CNS Whipple Disease 

Updated: Dec 11, 2014
Author: George C Bobustuc, MD; Chief Editor: Niranjan N Singh, MBBS, MD, DM, FAHS, FAANEM 



Whipple disease constitutes a rare, relapsing, slowly progressive, infectious, systemic illness characterized by fever of unknown origin, polyarthralgias, and chronic diarrhea.

Other manifestations include skin and ocular involvement (ie, uveitis, retinitis, optic neuritis); generalized lymphadenopathy; afebrile, blood culture-negative endocarditis which, reportedly, can be complicated with cardioembolic strokes; and a sarcoidosis-like syndrome with mediastinal lymph nodes and central nervous system (CNS) involvement (ie, dementia, sensory and motor deficits, ophthalmoplegia, myoclonus, stroke and hypothalamic damage with dysautonomia, emotional impairment, endocrinopathy).

Fewer than 1000 cases have been reported, and less than one half (6-43%) of these patients presented with neurological manifestations. This likely represents an underestimate due to both a low index of suspicion in some cases and difficulties in reaching a diagnosis in others.

This article, besides being a general presentation of Whipple disease, focuses on both the neurologic manifestations and specifics of diagnosis and treatment of Whipple disease with symptomatic CNS involvement (CNS-WD).[1]

Despite the slowly progressive course of most cases of Whipple disease, CNS-WD may have a fulminant course, and manifest isolated CNS-WD cases have been reported in the literature.[2] Prompt diagnosis is imperative, as very effective therapies are easy to employ with typically rapid limitation of CNS progression and even partial reversal of CNS symptoms. If left untreated, progression to death may come as quickly as 1 month after CNS involvement begins.


Historical perspective

1907: Whipple proposed the name of "intestinal lipodystrophy" for a new, distinctive clinical syndrome.[3] Whipple's case report presented a 36-year-old medical missionary with a 5-year history of episodes of relapsing-progressive polyarthritis subsequently complicated by weight loss, cough, fever, diarrhea, hypotension, abdominal swelling, increased skin pigmentation, and severe anemia. The hallmark of the pathologic report was the marked infiltration by foamy macrophages of joints and aortic valves, and prominent deposits of fat within intestinal mucosa and mesenteric lymph nodes, which made Whipple consider this case an obscure disease of fat metabolism and propose the name intestinal lipodystrophy. Whipple pointed out the existence of great numbers of peculiar rod-shaped bacteria found in extracts of lymph node tissue and lamina propria of the intestine.

1923: A second case of Whipple disease was reported in the literature.

1947: Peroral small bowel biopsy was used for the first time to make the first reported premortem diagnosis.

1949: Black-Schaffer advanced the diagnosis, proved the systemic nature of this disease, and raised the suspicion of an infectious cause for Whipple disease.[4]

  • He identified periodic acid-Schiff (PAS)–staining granules, most likely representing degenerating bacterial forms, within macrophages isolated from the small bowel as well as other tissue and fluid specimens (eg, pericardium, endocardium, lymph nodes, synovia, lung, brain, meninges) obtained from patients in whom Whipple disease was suspected.

  • The presence of PAS-positive granules containing macrophages is not pathognomonic. Intestinal lamina propria of AIDS patients with concomitant Mycobacterium avium-intracellulare complex (MAC) infection may be packed with PAS-positive granules containing macrophages, but the intracellular bacilli are acid fast. Concomitant Whipple disease and MAC have been reported.

1952: Paulley was first to report a case of a patient with histologically proven Whipple disease whose symptoms responded to chloramphenicol.[5] Other reports followed of successful attempts to treat patients with prolonged courses of antibiotics (12 mo or longer), particularly a combination of penicillin and streptomycin followed by trimethoprim-sulfamethoxazole (TMP-SMX).

1961: Electron microscopy (EM) studies by Yardley et al provided more evidence for an infectious cause of Whipple disease by finding bacillary bodies within membrane-bound vesicles in the cytoplasm of macrophages. Whipple disease bacillus has a characteristic trilamellar appearance on EM.

1985: A survey by Keinath et al[6] of 88 patients with Whipple disease whose symptoms responded to antibiotics revealed a high rate of relapse (40%, 31/88); many of the relapses were in the CNS, thus indicating the need to use antibiotics with adequate blood-brain barrier (BBB) penetrance.

1991-1992: Wilson et al[7] reported Whipple bacillus as a gram-positive bacterium rich in guanine and cytosine and likely an actinomycete. They used a gene that encodes for 16S ribosomal RNA (rRNA) in bacteria to characterize the nucleotide sequence of the bacillus from a patient with Whipple disease.

1992: Relman et al[8] confirmed the findings of Wilson et al and proposed a classification of the organism. Tropheryma whippelii, a previously uncharacterized organism, was, on the basis of phylogenetic analysis of a specific 16S rRNA gene sequence, a novel actinomycete.

  • Relman used polymerase chain reaction (PCR) to amplify this unique bacterial 1321-base sequence of the 16S rRNA gene obtained from tissues from 5 patients with Whipple disease. It could not be obtained from 10 control patients with other conditions.

  • Further PCR studies have been used successfully to confirm the systemic involvement of other tissues (eg, heart, vitreous fluid, peripheral blood cells, pleural effusion cells).

1997: Ramzan et al[9] used PCR to confirm Whipple disease in patients whose histologic studies of small intestine samples obtained by peroral biopsy were nonconfirmatory. PCR studies with 16S rDNA primers of T whippelii proved to be highly sensitive, specific, and useful for monitoring response to therapy and likelihood of relapse. Prior studies had shown no correlation between posttreatment histologic findings, clinical outcome, and likelihood of recurrence.

1997: A study by Herbay et al suggested that most, if not all, patients with Whipple disease have CNS involvement and only some develop clinical and radiologic evidence of CNS-WD; PCR analysis of cerebrospinal fluid (CSF) was proposed as routine in the diagnostic evaluation of patients in whom Whipple disease is suspected.

2000: Raoult et al[10] successfully cultivated T whippelii using a human fibroblast cell line (HEL). They completed 7 passages of an isolate obtained from the aortic valve of a patient with endocarditis caused by Whipple disease. The following findings confirmed that the isolates passaged were T whippelii: the amplified sequences of the 16S rRNA gene of the isolate were identical to those of T whippelii; transmission EM of the isolate revealed the distinctive trilamellar appearance of Whipple disease bacillus; PAS-positive bacilli (not acid fast but gram positive) were identified in an intracellular location in the cell-culture monolayer; and mice-produced polyclonal antibodies could detect the bacterium in the patient's excised heart valve.

  • Raoult et al[11, 12] developed an immunofluorescence serologic test with which they examined serum from a limited number of patients with Whipple disease (9 patients with Whipple disease and 40 control subjects). Both immunoglobulin G (IgG) and immunoglobulin M (IgM) antibodies against the bacillus were tested with cut-off values of 1:100 and 1:50, respectively. The sensitivity of the IgG antibody testing was high (9 of 9), but the specificity was quite low, as almost 75% of control subjects tested positive.

  • The IgM antibody testing revealed slightly lower sensitivity (7 of 9) but proved to be more specific (only 3 of 40 control subjects tested positive). A caveat is warranted in the interpretation of these results: both IgG and IgM antibody testing results may be distorted by a sampling effect, in that a small number of samples from patients with Whipple disease and the relatively small number of control subjects with a limited variety of other infectious diseases may underrate the IgM cross-reactivity.

  • The high frequency of IgG antibodies against Whipple disease isolate in samples from control subjects suggests that this pathogen is ubiquitous, causing illness only occasionally. This may be due to differences in host factors or virulence amongst strains or a result of the patient's exposure to other cross-reacting microorganisms.

2003: The genome sequencing of 2 different T whippelii strains (Twist and TW08/27) is achieved. It revealed interesting particularities, which could explain some of the clinical traits already observed. T whippelii genome encodes for around 800 protein coding genes. It lacks key biosynthetic pathways and has a reduced capacity for energy metabolism. It has a family of large surface proteins, some associated with large amounts of noncoding repetitive DNA, which appears to trigger frequent genome rearrangements, potentially resulting in the expression of different subsets of cell surface proteins. This could be the basis of a mechanism to evade host defenses.[13]

Host abnormalities

A variety of host abnormalities has been reported in patients with Whipple disease. They point to an anomalous cytokine-driven regulation of both phagocytosis and humoral and cellular immunity and specifically suggest a defect in the axis of interleukin-12 (IL-12) and gamma interferon.

  • IL-12 is a proinflammatory cytokine, rapidly produced by phagocytic cells, professional antigen-presenting cells such as dendritic cells and skin Langerhans cells, and B cells.

  • IL-12 production is triggered by intracellular pathogens, bacteria, fungi, viruses or their phagocytosis-induced breakdown products.

  • It is secreted in both a T-cell–dependent and –independent manner.

  • IL-12 elicits gamma interferon production by activating both T and natural killer cells and resting peripheral monocytes and thus enhances completion of phagocytosis.

  • IL-12 also functions as a growth factor for the activated cluster of differentiation (CD) 4 and CD8 lymphocytes and inhibits immunoglobulin E (IgE) production.

  • Patients with Whipple disease have shown a decreasing number of immunoglobulin A (IgA)–containing plasmacytes in the lamina propria of the bowel during the infection as the clinical symptoms worsened; this number returns to normal with treatment and inversely correlates with the number of foamy macrophages. The foamy macrophages represent transformed monocytes already engaged in the engulfing and phagocytosis of T whippelii bacilli. The number of foamy macrophages declines with treatment.

  • Accumulation of foamy macrophages in the advanced stages of Whipple disease has been supported by the decreased in vitro phagocytic ability of monocytes collected from patients with Whipple disease.

  • The reduced in vitro phagocytic ability of macrophages from patients with Whipple disease has been explained on the basis of anomalous cytokine-related regulation of this function. A decrease in the IL-12 secretory ability of the monocytes and subsequently gamma interferon by T cells has been identified in vitro in patients with Whipple disease. This is supported further by a case report of a patient with Whipple disease who had developed resistance to antibiotics on recurrence and subsequently had been treated successfully with gamma interferon.

  • Patients with Whipple disease have shown an increased number of lymphocytes in the lamina propria of the small intestine with a decrease in the CD4-to-CD8 ratio and a decrease in the CD11b (complement receptor 3alpha chain)–expressing subpopulation. This represents another immunological abnormality encountered in this disease, which could be explained by a defect in the axis of IL-12 and gamma interferon.

  • Specific molecular defects involving the axis of IL-12 and gamma interferon have been recognized as the basis for a variety of host anomalies responsible for the increased susceptibility to chronic inflammatory conditions caused by intracellular pathogens, including nontuberculous mycobacteria, vaccine associated bacille Calmette-Guérin (BCG) infections, Salmonella species, and some virus-induced infections.

  • The mutations described in these cases involved gamma interferon receptor, IL-12 receptor beta1, and IL-12 p40 genes.[14]

  • No reports have yet attempted to identify and clarify any specific genetic defect involving the proven IL-12–gamma-interferon axis functional deficit in patients with Whipple disease.

Humans remain the only known host for the disease. No evidence exists of person-to-person transmission, and no reported outbreaks have occurred. In Germany, an environmental source was suggested by findings of specific T whippelii DNA in sewage water and the saliva[15] and jejunal juice of some healthy controls.

The initial gastrointestinal (GI) involvement argues for this site as the entry portal of T whippelii and probable dissemination through the body by the lymphatics and bloodstream either directly or via a carrier (eg, monocytes). The brain ultimately represents a favored site, but the mechanism by which the BBB is breached is unclear and insidious, supporting the theory of carrier-mediated dissemination.

Whether the clinical manifestations of Whipple disease result from direct bacterial invasion or from the ensuing inflammatory response is not clear.

At this time, the inability to grow T whippelii in cell-free, medium-only culture prevents researchers from developing better testing procedures (eg, selection of more specific antigens for development of more specific serologic tests[16] ) and treatments (antimicrobial susceptibility testing) and from answering important questions about this pathogen (eg, what is T whippelii, a commensal intestinal organism or a saprobe [ie, an organism that lives in and derives its nourishment from organic matter in stagnant or foul water]? What are the differences in pathogenicity among various strains? Is the infection acquired primarily through the GI tract?). Furthermore, the 2 remaining Koch postulates are still to be fulfilled—the development of Whipple disease in an animal model infected with Whipple disease isolate and subsequent isolation of T whippelii from the animal.

Morphology of T whippelii

T whippelii has a specific morphology. The thick wall of this 1- to 2-mm rod gives it the appearance of encapsulation, and the inner layer is PAS positive.

Stages of Whipple disease

The clinical course of untreated Whipple disease can include the following 3 stages:

  • Nonspecific: This stage includes vague complaints of migratory polyarthralgias, abdominal fullness, low-grade fever, anorexia, and cough. This stage may last more than 5 years.

  • Abdominal: This stage involves weight loss, weakness, chronic diarrhea, and abdominal pain. This stage may last 10-20 years.

  • Generalized: This stage is characterized by steatorrhea; cachexia; lymphadenopathy; hyperpigmentation; and cardiovascular, pulmonary, neurological, and ocular dysfunction. This stage may last as long as 5 years until death if Whipple disease is not diagnosed and remains untreated.

This proposed staging had at its base a limited review of 15 patients. This review also showed that 50% of patients had symptoms for more than 5 years before presentation. Patients with Whipple disease who were left untreated had a 5-year survival rate of 80% after onset of arthralgias, but only 20% of patients survived 5 years after onset of diarrhea or abdominal pain.



United States

Whipple disease is a rare condition. No incidence and prevalence studies have been reported.

  • Several difficulties are encountered when these studies are contemplated, such as lack of a target population, low index of suspicion in the medical community, unavailability of diagnostic methods, and variations in diagnostic standards.

  • No reported cluster of cases indicates a target population. No specific natural habitat of the organism is known, and the specific mechanisms by which the infection takes place are not known.


Several comprehensive reviews of the literature have been conducted over the years, and the number of approximate reported cases evolved as follows: 300 cases in 1983; 800 cases in 1996; and 1000 cases in 1998. This may represent an increase in the index of suspicion, availability of new diagnostic techniques, and population increase. These numbers still are believed to represent an underestimate of the disease frequency.


Whipple disease left untreated is uniformly fatal.

  • Fewer than 5% of patients have signs and symptoms suggesting CNS involvement at the clinical onset of the disease, but the brain reportedly represents the final target organ in most patients.

  • The incidence of relapse may be quite high (approaching 40%) in patients in whom antibiotic treatment was terminated after 1 year but is not correlated with significant tissue findings on PCR studies (ie, tissue deriving from an organ accountable for clinical symptoms). Patients with negative PCR results in significant tissue at the time of completion of their antibiotic course had a very low relapse rate.

  • Human leukocyte antigen (HLA) B27 was reported in some studies as being more frequent in patients with Whipple disease than in the general population.


Most of the cases reported originated from Europe and North America, and some prior reports mentioned a preponderance of Whipple disease in white, middle-aged men. Still, the number of reported cases is too low to reveal any significant racial susceptibility.


The male-to-female ratio is 6-8:1.


Onset is usually in middle age (30-40 y). Age range at diagnosis reported in the literature is 3 months to 81 years.




Whipple disease commonly starts with GI complaints, but because of the multisystemic involvement, presentation can be quite variable. At any time in the course of the disease a constellation of symptoms relating to different organ systems may be present. The GI symptomatology is initially mild, and affected persons may go years before seeking medical attention, often for symptoms other than GI-related ones.

  • Symptoms at presentation (other than CNS related) include the following:

    • Weight loss (80-100% of patients) generally ranges from 20-40 pounds over several months to years.

    • Diarrhea, watery or fatty, occurs in 75% of patients. Occult blood loss was reported with almost all patients, but clinically relevant lower GI bleed (hematochezia or melena) is distinctly unusual.

    • Arthralgia (70% of patients) typically is the most common cause of initial concern for the patient. It is migratory in nature and involves large joints in an asymmetric fashion.

    • Abdominal (nonspecific) pain (50%) is usually more severe after eating.

    • Chest (ie, pleural and/or pericardial) pain with or without nonproductive cough occurs in 50% of patients.

    • Fever occurs in 45% of patients.

    • Leg swelling (25%) typically occurs after 1-2 years of progressive arthralgias and diarrhea.

    • Glossitis occurs in 20% of patients.

    • Abdominal fullness (ie, ascites and/or splenomegaly) occurs in 20% of patients.

  • Fewer than 10-15% of patients with Whipple disease eventually develop clinically significant CNS involvement. A review of patients with CNS-WD showed that by the time of CNS disease onset they often report previous systemic problems, including the following:

    • Chronic migratory arthralgias or polyarthralgias, sometimes for several years, preceding the onset of neurological symptoms (48% of patients)

    • Unexplained weight loss (46% of patients)

    • GI complaints (45% of patients), including chronic diarrhea in 39%, abdominal pain in 20%, steatorrhea in 13%, and abdominal distension in 8%

    • Fever of unknown origin (40%)

    • Malaise (29%)

    • Night sweats (4%)

    • Blurry vision (2%)

  • GI complaints usually precede CNS-WD symptoms by several years. A limited group of patients has manifestations other than GI at onset, and they seem to have a higher likelihood of developing CNS involvement. CNS symptoms are more frequent at the time of relapse (60-70% of relapse patients, most of whom do not have recurrence of intestinal symptoms).

  • CNS symptoms: Fewer than 100 cases of patients with confirmed CNS-WD have been reported. Most of these patients presented for a combination of neurological and psychiatric symptoms. Most commonly the symptoms included cognitive changes, movement disorders (eg, myoclonus), hypothalamus-related problems (eg, polydipsia, hyperphagia, decreased libido), and seizures.

  • Altered mentation was the most frequent presenting symptom, seen in almost 75% of patients with CNS-WD.

    • Almost 50% of these patients had a concomitant psychiatric illness (eg, depression, hypomania, anxiety, psychosis; mostly with delusional content, or change in personality).

    • Two thirds of patients with altered mentation (almost 50% of all patients with CNS-WD) had dementia.

    • Other mentation problems were limited loss of memory, change in reasoning, and change in attention span.[17]

  • Hypothalamus-related problems (eg, polydipsia, hyperphagia, decreased libido, amenorrhea, change in sleep-wake cycle with insomnia, but never with isolated somnolence), though clinically significant and largely disregarded by patients, represented initial manifestations of CNS involvement in 30% of patients.

  • Myoclonus, including brief, irregular, jerky movements, represented the presenting symptom in 10-15% of patients.

  • Seizures (ie, simple or complex partial seizures, generalized tonic-clonic seizures, either alone or complicating partial seizures) represented the presenting symptom in about 10-15% of patients.

  • A specific clinical triad noted in CNS-WD includes dementia, vertical ophthalmoplegia, and myoclonus.


Exhaustive physical examination should be performed to assess the extent of extraneuraxial involvement; systems and/or organs usually known to be affected in Whipple disease (ie, GI, cardiovascular, pulmonary, CNS, liver, skin) should be targeted.

  • Signs (other than CNS) seen at presentation are as follows:

    • Hypotension (defined as systolic blood pressure < 110 mm Hg and diastolic blood pressure < 60 mm Hg): This is usually a very late finding and has been described in more than 70% of patients with long-standing GI complaints (usually for several years). In some patients, hypotension is part of their dysautonomia. Another dysautonomic feature described in the late stages of Whipple disease is hypothermia.

    • Peripheral lymphadenopathy (more than 50% of patients)

    • Nonspecific abdominal tenderness (50% of patients)

    • Hyperpigmentation and/or skin photosensitivity (40% of patients; does not represent adrenal failure)

    • Low-grade fever (40% of patients)

    • Cardiac murmurs (more than 30% of patients)

    • Peripheral edema (25% of patients)

    • Splenomegaly (25% of patients)

    • Glossitis (25% of patients)

    • Ill-defined abdominal mass (20% of patients)

    • Pleural and pericardial friction rubs (less than 10% of patients)

    • Ascites (very infrequently chylous): This is rare.

    • Uveitis (2% of patients)

  • In patients with a high likelihood of having Whipple disease (based on historical data and general examination), the comprehensive neurologic examination should target cognition and (eye) movement abnormalities, especially signs with high pathognomonic value (eg, vertical supranuclear ophthalmoplegia [SNO] with or without oculomasticatory myorhythmia [OMM][18] and/or oculofacialskeletal myorhythmia [OFSM]).

  • Although dementia, ophthalmoplegia, and myoclonus represent a highly specific diagnostic triad, it is encountered in only approximately 10% of patients with CNS-WD.

    • Dementia - Present in more than 50% of patients with CNS-WD

    • Psychiatric illness (ie, depression; hypomania; anxiety; psychosis, mostly with delusional content; change in personality) - Present in almost 50% of patients

    • Ophthalmoplegia - Present in 50% of patients

    • Myoclonus - Present in 20% of patients

  • Eye findings in CNS-WD are as follows:

    • Supranuclear ophthalmoplegia

      • Both horizontal and vertical SNO are found in approximately two thirds of patients with SNO.

      • Pure vertical SNO is found in approximately one third of patients with SNO.

      • Pure horizontal SNO is never found in CNS-WD.

    • OMM and OFSM are synchronous movements combining pendular vergence oscillations (PVOs) with myorhythmia.

      • PVOs are characterized by 1-Hz, rhythmic, smooth, continuous, convergent eye movements (unilateral or bilateral), varying from 1-25 degrees of amplitude per eye. Return divergent movements never go beyond the primary position. The oscillations continue during sleep and may be subtle and asymmetric. Convergent and return divergent movements happen at the same speed (this does not represent nystagmus, as both phases are of the same speed) and are not accompanied by miosis or accommodation.

      • Myorhythmia represents regular, repetitive contractions (1-2 Hz) of the facial, masticatory, and pharyngeal muscles with or without limb involvement. It persists through sleep and should be distinguished from oculopalatal myoclonus, which most commonly is caused by brainstem strokes and demyelination with a similar average frequency, although it is slightly irregular.

      • PVOs are synchronous with myorhythmic movements of the masticatory muscles and in some patients may include facial and pharyngeal muscles (in the case of OMM) and more extensive myorhythmic movements involving facial and limb muscles (in the case of OFSM).

      • OMM and OFSM are very infrequent signs (< 20 cases have been reported in the literature), and they have been described only in CNS-WD. In all reported cases, OMM and OFSM always have been associated with vertical SNO. Although OMM and OFSM are highly specific for CNS-WD, they have a very low sensitivity of approximately 20%.

      • OMM/OFSM together with vertical SNO are highly pathognomonic for CNS-WD and should prompt timely diagnosis and treatment.

    • Isolated ptosis is found in 23% of patients.

    • Pupillary abnormalities (usually anisocoria or unreactive pupil) are found in only 18% of patients.

  • Other neurologic signs are as follows:

    • Hypothalamic manifestations (in 30% of patients) included polydipsia, hyperphagia, decreased libido, amenorrhea, changes in sleep-wake cycle, and disruptive insomnia. Somnolence as an isolated symptom has never been encountered. Hypothalamus-driven hypopituitarism as a presenting feature was reported in one patient with Whipple disease at relapse, with an MRI-proven rostral infundibular lesion and low levels of cortisol, free testosterone, and free thyroxine without an elevated thyroid-stimulating hormone.[19]

    • Ataxia (cerebellar) is found in 25% of patients.

    • Seizures are found in 25% of patients (ie, simple or complex partial seizures, generalized tonic-clonic seizures, either alone or complicating partial seizures).

    • Segmental myoclonus (which is irregular/nonrhythmic, to be distinguished from myorhythmic movements of OMM and OFSM) is found in 20% of patients.

    • Sensory deficits (usually in a central distribution) are found in 10% of patients.

    • Isolated cranial nerve (CN) palsy (CN III, IV, V, VI, or VII) is very rare (< 3% of patients).


Whether the clinical manifestations of Whipple disease result from direct bacterial invasion or from the ensuing inflammatory response is not yet clear.

More effective (ie, sensitive) diagnostic techniques (eg, PCR) have continued to provide more and more evidence of direct bacterial invasion at the various symptomatic-target organ sites, suggesting a combined mechanism of bacterial invasion and ensuing inflammatory response.





Laboratory Studies

Routine laboratory values are often abnormal. Initial laboratory studies routinely obtained include the following:

  • Baseline CBC with differential, platelets, and erythrocyte indexes

  • Electrolytes - Sodium, potassium, chloride, calcium, magnesium, iron

  • Vitamin levels (usually folate and vitamin B-12 with its metabolic derivatives, homocysteine and methylmalonic acid); thiamin (vitamin B-1), pyridoxine (vitamin B-6), and carotene (provitamin A) levels may help identify rapidly reversible deficits and rule out alternative diagnoses.

Other baseline labs should target some of the differential diagnosis variants and include rapid plasma reagent and HIV testing, erythrocyte sedimentation rate, complement C3 and C5, antinuclear antibody, and angiotensin-converting enzyme levels. Also obtain baseline liver function tests, thyroid function tests, and random cortisol level. Extend endocrinologic workup (free testosterone, luteinizing hormone, follicle-stimulating hormone) on an individual basis to rule out pituitary gland or, more commonly, hypothalamic involvement.

Nonspecific results supportive of steatorrhea include decreased serum cholesterol and carotene levels, decreased iron levels, elevated prothrombin time, and low serum albumin as a result of a combination of the protein-losing enteropathy, decreased absorption of amino acids, and decreased hepatic synthesis.

Multifactorial normocytic hypochromic anemia is present in 90% of patients and is caused by chronic infection and malabsorption with decreased iron, vitamin B-12, and folate.

Not infrequently, thrombocytosis is present with counts of greater than 1 million/mL. This may be the basis for a hyperviscosity syndrome (with exacerbation of cardiac disease and stroke). Both anemia and thrombocytosis reverse with specific antibiotic treatment for Whipple disease.

Imaging Studies


Of patients with Whipple disease with GI upset, 85% have abnormal results on upper GI series. More than 50% of these patients present with bowel dilation with or without prominent mucosal folds of the duodenum and jejunum. In 60% of these patients, flocculation and segmentation of barium also is reported.

Other findings include dilatation of the proximal ileum, stomach with thickened nodular folds, and possibly edema of the colon on barium enema. All of these findings are nonspecific; they also may be found in other diseases such as celiac sprue and lymphoma.

Various imaging studies reveal enlarged retroperitoneal and mediastinal lymph nodes in 10% of patients with Whipple disease.


Neuroimaging is largely nonspecific. CT scan and MRI with and without contrast may reveal nondiagnostic abnormalities including atrophy, hydrocephalus, mass lesions with contrast enhancement, ring-enhancing lesions, and other focal changes, including white matter changes with no mass effect, suggesting demyelination.

Other Tests

Specific tests for malabsorption and biliary salts metabolism

Specific tests indicative of steatorrhea (90% of patients with Whipple disease have malabsorption) include qualitative or 72-hour quantitative stool analysis and absorption of D-xylose (with decreased fractional absorption detected by low serum and urinary levels).[15]

Bile salt metabolism testing reveals an increase in the deoxycholic acid-to-cholic acid ratio after intravenous challenge with taurocholate (ie, the inverse of the normal response).

Electrodiagnostic testing

No diagnostic/specific EEG pattern has been described for CNS-WD. EEG may be normal, show focal or generalized slow wave activity, or identify irritative foci with spike activity.

Electromyography (EMG) of the muscles involved in OFSM has revealed 400-millisecond bursts of bilateral rhythmic activity. The activity originates from the level of CN VII and spreads rostrally to involve the muscles of mastication and caudally to involve muscles of the neck, arms, and legs.

Cerebrospinal fluid studies

CSF is frequently abnormal (high protein, > 5 WBC, intrathecal synthesis of IgA). In patients who carry a diagnosis of Whipple disease, nonspecific inflammatory markers in CSF have been reported as useful surrogates to monitor response and durability of remission. More specific testing of the CSF is required for a positive diagnosis, such as PAS staining, EM, and PCR.

CSF PCR should be considered in all patients with Whipple disease; as it gains acceptance as a useful tool for assessing CNS involvement, it represents the most sensitive tool in evaluating response to antibiotic treatment and early detection of relapse.[20] Termination of a long course antibiotic treatment in a patient with Whipple disease who experienced a good clinical response should be based ultimately on CSF PCR analysis alone.


The diagnosis of Whipple disease is made by demonstrating characteristic lesions in tissue obtained from biopsy of significant organs (defined as organs usually known to be involved in Whipple disease and responsible for some of the complaints and/or signs noted on examination).

The medical practitioner should plan significant organ biopsy with PAS staining, EM analysis, and PCR of the tissue sample.

For patients with GI complaints, peroral GI biopsy should be the initial diagnostic method of choice. It reveals foamy macrophages containing PAS-positive, gram-positive bacilli in the lamina propria of the mucosa of the small intestine.

Studies reviewing patients with Whipple disease and GI complaints revealed only a 50% sensitivity of multiple (2-3) peroral GI biopsies. The review of patients with CNS-WD revealed a higher sensitivity of 70%, which is due in part to a sampling effect, since these patients are less likely to be diagnosed with CNS-WD without a positive peroral GI biopsy.

Consider all patients with CNS-WD for a peroral GI biopsy as part of their initial diagnostic workup.

PAS-positive macrophages in the lamina propria of the colon and rectum are nondiagnostic, since they have been reported in healthy people and in patients with histiocytosis, melanosis coli, or pneumatosis intestinales.

PAS-positive macrophages in lymph node biopsy alone, not corroborated by other significant tissue biopsy results, is not diagnostic, since this can be a common finding in tuberculosis, sarcoidosis, Gaucher disease, and berylliosis.

Lumbar puncture: CNS involvement can be demonstrated quite well with CSF analysis; PAS staining of cellular material may be adequate, or the diagnosis may be confirmed with PCR when PAS staining is inconclusive.

Brain biopsy[21] is a last resort in a patient with a nonspecific focal lesion amenable to biopsy in whom the diagnosis cannot be made by testing other sites (eg, GI biopsy) or by CSF analysis.

Histologic Findings

Macrophages containing PAS-positive, gram-positive bacilli have been found in various body tissues of patients with Whipple disease, including heart and heart valves, CNS and CSF, lung, spleen, Kupffer cells (ie, cells with functional role of macrophages in the liver), pancreas, muscle, bone marrow, kidney, lymph nodes, and synovial membranes.

  • The PAS-positive material found in macrophages represents remnants of the cell walls of the phagocytosed bacilli. These bacilli stain gram positive and acid-fast negative. Whipple disease bacillus has a distinct wall and an outer capsule and measures approximately 0.5-1.5 mm. These bacilli, despite typically being regarded as intracellular pathogens, also have been found as PAS-positive, rod-shaped bacteria in the extracellular environment.

  • CNS-WD has a predilection for the periaqueductal gray matter, hypothalamus, hippocampus, basal ganglia, cerebellum, and cerebral cortex. Whipple disease bacillus usually is found at the site of involvement along with a specific inflammatory response.

  • The gross pathologic features of CNS-WD are generalized atrophy and small, chalky nodules or granulomas as large as 2 mm in diameter scattered diffusely in the grey matter of the cerebral and cerebellar hemispheres and in the periventricular and periaqueductal regions. The changes are patchy and usually are surrounded by totally normal areas of the brain.

  • The involvement of periaqueductal gray matter is thought to be responsible for hydrocephalus observed in some patients.

  • The microscopic appearance of the granulomatous changes reveals strongly PAS-positive macrophages surrounded by large reactive astrocytes. In more advanced cases, the PAS-positive cellular infiltrate may extend in the white matter and may be associated with demyelination and neuronal death with formation of large vacuoles; with more severe extension, the PAS-positive infiltrate may burst into the subarachnoid space. All strongly PAS-positive areas of the brain reveal specific trilamellar appearance of T whippelii bacilli and bacillary debris on EM.



Medical Care

Treatment of Whipple disease is less challenging than its diagnosis. Antibiotics are the mainstay of therapy. A good outcome relies on a timely diagnosis and initiation and completion of a long-term antibiotic course. The literature comprises a consensus about the need for completion of a lengthy antibiotic course of 1-2 years. Treatments of shorter duration have been associated with a high rate of relapse. The use of specific diagnostic techniques (eg, PCR) is important in establishing a diagnosis[22] and in evaluating response and adjusting the antibiotic therapy.

The following guidelines have been proposed for diagnostic screening, biopsy, and treatment of CNS-WD:

  • Definite CNS-WD must have any 1 of the following 3 criteria:

    • OMM or OFSM

    • Positive tissue biopsy

    • Positive PCR analysis

    • If histologic or PCR analysis was not performed on CNS tissue, then the patient also must demonstrate neurological signs. If histologic or PCR analysis was performed on CNS tissue, then the patient need not demonstrate neurological signs (ie, asymptomatic CNS infection).

  • Possible CNS-WD should include 1 of 4 systemic symptoms and 1 of 4 neurologic signs and/or symptoms.

    • Any 1 of the 4 systemic symptoms, not due to another known etiology, as follows:

      • Fever of unknown origin

      • GI symptoms (ie, steatorrhea, chronic diarrhea, abdominal pain)

      • Chronic migratory arthralgias or polyarthralgias

      • Unexplained lymphadenopathy, night sweats, or malaise

    • Any 1 of 4 neurological signs, not due to another known etiology, as follows:

      • Supranuclear vertical palsy

      • Rhythmic myoclonus

      • Dementia with psychiatric symptoms

      • Hypothalamic manifestations (dysautonomia, vigilance problems)

  • In a review of the literature, which represents the basis for these guidelines, 20% of patients with CNS-WD had no systemic symptoms or signs. Another 11% had only "soft" neurological features such as subtle cognitive changes or altered level of consciousness. Thus, even when these guidelines are followed, the diagnosis of some cases of CNS-WD may not be reached premortem.

A combination of antibiotics is preferable, particularly at the initiation of treatment. Antibiotics usually provide rapid resolution of extraneuraxial symptoms. Arthralgias and fever usually resolve within a few days. Diarrhea and malabsorption disappear within 2-4 weeks.

  • CNS symptoms can be limited by initiation of antibiotic treatment but remain difficult to reverse completely. This is especially true for focal deficits accompanied by positive corresponding imaging with obvious structural changes such as granulomas, infarcts, and atrophic changes. The cognitive abnormalities reverse more than other CNS symptoms and signs.

  • In some patients, worsening neurological symptoms have been noted even after onset of antibiotic therapy; this usually requires changing the antibiotic regimen.

Relapse is quite common in Whipple disease; the disease relapses in approximately one third of patients in whom cessation of treatment was made based on negative serial duodenal biopsies alone.

  • Symptoms at relapse usually take place several weeks to several years after treatment was discontinued.

  • Neurologic complications are usually more prominent in patients with relapsed Whipple disease.

  • Relapse commonly involves the CNS in almost all patients with Whipple disease. Patients with previously known CNS involvement have the highest relapse rate.

  • Relapse is usually more difficult to treat than the initial episode. Acquired resistance to previously known efficacious antibiotics has been reported at relapse. A higher rate of relapse occurs in patients treated with a single antibiotic, with antibiotics with very low BBB penetrance, and/or for an inadequate amount of time.

Monitor clinical response to treatment and complement it with other data obtained with biopsy and imaging studies.

  • PCR analysis is a great tool for documenting response. It should target various tissues obtained at biopsy of significant organs and CSF.

  • In all patients with Whipple disease, consider CSF PCR analysis prior to onset of treatment together with follow-up serial studies for monitoring response, for deciding when treatment should be stopped, and even after cessation of treatment for early detection of relapse based on the clinical progress of the patient and physician index of suspicion.

  • No mention is found in the literature as to when to stop antibiotic therapy after the first negative CSF PCR analysis. Most investigators recommend first confirming this result within 2-3 months and continuing the antibiotic therapy for an additional 8-12 months.

Surgical Care

Neurosurgical care is relevant for both obtaining diagnostic biopsy specimens in selected patients and placement of ventriculoperitoneal shunt (VPS) in patients with hydrocephalus.

  • Hydrocephalus

    • Both communicating and noncommunicating hydrocephalus has been reported in patients with CNS-WD. The role of shunting has to be assessed thoroughly.

      • WD bacillus has a predilection for the periaqueductal gray matter, making noncommunicating hydrocephalus (secondary to aqueductal stenosis) more likely in patients with advanced CNS-WD. Patients with aqueductal stenosis represent a neurosurgical emergency and VPS should be placed urgently.

      • In progressive CNS-WD with progressive, communicating hydrocephalus, some of the cognitive and motor deficits noted in these patients potentially could be limited rapidly and sometimes reversed by shunting. This is especially important as the reported reversal of other CNS symptoms and signs remains limited and requires several weeks to months despite otherwise efficient antibiotic treatment.

  • Rapid initiation of efficacious antibiotic treatment prior to, or at the time of, shunting is most important as placement of VPS without concomitant antibiotic treatment potentially could initiate dissemination of disease in the peritoneal cavity. No such cases have been reported in the literature.

  • Placement of a VPS (with Ommaya reservoir) raises an interesting question about a great therapeutic opportunity, intrathecal/intraventricular antibiotic therapy concomitant with systemic treatment. No reports of attempted intraventricular antibiotic treatment have been published. This is especially important as some reports emphasized the meningoependymitis (ventriculitis) pattern seen in patients with CNS-WD.

  • No attempts at concomitant systemic antibiotic treatment together with intrathecal antibiotic treatment through lumbar puncture have been reported.



Medication Summary

Various antibiotics have been used with different efficacies. The most commonly used antibiotics, as reported in the literature, are tetracycline, penicillin, TMP-SMX, and chloramphenicol. Others used include demeclocycline, doxycycline, oxytetracycline, minocycline, gentamicin, streptomycin, amoxicillin, ampicillin, pefloxacin, erythromycin, profloxacin, vancomycin, ceftriaxone, cephalexin, and rifampin.[23]

A consensus is found in the literature about using a combination of parenteral antibiotics at the initiation of therapy ("induction period") for 2-4 weeks, followed by long-term treatment (1-2 y) with an oral antibiotic. Options reported as efficacious for the induction period are a combination of parenteral antibiotics (penicillin and streptomycin) or third-generation cephalosporins such as ceftriaxone for at least 2-4 weeks.

Antibiotics for the induction period

Class Summary

TMP-SMX taken PO 2-3 times per day for long-term treatment (1-2 y) is the agent mostly reported as effective. An alternative reported as efficacious for long-term treatment is cefixime. Some treatment failures and acquired resistance at relapse have been reported with TMP-SMX.

Ceftriaxone (Rocephin)

Third-generation cephalosporin with broad-spectrum, gram-negative activity; lower efficacy against gram-positive organisms; higher efficacy against resistant organisms. Arrests bacterial growth by binding to 1 or more penicillin-binding proteins.

For treatment of WD, induction period of 2-4 wk should be pursued, followed by long-term oral antibiotics; role of induction period has not been studied extensively; it may not be necessary, as long as good combination antibiotic regimen (ie, one with high BBB penetrance) is initiated for long-term use.

Sulfamethoxazole and trimethoprim (Bactrim, Septra, Cotrim)

TMP-SMX inhibits bacterial growth by inhibiting synthesis of dihydrofolic acid.

Both sulfamethoxazole and trimethoprim diffuse into CSF.

For treatment of WD, long-term treatment (1-2 y) should be pursued

Cefixime (Suprax)

Semisynthetic third-generation cephalosporin for oral administration; can be administered in tab or susp form; chemically is (6R,7R)-7-[2-(2-Amino-4-thiazolyl) glyoxylamido]-8-oxo-3-vinyl-5-thia-1-azabicyclo[4.2.0]oct-2-ene-2-carboxylic acid, 72(Z)-[O-(carboxymethyl)oxime]-trihydrate.

Molecular weight is 507.50 as trihydrate.

Most commonly used in treatment of bronchitis, gonorrhea, urinary tract infections, otitis media, pharyngitis, and tonsillitis. Has been reported as successful long-term treatment alternative to TMP-SMX in patients with WD.

Arguably, any antibiotic with good BBB penetrance, benign toxicity profile, and activity against gram-positive bacteria can be used in long-term treatment of CNS-WD.

Approximately 40-50% absorbed from gut whether administered with or without food. Time to maximal absorption is increased approximately 0.8 h when administered with food. Single 200-mg tab produces average peak serum concentration of approximately 2 mcg/mL (range 1-4 mcg/mL); single 400-mg tab produces average peak concentration of approximately 3.7 mcg/mL (range 1.3-7.7 mcg/mL).

PO susp produces average peak concentrations approximately 25-50% higher than tab. Area under time versus concentration curve is greater by approximately 10-25% with PO susp than with tab after doses of 100-400 mg when tested in healthy adult volunteers. Consider this increased absorption if PO susp is to be substituted for tab.

Because of lack of bioequivalence, do not substitute tab for PO susp in treatment of otitis media.

Peak serum concentrations occur 2-6 h following administration of both tab and susp. Approximately 50% of absorbed dose excreted unchanged in urine in 24 h. Also believed to be excreted in bile in excess of 10% of administered dose.

Serum protein binding is concentration independent with bound fraction of approximately 65%.

In multiple dose regimen, little accumulation in serum or urine after dosing for >14 d. Serum half-life in healthy subjects is independent of dosage form and averages 3-4 h but may range to as long as 9 h.

Average AC at steady state in elderly patients is approximately 40% higher than average AC in other healthy adults.

In subjects with moderate impairment of renal function (CrCl 20-40 mL/min), average serum half-life prolonged to 6.4 h. In severe renal impairment (CrCl 5-20 mL/min), half-life increased to average of 11.5 h. Not cleared significantly from blood by hemodialysis or peritoneal dialysis. Patients undergoing hemodialysis have similar blood profiles as subjects with CrCl of 21-60 mL/min.

No evidence of metabolism in vivo. Penetrates BBB but adequate detailed data of correlative CSF levels not available.


Class Summary

Various antiepileptic drugs (AEDs) and benzodiazepines reportedly have been tried to control movement abnormalities encountered in some patients with CNS-WD. Carbamazepine, phenobarbital, and lorazepam demonstrated no notable success in control of OMM and/or OSFM. Valproic acid has been reported as effective in controlling OMM and/or OSFM in some patients.

Valproic acid (Depacon, Depakene, Depakote)

Indicated for seizures and myoclonus; chemically unrelated to other drugs that treat seizure disorders; although mechanism of action not established, activity may be related to increased brain levels of GABA or enhanced GABA action; also may potentiate postsynaptic GABA responses, affect potassium channel, or have direct membrane-stabilizing effect.

For conversion to monotherapy, concomitant AED dosage ordinarily can be reduced by approximately 25% every 2 wk; this reduction may start at initiation of therapy or be delayed by 1-2 wk if concern that seizures may occur with reduction; monitor patients closely during this period for increased seizure frequency.

As adjunctive therapy, divalproex sodium may be added to regimen at 10-15 mg/kg/d; may increase by 5-10 mg/kg/wk to achieve optimal clinical response; ordinarily, optimal clinical response achieved at daily doses < 60 mg/kg/d.

Has been used with some notable success for control of OMM and/or OSFM (reports exist of complete suppression of OMM and/or OSFM) once therapeutic levels were attained (75-100 mcg/mL).

Subtherapeutic levels or attempts to wean in most patients resulted in return of abnormal movements, in some patients even after completion of long-term antibiotic treatment.



Further Outpatient Care

See the list below:

  • Regular visits in various specialty clinics targeting monitoring clinical response to treatment of significant symptoms and laboratory tests should be pursued.

  • Significant symptom inventory should be reviewed with the patient at each clinic visit.

  • Treatment adherence should be emphasized continuously.

  • Guidelines for response assessment should be reviewed consistently with other specialty teams involved in the care of the patient.

  • CSF PCR analysis should be used as the ultimate tool in monitoring response and treatment decision making.

Further Inpatient Care

See the list below:

  • Diagnosis and treatment of patients with Whipple disease should be based on a multiteam approach, targeting early involvement of gastroenterology, neurology, ophthalmology, cardiology, and rheumatology specialists.

  • No patients in whom multiple target-organ routine inventory status has been pursued, unless indicated by symptomatology or clinical examination (except for CNS), have been reported.

  • Some have speculated that by the time the CNS involvement becomes clinically relevant in patients with Whipple disease, they also might have disseminated pulmonary, cardiovascular, hepatic, and/or ocular disease. Furthermore, the choice and duration of antibiotic treatment might be influenced by the presence of disease in these organs.

  • The role of routine echocardiogram and chest and abdomen imaging (CT scan or MRI) remains to be established; these should be pursued on an individual case basis and on the clinician's need to know for significant management decisions and prognostic evaluation.

Inpatient & Outpatient Medications

A long course of antibiotics (more than 1 y) that has good BBB penetrance represents the key in successful treatment of patients with Whipple disease.


See the list below:

  • Limited information suggests that the WD bacillus is a saprobe.

  • A limited number of cases have been reported of patients diagnosed with WD several months after spending vacation time in lake regions.

  • In patients who might have a specific IL-12–gamma-interferon axis defect, swimming in lakes may be hazardous, especially in those where accidental drainage of sewage water took place.


See the list below:

  • Potentially irreversible CNS symptoms have been reported at relapse together with an increased likelihood of onset of resistance to previously efficacious antibiotics.

  • Aggressive CSF PCR monitoring of response to treatment at relapse is of utmost importance.


See the list below:

  • WD left untreated is uniformly fatal.

  • WD may represent a diagnostic dilemma in some cases.

  • For the astute clinician with a high index of suspicion, treating patients with WD could end up being a very rewarding experience. Timely diagnosis and rapid institution of efficacious treatments are paramount in obtaining a potential cure. A long course of antibiotics (over 1 y) which have good BBB penetrance and treatment decisions based on PCR studies of both significant organs (accountable for the symptoms encountered) and CSF are key for successful treatment of WD.

  • In patients treated for less than 1 year, with antibiotics with low BBB penetrance, or without PCR studies to guide treatment decisions, the likelihood of relapse and potentially irreversible neurological deficits is very high (approaching 40%).

Patient Education

See the list below:

  • WD may represent a diagnostic challenge, but treatment is readily available and potentially curative.

  • Patient adherence to a long course of antibiotics is paramount in obtaining a cure; the importance of this should be emphasized repeatedly to the patient.

  • The alternative to poor antibiotic treatment compliance, as a rule, is worsening or early relapse with new or worse and potentially irreversible CNS symptoms. Resistance to previously, clinically proven, sensitive antibiotics also has been reported at relapse.

  • For excellent patient education resources, see eMedicineHealth's patient education article Stroke-Related Dementia.