as bowel for negative results before discontinuing antibiotic treatment. After cessation of antibiotic treatment further PCR checks of CSF should take place at intervals determined by the clinical progress of the patient and the acumen of the physician.

    Relapse after treatment of Whipple's disease is common and CNS involvement carries the highest relapse rate which may take place some years after antibiotic treatment has ceased.9 It is said that those cases with ophthalmoplegia and mental change respond best to antibiotic treatment, whereas those with more obvious structural lesions such as granulomas, infarcts, and atrophic change do less well. Long term follow up with repeated PCR tissue analysis is necessary for such cases. At this stage it is not possible to say if repeated CSF analysis is sufficient—probably not.

    Whipple's disease is now established as one of the more esoteric of the infectious diseases that affect the nervous system. Much remains unknown about its pathogenesis and how it interacts with host immune mechanisms. Great strides have taken place in the past few years with identification of the causal organism, Tropheryma whippelii, and the development of a highly specific diagnostic PCR which seems to be effective when applied to CSF. Treatment is possible with antibiotics and if the diagnosis is made early enough cure can be effected. It is important that neurologists remain alert to Whipple's disease as a potential differential diagnosis in a wide range of CNS disorders—encaphalopathies, chronic meningitides, cerebral vasculitides and granulomas, incipient dementias and the range of focal cerebral disturbances associated with solid and cavitating lesions on CT and MRI of the brain (much like AIDS). When there is already evidence of gastrointestinal disorder the diagnosis is perhaps a little easier but it is important to remember that neurological Whipple's disease can occur as the presenting manifestation of the disease; hence the very wide differential diagnosis. Perhaps in future a request for “PCR for T whippelii” will become part of the laboratory screening investigation for many of these conditions?


    View Abstract

    Request permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.