SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Appendix

Aliment Pharmacol Ther31, 1365–1370

Summary

Background  Distinguishing between inflammatory bowel disease (IBD) and irritable bowel syndrome (IBS) can be challenging.

Aims  To investigate the utility of faecal lactoferrin as a marker of inflammation in patients with IBD, IBS and controls.

Methods  Disease activity in IBD patients was assessed using the modified Harvey–Bradshaw Activity Index. Stool samples were analysed using an ELISA assay.

Results  We recruited 137 patients with IBS, 126 with ulcerative colitis (UC) and 104 with Crohn’s disease (CD), and 98 healthy volunteers. The median ± IQ lactoferrin concentration (μg/g faecal weight) was 0 ± 1.4 for IBS patients, 6.6 ± 42 for UC patients, 4 ± 12.7 for CD patients and 0.5 ± 2 for healthy controls. Lactoferrin levels were significantly higher in IBD patients compared with IBS/healthy controls (< 0.001). The median lactoferrin concentrations were significantly higher in active UC & CD patients compared with inactive patients (< 0.001 and = 0.002 respectively). The sensitivity, specificity, positive and negative predictive values of lactoferrin in distinguishing active IBD from IBS/healthy controls were 67% and 96%, 87% and 86.8% respectively.

Conclusions  Lactoferrin is useful to differentiate between IBD and IBS, and can be used as an adjunct to blood parameters to determine IBD patients who have ongoing inflammation.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Appendix

Inflammatory bowel disease (IBD) and irritable bowel syndrome (IBS) are common entities in the Western population.1–3 Both conditions may present with similar clinical features such as diarrhoea and abdominal pain. Patients with IBD oscillate between periods of active and inactive disease and may even present with concomitant functional IBS. The discrimination of IBS from active IBD can be resourcefully challenging for clinicians and may delay effective treatment. Some investigations may also be perceived as uncomfortable or invasive for the patient. Clinical criteria such as ROME II IBS have been devised to aid the diagnosis of IBS.4, 5 The determination of inflammatory activity is crucial for patients with IBD for the diagnosis, monitoring and step up of therapy. Clinical indices are widely used, but are hampered by the subjective nature of symptom reporting and have been shown to be poorly correlated with mucosal activity.6 Colonoscopy is the accepted gold standard for investigation of the colon, but is invasive and associated with risks.7 Whilst there is emerging evidence of activation of the mucosal innate defence system toward a pro-inflammatory response in IBS patients, the absence of endoscopic and histological inflammation remains an accepted approach to the diagnosis of IBS by the bedside.8

Lactoferrin (LF) is an iron binding glycoprotein secreted by most mucosal membranes and a major component of secondary granules of polymorphonuclear neutrophils, a component of the inflammatory response.9, 10 Elevated LF has been used as a marker of active IBD11–16 and for monitoring patients for response to treatment.17 Some studies report a high sensitivity of LF for active IBD in comparison with IBS. However, the use of LF for the distinction of inactive IBD and IBS is less clear.13, 14, 18, 19 In Table 1, the comparative studies of patients with IBD and IBS using LF are tabulated.11, 13–16, 19

Table 1.   Studies investigating the utility of faecal lactoferrin in comparison of inflammatory bowel disease and Irritable bowel syndrome patients
Author & YearCountryStudy comparative groupsTotal number of patients (Irritable bowel syndrome/IBS)Results
Walker 2007USAFaecal lactoferrin, blood parameters and clinical activity index148 (IBS = 7)Sensitivity 84% Specificity 97%
Langhorst 2008GermanyFaecal markers, activity index, blood parameters, endoscopy140 (IBS = 54)Sensitivity 85% Specificity 77%
Schoepfer 2008SwitzerlandFaecal markers, activity index, blood parameters & endoscopy136 (IBS = 30)Sensitivity 87% Specificity 96%
Schroder 2007GermanyFaecal markers & endoscopy88 (IBS = 31)Sensitivity 82% Specificity 100%
Kane 2003USAFaecal markers& activity index271 (IBS = 31)Sensitivity 90% Specificity 100%
Dai 2007ChinaFaecal markers, activity index & endoscopy177 (IBS = 25)Sensitivity 92% Specificity 88%

The aim of this study was to investigate the clinical utility of LF as a marker of GI inflammation in patients with active and inactive IBD compared with patients with diarrhoea predominant IBS and healthy controls.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Appendix

Patients

Consecutive patients were recruited from the out-patient clinic. Patients with inflammatory bowel disease were questioned about their general well being, the frequency of bowel habit, the presence/absence of abdominal pain or blood in the stool. Patients with established IBD were given a Harvey–Bradshaw Index (HBI) for Crohn’s disease and a (previously validated) modified HBI for Ulcerative colitis (Appendix 1).15 Patients with HBI of ≥4 were considered to have active disease. All patients who had diarrhoea with the presence of abdominal discomfort and who fulfilled the Rome II criteria for diarrhoea predominant IBS were also recruited.4, 5 All patients were investigated and treated according to the British Society of Gastroenterology guidelines.4 Colonoscopy was requested based on clinical need. Healthy controls were recruited after exclusion of disease with a questionnaire. All participants were requested to return a stool sample in a container provided. Ethical approval was obtained from the North Sheffield Ethics Committee.

Stool analysis

Analysis was performed blind to the clinical details of the patient. Stool samples were frozen at −20 °C immediately on receipt. Quantitative ELISA (IBD SCAN) faecal lactoferrin test was performed on each thawed sample. The stool analysis kits were provided by ScheBo Biotech UK Limited and Techlab, USA (Blacksburg, VA, USA). A cut off level of >7.25 μg/g was deemed positive based on the manufacturer’s guide.

Statistical analysis

The data were analysed using spss version 15 (SPSS Inc, Chicago, IL, USA). Nonparametric tests (Mann–Whitney U-test) were used to compare lactoferrin concentrations between groups as the data were not normally distributed. Kendal tau correlations were performed to assess the relationship between lactoferrin concentration and disease activity (HBI). Assistance was also sought from the University of Sheffield statistics department.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Appendix

Four hundred and sixty-five patients were recruited between November 2006 and October 2008. The mean age in the IBS, UC and Crohn’s group was 42 years, 58 years and 56 years respectively. The median LF levels were significantly higher in patients with IBD compared with patients with IBS (< 0.001) and healthy controls (< 0.001). In Table 2, the mean and median LF values for each group are tabulated, whilst in Figure 1, the distribution of LF values in all patients is shown.

Table 2.   Mean and median faecal lactoferrin concentration (μg/g) in all groups
 Ulcerative colitisCrohn’s diseaseIrritable bowel syndromeHealthy controls
No. of patients12610413798
Mean LF (μg/g) (±s.d.)69.5 (168)41.4 (139)1.39 (3.4)2.4 (7.2)
Median LF (μg/g) (±IQ)6.6 (42)4 (13)0 (1.4)0.5 (2)
image

Figure 1.  Faecal lactoferrin concentrations in all patients.

Download figure to PowerPoint

Among patients with IBD, there was a trend towards higher LF values in patients with UC compared with patients with CD (= 0.051). As for stratification based on severity of symptoms/disease activity, the median LF (μg/g) levels were significantly higher in patients with active disease (HBI ≥4) compared with patients with inactive disease for both UC and CD (< 0.001 and = 0.002 respectively). Analysis of LF levels in IBD patients based on disease activity is tabulated in Table 3.

Table 3.   Comparison of faecal lactoferrin values in patients with inflammatory bowel disease based on disease activity
 Ulcerative colitis (= 126) Median LF (IQ)Crohn’s disease (= 104) Median LF (IQ)
Active disease= 51 Median 26 (102)= 51 Median 8.4 (32)
Inactive disease= 75 Median 3 (8.5)= 53 Median 1 (6)
P values<0.001<0.002

Comparisons were also made between patients with inactive IBD (HBI <4) and those with diarrhoea predominant IBS. Patients with inactive IBD had significantly higher LF levels compared with patients with IBS. The median LF (μg/g) ± IQ for patients with inactive UC and that for patients with inactive CD was 3.1 μg/g (8.5) and 1 μg/g (5.8) respectively compared to 0 μg/g (1.4) for patients with IBS (< 0.001 and = 0.002 respectively).

The correlation between LF values and the disease activity (Harvey–Bradshaw Index) was fair. The correlation coefficient for patients with UC was 0.4, whereas it was 0.2 for patients with CD and 0.3 for any diagnosis of IBD.

The sensitivity and specificity of LF for active IBD vs. IBS patients were 67% and 96% respectively with positive and negative predictive values of 92% and 80% respectively. Similar calculations for active UC and CD patients are tabulated in Table 4. ROC curves were calculated to illustrate the trade off between the sensitivity and specificity for each group as shown in Figure 2a–c.

Table 4.   Comparison of patients with active inflammatory bowel disease and patients with Irritable bowel syndrome
 SensitivitySpecificityPositive predictive valueNegative predictive value
  1. UC, ulcerative colitis; CD, Crohn’s disease; IBS, irritable bowel syndrome.

Comparison of active UC against IBS78%96%86%92%
Comparison of active CD against IBS58%96%83%86%
image

Figure 2.  (a) ROC curve for all patients, area under the curve 0.75. (b) ROC curve for patients with active inflammatory bowel disease compared to patients with Irritable bowel syndrome (area under the curve 0.84). (c) ROC curve for patients with active vs. inactive inflammatory bowel disease (area under the curve 0.81).

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Appendix

This study, the largest to date on the use of LF (= 465), has shown that LF has a high sensitivity and specificity for the discrimination of patients with active IBD against patients with IBS and healthy controls. In addition, LF levels were significantly higher in patients with inactive IBD than in patients with IBS, making it a valuable investigative tool in patients where the differentiation is difficult, based on clinical history alone. Whilst the poor correlation between symptom reporting and disease activity in IBD has been demonstrated before,6 previous LF studies making similar comparisons have shown conflicting results with some studies failing to show a difference between inactive IBD and IBS.13, 18, 19

We have also demonstrated a significant difference in LF levels in patients with active IBD compared with patients with inactive IBD. This suggests that LF could be used in conjunction with other parameters (clinical and blood inflammatory markers) to determine the subset of patients who have active disease or who may require a step up of therapy. A paediatric study (= 5) showed that LF levels is potentially useful as a biomarker for response to anti-tumour necrosis factor (anti-TNF) therapy.20 It has also been suggested that the course of LF may be an early predictor of a relapse.11, 20 In a recent study, LF predicted post-operative recurrence in Crohn’s disease with greater accuracy than C-reactive protein, platelet count or endoscopic appearance.17

In our study, the median LF levels in IBD patients were comparatively lower than in other studies in the published literature.14, 15 This could be explained by the larger number of patients with inactive disease. The inclusion of patients into this study was based on recruitment from routine out-patient clinics from a single centre as opposed to specially selected patients with severe symptoms.

A perceived limitation of our study is the lack of correlation with endoscopic and histological grading. In our study, colonoscopy was only performed based on clinical need and represented less than 30% of the population group. In addition, the endoscopies which were performed as routine care were performed by a number of endoscopists. Similarly biopsies from these patients were also analysed by a number of histopathologists, which made meaningful comparisons difficult. Previous investigators have demonstrated that LF has a good correlation with endoscopic grading.14

LF is an inexpensive and non-invasive test that can provide the clinician with a marker to differentiate between IBD (particularly active disease) and IBS and stratify patients who require endoscopic investigations. In addition, LF can also be used as an adjunct to blood parameters and clinical symptoms to determine IBD patients who have ongoing inflammation.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Appendix

Declaration of personal interests: DSS designed the study. All authors contributed to the data collection. RS performed the initial analysis and wrote the initial draft, and all ten authors were involved in the subsequent revisions, critical analysis and final draft. David Sanders is the guarantor for this manuscript and he has served as a speaker for ScheBo Biotech UK Limited at the FOCUS meeting 2009. Declaration of funding interests: Funding assistance with provision of stool analysis kits from ScheBo Biotech UK Limited and TechLab Inc U.S.A. is acknowledged.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Appendix

Appendix

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Appendix

Appendix 1

Harvey–Bradshaw Index for Crohn’s disease
(1) Number of liquid stools per day
(2) Abdominal pain, sum of seven daily ratings: (0-none, 1-mild, 2-moderate, 3-severe)
(3) General well being (0-very well, 1-slightly below par, 2-poor, 3-very poor, 4-terrible)
(4) Complications (score 1 point per item) Arthritis/arthalgia Skin/mouth lesions Iritis/uveitis Anal fissure, fistula/perianal abscess
(5) Abdominal mass (0-none, 1-questionable, 2-definite, 3-definite & tender)
Modified Harvey–Bradshaw Index for Ulcerative Colitis
(1) Number of liquid stools per day
(2) Abdominal pain, sum of seven daily ratings: (0-none, 1-mild, 2-moderate, 3-severe)
(3) General well being (0-very well, 1-slightly below par, 2-poor, 3-very poor, 4-terrible)
(4) Complications (score 1 point per item) Arthritis/arthalgia Skin/mouth lesions Iritis/uveitis Anal fissure,fistula/perianal abscess
(5) Bleeding per rectum (0-none, 1-slight, 2-moderate, 3-severe)