Search

Background & Aims: An impaired monocyte function and impaired interferon (IFN)-γ production has been suggested as a possible pathogenetic factor in Whipple's disease (WD) and as a cause for the delayed elimination of Tropheryma whipplei in some patients. Methods: We studied, in a series of 20 WD patients with various degrees of disease activity, cellular immune functions. Results: We found an increased in vitro production of interleukin (IL)-4 by peripheral mononuclear blood cells as determined by enzyme-linked immunosorbent assay, but reduced secretion of IFN-γ and IL-2 as compared with age- and sex-matched controls. In addition, we observed a significantly reduced monocyte IL-12 production in response to various stimuli in WD patients whereas other cytokines were comparable with controls; these immunologic alterations were not significantly different in patients with various disease activities. At the mucosal level, we found decreased CD4 T-cell percentage and a significantly impaired IFN-γ secretion. Conclusions: Our data define a defective cellular immune response in a large series of WD patients and point to an important pathogenetic role of impaired Th1 responses. The decreased monocyte IL-12 levels may result in reduced peripheral and mucosal IFN-γ production and lead to an increased susceptibility to T. whipplei infection in certain hosts. GASTROENTEROLOGY 2002;123:1468-1477

BACKGROUND & AIMS: Diagnostic procedures in Whipple's disease usually focus on the intestine, but symptomatic central nervous system involvement is a major threat for patients. The aim of this study was to determine the diagnostic value of cerebrospinal fluid (CSF) analysis. METHODS: A total of 39 CSF samples and 2 brain biopsy specimens that were obtained from 24 patients with Whipple's disease at various intervals after diagnosis were examined. Five patients presented with neurological symptoms, 3 of them as relapses after therapy. Thirty-two CSF samples were examined by polymerase chain reaction for Tropheryma whippelli and 20 CSF samples by cytology. Brain biopsy specimens were examined histologically. RESULTS: Positive results were obtained in 4 of 5 patients (80%) with neurological symptoms, in 7 of 10 patients (70%) without neurological symptoms examined before therapy, and in 3 of 11 patients (27%) without neurological symptoms studied during or after therapy. Conversion from positive to negative was observed in 4 patients after antibiotic treatment. CONCLUSIONS: Testing of CSF in Whipple's disease yields a high rate of positive results, even in patients without neurological symptoms. Examination of CSF is therefore potentially useful for initial staging and for monitoring of the efficiency of therapy. (Gastroenterology 1997 Aug;113(2):434-41)

BACKGROUND & AIMS: The uncultured Whipple's disease bacterium (Tropheryma whippelii) was characterized in 1991-1992 by polymerase chain reaction (PCR) and sequencing of the bacterial 16S ribosomal RNA gene. The aim of this study was to develop a PCR assay for diagnostic purposes. METHODS: Modified primers for PCR and a specific probe for hybridization were designed. The specificity of this PCR assay was tested using 37 bacterial control strains and intestinal biopsy samples from 16 patients without Whipple's disease. The sensitivity was tested in 88 intestinal biopsy samples from 35 patients with Whipple's disease. RESULTS: PCR and hybridization were negative in all 37 bacterial controls and in all 16 patients without Whipple's disease. Before therapy, DNA of T. whippelii was detected in all 30 patients with Whipple's disease from whom formalin-fixed biopsy material was available, whereas Bouin-fixed material was negative. During and after treatment, PCR was negative in 23 of the 24 patients who were followed up. Generally, conversion to negative occurred within 1 year. Despite negative intestinal PCR, symptomatic cerebral Whipple's disease appeared in 3 patients. CONCLUSIONS: This PCR assay is specific and sensitive and is applicable as a diagnostic test. However, PCR from intestinal biopsy samples seems less helpful for monitoring the effect of treatment. (Gastroenterology 1996 Jun;110(6):1735-43)

A 28-year-old carpenter presented with diarrhea and abdominal pain for the past 4 months. He reported the passage of 4-6 loose stools daily, associated with crampy midabdominal pain and bloating. He also reported a 5-year history of fever. This occurred primarily in the evenings, with temperatures sometimes reaching 103’F associated with drenching sweats. In addition, there was a 5-year history of fatigue and migratory arthralgias involving the proximal interphalangeal joints, wrists, elbows, knees, ankles, tarsal areas, and metatarsophalangeal joints. The arthralgias would usually only involve a single joint at a time with pain lasting 1-4 days in a given joint area. He had been treated with enteric-coated aspirin up to 4 g/day, as wel1 as with prednisone in a dose ranging from 5 to 10 mg daily, with excellent suppression of the arthralgias, although not complete relief of the fevers. He denied use of other medications. He was married, and there was no history of sexually transmitted disease or illicit drug use. His father was Italian and mother German; there was no family history of gastrointestina1 disease. The review of systems revealed no history of gastrointestinal bleeding, weight loss, skin rash, or other symptoms. On examination, his temperature was 36.9”C. He was muscular and healthy appearing. The vita1 signs were normal. The skin and mucous membrane examinations did not reveal any abnormalities. NO pallor or clubbing was noted. Abdominal examination revealed no masses or tenderness, and recta1 examination was normal. There were several smal1 nontender axillary nodes bilaterally and several less prominent inguinal nodes, but no cervical lymphadenopathy. On cardiac auscultation, an inconstant soft systolic click was heard, but no murmurs or rubs were detected. Neurological examination was normal. Laboratory results were as fellows (normal values in parentheses): hemoglobin, 9.1 g/dL (12.9-16.6); mean corpus9.2 X lO”/L (4.1-10.9) with normal differential; platelet count, 593 X 109/L (184-370); erythrocyte sedimentation rate, 52 mm/h (0-22); reticulocytes, 1.1% (0.6-1.8); periphera1 blood smear showed microcytic red blood cells and slight abnormalities including regenerating macrocytes, schizocytes, keratocytes and stomatocytes; hemoglobin electrophoresis normal; serum iron, 9 pg/dL (50-150); total iron binding capacity, 266 pg/dL (250-400); iron saturation, 3% (14-50); vitamin B,,, 367 ng/L (281-1079); serum folate, 7.0 pg/L (2-20); sodium, 137 mEq/L (135145); potassium, 4.9 mEq/L (3.6-4.8); calcium, 9 mg/dL (8.9-10.1); phosph orus, 4.6 mg/dL (2.5-4.5); total protein, 6.1 g/dL (6.3-7.9); glucose, 78 mg/dL (70-100); alkaline phosphatase, 159 U/L (98-25 1); aspartate aminotransferase, 25 u/L (12-31); total bilirubin, 0.4 mg/dL (0.1-1.1); direct bilirubin, 0.1 mg/dL (0.0-0.3); uric acid, 4.2 mg/dL (4.3-8.0); creatinine, 1.0 mg/dL (0.8-1.2); albumin, 3.3 g/ dL (3.5-5.0); total thyroxine, 6.5 pg/dL (5.0-12.5); fecal hemoglobin, 1.6 mg/g stool (0-2); rheumatoid factor, ~30 IU/mL (0-39); antinuclear antibody negative; anti-double stranded DNA, 53 U (0-70); and rapid plasma reagin nonreactive. Stool for parasites was negative; there were some fatty crystals. Cultures of urine and blood were negative. Chest and spine radiographs were normal. An upper gastrointestinal barium series showed a normal esophagus and stomach but minimally thickened duodenal and proximal jejunal folds. A barium enema showed slightly prominent “lymphoid follicles” in the cecum and ascending colon.

Background & Aims Whipple's disease is a chronic infection caused by the actinomycete Tropheryma whipplei . We conducted a randomized controlled trial of the efficacy of antimicrobials that are able to cross the blood-brain barrier and to which T whipplei is susceptible. Methods Patients from central Europe with previously untreated Whipple's disease (n = 40) were assigned randomly to groups given daily infusions of either ceftriaxone (1 × 2 g, 20 patients) or meropenem (3 × 1 g, 20 patients) for 14 days, followed by oral trimethoprim–sulfamethoxazole for 12 months. The primary outcome measured was maintenance of remission for 3 years, determined by a composite index of clinical and laboratory data as well as histology. Results All patients were observed for the entire follow-up period (median, 89 mo; range, 71–128 mo); all achieved clinical and laboratory remission. Remission was maintained in all patients during the time of observation, except for 2 who died from unrelated causes. A single patient with asymptomatic cerebrospinal infection who was resistant to both treatments responded to chloroquine and minocycline. The odds ratio for the end point (remission for at least 3 years) was 0.95 (95% confidence interval, 0.05–16.29; P = 1.0). Conclusions This was a randomized controlled trial to show that treatment with ceftriaxone or meropenem, followed by trimethoprim–sulfamethoxazole, cures patients with Whipple's disease. One asymptomatic individual with infection of the cerebrospinal fluid required additional therapy.

Background & Aims Whipple's disease is a chronic multisystemic infection caused by Tropheryma whipplei . Host factors likely predispose for the establishment of an infection, and macrophages seem to be involved in the pathogenesis of Whipple's disease. However, macrophage activation in Whipple's disease has not been studied systematically so far. Methods Samples from 145 Whipple's disease patients and 166 control subjects were investigated. We characterized duodenal macrophages and lymphocytes immunohistochemically and peripheral monocytes by flow cytometry and quantified mucosal and systemic cytokines and chemokines indicative for macrophage activation. In addition, we determined duodenal nitrite production and oxidative burst induced by T whipplei and by other bacteria. Results Reduced numbers of duodenal lymphocytes, increased numbers of CD163 + and stabilin-1 + , reduced numbers of inducible nitric synthase+ duodenal macrophages, and increased percentages of CD163 + peripheral monocytes indicated a lack of inflammation and a M2/alternatively activated macrophage phenotype in Whipple's disease. Incubation with T whipplei in vitro enhanced the expression of CD163 on monocytes from Whipple's disease patients but not from control subjects. Chemokines and cytokines associated with M2/alternative macrophage activation were elevated in the duodenum and the peripheral blood from Whipple's disease patients. Functionally, Whipple's disease patients showed a reduced duodenal nitrite production and reduced oxidative burst upon incubation with T whipplei compared with healthy subjects. Conclusions The lack of excessive local inflammation and alternative activation of macrophages, triggered in part by the agent T whipplei itself, may explain the hallmark of Whipple's disease: invasion of the intestinal mucosa with macrophages incompetent to degrade T whipplei .

Background & Aims Whipple's disease is a systemic, chronic, relapsing disorder caused by a combination of environmental ( Tropheryma whipplei ) and unknown host factors. Because it is a rare disease, the association between HLA type and Whipple's disease has been studied in only small numbers of patients; these studies have led to conflicting results. We aimed to investigate whether disease phenotype and outcome are associated with HLA type in 122 patients with Whipple's disease. Methods Genomic DNA was collected from 103 German, 11 Italian, and 8 Austrian patients with Whipple's disease, along with 62 healthy Austrian workers exposed to T whipplei (14 stool samples contained the bacterium). HLA class I and II alleles were identified by polymerase chain reaction analysis. Patient genotypes were compared with those of healthy German and Austrian populations; data for Italian controls were obtained from the Pavia HLA bone marrow donors' bank. Results HLA-DRB1*13 and DQB1*06 alleles occurred significantly more frequently in patients with Whipple's disease but not in healthy individuals who had been exposed to T Whipplei . The cumulative odds ratios for disease were 2.23 for the DRB1*13 allele ( P ) and 2.25 for the DQB1*06 allele ( P ). Conclusions DRB1*13 and DQB1*06 alleles were found to be risk factors in the largest HLA study ever performed in patients with Whipple's disease.

BACKGROUND & AIMS: Whipple's disease (WD) is a systemic infection in which the causative bacteria typically accumulate within macrophages. The aim of this study was to test whether this macrophage dysfunction is the cause or result of previously shown T-cell defects. METHODS: In vitro production of interleukin (IL)-12, IL-10, tumor necrosis factor alpha, interferon gamma (IFN-gamma), and transforming growth factor beta (TGF-beta) from purified monocytes and peripheral blood mononuclear cells, cytokine expression on duodenal biopsy specimens, and serum cytokine and immunoglobulin (Ig) levels were tested in 9 patients with WD. RESULTS: Reduced monocyte IL-12 production and decreased IFN-gamma secretion by peripheral blood mononuclear cells in vitro were found, as well as reduced immunohistological staining for IL- 12 and IFN-gamma, but no decrease in other cytokines in patients with WD. A similar but less severe defect in 2 relatives with WD argued for a genetic basis of this abnormality. Serum IgG2, an IFN-gamma-dependent Ig subclass, and serum TGF-beta levels were reduced in patients with WD. CONCLUSIONS: The described monocyte defects in WD may result in a secondary reduction of IFN-gamma production and IgG2 serum levels. This provides a rationale for additive immunotherapy in patients with antibiotic-refractory WD. (Gastroenterology 1997 Aug;113(2):442-8)

A previously healthy 67-yr-old man presented with progressive dementia over an 11-mo period. Evaluation revealed evidence of malabsorption. Jejunal biopsy established the diagnosis of Whipple's disease. No other etiology for the patient's dementia was uncovered. Treatment with trim ethoprim-sulfamethoxazole resulted in rapid elimination of Whipple's bacilli from the jejunum and complete reversal of the patient's dementia over a 6-mo period. Significant levels of trimethoprim and sulfamethoxazole were easily quantitated in the cerebrospinal fluid during therapy. There is increasing recognition of progressive neurologic disease in patients with Whipple's disease who were treated with tetracycline. The reversal of presumed central nervous system disease in this case suggests that drugs that penetrate the blood-brain barrier might be preferable for the initial treatment of Whipple's disease.

Reports of clinical relapse occurring after apparently successful antibiotic treatment of Whipple's disease prompted this review of long-term follow-up of treated patients. Follow-up of at least 1 yr after completion of treatment or 2 yr after diagnosis was obtained on 88 patients with documented Whipple's disease by a review of the medical literature, correspondence with the authors as needed, and questionnaires mailed to academic gastroenterology programs in the United States. Relapse was defined on the basis of morphology (preferably) or clinically, or both. Thirty-one patients relapsed, 6 of whom relapsed twice. Fifty-seven patients did not relapse. The mean time to relapse was 4.2 yr. The mean follow-up period of patients who did not relapse was 8.2 yr. The type and number of relapses were as follows: clinical, 16; central nervous system, 13; arthralgia, 5; gastrointestinal, 1; and cardiac, 2. The clinical, arthralgia, and gastrointestinal relapses were evenly distributed between early relapses (occurring 2 yr after diagnosis). All cardiac and central nervous system relapses were late. Twenty-one of 49 patients treated with tetracycline alone relapsed. Two relapses were reported in 15 patients treated with penicillin and streptomycin followed by tetracycline. Three relapses developed in 8 patients treated with penicillin alone. Five of the 16 patients treated with other regimens relapsed. Nine of the 13 patients with central nervous system relapse had been initially treated with tetracycline, 2 were treated with penicillin, and 2 were treated with combinations of antibiotics. Results of treatment of central nervous system relapse were poor in 10 of the 11 patients for whom details were available. Results of treatment of non-central-nervous-system relapse were excellent in 19 of 20 patients. It is concluded that tetracycline alone, or penicillin alone, is not adequate initial therapy for Whipple's disease and that central nervous system relapse is resistant to antibiotic therapy. The authors recommend parenteral penicillin and streptomycin followed by 1 yr of oral trimethoprim-sulfamethoxazole therapy or oral trimethoprim-sulfamethoxazole alone for 1 yr as initial therapy for Whipple's disease. Relapse should be defined by demonstration of recurrence of bacilli whenever possible.

We report a patient with acquired immunodeficiency syndrome whose clinical presentation and histologic features resembled Whipple's disease. The unique feature of this case was the absence of Whipple's bacillus and the presence of Mycobacterium avium-intracellulare within macrophages infiltrating the small intestinal lamina propria.

Hepatic epithelioid cell granulomas that were negative for periodic acid-Schiff stain after diastase digestion were found in a 32-yr-old man who presented with painless hepatomegaly and slight fever. The patient never complained of intestinal symptoms, which in part explains why the diagnosis of Whipple's disease was made only 3 mo later, at a time when severe neurologic manifestations had appeared. The definitive diagnosis was made on the basis of the characteristic histologic findings in biopsy material obtained from jejunum and abdominal lymph nodes at laparotomy. The patient's condition, especially neurologic manifestations, rapidly improved after antibiotic therapy. It is noteworthy that Whipple's disease is generally not included among the causes of hepatic epithelioid cell granulomas. It is suggested that its possibility should be considered in patients with hepatic granulomas without obvious etiology, even in the absence of intestinal symptoms.

A patient with Whipple's disease was studied for 56 wk from diagnosis, during which time he received continuous antibiotic therapy. Intramucosal bacillary bodies detected by electron microscopy disappeared within 12 wk and a threefold fall in antibody titer to Hemophilus influenza type B bacillus occurred during this period. Circulating immune complexes of IgG class were consistently detected during the first 28 wk of treatment but not subsequently. IgM class immune complexes were detected at a time when mucosal recovery had occurred and when IgG complexes were no longer detectable. A further rise of IgM immune complexes could be induced by enteric challenge with bovine serum albumin in our patient but not in control subjects. The detection of serum immune complexes in Whipple's disease may reflect the entry of foreign antigen through intestinal mucosa. These observations also support the possibility of an underlying defect of antigen exclusion in this disorder, which persists despite apparent mucosal recovery.

A middle-aged white male presented with weight loss, low grade fever, and increasing abdominal girth. A small right-sided pleural effusion and moderate ascites were noted on physical examination. The pleural fluid reportedly contained an acid-fast bacillus, but it was sterile. The ascitic fluid was a sterile exudate. Following antituberculous therapy his pleural effusion and ascites resolved. Laboratory studies returned to normal. When streptomycin was discontinued, while isoniazide and p-aminosalicylic acid were continued, his previous symptoms, abnormal physical findings, and deranged laboratory studies recurred. Light and electron microscopy of peroral small bowel biopsy established the diagnosis of Whipple's disease. Peritoneoscopy, prior to starting any therapy, disclosed multiple raised nodules over the upper parietal peritoneum. Light microscopic examination of biopsies of the peritoneal nodules revealed diastase-resistant periodic acid- Schiff-positive material. After 7 months of tetracycline therapy, the nodules almost totally resolved. The presence of exudative ascites and multiple small peritoneal nodules or both should alert the clinician to the possibility of Whipple's disease.

Whipple's disease (intestinal lipodys­ trophy), from the time of its first descrip­ tion in 1907/ was considered, universally, to be a fatal disease that could be diagnosed only by laparotomy or at autopsy until about 1950. Such concepts are now known to be erroneous. The rapidly growing literature, during the preceding 10 years, clearly in­ dicates a renewed interest in this unusual pathologic entity. This is largely due to the ease of diagnosis once suspected, to new research techniques, especially peroral small bowel biopsy and electron microscopy, and to the gratifying response to appropriate treatment. Recent reviews by Gross,2 Eng­ land,3 and HoW emphasize the inconsistent therapeutical effect of adrenocorticosteroids and antibiotics and leave one uncertain as to which of these agents was responsible for the recovery observed. Although clinical remission after treatment has been reported in a number of patients, the period of follow­ up has been too brief to justify the conclu­ sion that complete recovery had taken place. In fact, entirely normal tissue obtained by small bowel biopsy after therapy has been reported in only 3 eases. 4 - 6 The purpose of this report is to review our experience in the treatment of this disease with steroids and antibiotics, to show the extent of the clinical, biochemical, radiologic, and histo­ logic remission observed in our patients and finally to present data whieh would allow one to form an opinion as to which of these agents was responsible for recovery.

Five consecutive patients with Whipple's disease exhibited intestinal blood loss at the time of their initial presentation. Three had gross intestinal bleeding and the other 2 had occult bleeding with microcytic anemia. Although not generally emphasized, Whipple's disease needs to be considered in the differential diagnosis of acute and chronic gastrointestinal bleeding.

A patient with Whipple's disease is described, and multiparameter flow cytometric examinations of several of the patient's phagocyte functions 3 and 9 mo after the start of oxytetracycline therapy are reported. Almost no intracellular degradation of Escherichia coli or Streptococcus pyogenes proteins and DNA occurred after ingestion by the patient's monocytes and macrophages. In addition, only minor digestion of phagocytozed zymosan particles was detected. The mononuclear intracellular degradation was equally impaired 3 and 9 mo after the start of therapy. The monocyte and macrophage phagocytosis and intracellular killing, and all granulocyte phagocyte functions tested, were normal. The impaired mononuclear degradation of ingested material that was measured is consistent with the accumulation of periodic acid-Schiff-positive bacterial degradation products seen in macrophages of affected tissues in vivo, and suggests a key role of macrophage dysfunction in the pathogenesis of Whipple's disease.

Three men with Whipple's disease had platelet counts of 729,000-1,142,000 per mm3, which fell to normal as their illness responded to antibiotic therapy. Reports on Whipple's disease and hospital charts from the largest published series reveal thrombocytosis in 11 of the 24 patients whose counts were recorded. The prevalence of anemia, marrow infiltration, and classic manifestations of Whipple's disease did not differ between patients with and without thrombocytosis. The thrombocytosis of Whipple's disease is similar to that of celiac sprue, Crohn's disease, and ulcerative colitis.

Reports of clinical relapse occurring after apparently successful antibiotic treatment of Whipple's disease prompted this review of long-term follow-up of treated patients. Follow-up of at least 1 yr after completion of treatment or 2 yr after diagnosis was obtained on 88 patients with documented Whipple's disease by a review of the medical literature, correspondence with the authors as needed, and questionnaires mailed to academic gastroenterology programs in the United States. Relapse was defined on the basis of morphology (preferably) or clinically, or both. Thirty-one patients relapsed, 6 of whom relapsed twice. Fifty-seven patients did not relapse. The mean time to relapse was 4.2 yr. The mean follow-up period of patients who did not relapse was 8.2 yr. The type and number of relapses were as follows: clinical, 16; central nervous system, 13; arthralgia, 5; gastrointestinal, 1; and cardiac, 2. The clinical, arthralgia, and gastrointestinal relapses were evenly distributed between early relapses (occurring less than 2 yr after diagnosis) and late relapses (occurring greater than 2 yr after diagnosis). All cardiac and central nervous system relapses were late. Twenty-one of 49 patients treated with tetracycline alone relapsed. Two relapses were reported in 15 patients treated with penicillin and streptomycin followed by tetracycline. Three relapses developed in 8 patients treated with penicillin alone. Five of the 16 patients treated with other regimens relapsed. Nine of the 13 patients with central nervous system relapse had been initially treated with tetracycline, 2 were treated with penicillin, and 2 were treated with combinations of antibiotics. Results of treatment of central nervous system relapse were poor in 10 of the 11 patients for whom details were available. Results of treatment of non-central-nervous-system relapse were excellent in 19 of 20 patients. It is concluded that tetracycline alone, or penicillin alone, is not adequate initial therapy for Whipple's disease and that central nervous system relapse is resistant to antibiotic therapy. The authors recommend parenteral penicillin and streptomycin followed by 1 yr of oral trimethoprim-sulfamethoxazole therapy or oral trimethoprim-sulfamethoxazole alone for 1 yr as initial therapy for Whipple's disease. Relapse should be defined by demonstration of recurrence of bacilli whenever possible.

A patient with Whipple's disease was studied for 56 wk from diagnosis, during which time he received continuous antibiotic therapy. Intramucosal bacillary bodies detected by electron microscopy disappeared within 12 wk and a threefold fall in antibody titer to Hemophilus influenza type B bacillus occurred during this period. Circulating immune complexes of IgG class were consistently detected during the first 28 wk of treatment but not subsequently. IgM class immune complexes were detected at a time when mucosal recovery had occurred and when IgG complexes were no longer detectable. A further rise of IgM immune complexes could be induced by enteric challenge with bovine serum albumin in our patient but not in control subjects. The detection of serum immune complexes in Whipple's disease may reflect the entry of foreign antigen through intestinal mucosa. These observations also support the possibility of an underlying defect of antigen exclusion in this disorder, which persists despite apparent mucosal recovery.

We describe an unusual case of Whipple's disease, in a patient previously treated for tuberculosis, in which involvement of the small intestine was restricted to the submucosa. This is of diagnostic importance since the presence of Whipple's disease cannot be established by jejunal biopsy in such cases unless the submucosa is adequately sampled. It is possible that this unusual distribution of intestinal infection is related to the effects of e antibiotic therapy. Retrospective study of the patient's serum cholesterol, carotene, and albumin suggests that malabsorption may occur in Whipple's disease despite the absence of infection in the mucosa.

Whipple's disease is often accompanied by a long, preintestinal phase of vague symptoms, such as weight loss, fever, and migratory arthralgia, which may delay diagnosis and proper treatment. We report a patient who presented with sarcoidlike granulomas in the lung 1.5 yr before the development of gastrointestinal symptoms. He was treated with prednisone and his lung lesions improved dramatically. However, steroids could not be discontinued until the diagnosis of Whipple's disease was made and he was started on antibiotic treatment. Whipple's disease was diagnosed from a small intestinal biopsy specimen by electron microscopic demonstration of characteristic bacillary bodies. Liver biopsy specimens also demonstrated a few Kupffer cells containing degenerative bacillary bodies. Based on this case and other reported cases of Whipple's disease with sarcoidlike lesions in various organs, we suggest that sarcoidlike tissue reaction can be an early manifestation of Whipple's disease, recognition of which may have practical value in facilitating an early diagnosis and treatment.

Extensive electron microscopic observations of 19 intestinal biopsies obtained from 13 patients with untreated Whipple's disease are reported. The outstanding feature is the profuse presence of bacilli free within the lamina propria of the intestinal mucosa and the presence of numerous macrophages containing ingested bacilli. Intestinal epithelial cell invasion by bacilli is identified in 11 patients, and evidence of bacillary invasion is identified within the lymphatic endothelium of 4 patients, within capillary endothelium of 3 patients, within polymorphonuclear leukocytes of 5 patients, and within plasma cells of 2 patients. Reported for the first time is the presence of bacilli within intrinsic smooth muscle of the lamina propria of 2 patients, within intraepithelial lymphocytes of 1 patient, and in mast cells of 1 patient. These observations suggest that the Whipple bacillus is an intracellular pathogen. Intracellular pathogens, unlike pyogenic bacteria, may survive within macrophages. The morphologic appearance of the Whipple bacillus is reviewed. Emphasis is placed upon the unique nature of an outer "membrane" external to the cell wall of Whipple bacilli, a feature that is one of the identifying characteristics of the Whipple bacillus and that may explain the inability to culture the bacillus in vitro.

A previously healthy 67-yr-old man presented with progressive dementia over an 11-mo period. Evaluation revealed evidence of malabsorption. Jejunal biopsy established the diagnosis of Whipple's disease. No other etiology for the patient's dementia was uncovered. Treatment with trimethoprim-sulfamethoxazole resulted in rapid elimination of Whipple's bacilli from the jejunum and complete reversal of the patient's dementia over a 6-mo period. Significant levels of trimethoprim and sulfamethoxazole were easily quantitated in the cerebrospinal fluid during therapy. There is increasing recognition of progressive neurologic disease in patients with Whipple's disease who were treated with tetracycline. The reversal of presumed central nervous system disease in this case suggests that drugs that penetrate the blood-brain barrier might be preferable for the initial treatment of Whipple's disease.

A middle-aged man was found to have Whipple's disease after episodes of nephritis and arthralgia. While on antibiotic therapy, and in the absence of worsening of the histologic appearances of the jejunum, aortic valve endocarditis developed, presumably due to Whipple's disease. Observations during the course of his illness included the isolation of Corynebacterium bovis from an inguinal lymph node, and detection of circulating antibodies against material within the characteristic abnormal macrophages present in the gastrointestinal mucosa. This antigen-antibody reaction was specifically blocked by the monosaccharide rhamnose, a component of the polysaccharide surface coat of many bacteria, including C. bovis.

A patient with Whipple's disease is described whose small bowel biopsy demonstrated antigenicity to rabbit antisera to groups A, B, C, and G streptococci and Shigella flexneri, but not to antisera to pneumococcus or Shigella sonnei by immunofluorescent techniques. In addition, the patient's lymphocytes responded normally to phytohemagglutinin and to concanavalin A. These studies support the idea what Whipple's disease is mediated by a bacteria-like agent which shares certain antigenic similarities with groups A, B, C, and G streptococci and S. flexneri, and that T cell dysfunction does not appear to be an essential prerequisite for the disease.