Abbreviations used in this article: CSF, cerebrospinal fluid, CT, computed tomography, ESR, erythrocyte sedimentation rate, MRI, magnetic resonance imaging, PAS, periodic acid Schiff, PCR, polymerase chain reaction, TMP-SMX, trimethoprim-sulfamethoxazole
Whipple disease is a rare multisystemic infectious disease caused by Tropheryma whippelii. This bacteria is probably a commensal organism of the gastrointestinal tract but may cause infection if underlying immunologic abnormalities are present, especially of macrophages (6,30,34). The disease is multifocal but usually affects the gastrointestinal tract, and duodenal biopsy remains the standard diagnostic procedure. T. whippelii cannot be cultured by traditional methods (38,39,46), so the diagnosis is based on typical histologic lesions and identification of the causative organism by polymerase chain reaction (PCR). The disease can be cured with prolonged antibiotic treatment.
Involvement of the central nervous system is a classical feature of Whipple disease (20%–40%) (17,31,52) but clinical symptoms usually develop in the later stage of the disease. Neurologic involvement has been reported particularly in patients who previously received antibiotic treatments that could not cross the blood-brain barrier. We conducted the present study to describe neurologic aspects of Whipple disease, based on the analysis of 12 original cases and a review of the published literature.
Patients and Methods
All patients included in the study had Whipple disease diagnosed on the basis of histologic and/or PCR assay data. The cases of 12 patients with Whipple disease with involvement of the central and/or peripheral nervous system were analyzed thoroughly. The cases were reported from 1982 to 2000 by physicians from the internal medicine and neurology departments of various French hospitals (Lyon, Grenoble, Saint-Etienne, Paris, Limoges, Angoulême, Roanne, Amiens, and Saint-Quentin).
We performed a MEDLINE (National Library of Medicine, Bethesda, MD) search of the literature from 1966 to 2000 to compare our cases with those previously published. The following terms were used both alone and in association:Whipple disease, Tropheryma whippelii, CNS, neurological, brain, cerebral, dementia, OMM, CSF, psychiatric, treatment, and antibiotic. We also researched the extensive bibliography obtained from the references of articles found, which included cases from 1962.
As we were mainly interested in the frequency of presentation of neurologic signs, we excluded patients in whom neurologic involvement of Whipple disease had been discovered postmortem but who had not shown any neurologic signs during their life. We also excluded patients with isolated posterior ocular involvement such as posterior uveitis, retinitis, or optic neuritis. We identified 122 cases published in the literature and excluded 43 cases that were in fact neurologic relapses of Whipple disease (see Discussion section).
We divided neurologic involvement into 3 categories: encephalic involvement, cord involvement, and peripheral involvement, and each category was analyzed separately. Morphologic examinations were thoroughly reported. Previous, contemporary, or subsequent extraneurologic manifestations were systematically noted. The different treatments used in the case of neurologic symptoms and the response to treatment were also recorded.
The following illustrative patient reports demonstrate 6 important points. Further data regarding our 12 cases are shown in Table 1.
Mrs. M. was born in 1955. Her medical history began in 1975, with the sudden onset of left hemiparesis sparing the face. Neither sensitivity disorders nor Babinski sign was noticed. Examination of the ocular fundus was normal. The erythrocyte sedimentation rate (ESR) was 11 mm/h. Cerebrospinal fluid (CSF) was normal. Computed tomography (CT) of the brain showed an expansive lesion in the posterior part of the right frontal lobe. A neurosurgical resection of the tumor was performed. The histologic analysis showed a necrotic lesion with no evidence of malignancy but with inflammatory granulomatous tissue infiltrated by numerous cells. The macrophages presented only a few periodic acid Schiff (PAS)- positive granules. In August 1998, the patient presented with agitation, behavior disorders, and confusion associated with clonic movements on the right side and decreased bilateral visual acuity. Physical examination revealed bilateral upper motor neuron disorder. Biologic investigations showed a hemoglobin of 107 g/L and an ESR of 25 mm/h. The brain CT scan revealed prominent edema of the left hemisphere associated with a hypodense lesion in the left occipital lobe with a mass effect (Figure 1). The patient was transferred to the neurosurgery department in September 1998 with total dementia with mutism. Stereotactic biopsy of the left occipital lobe was performed. The histology was similar to that observed in 1975. The lesions involved both white and gray matter, mimicking encephalitis. There were also associated perivascular and intraparenchymatous macrophages exhibiting PAS-positive inclusions suggestive of Whipple disease (Figure 2). Analysis of the CSF did not show any cell with PAS-positive cytoplasm. The specific search for T. whippelii with DNA extraction and specific PCR controlled with hybridization in the CSF was negative. At the end of October 1998, because of delirium, a gastroscopy was performed. The gastroscopy did not reveal any macroscopic abnormality. Duodenal and jejunal biopsies were normal. Treatment with trimethoprim-sulfamethoxazole (TMP-SMX) 320 + 1,600 mg/d was introduced and significant neurologic improvement was noticed. The deficit in the right part of the body decreased, delirium disappeared, and the patient was able to talk and perform daily activities again. A CT scan of the brain was performed for follow-up in December 1998: the hypodense edematous lesions had clearly shrunk, especially in the left part of the brain (Figure 3).
Mrs. M. was seen again in January 2000. Her neurologic condition was perfectly stable, as was the brain imaging. The treatment with TMP-SMX had been stopped more than a year after introduction. The patient complained mainly of persistent bilateral blindness and of pain in the left part of her body that was successfully relieved with gabapentin. She did not develop epilepsy.
Comment: Whipple disease can present like a brain tumor without any extraneurologic signs.
Mr. D. was referred to the neurology department in March 1998, at the age of 67 years. His medical history included migratory recurring polyarthritis episodes in 1986 and 1988, which had been diagnosed as seronegative rheumatoid arthritis. He had been treated successively with synthetic antimalarials, gold salts, and then methotrexate. Mr. D. had complained of gait disturbances since December 1997. Physical examination revealed left spasticity without any motor deficit. Magnetic resonance imaging (MRI) of the brain showed high signals in the right middle cerebral peduncle and in the right temporal horn associated with cortical atrophy.
Mr. D. was admitted to hospital in March 1998 because he had fainted on exertion. He suffered from headaches, memory impairment, and urinary incontinence in addition to the previous gait disturbances. He also complained of asthenia, severe weakness, and depression, the onset of which coincided with the beginning of the neurologic disorders. Physical examination revealed fever spikes (38 °C) in the morning and spastic paresis of the extremities. ESR was 80 mm/h and the C-reactive protein level was 54 mg/L. Lumbar puncture showed lymphocytic and histiocytic reaction of the CSF (5/mm3), associated with high protein levels (1.95 g/L) and normal glucose levels. On the brain MRI, high signals could be seen in the right cerebral peduncle and in the right temporal white matter, with several diffuse high signals in both hemispheres. MRI of the thoracic and lumbar cord (Figure 4) showed high signals in the white matter.
The esophago-gastro-duodenoscopy performed at the end of April 1998 was normal. Nevertheless, histologic examination of the duodenal biopsies revealed infiltration of the lamina propria with histiocytic cells with a round nucleus and a fine granular cytoplasm. These granules were positive when stained with PAS reagent. The PCR analysis showed the presence of DNA from T. whippelii. The PCR assay performed on the CSF was negative. There were no gastrointestinal signs, and the patient did not complain of arthralgia. The corticosteroids that had been previously administered without any response were stopped and replaced by an antibiotic treatment. Amoxicillin (3 g/d) combined with amikacin (750 mg/d) was infused intravenously for the first 6 days. Then TMP-SMX (320 + 1,600 mg/d) was prescribed on discharge. The clinical picture was unchanged in December 1999. Brain and cord MRI with FLAIR (Fluid Attenuation Inversion Recuperation) and T2-weighted images showed remission of the high signals that initially involved the white matter of the brain and the spinal cord (Figure 5). TMP-SMX was continued at the same dosage.
Comment: Whipple disease, although rare, must be considered in a case of clinical and radiologic (MRI) inflammatory spinal cord involvement.
In January 1985, Mr. B., 33 years old, was admitted to hospital because of loss of consciousness and subsequent cranial trauma. He also presented with sudden onset of confusion. He had been followed for fever spikes with a temperature of 40 °C of unknown origin since 1982. These fever spikes lasted 3–4 days and occurred every 6–8 weeks. They responded to aspirin, and were associated with sweating, diffuse joint pain, tremor, and tendency to hypersomnia. Every spike ended with hypothermia (36 °C). No other functional sign was observed and the general condition remained stable. On examination he had splenomegaly and inguinal lymphadenopathy. Additional tests showed a white blood count of 10.7 × 109/L with 77% neutrophils, an ESR of 41 mm/h, with no other biologic abnormality. An abdominal CT scan showed right lateral aortic and iliac lymphadenopathy. This led to laparotomy in November 1984 with pathologic examination that revealed only nonspecific inflammatory reaction of the lymph nodes. Liver and bone marrow biopsies were normal. The patient’s neurologic history began in January 1985 with a general seizure. Time and space disorientation persisted, as well as mainly anterograde amnesia. The examination showed paralysis of the vertical gaze with limitation of voluntary and automatic upward gaze associated with vertical nystagmus. Paralysis of convergence was also noticed. After 1 month, neurologic presentation worsened with signs of hypothalamic involvement including diurnal and nocturnal polyphagia, polyuria, and polydipsia, tendency to hypothermia (35.5 °C), and reversal of the sleep rhythm associated with confusion, apathy, and personality disorders. Lumbar puncture, brain CT, and MRI were initially normal. Repeated lumbar puncture showed chronic meningitis with 20/mm3 PAS-negative white blood cells. Acute phase response (ESR: 50 mm/h) and arthralgia persisted but no gastrointestinal sign was observed. Small bowel biopsies were normal with PAS-negative examination.
The lymph node biopsies of 1984 and the liver biopsy performed because of minor abnormalities were analyzed once again. This new analysis showed granules with histiocytes, whose cytoplasm was packed with granular PAS-positive material. The diagnosis of Whipple disease could thus be established. Antibiotic treatment with rifampin (1,200 mg/d) and TMP-SMX (320 + 1,600 mg/d) was administered in association with phenobarbital (150 mg/d). There was little improvement in the signs despite treatment in July 1985, although memory impairment seemed to have lessened. After a year of antibiotic treatment including chloramphenicol (3 g/d), rifampin (1,200 mg/d), and ampicillin (3 g/d), fugue states and uncontrolled polydipsia developed. Lumbar puncture performed in July 1986 showed 3 PAS-positive macrophages. The patient died in November 1986 in the context of metabolic disorders with hyponatremia (Na+ = 128 mM/L).
Comment: Considering Whipple disease in a patient presenting with encephalopathy with fever, associated arthralgia and lymphadenopathy can be important.
Mr. Z., 40 years old, was admitted to hospital in April 1987 as a result of a 15-kg weight loss (52 kg for 1.70 m) in a few months and a long-standing fever (39 °C). His history included intermittent arthralgia of the wrists and ankles for 3 years. Axillary, inguinal, and cervical lymphadenopathy were noticed on examination; ESR was 72 mm/h. Blood tests revealed a hemoglobin level of 105 g/L, a white blood cell count of 13.2 × 109/L with 83% neutrophils, an albumin level of 32.8 g/L and an anicteric cholestasis (alkaline phosphatase = 500 UI/L [normal = 30–120 UI/L]). Histologic examination of 1 lymph node suggested non-Hodgkins lymphoma. The patient then developed behavioral disorders, impairment of recent memory, and psychomotor retardation. Dilatation of the ventricles was observed on the brain CT scan. Ventriculography clearly showed stenosis of the aqueduct of Sylvius, leading to ventriculostomy of the floor of the third ventricle. Surgical intervention improved behavioral disorders and speech disturbances disappeared, but memory impairment persisted. At the beginning of May 1987, the patient presented with seizures followed by coma. CT scan of the brain showed major ventricular dilatation, and ventriculography showed another stenosis of the aqueduct of Sylvius. Another ventriculostomy was then performed. In August 1987, the patient presented with Korsakoff syndrome and intracranial hypertension. As the ventriculostomy did not seem to be effective, a ventriculoperitoneal shunt was performed and produced good results regarding the neurologic symptoms. The neurologic examination was nearly normal. In April 1988, the patient was admitted for weight loss, fever, polyadenopathy, intellectual and motor slowing, and hyponatremia. Exploratory laparotomy was performed in August 1988 and showed multiple retroduodenal, pancreatic, and mesenteric lymphadenopathy associated with hepatomegaly and splenomegaly. Histologic analyses of nodal and hepatic lesions, showing PAS-positive foamy histiocytes, were strongly suggestive of Whipple disease. The esophago-gastroduodenoscopy with duodenal biopsies also showed numerous macrophages that were PAS-positive on pathologic examination. Antibiotic treatment with TMP-SMX (320 + 1,600 mg/d) was started on August 1988. At the beginning of September 1988, the intellectual and motor slowing had disappeared, as had the hyponatremia, and the liver function tests had improved. MRI of the brain showed stenosis of the aqueduct of Sylvius, atrophy of the corpus callosum, a region with low signals in the frontal white matter and high signals in the white matter of the left hemisphere and in the body of the left caudate nucleus. In February 1989, after 6 months of treatment, the patient presented without fever and was in good condition. Hepatic and duodenal follow-up biopsies were normal at the histologic examination (PAS-negative).
The patient was emergently admitted to hospital in March 1989 because of acute intracranial hypertension associated with coma, signs of decerebration, and Parinaud syndrome. Brain CT scan showed dilatation of the ventricles, which had been normal since the shunt was placed in 1987. The ventriculoperitoneal shunt was then replaced by ventriculocisternostomy of the floor of the third ventricle. This ventriculostomy was effective but only mildly improved the clinical condition. In March 1989, histologic examination of a lymph node of the jugular chain showed numerous PAS-positive histiocytes. Analysis of the CSF revealed 5 lymphocytes but no histiocytes. The condition of the patient remained stable until June 1989, when the patient died suddenly of pulmonary embolism.
Comment: Sylvius aqueduct stenosis and ventricular dilatation can be either late complications or, more rarely, presenting manifestations of Whipple disease.
Mrs. C. was admitted to hospital in 1982, aged 58 years, because of a bout of inflammatory arthropathy of the ankle. She had been followed for 3 years for episodes of fever and arthralgia involving the interphalangeal joints, wrists, elbows, and knees. In 1982, the bout of arthritis was associated with recurring fever of 39 °C, nausea, vomiting, and diarrhea. All tests performed were normal at that time. Moreover, the patient had complained for 2 months of sudden but prolonged and rapidly progressing memory disorders associated with disorientation for time and place. Alteration of memory and orientation progressed rapidly. An assessment was performed in March 1983 because of a recent episode of polydipsia with polyuria. The brain CT scan and electroencephalography were normal. Analysis of the CSF showed neutrophilic pleocytosis, lymphocytes, and numerous abnormal mononuclear cells. In a second specimen, granulomatous cells were found and the diagnosis of sarcoidosis was considered. The tuberculin skin test was negative. The angiotensin-converting enzyme level in the blood was increased. A liver biopsy and bronchoalveolar lavage were both normal. Treatment with 1 mg/kg daily of prednisone was started, but no effect was noticed and the patient was discharged. One month later, the patient was admitted to hospital because of worsening memory loss, onset of Parinaud syndrome, nystagmus, and drowsiness. The improbable hypothesis of cerebral lymphoma was considered and irradiation was performed (45 grays in 5 weeks), but with no improvement in the clinical condition. The patient had decreasing levels of consciousness and died in September 1983. Diagnosis was determined by the postmortem examination. Lumpy necrotic foci were found bilaterally and symmetrically in the anterior and internal part of the amygdalae and the anterior and internal part of the right putamen. Necrotic lesions were also observed around the aqueduct of Sylvius. Microscopic examination showed numerous macrophages with PAS-positive granules in the necrotic foci, which are characteristic of Whipple disease. Visceral examination showed chronic nephritis and infiltration of the lamina propria by macrophages with PAS-positive material in the small bowel.
Comment: The history of arthritis, fever, and diarrhea could have suggested that the patient’s dementia was caused by Whipple disease.
Mrs. L., 64 years old, had a history of axial and peripheral HLA B27-negative spondyloarthropathy for more than 10 years. She also had suffered from depression and had had febrile gastroenteritis in April 1997. She was admitted to hospital in September 1998 because of longstanding fever, asthenia, anorexia, and weight loss which had developed over 18 months and which were associated with intellectual impairment. Radiologic and microbiologic examinations and a temporal artery biopsy had been negative in March 1998.
On admission, the general condition of the patient was gradually worsening. She had lost 12 kg; her temperature was 38.2 °C. She had diffuse hyperpigmentation, moderate hepatomegaly, and dry eyes and mouth. Examination mainly revealed intellectual disorders including time disorientation, auditory hallucinations, perseveration, aphasia, acalculia, agraphia, and apraxia. The mini-mental test score was 21/30. Upper motor neuron disorder, extrapyramidal rigidity, grasping, and several episodes with myoclonic movements of the limbs were also noticed. CT scan and MRI of the brain showed only cortical and subcortical atrophy. ESR was 70 mm/h, C-reactive protein level was 62 mg/L, and fibrinogen level was 6.6 g/L. The platelet count was 712 × 109/L. The blood neutrophil count was high. Lumbar puncture showed 13 cells/mm3, 63% of which were neutrophils, and a protein level of 0.4 g/L.
Diagnosis of Whipple disease could be established based on duodenal and jejunal biopsies showing infiltration of the submucosa and the lamina propria by PAS-stained foamy macrophages, with inclusion of bacillary bodies. Treatment with intravenous TMP-SMX (320 + 1,600 mg/d) was started on 15 October 1998, which led to dramatic neurologic and general improvement in a few days with apyrexia and disappearance of the inflammatory syndrome. Treatment was switched to oral TMP-SMX (320 + 1,600 mg/d) from 30 October to 29 November, but this led to a severe early relapse. The patient was bedridden with major consciousness disorders, a severe diffuse extrapyramidal movement disorder, and myoclonic seizures, all these manifestations being refractory to treatment with intravenous TMP-SMX (160 + 800 mg then 640 + 3,200 mg/d). Lumbar puncture showed 21 cells/mm3, 64% of which were lymphocytes. Treatment with 3 g daily of intravenous ceftriaxone and 0.9 mg/kg daily of methylprednisolone was started at the beginning of December, in the hypothesis of cerebral vasculitis, and led to dramatic improvement that lasted until March 1999. Corticosteroids and ceftriaxone were stopped and changed to oral cephalosporin (cefixime 400 mg/d). Three weeks later, the patient presented with fever again associated with anorexia, weight loss, increase in akinesia and rigidity, depression, and relapse of the severe inflammatory syndrome. Treatment with ceftriaxone (2 g/d) was started again, providing good results for 3 months. Two attempts were made in June 1999 and March 2000 to treat with oral TMP-SMX (320 + 1,600 mg/d) but they were followed by immediate failure with fever and neurologic signs. Thereafter, treatment with ceftriaxone (2 g/d) was maintained with no specific problem and no relapse. After 2 years, moderate neuropsychologic sequelae were noticed. The last lumbar puncture performed in January 1999 was negative for T. whippelii.
Comment: This observation underlines the importance of third-generation cephalosporins as an alternative treatment for the neurologic forms of Whipple disease.
Whipple disease can be defined as a relapsing chronic multisystemic disease. Neurologic presentation mainly occurs late in the disease (3,14,18,24,25,33,56). Neurologic signs of Whipple disease can be protean (8,47). They were first observed at necropsy of patients who died of Whipple disease (45,48). Neurologic presentation was then described as involving the brain as well as the spinal cord or the peripheral nervous system.
In Table 2 we summarize the frequency of neurologic signs of Whipple disease at diagnosis found in 4 major reviews of the literature. Neurologic presenting signs of the disease in 2 of these major reviews are summarized in Table 3. Neurologic involvement seems to be relatively common when routinely sought in cases progressing over several months (20%–43% of cases). On the contrary, Whipple disease rarely presents with isolated neurologic involvement (4% in the 2 major reviews of the literature). We report 12 original cases of Whipple disease with neurologic presenting signs. We compared them to 122 cases published in the literature from 1962 to 2000 concerning neurologic presentation of the disease. We first had to exclude 43 of these 122 cases that were in fact neurologic relapses of Whipple disease. These cases with neurologic relapse were reported from the 1960s to the 1980s and had occurred 1–4 years after the initial treatment (22). Indeed, such relapses were frequently observed when treatments were based on antibiotics, mainly tetracyclines, which were unable to cross the blood-brain barrier (10). Now, since current treatment is based on antibiotics able to cross the blood-brain barrier, these neurologic relapses are less commonly reported. We finally limited our study to the 79 remaining published cases plus our 12 cases. This led to a total of 91 cases with neurologic presentation of Whipple disease, which were classified into 2 groups:
- The first group (n = 70) included patients with predominantly but not exclusively neurologic signs. This group included 59 cases from the literature and 11 cases from our study.
- The second group (n = 21) included patients with clinically isolated neurologic presentation of the disease, including 20 cases from the literature and 1 case from our study. Some of the patients may have developed extraneurologic features later.
We analyzed the clinical signs and results of additional examinations in both groups and compared them to see what differences could be noted and what conclusions could be drawn. No statistical analysis could be applied to these disparate observations, but we were able to draw several conclusions; these are reported below.
Age and sex: The mean patient age in the first group (neurologic presentation) was 50 years, and the male:female ratio was 0.80. These characteristics are comparable with major reviews (15). In comparison, the second group (isolated neurologic manifestations) was characterized by a younger mean age (42 yr) and more women (male:female ratio = 0.5).
Neurologic manifestations: All observations showed diffuse neurologic signs. On average, 4–5 signs were observed (each patient presenting 1–13 symptoms). Various locations were observed, including involvement of the cerebral cortex, the hypothalamus, and the subcortical region, as well as cord involvement and peripheral nervous system involvement. The observations were classified according to the predominant clinical feature, in decreasing order of frequency: supranuclear ophthalmoplegia, confusion and dementia, psychiatric signs, myoclonic signs, seizures, hypothalamic involvement, cerebellar forms and myorhythmic forms (oculomasticatory myorhythmia and oculofacial-skeletal myorhythmia), cranial nerve involvement, peripheral neuropathies and cord involvement (Table 4). These forms of Whipple disease are traditionally isolated but meningeal involvement is often associated with them. Thus 2 forms should be added: meningoencephalitis and meningo-uveitis.
Globally, irrespective of group, diagnosis of neurologic involvement during Whipple disease was based on a combination of various neurologic symptoms underlining the spreading of the disease. Focal neurologic forms of Whipple disease, limited, for example, to the hypothalamus or the cortex, are in fact rare. Psychiatric signs and dementia are nearly always associated with neurologic signs.
Extraneurologic manifestations: Extraneurologic signs of Whipple disease included, in decreasing order of frequency (Table 5): joint involvement (arthralgia, arthritis), weight loss, fever, gastrointestinal signs (diarrhea, abdominal pain), lymphadenopathy, malabsorption, cutaneous signs (skin darkening), ocular signs (conjunctivitis, keratitis, uveitis), and serositis (pleuritis, pericarditis). Acute phase reaction was observed in 29/70 patients (41%).
When a patient presents with diffuse neurologic signs, different manifestations should be sought: rheumatic manifestations (two-thirds of the cases), constitutional symptoms (half the cases), fever (half the cases), lymphadenopathy, digestive signs, and acute phase reaction (in slightly more than one-third of the cases) and some times malabsorption, ocular manifestations, dermatologic manifestations, and involvement of serous membranes (in less than one-fifth of cases).
Functional and morphologic examinations
Electroencephalography, performed in 50 cases, did not reveal any specific pattern. Brain MRI, performed in 52 cases, appeared more sensitive than brain CT scan, performed in 62 cases. In some cases, brain MRI and brain CT were normal (12/52 and 25/62, respectively) or showed only atrophy. In other cases, they showed various lesions (high signals on T2-weighted images on MRI, hypodense lesion on CT scan), affecting, in decreasing order of frequency, the frontal cerebral cortex, the basal ganglia, the periventricular white matter, the hypothalamus, the temporal cortex, and the parietal cortex. Cerebral MRI appears to be the best imaging modality for Whipple disease, even if it can be nonspecific or normal in some cases. Nevertheless, it can be useful to guide biopsy (35,51) and to follow evolution during treatment, especially in cases with tumorlike brain involvement.
CSF was abnormal in half of the cases studied (n = 73 cases), compared to two-thirds of cases with neurologic relapses (n = 29). PAS-positive granules could be seen with light microscopy in half the cases even if the CSF was normal. Analysis of CSF by PCR assay could not be assessed as only 12 cases were studied. However, PCR assay in CSF now appears to be an essential tool for confirming neurologic involvement in patients with Whipple disease. It also helps to assess response to treatment in parallel with the clinical evolution and to consider discontinuing treatment when the PCR becomes negative (53).
Lumbar puncture is a cost-effective diagnostic test of neurologic Whipple disease since it is abnormal in more than half the cases. Lumbar puncture is also essential for following patients suffering from Whipple disease with neurologic involvement. Analysis of the CSF should be routinely performed and should include light microscopy with PAS staining, electron microscopy, and PCR assay for diagnosis and follow-up of patients. If the CSF PCR is positive there is a strong argument for the diagnosis of Whipple disease. Nevertheless it appears essential to look for histologic involvement of the duodenum to have the formal proof of the disease. The search for other sites depends on the clinical presentation but a total body CT scan is justified to look for lymphadenopathy.
In the first group of patients with neurologic presentation (n = 70), the histologic aspect of the duodenal biopsies was suggestive of Whipple disease in 10 of 11 of our cases, showing that the disease was widespread. The literature is consistent with these data and reports 42 cases including 75% of positive biopsies with light microscopy. Electron microscopy appeared to be positive in some specimens that were negative with light microscopy. PCR assay was performed only once in our observations and was positive. It was carried out 11 times in the literature data and was always positive.
In the second group, with isolated neurologic signs (n = 21), duodenal biopsies were always negative with light microscopy. However, the 3 times PCR assay was performed on these biopsies, it was positive, proving gastrointestinal involvement (5,9,36).
In the first group, abdominal lymphadenopathy could be seen on CT scan. One-third of the examined patients had mesenteric lymphadenopathy (22 cases in the literature and 3 of our cases). Biopsies of these lymph nodes were positive with light microscopy 21 of 25 times (84%). CT was probably underused in the second group because of isolated neurologic presentation, and no abdominal CT scan was reported.
In the literature, 6 postmortem examinations were performed on patients of the second group. Analysis of organs other than the brain was reported in 5 cases. Three of these 5 patients suffered from isolated neurologic involvement (1,40) or neuroophthalmologic involvement (23). In the latter case (23), light microscopy and electron microscopy of the retina were positive. In 2 patients, the postmortem light microscopy examination showed multisystemic involvement. Indeed, the liver, lungs, spleen, and 1 mesenteric lymph node were involved in a 32-year-old woman. She suffered from meningoencephalitis of unknown origin, ophthalmoplegia, and facioscapular myopathy. She died of acute right heart ventricular failure secondary to recurrent pulmonary embolism (19). In a 28-year-old woman, the liver, lungs, spleen, and lymph nodes were involved. She suffered from meningoencephalitis and also died of acute right heart ventricular failure secondary to recurrent pulmonary embolisms (28).
When a neurologic presentation of Whipple disease is suspected but the analysis of the duodenal biopsy with light microscopy is negative, the duodenal biopsy should be analyzed with PCR assay (29) and the CSF should be analyzed with light microscopy, electron microscopy, and PCR. Moreover, an abdominal CT scan to search for lymph node involvement should be performed before every brain biopsy.
Patients must undergo thorough examination for any sign of systemic involvement (arthritis, serositis, lymphadenopathy). In fact, it can be useful to tap or biopsy these lesions if the brain biopsy seems difficult because of the position of the lesion. Neurologic presentation of the disease in which no systemic lesions can be found on postmortem examination is rare, but these cases show that selective involvement of the nervous system is possible (1,23,40). In the case of peripheral neuropathy, examination of other sites included muscle biopsies analyzed with light microscopy (3 times) and with electron microscopy (twice) (11,12,20), 2 peripheral nerve biopsies analyzed with light microscopy (twice) and with electron microscopy (once) (11,20). None of these biopsies was positive. Specific involvement of the peripheral nervous system could not be confirmed with histologic analysis. PCR assays would be necessary.
When Whipple disease presents with neurologic signs the prognosis is generally poor. The severity of the disease and the poor response to treatment explain why neurologic symptoms are such a common cause of death in patients with the disease. We reanalyzed the 79 observations of the literature in combination with our own 12 cases to compare the progression of the disease in each group according to the treatment used. For the 87 cases in whom the outcome was known, we studied the different treatments used: corticosteroids alone, tetracyclines, TMP-SMX, cephalosporins, other antibiotics (penicillin, macrolides, chloramphenicol, antimycobacterial drug, etc), and other treatments (surgery, vitamins, neuroleptics, antiepileptics, etc). Two patients whose treatment was not known died.
We divided the prognosis of the 87 patients whose outcome and treatment were known into 3 subgroups: clinical and/or radiologic improvement, clinical and/or radiologic stabilization, and worsening and death (Table 6). We observed that corticosteroids were ineffective and even dangerous in the long term, as was already known. Treatment with TMP-SMX, irrespective of the group to which the patient belonged, was highly effective, with 3 deaths in 25 patients in the first group and no deaths in 9 patients in the second group. The outcome was improved for 22 of the 37 patients treated. Dramatic improvement might occur, as in 1 of our patients who started with mutism and recovered normal daily life within a month. Other antibiotics appeared less effective: they yielded a nearly equal distribution of patients who improved, stabilized, or died. One patient improved after surgical treatment for hydrocephalus.
The prognosis of the neurologic presentation of Whipple disease is the same in both groups, whether the neurologic manifestations are isolated or not. The prognosis mainly depends on the treatment used. The effectiveness of third-generation cephalosporins was established by Adler and Galetta (2) and Cooper and colleagues (10) after the failure of TMP-SMX. Many authors also report the effectiveness of third-generation cephalosporins (5,13,16,27,32,49,51) particularly for neurologic relapses of Whipple diseases previously diagnosed (7,26,37,43,44,55). (Neurologic relapses were not included in our study, however, as previously described.) The most effective drugs are TMP-SMX and cephalosporins (42,43).
Proposition for a Diagnostic and Therapeutic Strategy for Whipple Disease with Neurologic Involvement
Strategy for additional examinations
Diagnosis can be established only by associating clinical, histologic, and bacteriologic data, as the germ can be found in healthy individuals in the gastric fluid or in the saliva by PCR assay. We concluded from our observations and the literature review that several tests are needed when neurologic signs suggestive of Whipple disease are present. These tests should be performed in the following order:
- Analysis of the CSF with light microscopy, electron microscopy, and PCR as soon as neurologic signs begin, associated with MRI. Analysis of the CSF should be repeated during follow-up, and treatment should be discontinued only when the results of the PCR assay performed on the CSF are negative.
- Search for gastrointestinal tract involvement with multiple duodenal biopsies at different levels and analysis of the biopsies with light microscopy, electron microscopy, and PCR.
- Search for systemic involvement, with biopsies of superficial lymphadenopathy, puncture of joint effusion, and use of abdominal CT scan to guide biopsies or surgical resection of abdominal lymphadenopathy which should be analyzed with light microscopy, electron microscopy, and PCR.
- When all these examinations are negative, if Whipple disease is suspected and a lesion is found on the brain MRI, a stereotactic cerebral biopsy should be performed.
We were disappointed by the biopsy of peripheral nerves even in the case of clinical peripheral neuropathy. The significance of this involvement is still an open question.
The optimal treatment of Whipple disease remains controversial, as far as both the antibiotic chosen and the length of treatment. The treatment currently recommended for patients without neurologic involvement is a combination of procaine, penicillin G 1.2 million units daily, and streptomycin 1 g daily intramuscularly for 14 days followed by oral TMP-SMX (320 mg + 1,600 mg daily) for 1 year (22,50). However, neurologic symptoms may develop or persist despite this treatment (10,16,27,54). The particularly resistant character of the neurologic involvement led authors like Adler and Cooper to propose an alternative treatment:
Parenteral treatment starting with 2 g of ceftriaxone twice daily for 14–30 days, depending on the authors, followed by oral TMP-SMX twice daily (160 mg/800 mg) or cefixime (400 mg/day) for 1–2 years (4,21).
For patients whose pretreatment CSF PCR is positive, antibiotics should be continued until the CSF PCR becomes negative (4,53).
We note 1 case of neurologic deterioration with antibiotics for which the administration of antibiotics in combination with gamma-interferon was found to be effective (41).
Whipple disease is due to infection by T. whippelii in people with a probable underlying immunosuppression condition. It is multisystemic and may involve the central nervous system, peripheral involvement being much rarer and nonspecific. Clinical signs including multisystemic disease with slow progression, with selective involvement, for example involvement of the hypothalamus or dementia, may be suggestive of Whipple disease. However, it may be much more difficult to confirm the diagnosis, especially when neurologic signs seem isolated. In that case, extraneurologic signs, especially gastrointestinal tract or rheumatic involvement, should be systematically sought. Routine duodenal biopsy is positive in 75% of cases. Analysis of the CSF is the additional neurologic test that may provide most information. The CSF is abnormal in 2 of 3 cases, with PAS-positive granules in half the cases. Analysis with PCR assay to look for T. whippelii DNA is promising.
Treatment of Whipple disease with long-term TMP-SMX for at least 1 year has proved effective, but further studies are needed to assess the value of initial treatment and/or alternative treatment with third-generation cephalosporins when other treatments are not effective. Gamma-interferon might be added in case of resistance to antibiotic treatment.
1. Adams M, Rhyner PA, Day J, De Armond S, Smuckler EA. Whipple disease confined to the central nervous system. Ann Neurol 21: 104–8, 1987.
2. Adler CH, Galetta SL. Oculo-facial-skeletal myorhythmia in Whipple disease: Treatment with ceftriaxone. Ann Intern Med 112: 467–9, 1990.
3. Albers JW, Nostrant TT, Riggs JE. Neurologic manifestations of gastrointestinal disease. Neurol Clin 7: 525–48, 1989.
4. Anderson M. Neurology of Whipple’s disease. J Neurol Neurosurg Psychiatry 68: 2–5, 2000.
5. Averbuch-Heller L, Paulson GW, Daroff RB, Leigh RJ. Whipple’s disease mimicking progressive supranuclear palsy: The diagnostic value of eye movement recording. J Neurol Neurosurg Psychiatry 66: 532–5, 1999.
6. Bjerknes R, Laerum OD, Degaard S. Impaired bacterial degradation by monocytes and macrophages from a patient with treated Whipple’s disease. Gastroenterology 89: 1139–46, 1985.
7. Brandle M, Ammann P, Spinas GA, Dutly F, Galeazzi RL, Schmid C, Altwegg M. Relapsing Whipple’s disease presenting with hypopituitarism. Clin Endocrinology 50: 399–403, 1999.
8. Bruyn GW. Whipple disease. In: Harris AA, ed. Handbook of clinical neurology. Vol 8(52): Microbial disease. New York: Elsevier Science, pp 135–42, 1988.
9. Clarke CE, Falope ZF, Abdelhadi HA, Franks AJ. Cervical myelopathy caused by Whipple’s disease. Neurology 50: 1505–6, 1998.
10. Cooper GS, Blades EW, Rembler BF, Salata RA, Bennert KW, Jacobs GH. Central nervous system Whipple disease: Relapse during therapy with trimethoprim-sulfamethoxazole and remission with cefixime. Gastroenterology 106: 782–6, 1994.
11. Cooper TJ, Bird G, White B, Ferguson IT. Recovery from bilateral wrist-drop in Whipple disease. J Neurol Neurosurg Psychiatry 51: 152–4, 1988.
12. Cruz Martinez A, Gonzalez P, Garza E, Bescansa E, Anciones B. Electrophysiologic follow-up in Whipple Disease. Muscle Nerve 10: 616–20, 1987.
13. Delanty N, Georgescu L, Lynch T, Paget S, Stubgen JP. Synovial fluid polymerase chain reaction as an aid to the diagnosis of central nervous system Whipple’s disease. Ann Neurol 45: 137–8, 1999.
14. Dobbins III. WO Whipple disease. Mayo Clin Proc 63: 623–4, 1988.
15. Dobbins WO III. Whipple disease. Springfield, IL: Charles C Thomas, pp 113–8, 1987.
16. Dykman DD, Cuccherini BA, Fuss IJ, Blum LW, Woodward JE, Strober W. Whipple disease in a father-daughter pair. Dig Dis Sci 44: 2542–4, 1999.
17. Fleming JL, Wiesner RH, Shorter RG. Whipple disease: Clinical, biochemical and histopathologic features and assessment of treatment in 29 patients. Mayo Clin Proc 63: 539–51, 1988.
18. Gaist D, Ladefoged K. Whipple disease. Scand J Gastroenterol 29: 97–101, 1994.
19. Gartner J. Whipple disease of the central nervous system associated with ophthalmoplegia externa and severe asteroid hyalitis. A clinicopathologic study. Doc Ophthalmol 49: 155–87, 1980.
20. Halperin JJ, Landis DMD, Kleinman GM. Whipple disease of the nervous system. Neurology 32: 612–7, 1982.
21. Kaufmann M, Risti B, Fried M. Die Whipple-Erkrankung. Internist 37: 895–902, 1996.
22. Keinath RD, Merrell DE, Vliestra R, Dobbins III. WO Antibiotic treatment and relapse in Whipple disease. Long-term follow-up of 88 patients. Gastroenterology 88: 1867–73, 1985.
23. Knox DL, Green WR, Troncoso JC, Yardley JH, Hsu J, Zee DS. Cerebral ocular Whipple disease: A 62-year odyssey from death to diagnosis. Neurology 45: 617–25, 1995.
24. Lange U, Teichmann J, Doppl W, Klor H. Whipple disease—Current status of diagnostics and therapy. Eur J Med Res 3: 331–9, 1998.
25. Le Thi Huong DU, Wechsler B, Gray F, Poisson M, Brunet P, Signoret JL, Godeau P. Maladie de Whipple a forme neurologique. A propos d’une observation avec verification anatomique. Arch Anat Cytol Pathol 34: 175–80, 1986.
26. Lieb K, Maiwald M, Berger M, Voderholzer U. Insomnia for 5 years. Lancet 354: 1966, 1999.
27. Louis ED, Lynch T, Kaufmann P, Fahn S, Odel J. Diagnostic guidelines in central nervous system Whiple’s disease. Ann Neurol 40: 561–8, 1996.
28. Ludwig B, Bohl J, Haferkamp G. Central nervous system involvement in Whipple disease. Neuroradiology 21: 289–93, 1981.
29. Lynch T, Odel J, Fredericks DN, Louis ED, Forman S, Rotterdam H, Fahn S, Relman DA. Polymerase chain reaction-based detection of Tropheryma whippelii in central nervous system Whipple’s disease. Ann Neurol 42: 120–4, 1997.
30. Maiwald M, Schuhmacher F, Ditton HJ, von Herbay A. Environmental occurrence of the Whipple disease bacterium (Tropheryma whippelii). App Environ Microbiol 64: 760–2, 1998.
31. Maizel H, Ruffin JM, Dobbins III. WO Whipple disease: A review of 19 patients from one hospital and a review of the literature since 1950. Medicine (Baltimore) 49: 175–205, 1970.
32. Manzel K, Tranel D, Cooper G. Cognitive and behavioral abnormalities in a case of central nervous system Whipple disease. Arch Neurol 57: 399–403, 2000.
33. Marbet UA, Stalder GA, Gyr KE. Whipple disease: A multisystemic disease with changing presentation. Dig Dis Sci 4: 119–28, 1986.
34. Marth T, Neurath M, Cuccherini BA, Strober W. Deffects of monocyte interleukin 12 production and humoral immunity in Whipple disease. Gastroenterology 113: 442–8, 1997.
35. Mendel E, Khoo LT, Go JL, Hinton D, Zee CS, Apuzzo MLJ. Intracerebral Whipple’s disease diagnosed by stereotactic biopsy: A case report and review of the literature. Neurosurgery 44: 203–9, 1999.
36. Misbah SA, Ozols B, Franks A, Mapstone N. Whipple’s disease without malabsorption: New atypical features. Q J Med 90: 765–72, 1997.
37. Peters FPJ, Wouters RSME, De Bruine AP, Stockbrugger RW. Cerebral relapse of Sarcoidlike Whipple’s disease. Clin Infect Dis 24: 1252–5, 1997.
38. Raoult D, Birg ML, La Scola B, Fournier PE, Enea M, Lepidi H, Roux V, Piette JC, Vandenesch F, Vital Durand D, Marrie TJ. Cultivation of the bacillus of Whipple disease. N Engl J Med 342: 620–5, 2000.
39. Relman DA, Schmidt TM, MacDermott RP, Falkow S. Identification of the uncultured bacillus of Whipple disease. N Engl J Med 327: 293–301, 1992.
40. Romanul FC, Radvany J, Rosales RK. Whipple disease confined to the brain: A case studied clinically and pathologically. J Neurol Neurosurg Psychiatry 40: 901–9, 1977.
41. Schneider T, Stallmach A, von Herbay A, Marth T, Strober W, Zeitz M. Treatment of refractory Whipple disease with interferon-gamma. Ann Intern Med 129: 875–7, 1998.
42. Schnider PJ, Reisinger EC, Berger T, Krejs GJ, Auff E. Treatment guidelines in central nervous system Whipple’s disease. Ann Neurol 41: 561–2, 1997.
43. Schnider PJ, Reisinger EC, Gerschlager W, Muller C, Berger T, Krejs GJ, Auff E. Long-term follow-up in cerebral Whipple’s disease. Eur J Gastroenterol Hepatol 8: 899–903, 1996.
44. Schnider PJ, Trattnig S, Kollegger H, Auff E. MR of cerebral Whipple disease. AJNR Am J Neuroradiol 16: 1328–9, 1995.
45. Schochet Jr, SS Lampert PW. Granulomatous encephalitis in Whipple disease. Electron microscopic observations. Acta Neuropathol (Berl) 13: 1–11, 1969.
46. Schoedon G, Goldenberger D, Forrer R, Gunz A, Dutly F, Hochli M, Altwegg M, Schaffner A. Deactivation of macrophages with interleukin-4 is the key to the isolation of Tropheryma whippelii. J Infect Dis 176: 672–7, 1997.
47. Selhorst JB, Schwartz MA. Cerebral manifestations of Whipple disease. Mayo Clin Proc 63: 1057–60, 1988.
48. Sieracki JC, Fine G, Horn RC, Bebin J. Central nervous system involvement in Whipple disease. J Neuropathol Exp Neurol 19: 68–74, 1960.
49. Simpson DA, Wishnow R, Gargulinski RB, Pawlak AM. Oculofacial-skeletal myorhythmia in central nervous system Whipple’s disease: Additional case and review of the literature. Mov Disord 10: 195–200, 1995.
50. Singer R. Diagnosis and treatment of Whipple disease. Drugs 55: 699–704, 1998.
51. Suzer T, Demirkan N, Tahta K, Coskun E, Cetin B. Whipple’s disease confined to the central nervous system: Case report and review of the literature. Scand J Infect Dis 31: 411–4, 1999.
52. Vital Durand D, Lecomte C, Cathebras P, Rousset H, Godeau P, SNFMI Research Group on Whipple Disease. Whipple disease. Clinical review of 52 cases. Medicine (Baltimore) 76: 170–84, 1997.
53. Von Herbay A, Ditton HJ, Schuhmacher F, Maiwald M. Whipple disease: Staging and monitoring by cytology and polymerase chain reaction analysis of cerebrospinal fluid. Gastroenterology 113: 434–41, 1997.
54. Von Herbay A, Otto HF, Stolte M, Borchard F, Kirchner T, Ditton HJ, Maiwald M. Epidemiology of Whipple disease in Germany. Analysis of 110 patients diagnosed in 1965–95. Scand J Gastroenterol 32: 52–7, 1997.
55. Weeks RA, Scott J. Cerebral Whipple’s disease: Relapse presenting with spinal myoclonus. BMJ 312: 371–3, 1996.
© 2002 Lippincott Williams & Wilkins, Inc.
56. Wilson KH, Wilson JAP. Whipple’s disease. Curr Opin Infect Dis 8: 380–3, 1995.