WFSTO BRAR
`
`PilAY 1 8 2001
`4/120 CLINICAL zliENCE CENTER.
`600 HIGHLAND AVE MADISON WI 63792
`
`Sandoz v. AbbVie
`Sandoz Ex. 1042
`
`Ex. 1042 - Page 1
`
`

`

`Volume 5
`
`Number
`
`May 200
`
`The
`American
`,journal of
`Gastroenterology
`
`CONTENTS
`
`EDrORIALS
`
`1321 p53 and Neoplastic Progression in Barrett's Esophagus
`B. J. Reid
`
`1323 Endosonography for Differentiating Benign From Malignant Intraductal Mutinous Tumors of
`the Pancreas: Is the Jury Out?
`I. Waxman
`
`1325 Unraveling the Mechanisms of Thrombosis in Inflammatory Bowel Disease
`I. E. Koutrottbctkis
`
`1327 Controlling Pain in Liver Biopsy, or "We Will Probably Need to Repeat the Biopsy in a Year
`or Two to Assess the Response"
`S. H. Caklwell
`
`•i"-INHAT'S NEW IN GI
`1331 Coming Soon to Your Local Magnet: MR Enteroclysis
`A. Matamoros, Jr.
`
`1331 Don't Ask (Details), Don't Tell (Details)
`T. M. McCashland
`
`1332 The Outcome TIPS in Favor of Surgical Treatment
`J. S. Thompson
`
`1332 Plasma Citrulline as a Marker of Intestinal Failure in Short Bowel Syndrome
`J. K. DiBaise
`
`(continued)
`
`The American Journal of Gastroenterology (ISSN 0002-9270) is the Official Journal of the American College of Gastroenterology, and is published monthly by
`Elsevier Science Inc., 655 Avenue of the Americas, New York, NY 10010-5107. Periodicals postage paid at New York, NY and at additional mailing offices.
`POSTMASTER Send address changes to The American Journal of Gastroenterology, Elsevier Science Inc., 655 Avenue of the Americas. New York,
`NY 10010-5107. 2001 Subscription Rates: For customers in the U.S. only: Institutional Rate: US$350.00; Personal Rate: US$235.00; Student/Resident
`Rate: US$85.00. For surface mail delivery to customers in Europe: Institutional Rate: NLG 839.00; Personal Rate: NLG 594.00; Student/Resident Rate:
`NLG 307.00. For surface mail delivery to customers in Japan: Institutional Rate: JPY 52,700.00; Personal Rate: JPY 37.300.00; Student/Resident Rate:
`JPY 19,300.00. For surface mail delivery to customers in all other countries: Institutional Rate: US$426.00; Personal Rate: US$302.00; Student/Resident
`Rate: US$156.00, For airmail delivery to customers inn Europe: Institutional Rate; NLG 997.00; Personal Rate: NLG 752.00. For airmail delivery to
`customers in Japan: Institutional Rate: JPY 62,600.00;• Personal Rate: WY 47,300.00. The GST number for Canadian subscribers is 137135919. For
`additional subscription information, contact Elsevier Science Customer Support Department, P.O. Box 945, New York, NY 10010, Phone: 212-633-3730;
`Toll free (for customers in North America): 1-888-4ES-INFO; Fax: 212-633-3680; email: usinfo-f@elsevier.corn. Advertising inquiries should be addressed
`to Iv!. J. Mrvica Associates, 2 West Taunton Ave., Berlin, NJ 08009, Phone: 856-768-9360; Fax: 856-753-0064. For recruitment and classified advertising
`in North and South America: John Baltazar, Elsevier Science Inc., 655 Avenue of the Americas, New York, NY 10010, Phone 212-633-3829; Fax:
`212-633-3820; j.baltazar@elsevier.com. For international advertising (excluding Japan), contact Advertising Department, Elsevier Science, The Boulevard,
`Lang.ord Lane, Kidlington. Oxford, OX5 IGB, UK, Phone: 44-1865-843565; Fax: 44-1865-843976; media@elsevier.co.uk. For advertising in Japan:
`Advertising Department, Elsevier Science Japan, Phone: 81-3-5561-5033; Fax 81-3-5561-5047. Commercial reprints inquiries should be addressed to
`I Anne Rosenthal, Elsevier Science Inc., 655 Avenue of the Americas, New York, NY 10010, Phone: 212-633-3813; Fax: 212-633-3820;
`) arosenthal@elsevier.com. For information on Society Membership, contact: Thomas S. Fise, American College of Gastroenterology. 4900 B South 31st
`1 Stree. Arlington, VA 22206, Phone: 703-820-7400; Fax: 703-931-4520. Indexed by Index Medicus, Current Contents (Clinical Medicine, Science Citation
`I Index, Research Alert, ISI/BioMed), BIOSIS, EMBASE/Excerpta Medica, and Elsevier BIOBASE/Current Awareness in Biological Sciences. Copyright
`2001 by the American College of Gastroenterology.
`
`II
`W
`t
`
`t
`( (cid:9)
`
`1 1 011 111
`11 I 11 (cid:9)
`10 111 I (cid:9)
`0002-9270(200105)96 5;1-A
`
`Ex. 1042 - Page 2
`
`

`

`.r„E AI, RICAN JOURNAL OF GASTROENTEROLOGY (cid:9)
`0 2001 by Am. Coll. of. Gastroenterology (cid:9)
`Wished by Elsevier Science Inc.
`
`,.••••••••"`"
`
`This material may be protected by CopylIght law (idle 17 U.S. Code)
`
`Vol. 96, No. 5, 2001
`ISSN 0002-9270/01420.00
`Pit S0002-9270(01)02314-0
`
`Detection of Mycobacterium avium Subspecies
`paratuberculosis in. Crohn's Diseased Tissues by
`in Situ Hybridization
`
`f(sistiaa }Julien, Ph.D., Hala M. T. El-Zimaity, M.D., Toumo J. Karttunen, M.D., Ashraf Almashhrawi, M.D.,
`Mary R. Schwartz, M.D., David. Y. Graham, M.D., and Fouad A. K. El-Zaatari, Ph.D.
`milannnatoty Bowel Disease Laboratory, Veterans. Affairs Medical. Center,. Departments of Medicine and
`Pathology, and Division of Molecular Virology, Baylor College of Medicine, Houston, Texas; and
`Depar tment of Pathology, University of Oulu, Oulu, Finland
`
`OBJECTIVES: Reports about the association between Crohn's
`disease (CD) and cell wall-deficient (CWD) forms of My-
`tobacteritint aviurn subspecies paratuberculosis (M. para-
`tuberculosis) are controversial. This may be due to the
`heteroeneous nature of CD where only about 50% of the
`patients show granulomatous inflammation. Detection of
`CWD forms of M. paratuberculosis in tissues from patients
`with CD would support its association with the. disease. To
`help identify these forms in inflamed tissues, a previously
`developed and optimized nonradioactive in situ hybridiza-
`lion method was applied on well-defined tissue materials
`obtained from patients with CD, ulcerative colitis (UC), and
`controls.
`
`METHODS: Specimens from. 37 patients with CD (15 with
`epitheloid cell granulomas and 22 without granulomas), 21
`UC, and 22 noninflammatory bowel disease (IBD) patients
`were analyzed by the in situ hybridization method based on
`the digoxigenin-labeled M. paratuberculosis IS900 frag-
`Ilent, previously shown to be species specific. Samples
`sere counterstained with hematoxylin and eosin to show the
`location of the positive signal. Positive controls made of
`leof cubes injected with CWD and acid.-fast M. parattiber-
`'olosis and negative controls were included in each exper-
`lent to monitor for nonspecific hybridization or staining.
`
`iESULTS: Six of 15 (40%)patients with CD and granulomas
`thawed positive signals in myofibroblasts and macrophages.
`'iaerestingly, no positive signals were observed within gran-
`Aamas. Only 4.5% of 22 CD samples from patients with
`Agranulomatous disease,. 9.5% of 21 UC, and remarkably,
`iSile of the 22 non-IBD patients were M. paratuberculosis
`'s)sitive.
`
`°Ne1.11SION: The demonstration of DNA from CWD
`'ins of M. paratuberculosis in this limited number of CD
`'sues further supports and confirms previous reports of its
`leiation with the granulomatous type of the disease. (Am
`Gastroenterol 2001;96:1529-1535. © 2001 by Am. Coll.
`Gastroenterology)
`
`INTRODUCTION
`
`An association of mycobacteria with. Crohn's disease (CD)
`has been suspected since the early 1900s when Dalziel
`described it as chronic enteritis in humans (1). His descrip-
`tion was based on its similarity to a disease that was earlier
`described as chronic enteritis in animals (2). The animal
`disease is now known as. Johne's disease or paratuberculo-
`sis, a disease also found in subhuman primates, that is
`caused by Mycobacterium avium subspecies paratubercu-
`losis (M. paratuberculosis) (3, 4). These intestinal human
`and animal diseases share similar histological and micro-
`scopic features with tuberculosis; diseased tissues typically
`contain granulomas (5, 6) Although controversial, further
`support of the mycobacterial etiology in CD disease are: 1)
`the recent emergence of cell wall-deficient (CWD) forms of
`M. paratuberculosis in cultures from tissue specimens and
`milk of a portion of the patients with CD (7-13), 2) the high
`frequency in detecting M. paratuberculosis DNA in CD
`tissues (14-18), 3) the successful therapy of CD patients
`with antirnycobacterial treatment containing macrolide an-
`tibiotics (19-21), and 4) the specific seroreactivity of the
`majority of CD patient serum samples specifically against
`M. paratuberculosis antigens (22-24). The variation of dis-
`ease severity, relapses, and remissions, and response to
`treatment between individual patients, however, makes
`these results inconsistent among all investigators and creates
`controversy about an etiology of this disease (25-27).
`Hence, it is tempting to believe of CD as a group of disor-
`ders, of different etiologies, but with similar presentation,
`and that one of these disorders is triggered or caused by M.
`paratuberculosis, specifically its CWD form.
`Distinction between acid-fast mycobacteria and the cul-
`tured CWD forms in tissue of CD patients is difficult.
`Although M. paratuberculosis CWD forms have been cul-
`tured from CD tissue, immunohistochemical methods and
`electron microscopic surveys have not revealed their actual
`localization within CD tissues; these organisms do not stain
`with the Zeil-Nielsen method and have no structure that
`would bring any contrast against host tissue components. To
`
`Ex. 1042 - Page 3
`
`

`

`1530 (cid:9)
`
`Hulten et al. (cid:9)
`
`AJG - Vol. 96, No. 5, ';.001
`
`date, neither nnycobacteria nor their antigens have been
`convincingly demonstrated in inflamed CD tissue (28). In
`situ hybridization has been adapted for detection of infec-
`tious agents that are difficult to visualize with conventional
`methods (29, 30). With the aim of detecting M. paratuber,
`culosis CWD forms within tissue, we previously developed
`a novel and specific in situ hybridization method (31). The
`method selectively provides a way to lyse only the CWD
`form of M. paratuberculosis in tissue specimens and hence,
`distinguish it from its acid-fast form; it is impossible to lyse
`mycobacterial bacilli without destroying the host tissue
`structure (31). It was previously used for the detection of
`CWD M.. paratuberculosis in tissue specimens from animals
`with Johne's disease (31). This method was first tested on a
`number of positive and negative controls to ascertain the
`specificity of hybridization and rule out nonspecific anti-
`body binding and staining (30). In this communication, we
`describe its application on tissue specimens from patients
`with CD and controls.
`
`MATERIALS AND METHODS
`
`Clinical Samples
`Coded paraffin blocks containing biopsy and resected tissue
`specimens from 80 patients, consisted of 58 cases of inflam-
`matory bowel disease (IBD) and 22 age-matched controls,
`were studied. The 58 patients were classified into main
`categories of IBD (CD and ulcerative colitis; UC) according
`to endoscopic, radiological, histological, operative findings,
`and standard clinical criteria (32). Colonic resected tissues
`from the majority of samples (only five colonic biopsy
`specimens of each category of CD with granulomas, CD
`without granulomas, and UC) were used. All samples were
`selected from archival collection at the Department of Pa-
`thology, University of Oulu, Oulu, Finland, and the Depart-
`ment of Pathology, Baylor College of Medicine, Houston,
`TX. Paraffin blocks with resected tissues from non-IBD
`(included three patients with diverticular disease, three di-
`verticulitis, 10 ischemic colitis, one serosal abscess, two
`cytomegalovirus colitis, one acute nonspecific colitis, and
`two adhesions) patients who came to the Digestive Diseases
`Clinic, Baylor College of Medicine and the Veterans Affairs
`Medical Center, Houston, were also tested. Although no
`documentation of treatment was available, all the samples
`were selected as definite cases of these diseases and all
`specimens included surface epithelium and muscularis mu-
`cosae and were tested blindly.
`
`In situ Hybridization
`The procedure used in this study was a modification of a
`previously described method (31). Briefly, sections of 4 p.m
`in thickness were placed on silane (Silane Prep slides,
`Sigma, St, Louis, MO)-treated slides, baked for 60 min at
`60°C, deparaffinized in xylene (2 X 10 min), rehydrated
`through graded (100%, 80%, 50%, 3 min each) ethanol, and
`equilibrated in phosphate buffered saline (PBS) at pH 7.4
`
`for 3 min. To increase cell permeability and accessibility to
`target DNA and to quench the background signals caused by
`factors like endogenous alkaline phosphatase and formalin
`fixation, sections were subjected to proteinase K (Boeh
`er-Mannheim, Indianapolis, IN) treatment (0.01 mg/ml in
`PBS) for 20 min at 37°C in a humid chamber. Sections were
`washed twice (5 min each) in. PBS, dehydrated in gi aded
`ethanol (50%, 80%, 100%, 1 min each) and air-dried. Each
`section was covered with 20-40 p.1 (depending on the size
`of the section) of hybridization solution (50% deionized
`formamide, 2x SSC [I X SSC = 0.14 mol/L sodium chlo-
`ride, 0.014 mol/L sodium citrate], 10% dextran sulfate, 0.25
`p4/µ1 yeast t-RNA, 0.5 p,g/p,1 heat denatured salmon sperm
`DNA, and 1 X Denhart's solution) containing 1 ng/p,1
`digoxigenin-labeled IS900-specific M. paratuberculosis
`probe of 251-bp DNA fragment. The hybridization solution
`was boiled for 10 min and chilled on ice before it was added
`to the samples. Hybri-slip (Sigma, St. Louis, MO; were
`mounted with rubber cement (Best-test, Union Rubber Inc,
`NJ) and the probe was denatured at 95°C for 10 min and
`chilled on ice for 10 min. Hybridization was perfDnued
`overnight at 37°C in a humidified chamber. Posthybridiza-
`tion stringency washes were performed in three 10 min
`steps; once in 2 X SSC at room temperature and twice in
`0.3 X SSC at 40°C followed by a brief rinse in 0.3..X SSC
`at room temperature.
`To visualize the hybridized probe, the immunological
`detection system was used with antidigoxigenin antibodies
`conjugated with alkaline phosphatase following the manu-
`facturer's instructions (Boehringer-Mannheim). Briefly,
`samples were blocked with buffer 1 (100 mmol/L Tris-CI,
`150 mmol/L NaCl at pH 7.5), containing 2% normal sheep
`serum and 03% Triton X-100 for 30 min at room temper-
`ature. Antibody solution (buffer 1, 0.3% Triton X-100, I% (
`normal sheep serum and alkaline phosphatase-conjugated
`antidigoxigenin antibodies [1:300 dilution]) was added and
`incubated for 3 h at room temperature. The slides were then
`washed sequentially in buffer 1 and in buffer 2 (100 rnmol/L
`Tris-CI, 100 mmol/L NaC1, 50 mmol/L MgCl2 at pH 9.5) for
`10 min each at room temperature. Hybridization producl
`was detected using nitro-blue-tetrazolium chloride and
`5-bromo-4-chloro-3-indolyl-phosphate in buffer 2 as sug-
`gested by the manufacturer (Boehringer Mannheim). After
`overnight incubation at room temperature, slides were
`washed in distilled water for 2 min and coverslips welc
`mounted using water-based mounting medium (GelMouul.
`Biomeda,. Foster City, CA). Slides were coded and blindl
`read by a pathologist (H.M.T. E-Z) and representative Si
`ples were photographed. To verify the exact location of the
`probe in the tissue, the coverslips were removed and the
`sections washed in distilled water. Tissues were then coon:,
`terstained with hematoxylin and eosin, washed ir; distilled
`
`water, coverslipped, mounted, and examined under the r1)
`croscope. The location of M. paratuberculosis in the tissue
`was assessed morphologically based on the presel ce oftUe
`hybridized probe.
`
`Ex. 1042 - Page 4
`
`

`

`AjG (cid:9) May, 2001 (cid:9)
`
`M. avium Subspecies paratuberculosis in CD (cid:9)
`
`1531
`
`'rabic 1. Results of In Situ Hybridization and PCR Assays Applied on. Clinical Samples
`No. Positives/Total Tested
`PCR
`paratuherculosis
`3/14
`6/22
`3/21
`0/22
`
`Disease
`granulomas)
`CP (no granulomas)
`UC
`Non-1BD
`c Al: avium complex.
`t Five biopsies of each category were from colon or rectum; din teat of '.the samples were colonie resected, tissues..
`
`M. tuberculosis (cid:9)
`0/6
`0/6
`0/5
`ND
`
`MAC* (cid:9)
`4/6
`4/6
`3/5
`ND
`
`In Situ
`Hybridization (%)
`6/15 (40)t
`1/22 (4,5)t
`2/21 (9.5)t
`0/22 (00)
`
`Control Specimens
`positive and negative controls made of paraffin blocks with
`beef tissues injected with. CWD and acid-fast forms of M.
`paratuberculosis strain Linda (ATCC 43015; originally iso-
`lated from CD tissue), respectively, were used in parallel
`with each clinical sample (31). To monitor for nonspecific
`hybiidization or staining, more negative controls with acid-
`fast MI paratuberculosis, both forms (acid-fast and CWD)
`of M. tuberculosis strain ATCC 25177, both forms (acid-
`fast and CWD) of M. smegmatis strain ATCC 607, and
`clinical isolates of Escherchia coli (E colt) and Helicobac-
`ter pylori (H. pylori) RD26 obtained from the Veterans
`Affairs Medical Hospital, Houston, TX, were also included.
`• The intact acid-fast bacillary and CWD forms of the above
`I
`) mycobacterial controls were prepared in vitro as described
`) previously (31). The presence of both forms was confirmed
`by staining the sections with either Kinyoun's stain or
`acridine orange stain, respectively (31). Helicobacter genus-
`speciic probe derived from 16S rRNA gene fragment
`(about 500 bp) was used as an additional negative control
`probe (33). The specificity of these probes has previously
`been established (31, 33-35). Nonspecific, binding of anti-
`body or color development solution was tested by perform-
`ing experiments excluding probe, antibody, or both.
`
`Preparation of DNA From Paraffin-Embedded Tissue
`and Polymerase Chain Reaction
`Five 4- Ain paraffin-embedded tissue sections, adjacent to a
`section that was examined after staining with hematoxylin
`and eosin, were prepared per sample and placed in a sterile
`Eppendorf tube. To avoid contamination, a disposable blade
`was used for each sample. Each sample was deparaffinized
`twice in 1.2 ml of xylene for 30 min each. After being
`washed twice in absolute alcohol for 10 min each, samples
`were dried in a speed vacuum concentrator for around 2-5
`min. The samples were then prepared with the QiaAmp
`lissue kit as instructed by the manufacturer with some mod-
`4cations (Qiagen, Valencia, CA). Each sample was di-
`gested in 200 Al of lysis buffer (supplied in the kit) con
`'raining proteinase K at a final concentration of 360 Ag/m1 at
`55°C overnight or until clearly digested. Proteinase K was
`Inactivated by boiling for 10 min. Each DNA pellet was then
`.Suspended in 50 Al of TE (10 mM Tris HC1, LmM EDTA,
`P118.0) and stored at 4°C until use.
`DNA prepared from paraffin-embedded biopsy speci-
`
`mens were evaluated by the M. paratuberculosis-, M. tu-
`berculosis-, and M. avium complex (MAC)- specific poly-
`merase chain reaction (PCR) assays and Southern
`hybridization techniques using, specific probes as described
`in detail elsewhere (11, 36). Because M. paratuberculosis is
`a member of MAC group, its DNA can be amplified and
`hybridized by the MAC-specific PCR assay and MAC-
`specific probe, respectively. The IS900-specific PCR assay
`and IS900-specific probe, however, amplifies and hybridizes
`only to DNA from M. paratuberculosis, respectively (36).
`The origin and specificity of these probes are discussed in
`detail elsewhere (36).
`
`Statistical Analysis
`X2 Test was used to estimate significance of differences in
`frequencies. The p value of < 0.05 was regarded as being
`statistically significant.
`
`RESULTS
`
`Patients
`The median age of the patients, 34 men and 46 women, was
`47 yr with the range of 19 to 87 yr. These cases consisted of
`22 patients with non-IBD, 21 patients with UC, 37 with
`active CD including 15 with epitheloid cell granulomas
`(only two perforated) and 22 without granulomas (only one
`perforated). All CD patients had been followed up for sev-
`eral years with consistent absence pr presence of granulo-
`mas. None of the patients had received antimycobacterial
`therapy.
`
`In Situ Hybridization and PCR on Clinical Samples
`The results of the clinical application of the in situ hybrid-
`ization and PCR assays are summarized in Table 1. All the
`M. paratuberculosis-positive samples by in situ hybridiza-
`tion and IS900-specific PCR assay were also tested by M.
`tuberculosis-specific and MAC-specific PCR assays. Two
`additional M. paratuberculosis-negative samples from UC
`patients were also tested by M. tuberculosis-specific and
`MAC-specific PCR assays.
`Six (three resected tissues and three biopsies) of 15 (40%)
`patients with CD and granulomas, including one patient
`with perforated disease, were positive by in situ hybridiza-
`tion; three of which were also positive by M. paratubercu-
`
`Ex. 1042 - Page 5
`
`

`

`1532 (cid:9)
`
`Hulten et al. (cid:9)
`
`AJG - Vol. 96, No. 5, 2001.
`
`losis- and MAC-specific PCR assays and one additional by
`MAC-specific PCR assay only (Table '1). One (resected
`tissue) of 22 (4.5%) patients with CD and without granulo-
`mas was positive by in situ hybridization and by M. para-
`tuberculosis- and MAC-specific PCR assays; five additional
`samples were positive by M. paratuberculosis-specific PCR
`assay of which three samples were positive by MAC-spe-
`cific PCR assay. Two (resected tissues) of 21 (9.5%) pa-
`tients with UC were also positive by in situ hybridization
`and by M. paratuberculosis- and MAC-specific PCR assays.
`One additional sample was positive only by M. paratuber-
`culosis- and MAC- specific PCR assays. All M. paratuber-
`culosis-positive samples were negative by M. tuberculosis-
`specific PCR assay but, as expected, the majority of M.
`paratuberculosis-positive samples were also positive by
`MAC-specific PCR assay; M. paratuberculosis is a member
`of MAC (36).
`The positive signals were found in myofibroblasts and in
`macrophages (Figs. 1 and 2, respectively). Although the
`majority of the positive signals were seen in patients with
`CD and granulomas, no signal was seen within granulomas.
`One patient with CD and without granulomas but with
`perforated disease was negative by all tests. Samples from
`the 22 non-IBD patients were all negative by in situ hybrid-
`ization and by M. paratuberculosis-specific PCR assays.
`To ensure specificity, no signal was observed when the
`probe or the antidigoxigenin antibodies were removed from
`the hybridization solutions indicating that the positive sig-
`nals were specific. The age, gender, the origin of samples
`(colon or rectum), and the perforation status of the three
`tested samples had no effect on the outcome of the results.
`These experiments were repeated at least three times with
`identical outcome.
`
`Specificity and Sensitivity of In Situ Hybridization and
`PCR Assays
`The specificity of in situ hybridization was confirmed pre-
`viously by the lack of hybridization signal with several
`different controls (31). No hybridization with the IS900
`probe was observed on sections of beef injected with acid-
`fast or CWD forms of M. tuberculosis or M. smegmatis. In
`addition, the probe did not bind to sections containing H.
`pylori or E. coll. Sections containing acid-fast forms of M.
`paratuberculosis did not hybridize with the probe, probably
`because of the impermeability of the cell wall.
`Negative results were also achieved when the probe was
`omitted from the hybridization solution or an irrelevant
`Helicobacter-specific probe was used. The sensitivity of the
`assay was difficult to assess because of the formation of
`leaky cell walls (partial CWD) of some mycobacterial or-
`ganisms as a result of the glycine/lysozyme treatment (31).
`This caused the mycobacterial cells to become fragile and
`some of them to aggregate, which made them difficult to
`disperse during their in vitro preparation process (31). This
`leads to the inability to detect or count single organisms
`(31).
`
`Figure 1. Representative application of in situ hybridization on
`paraffin tissue sections obtained from clinical specimens of
`Crohn's disease with granulomas. (A) In situ hybridization ,with no
`counterstain showing M. paratuberculosis DNA along the base-
`ment membrane of a gland (arrow). (Original magnification X40).
`(B) Same as in (A) with hematoxylin and eosin as a counterstain
`showing the M. paratuberculosis DNA as brown positive spots
`within myofibroblasts (arrow). (Original magnification X 1000).
`
`The specificity and sensitivity of the PCR assays z nd the
`probes used in this study were previously proven to be
`species specific and can detect one to 50 organisms Per
`reaction (36).
`
`Statistical Significance
`Although it is difficult to assess the significance of the
`positive number of samples by in situ hybridization because
`of the low number of organisms per sample, the frequeneY
`of positive signals between patients with IBD and non-IBD
`was significant (9 of 57 IBD vs 0 of 22 non-IBD; p = 0.047).
`A higher statistical significance was achieved, however'
`when CD with granulomas was considered as a third disease
`entity in IBD patients; 6 of 15 patients with CD and gran"
`ulomas versus 1 of 22 without granulomas, p = 0.007; 6 °I.
`15 patients with CD and granulomas versus 2 of 21 patients
`with UC, p = 0.03; 6 of 15 patients with CD and &ran111°-
`mas versus non-IBD, p = 0.001 (Table 1).
`
`Ex. 1042 - Page 6
`
`

`

`AJG - May, 2001 (cid:9)
`
`M. avium Subspecies paratuberculosis in CD (cid:9)
`
`1533
`
`water, or dairy products; pasteurization sometimes does not
`destroy it (41-43).
`The repeated failure to detect mycobacteria in gut wall
`tissues using the Ziel-Nielsen staining method (to detect
`acid-fast bacilli) and immunohistochemical staining (by an-
`tibodies and peroxiclase-antiperoxidase) is probably related
`to the low concentrations of M. paratuberculosis organisms
`within tissue (around 300 organisms per gram of tissue (14))
`or perhaps to its presence in a CWD forrn. Although CWD
`forms of M. paratuberculosis have never been primarily
`isolated from humans other than patients with CD and
`perhaps tissues from patients with sarcoidosis (10, 11, 36),
`no in situ hybridization experiments to detect M. paratu-
`berculosis DNA in CD tissues have been published.
`Only recently have mycobacterial genus- and species-
`specific probes become available. Because of the lack of
`positive controls to assess the validity of this in situ proce-
`dure, positive and negative controls including the acid-fast
`and CWD forms of mycobacteria had to be created and
`tested in parallel with each clinical sample (31). Specificity
`of the in situ hybridization assay was previously demon-
`strated in several ways including the negative findings with
`control tissue preparations containing CWD or acid-fast
`forms of unrelated mycobacteria (M. tuberculosis and Al.
`smegmatis). The probe was neither binding nonspecifically
`to the cell membrane of CWD organisms, nor was any
`binding seen in sections containing H pylori or E. colt (31).
`IS900 is a well-suited target for in situ hybridization, as it is
`a repetitive gene with approximately 15 to 20 copies in each
`cell (44), and quite within the detection limits for in situ
`hybridization (45). By applying this novel in situ hybridiza-
`tion procedure, it was possible for the first time to demon-
`strate the existence of CWD forms of M. paratuberculosis
`within diseased tissue from patients with CD.
`This study shows a statistical significance in the associ-
`ation between granulmatous CD and M. paratuberculosis,
`supporting previous speculations (46, 47). Interestingly,
`these organisms were found only'in the myofibroblasts and
`macrophages in the lamina propria of those Crohn's patients
`with granulomas but not (as would be expected) within
`granulomas. Similar findings were also observed in tissue
`specimens from animals with Johne's disease (31, 48-50).
`The detection of M. paratuberculosis in the large proportion
`of granulomatous CD patients and in a small proportion of
`patients with UC and CD without granulomas may empha-
`size the lack of clinical disease markers to reliably assess
`IBD disease (51, 52). Definitive diagnosis to assess the
`inflammatory activity of IBD is usually inaccurate and re-
`quires the reevaluation of the initial diagnosis to confirm
`specific disease category; CD, UC, indeterminate colitis, or
`probable CD (51, 52). Hence, it is possible that a group of
`patients with IBD, including a large proportion of CD with
`granulomas and small proportions of CD without granulo-
`mas and UC, may constitute a disease category of IBD with
`known etiology (M. paratuberculosis). Similar findings
`were recently reported in samples from patients with sar-
`
`Figure 2. Representative application of in situ hybridization on
`paraffin tissue sections obtained from clinical specimens of
`Crohn's disease with granulomas. (A) In situ hybridization with no
`counterstain showing Al. paratuberculosis DNA in the lamina
`provia and occasionally infiltrating gland (arrow). (Original mag-
`nification. X40), (B) Same as in (A) with hematoxylin and eosin as
`a Co interstain showing the. M. paratuberculosis DNA as brown
`positive spots within macrophages in the lamina propria. (Original
`magnification X1000).
`
`DISCUSSION.
`
`Although there is no recognized etiology for CD, the suc-
`cessful recovery of the CWD variants of M. paratubercu-
`losis from CD tissues in several laboratories including ours
`makes it a probable candidate for causing or exacerbating at
`least some cases of CD (7-11, 37-40). This is consistent
`With the most striking recent reports documenting: 1) the
`isoh.tion of the organism from the breast milk of two moth-
`ers with CD (12), 2) the identification of M. paratubercu-
`losis in a cervical lymphadenitis lesion of a patient who later
`developed CD (19), and 3) the specific seroreactivities
`found in a significant proportion of sera from patients with
`CD compared to controls (22-24). Its isolation from tissues
`from patients with CD by culture suggests that it may be
`transmitted from animals to humans by ingestion of meat,
`
`Ex. 1042 - Page 7
`
`

`

`1534 (cid:9)
`
`Hulten et al. (cid:9)
`
`AJG - Vol. 96, No. 5, 2001
`
`coidosis (36); M. paratuberculosis or unknown MAC or-
`ganisms were identified in these patients (no M. tuberculosis
`was found in any of the tested samples) (36). Whether
`detecting M. paratuberculosis is specific for caseating or
`noncaseating granulornatous diseases would be interesting
`to investigate.
`The detection of the IS900 sequence by PCR in UC
`individuals including some CD patients without granulomas
`suggests that the organism may be present transiently in the
`acid-fast and not the CWD form, although colonization does
`not occur. Alternatively, the difference in the number of
`positive samples by PCR and in situ hybridization methods
`may be due to their technical differences and may be ex-
`plained as follows: 1) the PCR sensitivity or the low number
`of organisms within tissue (14) may have contributed to the
`detection of the IS900 DNA sequence by PCR in more UC
`and CD patients without granuloma8; and 2) inhibitory
`factors may have contributed to the PCR negativity of some
`CD samples with granulomas. However, our data may in-
`dicate that some patients with IBD (especially with granu-
`lomatous CD) may have a genetic predisposition that makes
`them susceptible to M. paratuberculosis infection as com-
`pared to control patients. The detection of the IS900 element
`within some IBD tissues, and not in the non-IBD controls
`further indicates that some cases of IBD may be associated
`directly with or perhaps triggered or caused by a pathogenic
`mycobactin-dependent mycobacterium: M. paratuberculo-
`sis or closely related but unidentified, pathogenic mycobac-
`terial species or subspecies. This also would suggest a
`causal role for M. paratuberculosis in these cases of IBD,
`specifically the granulomatous group.
`In conclusion, the isolation of M. paratuberculosis by
`culture from a high proportion (37-40%) of CD tissues (10,
`11, 13) and the detection of the same organism by in situ
`hybridization in tissues of a similar high proportion (40%)
`of granulomatous CD patients further confirm its potential
`etiological role in this human disease. This causal relation-
`ship is also implied by the broad host range of M. parcitu-
`berculosis and its well-established enteric pathogenicity.
`Hence, the hypothesis of M. paratuberculosis involvement
`in a group of CD patients (or perhaps IBD) is further
`supported. Although, further studies are needed to substan-
`tiate our findings, studies to correlate a specific stratification
`of IBD (especially the granulomatous ones) with M. para-
`tuberculosis infection are urgently needed. Studies to elu-
`cidate disease pathogenesis at the cellular level and to