Meta-analysis: serological markers and the risk of acute and chronic pouchitis

Authors


  • As part of AP&T's peer-review process, a technical check of this meta-analysis was performed by Mr M. Siddiqui.

Correspondence to:

Dr D. S. Pardi, Division of Gastroenterology and Hepatology, Department of Internal Medicine, Mayo Clinic, 200 First Street SW, Rochester, MN 55905, USA.

E-mail: pardi.darrell@mayo.edu

Summary

Background

Serological markers such as anti-neutrophil cytoplasmic antibody (ANCA) and anti-Saccharomyces cerevisiae antibody (ASCA) may be associated with pouchitis after ileal pouch-anal anastomosis (IPAA).

Aim

To perform a systematic review with meta-analysis of studies evaluating the association of ANCA and ASCA status with risk of acute and chronic pouchitis after IPAA.

Methods

We searched multiple databases (upto September 2012) for studies reporting ANCA and/or ASCA status along with risk of acute or chronic pouchitis after IPAA in adults with ulcerative colitis (UC). We abstracted odds ratio (OR) or raw data from the individual studies to calculate summary OR estimates with 95% CIs using random-effects model.

Results

Eight studies reporting 184 cases of acute pouchitis and six studies reporting 151 cases of chronic pouchitis were included. The odds of chronic pouchitis were 76% higher in ANCA-positive patients than ANCA-negative (six studies; OR: 1.76; 95% CI: 1.19–2.61; P < 0.01). ASCA-positivity was not associated with the risk of chronic pouchitis (three studies; OR: 0.89; 95% CI: 0.49–1.59; P = 0.68). Neither ANCA (eight studies; OR: 1.54; 95% CI: 0.79–3.02; P = 0.21) nor ASCA-positivity (two studies; OR: 1.28; 95% CI: 0.25–6.54; P = 0.77) were associated with the risk of acute pouchitis.

Conclusions

The risk of chronic pouchitis after IPAA is higher in ANCA-positive patients, but the risk of acute pouchitis is unaffected by ANCA status. ASCA status was not associated with the risk of acute or chronic pouchitis. This information may be used to counsel UC patients regarding their risk of pouchitis after IPAA.

Introduction

Restorative proctocolectomy with ileal pouch-anal anastomosis (IPAA) is the surgical procedure of choice for patients with medically refractory ulcerative colitis (UC) or UC-associated colorectal neoplasia. The lifetime risk of colectomy can be as high as 10–30% in patients with UC,[1, 2] although the rates may be declining.[3] Pouchitis is the most common complication in patients who undergo IPAA for UC, with reported cumulative frequencies of 23–59% after 10 years of surgery,[4-7] with some studies reporting even higher rates.[8] It is believed to be related to an abnormal mucosal immune response to bacterial flora in the pouch in a genetically predisposed individual.[9, 10] Acute pouchitis typically responds well to a course of oral antibiotics. About 10–38% of patients with acute pouchitis develop a chronic form of pouchitis which is antibiotic-dependent or antibiotic-refractory and may increase the risk of pouch failure, requiring pouch excision.[11-14]

Antibodies directed against a variety of microbial antigens, although not pathogenic, may occasionally be useful in distinguishing patients with Crohn's disease (CD) from UC, and may also be useful in predicting disease course, complications, response to medications, and need for surgery.[15] Perinuclear anti-neutrophil cytoplasmic antibody (ANCA) has been identified in 41–73% patients with UC and only 6–38% with CD.[15] On the other hand, anti-Saccharomyces cerevisiae antibody (ASCA) directed against a cell-wall component of the yeast, S. cerevisiae, is more specific for CD (seen in 29–69% patients) and uncommon in UC (0–29%). Some studies have shown that the presence of ANCA, especially in high titre, may increase the risk of pouchitis after IPAA,[16] but the results have been conflicting.[17, 18] ASCA-positivity has been associated with an increased risk of post-operative fistulae and CD-like condition of the pouch,[19, 20] but it is unclear whether it influences the risk of pouchitis. In addition, studies have not systematically evaluated the relationship between ANCA/ASCA and acute or chronic pouchitis.[21]

To better understand this relationship, we performed a systematic review with meta-analysis of all studies that investigated the association of ANCA and ASCA with the risk of developing acute and chronic pouchitis in adults following IPAA for UC.

Methods

We performed the meta-analysis according to the guidelines proposed by the Meta-analysis of Observational Studies in Epidemiology group.[22]

Search strategy

A systematic literature search of Medline (1966 through 30 September 2012), Embase (1988 through 30 September 2012) and Web of Science (1993 through 30 September 2012) bibliographic databases was conducted for all relevant articles on the association of ANCA- and ASCA-positivity with the risk of developing acute and/or chronic pouchitis after IPAA in adult patients with UC. Keywords used in the search included ‘anti-neutrophil cytoplasmic antibody’, ‘ANCA’, ‘anti-Saccharomyces antibody’, ‘ASCA’, ‘serology’, ‘risk factors’ AND ‘pouchitis’. The title and abstract of studies identified in the search were reviewed by two authors independently (S. S., gastroenterology fellow and P. K. S., final year medical student) to exclude studies that did not answer the research questions of interest. The full text of each remaining article was read to determine whether it contained relevant information. The reference list of articles with information on the topic was also reviewed for additional pertinent studies. To identify any additional studies published only in the abstract form, we manually searched the meeting abstracts from major gastroenterological conferences (Digestive Diseases Week, annual meeting of the American College of Gastroenterology and Advances in IBD, the annual meeting of the Crohn's and Colitis Foundation of America), from 2005 to 2012.

Selection criteria

The studies considered in this meta-analysis were observational studies (case–control or cohort) or randomised controlled trials that met the following inclusion criteria: (i) evaluated ANCA and/or ASCA status; (ii) reported pouchitis (acute and/or chronic) outcomes in adults patients (>18 years) who underwent IPAA for UC; and (iii) reported relative risks (RR) or odds ratio (OR) or provided data for their calculation. Inclusion was not otherwise restricted by study size, publication type or language. Articles were excluded from the meta-analysis if there was insufficient published data to estimate the RR/OR and 95% CI. When there were multiple publications from the same population, only data from the most recent comprehensive report were included. The flow diagram of the studies excluded from this analysis is shown in Figure 1.

Figure 1.

Flowsheet summarising study identification and selection.

Data abstraction

Data from each study were independently abstracted onto a standardised form by two investigators (S. S. and P. K. S.). The following data were collected from each study: study design, time period of study, year of publication, location of the population studied, duration of follow-up, primary outcome reported including definition of pouchitis and method of assessment, method of measurement and titre of ANCA/ASCA [indirect immuno-flourescence (IIF) and/or enzyme-linked immunosorbent assay (ELISA)], total number of persons in each group (antibody-positive vs. antibody-negative) and RR/OR and 95% CIs. Conflicts in data abstraction were resolved by consensus, referring back to the original article.

The quality of observational studies was assessed primarily based on representativeness of the population being studied, case-finding and case-ascertainment strategies and comparability of outcome assessment and follow-up. A formal quality score was not used since these have not been validated.[23]

Outcomes assessed

Our primary analysis focused on assessing the risk of acute and chronic pouchitis after IPAA depending on ANCA and ASCA status. Acute pouchitis was defined as single or infrequent episodes of antibiotic-responsive pouchitis; chronic pouchitis was defined as antibiotic-dependent or antibiotic-refractory pouchitis. As a secondary outcome, we also assessed the relationship between ANCA titres and risk of pouchitis.

Statistical analysis

We used the random-effects model described by DerSimonian and Laird to calculate the meta-analytic OR and 95% CI.[24] We assessed heterogeneity between study-specific estimates using two methods.[25, 26] First, the Cochran's Q-test for heterogeneity was measured. Since this test is underpowered to detect moderate degrees of heterogeneity,[27] a P-value of <0.10 was considered evidence of significant heterogeneity. Second, to estimate what proportion of total variation across studies was due to heterogeneity rather than chance, the inconsistency index (I2 statistic) was measured. Values of <30%, 30–60%, 61–75% and >75% were considered suggestive of low, moderate, substantial and considerable heterogeneity respectively.[26, 28] If heterogeneity was noted, we explored potential causes by stratifying for methodological and clinical features of studies. This included study design (case–control vs. cohort) and mean duration of follow-up in the study (≤4 years vs. >4 years). In this analysis, a P-value for differences between subgroups (Pinteraction) of <0.10 was considered statistically significant. Due to the small number of published studies, a formal statistical analysis of funnel plot asymmetry for assessment of publication bias was not performed.[29] All P-values were two tailed. For all tests (except for heterogeneity), a probability level <0.05 was considered statistically significant. All calculations and graphs were performed using comprehensive meta-analysis.[30]

Results

Of 250 unique studies identified using the search strategy, 11 studies met the inclusion criteria and were included in the meta-analysis (eight cohort, three case–control).[17-19, 31-38] There were no randomised controlled trials that addressed this question. Five studies with overlapping populations already included in other studies were excluded.[16, 39-42] For three studies, there were insufficient data for calculation of OR.[6, 43, 44] Of the 11 included studies, five reported only the relationship between ANCA/ASCA and acute pouchitis, three reported only the relationship with chronic pouchitis and three reported on the relationship with both acute and chronic pouchitis. Hence, we analysed eight studies reporting 184 cases of acute pouchitis in 669 adults,[17, 18, 31-34, 37, 38] and six studies reporting 151 cases of chronic pouchitis in 891 adults who underwent IPAA for UC.[17-20, 33, 35]

Characteristics of included studies

The characteristics of the included studies are shown in Table 1. The earliest study was published in 1994 and the most recent in 2012. All of these studies reported the relationship between ANCA and pouchitis; four studies measured ASCA (three studies on chronic pouchitis, two on acute pouchitis).[17, 19, 20, 32] The average follow-up after IPAA in studies reporting chronic pouchitis ranged from 4 to 9 years; the average follow-up in studies on acute pouchitis ranged from 1.5 to 6.5 years. The duration of follow-up was comparable in both cases and controls in most studies, except in one study in which cases of acute pouchitis were followed for ~3.5 years and controls (without pouchitis) were followed for only 0.7 years.[32] In 10 studies, pouchitis was diagnosed based on a combination of clinical and endoscopic and/or histological findings; in one study, diagnosis was based only on clinical features.[17] Antibodies were measured using IIF alone in five studies,[18, 31, 34, 37, 38] ELISA alone in three studies,[17, 19, 32] and both techniques in three studies.[33, 35, 36] In all studies, serological markers were measured post-operatively (at variable time periods) after IPAA, except one study in which half of the measurements were performed prior to surgery.[33]

Table 1. Characteristics of included studies assessing ANCA/ASCA and the risk of acute and chronic pouchitis
StudyDesignLocation, time periodMean follow-up, years (range)Total subjectsAcute pouchitis (AP) %Chronic pouchitis (CP) %ANCA+ANCA−ASCA+ASCA−
Pouchitis presentTotal ANCA +Pouchitis presentTotal ANCA −Pouchitis presentTotal ASCA +Pouchitis presentTotal ASCA −
  1. ANCA, perinuclear anti-neutrophil cytoplasmic antibody; ASCA, anti-Saccharomyces cerevisiae antibody; C-C, case–control; NR, not reported.

  2. a

     Mean, standard deviation.

  3. b

     ANCA status available for 341/399 patients, ASCA status available for 391/399 patients.

  4. c

     This study did not study association between ANCA status and risk of AP.

Aisenberg et al.[17]CohortNew York, NY (1987–1995)5 (5–12)10236 (35.3)14 (13.7)AP: 18; CP: 955AP: 18; CP: 547AP: 5; CP: 215AP: 13; CP: 1287
Brett et al.[31]C-CLondon, UK (NR)6.5 (0.3–17)5525 (45.5)AP: 1222AP: 1326  
Coukos et al.[19]CohortBoston, MA (2006–2010)7.2 (0.6–26.6)142 (5 IC)18 (12.7)CP: 1181CP: 761CP: 229CP: 16113
Dendrinos et al.[32]CohortBoston, MA (1980–2002)Pouchitis: 3.5 (0.3–10) No pouchitis: 0.7 (0.3–12)34 (4 IC)20 (58.8)AP: 1021AP: 1013AP: 714AP: 1320
Esteve et al.[18]C-CCatalonia, Spain (NR)4 (2–9)3111 (35.5)4 (12.9)AP: 4; CP: 211AP: 7; CP: 220  
Fleshner et al.[33]CohortLos Angeles, CA (1994–2006)4 (0.3–12)238 (70 IC)43 (18.1)29 (12.2)AP: 37; CP: 24171AP: 6; CP: 567  
Reumaux et al.[34]CohortFrance (NR)1.69812 (12.2)AP: 853AP: 445  
Sandborn et al.[35]C-CRochester, MN (NR)5 (3)a3719 (51.4)CP: 1928CP: 09  
Tyler et al.[20]CohortToronto, Canada (2007)9.2 (6.8)a399b112 (28.1)c67 (16.8)CP: 37152CP: 30189CP: 1270CP: 55321
Vecchi et al.[37]CohortBologna, Italy (NR)3.4 (0.3–9)429 (21S4)AP: 814AP: 128  
Yang et al.[38]CohortGoteberg, Sweden (1983–1993)4 (0.5–11)7628 (36.8)AP: 2149AP: 727  

Of the total 1254 patients included in the studies, 657 were ANCA-positive (54.9%, range: 33.3–75.7% in individual studies). In four studies that measured ASCA, 128 of 669 patients (19.1%; range: 14.7–41.2% in individual studies) were positive. In pooled results from cohort studies, 22.4% of patients (260 of 1131 patients, seven studies) developed acute pouchitis, and 14.5% of patients (128 of 881 patients, four studies) developed chronic pouchitis.

Risk of acute pouchitis

Of the eight studies which reported the association between ANCA status and the risk of acute pouchitis, two observed a higher risk of acute pouchitis in ANCA-positive patients,[33, 37] and six did not observe any association.[17, 18, 31, 32, 34, 38] Meta-analysis of these studies did not show any relation between ANCA-positivity and risk of acute pouchitis (Figure 2). There was moderate heterogeneity across studies (Cochran's Q-test P = 0.02, I2 = 60%). This heterogeneity could not be explained based on study design (case–control vs. cohort) (OR: 1.15; 95% CI: 0.46–2.86 vs. OR: 1.74; 95% CI: 0.72–4.25) (Pinteraction = 0.52) or duration of follow-up (≤4 years vs. >4 years) (OR: 2.17; 95% CI: 0.76–6.22 vs. OR: 0.93; 95% CI: 0.51–1.71) (Pinteraction = 0.17).

Figure 2.

Summary of odds ratios (ORs) of studies assessing the risk of acute pouchitis depending on anti-neutrophil cytoplasmic antibody (ANCA) status.

Meta-analysis of two studies which evaluated the association between ASCA status and risk of acute pouchitis did not observe any association (OR: 1.28; 95% CI: 0.25–6.54).

Risk of chronic pouchitis

Of the six studies which reported the association between ANCA status and risk of chronic pouchitis, two observed a higher risk of chronic pouchitis in ANCA-positive patients,[20, 35] and four did not observe any association.[17-19, 33] Meta-analysis of these studies revealed that the odds of chronic pouchitis were 76% higher in ANCA-positive patients compared with ANCA-negative patients (OR: 1.76; 95% CI: 1.19–2.61) (Figure 3). The results were homogenous (Cochran's Q-test = 0.43, I2 = 0%), and stable across study design. To assess whether any one study had a dominant effect on the meta-analytic OR, each study was excluded and its effect on the main summary estimate was evaluated. The pooled OR remained significantly increased (range of pooled OR: 1.66–1.91), with the corresponding 95% CI remaining above 1.

Figure 3.

Summary of ORs of studies assessing the risk of chronic pouchitis depending on ANCA status.

On meta-analysis of three studies, ASCA-positive patients were not more likely to develop chronic pouchitis as compared with ASCA-negative patients (OR: 0.89; 95% CI: 0.49–1.59).

Subgroup analysis

ANCA titres were categorised as ‘high’ and ‘low’ and its relationship with the risk of pouchitis was reported in five studies.[17, 18, 33, 34, 38] In studies which used IIF for ANCA measurement, ‘high’ titres were defined as >1:40 (one study)[18] or >1:80 (two studies)[34, 38]; ‘low’ titres were defined as less than these values. In two studies which measured ANCA using ELISA, ‘high’ titres were defined as >100 EU/mL[33] or >40 EU/mL[17]; ‘low’ titres were defined as less than these values.

The association between ANCA-positivity and chronic pouchitis was seen with both ‘high’ titres (three studies; OR: 3.60; 95% CI: 1.47–8.82) and ‘low’ titres (three studies; OR: 2.54; 95% CI: 1.02–6.37). On the other hand, there was no association between ANCA titres and risk of acute pouchitis (‘high’ titres vs. ‘low’ titres, OR: 1.16; 95% CI: 0.40–3.38 vs. OR: 1.33; 95% CI: 0.80–2.23).

Discussion

While acute pouchitis is typically easily treatable and does not significantly impact functional pouch outcomes, chronic pouchitis is associated with significant morbidity, need for chronic antibiotic therapy and occasionally budesonide or immunomodulator therapy, and may lead to pouch failure.[45] In this meta-analysis, we identified that ANCA-positivity was associated with 76% higher odds of chronic pouchitis after IPAA for UC (as compared with ANCA-negativity), but did not modify the risk of acute pouchitis. ASCA status did not influence the risk of acute or chronic pouchitis. These results suggest that ANCA status may be used to counsel patients undergoing IPAA for UC regarding their risk of chronic pouchitis, although the presence or absence of ANCA should not alter the choice of surgical procedure.

Pouchitis likely represents a dysregulated mucosal immune response to intestinal microbes.[9, 10] ANCA is an autoantibody that cross-reacts with a number of commensal bacterial antigens, with the target antigen being located within the neutrophil nucleus. The presence of ANCA, especially in high-titres, may represent a vigorous immune response to these commensal bacteria, thereby predisposing to pouchitis.[16, 21] It is unclear why ANCA is associated with chronic, but not acute, pouchitis. There is increasing evidence that pouchitis represents a heterogenous inflammatory disorder. While there are similarities in the clinical presentation of acute and chronic pouchitis, several factors point to differences in pathogenesis.[10, 33, 46] For example, only a small fraction of patients with acute pouchitis develop chronic pouchitis.[11-13] Furthermore, the risk factors for acute and chronic pouchitis are different. While smoking may promote acute pouchitis,[33] it may be protective against chronic pouchitis.[47] The presence of co-existing autoimmune disorders increases the risk of chronic pouchitis, but not acute pouchitis.[48] It is likely that patients with acute pouchitis develop a transient reaction to endo-luminal antigens, while patients with chronic pouchitis develop persistent inflammation to these same bacterial products, perhaps reflecting a more profound host-immune response.[33] This may explain the presence of an association between ANCA and chronic pouchitis, but not with acute pouchitis. Besides ANCA and ASCA, some recent studies have suggested that newer anti-microbial antibodies (e.g. anti-CBir1 and anti-OmpC) may also be associated with adverse outcomes after IPAA, including pouchitis, post-operative fistulae and CD of the pouch).[19, 32, 41]

Based on available data, a causal association between ANCA and ASCA status and the risk of pouchitis is difficult to draw. While some studies have shown that ANCA status is not associated with disease activity in UC,[15] others have shown that ANCA-positivity may be associated with an increased risk of treatment-resistant severe left-sided colitis,[49] more aggressive disease requiring intensive immunosuppression with azathioprine,[50] nonresponse to infliximab therapy[51] and increased risk of colectomy.[49, 52, 53] All of these factors may also be independently associated with increased risk of pouchitis after IPAA.[10] The included studies in this meta-analysis did not adjust for these important confounding variables, limiting any inference on the independent role of ANCA status in predicting the risk of pouchitis. In addition, in most studies, ANCA was measured after colectomy; since antibody titres may decline after surgery,[54] these titres may not be reflective of the preoperative level. This limits any inference drawn on the utility of the preoperative ANCA status on the risk of post-operative pouchitis.

The strengths of our analysis include well-defined inclusion and exclusion criteria and inclusion of all available studies fulfilling criteria and not restricting analysis based on study design, publication type or language. Previous studies have been conflicting as to whether ANCA status modifies the risk of subsequent pouchitis. Individually, these studies have been limited by small sample size, limited duration of follow-up and variable source population. With this pooled analysis, we have been able to more clearly define the association between these anti-microbial antibodies and pouchitis.

Our study has limitations that merit further discussion. In some of the retrospective studies, it is unclear whether measurement of serological markers was part of routine clinical practice in the management of patients with UC, or was prompted by certain clinical characteristics, which could have selected for patients deemed to be at higher risk for complicated disease. This is evident in the relatively small number of patients who participated in the studies (and whose antibody status was known), in comparison to the significantly greater number of patients who had undergone IPAA during the same time period in the same institution. For example, out of 500 patients who had undergone IPAA for UC in one study, only 102 participated in the study.[17] Likewise, of the 1600 patients who had undergone IPAA for UC at Mount Sinai Hospital in Toronto, Canada, during the study period, only 399 agreed to participate, and ANCA status was known for only 341 of those.[20] In addition, the duration of follow-up was variable in different studies, which is a significant limitation when performing meta-analysis of studies of a time-dependent outcome such as pouchitis. Over the three decades that these studies were performed, the assessment and classification of pouchitis, as well methods of assessing and reporting titres of ANCA and ASCA have undergone changes. Titres of ANCA were not routinely reported in the studies, and where reported, were not uniform in definition of ‘high’ and ‘low’ titres. Thus, subgroup analysis was limited. Besides disease severity and extent (which may impact ANCA/ASCA titres), the included studies did not adjust for other variables that may be associated with both ANCA status and risk of pouchitis, such as family history of inflammatory bowel disease or the presence of primary sclerosing cholangitis.[10] Future studies aimed at assessing the clinical utility of preoperative serological markers in predicting post-operative outcomes after IPAA should adjust for these important covariates.

In conclusion, ANCA-positivity may be associated with an increased risk of chronic pouchitis in adult patients undergoing IPAA for UC, but does not seem to modify the risk of acute pouchitis. This finding may be used to counsel patients prior to surgery, but should not influence the choice of surgery. In our adult practice, we do not routinely assess ANCA and ASCA serology. Future prospective evaluation of ANCA and ASCA serology prior to surgery may help better understand its utility in patients undergoing IPAA for UC. Before recommending routine use of serological markers for predicting acute and chronic pouchitis after IPAA, diagnostic studies reporting sensitivity and specificity of serological markers in predicting pouchitis are warranted. If found to be useful as independent predictors of chronic pouchitis, these markers could potentially be added to clinical nomograms aimed at predicting the risk of adverse pouch outcomes after IPAA.[55]

Authorship

Guarantor of the article: Darrell S. Pardi.

Author contributions: S. S. and D. S. P. were involved in study concept and design; S. S. and P. K. S. were involved in acquisition of data; S. S. was involved in statistical analysis; S. S. and P. K. S. were involved in drafting of the manuscript; D. S. P. and E. V. L. were involved in critical revision of the manuscript for important intellectual content. All authors have approved the final version of the article, including the authorship list.

Acknowledgement

Declaration of personal interests: Dr Pardi has served as a consultant for Optimer and Salix within the past 2 years and has received research support from Viropharma Incorporated, Jenssen Biotech, Inc. and Merck Research Laboratories. Dr Loftus has served as a consultant for Abbott, UCB, Janssen, Elan, Hospira, Eisai and Given within the past 2 years, and has received research support from Amgen, Abbott, Bristol-Myers Squibb, GlaxoSmithKline, Janssen, UCB, Genentech, Millenium-Takeda, Santarus, Braintree and Pfizer.

Declaration of funding interests: None.

Ancillary