• 
• 

Three Part Question

In [children attending the emergency department with right sided abdominal pain] can [measurement of plasma calprotectin (S100A8/A9)] exclude the diagnosis of [acute appendicitis]

Clinical Scenario

A young child attends the emergency department with her concerned parents. She has been complaining of right sided lower abdominal pain for 2 days, with several associated episodes of vomiting. There is no fever, the child looks well and urinalysis is normal. On further questioning, several members of the family have had a recent viral illness.
On examination the child has some generalised discomfort around the umbilical and right iliac regions, but no localised peritonism. Observations are within normal limits.
After you have finished your examination, the father declares himself to be a barrister with a medicolegal firm. He mentions appendicitis and is very insistent that further tests be done to refute the diagnosis. You have recently read a research article about a new biomarker for acute appendicitis and wonder whether the test would go some way towards reassuring you and/or the family. You resolve to consult the literature.

Search Strategy

Medline (1946 to May week 3, 2013) and Embase (1980 to 2013 week 20), the Cochrane Database of Systematic Reviews (2005 to April 2013), ACP Journal Club (1991 to April 2013), the Database of Abstracts of Reviews (2nd quarter 2013), the Cochrane Central Register of Controlled Trials (April 2013), the Cochrane Methodology Register (3rd quarter 2012), Health Technology Assessment (2nd quarter 2012) and the NHS Economic Evaluation Database (2nd quarter 2013) were searched using the Ovid interface. The reference lists of relevant articles were also hand searched.
(exp abdominal pain/OR Abdominal pain.mp OR exp appendicitis/OR exp acute appendicitis/OR appendicitis.mp OR exp appendectomy/OR Right iliac fossa pain.mp) AND (exp biological marker/OR calprotectin.mp OR biomarker.mp OR S100A8A9) AND (exp diagnosis/). LIMIT to human and English language.

Search Outcome

Six hundred and twenty-nine abstracts were retrieved and screened for relevance. Six papers were deemed to be directly relevant and assessed in full. One of these was a conference abstract repeating the results of a previous paper. Another was an editorial review.

Four papers were deemed to be independently directly applicable to the three-part question. These papers are presented in the table

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
Mills et al 2012 USA	A convenience sample of 1052 adults and children (848 analysed) presenting to the emergency department with acute right lower quadrant abdominal pain.	Prospective multicentre diagnostic cohort study.	Sensitivity for predefined cutpoint of 14 units	96.2% (95% CI 92.9 to 98.2)	Potential enrollment bias with no attempt at screening log. Variable pretest probability by site (5.9 to 43.3%) Potential shipping/delay in sample analysis with impact on results.
			Specificity for predefined cutpoint of 14 units	16% (95% CI 13.3 to 19.1)
			Area under the curve (ROC curve analysis)	0.66
			Pretest probability of appendicitis within the overall cohort	27.5%
Kharbanda et al 2012 USA	A consecutive sample of 192 eligible (176 enrolled) children aged 3-18 presenting to a single emergency department, with acute abdominal pain of less than 96 hours duration, undergoing evaluation for possible appendicitis.	Prospective, cross sectional diagnostic cohort study. Diagnostic serum samples were obtained at time of enrollment in the ED, centrifuged locally and assayed at a later stage. Appendicitis was defined by histopathology, postoperative surgical report, or telephone follow up ar 14-21 days for discharged patients.	Sensitivity of plasma calprotectin (using a cut off of <159ng/mL) for the diagnosis of appendicitis	100% (95% CI 91 to 100)	Patients only enrolled if clinicians had made an active decision to investigate either via blood samples or imaging. No assessor blinding. In this cohort of patients, the white cell count actually performed as a superior biomarker (using a cut off of <8.85), with a sensitivity of 100% (95% CI 92 to 100) and a specificity of 42% (95% CI 38 to 56)
			Specificity of plasma calprotectin for the diagnosis of appendicitis	27% (95% CI 19 to 37)
			AUC for plasma calprotectin	0.68
			Pretest probability of appendicitis within the cohort	34%
Thuijls et al 2009 Netherlands	51 consecutive patients undergoing appendicectomy, with additional data from 27 healthy volunteers	Case control study. Median plasma concentrations for lactoferrin and calprotectin were determined in 51 patients with histopatholgically proven appendicitis, and aditionally in 27 healthy volunteers.	Median plasma calprotectin levels in the appendicitis group compared to healthy controls	766ng/mL vs 239ng/mL respectively (p<0.001	No details/demographic data available for patients therefore uncertain if any paediatric participants. Retrospective analysis to assess for correlation only. Small sample. No attempt at pragmatic use of the test. Pilot data only. 4 hours for analysis of sample - impractical for routine clinical use.
			Correlation of calprotectin to CRP	Significant correlation (r2 0.114, p=0.015)
			Correlation of calprotectin to WBC	Significant correlation (r2 0.099, p=0.024)
Bealer and Colgin 2010 USA	181 adult and paediatric patients presenting to one of three community emergency departments with acute abdominal pain and the following inclusions: acute complaint; duration of symptoms less than 2 weeks; pain located primarily to right side of the body.	Prospective diagnostic cohort pilot study. Appendicitis was defined as positive histopathological report following appendicectomy. Patients who had not undergone surgery were contacted at 1-4 weeks following presentation to determine outcome.	Sensitivity of plasma calprotectin for the diagnosis of appendicitis using a cut off of 20 units	92.7% (95% CI 80.57 to 97.48%)	No assessor blinding. Homogenous cohort limiting generalisability. Broad inclusion criteria. No power calculation, with relatively small sample size. Both authors have a significant conflict of interest.
			Specificity of plasma calprotectin for the diagnosis of appendicitis using a cut off of 20 units	53.6% (95% CI 45.33 to 61.63%)
			Negative likelihood ratio	0.14 (95% CI 0.04 to 0.39)
			Area under the ROC curve	0.71
			Pretest probability of acute appendicitis in this cohort	22.7% (41/181)

Comment(s)

Abdominal pain is a common paediatric presentation to the emergency department with a large differential diagnosis. Appendicitis is often considered, but remains difficult to confirm or refute on clinical grounds alone. Current clinical decision tools such as the modified Alvarado score have shown limited sensitivity (Heineman and Drake, 2012). Imaging strategies such as CT or ultrasound scan are not without their limitations. Patients are often referred for inpatient surgical opinion and short-term admission. If suspicion remains, laparoscopic or open surgical evaluation of the appendix remains the gold standard. A sensitive diagnostic test has the potential to reduce the incidence of perforation by early diagnosis, while also decreasing unnecessary admissions and potentially negative laparoscopy/laparotomy rates and surgical complications. As yet, standard inflammatory markers as a diagnostic tool have proved unreliable in isolation or in tandem with standardised clinical assessment. Calprotectin is a protein complex found in the cytoplasm of neutrophils, released during degranulation. It has already been used for monitoring of rheumatoid arthritis, with documented advantages over standard tests (García-Arias et al, 2013). There is also work suggesting benefit from use in inflammatory bowel conditions and gastrointestinal neoplasia. Indeed, some authors suggest the use of fecal calprotectin to be sensitive and specific for monitoring of inflammatory bowel disease, including relapse prediction and response to treatment (Van Rheenen et al, 2010). This accepted use in inflammatory bowel conditions has led to the theory that calprotectin may serve as a superior marker of appendicitis compared to current options. Unfortunately, in current studies plasma calprotectin appears to show only reasonable sensitivity and limited specificity for the diagnosis of acute appendicitis. At present, there is little to suggest benefit over the use of standard inflammatory markers (C-reactive protein and white cell count). The CI range combined with the serious consequences of a false negative result preclude its use as a standalone ‘rule out’ test. The low specificity would also preclude discharge and lead to increased admission rates, also with the potential for higher negative laparotomy rates. Combination of the biomarker with a clinical decision rule predicting low pre-test probability could potentially be used to exclude disease, but the current estimates of specificity would be likely to preclude this due to the cost increase associated with unnecessary admissions and further investigation. No studies have examined this combination to date.

Editor Comment

CRP. C-reactive protein; ED, emergency department; ROC, receiver operating characteristic; WCC, white cell count.

Clinical Bottom Line

There is currently no evidence to suggest that serum calprotectin is superior to standard inflammatory markers for the exclusion or confirmation of suspected appendicitis. The area under the receiver operating characteristic curve ranges from 0.66 to 0.71 in the best studies. Clinical examination findings remain the cornerstone of surgical decision-making, with standard inflammatory markers, non-invasive imaging and inpatient observation utilised appropriately in borderline cases.

Level of Evidence

Level 2 - Studies considered were neither 1 or 3.

References

1. Mills A, Huckins D, Kwok H et al. Diagnostic characteristics of S100A8/A9 in a multicenter study of patients with acute right lower quadrant abdominal pain Academic Emergency Medicine 2012;19:48-55
2. Kharbanda A, Rai A, Cosme Y et al. Novel serum and urine markers for paediatric appendicits Academic Emergency Medicine 2012;19:56 - 62
3. Thuijls G, Derikx J, Prakken F et al. A pilot study on potential new plasma markers for diagnosis of acute appendicitis American Journal of Emergency Medicine 2011;29:256-260
4. Bealer J and Colgin M. S100A8/A9: A potential new diagnostic aid for acute appendicitis Academic Emergency Medicine 2010;17:333-336
5. Van Rheenen PF, Van de Vijver E, Fidler V. Faecal calprotectin for screening of patients with suspected inflammatory bowel disease: diagnostic meta-analysis BMJ 2010;341:c3369
6. García-Arias M, Pascual-Salcedo D, Ramiro S, et al. Calprotectin in rheumatoid arthritis : association with disease activity in a cross-sectional and a longitudinal cohort. Mol Diagn Ther 2013;17:49–56.
7. Heineman J, Drake D. BET 1: an evaluation of the Alvarado score as a diagnostic tool for appendicitis in children. Emerg Med J 2012;29:1013–14.