Medscape	Home   Site Map   Marketplace   My Medscape   CME Center   Feedback   Help Desk

Mycobacteria and Crohn's Disease

Abstract: Crohn's disease is a granulomatous disease of the intestinal tract. Its cause is unknown, but the disease has features in common with Johne's disease, an intestinal infection of animals caused by Mycobacterium paratuberculosis. Molecular biologic techniques have been used to find a link between Crohn's disease and mycobacteria, but so far these studies have yielded conflicting results. [Infect Med 14(7):569-573, 1997]

Interact with Medscape

• Communicate with the Medscape Editor
  • Send email

Contents

• Email this article to a colleague.

• Online Resources
  • Related articles and abstracts

• Abstract & Keywords
• Introduction
• Johne's Disease
• Crohn's Disease
• Does Crohn's Disease Have an Infectious Etiology?

• References

Introduction

Nineteen years later, Burrill B. Crohn and colleagues described a granulomatous disease of the terminal ileum.^[2] Granulomas are focal areas of granulomatous inflammation, the latter defined by a typical pattern of chronic inflammation in which activated macrophages predominate. There are 2 types of granulomas: foreign-body granulomas are initiated by inert foreign bodies, whereas immune granulomas represent a T-cell-mediated reaction to immunogenic particles. Diseases characterized by granulomatous inflammation include sarcoidosis, berylliosis, cat-scratch disease, Hansen's disease, syphilis, brucellosis, and infection with M tuberculosis, nontuberculous mycobacteria, and a variety of fungi and parasites. The presence of granulomas in such idiopathic disease as sarcoidosis continues to suggest exogenous etiologies, and the prototype of M tuberculosis leads researchers to continue to look for other mycobacteria as potential causes.

Crohn, Ginsburg, and Oppenheimer^[2] described 14 patients (ages 17-52 years, average age 32 years) with a chronic illness characterized by "fever, diarrhea, continuous loss of weight, and a progressive anemia." The most frequent physical findings were described as: "(1) a mass in the right iliac region; (2) evidences of fistular formation; (3) emaciation and anemia; (4) the scar of a previous appendectomy; and (5) evidences of intestinal obstruction." Pathology was limited to the terminal ileum and consisted primarily of oval mucosal ulcerations; examination showed acute, subacute, and chronic inflammation, accompanied by occasional areas of focal inflammation with giant cells. The authors assumed these to be foreign body granulomas.

Although a tuberculous etiology was considered, none of the pathologic specimens was positive by smear or culture, and none of the patients had a history of pulmonary tuberculosis. What later came to be known as Crohn's disease was thought to be distinct from intestinal tuberculosis, which occurs when the tubercle bacillus invades the bowel. In the prechemotherapy era, the majority of cases accompanied overt pulmonary disease, but today the majority have no evidence of pulmonary involvement. Pathology consists of granulomas with giant cells and caseation necrosis. Acid-fast stains usually demonstrate the infecting organism, and cultures are positive. As described above, Crohn's disease is characterized by ulceration and fistula formation. The granulomas usually do not have caseation necrosis. Acid-fast stains and cultures are negative. For these reasons, M tuberculosis has not been considered as the etiology of Crohn's disease.

These 2 diseases--Crohn's disease (human) and Johne's disease (animal)--share some common features. In recent years, the possibility of a mycobacterial etiology of Crohn's disease has again been raised. Modern molecular techniques such as polymerase chain reaction (PCR) have been employed in an attempt to answer this old question. This article reviews the evidence that M paratuberculosis may play a role in Crohn's disease.

Johne's Disease

There are 3 stages of the disease. Stage I is subclinical, with no detectable shedding of the microorganism. Stage II is also subclinical, but with detectable mycobacterial shedding. In stage III, animals develop intractable diarrhea and massive mycobacterial shedding. In experimental infection, most animals develop stage II disease; only a minority progress to stage III.^[3]

Johne's disease is widespread. Nationwide, it has been shown to infect 2.9% of US dairy herds in one study.^[5] In a Pennsylvania slaughterhouse survey, 7.2% of the cattle were culture positive for M paratuberculosis. It is estimated that at least twice as many are infected as can be identified by current diagnostic techniques.^[6] In Wisconsin, it is estimated that 10% of dairy cattle and 30% of dairy herds are infected with M paratuberculosis.^[7] Infection usually spreads from mother to calf via infected colostrum and milk,^[8,9] but it is also spread prenatally.^[7] Theoretically, humans can acquire the microorganism through consumption of contaminated water or dairy products. Two reports demonstrate that under certain non-industrial laboratory conditions, pasteurization may not completely eradicate M paratuberculosis from milk.^[10,11] A report from England found PCR evidence for the presence of M paratuberculosis in 7% of pasteurized milk samples obtained from retail outlets.^[12] The authors were unable to determine if these were viable or inactivated organisms.

M paratuberculosis is an acid-fast microorganism ranging from 0.5 to 1.5 micrometers in length. It is extremely slow-growing and may require up to 4 months of incubation to form visible colonies on solid media.^[8,13] It is fastidious in its growth requirements and often requires unique molecules, known as mycobactins, for growth. Mycobactins are high-molecular-weight complex lipids with alkyl side chains of different lengths and a core to which ferric iron (Fe³⁺) is coordinately linked. Mycobactins are found in the cell membranes of Mycobacterium and Nocardia species. Together with small proteins found in the extracellular fluid known as exochelins, mycobactins facilitate the transport of iron across the bacterial cell wall.^[6] Exochelins, made by the mycobacterial cell, are released extracellularly and remove ferric (Fe³⁺) iron from animal transport molecules, transferrin, or lactoferrin. These exochelins then transport the iron across the cell wall where the intracellular mycobactin receives it. The possession of both mycobactins and exochelins is a property unique to mycobacteria. M paratuberculosis requires mycobactin supplementation for cultivation in vitro.^[6]

Dependence on mycobactin supplementation for in vitro growth was once thought to be pathognomonic for M paratuberculosis, but some strains of Mycobacterium avium and the wood pigeon mycobacterium, M avium subspecies silvaticum, are also mycobactin-dependent.^[14,15] M paratuberculosis is closely related to M avium and the wood pigeon mycobacterium; there is nearly 100% homology between these strains by DNA-DNA hybridization analysis.^[16]

Crohn's Disease

Does Crohn's Disease Have an Infectious Etiology?

In 1984, Chiodini and others isolated a mycobacterium from 3 of 14 patients with Crohn's disease.^[20,21] These strains--designated Linda, Ben, and Dominic--have become the prototype strains of mycobacteria associated with Crohn's disease. Inoculation of infant goats with the Linda strain produced a granulomatous disease of the ileum accompanied by lesions that were similar to those seen in Crohn's disease.^[22] DNA-DNA hybridization studies later showed this organism to be M paratuberculosis.^[23] However, the organism has only been isolated from a small percentage of patients with Crohn's disease--2% to 15%.^[3] This might reflect either its fastidious nature and prolonged incubation time or the fact that it is merely a commensal.

PCR technology has greatly enhanced the ability of researchers to detect M paratuberculosis in clinical specimens, but this method has yielded conflicting data from different laboratories (Table I). Using PCR based on the 5' region of IS900 (a multicopy genomic DNA insertion element that is highly specific for M paratuberculosis), Sanderson and colleagues^[24] identified the microorganism in 65% of patients with Crohn's disease, 4.3% of patients with ulcerative colitis, and 12% of control patients. The IS900 does not cross react with other mycobacteria, even M avium. Similar results were obtained by Dell'Isola and coworkers,^[25] who detected M paratuberculosis in 72% of patients with Crohn's disease, 33% of patients with ulcerative colitis, and 24% of control patients. Results from these studies suggest a strong statistical association between the microorganism and Crohn's disease. On the other hand, Suenaga and associates,^[26] also using the IS900 sequence, detected M paratuberculosis in 100% of Crohn's patients, 61% of patients with ulcerative colitis, and 87.5% of controls. The high percentage of carriers in the control population cast doubt on any etiologic role for the microorganism. A recent study by Dumonceau and coworkers^[27] also failed to find PCR evidence for M paratuberculosis in any sample from patients with Crohn's disease or ulcerative colitis, or from controls. Using the 16S rRNA gene--which is detected in all mycobacterial species (with the exception of M nonchromogenicum)--as a probe, they found approximately 50% positivity in all 3 groups.^[27] This latter finding implies that commensal mycobacteria may be fairly common in humans.^[27] Cell-wall-deficient spheroplasts, some of which were later identified by PCR as M paratuberculosis, have been detected in specimens from Crohn's patients.^[28] This provides further evidence that M paratuberculosis is associated with at least some cases of Crohn's disease.

To date, the significance of the PCR evidence is unclear. Why these studies yielded different results is uncertain. All used the 5' IS900 probe, all used tissue from surgical intestinal resections or biopsies, and all took particular care to avoid environmental contamination. Is M paratuberculosis truly associated with disease, or is it simply a ubiquitous gastrointestinal commensal? Even if a strong statistical association between Crohn's disease and M paratuberculosis were to be found, we would still need to know whether the association implied a pathologic or a commensal role for the organism.

Immunologic techniques have also been used to explore potential links between mycobacteria and Crohn's disease (Table II). Kreuzpaintner and others^[29] demonstrated that antibodies to the mycobacterial 45/48 kilodalton doublet antigen, shared by M tuberculosis and M paratuberculosis, were found in 64.7% (17) of Crohn's patients, 10% (10) of ulcerative colitis patients, 5% (20) of patients with colon cancer, and none of the 10 healthy controls. Notably, 5 of the Crohn's patients showed a reduced antibody response 180 days after undergoing intestinal resection for control of their symptoms. Two of these had levels that were undetectable.

On the other hand, the study by Tanaka and associates^[30] showed no difference in antibody response to M paratuberculosis protoplasmic antigens among Crohn's patients, ulcerative colitis patients, and healthy controls. Humoral antibody responses to mycobacteria are common. The strength of the response depends on the size of the bacterial load and on the health of the patients' pre-exposure immune function. A variety of studies of anti-mycobacterial antibody response in Crohn's patients have been performed. No consistent pattern has been discovered.

Guidelines for clinicians. No recommendation for antimycobacterial agents in the treatment of Crohn's disease can be made. In 1989, Hampson and colleagues^[31] published a pilot study in which 20 patients with active Crohn's disease were treated with rifampicin, ethambutol, isoniazid and pyrazinamide, or clofazamine for 9 months. At the completion of this project, 10 patients were in remission. The authors believed these results suggested that quadruple antimycobacterial therapy for Crohn's disease should be studied in controlled trials.

In 1994, Prantera and coworkers^[32] published a placebo-controlled, double-blind trial to determine the efficacy of antimycobacterial drugs in maintaining remission of Crohn's disease. In this study, 40 steroid-dependent Crohn's disease patients were randomized to receive 2 months of tapering steroids plus a 1-time dose of rifampin and 9 months of ethambutol, clofazimine and dapsone, or placebo. The treatment group was more likely to remain in remission than the placebo group. Three of 19 patients on active drug treatment relapsed, while 11 of 17 patients on placebo relapsed during the study period. The authors concluded that this treatment regimen was effective for some Crohn's disease patients for the relief of symptoms and the maintenance of remission. Although studies have shown patient improvement on various antibiotic regimens, it is unclear whether this is the result of an effect on a putative microorganism or an anti-inflammatory effect characteristic of many antibiotics.

Summary

About the Authors

Online Resources

• Search Medscape's full text articles and MEDLINE

• Clinical Abstracts - Molecular Epidemiology of Crohn's Disease

Table

Table I - Polymerase Chain Reaction Studies of the Link Between M paratuberculosis and Crohn's Disease

*	IS900 is specific for M paratuberculosis; 16S rRNA detects Mycobacteria species.
UC = ulcerative colitis.

Table II - Serologic Studies of the Link Between M paratuberculosis and Crohn's Disease

*	Tanaka and colleagues used an ELISA technique and reported mean optical densities in patients with colon cancer, UC, and controls; therefore, there are no numbers of "positive" or "negative" patients.
UC = ulcerative colitis.

Drugs Mentioned in This Article

Home   Site Map   Marketplace   My Medscape   CME Center   Feedback   Help Desk

Medscape Search Options Clinical Content News Info for Patients Medical Images MEDLINE AIDSLINE Drug Info Bookstore Dictionary Whole Web Dow Jones Library ($) Select a database to search, enter a search term, then click “go.”    Advanced Search Forms

All material on this website is protected by copyright. Copyright © 1994-2000 by Medscape Inc. All rights reserved. This website also contains material copyrighted by 3rd parties. Medscape requires 3.x browsers or better from Netscape or Microsoft.