Presented at the 2009 Society for Academic Emergency Medicine annual meeting, New Orleans, LA, May 2009.
S100A8/A9: A Potential New Diagnostic Aid for Acute Appendicitis
Article first published online: 1 MAR 2010
© 2010 by the Society for Academic Emergency Medicine
Academic Emergency Medicine
Volume 17, Issue 3, pages 333–336, March 2010
How to Cite
Bealer, J. F. and Colgin, M. (2010), S100A8/A9: A Potential New Diagnostic Aid for Acute Appendicitis. Academic Emergency Medicine, 17: 333–336. doi: 10.1111/j.1553-2712.2010.00663.x
Drs. Bealer and Colgin have financial relationships with AspenBio Pharma, Inc.: Dr. Colgin is the Chief Scientific Officer and Dr. Bealer is a consultant for the company.
AspenBio Pharma, Inc., is the company that developed the enzyme-linked immunosorbent assay tested in this article.
- Issue published online: 1 MAR 2010
- Article first published online: 1 MAR 2010
- Received May 15, 2009; revision received August 10, 2009; accepted August 11, 2009.
Objectives: Diagnosing acute appendicitis is a daunting clinical challenge, as there is no single test that reliably distinguishes acute appendicitis from other etiologies of acute abdominal pain. In this study, the authors examined whether circulating levels of S100A8/A9 could be useful as a marker to aid in the diagnosis of acute appendicitis.
Methods: Plasma samples from emergency department (ED) patients with acute abdominal pain (n = 181) were tested using an immunoassay for S100A8/A9.
Results: The sensitivity and specificity for S100A8/A9 in diagnosing acute appendicitis were estimated to be 93% (95% confidence interval [CI] = 81% to 97%) and 54% (95% CI = 45% to 62%), respectively. Negative predictive value (NPV) was 96% (95% CI = 89% to 99%), and positive predictive value (PPV) was 37% (95% CI = 28% to 47%). Performance characteristics of elevated white blood cell (WBC) count were also estimated: sensitivity 63% (95% CI = 47% to 76%), specificity 67% (95% CI = 59% to 75%), NPV 86% (95% CI = 78% to 91%), and PPV 36% (95% CI = 26% to 47%).
Conclusions: This is the first report exploring the relationship between circulating S100A8/A9 and acute appendicitis and establishes proof of concept for this biomarker as a diagnostic test for acute appendicitis. Further studies are indicated to optimize the use of this biomarker, in conjunction with other established approaches.
ACADEMIC EMERGENCY MEDICINE 2010; 17:333–336 © 2010 by the Society for Academic Emergency Medicine
Acute appendicitis is the most common surgical condition that produces abdominal pain. The diagnostic differentiation of acute appendicitis from other causes of abdominal pain has increasingly utilized computed tomography (CT) scanning. However, a number of authors have raised significant questions regarding the utility of CT in this clinical setting.1–4 Ideally, CT would only be used as a diagnostic test for appendicitis, after a screening test has defined a patient at risk. Given these concerns, it would be desirable to have a noninvasive and inexpensive approach to ruling out acute appendicitis, which could eventually be utilized as a screening tool. S100A8 (also named calgranulin A; myeloid-related protein 8 [MRP8]) and S100A9 (calgranulin B; MRP14) are intracellular calcium-binding proteins, which are key to the transduction of calcium signaling during inflammation.5 These two independent proteins have a tissue-specific pattern of expression and readily form dimers that, when combined, are commonly known as either S100A8/A9 or calprotectin.6 S100A8/A9 is constitutively expressed in neutrophils, monocytes, some epithelial cells, and the keratinocytes of inflamed tissues, while not generally expressed in tissue macrophages or lymphocytes.5,6 Most of S100A8/A9’s pro-inflammatory functions require extracellular release, but the exact secretory mechanism is not fully understood. What is known about this mechanism is that the secretion of S100A8/A9 in vivo is tightly controlled and requires concomitant activation of two independent signal pathways. Signal one is activation of protein kinase C that can be induced by many different inflammatory stimuli.7 The second signal is provided by contact of phagocytes with activating surfaces, such as extracellular matrix (ECM) proteins or tumor necrosis factor (TNF)-stimulated endothelium, but not by interaction with resting endothelial cells. Thus, secretion of S100A8/A9 is restricted to the sites of monocyte–endothelial or monocyte–ECM interactions during inflammatory conditions.7,8 The overexpression of S100A8 and S100A9 at these types of inflammatory sites are well recognized and there is growing evidence that S100A8/A9 could be a biomarker for a number of inflammatory conditions.5,6,9
In this study, we hypothesized that circulating levels of S100A8/A9 (calgranulin A and B, MRP8/14, calprotectin) are increased in acute appendicitis. S100A8/A9 is a demonized calcium-binding protein of the S100 family that has a proven and pivotal role in gastrointestinal inflammation and could potentially differentiate acute appendicitis from noninflammatory causes of acute abdominal pain. The aim of this pilot study was to provide a preliminary characterization of the clinical performance characteristics of a blood test for S100A8/A9.
This was a prospective pilot study of human serum samples investigating the potential relationship between plasma levels of S100A8/A9 and acute appendicitis. The study protocol was approved and monitored by the individual institutional review boards at the participating hospitals.
Study Setting and Population
This study was conducted in the emergency departments (EDs) of three separate community hospitals. The study population of 181 patients included both adults and children presenting with acute abdominal pain. Inclusion criteria were defined as follows: 1) chief complaint was abdominal pain, 2) chief complaint was a new complaint for the patient, 3) duration of symptoms was less than 2 weeks, and 4) pain was located primarily on the right side of the body (iliac fossa) and/or primarily below the umbilicus. Exclusion criteria were as follows: 1) not meeting the above inclusion criteria; 2) patient had dysuria with burning, stinging, or itching at the urethral meatus; 3) a history of end-stage or metastatic cancer; 4) a history of recent trauma; and 5) previous appendectomy.
Between February 2008 and May 2008, patients complaining of right lower quadrant abdominal pain presenting to the study EDs were offered entry into our study. After informed consent was obtaining, patients were enrolled and received clinical care as directed by the treating physicians. As part of their clinical management, white blood cell (WBC) counts were determined for most patients, although the protocol did not require any specific diagnostic approach. In addition, all of these patients had a separate blood sample drawn for the determination of plasma S100A8/A9 levels. The S100A8/A9 concentration was determined using a sandwich enzyme-linked immunosorbent assay (AspenBio Pharma, Inc., Castle Rock, CO), read by the Bio-Rad model 680 microplate reader (Bio-Rad Laboratories, Inc., Hercules, CA). A preliminary normal range was determined by measuring S100A8/A9. The cutoff for this analysis was determined retrospectively based on a receiver operating characteristic (ROC) analysis of this patient population. The results of these tests were collected, but not reported to the physicians. Clinical outcome was defined on the basis of two separate data sources obtained following the ED visit. The first data source was histologic evaluation of the appendix specimen in those patients who underwent an appendectomy. For those patients who did not have an appendectomy, the second source of information was a telephone interview performed 1–4 weeks following presentation to the ED. Patients who had not undergone appendectomy and those with normal appendix histology were defined as not having appendicitis, while those with pathologic confirmation of the disease were defined as having appendicitis.
Performance characteristics (to predict acute appendicitis) were estimated and 95% confidence intervals (CIs) were calculated using SAS (SAS Institute, Cary, NC). Clinical performance characteristics of S100A8/A9 were evaluated using ROC data. Given that the goal of S100A8/A9 is to help identify true positives, the ROC curve generated indicated a potential cutoff value between 14–20 units (Figure 1). The WBC cutoff was 10 × 103/μL.
A total of 181 patients were evaluated in the study; 118 patients were female and 63 were male. The mean age was 31.5 years, with a range of 8–76 years. The prevalence of appendicitis was 22.7% (41/181) for the patient population. Idiopathic abdominal pain accounted for 84 patients (46%), and other significant diagnoses were found in 56 patients (31%). There were 31 patients who had an appendectomy, with a negative appendectomy rate of 9.7% (3/31). Table 1 presents performance characteristics of S100A8/A9 with a cutoff of 20 units.
|Diagnostic Method||Negative Likelihood Ratio||Positive Likelihood Ratio||NPV||PPV||Sensitivity||Specificity|
|WBC count (n = 178)*||0.56 (0.37–0.84)||1.92 (1.37–2.69)||86.1 (78.34–91.4)||35.7 (25.5–47.41)||62.5 (47.03–75.78)||67.4 (59.19–74.65)|
|S100A8/A9 (n = 181)||0.14 (0.04–0.39)||2.00 (1.64–2.43)||96.2 (89.29–98.68)||36.9 (28.2–46.53)||92.7 (80.57–97.48)||53.6 (45.33–61.63)|
This study shows a correlation between circulating S100A8/A9 and acute appendicitis, suggesting that the test has the potential to be sensitive for detecting the disease in patients with acute, right-lower-quadrant abdominal pain.
While there are no previous reports linking circulating S100A8/A9 and appendicitis, there is growing experience using this marker to detect other gastrointestinal inflammatory conditions. For example, fecal S100A8/A9 levels are elevated in patients with inflammatory bowel disease (IBD) and these levels are used clinically to quantify intestinal inflammation.10–16 The increased S100A8/A9 in fecal samples of IBD patients is partially explained by an excellent correlation between fecal excretion of indium-111–labeled granulocytes and fecal S100A8/A9 in these patients, suggesting that fecal S100A8/A9 is a reflection of granulocyte migration through the gut wall into the feces.17 Other studies further support the connection between S100A8/A9 and the gut. For example, global gene expression profiles of inflamed colonic tissue using DNA microarrays have shown S100A9 to be highly up-regulated compared to healthy tissue.18 In addition, several immunohistologic studies of inflamed bowel have confirmed an elevated expression of S100A8 and S100A9 by infiltrating monocytes, neutrophils, and epithelial cells.19 Similar to our findings in appendicitis, there is also a good correlation between S100A8 and S100A9 serum concentrations and IBD.9 This evidence, including our findings between S100A8/A9 and appendicitis, supports a potentially powerful role for S100A8/A9 as a biomarker of gastrointestinal disease.
Our study suggests that the known physiology of S100A8/A9 in gut inflammation also applies to appendicitis. While not a specific marker of appendicitis, S100A8/A9’s dual-activation pathway fits well with the known pathogenesis of appendicitis. In classically described appendicitis, luminal obstruction leads to increased transmural pressure that ultimately leads to venous occlusion and ischemia of the appendix. It is possible that ischemia would affect and damage the mucosa of the appendix first, as it is the most oxygen-sensitive tissue, and expose infiltrating neutrophils to ECM proteins or TNF-stimulated endothelium. If true, this would suggest that S100A8/A9 is being secreted at the earliest stages of appendicitis and may help explain the sensitivity we demonstrated in our study. It would also suggest that secretion of S100A8/A9 may increase in a crescendo fashion as appendicitis progresses, because greater disease progression would lead to greater tissue damage with increasing amounts of ECM exposure and neutrophil infiltration. This could suggest a role for S100A8/A9 in quantifying the severity of appendicitis far more accurately than the current nonspecific descriptive terminology.
WBC and C-reactive protein (CRP) measurements have been described as having some utility in the diagnosis of acute appendicitis. However, the sensitivity of these two tests to identify patients with acute appendicitis is relatively low.20 Past studies have found that increased levels of CRP show a greater relationship to the severity of appendicitis than that of increased WBC levels.21 Both tests show a correlation to acute appendicitis, although ultimately, the sensitivity of these tests is insufficient to achieve reliable rule-out.
There are limitations to this study that should be taken into consideration. First, this was a pilot study representing an initial investigation into the relationship between appendicitis and S100A8/A9. We did not prospectively require any specific diagnostic work-up. In addition, our patient population was fairly homogenous, being representative of the private suburban hospitals where the study was conducted. This limits our ability to make comments about specific demographic subpopulations, such as women and children, or how the test would perform among other patient populations in different hospital settings.
We report the first use of circulating S100A8/A9 to aid in the diagnosis of appendicitis, with a sensitivity of 93%, a specificity of 54%, and a positive predictive value (PPV) and a negative predictive value (NPV) of 39 and 96%, respectively, in this pilot study. If proven safe and effective in a statistically sized study, S100A8/A9 could possibly be a new diagnostic to aid in the work-up of appendicitis. Our findings have encouraged us to design a larger study, sized sufficiently to provide the statistical power to make specific conclusions about the utility of the test.
- 8Myeloid-related proteins 8 and 14 are specifically secreted during interaction of phagocytes and activated endothelium and are useful markers for monitoring disease activity in pauciarticular-onset juvenile rheumatoid arthritis. Arthritis Rheum. 2000; 43:628–37., , , et al.