- Email: [email protected]
Increased Detection of Celiac Disease With Measurement of Deamidated Gliadin Peptide Antibody Before Endoscopy
Accepted Manuscript Increased Detection of Celiac Disease With Measurement of Deamidated Gliadin Peptide Antibody Before Endoscopy Dr Peter D. Mooney, MBChB, Simon H. Wong, Alexander J. Johnston, Dr Matthew Kurien, MD, Dr Anastasios Avgerinos, MD, Prof David S. Sanders, MD
PII: DOI: Reference:
S1542-3565(15)00096-8 10.1016/j.cgh.2015.01.010 YJCGH 54140
To appear in: Clinical Gastroenterology and Hepatology Accepted Date: 9 January 2015 Please cite this article as: Mooney PD, Wong SH, Johnston AJ, Kurien M, Avgerinos A, Sanders DS, Increased Detection of Celiac Disease With Measurement of Deamidated Gliadin Peptide Antibody Before Endoscopy, Clinical Gastroenterology and Hepatology (2015), doi: 10.1016/j.cgh.2015.01.010. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. All studies published in Clinical Gastroenterology and Hepatology are embargoed until 3PM ET of the day they are published as corrected proofs on-line. Studies cannot be publicized as accepted manuscripts or uncorrected proofs.
ACCEPTED MANUSCRIPT Increased Detection of Celiac Disease With Measurement of Deamidated Gliadin Peptide Antibody Before Endoscopy
1,2
Dr Peter D Mooney* (MBChB), 2Simon H Wong, 2Alexander J Johnston,
1,2
Dr Matthew
RI PT
Kurien (MD), 1Dr Anastasios Avgerinos (MD), 1,2Prof David S Sanders (MD) Affiliations: 1Royal Hallamshire Hospital, Sheffield, UK, 2Univeristy of Sheffield, UK
SC
*Corresponding author: P floor Royal Hallamshire Hospital, Glossop Rd, Sheffield, S10 2JF. Tel: 0114 2261179 Fax: 0114 2712692. Email [email protected]
Acknowledgements: This investigator led study was partially funded by research grants
M AN U
from Tillotts Pharma, BHR Pharmaceuticals, and Coeliac UK. The study was adopted to the National Institute for Health Research portfolio, study ID 14040. None of the funding sources had any input into study design, access to study data or interpretation. Conflict of interest statement: DSS has received funding from Tillotts Pharma, BHR pharmaceuticals and Coeliac UK for investigator led studies into point of care testing for
TE D
celiac disease. PDM, SHW, AJJ, MK and AA declare no conflicts of interest.
Authorship statement: PDM recruited patients, obtained duodenal biopsies, analyzed the data and drafted the manuscript. SHW and AJJ performed point of care tests and collected
EP
patient data, MK and AA recruited patients and obtained duodenal biopsies, DSS conceived the study and edited the final manuscript and acts as guarantor for the article. All authors
AC C
had access to the study data and approved the final manuscript.
Abbreviations
DGP - De-amidated gliadin peptide antibody EMA - Endomysial antibody NPV - Negative predictive value PPV - Positive predictive value tTG - Tissue transglutaminase antibody
1
ACCEPTED MANUSCRIPT Abstract Background & Aims: Celiac disease is under-diagnosed. Many patients are examined by endoscopy, but celiac disease is missed or not detected. We evaluated the accuracy of finger prick-based point of care tests in detection of celiac disease and developed an
RI PT
algorithm for diagnosis.
SC
Methods: We performed a prospective study of 2 groups of patients with celiac disease evaluated at the Royal Hallamshire Hospital in Sheffield UK from March 2013 through
M AN U
February 2014. In group 1, patients at high risk of celiac disease who tested positive for endomysial antibody (n=55) were evaluated using the Biocard test and Celiac Quick Test, which measure antibodies to tissue transglutaminase (anti-tTG), and the Simtomax test, which measures deamidated gliadin peptide antibodies (DGP). Patients in group 2 (508
TE D
consecutive patients who underwent endoscopy examinations for any indication) received the DGP test, and were also evaluated using a diagnostic algorithm that incorporated results from the DGP test and data on symptoms. In both groups point of care tests were taken at
EP
the time of endoscopy and results were compared with results from histologic analyses of
AC C
duodenal biopsies from all patients.
Results: In Group 1, the DGP test identified patients with celiac disease with 94.4% sensitivity, the Celiac Quick Test with 77.8% sensitivity (P=.03 vs the DGP test), and the Biocard test with 72.2% sensitivity (P=.008 vs the DGP test). In Group 2, the DGP test identified patients with celiac disease with 92.7% sensitivity (95% confidence interval [CI], 83.0–97.3), 85.2% specificity (95% CI, 81.5–88.3), a positive predictive value of 49.2% (95% CI, 40.3–58.2), and a negative predictive value of 98.7% (95% CI, 96.8–99.5). Measurement 2
ACCEPTED MANUSCRIPT of serum anti-tTG identified patients with celiac disease with 91.2% sensitivity (95% CI, 81.1–96.4), 87.5% specificity (95% CI, 84.0–90.4), a positive predictive value of 53.0% (43.6– 62.2), and a negative predictive value of 98.5% (95% CI, 96.5–99.4). The algorithm identified patients with celiac disease with 98.5% sensitivity; its use could reduce duodenal biopsies by
RI PT
35%.
SC
Conclusion: In a prospective study, a test for DGP identified patients with celiac disease with similar levels of sensitivity and specificity as standard serologic analysis of anti-tTG. Use of
M AN U
the DGP test before endoscopy could increase the accuracy of diagnosis of celiac disease. Further studies, in lower prevalence populations, are required to assess the impact of the test in clinical practice.
AC C
EP
TE D
KEY WORDS: Finger prick test; case finding; missed diagnosis; diagnostic assay
3
ACCEPTED MANUSCRIPT Introduction Adult celiac disease is a common chronic inflammatory bowel condition caused by a heightened immunological response to dietary gluten. Internationally the prevalence of celiac disease is now recognised to be 0.2-1.2%.1-4 However, despite increased awareness of celiac disease worldwide meta-analyses have demonstrated that for every patient diagnosed 3 - 4
ranging from 4 to 13 years.6,
7
RI PT
patients go undetected.5 International cohorts have reported significant delays in diagnosis Furthermore, when specifically considering endoscopy, the
failure to recognise celiac disease (and perform a duodenal biopsy) ranges from 5-13%.
SC
Therefore, current national and international guidelines recommend routine duodenal biopsy in patients with high risk symptoms of weight loss, chronic diarrhea and anemia.8, 9 A recent large observational study however has demonstrated that this practice is highly variable with
M AN U
duodenal biopsy rates for these conditions ranging from 19 – 59%.10 One explanation for this under-use of duodenal biopsy may be reliance by clinicians on the presence of endoscopic markers of celiac disease to guide biopsy. However, endoscopic markers of celiac disease lack the required sensitivity.11 Furthermore many patients with celiac disease present with nonspecific symptoms such as IBS or dyspepsia that may not prompt a physician to consider a 13
Therefore to minimise the risks of missed diagnosis some
TE D
diagnosis of celiac disease.12,
international specialist centres have advocated a routine duodenal biopsy approach.14,
15
However, a routine biopsy approach has significant cost implications. In an effort to positively identify appropriate patients for duodenal biopsy we have previously described a clinical
EP
algorithm which combines symptoms and serology (Figure 1).16 This algorithm advocates biopsy sampling from ‘high-risk’ individuals with symptoms of diarrhea, weight loss or anemia
AC C
and also from patients with a positive tissue transglutaminase (tTG) result.16 By adopting this strategy, our group demonstrated a sensitivity of 100% for detecting patients who had undiagnosed celiac disease presenting to endoscopy, and confidently identified those patients who did not require a duodenal biopsy. However the limitation to using this clinical decision tool in practice is that serology results are not always available at the time of the endoscopic procedure. 17 Thus the use of an accurate point of care test that could provide a rapid result prior to endoscopy in appropriate patient groups could guide biopsy, avoid missed diagnosis and potentially be cost saving.
4
ACCEPTED MANUSCRIPT Three commercially available point of care tests for celiac disease are widely available internationally for both professional and personal use. So far only 1 test, Biocard, has been marketed in the USA but the tests are available online and increased availability is anticipated.18 A fourth point of care test has also been developed but not marketed. This is only available from the manufacturer and has no published data to support its validity. 3 of
RI PT
the 4 point of care tests available are based on tTG. The studies are limited by size and when compared against the gold standard of duodenal biopsy for all studied patients sensitivity has been disappointing (a summary of the available studies is shown in the supplemental file).19 More recently a novel point of care test (Simtomax), which utilises de-amidated gliadin
SC
peptide antibodies (DGP), has become available. This test may have several advantages including incorporating IgA and IgG antibodies as well as a total IgA assay. No previous studies
M AN U
have compared existing point of care tests. We aimed to evaluate which point of care test was optimal and assess point of care test in the setting of endoscopy. This approach could guide duodenal biopsy and validate our previously published diagnostic algorithm.
Methods
TE D
Patients were prospectively recruited from a specialist endoscopy list between March 2013 and February 2014 at the Royal Hallamshire Hospital in Sheffield. The endoscopy list is open to primary and secondary care referrals for patients with suspected celiac disease but also includes patients who are referred for open access endoscopy from primary care with for
EP
example, upper gastrointestinal symptoms, dyspepsia, gastro-oesophageal reflux disease,
AC C
weight loss, anemia or diarrhea.
All recruited patients were concurrently serologically tested for IgA tTG, IgA endomysial antibody (EMA) and a total IgA level. Duodenal biopsy was performed as the ‘gold standard’ in all cases. Patient consent was obtained prior to all gastroscopy examinations, with quadrantic biopsies from the second part of the duodenum and at least a single bulb biopsy taken in all patients. Patients were excluded from the study if they had known celiac disease or were on a gluten free diet at the time of the test. Patients with a coagulopathy, active gastrointestinal bleeding or a suspected carcinoma observed during the examination were also excluded.
5
ACCEPTED MANUSCRIPT Point of Care Tests In Group 1, we compared three commercially available point of care tests for celiac disease in patients referred with a positive EMA. EMA has been shown to have a high positive predictive value for celiac disease by comparison to other serological tests. The three tests
RI PT
analysed were: Biocard (BHR pharmaceuticals, Nuneaton, UK) based on IgA tTG, Celiac Quick Test (Biohit Healthcare UK, Ellesmere Port, UK) which measures IgA, IgG and IgM tTG, and Simtomax (Tillotts Pharma, Rheinfelden, Switzerland) that detects IgA and IgG anti-DGP. A fourth test, CeliacScreen Professional (Personal Diagnostics, Dorking, UK), was not available
SC
when the study was commenced. All of these tests are immunochromatographic tests and are performed in a similar way with whole blood and buffer placed on a test field which diffuses down a test strip. If antibodies are detected a visible line is seen. All of the tests
M AN U
were carried out as per the individual manufacturers’ instructions by experienced endoscopy staff and were read independently of the endoscopy findings. In Group 2, we evaluated a single point of care test, Simtomax in all patients attending a single endoscopy list. Figure 2 demonstrates the appearances of both a positive and
TE D
negative Simtomax test result.
In all patients the point of care test was compared to the presence of villous atrophy on duodenal biopsy as the gold standard for diagnosing celiac disease.
EP
Serology
Total IgA was measured on a Behring BN2 nephelometer. IgA tTG antibodies were assayed
AC C
using enzyme-linked immunosorbent assay kits (Aesku Diagnostics, Wendelsheim, Germany). A tTG titre of > 15 U/ml was regarded as positive. IgA EMA was detected by immunofluorescence on primate oesophagus sections (Binding Site, Birmingham, UK). Biopsies and histology
In total, at least five biopsies were taken from the duodenum, including at least one biopsy from the bulb, with each biopsy fixed in formalin at the time of the endoscopy. Specimens were then processed, orientated and embedded in paraffin wax. Standard 3 µm thick sections at three levels were stained with haematoxylin and eosin and reported routinely by
6
ACCEPTED MANUSCRIPT histopathologists without knowledge of the point of care test results. We graded villous atrophy according to the modified Marsh criteria. The presence of villous atrophy (Marsh 3a-3c) on histology and supporting positive EMA or tTG was required for a diagnosis of celiac disease. In cases of villous atrophy in the absence of a positive standard serological test HLA genotyping was performed with a negative HLA DQ2 or DQ8 phenotype used to
RI PT
rule out celiac disease. Supporting information such as family history and response to gluten free diet was also considered. Ethics
SC
This study protocol was approved by the Yorkshire and the Humber Research Ethics committee and registered with local research and development department of Sheffield
M AN U
Teaching Hospital NHS Foundation Trust under the registration number STH15416. The study was adopted to the National Institute of Health Research portfolio, study ID 14040. Statistics
Diagnostic value of the point of care test, EMA and tTG were calculated using sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) as well as positive
TE D
and negative likelihood ratios all with 95% confidence intervals. Sensitivity of the point of care tests was compared using the McNemar test for correlated proportions. Univariate analysis of presenting characteristics was performed using a chi square and odds ratios with
EP
95% confidence intervals were calculated. Statistical analysis was undertaken using SPSS version 19.0 (IBM). All authors had access to the study data and reviewed and approved the
AC C
final manuscript. Results
In Group 1, 55 patients (56.4% female, mean age 39.5 SD18.3) referred for positive EMA were tested with all 3 point of care test kits. In 9 patients the EMA had normalised on repeat testing. None of these patients had villous atrophy. Of the remaining 46 patients 36 had evidence of villous atrophy and were diagnosed with celiac disease. Of all the point of care tests tested Simtomax had the best sensitivity at 94.4% (80.0 – 99.0) which was superior to Celiac Quick Test 77.8% (60.4 – 89.3) (p 0.03) and Biocard 72.2% (54.6 – 85.2) (p 0.008). As all cases in Group 1 were identified on the basis of a positive IgA EMA there were 7
ACCEPTED MANUSCRIPT no cases of IgA deficiency in this cohort. However, in clinical practice IgA deficiency could reduce the sensitivity of the Biocard test and serum tTG. A full comparison of the point of care tests compared to tTG is shown in table 1. In Group 2, 508 patients (55.9% female, mean age 53.4 SD18.2) were recruited. 68 (59.7%
RI PT
female) patients investigated were diagnosed with celiac disease during the study period giving a prevalence of celiac disease in the study population of 13.4%.
6 (9%) of the newly diagnosed celiac disease patients had undergone a previous endoscopy in the last 10 years with a median delay to diagnosis from first endoscopy of 19 months
SC
(range 6 – 75 months). 1 patient had undergone 2 previous endoscopies.
Mean age was lower in the patients diagnosed with celiac disease (39.0 SD16.2) compared
M AN U
to controls (55.6 SD17.5) (p<0.0001). There was no significant difference in the gender profile of the celiac patients compared to controls. 278 patients were defined as high risk with symptoms of anemia (103 patients), diarrhea (122 patients) or weight loss (141 patients). The most common single reason for referral was gastro-oesophageal reflux (274 patients). On direct questioning, 94 patients fulfilled the Rome III criteria for IBS. 28 had
TE D
diarrhea predominant, 18 constipation predominant, 44 mixed type and 3 unclassified IBS. Lethargy was common in both celiac patients and controls with 113 patients complaining of feeling ‘tired all the time’. On univariate analysis patients with celiac disease were more likely to present with diarrhea (OR 2.6), feeling tired all the time (OR 2.3) or to have a first
EP
degree relative with celiac disease (OR 11.0). A comprehensive list of the presenting factors
AC C
is shown in Table 2.
Accuracy of the DGP point of care test Of the 68 new cases of celiac disease 63 tested positive using the DGP test giving a sensitivity of 92.7%. A full analysis of the accuracy of the point of care test compared to standard serology is shown in table 3. 5 patients with a negative point of care test were diagnosed with celiac disease. Of these, 1 patient was positive for both tTG and EMA, 1 patient was positive for EMA only and 3 patients had seronegative celiac disease confirmed by a compatible HLA genotype. 4 of the 5 patients with a false negative point of care test had high risk symptoms giving a sensitivity 8
ACCEPTED MANUSCRIPT of 98.5% (91.0 – 100), specificity 40.7% (36.1 – 45.4), PPV 20.4% (16.3 – 25.3) and NPV 99.4% (96.5 – 100) for our previously published diagnostic algorithm using the point of care test in place of tTG. The single patient missed by the point of care test algorithm presented with non-specific dyspepsia and had seronegative villous atrophy but was confirmed with a compatible HLA type. The point of care test had marginally superior sensitivity to tTG in the
RI PT
same algorithm which demonstrated sensitivity, specificity, PPV and NPV of 95.6% (86.8 – 98.9), 40.9% (36.3 – 45.7), 20.0% (15.9 – 24.9) and 98.4% (94.9 – 99.6) respectively. Economic impact
SC
The current NHS tariff (financial year 2013-14) for an OGD with duodenal biopsy is currently £387.00 ($634.76) and the cost of an OGD without biopsy is currently £349.00 ($572.43). The added cost of a duodenal biopsy is therefore £38.00 ($62.33). The additional cost of a
M AN U
routine biopsy strategy in our cohort would therefore be £19,304.00 ($31,662.52). According to the manufacturers of Simtomax the cost of a single Simtomax kit to the healthcare provider would be £17.00 ($27.88). The cost of point of care test for all patients with subsequent duodenal biopsy only in patients with a positive point of care test would cost £13,500.00 ($22,142.77). Therefore if we had used only a positive point of care test
TE D
result to dictate a duodenal biopsy there could have been a potential saving of £5804.00 ($9519.75) (£11,425.00 ($18,739.34) per 1000 endoscopies) compared to a routine biopsy strategy although this would have led to 5 missed cases of celiac disease in our cohort. The
EP
cost of the point of care test in place of tTG in our previously published clinical algorithm would have equated to £16,374.00 ($26,856.72) in this cohort leading to a potential saving of £2,930.00 ($4805.80) over routine duodenal biopsy (£5768.00 ($9460.70) per 1000
AC C
endoscopies). It is important to note that the NHS tariff does not take into account the time saving of avoiding duodenal biopsy for endoscopists and pathologists that could increase productivity and reduce overhead costs. Discussion
This is the first study to compare existing point of care test against each other using the gold standard of conventional serology and duodenal biopsy. We also validated a DGP point of care test in the endoscopy setting. Simtomax had a greater sensitivity than the other point of care tests that were tested in patients referred with a positive EMA. In Group 2, the 9
ACCEPTED MANUSCRIPT general endoscopy cohort, the performance of Simtomax was comparable to standard serology as a screening test (92.7% versus 91.2% for tTG). This equivalent sensitivity of serum tTG and the DGP test validates its potential as an office based tool for rapid serological testing. This has potential cost saving implications with fewer office visits required for patients and a potential reduction in ordering of unnecessary investigations.
RI PT
Furthermore endoscopy is carried out in both primary and secondary care in the US and the use of this point of care test in endoscopy prior to all procedures has the potential to significantly reduce the numbers of missed cases at endoscopy. The prevalence of celiac disease in our cohort was 13% and represents referral bias to a tertiary celiac center. In
SC
lower prevalence populations, such as the 2-4% expected for all-comers to endoscopy, the PPV will inevitably fall and this will have an impact on cost savings. However crucially for a
missed diagnoses.
M AN U
screening test the NPV will improve in lower prevalence populations resulting in fewer
A missed diagnosis of celiac disease may have several implications for the patient. Firstly it may result in patients experiencing persistent symptoms that may be potentially treatable with a gluten-free diet, avoidance of unnecessary further investigations and clinic visits with 21
Finally it may result in a delay in starting an effective
TE D
associated cost implications.20,
gluten-free diet which could reduce the risk of complications of celiac disease such as osteoporosis, anemia and lymphoma. More recently this has become a litigious issue for
EP
patients and clinicians.
Our previously published clinical algorithm was limited in ‘real clinical practice’ by the
AC C
unavailability of standard serology. This issue was highlighted by a recent US study demonstrating that in 68% of duodenal biopsy specimens received by the pathology department to rule out celiac disease no celiac serology had been performed.17 The diagnostic yield was 35.3% in patients with positive pre-endoscopy serology compared to 1.6% in those who hadn’t been tested prior to their endoscopy. Our present study has a crucial advantage over our previous pre-endoscopy testing study in that a rapid point of care test provides a feasible solution to the problem of missing serology results at the time of endoscopy. The DGP test performed as well as tTG within the diagnostic algorithm with a sensitivity of 98.5% (91.0 – 100) and a diagnostic yield of 20.2% compared to a diagnostic yield of 13.2% with a routine biopsy strategy. With the use of a rapid point of care test 10
ACCEPTED MANUSCRIPT serology results are available within 10 minutes and could be used routinely prior to a patient’s gastroscopy to guide duodenal biopsy and greatly increase the diagnostic yield. This point of care test approach could also be cost-effective particularly when offset against avoiding duplicate endoscopy, reducing biopsy in low risk groups, or the practice of routine
RI PT
duodenal biopsy. This approach also has the potential to reduce the time-burden and cost of preparation and reporting of duodenal biopsies that are likely to be normal. By incorporating the DGP test into our previously published clinical decision tool [13] 35.4% (180/508) patients would have avoided duodenal biopsy resulting in only a single missed
SC
diagnosis.22-25 The DGP test however performed equally well in both high and low risk patients. The most cost effective approach may therefore be to perform the point of care test in all patients and accept a small false negative rate. This would have resulted in the
M AN U
greatest saving in our cohort (£11,425.00 ($18,739.34) per 1000 endoscopies). Conclusions
In conclusion we have demonstrated that Simtomax has similar diagnostic accuracy to standard serological markers when used in an endoscopic setting. We have also validated a
TE D
clinically feasible diagnostic algorithm for identifying celiac disease in a endoscopy. Although the DGP test performed well in our celiac enriched population, further studies are required in lower prevalence populations, such as an open access setting, to assess the true impact of rapid testing in endoscopy. The results are however promising and this has the potential to
EP
improve the investigation of celiac disease by reducing missed diagnosis, unecessary
AC C
investigations and associated costs.
References
1. Ludvigsson JF, Rubio-Tapia A, van Dyke CT, et al. Increasing incidence of celiac disease in a North American population. Am J Gastroenterol. 2013;108(5):818-24. 2. Mustalahti K, Catassi C, Reunanen A, et al. The prevalence of celiac disease in Europe: results of a centralized, international mass screening project. Ann Med. 2010;42(8):587-95. Epub 2010/11/13. 11
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
3. Sanders DS, Patel D, Stephenson TJ, et al. A primary care cross-sectional study of undiagnosed adult coeliac disease. Eur J Gastroenterol Hepatol. 2003;15(4):407-13. 4. Fasano A, Berti I, Gerarduzzi T, et al. Prevalence of celiac disease in at-risk and not-at-risk groups in the United States: a large multicenter study. Arch Intern Med. 2003;163(3):286-92. 5. West J, Fleming KM, Tata LJ, et al. Incidence and prevalence of celiac disease and dermatitis herpetiformis in the UK over two decades: population-based study. Am J Gastroenterol. 2014;109(5):757-68. 6. Sanders DS, Hurlstone DP, Stokes RO, et al. Changing face of adult coeliac disease: experience of a single university hospital in South Yorkshire. Postgrad Med J. 2002;78(915):31-3. 7. Lo W, Sano K, Lebwohl B, et al. Changing presentation of adult celiac disease. Dig Dis Sci. 2003;48(2):395-8. 8. Ciclitira P, Dewar DH, McLaughlin SD, et al. The Management of Adults with Coeliac Disease. British Society of Gastroenterology. 2010. 9. Rubio-Tapia A, Hill ID, Kelly CP, et al. ACG clinical guidelines: diagnosis and management of celiac disease. Am J Gastroenterol. 2013;108(5):656-76; quiz 77. 10. Lebwohl B, Tennyson CA, Holub JL, et al. Sex and racial disparities in duodenal biopsy to evaluate for celiac disease. Gastrointest Endosc. 2012;76(4):779-85. 11. Olds G, McLoughlin R, O'Morian C, et al. Celiac disease for the endoscopist. Gastrointest Endosc. 2002;56(3):407-15. 12. Bardella MT, Minoli G, Ravizza D, et al. Increased prevalence of celiac disease in patients with dyspepsia. Arch Intern Med. 2000;160(10):1489-91. 13. Ford AC, Chey WD, Talley NJ, et al. Yield of diagnostic tests for celiac disease in individuals with symptoms suggestive of irritable bowel syndrome: systematic review and meta-analysis. Arch Intern Med. 2009;169(7):651-8. 14. Green PH, Murray JA. Routine duodenal biopsies to exclude celiac disease? Gastrointest Endosc. 2003;58(1):92-5. 15. Lebwohl B, Bhagat G, Markoff S, et al. Prior endoscopy in patients with newly diagnosed celiac disease: a missed opportunity? Dig Dis Sci. 2013;58(5):1293-8. 16. Hopper AD, Cross SS, Hurlstone DP, et al. Pre-endoscopy serological testing for coeliac disease: evaluation of a clinical decision tool. BMJ. 2007;334(7596):729. 17. Wiland HOt, Henricks WH, Daly TM. Limited utilization of serologic testing in patients undergoing duodenal biopsy for celiac disease. BMC Gastroenterol. 2013;13:156. 18. Mooney PD, Kurien M, Sanders DS. Simtomax, a novel point of care test for coeliac disease. Expert Opin Med Diagn. 2013. 19. Mooney PD, Kurien M, Evans KE, et al. Point-of-care testing for celiac disease has a low sensitivity in endoscopy. Gastrointest Endosc. 2014. 20. Dickey W, McConnell JB. How many hospital visits does it take before celiac sprue is diagnosed? J Clin Gastroenterol. 1996;23(1):21-3. Epub 1996/07/01. 21. Violato M, Gray A, Papanicolas I, et al. Resource use and costs associated with coeliac disease before and after diagnosis in 3,646 cases: results of a UK primary care database analysis. PLoS One. 2012;7(7):e41308. 22. Long KH, Rubio-Tapia A, Wagie AE, et al. The economics of coeliac disease: a populationbased study. Aliment Pharmacol Ther. 2010;32(2):261-9. 23. Green PH, Neugut AI, Naiyer AJ, et al. Economic benefits of increased diagnosis of celiac disease in a national managed care population in the United States. J Insur Med. 2008;40(3-4):21828. 24. Ukkola A, Kurppa K, Collin P, et al. Use of health care services and pharmaceutical agents in coeliac disease: a prospective nationwide study. BMC Gastroenterol. 2012;12:136. 25. Norstrom F, Sandstrom O, Lindholm L, et al. A gluten-free diet effectively reduces symptoms and health care consumption in a Swedish celiac disease population. BMC Gastroenterol. 2012;12:125. 12
ACCEPTED MANUSCRIPT
What is current knowledge There is significant under-diagnosis of celiac disease and many patients with celiac disease experience significant delay in diagnosis
•
A significant proportion of patients with celiac disease have undergone a previous gastroscopy where the opportunity to take duodenal biopsies had been missed
•
A previously published algorithm for diagnosing celiac disease at endoscopy reported a sensitivity of 100% but required celiac serology prior to endoscopy which is not always available
SC
RI PT
•
What is new here
Simtomax a novel point of care test for celiac disease that detects deamidated gliadin antibodies has equivalent sensitivity and specificity to serum tissue transglutaminase antibodies
•
In an endoscopic setting Simtomax in conjunction with a clinical algorithm could be used to direct duodenal biopsy which could result in significant cost savings and result in fewer missed diagnoses at endoscopy
AC C
EP
TE D
M AN U
•
Table 1: Analysis of the point of care tests in Group 1: Sensitivity for detecting villous atrophy and agreement of the available tests with serum tTG (tissue transglutaminase) and EMA (endomysial antibody); CI (confidence interval)
Test
Sensitivity % (95%CI)
Agreement with EMA
Kappa (95%CI)
Agreement with tTG
Kappa (95% CI)
Serum tTG
97.2 (83.8 – 99.9)
90.9 %
0.652 (0.371 - 0.933)
X
X
Biocard
72.2 (54.6 – 85.2)
70.9 %
0.346 (0.128 – 0.565)
69.1 %
0.346 (0.128 – 0.565)
13
ACCEPTED MANUSCRIPT Celiac Quick Test
77.8 (60.4 – 89.3)
76.4 %
0.384 (0.130 – 0.639)
78.2 %
0.384 (0.130 – 0.639)
Simtomax
94.4 (80.0 – 99.0)
87.3 %
0.513 (0.199 – 0.827)
89.1 %
0.561 (0.248 – 0.874)
Table 2: Univariate analysis of presenting characteristics in Group 2
508 278 274 150 141 122 113 103 93
%CELIAC DISEASE 13.4 16.5 6.2 8.7 14.9 23.0 22.1 19.4 19.4
43
18.6
1.5
Osteoporosis 37 Family history celiac disease 32 (1st degree relative) *Significantly reduced risk of celiac disease
10.8
0.8
n/a 1.1 - 3.2 0.1 - 0.4 0.3 - 1.0 0.7 - 2.1 1.5 - 4.4 1.3 - 4.0 1.0 - 3.2 1.0 - 3.2 0.7 – 3.5 0.3 - 2.2
56.3
11
5.1 - 23.4
TE D
Autoimmune thyroid disease
95% CI
p
n/a 0.03 <0.0001* 0.06 0.6 0.0007 0.003 0.06 0.09 0.4 0.8 < 0.0001
RI PT
All patients High risk Reflux Non-specific dyspepsia/abdo pain Weight loss Diarrhea Tired all the time Anemia IBS
Odds ratio n/a 1.9 0.2 0.5 1.2 2.6 2.3 1.8 1.8
SC
N
M AN U
Reason
Table 3: Analysis of the point of care test to standard serology. tTG (tissue transglutaminase), EMA (endomysial antibody), PPV (positive predictive value), NPV (negative predictive value), CI (confidence interval)
Sensitivity % (95% CI)
Specificity % (95% CI)
PPV % (95% CI)
NPV % (95% CI)
Simtomax
92.7 (83.0 – 97.3)
85.2 (81.5 – 88.3)
49.2 (40.3 – 58.2)
98.7 (96.8 – 99.5)
tTG
91.2 (81.1 – 96.4)
87.5 (84.0 – 90.4)
53.0 (43.6 – 62.2)
98.5 (96.5 – 99.4)
89.7 (79.3 -95.4)
96.6 (94.3 – 98.0)
80.3 (69.2 – 88.2)
98.4 (96.5 – 99.3)
AC C
EMA
EP
Test
Figure 1: Diagnostic algorithm for celiac disease in endoscopy Figure 2: Negative (left) and positive (right) Simtomax tests. Line A - IgA/IgG deamidated gliadin peptide, Line B - Total IgA, CT (control line)
14
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT Supplemental file 1: Summary of available point of care studies
Test
Ref
Country
Year
n
Population Sensitivity Specificity
20
165 Paediatric
21
Biocard
2006
Adult / 151 Paediatric
Hungary and Finland 2006
150 Paediatric
Italy
Biocard
Hungary
2007 2676 Paediatric
24
Biocard
Finland
2007
Biocard
Brazil
UK
2011
Adult / 254 Paediatric
Adult / 196 Paediatric
AC C
27
2009
EP
26
Biocard
43 Adult
TE D
25
Biocard
95.5%
M AN U
23
Biocard
90.2%
SC
22
97.4%
Libya
2011 2920 Paediatric
Romania
2013
97.6%
Antibody negative patients not 100.0% biopsied Compared to serology only
RI PT
Hungary and Finland 2005
Biocard
78.1%
91.7%
97.1%
100.0%
Adult/ 148 Paediatric
Antibody negative patients not biopsied
High pre-test probability 56% of patients had celiac 78.9% disease Only positive POCT biopsied
No sensitivity specificity data given Only positive POCT biopsied
No sensitivity specificity data given
50 patients had positive POCT of these 19 had biopsy confirmed CD. 4 patients with negative POCT had CD
28
Biocard
Study Limitations Antibody neg. without GI symptoms not biopsied
92.3%
100.0%
Not all patients tested with standard serology Only positive POCT or standard serology biopsied Compared to EMA only
ACCEPTED MANUSCRIPT 29
Biocard
UK
2014
523 Adult
70.1%
Italy
2006
Adult / 329 Paediatric
Switzerland 2012
250 Paediatric
Celiac Quick test (Stick CD-1)
Simtomax
Switzerland 2013 No published data.
Adult/ 66 Paediatric
AC C
EP
TE D
Celiac Screen Professional
93.1%
M AN U
31
Simtomax
100%
SC
30
RI PT
21
High pre-test probability 22.4% patients had celiac 96.6% disease Antibody negative patients not biopsied. Serum not whole blood 94.90% used. Only compared to serology
100.0%
95.0% Only compared to serology
93.1%
PII: DOI: Reference:
S1542-3565(15)00096-8 10.1016/j.cgh.2015.01.010 YJCGH 54140
To appear in: Clinical Gastroenterology and Hepatology Accepted Date: 9 January 2015 Please cite this article as: Mooney PD, Wong SH, Johnston AJ, Kurien M, Avgerinos A, Sanders DS, Increased Detection of Celiac Disease With Measurement of Deamidated Gliadin Peptide Antibody Before Endoscopy, Clinical Gastroenterology and Hepatology (2015), doi: 10.1016/j.cgh.2015.01.010. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. All studies published in Clinical Gastroenterology and Hepatology are embargoed until 3PM ET of the day they are published as corrected proofs on-line. Studies cannot be publicized as accepted manuscripts or uncorrected proofs.
ACCEPTED MANUSCRIPT Increased Detection of Celiac Disease With Measurement of Deamidated Gliadin Peptide Antibody Before Endoscopy
1,2
Dr Peter D Mooney* (MBChB), 2Simon H Wong, 2Alexander J Johnston,
1,2
Dr Matthew
RI PT
Kurien (MD), 1Dr Anastasios Avgerinos (MD), 1,2Prof David S Sanders (MD) Affiliations: 1Royal Hallamshire Hospital, Sheffield, UK, 2Univeristy of Sheffield, UK
SC
*Corresponding author: P floor Royal Hallamshire Hospital, Glossop Rd, Sheffield, S10 2JF. Tel: 0114 2261179 Fax: 0114 2712692. Email [email protected]
Acknowledgements: This investigator led study was partially funded by research grants
M AN U
from Tillotts Pharma, BHR Pharmaceuticals, and Coeliac UK. The study was adopted to the National Institute for Health Research portfolio, study ID 14040. None of the funding sources had any input into study design, access to study data or interpretation. Conflict of interest statement: DSS has received funding from Tillotts Pharma, BHR pharmaceuticals and Coeliac UK for investigator led studies into point of care testing for
TE D
celiac disease. PDM, SHW, AJJ, MK and AA declare no conflicts of interest.
Authorship statement: PDM recruited patients, obtained duodenal biopsies, analyzed the data and drafted the manuscript. SHW and AJJ performed point of care tests and collected
EP
patient data, MK and AA recruited patients and obtained duodenal biopsies, DSS conceived the study and edited the final manuscript and acts as guarantor for the article. All authors
AC C
had access to the study data and approved the final manuscript.
Abbreviations
DGP - De-amidated gliadin peptide antibody EMA - Endomysial antibody NPV - Negative predictive value PPV - Positive predictive value tTG - Tissue transglutaminase antibody
1
ACCEPTED MANUSCRIPT Abstract Background & Aims: Celiac disease is under-diagnosed. Many patients are examined by endoscopy, but celiac disease is missed or not detected. We evaluated the accuracy of finger prick-based point of care tests in detection of celiac disease and developed an
RI PT
algorithm for diagnosis.
SC
Methods: We performed a prospective study of 2 groups of patients with celiac disease evaluated at the Royal Hallamshire Hospital in Sheffield UK from March 2013 through
M AN U
February 2014. In group 1, patients at high risk of celiac disease who tested positive for endomysial antibody (n=55) were evaluated using the Biocard test and Celiac Quick Test, which measure antibodies to tissue transglutaminase (anti-tTG), and the Simtomax test, which measures deamidated gliadin peptide antibodies (DGP). Patients in group 2 (508
TE D
consecutive patients who underwent endoscopy examinations for any indication) received the DGP test, and were also evaluated using a diagnostic algorithm that incorporated results from the DGP test and data on symptoms. In both groups point of care tests were taken at
EP
the time of endoscopy and results were compared with results from histologic analyses of
AC C
duodenal biopsies from all patients.
Results: In Group 1, the DGP test identified patients with celiac disease with 94.4% sensitivity, the Celiac Quick Test with 77.8% sensitivity (P=.03 vs the DGP test), and the Biocard test with 72.2% sensitivity (P=.008 vs the DGP test). In Group 2, the DGP test identified patients with celiac disease with 92.7% sensitivity (95% confidence interval [CI], 83.0–97.3), 85.2% specificity (95% CI, 81.5–88.3), a positive predictive value of 49.2% (95% CI, 40.3–58.2), and a negative predictive value of 98.7% (95% CI, 96.8–99.5). Measurement 2
ACCEPTED MANUSCRIPT of serum anti-tTG identified patients with celiac disease with 91.2% sensitivity (95% CI, 81.1–96.4), 87.5% specificity (95% CI, 84.0–90.4), a positive predictive value of 53.0% (43.6– 62.2), and a negative predictive value of 98.5% (95% CI, 96.5–99.4). The algorithm identified patients with celiac disease with 98.5% sensitivity; its use could reduce duodenal biopsies by
RI PT
35%.
SC
Conclusion: In a prospective study, a test for DGP identified patients with celiac disease with similar levels of sensitivity and specificity as standard serologic analysis of anti-tTG. Use of
M AN U
the DGP test before endoscopy could increase the accuracy of diagnosis of celiac disease. Further studies, in lower prevalence populations, are required to assess the impact of the test in clinical practice.
AC C
EP
TE D
KEY WORDS: Finger prick test; case finding; missed diagnosis; diagnostic assay
3
ACCEPTED MANUSCRIPT Introduction Adult celiac disease is a common chronic inflammatory bowel condition caused by a heightened immunological response to dietary gluten. Internationally the prevalence of celiac disease is now recognised to be 0.2-1.2%.1-4 However, despite increased awareness of celiac disease worldwide meta-analyses have demonstrated that for every patient diagnosed 3 - 4
ranging from 4 to 13 years.6,
7
RI PT
patients go undetected.5 International cohorts have reported significant delays in diagnosis Furthermore, when specifically considering endoscopy, the
failure to recognise celiac disease (and perform a duodenal biopsy) ranges from 5-13%.
SC
Therefore, current national and international guidelines recommend routine duodenal biopsy in patients with high risk symptoms of weight loss, chronic diarrhea and anemia.8, 9 A recent large observational study however has demonstrated that this practice is highly variable with
M AN U
duodenal biopsy rates for these conditions ranging from 19 – 59%.10 One explanation for this under-use of duodenal biopsy may be reliance by clinicians on the presence of endoscopic markers of celiac disease to guide biopsy. However, endoscopic markers of celiac disease lack the required sensitivity.11 Furthermore many patients with celiac disease present with nonspecific symptoms such as IBS or dyspepsia that may not prompt a physician to consider a 13
Therefore to minimise the risks of missed diagnosis some
TE D
diagnosis of celiac disease.12,
international specialist centres have advocated a routine duodenal biopsy approach.14,
15
However, a routine biopsy approach has significant cost implications. In an effort to positively identify appropriate patients for duodenal biopsy we have previously described a clinical
EP
algorithm which combines symptoms and serology (Figure 1).16 This algorithm advocates biopsy sampling from ‘high-risk’ individuals with symptoms of diarrhea, weight loss or anemia
AC C
and also from patients with a positive tissue transglutaminase (tTG) result.16 By adopting this strategy, our group demonstrated a sensitivity of 100% for detecting patients who had undiagnosed celiac disease presenting to endoscopy, and confidently identified those patients who did not require a duodenal biopsy. However the limitation to using this clinical decision tool in practice is that serology results are not always available at the time of the endoscopic procedure. 17 Thus the use of an accurate point of care test that could provide a rapid result prior to endoscopy in appropriate patient groups could guide biopsy, avoid missed diagnosis and potentially be cost saving.
4
ACCEPTED MANUSCRIPT Three commercially available point of care tests for celiac disease are widely available internationally for both professional and personal use. So far only 1 test, Biocard, has been marketed in the USA but the tests are available online and increased availability is anticipated.18 A fourth point of care test has also been developed but not marketed. This is only available from the manufacturer and has no published data to support its validity. 3 of
RI PT
the 4 point of care tests available are based on tTG. The studies are limited by size and when compared against the gold standard of duodenal biopsy for all studied patients sensitivity has been disappointing (a summary of the available studies is shown in the supplemental file).19 More recently a novel point of care test (Simtomax), which utilises de-amidated gliadin
SC
peptide antibodies (DGP), has become available. This test may have several advantages including incorporating IgA and IgG antibodies as well as a total IgA assay. No previous studies
M AN U
have compared existing point of care tests. We aimed to evaluate which point of care test was optimal and assess point of care test in the setting of endoscopy. This approach could guide duodenal biopsy and validate our previously published diagnostic algorithm.
Methods
TE D
Patients were prospectively recruited from a specialist endoscopy list between March 2013 and February 2014 at the Royal Hallamshire Hospital in Sheffield. The endoscopy list is open to primary and secondary care referrals for patients with suspected celiac disease but also includes patients who are referred for open access endoscopy from primary care with for
EP
example, upper gastrointestinal symptoms, dyspepsia, gastro-oesophageal reflux disease,
AC C
weight loss, anemia or diarrhea.
All recruited patients were concurrently serologically tested for IgA tTG, IgA endomysial antibody (EMA) and a total IgA level. Duodenal biopsy was performed as the ‘gold standard’ in all cases. Patient consent was obtained prior to all gastroscopy examinations, with quadrantic biopsies from the second part of the duodenum and at least a single bulb biopsy taken in all patients. Patients were excluded from the study if they had known celiac disease or were on a gluten free diet at the time of the test. Patients with a coagulopathy, active gastrointestinal bleeding or a suspected carcinoma observed during the examination were also excluded.
5
ACCEPTED MANUSCRIPT Point of Care Tests In Group 1, we compared three commercially available point of care tests for celiac disease in patients referred with a positive EMA. EMA has been shown to have a high positive predictive value for celiac disease by comparison to other serological tests. The three tests
RI PT
analysed were: Biocard (BHR pharmaceuticals, Nuneaton, UK) based on IgA tTG, Celiac Quick Test (Biohit Healthcare UK, Ellesmere Port, UK) which measures IgA, IgG and IgM tTG, and Simtomax (Tillotts Pharma, Rheinfelden, Switzerland) that detects IgA and IgG anti-DGP. A fourth test, CeliacScreen Professional (Personal Diagnostics, Dorking, UK), was not available
SC
when the study was commenced. All of these tests are immunochromatographic tests and are performed in a similar way with whole blood and buffer placed on a test field which diffuses down a test strip. If antibodies are detected a visible line is seen. All of the tests
M AN U
were carried out as per the individual manufacturers’ instructions by experienced endoscopy staff and were read independently of the endoscopy findings. In Group 2, we evaluated a single point of care test, Simtomax in all patients attending a single endoscopy list. Figure 2 demonstrates the appearances of both a positive and
TE D
negative Simtomax test result.
In all patients the point of care test was compared to the presence of villous atrophy on duodenal biopsy as the gold standard for diagnosing celiac disease.
EP
Serology
Total IgA was measured on a Behring BN2 nephelometer. IgA tTG antibodies were assayed
AC C
using enzyme-linked immunosorbent assay kits (Aesku Diagnostics, Wendelsheim, Germany). A tTG titre of > 15 U/ml was regarded as positive. IgA EMA was detected by immunofluorescence on primate oesophagus sections (Binding Site, Birmingham, UK). Biopsies and histology
In total, at least five biopsies were taken from the duodenum, including at least one biopsy from the bulb, with each biopsy fixed in formalin at the time of the endoscopy. Specimens were then processed, orientated and embedded in paraffin wax. Standard 3 µm thick sections at three levels were stained with haematoxylin and eosin and reported routinely by
6
ACCEPTED MANUSCRIPT histopathologists without knowledge of the point of care test results. We graded villous atrophy according to the modified Marsh criteria. The presence of villous atrophy (Marsh 3a-3c) on histology and supporting positive EMA or tTG was required for a diagnosis of celiac disease. In cases of villous atrophy in the absence of a positive standard serological test HLA genotyping was performed with a negative HLA DQ2 or DQ8 phenotype used to
RI PT
rule out celiac disease. Supporting information such as family history and response to gluten free diet was also considered. Ethics
SC
This study protocol was approved by the Yorkshire and the Humber Research Ethics committee and registered with local research and development department of Sheffield
M AN U
Teaching Hospital NHS Foundation Trust under the registration number STH15416. The study was adopted to the National Institute of Health Research portfolio, study ID 14040. Statistics
Diagnostic value of the point of care test, EMA and tTG were calculated using sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) as well as positive
TE D
and negative likelihood ratios all with 95% confidence intervals. Sensitivity of the point of care tests was compared using the McNemar test for correlated proportions. Univariate analysis of presenting characteristics was performed using a chi square and odds ratios with
EP
95% confidence intervals were calculated. Statistical analysis was undertaken using SPSS version 19.0 (IBM). All authors had access to the study data and reviewed and approved the
AC C
final manuscript. Results
In Group 1, 55 patients (56.4% female, mean age 39.5 SD18.3) referred for positive EMA were tested with all 3 point of care test kits. In 9 patients the EMA had normalised on repeat testing. None of these patients had villous atrophy. Of the remaining 46 patients 36 had evidence of villous atrophy and were diagnosed with celiac disease. Of all the point of care tests tested Simtomax had the best sensitivity at 94.4% (80.0 – 99.0) which was superior to Celiac Quick Test 77.8% (60.4 – 89.3) (p 0.03) and Biocard 72.2% (54.6 – 85.2) (p 0.008). As all cases in Group 1 were identified on the basis of a positive IgA EMA there were 7
ACCEPTED MANUSCRIPT no cases of IgA deficiency in this cohort. However, in clinical practice IgA deficiency could reduce the sensitivity of the Biocard test and serum tTG. A full comparison of the point of care tests compared to tTG is shown in table 1. In Group 2, 508 patients (55.9% female, mean age 53.4 SD18.2) were recruited. 68 (59.7%
RI PT
female) patients investigated were diagnosed with celiac disease during the study period giving a prevalence of celiac disease in the study population of 13.4%.
6 (9%) of the newly diagnosed celiac disease patients had undergone a previous endoscopy in the last 10 years with a median delay to diagnosis from first endoscopy of 19 months
SC
(range 6 – 75 months). 1 patient had undergone 2 previous endoscopies.
Mean age was lower in the patients diagnosed with celiac disease (39.0 SD16.2) compared
M AN U
to controls (55.6 SD17.5) (p<0.0001). There was no significant difference in the gender profile of the celiac patients compared to controls. 278 patients were defined as high risk with symptoms of anemia (103 patients), diarrhea (122 patients) or weight loss (141 patients). The most common single reason for referral was gastro-oesophageal reflux (274 patients). On direct questioning, 94 patients fulfilled the Rome III criteria for IBS. 28 had
TE D
diarrhea predominant, 18 constipation predominant, 44 mixed type and 3 unclassified IBS. Lethargy was common in both celiac patients and controls with 113 patients complaining of feeling ‘tired all the time’. On univariate analysis patients with celiac disease were more likely to present with diarrhea (OR 2.6), feeling tired all the time (OR 2.3) or to have a first
EP
degree relative with celiac disease (OR 11.0). A comprehensive list of the presenting factors
AC C
is shown in Table 2.
Accuracy of the DGP point of care test Of the 68 new cases of celiac disease 63 tested positive using the DGP test giving a sensitivity of 92.7%. A full analysis of the accuracy of the point of care test compared to standard serology is shown in table 3. 5 patients with a negative point of care test were diagnosed with celiac disease. Of these, 1 patient was positive for both tTG and EMA, 1 patient was positive for EMA only and 3 patients had seronegative celiac disease confirmed by a compatible HLA genotype. 4 of the 5 patients with a false negative point of care test had high risk symptoms giving a sensitivity 8
ACCEPTED MANUSCRIPT of 98.5% (91.0 – 100), specificity 40.7% (36.1 – 45.4), PPV 20.4% (16.3 – 25.3) and NPV 99.4% (96.5 – 100) for our previously published diagnostic algorithm using the point of care test in place of tTG. The single patient missed by the point of care test algorithm presented with non-specific dyspepsia and had seronegative villous atrophy but was confirmed with a compatible HLA type. The point of care test had marginally superior sensitivity to tTG in the
RI PT
same algorithm which demonstrated sensitivity, specificity, PPV and NPV of 95.6% (86.8 – 98.9), 40.9% (36.3 – 45.7), 20.0% (15.9 – 24.9) and 98.4% (94.9 – 99.6) respectively. Economic impact
SC
The current NHS tariff (financial year 2013-14) for an OGD with duodenal biopsy is currently £387.00 ($634.76) and the cost of an OGD without biopsy is currently £349.00 ($572.43). The added cost of a duodenal biopsy is therefore £38.00 ($62.33). The additional cost of a
M AN U
routine biopsy strategy in our cohort would therefore be £19,304.00 ($31,662.52). According to the manufacturers of Simtomax the cost of a single Simtomax kit to the healthcare provider would be £17.00 ($27.88). The cost of point of care test for all patients with subsequent duodenal biopsy only in patients with a positive point of care test would cost £13,500.00 ($22,142.77). Therefore if we had used only a positive point of care test
TE D
result to dictate a duodenal biopsy there could have been a potential saving of £5804.00 ($9519.75) (£11,425.00 ($18,739.34) per 1000 endoscopies) compared to a routine biopsy strategy although this would have led to 5 missed cases of celiac disease in our cohort. The
EP
cost of the point of care test in place of tTG in our previously published clinical algorithm would have equated to £16,374.00 ($26,856.72) in this cohort leading to a potential saving of £2,930.00 ($4805.80) over routine duodenal biopsy (£5768.00 ($9460.70) per 1000
AC C
endoscopies). It is important to note that the NHS tariff does not take into account the time saving of avoiding duodenal biopsy for endoscopists and pathologists that could increase productivity and reduce overhead costs. Discussion
This is the first study to compare existing point of care test against each other using the gold standard of conventional serology and duodenal biopsy. We also validated a DGP point of care test in the endoscopy setting. Simtomax had a greater sensitivity than the other point of care tests that were tested in patients referred with a positive EMA. In Group 2, the 9
ACCEPTED MANUSCRIPT general endoscopy cohort, the performance of Simtomax was comparable to standard serology as a screening test (92.7% versus 91.2% for tTG). This equivalent sensitivity of serum tTG and the DGP test validates its potential as an office based tool for rapid serological testing. This has potential cost saving implications with fewer office visits required for patients and a potential reduction in ordering of unnecessary investigations.
RI PT
Furthermore endoscopy is carried out in both primary and secondary care in the US and the use of this point of care test in endoscopy prior to all procedures has the potential to significantly reduce the numbers of missed cases at endoscopy. The prevalence of celiac disease in our cohort was 13% and represents referral bias to a tertiary celiac center. In
SC
lower prevalence populations, such as the 2-4% expected for all-comers to endoscopy, the PPV will inevitably fall and this will have an impact on cost savings. However crucially for a
missed diagnoses.
M AN U
screening test the NPV will improve in lower prevalence populations resulting in fewer
A missed diagnosis of celiac disease may have several implications for the patient. Firstly it may result in patients experiencing persistent symptoms that may be potentially treatable with a gluten-free diet, avoidance of unnecessary further investigations and clinic visits with 21
Finally it may result in a delay in starting an effective
TE D
associated cost implications.20,
gluten-free diet which could reduce the risk of complications of celiac disease such as osteoporosis, anemia and lymphoma. More recently this has become a litigious issue for
EP
patients and clinicians.
Our previously published clinical algorithm was limited in ‘real clinical practice’ by the
AC C
unavailability of standard serology. This issue was highlighted by a recent US study demonstrating that in 68% of duodenal biopsy specimens received by the pathology department to rule out celiac disease no celiac serology had been performed.17 The diagnostic yield was 35.3% in patients with positive pre-endoscopy serology compared to 1.6% in those who hadn’t been tested prior to their endoscopy. Our present study has a crucial advantage over our previous pre-endoscopy testing study in that a rapid point of care test provides a feasible solution to the problem of missing serology results at the time of endoscopy. The DGP test performed as well as tTG within the diagnostic algorithm with a sensitivity of 98.5% (91.0 – 100) and a diagnostic yield of 20.2% compared to a diagnostic yield of 13.2% with a routine biopsy strategy. With the use of a rapid point of care test 10
ACCEPTED MANUSCRIPT serology results are available within 10 minutes and could be used routinely prior to a patient’s gastroscopy to guide duodenal biopsy and greatly increase the diagnostic yield. This point of care test approach could also be cost-effective particularly when offset against avoiding duplicate endoscopy, reducing biopsy in low risk groups, or the practice of routine
RI PT
duodenal biopsy. This approach also has the potential to reduce the time-burden and cost of preparation and reporting of duodenal biopsies that are likely to be normal. By incorporating the DGP test into our previously published clinical decision tool [13] 35.4% (180/508) patients would have avoided duodenal biopsy resulting in only a single missed
SC
diagnosis.22-25 The DGP test however performed equally well in both high and low risk patients. The most cost effective approach may therefore be to perform the point of care test in all patients and accept a small false negative rate. This would have resulted in the
M AN U
greatest saving in our cohort (£11,425.00 ($18,739.34) per 1000 endoscopies). Conclusions
In conclusion we have demonstrated that Simtomax has similar diagnostic accuracy to standard serological markers when used in an endoscopic setting. We have also validated a
TE D
clinically feasible diagnostic algorithm for identifying celiac disease in a endoscopy. Although the DGP test performed well in our celiac enriched population, further studies are required in lower prevalence populations, such as an open access setting, to assess the true impact of rapid testing in endoscopy. The results are however promising and this has the potential to
EP
improve the investigation of celiac disease by reducing missed diagnosis, unecessary
AC C
investigations and associated costs.
References
1. Ludvigsson JF, Rubio-Tapia A, van Dyke CT, et al. Increasing incidence of celiac disease in a North American population. Am J Gastroenterol. 2013;108(5):818-24. 2. Mustalahti K, Catassi C, Reunanen A, et al. The prevalence of celiac disease in Europe: results of a centralized, international mass screening project. Ann Med. 2010;42(8):587-95. Epub 2010/11/13. 11
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
3. Sanders DS, Patel D, Stephenson TJ, et al. A primary care cross-sectional study of undiagnosed adult coeliac disease. Eur J Gastroenterol Hepatol. 2003;15(4):407-13. 4. Fasano A, Berti I, Gerarduzzi T, et al. Prevalence of celiac disease in at-risk and not-at-risk groups in the United States: a large multicenter study. Arch Intern Med. 2003;163(3):286-92. 5. West J, Fleming KM, Tata LJ, et al. Incidence and prevalence of celiac disease and dermatitis herpetiformis in the UK over two decades: population-based study. Am J Gastroenterol. 2014;109(5):757-68. 6. Sanders DS, Hurlstone DP, Stokes RO, et al. Changing face of adult coeliac disease: experience of a single university hospital in South Yorkshire. Postgrad Med J. 2002;78(915):31-3. 7. Lo W, Sano K, Lebwohl B, et al. Changing presentation of adult celiac disease. Dig Dis Sci. 2003;48(2):395-8. 8. Ciclitira P, Dewar DH, McLaughlin SD, et al. The Management of Adults with Coeliac Disease. British Society of Gastroenterology. 2010. 9. Rubio-Tapia A, Hill ID, Kelly CP, et al. ACG clinical guidelines: diagnosis and management of celiac disease. Am J Gastroenterol. 2013;108(5):656-76; quiz 77. 10. Lebwohl B, Tennyson CA, Holub JL, et al. Sex and racial disparities in duodenal biopsy to evaluate for celiac disease. Gastrointest Endosc. 2012;76(4):779-85. 11. Olds G, McLoughlin R, O'Morian C, et al. Celiac disease for the endoscopist. Gastrointest Endosc. 2002;56(3):407-15. 12. Bardella MT, Minoli G, Ravizza D, et al. Increased prevalence of celiac disease in patients with dyspepsia. Arch Intern Med. 2000;160(10):1489-91. 13. Ford AC, Chey WD, Talley NJ, et al. Yield of diagnostic tests for celiac disease in individuals with symptoms suggestive of irritable bowel syndrome: systematic review and meta-analysis. Arch Intern Med. 2009;169(7):651-8. 14. Green PH, Murray JA. Routine duodenal biopsies to exclude celiac disease? Gastrointest Endosc. 2003;58(1):92-5. 15. Lebwohl B, Bhagat G, Markoff S, et al. Prior endoscopy in patients with newly diagnosed celiac disease: a missed opportunity? Dig Dis Sci. 2013;58(5):1293-8. 16. Hopper AD, Cross SS, Hurlstone DP, et al. Pre-endoscopy serological testing for coeliac disease: evaluation of a clinical decision tool. BMJ. 2007;334(7596):729. 17. Wiland HOt, Henricks WH, Daly TM. Limited utilization of serologic testing in patients undergoing duodenal biopsy for celiac disease. BMC Gastroenterol. 2013;13:156. 18. Mooney PD, Kurien M, Sanders DS. Simtomax, a novel point of care test for coeliac disease. Expert Opin Med Diagn. 2013. 19. Mooney PD, Kurien M, Evans KE, et al. Point-of-care testing for celiac disease has a low sensitivity in endoscopy. Gastrointest Endosc. 2014. 20. Dickey W, McConnell JB. How many hospital visits does it take before celiac sprue is diagnosed? J Clin Gastroenterol. 1996;23(1):21-3. Epub 1996/07/01. 21. Violato M, Gray A, Papanicolas I, et al. Resource use and costs associated with coeliac disease before and after diagnosis in 3,646 cases: results of a UK primary care database analysis. PLoS One. 2012;7(7):e41308. 22. Long KH, Rubio-Tapia A, Wagie AE, et al. The economics of coeliac disease: a populationbased study. Aliment Pharmacol Ther. 2010;32(2):261-9. 23. Green PH, Neugut AI, Naiyer AJ, et al. Economic benefits of increased diagnosis of celiac disease in a national managed care population in the United States. J Insur Med. 2008;40(3-4):21828. 24. Ukkola A, Kurppa K, Collin P, et al. Use of health care services and pharmaceutical agents in coeliac disease: a prospective nationwide study. BMC Gastroenterol. 2012;12:136. 25. Norstrom F, Sandstrom O, Lindholm L, et al. A gluten-free diet effectively reduces symptoms and health care consumption in a Swedish celiac disease population. BMC Gastroenterol. 2012;12:125. 12
ACCEPTED MANUSCRIPT
What is current knowledge There is significant under-diagnosis of celiac disease and many patients with celiac disease experience significant delay in diagnosis
•
A significant proportion of patients with celiac disease have undergone a previous gastroscopy where the opportunity to take duodenal biopsies had been missed
•
A previously published algorithm for diagnosing celiac disease at endoscopy reported a sensitivity of 100% but required celiac serology prior to endoscopy which is not always available
SC
RI PT
•
What is new here
Simtomax a novel point of care test for celiac disease that detects deamidated gliadin antibodies has equivalent sensitivity and specificity to serum tissue transglutaminase antibodies
•
In an endoscopic setting Simtomax in conjunction with a clinical algorithm could be used to direct duodenal biopsy which could result in significant cost savings and result in fewer missed diagnoses at endoscopy
AC C
EP
TE D
M AN U
•
Table 1: Analysis of the point of care tests in Group 1: Sensitivity for detecting villous atrophy and agreement of the available tests with serum tTG (tissue transglutaminase) and EMA (endomysial antibody); CI (confidence interval)
Test
Sensitivity % (95%CI)
Agreement with EMA
Kappa (95%CI)
Agreement with tTG
Kappa (95% CI)
Serum tTG
97.2 (83.8 – 99.9)
90.9 %
0.652 (0.371 - 0.933)
X
X
Biocard
72.2 (54.6 – 85.2)
70.9 %
0.346 (0.128 – 0.565)
69.1 %
0.346 (0.128 – 0.565)
13
ACCEPTED MANUSCRIPT Celiac Quick Test
77.8 (60.4 – 89.3)
76.4 %
0.384 (0.130 – 0.639)
78.2 %
0.384 (0.130 – 0.639)
Simtomax
94.4 (80.0 – 99.0)
87.3 %
0.513 (0.199 – 0.827)
89.1 %
0.561 (0.248 – 0.874)
Table 2: Univariate analysis of presenting characteristics in Group 2
508 278 274 150 141 122 113 103 93
%CELIAC DISEASE 13.4 16.5 6.2 8.7 14.9 23.0 22.1 19.4 19.4
43
18.6
1.5
Osteoporosis 37 Family history celiac disease 32 (1st degree relative) *Significantly reduced risk of celiac disease
10.8
0.8
n/a 1.1 - 3.2 0.1 - 0.4 0.3 - 1.0 0.7 - 2.1 1.5 - 4.4 1.3 - 4.0 1.0 - 3.2 1.0 - 3.2 0.7 – 3.5 0.3 - 2.2
56.3
11
5.1 - 23.4
TE D
Autoimmune thyroid disease
95% CI
p
n/a 0.03 <0.0001* 0.06 0.6 0.0007 0.003 0.06 0.09 0.4 0.8 < 0.0001
RI PT
All patients High risk Reflux Non-specific dyspepsia/abdo pain Weight loss Diarrhea Tired all the time Anemia IBS
Odds ratio n/a 1.9 0.2 0.5 1.2 2.6 2.3 1.8 1.8
SC
N
M AN U
Reason
Table 3: Analysis of the point of care test to standard serology. tTG (tissue transglutaminase), EMA (endomysial antibody), PPV (positive predictive value), NPV (negative predictive value), CI (confidence interval)
Sensitivity % (95% CI)
Specificity % (95% CI)
PPV % (95% CI)
NPV % (95% CI)
Simtomax
92.7 (83.0 – 97.3)
85.2 (81.5 – 88.3)
49.2 (40.3 – 58.2)
98.7 (96.8 – 99.5)
tTG
91.2 (81.1 – 96.4)
87.5 (84.0 – 90.4)
53.0 (43.6 – 62.2)
98.5 (96.5 – 99.4)
89.7 (79.3 -95.4)
96.6 (94.3 – 98.0)
80.3 (69.2 – 88.2)
98.4 (96.5 – 99.3)
AC C
EMA
EP
Test
Figure 1: Diagnostic algorithm for celiac disease in endoscopy Figure 2: Negative (left) and positive (right) Simtomax tests. Line A - IgA/IgG deamidated gliadin peptide, Line B - Total IgA, CT (control line)
14
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT Supplemental file 1: Summary of available point of care studies
Test
Ref
Country
Year
n
Population Sensitivity Specificity
20
165 Paediatric
21
Biocard
2006
Adult / 151 Paediatric
Hungary and Finland 2006
150 Paediatric
Italy
Biocard
Hungary
2007 2676 Paediatric
24
Biocard
Finland
2007
Biocard
Brazil
UK
2011
Adult / 254 Paediatric
Adult / 196 Paediatric
AC C
27
2009
EP
26
Biocard
43 Adult
TE D
25
Biocard
95.5%
M AN U
23
Biocard
90.2%
SC
22
97.4%
Libya
2011 2920 Paediatric
Romania
2013
97.6%
Antibody negative patients not 100.0% biopsied Compared to serology only
RI PT
Hungary and Finland 2005
Biocard
78.1%
91.7%
97.1%
100.0%
Adult/ 148 Paediatric
Antibody negative patients not biopsied
High pre-test probability 56% of patients had celiac 78.9% disease Only positive POCT biopsied
No sensitivity specificity data given Only positive POCT biopsied
No sensitivity specificity data given
50 patients had positive POCT of these 19 had biopsy confirmed CD. 4 patients with negative POCT had CD
28
Biocard
Study Limitations Antibody neg. without GI symptoms not biopsied
92.3%
100.0%
Not all patients tested with standard serology Only positive POCT or standard serology biopsied Compared to EMA only
ACCEPTED MANUSCRIPT 29
Biocard
UK
2014
523 Adult
70.1%
Italy
2006
Adult / 329 Paediatric
Switzerland 2012
250 Paediatric
Celiac Quick test (Stick CD-1)
Simtomax
Switzerland 2013 No published data.
Adult/ 66 Paediatric
AC C
EP
TE D
Celiac Screen Professional
93.1%
M AN U
31
Simtomax
100%
SC
30
RI PT
21
High pre-test probability 22.4% patients had celiac 96.6% disease Antibody negative patients not biopsied. Serum not whole blood 94.90% used. Only compared to serology
100.0%
95.0% Only compared to serology
93.1%