WO1992002819A2

WO1992002819A2 - Assay for ulcerative colitis and primary sclerosing cholangitis

Info

Publication number: WO1992002819A2
Application number: PCT/US1991/005276
Authority: WO
Inventors: Stephan R. Targan; Fergus L. J. Shanahan; Carol J. Landers
Original assignee: The Regents Of The University Of California
Priority date: 1990-08-10
Filing date: 1991-07-25
Publication date: 1992-02-20
Also published as: WO1992002819A3

Abstract

The present invention provides a method for detecting the presence of diseases including ulcerative colitis or primary sclerosing cholangitis in a patient. The method consists of contacting a body sample of the patient with immobilized neutrophils so as to allow the antineutrophilic cytoplasmic antibodies to bind to the neutrophils. Further, the method provides detecting bound anti-neutrophilic cytoplasmic antibodies, the presence of the antibodies indicative of the disease state.

Description

ASSAY FOR ULCERATIVE COLITIS AND PRIMARY SCLEROSING CHOLANGITIS

This work was supported by the National Foundation for Ileitis & Colitis and the Harbor/UCLA Inflammatory Bowel Disease Center, NIH Grant DK 36200. The government has certain rights in the invention.

BACKGROUND OF THE INVENTION

This invention relates to the diagnosis of disease and, more specifically to the detection of anti- neutrophilic cytoplasmic antibodies.

Inflammatory Bowel Disease (IBD) is the collective term used to describe the two gastrointestinal disorders, ulcerative colitis and Crohn's disease. Although these diseases have distinctive pathophysiological characteristics, they are freguentl" considered together due to several clinical and therapeutic similarities. Excluded from this category, however, are gastrointestinal inflammatory disorders of known infectious, toxic or ischemic etiology which may mimic IBD acutely, but do not cause a chronic relapsing and remitting syndrome. Inflammatory bowel disease poses a clinical and scientific challenge to physicians and researchers. Confirming a diagnosis may prove a long-term endeavor. As indicated above, implication of a causal factor remains elusive, and there is no known cure. IBD, and quite often its treatment, affects the lifestyle and functional capabilities of those afflicted. Treatment courses often result in adverse physiologic manifestations which must be balanced against the therapeutic benefit. Any intervention which can improve patients' toleration of their disease and therapeutic program is welcome.

Some estimates suggest the number of people afflicted with IBD may be as high as two million. Inflammatory bowel disease occurs world-wide, but is more common in North America, Northern Europe, and among Caucasians in Australia and New Zealand. There is some evidence that risk for females may be slightly higher, that there is increased risk for IBD among Ashkenazi Jews, and is more common among whites than blacks. The incidence rate of IBD appears to be stabilizing although data for ulcerative colitis differs from that for Crohn's disease. Onset has been documented at all ages; however, IBD predominantly affects young adults. The risk is greatest for family members of a patient, though the majority of patients have no known affected relative.

The course and prognosis of IBD is widely variable. Ten to 15% of all patients with IBD will require surgery over a ten-year period. The risk for the development of cancer is increased in patients with IBD as well, particularly in those with ulcerative colitis. The longer the duration of disease, the higher the risk of developing carcinoma. Patients with ulcerative colitis regularly undergo cancer surveillance by endoscopy after ten years of disease.

Differentiating Crohn's disease (CD) from ulcerative colitis (UC) is difficult, particularly with Crohn's disease of the colon. Diagnosis is based upon a variety of clinical and pathological indicators. Ulcerative colitis almost always involves the rectum as well as the colon. Crohn's disease may affect any area of the gastrointestinal tract from the mouth to the anus, most commonly, however, the terminal ileum. Inflammation associated with ulcerative colitis is generally superficial and continuous, whereas the inflammatory pattern of Crohn's disease is transmural and is characterized by "skip" areas. Perianal lesions are rare in ulcerative colitis and fistula formation should raise suspicion of Crohn's disease. Difficulty establishing the diagnosis of UC may arise at two levels. First, acute idiopathic UC of recent onset must be distinguished from a variety of disorders that may closely mimic it. These include the infectious and ischemic colitides. Second, UC must be distinguished from CD. This may be particularly difficult when CD is limited to the colon. Indeed, depending on the period of follow-up time, in many patients the colitis must be regarded as indeterminate or cannot be definitively diagnosed because of overlapping features of UC and CD. The distinction between UC and CD carries important prognostic and therapeutic implications. For example, when colectomy is indicated, the type of inflammatory bowel disease involved determines which surgical options are appropriate. Continent procedures such as the ileorectal pull-through (mucosal proctectomy) or the Kock pouch may be desirable in UC but are contraindicated in CD. The availability of a diagnostic marker which could help distinguish UC from CD of the colon would be an important clinical advance. There are standard clinical approaches to diagnosis, including endoscopy, biopsy and x-rays, but these tend to be expensive, invasive and time consuming. Once the diagnosis of UC is established, an additional difficulty is the assessment of disease activity. A convenient and reliable blood test which might parallel disease activity or even predict an impending flare of activity is unfortunately not available.

Primary sclerosing cholangitis (PSC) , a chronic, progressive inflammatory disorder characterized by inflammation and fibrosis of the intrahepatic extrahepatic tile ducts, is a disease which commonly, but not always, occurs in patients who also have a history of UC. The explanation for this clinical association is not known. The course of the disease is generally one of slow progression to cirrhosis, portal hypertension and death from liver failure. Diagnosis is usually based on biochemical, radiologic and histologic criteria.

Serum immunoglobulin G (IgG) antibodies directed against cytoplasmic components of neutrophils were first described in patients with glomerulonephritis and systemic vasculitis. Subsequently, the importance of anti- neutrophil cytoplasmic antibodies (ANCA) in the diagnosis and management of patients with Wegener's granulomatosis ( G) has been recognized.

Recently, studies using an immunofluorescent (IF) technique to detect ANCA, showed that 25% of UC and 3% of CD patients have ANCA. (Nielsen et al., Acta. Path. Microbiol. Immunol. Scan. 91:23-26, 1983, which is incorporated herein by reference.) However, the immunofluorescent method may not be highly sensitive and additionally, may give a high background reactivity.

There thus exists a need for a rapid sensitive test to determine the presence of ANCA, and determine its specificity for UC and PSC as compared to Crohn's disease and other colitides and diarrheal illnesses. The test should have a correlation with the disease states and a low or negative correlation with other diseases. The invention therefore provides a method for detection of ANCA in several disease states and allows for the screening of samples on a large scale.

SUMMARY OF THE INVENTION

The present invention provides a method for detecting the presence of diseases including ulcerative colitis or primary sclerosing cholangitis in a patient. The method consists of contacting a body sample of the patient with immobilized neutrophils so as to allow the anti¬ neutrophilic cytoplasmic antibodies to bind to the neutrophils. Further, the method provides detecting bound anti-neutrophilic cytoplasmic antibodies, the presence of the antibodies indicative of the disease state.

BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 shows levels of neutrophil binding by anti- neutrophil IgG at a dilution of 1:100. Coded sera were assayed in the fixed neutrophil ELISA at a dilution of 1:100. The dashed line represents the level of 2 SD above the mean for the normal controls. The solid line in each column represents that group's mean. Differences between groups are as follows: UC versus UCPC, p = 0.239 (not significant); UC or UCPC versus normal, p < 0.001 in each case; UC or UCPC versus each of the disease controls, p < 0.001 in each case; normal versus each of the disease controls, p > 0.05 in each case (not significant), except for collagenous colitis where p = 0.003.

Figure 2 shows titers of anti-neutrophil IgG. Coded test sera were titered in the fixed neutrophil ELISA. A level of binding exceeding 2 SD above the mean for the normal controls diluted 1:100 was considered to be a positive value. Titers for UC and UCPC were each significantly greater than the titers for each of the other colitides and diarrheal illnesses (p < 0.001 in each case).

DETAILED DESCRIPTION OF THE INVENTION

Serum anti-neutrophil cytoplasmic antibodies (ANCA) , distinct from those associated with active Wegener's granulomatosis (WG) , are present in the majority of patients with ulcerative colitis (UC) . The specificity of anti-neutrophil cytoplasmic antibodies for ulcerative colitis or primary sclerosing cholangitis as compared to other colitides and diarrheal illnesses has not been studied. This invention provides a method for detecting the presence of ulcerative colitis or primary sclerosing cholangitis by detecting the presence of anti-neutrophilic cytoplasmic antibodies. By using an enzyme-linked immunoabsorbant assay (ELISA) , the sensitivity and specificity in the detection of ANCA is greatly increased over previous methods using immunofluorescence. In addition, ELISA allows a large number of samples to be screened at one time. In a blinded study, test sera were screened for anti-neutrophil immunoglobulin G in a fixed neutrophil ELISA. Levels of neuti-oph.il binding by immunoglobulin G in titers of anti-neutrophil immunoglobulin G in sera from patients with ulcerative colitis and patients with ulcerative colitis post colectomy were each significantly greater than the levels and titers for normal controls and patients with a variety of other colitides and diarrheal illnesses. Levels of neutrophil binding for colonic Crohn's disease (CD), bacterial/amoebic colitis, the irritable bowel syndrome with diarrhea, and other miscellaneous diarrheal illnesses were not significantly different than the levels for normal controls. Although the levels of binding for collagenous colitis were significantly less than the levels for ulcerative colitis, they were significantly greater than the levels for normal controls. Patterns of neutrophil binding by sera that were positive in the enzyme-linked immunosorbant assay were determined by indirect immunofluorescence. Perinuclear staining was the predominant pattern exhibited by sera from patients with ulcerative colitis. The combination of a positive value in the ELISA and a perinuclear immunofluorescence pattern was 60% sensitive and 94% specific for ulcerative colitis. Most sera from Crohn's patients are ELISA negative. Of those which are positive, most do not exhibit a perinuclear immunofluorescence pattern.

Serum IgG ANCA, detected in a fixed neutrophil, enzyme-linked immunosorbant assay are present in the majority of patients with ulcerative colitis and a much smaller percentage of patients with colonic Crohn's disease. In contrast to the granular, diffuse cytoplasmic immunofluorescence pattern exhibited by ANCA that are characteristic of active WG, a perinuclear immunofluorescence pattern is exhibited by sera from patients with UC. These results confirm the presence of ANCA which exhibit perinuclear immunofluorescence staining in 60% of patients with UC. Perinuclear ANCA were also found in 63% of patients with ulcerative colitis post colectomy (UCPC) but in only 6% of patients with a variety of other colitides and diarrheal illnesses (the latter 6% comprised primarily of patients with collagenous colitis) .

The results indicate that ANCA in UC are not simply an epiphenomenon related to active colonic inflammation because (a) levels of neutrophil binding by IgG in sera from patients with a variety of acute and chronic colitides other than UC are not significantly different than the levels for normal controls, (b) perinuclear ANCA are highly specific for UC when compared to other colitides, and (c) ANCA are also present in patients with UCPC and PSC. The presence of ANCA in UC may reflect a fundamental disturbance of immune regulation. Similar ANCA are present in the majority of patients with PSC and the subset of patients with PSC and no endoscopic or histologic evidence for UC. Whether the UC- an: the PSC-associated ANCA are reactive to the same or different antigen(s) and whether they are of pathogenic importance is not yet known. However, the presence of similar ANCA in PSC and UC raises the possibility of shared immunopathogenic mechanisms.

In contrast to colonic CD, bacterial/amoebic colitis,

IBS-diarrhea, and other miscellaneous (primarily inflammatory) diarrheal diseases, sera from a subset of patients with collagenous colitis exhibited very high levels of binding and extremely elevated titers in the fixed neutrophil ELISA. Some of these sera also exhibited a perinuclear IF pattern that is indistinguishable from the predominant IF pattern exhibited by sera from patients with UC. Collagenous colitis is a clinicopathologic syndrome characterized by chronic, watery diarrhea and a distinctive subepithelial collagen layer in colonic mucosa. The etiology of this syndrome is not known. Although no clear relationship to UC or CD has been recognized, some investigators suggest that the ultrastructural changes seen in collagenous colitis can not be distinguished from those seen in chronic, fibrotic UC. Whether there are any pathogenic mechanisms common to UC and some cases of collagenous colitis is unknown. Although it is possible that ANCA in patients with UC and ANCA in a subset of

-patients with collagenous colitis may be directed against the same antigen(s) , it is much more likely that ANCA in these diseases are heterogenous. Purification of antigen identified by the ANCA associated with active WG has allowed development of a highly specific ELISA for diagnosis of WG.

Elevated levels of neutrophil binding and elevated titers of anti-neutrophil IgG are observed with some sera from patients with forms of chronic liver disease other than PSC, however, the levels and titers are generally lower than in the PSC, and the IF patterns can be clearly distinguished from the pattern typical of PSC. The results suggest that the combination of a positive value in the fixed neutrophil ELISA and a perinuclear IF pattern will distinguish PSC from another chronic inflammatory liver disease (provided, of course, that ulcerative colitis does not coexist with the other chronic liver disease) . Isolation of the specific antigen(s) recognized by the UC- and the PSC-associated .ANCA may permit development of an improved assay(s) for a disease marker(s) in UC and PSC, as has been achieved in WG. The UC- and PSC-associated ANCA differ from those that are specific for active WG in several respects. In contrast to the WG-associated ANCA which bind to cytoplasmic components of both neutrophils and monocytes, the UC-associated ANCA do not bind to monocytes. Additionally, in UC and PSC, ANCA is distributed in perinuclear IF patterns while in WG ANCA is seen as a diffuse, granular cytoplasmic pattern. The WG-associated ANCA appear to be a marker of disease activity, but levels of neutrophil binding by ANCA in UC and PSC do not correlate with disease activity, extent of disease, or duration of disease. Finally, the IF pattern exhibited by the majority of sera from patients with UC and PSC is distinct from the pattern characteristic of active WG.

Perinuclear staining of alcohol-fixed neutrophils has been attributed, in some instances, to antibodies directed against myeloperoxidase, elastase, or cathepsin G. Previously, antibodies to myeloperoxidase in some but not all sera from patients with UC and PSC have been observed. The frequency of antibodies to myeloperoxidase and their relationship to ANCA in UC and PSC is currently being examined. Significant anti-elastase or anti-cathepsin G activity has not been found in sera from patients with UC or PSC which exhibit high levels of binding in the fixed neutrophil ELISA and a perinuclear IF pattern, suggesting that the UC- and PSC-associated ANCA are not directed against elastase or cathepsin G. Other studies have not detected reactivity of UC sera with myeloperoxidase, elastase, or neutrophil primary granules.

ANCA that are distinct from those specific for active WG are present in the majority of sera from patients with UC, UCPC, the majority of patients with PSC and the majority of patients with PSC and no endoscopic or histologic evidence for UC. Levels of neutrophil binding by IgG and titers of anti-neutrophil IgG in sera from patients with UC and UCPC are significantly greater than the mean levels for sera from normal controls and patients with other colitides and diarrheal illnesses, and, with the exception of collagenous colitis, the levels of neutrophil binding for these other diarrheal diseases are not significantly different than the levels for normal controls. The pathophysiologic importance of the UC-and PSC associated ANCA is not known, but the results support the hypothesis that they are not simply an epiphenomenon related to active colonic inflammation. They may be a manifestation of altered immune regulation in these diseases.

The following examples are intended to illustrate but not limit the invention.

EXAMPLE I

Study population and serum samples

a. Ulcerative Colitis/Diarrheal Disease

The study population consisted of 40 patients with UC (including a subset of 23 patients with recent-onset UC; that^"is, with symptoms of UC for less than 1 year), 27 with UCPC (median of 1.3 years and range of 0.1-17.9 years post colectomy) , 18 with colonic CD, 35 with collagenous colitis, 19 with bacterial or amoebic colitis, 27 with the irritable bowel syndrome and diarrhea (IBS-diarrhea) , and 18 with other miscellaneous diarrheal illnesses (2 with pseudomembranous colitis, 1 with diverticulitis, 3 with radiation colitis or proctitiε, 2 with subacute colonic schistoso iasis, 2 with eosinophilic colitis, 1 with systemic mastocytosis, 1 with neutropenic colitis, 1 with persistent diarrhea following infection with Gi^rdia Iambiia, 1 with diarrhea secondary to laxative abuse, and 4 with acute self-limited diarrhea) . Twenty-five individuals without acute or chronic gastrointestinal or inflammatory disease served as normal controls. All sera were stored at -70^βC until assayed.

b. Primary Sclerosing Cholangitis

The study population consisted of 19 patients with UC only, 51 with PSC, and 49 liver disease controls (20 with primary biliary cirrhosis, 15 with chronic hepatitis B, 13 w-th chronic non-A, non-B (NANB) hepatitis, and 1 with chronic autoimmune hepatitis) . All patients were evaluated at the UCLA Center for the Health Sciences or the Mayo Clinic. Diagnosis were based upon appropriate clinical, endoscopic, serum biochemical, serological, cholangiographic and/or histologic criteria. Of the 51 patients with PSC, 37 had well documented UC and 14 had no endoscopic or histologic evidence for UC based on recent sigmoidoscopy or colonoscopy with biopsies. Liver histology was available for 44 of the patients with PSC, allowing subgrouping into early (stage I and II) or advanced (stage III and IV) disease categories based on previously published criteria. Thirty-two individuals without chronic gastrointestinal, hepatic, or inflammatory disease served as normal controls. All sera were stored at -70"C until assayed.

EXAMPLE II Neutrophil isolation

All studies were performed with neutrophils from a single normal individual. Neutrophils were isolated from peripheral blood by Ficoll-Hypaque (specific gravity 1.080) density centrifugation followed by dextran sedimentation of the cell pellet as detailed in Boyum et al., Scand J. Clin. Lab. Invest. 21:31-50, 1968, which is incorporated herein by reference. The buoyant neutrophils were recovered and washed, and residual red blood cells were lysed by hypotonic shock. Monocytes were purified by adherence to gelatin/plasma coated plastic dishes as described in Anton et al., J. Clin. Immunol., 8:148-156, 1988, which is incorporated herein by reference.

EXi^MPLE III Fixed neutrophil ELISA for anti-neutrophil antibodies

Test sera were thawed at room temperature, centrifuged for 15 minutes at 15,600 g in 0.2 μm Spin-X microcentrifuge filter tubes (Costar, Cambridge, MA) , and coded so that the investigators performing the assays were blinded to the diagnosis. Microtiter plates (Immulon 2, Dynatech Laboratories, Alexandria, VA) were coated with a monolayer of neutrophils by the addition of 100 μl/well of Hanks' balanced buffered salt solution (HBSS) which contained 250,000 neutrophils. After the cells had settled and spread for 30 minutes at room temperature, the plates were centrifuged at 1000 rpm (300 g) for 5 minutes, the supernatant was aspirated from the wells, and the plates were air dried. The cells were fixed with 100% methanol for 10 minutes after which the plates were air dried and stored at -20°C. For use, the plates were brought to room temperature, and 150 μl of 0.25% bovine serum albumin in phosphate buffered saline (BSA/PBS) were added to each microtiter well for 1 hour to block non-specific binding. The blocking material was discarded, and 100 μl of test serum diluted in BSA/PBS (or BSA/PBS alone for blank wells) were added. To standardize the assay, a positive sera pool consisting of a mixture of sera from 6 individuals with UC

(3 with very high and 3 with intermediate levels of ANCA) was used. This was included on each microtiter plate at dilution of 1:100. The plates were incubated for 1 hour at room temperature in a humidified box. They were then washed 3 times with 0.05% Tween 20 in PBS (PBS/Tween) , and 100 μl/well of 1:750 dilution of alkaline phosphatase coupled goat anti-human gamma chain specific antibody (Tago, Burlingame, CA) in BSA/PBS were added for 1 hour. This antibody was discarded, and the wells were washed 3 times with PBS/Tween and 4 times with 0.05 M Tris base in 0.9 M NaCl, pH 7.5. Substrate solution (1.5 mg/ml disodium p-nitrophenol phosphate in 0.01 M Tris base, 0.0025 M MgCl₂, pH 8.6, 100 μl/well) was added, and the color development was allowed to proceed until absorbance at 405 nm in the positive control wells was 0.8-1.0 optical density units greater than that in blank wells. Results for test sera were expressed in terms of percent of positive control after correction for background as measured in blank wells. A result exceeding 2 SD above the mean for the normal control sera diluted 1:100 was considered positive.

The use of a fixed neutrophil ELISA as a screening test to detect serum anti-neutrophil IgG for ANCA in inflammatory bowel disease has been validated (Saxon et al., in press). Diluting test sera to 1:100 gave the best separation between sera from patients with UC and sera from normal controls. At dilutions more concentrated than 1:100, the levels of neutrophil binding by UC sera tended to plateau. In the present study, a dilution of 1:100 yielded the greatest sensitivity for UC and a good separation between sera from patients with UC and sera from patients with other colitides and diarrheal illnesses.

At a dilution of 1:100, the levels of neutrophil binding (Figure 1) by sera from patients with UC and UCPC were not statistically different from each other (p = 0.239), were each significantly greater than the levels for the normal controls (p < 0.001 in each case), and were each significantly greater than the levels _υr collagenous colitis, colonic CD, bacterial/amoebic colitis, IBS- diarrhea, and miscellaneous (primarily inflammatory) diarrϊ* al illnesses (p < 0.001 in each case). Similarly, the levels of neutrophil binding by sera from the subset of patients with recent-onset UC were significantly greater than the levels for the normal controls and each of the I disease controls (p < 0.001 in each case except for collagenous colitis where p = 0.005). There were no significant differences between the levels of binding for the normal controls and the levels for colonic CD, bacterial/amoebic colitis, IBS-diarrhea, and miscellaneous diarrheal illnesses (p > 0.05 in each case) . The levels of binding by sera from patients with collagenous colitis were significantly greater than the levels for the normal controls (p = 0.003).

All sera were also titered in the fixed neutrophil ELISA. A level of binding in the ELISA exceeding 2 SD above the mean for the normal control sera diluted 1:100 was defined as a positive value (Figure 2) . There was a significant correlation between levels of binding at a dilution of 1:100 (Figure 1) and titers (Figure 2) in the ELISA (r = 0.87, p < 0.001). Titers for sera from patients with UC, recent-onset UC, and UCPC were significantly greater than the titers for collagenous colitis, colonic CD, bacterial/amoebic colitis, IBS-diarrhea, and miscellaneous (primary inflammatory) diarrheal illnesses (p < 0.001 in each case).

For sera from patients with UCPC, there was no significant correlation between the number of years post colectomy and either the levels of neutrophil binding at a dilution of 1:00 (r = 0.10, p = 0.614) or anti-neutrophil

IgG titers (r = 0.10, p = 0.618).

The majority of sera from patients with UC (15/19, 79%), PSC (42/51, 82%) and PSC without UC (11/14, 79%) were positive in the ELISA at a dilution of 1:100, and the levels of neutrophil binding for each of these groups were significantly greater than the levels for normal cont^rols (p < 0.001 in each case). The levels of binding for UC were not significantly different than the levels for PSC (p = 0.181) or the subset of PSC without UC (p = 0.358). Although 31/49 (63%) liver disease control sera were also positive in the ELISA, the levels of binding for primary biliary cirrhosis, chronic hepatitis B, and chronic NANB hepatitis were significantly less than the levels for PSC (p < 0.001 in each case). Similarly, the levels of binding for sera from the subset of patients with PSC and no endoscopic or histologic evidence for UC significantly exceeded the levels for primary biliary cirrhosis (p = 0.032), chronic hepatitis B (p = 0.008), and chronic N.ANB hepatitis (p = 0.011).

All sera were also titered in the ELISA. Levels of binding at a dilution of 1:100 and titers (Figure 1) in the ELISA for individual serum samples were compared, and there was a significant correlation (r = 0.92 p < 0.001). Titers for sera from patients with UC were significantly different than the titers for PSC (p = 0.128) or PSC without UC (p = 0.432). Titers for sera from patients with PSC were significantly greater than the titers for primary biliary cirrhosis, chronic hepatitis B, and chronic NANB hepatitis (p < 0.001 in each case). Similarly, titers for sera from the subject of patients with PSC and no endoscopic or histologic evidence for UC exceed the titers for primary biliary cirrhosis with a trend towards significance (p = 0.054) and significantly exceeded the titers for chronic hepatitis B (p = 0.027) and chronic NANB hepatitis (p = 0.004) .

There were no significant differences between sera from patients with PSC and histologically early disease

(n=14) and patients with PSC and histologically advanced disease (n=30) in either levels of binding at a dilution of 1:100 (p = 0.0376) or titers of anti-neutrophil IgG (p = 0.311) .

The combination of a positive value in the fixed neutrophil ELISA and a perinuclear IF pattern was found with 60% of sera from patients with UC and 63% of sera from patients with UCPC. In contrast, this combination was found with only 6% of sera from patients, with other colitides and diarrheal illnesses, and most of these 6% were sera from patients with collagenous colitis. Therefore, the combination of a positive value in the ELISA and a perinuclear IF pattern was 60% sensitive and 94% specific for UC. The combination of a positive value in the fixed neutrophil ELISA and a perinuclear IF pattern was 65% sensitive and 100% specific for PSC when compared to the other chronic inflammatory liver diseases.

EXAMPLE IV Indirect immunofluorescence assay

Glass slides with approximately 100,000 neutrophils/slide were prepared by cytocentrifugation (Shandon Cytospin, Cheshire, U.K.), fixed in 100% methanol at room temperature for 10 minutes, air dried, and stored at -20°C. Coded sera were tested at a dilution of 1:20 and stained with fluorescein-labelled F(ab')2 gamma chain specific antibody as described in Chng et al., Ann. Allergy 63:411-416, 1989, which is incorporated herein by reference. Slides were examined with an epifluorescence- equipped microscope. Monocytes were used as the substrate for indirect immunofluorescence (IF) staining in one experiment.

Because the fixed neutrophil ELISA detects any IgG antibodies directed against neutrophils and does not yield any information as to the component(s) of the neutrophil recognized by these antibodies, all sera that exhibited levels of binding in the ELISA exceeding 2 SD above the mean for the normal controls at a dilution of 1:100 (considered to be a positive value in the ELISA) were also examined in the indirect immunofluorescence assay (IF) (Table I) . The individuals performing the assay remained blinded to the diagnoses of sera donors. Perinuclear staining was the predominant IF pattern exhibited by ELISA- positive sera from patients with UC (25/35) , recent-onset UC (12/18) , and UCPC (17/21) . Perinuclear staining was also seen with some ELISA-positive sera from patients with collagenous colitis (5/14), colonic CD (1/5), amoebic/bacterial colitis (1/1) , PSC (33/42) , and ELISA- positive PSC without UC sera (8/11) . One serum from a patient with IBS-diarrhea exhibited IF staining that appeared to be homogenous cytoplasmic in some areas of the slide and perinuclear in other areas of the slide. In contrast, none of the ELISA-positive liver disease control sera exhibited the perinuclear IF pattern. Homogenous nuclear staining was exhibited by serum from one patient with UCPC. This patient suffers from a deforming polyarthritis, and his serum is positive for anti-nuclear antibodies. The majority of the remaining sera exhibited ho ogenr is, smooth, cytoplasmic staining.

TABLE I

INDIRECT IMMUNOFLUORESCENCE PATTERNS

EXHIBITED BY ELISA-POSITIVE SERA

Ulcerative colitis 15/19 13 2 (cytoplasmic)

Primary sclerosing 42/51 33 7 (cytoplasmic) cholangitis 2 (cytoplasmic/ dim nuclear)

Primary biliary 12/20 4 (cytoplasmic cirrhosis specks) 4 (cytoplasmic) 2 (cytoplasmic/ dim nuclear) 2 (nuclear)

Chronic hepatitis B 11/15 8 (cytoplasmic) 2 (cytoplasmic/ dim nuclear) 1 (nuclear)

Chronic NANB 7/13 4 (cytoplasmic) hepatitis 2 (cytoplasmic/ dim nuclear) 1 (nuclear)

Autoimmune hepatitis 1/1 1 (nuclear)

* Positive ELISA defined as a value greater than 2 SD above the mean for the normal controls. E.XAMPLE V Statistical analysis

Differences between groups in levels of binding, and titers, of anti-neutrophil IgG were assessed using the Mann-Whitney U test. The Spearman rank order correlation test was used to correlate the titers and levels of binding in the fixed neutrophil ELISA, and was also used to correlate the titers and levels of binding of anti- neutrophil IgG with the number of years post colectomy for sera from patients with UCPC.

E.XAMPLE VI Elastase and Cathepsin G Binding Assays

Because elastase and cathepsin G antibodies have been reported to exhibit perinuclear IF staining of alcohol- fixed neutrophils, sera from 10 patients with UC and 8 patients with PSC and no endoscopic or histologic evidence for UC which had previously exhibited high levels of binding in the fixed neutrophil ELISA and a perinuclear IF pattern were selected for examination in elastase and cathepsin G binding assays. Microtiter plates were incubated for 1 hour at room temperature and then overnight at 4^βC with 100 μl/well of a solution of 10 μg/ml of purified human neutrophil elastase (Elastin Products, Pacific, MO) or 10 μg/ml purified human neutrophil cathepsin G (Biodesign, Kennebunkport, ME) in PBS. They were then washed 3 times with PBS and blocked with 150 μl/well of BSA/PBS. After discarding the blocking solution, 100 μl of test serum diluted 1:100 in BSA/PBS (or BSA/PBS alone for blank wells) were added. Rabbit anti- human neutrophil elastase immunoglobulin (Biodesign, Kennebunkport, ME) or rabbit anti-human cathepsin G immunoglobulin (Biodesign, Kennebunkport, ME) was diluted in BSA/PBS in 10-fold serial dilutions ranging from 1:10 to 1:100,000, and 100 μl/well of each dilution were placed into the corresponding positive control wells. The plates were incubated for 2 hours at room temperature in a humidified box. They were then washed 3 times with PBS/Tween, 100 μl/well of a 1:750 dilution of alkaline phosphatase coupled goat anti-human gamma chain specific antibody (Tago, Burlingame, CA) in BSA/PBS were added to test sera wells, and 100 μl/well of a 1:500 dilution of alkaline phosphate coupled mouse anti-rabbit IgG (H & L) antibody (Jackson Immuno Research Laboratories, Avondale, PA) in BSA/PBS were added to positive control wells. The plates were incubated for 2 hours at room temperature in a humidified box. The alkaline phosphatase coupled antibodies were discarded, and the wells were washed 3 times with PBS/Tween and 4 times with Tris/NaCl. Substrate solution (1.5 mg/ml disodium p-nitrophenol phosphate in 0.01 M Tris base, 0.0025 M MgCl₂, pH 8.6, 100 μl/well) was added, and color development was allowed to proceed until absorbance at 405 nm in the lowest dilution positive control wells exceeded 1.5 optical density units greater than that in blank wells.

EXi^MPLE VII UC- and PSC-Associated Anti-Neutrophil Cytoplasmic Antibodies do not Bind to Monocytes. Elastase. or Cathepsin G

Sera from 5 patients with UC and 5 patients with PSC and no endoscopic or histologic evidence for UC which exhibited high levels of binding in the fixed neutrophil ELISA and a perinuclear IF pattern with neutrophils as the substrate did not exhibit detectable staining in an IF assay using monocytes as the substrate. In contrast, serum from a patient with WG, previously known to exhibit the diffuse, granμlar, cytoplasmic IF pattern of active WG, stained both neutrophils and monocytes.

Sera from 10 patients with UC and 8 patients with PSC and no endoscopic or histologic evidence for UC which were known to exhibit high levels of binding in the fixed neutrophil ELISA and a perinuclear IF pattern did not exhibit significant binding in either the elastase or cathepsin G binding assays, while the respective positive control antibodies exhibited detectable binding to the antigen-coated plates at all dilutions.

Although the invention has been described with reference to the presently-preferred embodiment, it should be understood that various modifications can be made without departing from the spirit of the invention. Accordingly, the invention is limited only by the following claims.

Claims

WE CLAIM:

1. A method for detecting the presence of a disease state selected from the group consisting of ulcerative colitis or primary sclerosing cholangitis in a patient, comprising the steps of:

a. contacting a body sample of said patient with immobilized neutrophils so as to allow anti¬ neutrophilic cytoplasmic antibodies to bind to said neutrophils; b. detecting the presence of bound anti¬ neutrophilic cytoplasmic antibodies, wherein the presence of said bound anti-neutrophilic cytoplasmic antibodies indicates the presence of said disease state.

2. The method of claim 1, wherein said detecting step further comprises contacting said bound anti¬ neutrophilic cytoplasmic antibodies with labelled antibodies specifically reactive with anti-neutrophilic cytoplasmic antibodies.

3. The method of claim 2, wherein said label is a component of an enzymatic reaction.

4. The method of claim 3, wherein said component of an enzymatic reaction is alkaline phosphatase.

5. The method of claim 1, wherein said body sample is serum.

6. The method of claim 1 further comprising the steps of analyzing by immunofluorescent staining the location of binding of said anti-neutrophilic cytoplpsmic antibodies, wherein the presence of perinuclear binding of said anti-neutrophilic cytoplasmic antibodies further confirms the presence of said disease.