Abstract

Objectives: Perinuclear anti-neutrophil cytoplasmic antibody (p-ANCA) and anti-Saccharomyces cerevisiae antibody (ASCA) have been proposed as markers for diagnosis and for subtyping of inflammatory bowel disease (IBD). The aim of this study was to investigate the association of p-ANCA and ASCA with a 10-year disease outcome in terms of cumulative rate of colectomy and relapse in a population-based European inception cohort of ulcerative colitis (UC) patients.

Methods: Serum samples from 432 consenting patients were analysed for p-ANCA and ASCA. The results were compared with the cumulative colectomy rate, relapsing disease and total number of relapses. We used multiple regression analyses adjusted for age, sex, residence, disease extent at diagnosis, smoking, familial IBD and drug treatment to study the relationship between serological values and disease course.

Results: The relapse rate was higher in the p-ANCA-positive patients: 82% (95% confidence interval [CI] 75–89%) compared with 67% (CI 62–72%, p = 0.011) in the p-ANCA-negative patients. The risk of relapsing disease course was higher by a factor of 1.4 (CI 1.1–1.8, p = 0.009) for p-ANCA-positive patients than for p-ANCA-negative patients, and the corresponding relative risk (RR) for the total number of relapses was 1.9 (CI 1.7–2.1, p < 0.001). In ASCA-positive patients RR for the total number of relapses was 1.8 (CI 1.5–2.1, p < 0.001). No significant association with colectomy rate was found for the presence of either p-ANCA or ASCA.

Conclusion: UC patients positive for p-ANCA and possibly for ASCA may have a more unfavourable long-term disease outcome in terms of relapse than UC patients without these markers.

Introduction

It has been proposed that serological markers such as the perinuclear anti-neutrophil cytoplasmic antibody (p-ANCA) and the anti-Saccharomyces cerevisiae antibody (ASCA) are potential diagnostic tools in inflammatory bowel disease (IBD). P-ANCA has been detected in 50–80% of patients with ulcerative colitis (UC) and in 10–30% with Crohn's disease (CD).1 ASCA has been found in 40–70% in CD and 5–15% in UC patients. In a population-based cohort of IBD patients, the combination of a positive p-ANCA and a negative ASCA had a positive predictive value of 75% for the diagnosis of UC.2 However, although the predictive value of these serological markers for defining sub-groups of IBD patients1 has been investigated, there are diverging opinion with regard to their association with a long-term disease outcome.

The aims of the present study were to evaluate p-ANCA and ASCA as markers of disease outcome by determining the cumulative colectomy rate, the cumulative number of patients with first relapse after diagnosis and the total number of relapses in a population-based multicentre European–Israeli inception cohort of UC patients followed for 10 years.

Methods

Study population

Between October 1 1991 and September 30 1993, a total of 2201 patients with IBD at 20 treatment centres in 12 European countries and Israel were recruited at inception in a prospective population-based study conducted by the European Collaborative Study Group of Inflammatory Bowel Disease (EC-IBD).3 UC was diagnosed in 1379 of the patients according to the criteria established by Lennard-Jones and Truelove and Witts.4 The patients were treated and followed according to the established procedures at the individual centre. After 10 years all the treatment centres were invited to participate in a follow-up study. Thirteen of the centres were able to participate; the others dropped out for technical or logistical reasons. Four of the 13 did not meet the criterion of at least 60% patient follow-up, which was arbitrarily set in order to maintain the population-based nature of the study.

Data collection

The patients were invited to a 10-year follow-up visit, which took place between August 1 2002 and January 31 2004. Data from a patient questionnaire and a patient-per-physician follow-up form were registered in a common database, which could be accessed by all the participating centres. Supplementary information was gathered from hospital records when necessary. The study population, including the follow-up, has previously been described in more detail.5,6

Definitions

Relapse

A relapse was defined as an increase in UC-related symptoms leading to consultation with a physician and resulting in an increase in the dose of ongoing medical treatment, the introduction of new medication, or surgery. Patients without symptoms or whose symptoms were too slight to fulfill these criteria were considered to be in clinical remission.

Colectomy

Data on total or partial resection of the colon and/or rectum with or without additional surgical procedures were noted at the10-year interview or obtained from hospital records.

Geographical location

The centres in Greece, Israel, Italy and Spain were classified as southern centres and those in Denmark, the Netherlands and Norway as northern centres.

Education level

The highest level of completed education was noted at the 10-year interview, and the patients were divided into two categories for comparison: primary (0–9 years) combined with lower secondary school (10–11 years), and upper secondary school (12 years or more) combined with higher education.

Disease extent

The extent of inflammation at diagnosis was assessed mainly on the basis of colonoscopy, but in a small number of cases distal colonoscopy was combined with double contrast barium enema. Disease activity proximal to the splenic flexure was classified as extensive colitis, colonic inflammation distal to the splenic flexure as distal colitis, and disease limited to the rectum 15 cm above the linea dentata as proctitis.

Smoking

At the 10-year interview the patients were asked about their smoking history. Patients who had never been daily cigarette smokers were classified as non-smokers; patients who had stopped smoking in any calendar year prior to the year they entered the study were classified as ex-smokers before study inclusion; and those who had stopped smoking in any year including the calendar year they entered the study were classified as ex-smokers after study inclusion. All patients who at the 10-year interview had an average consumption of at least one cigarette daily, were classified as current smokers regardless of when they started smoking. The cumulative cigarette consumption in pack years throughout the smoking period was assessed for the current and ex-smokers.

Relatives with IBD

Familial occurrence of one or more first-degree relatives with IBD was noted at the 10-year interview.

Medical treatment

Information on medical treatment was obtained from the medical records that had been kept up to date during follow-up. For patients who had taken medication for one or more periods, the proportion of the follow-up period during which they had used the drug (referred to as the use period) was calculated. The most commonly used systemic drugs were chosen for analysis and divided into three groups: aminosalicylates (5-ASA), including sulfasalazine and 5-aminosalicylic acid, glucocorticosteroids (GCS), and azathioprine or 6-mercaptopurine (AZA/6MP). The use of 5-ASA and GCS was closely related to relapse as defined in this study, but otherwise we had no information about whether these drugs were used as maintenance or to treat relapse. We therefore excluded them from the regression analyses. AZA/6MP was included in regression analysis because it is mainly used for maintenance of remission. The use of other drugs for the treatment of UC was considered negligible due to the small number of patients concerned, and these drugs were therefore excluded from further analysis.

Serological analysis

At the end of the follow-up whole blood and serum samples for serological testing were obtained at the 10-year visit from the consenting patients. The analysis of ASCA was performed by a standard ELISA method (Medipan Diagnostica, Germany)7 and interpreted by calculating the binding index (BI). ASCA (either IgG or IgA) were considered positive at BI > 1. P-ANCA IgG was determined by indirect immunofluorescence using ethanol-fixated neutrophil slides (Inova Diagnostics, San Diego, Calif).8 All samples were frozen and shipped to the University Hospital Gasthuisberg, Leuven, Belgium, for analysis.

Statistical analysis

The relapse rate was calculated as the cumulative rate for the first relapse in patients with at least one relapse after diagnosis. The number of relapses was calculated as the total number of relapses per unit of time at risk, which was defined as time from diagnosis until time of the 10-year interview, death, most recent consultation or colectomy. The cumulative number of patients with a first relapse was calculated by the Kaplan-Meier method, and relapse rates between sub-groups were compared using log-rank tests. The Mann–Whitney, Kruskal–Wallis, chi-squared and Fisher's exact tests were used when appropriate to identify possible differences in sex, age, geographical residence, smoking status, drugs and extent of disease between patients with and without serological test results. Adjustments for co-factors associated with the first relapse or with colectomy were made by Cox regression using the SPSS statistical software package version 14.0 (SPSS Inc., Chicago IL). On the assumption that the number of relapses would have a Poisson distribution, we used a Poisson regression model to adjust for co-factors for the total number of relapses for each participant. The follow-up time was included as an offset term to allow for different at-risk times for each patient. The Poisson model was fitted using the R software version 2.2.0 (the R Project for Statistical Computing). The two models provide information on two related but supplementary aspects of the development of the disease. Cox regression models the time to first relapse and important factors during this initial period. Poisson regression models the total number of relapses during the follow-up period and factors associated with the total burden of disease. P values less than 0.05 were considered statistically significant.

Ethical considerations

The study protocol was approved by the local committee for medical research ethics for all the participating centres and informed consent was obtained from all patients before they entered the study.

Results

Patients and centres

Data for the patients included in the 10-year follow-up are presented in Table 1. Of the 20 centres in the original cohort, nine centres in seven countries met the criterion of at least 60% patient follow-up. Status of the 781 patients from these centres diagnosed with UC has been outlined in Fig. 1. Ten patients had not been followed up and were therefore excluded, while 771 had been followed up for varying time periods. Blood samples were collected from 432 of the patients who attended the 10-year visit. The median follow-up period for these patients was 123 months (range 107–133). Of the 339 patients who were not tested for serology, 73 had died, 81 were lost during follow-up, and 185 patients were not tested for other reasons.

There were no significant differences in sex, north–south residence, number of colectomies or disease extent at diagnosis between patients tested for serology (n = 432) and non-tested (n = 339) patients. The tested patients had a median age (inter-quartile range, IQR) of 38 years (29–49) at diagnosis, which was significantly lower than the median age of the non-tested patients, which was 42 years (28–64) (p = 0.001). However, when the patients who died were excluded, no significant difference in age between the two groups was found. With regard to outcome variables, there were no significant differences in colectomy rates, but a significantly higher cumulative rate of relapsing disease of 0.71 (CI 0.67–0.75) versus 0.61(CI 0.55–0.67, p = 0.009) and a significantly higher median number of relapses per year at risk of 0.19 versus 0.10 (p = 0.002) were found for the 432 tested patients than for the 339 not tested. Among the 617 patients who attended the 10-year visit, no significant differences in relapse rate or number of relapses were found between tested and non-tested subjects.

Serological results

Of the 432 UC patients who were tested, 105 (24%) were p-ANCA-positive, with a variation between the centres of 7% to 36%. Twenty-eight (6%) of the UC patients were ASCA-positive, with a variation between the centres of 3% to 25%.

Disease outcomes

The cumulative colectomy rate, cumulative rate of first relapse and median number of relapses during disease course are presented in Table 2. The status of p-ANCA was significantly related to first relapse. In p-ANCA-positive patients the 10-year cumulative rate of first relapse was significantly higher than in p-ANCA-negative patients. The difference in relapse rates between p-ANCA-positive and p-ANCA-negative patients increased with time, as shown by the Kaplan-Meier plot in Fig. 2. No significant difference was found between ASCA-positive and ASCA-negative patients with regard to risk of first relapse. The total number of relapses during follow-up was significantly higher in p-ANCA-positive than in p-ANCA-negative patients. We found no significant relationship to total number of relapses for ASCA-positive patients compared with ASCA-negative patients with the Mann–Whitney test (p = 0.09), but the relationship was significant with a Poisson regression model (p < 0.001). ASCA-positive patients had a higher risk of colectomy than ASCA-negative patients, but the difference was not statistically significant (p = 0.09). Of the 28 ASCA-positive patients, five had undergone colectomy, and of the eight patients who were positive for both ASCA and p-ANCA, three had undergone colectomy. No significant difference in colectomy rate was found between p-ANCA-positive and p-ANCA-negative patients.

Medical treatment

For all the 432 patients tested, the median drug period was compared between the p-ANCA-positive and p-ANCA-negative patients, and between the ASCA-positive and ASCA-negative patients (Table 3). A significantly higher median use percentage for 5-ASA (p < 0.001) and GCS (p = 0.005) was seen in the p-ANCA-positive patients. The median use percentage for AZA/6MP was higher in the ASCA-positive than the ASCA-negative patients (p < 0.001).

Smoking

A history of cigarette smoking was obtained from 407(93%) of the 432 patients (Table 1). The median (IQR) for cigarette exposure in terms of number of pack years was 12 (5–20) for ex-smokers before study inclusion, 12 (5–31) for ex-smokers after study inclusion and 15 (6–28) for current smokers. The differences were not statistically significant (p = 0.5). The prevalence of p-ANCA was 12% in patients who were smokers at diagnosis versus 28% in non-smokers (p = 0.001), and related to smoking status at the end of the study the figures were 11% and 26% respectively (p = 0.009). The prevalence of ASCA was higher among smokers than among non-smokers both at diagnosis and at the 10-year follow-up, but the difference was not statistically significant (9% versus 5% at diagnosis, p = 0.12, and 11% versus 5% at the 10-year follow-up, p = 0.10).The prevalence of p-ANCA and ASCA respectively among patients who quit smoking during follow-up was not significantly different from the prevalence among current smokers.

Regression analyses

We used multivariate regression analysis to study the relationships between the serological variables and disease course, accommodating any modulating effect of potential confounding variables. The potential confounding variables were age, sex, level of education, disease extent, north–south residence, familial IBD, treatment with azathioprine and smoking status. Multivariate Cox regression analysis showed an adjusted RR for first relapse of 1.4 (CI 1.1–1.8, p = 0.009) for p-ANCA-positive patients and a non-significant difference between ASCA-positive and ASCA-negative patients. Multivariate Poisson regression analysis of the total number of relapses during follow-up showed significant differences in adjusted RR both for p-ANCA, 1.9 (CI 1.7–2.1, p < 0.001), and for ASCA, 1.8 (CI 1.5–2.1, p < 0.001).

Discussion

In this study we found a low prevalence of p-ANCA among UC patients, and a significantly higher rate of first relapse and total number of relapses in the p-ANCA-positive than in the p-ANCA-negative patients after 10 years of follow-up. A significantly higher total number of relapses among ASCA-positive patients was also shown by Poisson regression.

The reported prevalence of p-ANCA among UC patients is generally higher than what we found.9 The reason for this could be that most of these studies have been performed in referral populations and not in population-based cohorts, as in our study. Differences in the methods of analysis could also explain this discrepancy. The combination of IgG with IgA antibody titre could have revealed a higher prevalence of p-ANCA.10 In a population-based cohort from Rochester, Minnesota, serum samples from 83 UC patients were tested for p-ANCA using five different assays, and the prevalence of p-ANCA varied from 0 to 63%.2 The test used in our study was based on IgG detection, which showed a positive p-ANCA in 49.7% of 147 consecutive UC patients at a university referral centre.8 We therefore believe that the low prevalence of p-ANCA found in our cohort is due to the population-based nature of the study and to geographical variations between the centres.

Our findings support the hypothesis that p-ANCA is associated with a more aggressive disease during follow-up. Furthermore, the larger difference in relapse rates over time in these patients is consistent with p-ANCA being a marker of a poor long-term disease outcome. The combination of p-ANCA and ASCA has been proposed as an adjunctive diagnostic tool in IBD.11 In studies of sub-groups of UC patients, a positive p-ANCA has been associated with therapy-resistant left-sided disease,12 more aggressive disease, a more frequent need for early surgery1 and the development of pouchitis in ileal pouch anal anastomosis.13,14 However, the implication of these results for long-term disease outcome in UC has not been investigated before.

A number of studies have failed to show any correlation between disease activity and the presence of p-ANCA.1517 However, in others a lower prevalence of p-ANCA among UC patients in remission has been reported,18 and a positive titre of p-ANCA has been found to be associated with more aggressive UC.19,20 However, most of these studies have been carried out on a small number of patients, using different laboratory methods. Furthermore, no uniform definitions of disease activity, study outcome or follow-up routines have been used. These factors could explain the inconsistencies in the results reported so far.

Colectomy risk was not related to p-ANCA status. Extent of disease in our cohort was shown in a previous study to be related to the risk of colectomy,21 but not to the risk of relapse.22 The formation of p-ANCA has been proposed to be a consequence of a mucosal immune response with high specificity for UC.23 If so, the immune response could result in an increase in inflammatory activity and p-ANCA formation irrespective of disease extent or colectomy risk. Furthermore the decision to perform colectomy may be influenced by local treatment strategies for disease resistant to initial medical therapy or for prevention of colorectal cancer.24 Therefore, when investigating the impact of p-ANCA status, colectomy rate may be a less reliable marker of disease activity than relapse.

Positive titre of ASCA has been associated with CD, especially stricturing and penetrating disease,25 and with surgery26 and higher health care costs.27 However, there are few if any reports on the implications of positive ASCA in UC patients. Our results indicate that positive ASCA may be a marker of more aggressive disease in UC as well, as measured by the overall risk of relapse in the first 10 years of disease course and the relationship to the use of azathioprine/6MP. However, these results must be interpreted with great caution due to the low prevalence of positive ASCA among UC patients. Given the broad confidence interval in the Mann–Whitney U-test, the discrepancy between the results of the two statistical analyses of ASCA as a risk factor for total number of relapses could be due to a type-II statistical error.

The use period for both GCS and 5-ASA represented a significantly higher percentage in p-ANCA-positive than in negative patients. However, this should be regarded as an epiphenomenon owing to the frequent use of these drugs in the treatment of relapses that we found to be associated with positive p-ANCA. The higher levels of AZA/6MP in our ASCA-positive patients may also reflect more aggressive disease, but due to the small number of patients the higher levels could also be coincidental. The results of the multivariate regression analyses showed that the use of AZA did not contribute significantly explain the relation between disease outcome and p-ANCA status.

Cigarette smoking has been shown to protect against UC and may also alleviate the clinical course.28 The low prevalence of p-ANCA in smokers compared with non-smokers at the time of diagnosis in our study is consistent with this reported finding. Factors responsible for the effect of smoking on the development of UC may also reduce the formation of p-ANCA as an epiphenomenon. Since the prevalence of p-ANCA did not differ between patients who stopped smoking during follow-up and those who continued smoking, it is not likely that the protective effect of smoking on UC is mediated by a change in the status of p-ANCA. Other environmental factors likely to influence disease activity, such as treatment with GCS or colectomy, may also change p-ANCA status from positive to negative.19 This could explain the difference in serological status, even though it was not supported by our data.

The design of cohort formation and follow-up used in our study has a strong methodological advantage. However, although we found that a number of demographic and phenotypic variables were comparable in tested and non-tested patients, this does not entirely exclude a selection bias. We have previously demonstrated an inverse relationship between age at diagnosis and risk of relapse.22 The lower risk of relapse found in the non-tested group may have been caused by the greater age of the patients who had died. If there was a low risk of relapse among the patients who did not attend follow-up, this would also result in a selection bias. Since the risk of relapse in the non-tested patients was lower than in those tested, the association between a positive serological test and the risk of relapse for the total cohort may in fact be lower than what we found.

Despite the prospective inclusion and follow-up, the data on smoking were obtained retrospectively, and this may have resulted in recall biases. Our definition of relapse, although clinically relevant, did not include any clinical or endoscopical scoring systems, which have been used in other studies to measure disease activity.29,30 Although some minor relapses may not have been registered, we believe that we were able to record most of the clinically important relapses during follow-up.

The tests for ASCA and p-ANCA were performed at the end of the study. We were therefore not able to analyse the possibility of a relationship between these markers and mortality as an outcome. In addition the prognostic value of these tests in our study depended on the stability of the serological status from the point of inclusion throughout follow-up. A decline in the number of seropositive patients during the follow-up could have weakened the impact of these markers on disease prognosis. However, ASCA titres have previously been shown to be consistent over time, independently of medical or surgical treatment,3133 which indicates that they are a stable genetic marker of disease. P-ANCA status, on the other hand has been found to change from positive to negative after treatment with steroids or colectomy.19 This was not supported by our results, which showed neither a higher rate of colectomy nor a longer period of treatment with GCS in p-ANCA-negative than in p-ANCA-positive patients. Thus although fluctuations in p-ANCA titres over time have been reported, especially in patients with relapsing or chronic active UC, p-ANCA seems to remain stable in most patients during follow-up.20 This is supported by a study by Israeli et al., who showed that p-ANCA was positive in 25% of UC patients even before the clinical onset of the disease.34 These data indicate that p-ANCA remains serologically stable over time and are consistent with our finding that p-ANCA is a reliable indicator of long-term disease course. However, its serological stability means that the use of p-ANCA for monitoring disease activity and the effect of treatment during the course of disease could not be established by our data.

In conclusion, a positive p-ANCA status in a population-based cohort of UC patients followed for 10 years was a possible predictor of a relapsing disease and a larger number of relapses, information that could be useful when deciding on medical treatment. The presence of ASCA as an indicator of a more aggressive disease course in UC could not be excluded. However, the low prevalence of both markers in this population-based cohort limits their use in clinical practice, and better tests for both prediction and monitoring of disease activity in UC are still needed.35 Further studies should be performed of both established and new serological markers during long-term follow-up.

Conflict of interest

Guarantor of the article: Björn Moum. Specific author contribution: All of the authors of this paper have contributed in planning of the study and writing of the manuscript. With the exception of Geir Aamodt, all of the authors have also participated in the collection of the data. Financial support: The study was supported by the European Commission (QLG4-CT-2000-01414) and by a research fellowship grant from the Department of Medical Research, Sörlandet Hospital, Norway. Potential competing interests: None.

Acknowledgement

The authors wish to thank the members of the EC-IBD Study Group for their contributions: Vibeke Binder, Paolo Bodini, Mia Cilissen, Greet Claessens, Juan Clofent, Claudio Cortelezzi, Guiseppe d'Albasio, Michel Economou, Giovanni Fornaciari, Angie Harrington, Sofie Joosens, Kostas Katsanos, Ioannis Koutroubakis, Ebbe Langholz, Charles Limonard, Idar Lygren, Andrea Messori, Estella Monteiro, Maria Grazia Mortilla, Marius Nap, Paula Borrhalho Nunes, Victor Ruiz Ochoa, Colm O`Morain, Ioannis Mouzas, Athanasios Pallis, Angelo Pera, Santos Pereira, Asghar Quasim, Tullio Ranzi, Marielle Romberg, Maurice G. Russel, Jildou Sijbrandij, Camilla Solberg, Maria Tzardi, Ed van Hees, Robert van Hees, Gilbert van Zeijl, Ioannis Vlachonikolis, Frederik Wessels, Hagit Yona.

Abbreviations

    Abbreviations
  • p-ANCA

    Perinuclear anti-neutrohphil cytoplasmic antibody

  • ASCA

    Anti-Saccharomyces cerevecias antibody

  • AZA/6MP

    Azathioprine and 6-mercaptopurine

  • 5-ASA

    Sulfasalazine and 5-aminosalicylic acid

  • BI

    Binding index

  • CD

    Crohn's disease

  • CI

    Confidence interval

  • EC-IBD

    European Collaborative Study Group of Inflammatory Bowel Disease

  • GCS

    Glucocorticosteroids

  • IBD

    Inflammatory bowel disease

  • IQR

    Inter-quartile range

  • RR

    Relative risk

  • UC

    Ulcerative colitis

References

1
Nakamura
R.M.
Matsutani
M.
Barry
M.
Advances in clinical laboratory tests for inflammatory bowel disease
Clin Chim Acta
 
335
1–2
2003
9
20
2
Sandborn
W.J.
Loftus
E.V.
Jr.
Colombel
J.F.
et al
Evaluation of serologic disease markers in a population-based cohort of patients with ulcerative colitis and Crohn's disease
Inflamm Bowel Dis
 
7
3
2001
192
201
3
Stockbrugger
R.W.
Russel
M.G.
van Blankenstein
M.
et al
EC-IBD: a European effort in inflammatory bowel disease
Eur J Intern Med
 
11
4
2000
187
190
4
Lennard-Jones
J.E.
Classification of inflammatory bowel disease
Scand J Gastroenterol
 
170
1989
2
6
Suppl
5
Shivananda
S.
Lennard-Jones
J.
Logan
R.
et al
Incidence of inflammatory bowel disease across Europe: is there a difference between north and south? Results of the European Collaborative Study on Inflammatory Bowel Disease (EC-IBD)
Gut
 
39
5
1996
690
697
6
Wolters
F.L.
van Zeijl
G.
Sijbrandij
J.
et al
Internet-based data inclusion in a population-based European collaborative follow-up study of inflammatory bowel disease patients: description of methods used and analysis of factors influencing response rates
World J Gastroenterol
 
11
45
2005
7152
7158
7
Vermeire
S.
Joossens
S.
Peeters
M.
et al
Comparative study of ASCA (Anti-Saccharomyces cerevisiae antibody) assays in inflammatory bowel disease
Gastroenterology
 
120
4
2001
827
833
8
Peeters
M.
Joossens
S.
Vermeire
S.
et al
Diagnostic value of anti-Saccharomyces cerevisiae and antineutrophil cytoplasmic autoantibodies in inflammatory bowel disease
Am J Gastroenterol
 
96
3
2001
730
734
9
Roozendaal
C.
Kallenberg
C.G.
Are anti-neutrophil cytoplasmic antibodies (ANCA) clinically useful in inflammatory bowel disease (IBD)?
Clin Exp Immunol
 
116
2
1999
206
213
10
Gigase
P.
De Clerck
L.S.
Van Cotthem
K.A.
et al
Anti-neutrophil cytoplasmic antibodies in inflammatory bowel disease with special attention for IgA-class antibodies
Dig Dis Sci
 
42
10
1997
2171
2174
11
Sandborn
W.J.
Serologic markers in inflammatory bowel disease: state of the art
Rev Gastroenterol Disord
 
4
4
2004
167
174
12
Sandborn
W.J.
Landers
C.J.
Tremaine
W.J.
et al
Association of antineutrophil cytoplasmic antibodies with resistance to treatment of left-sided ulcerative colitis: results of a pilot study
Mayo Clin Proc
 
71
5
1996
431
436
13
Fleshner
P.R.
Vasiliauskas
E.A.
Kam
L.Y.
et al
High level perinuclear antineutrophil cytoplasmic antibody (pANCA) in ulcerative colitis patients before colectomy predicts the development of chronic pouchitis after ileal pouch-anal anastomosis
Gut
 
49
5
2001
671
677
14
Kuisma
J.
Jarvinen
H.
Kahri
A.
et al
Factors associated with disease activity of pouchitis after surgery for ulcerative colitis
Scand J Gastroenterol
 
39
6
2004
544
548
15
Reumaux
D.
Colombel
J.F.
Masy
E.
et al
Anti-neutrophil cytoplasmic auto-antibodies (ANCA) in ulcerative colitis (UC): no relationship with disease activity
Inflamm Bowel Dis
 
6
4
2000
270
274
16
Roozendaal
C.
Pogany
K.
Hummel
E.J.
et al
Titres of anti-neutrophil cytoplasmic antibodies in inflammatory bowel disease are not related to disease activity
QJM
 
92
11
1999
651
658
17
Zholudev
A.
Zurakowski
D.
Young
W.
et al
Serologic testing with ANCA, ASCA, and anti-OmpC in children and young adults with Crohn's disease and ulcerative colitis: diagnostic value and correlation with disease phenotype
Am J Gastroenterol
 
99
11
2004
2235
2241
18
Lindgren
S.
Floren
C.H.
Lindhagen
T.
et al
Low prevalence of anti-neutrophil cytoplasmic antibodies in ulcerative colitis patients with long-term remission
Eur J Gastroenterol Hepatol
 
7
6
1995
563
568
19
Rump
J.A.
Worner
I.
Roth
M.
et al
p-ANCA of undefined specificity in ulcerative colitis: correlation to disease activity and therapy
Adv Exp Med Biol
 
336
1993
507
513
20
Vecchi
M.
Bianchi
M.B.
Calabresi
C.
et al
Long-term observation of the perinuclear anti-neutrophil cytoplasmic antibody status in ulcerative colitis patients
Scand J Gastroenterol
 
33
2
1998
170
173
21
Hoie
O.
Wolters
F.L.
Riis
L.
et al
Low colectomy rates in ulcerative colitis in a European cohort followed for ten years
Gastroenterology
 
132
2
2007
507
515
22
Hoie
O.
Wolters
F.L.
Riis
L.
et al
Ulcerative Colitis: Patient characteristics may predict 10-year disease recurrence in a European-wide population-based cohort
Am J Gastroenterol
 
102
2007
1692
1701
23
Targan
S.R.
Landers
C.J.
Cobb
L.
et al
Perinuclear anti-neutrophil cytoplasmic antibodies are spontaneously produced by mucosal B cells of ulcerative colitis patients
J Immunol
 
155
6
1995
3262
3267
24
Langholz
E.
Munkholm
P.
Davidsen
M.
et al
Course of ulcerative colitis: analysis of changes in disease activity over years
Gastroenterology
 
107
1
1994
3
11
25
Vasiliauskas
E.A.
Plevy
S.E.
Landers
C.J.
et al
Perinuclear antineutrophil cytoplasmic antibodies in patients with Crohn's disease define a clinical subgroup
Gastroenterology
 
110
6
1996
1810
1819
26
Walker
L.J.
Aldhous
M.C.
Drummond
H.E.
et al
Anti-Saccharomyces cerevisiae antibodies (ASCA) in Crohn's disease are associated with disease severity but not NOD2/CARD15 mutations
Clin Exp Immunol
 
135
3
2004
490
496
27
Odes
S.
Friger
M.
Vardi
H.
et al
Role of ASCA and the NOD2/CARD15 mutation Gly908Arg in predicting increased surgical costs in Crohn's disease patients: a Project of the European Collaborative Study Group on Inflammatory Bowel Disease
Inflamm Bowel Dis
 
13
7
2007
874
881
28
Birrenbach
T.
Bocker
U.
Inflammatory bowel disease and smoking: a review of epidemiology, pathophysiology, and therapeutic implications
Inflamm Bowel Dis
 
10
6
2004
848
859
29
Bresci
G.
Parisi
G.
Bertoni
M.
et al
Long-term maintenance treatment in ulcerative colitis: a 10-year follow-up
Dig Liver Dis
 
34
6
2002
419
423
30
Edwards
F.C.
Truelove
S.C.
The course and prognosis of ulcerative colitis
Gut
 
41
1963
299
315
31
Oudkerk
P.M.
Ellerbroek
P.M.
Ridwan
B.U.
et al
Serum antineutrophil cytoplasmic autoantibodies in inflammatory bowel disease are mainly associated with ulcerative colitis. A correlation study between perinuclear antineutrophil cytoplasmic autoantibodies and clinical parameters, medical, and surgical treatment
Gut
 
34
1
1993
46
50
32
Teml
A.
Kratzer
V.
Schneider
B.
et al
Anti-Saccharomyces cerevisiae antibodies: a stable marker for Crohn's disease during steroid and 5-aminosalicylic acid treatment
Am J Gastroenterol
 
98
10
2003
2226
2231
33
Muller
S.
Styner
M.
Seibold-Schmid
B.
et al
Anti-Saccharomyces cerevisiae antibody titers are stable over time in Crohn's patients and are not inducible in murine models of colitis
World J Gastroenterol
 
11
44
2005
6988
6994
34
Israeli
E.
Grotto
I.
Gilburd
B.
et al
Anti-Saccharomyces cerevisiae and antineutrophil cytoplasmic antibodies as predictors of inflammatory bowel disease
Gut
 
54
9
2005
1232
1236
35
Nielsen
O.H.
Vainer
B.
Madsen
S.M.
et al
Established and emerging biological activity markers of inflammatory bowel disease
Am J Gastroenterol
 
95
2
2000
359
367

Figures

Figure 1

Flow chart demonstrating the status of the patients at the end of the 10-year follow-up study.

Figure 1

Flow chart demonstrating the status of the patients at the end of the 10-year follow-up study.

Figure 2

Ten year cumulative percentage (cum %) of p-ANCA positive (–) and negative (-----) patients having at least one relapse. The difference is statistically significant by the Log-rank test (p = 0.01).

Figure 2

Ten year cumulative percentage (cum %) of p-ANCA positive (–) and negative (-----) patients having at least one relapse. The difference is statistically significant by the Log-rank test (p = 0.01).

Tables

Table 1

Demographic details, serological results, potential risk factors and therapeutic data for the patient cohort

  p-ANCA  ASCA  No test Total cohort 
  Negative Positive Negative Positive   
Centre 
 Beer Sheva 15 17 21 39 
 Copenhagen 36 15 45 38 89 
 Cremona 16 19 22 41 
 Ioannina 25 27 10 37 
 Heraklion 41 53 
 Oslo 94 53 141 122 269 
 Reggio Emilia 31 37 12 50 
 South Limburg 64 12 72 64 140 
 Vigo 37 37 19 63 
 
Sex 
 Male 181 48 213 16 180 409 
 Female 146 57 191 12 169 372 
 
Age at diagnosis 
 Below 20 years 16 20 19 42 
 20–29 years 77 29 97 87 193 
 30–39 years 88 26 111 60 174 
 40–49 years 67 17 79 44 128 
 50–59 years 48 11 57 39 98 
 60–69 years 25 11 32 47 83 
 70–79 years 42 52 
 80 years and above 11 11 
 
Disease extent at diagnosis 
 Proctitis 101 29 120 10 95 225 
 Distal colitis 136 44 171 158 338 
 Extensive colitis 78 27 98 85 190 
 No data available 12 15 11 28 
 
Smoking status 
 Never smoked [ref] 136 52 180 45 233 
 Ex-smoker before inclusion 73 28 94 25 126 
 Ex-smoker after inclusion 41 44 53 
 Current smoker 63 63 26 97 
 Unknown 14 11 23 247 272 
 
First-degree relative(s) 
 None 277 84 334 27 90 451 
 Yes 37 11 48 17 65 
 No data available 13 10 22 242 265 
 
Education 
 Primary and lower secondary school 156 37 180 13 50 243 
 Higher secondary school and above 158 58 202 14 57 273 
 No data available 13 10 22 242 265 
 
Drugs 
5-ASA Drug used 273 95 342 26 271 639 
 Drug not used 45 53 58 112 
 No data available 20 30 
GCS Drug used 137 61 184 14 131 329 
 Drug not used 181 43 211 13 198 422 
 No data available 20 30 
AZA/6MP Drug used 28 11 31 16 55 
 Drug not used 288 92 361 19 313 693 
 No data available 11 12 20 33 
 
Colectomy 
 Yes 27 12 34 39 65 
 No 300 93 370 23 393 716 
Total  327 105 404 28 349 781 
  p-ANCA  ASCA  No test Total cohort 
  Negative Positive Negative Positive   
Centre 
 Beer Sheva 15 17 21 39 
 Copenhagen 36 15 45 38 89 
 Cremona 16 19 22 41 
 Ioannina 25 27 10 37 
 Heraklion 41 53 
 Oslo 94 53 141 122 269 
 Reggio Emilia 31 37 12 50 
 South Limburg 64 12 72 64 140 
 Vigo 37 37 19 63 
 
Sex 
 Male 181 48 213 16 180 409 
 Female 146 57 191 12 169 372 
 
Age at diagnosis 
 Below 20 years 16 20 19 42 
 20–29 years 77 29 97 87 193 
 30–39 years 88 26 111 60 174 
 40–49 years 67 17 79 44 128 
 50–59 years 48 11 57 39 98 
 60–69 years 25 11 32 47 83 
 70–79 years 42 52 
 80 years and above 11 11 
 
Disease extent at diagnosis 
 Proctitis 101 29 120 10 95 225 
 Distal colitis 136 44 171 158 338 
 Extensive colitis 78 27 98 85 190 
 No data available 12 15 11 28 
 
Smoking status 
 Never smoked [ref] 136 52 180 45 233 
 Ex-smoker before inclusion 73 28 94 25 126 
 Ex-smoker after inclusion 41 44 53 
 Current smoker 63 63 26 97 
 Unknown 14 11 23 247 272 
 
First-degree relative(s) 
 None 277 84 334 27 90 451 
 Yes 37 11 48 17 65 
 No data available 13 10 22 242 265 
 
Education 
 Primary and lower secondary school 156 37 180 13 50 243 
 Higher secondary school and above 158 58 202 14 57 273 
 No data available 13 10 22 242 265 
 
Drugs 
5-ASA Drug used 273 95 342 26 271 639 
 Drug not used 45 53 58 112 
 No data available 20 30 
GCS Drug used 137 61 184 14 131 329 
 Drug not used 181 43 211 13 198 422 
 No data available 20 30 
AZA/6MP Drug used 28 11 31 16 55 
 Drug not used 288 92 361 19 313 693 
 No data available 11 12 20 33 
 
Colectomy 
 Yes 27 12 34 39 65 
 No 300 93 370 23 393 716 
Total  327 105 404 28 349 781 

No test: patients not tested for p-ANCA or ASCA, 5-ASA: sulfasalazine and 5-aminosalicylates, GCS: glucocorticosteroids, AZA/6MP: azathioprine and 6-mercaptopurine, CI: 95% confidence interval.

Table 2

Disease outcome for p-ANCA and ASCA-positive and negative UC patients

 10-year cumulative colectomy rate 10-year cumulative relapse rate Total number of relapses per 10 years at risk 
 Cum % CI P value Cum % CI P value Median IQR P value 
p-ANCA 
Positive 11.4 5.3–17.5 0.32 81.7 74.3–89.1 0.01 2.1 1.0–5.8 < 0.001 
Negative 5.4 5.3–11.5  67.1 62.0–72.2  1.1 0–3.2  
 
ASCA 
Positive 17.9 3.8–32.0 0.09 77.8 62.1–93.5 0.23 2.6 0.3–5.8 0.09 
Negative 8.2 5.5–10.9  70.2 65.7–74.7  1.9 0–3.7  
 10-year cumulative colectomy rate 10-year cumulative relapse rate Total number of relapses per 10 years at risk 
 Cum % CI P value Cum % CI P value Median IQR P value 
p-ANCA 
Positive 11.4 5.3–17.5 0.32 81.7 74.3–89.1 0.01 2.1 1.0–5.8 < 0.001 
Negative 5.4 5.3–11.5  67.1 62.0–72.2  1.1 0–3.2  
 
ASCA 
Positive 17.9 3.8–32.0 0.09 77.8 62.1–93.5 0.23 2.6 0.3–5.8 0.09 
Negative 8.2 5.5–10.9  70.2 65.7–74.7  1.9 0–3.7  

Cum %: cumulative percentage, CI: 95% confidence interval, IQR: inter-quartile range. Log-rank test and Mann–Whitney U-test were used.

Table 3

Use of systemic drugs during follow-up

  5-ASA    GCS    AZA/6MP    
  User n Use period Non-user n No data n User n Use period Non-user n No data n User n Use period Non-user n No data n 
p-ANCA Negative 273 0.8 45 137 0.1 181 28 0.3 288 11 
 Positive 95 1.0 61 0.1 43 11 0.4 92 
ASCA Negative 342 0.8 53 184 0.1 211 31 0.5 361 12 
 Positive 26 0.8 14 0.2 13 0.1 19 
  5-ASA    GCS    AZA/6MP    
  User n Use period Non-user n No data n User n Use period Non-user n No data n User n Use period Non-user n No data n 
p-ANCA Negative 273 0.8 45 137 0.1 181 28 0.3 288 11 
 Positive 95 1.0 61 0.1 43 11 0.4 92 
ASCA Negative 342 0.8 53 184 0.1 211 31 0.5 361 12 
 Positive 26 0.8 14 0.2 13 0.1 19 

5-ASA, sulfasalazine/5-aminosalisylic acid; GCS, glucocorticosteroids; AZA/6MP, azathioprine/6-mercaptopurine; No data, no available information on drug use; Non-user, drug not used during follow-up; User, drug used during follow-up; Use period, proportion of time on medication during follow-up, median value calculated for drug users only.