CONTINUING MEDICAL EDUCATION

Continuing Medical Education Activity

in Academic Emergency Medicine
CME Editor: Hal Thomas, MD
Authors: Anupam B. Kharbanda, MD, MS, Yohaimi Cosme, MD, Khin Liu, Steven L. Spitalnik, MD, and Peter S.
Dayan, MD, MS
Article Title: Discriminative Accuracy of Novel and Traditional Biomarkers in Children With Suspected Appendicitis
Adjusted for Duration of Abdominal Pain
If you wish to receive credit for this activity, please refer to the website: http://www.wileyblackwellcme.com.
Accreditation and Designation Statement:
Blackwell Futura Media Services designates this
journal-based CME activity for a maximum of 1 AMA
PRA Category 1 Credit. Physicians should only claim
credit commensurate with the extent of their partici-
pation in the activity.
Blackwell Futura Media Services is accredited by
the Accreditation Council for Continuing Medical
Education to provide continuing medical education for
physicians.
Educational Objectives
After completing this exercise, the participant will be
able to discuss the indications for and usefulness of
various biomarkers to aid in diagnosis of appendicitis.
Activity Disclosures
No commercial support has been accepted related to
the development or publication of this activity.
Faculty Disclosures:
CME editor Hal Thomas, MD: no relevant nancial
relationships to disclose
Authors Anupam B. Kharbanda, MD, MS, Yohaimi
Cosme, MD, Khin Liu, Steven L. Spitalnik, MD, Peter S.
Dayan, MD, MS: No relevant nancial relationships to
disclose.
This manuscript underwent peer review in line with
the standards of editorial integrity and publication
ethics maintained by Academic Emergency Medicine.
The peer reviewers have no relevant nancial relation-
ships. The peer review process for Academic Emer-
gency Medicine is double-blinded. As such, the
identities of the reviewers are not disclosed in line with
the standard accepted practices of medical journal peer
review.
Conicts of interest have been identied and resolved
in accordance with Blackwell Futura Media Servicess
Policy on Activity Disclosure and Conict of Interest.
No relevant nancial relationships exist for any individ-
ual in control of the content and therefore there were
no conicts to resolve.
Instructions on Receiving Credit
For information on applicability and acceptance of
CME credit for this activity, please consult your profes-
sional licensing board.
This activity is designed to be completed within an
hour; physicians should claim only those credits that
reect the time actually spent in the activity. To suc-
cessfully earn credit, participants must complete the
activity during the valid credit period, which is up to
two years from initial publication.
Follow these steps to earn credit:
Log on to http://www.wileyblackwellcme.com
Read the target audience, educational objectives,
and activity disclosures.
Read the article in print or online format.
Reect on the article.
Access the CME Exam, and choose the best
answer to each question.
Complete the required evaluation component of
the activity.
This activity will be available for CME credit for
twelve months following its publication date. At that
time, it will be reviewed and potentially updated and
extended for an additional twelve months.
CLINICAL INVESTIGATIONS
Discriminative Accuracy of Novel
and Traditional Biomarkers in Children
With Suspected Appendicitis Adjusted
for Duration of Abdominal Pain
Anupam B. Kharbanda, MD, MS, Yohaimi Cosme, MD, Khin Liu, Steven L. Spitalnik, MD,
and Peter S. Dayan, MD, MS
Abstract
Objectives: The objective was to assess the accuracy of novel and traditional biomarkers in patients with
suspected appendicitis as a function of duration of symptoms.
Methods: This was a prospective cohort study, conducted in a tertiary care emergency department
(ED). The authors enrolled children 3 to 18 years old with acute abdominal pain of less than 96 hours
and measured serum levels of interleukin-6 (IL-6), interleukin-8 (IL-8), C-reactive protein (CRP), white
blood cell (WBC) count, and absolute neutrophil count (ANC). Final diagnosis was determined by histo-
pathology or telephone follow-up. Trends in biomarker levels were examined based on duration of
abdominal pain. The accuracy of biomarkers was assessed with receiver operating characteristic (ROC)
curves. Optimal cut-points and test performance characteristics were calculated for each biomarker.
Results: Of 280 patients enrolled, the median age was 11.3 years (interquartile range [IQR] = 8.6 to 14.8),
57% were male, and 33% had appendicitis. Median IL-6, median CRP, mean WBC count, and mean
ANC differed signicantly (p < 0.001) between patients with nonperforated appendicitis and those with-
out appendicitis; median IL-8 levels did not differ between groups. In nonperforated appendicitis, med-
ian IL-6, WBC, and ANC levels were maximal at less than 24 hours of pain, while CRP peaked between
24 and 48 hours. In perforated appendicitis, median IL-8 levels were highest by 24 hours, WBC count
and IL-6 by 24 to 48 hours, and CRP after 48 hours of pain. The WBC count appeared to be the most
useful marker to predict appendicitis in those with fewer than 24 or more than 48 hours of pain, while
CRP was the most useful in those with 24 to 48 hours of pain.
Conclusions: In this population, the serum levels and accuracy of novel and traditional biomarkers
varies in relation to duration of abdominal pain. IL-6 shows promise as a novel biomarker to identify
children with appendicitis.
ACADEMIC EMERGENCY MEDICINE 2011; 18:568574 2011 by the Society for Academic Emergency
Medicine
A
ppendicitis remains a challenging diagnosis in
the pediatric population.
1
Clinicians increas-
ingly rely on diagnostic imaging, such as com-
puted tomography (CT), to identify patients with
appendicitis.
2
Although CT is highly accurate, substantial
concerns have been raised regarding the overuse of CT,
particularly the adverse health effects resulting from
exposure to ionizing radiation.
3,4
Biomarkers may serve as alternative diagnostic tools
to identify patients at high or low risk for appendici-
tis.
57
To date, there is no optimal biomarker, or combi-
nation of biomarkers, for diagnosing appendicitis. Two
recent studies have suggested the prominent role of
cytokines (e.g., interleukins [IL-6 and IL-8]) in the
inammatory response of patients with appendicitis.
6,8
Previous authors have demonstrated that serum IL-6, a
ISSN 1069-6563 2011 by the Society for Academic Emergency Medicine
568 PII ISSN 1069-6563583 doi: 10.1111/j.1553-2712.2011.01095.x
From the Division of Pediatric Emergency Medicine (ABK, YC,
PSD) and the Department of Pathology and Cell Biology (KL,
SLS), Columbia University College of Physicians and Surgeons,
New York, NY. Dr. Kharbanda is currently with the Division of
Pediatric Emergency Medicine, University of Minnesota, Min-
neapolis, MN.
Received September 9, 2010; revision received December 23,
2010; accepted January 11, 2011.
Presented in part at the annual meeting of Pediatric Academic
Societies, Toronto, Ontario, Canada, May 2010.
Supported by Grant UL1 RR024156 from the National Center for
Research Resources (NICRR), a component of the National Insti-
tute of Health (NIH) and NIH Roadmap for Medical Research.
The authors have no potential conicts of interest to disclose.
Supervising Editor: Bema Bonsu, MD.
Address for correspondence and reprints: Anupam B. Khar-
banda, MD, MS; e-mail: abk@umn.edu.
potent inducer of the systemic immune response to a
bacterial infection, is up-regulated in appendicitis.
9
IL-8,
a chemoattractant for neutrophils, is also up-regulated
in proportion to the degree of neutrophil invasion in
the affected appendix.
10,11
Prior studies have revealed conicting and variable
performance of traditional biomarkers such as the
white blood cell (WBC) count to diagnose appendicitis,
emphasizing the need to explore and study novel bio-
markers.
1,9,1214
The limited utility of traditional biomar-
kers should not be surprising, as appendicitis is a
dynamic, evolving process, for which uctuations in
levels might be expected. This theory has been explored
by researchers who have demonstrated the diagnostic
benet of repeat laboratory measures (WBC, C-reactive
protein [CRP]) in hospitalized, adult patients with acute
abdominal pain.
1517
We hypothesized that novel and standard biomarkers
would have improved test performance if correlated
with a patients evolving, inammatory response to
appendicitis. Therefore, we aimed to describe the
trends and accuracy in serum levels of novel (IL-6, IL-8)
and traditional (WBC, absolute neutrophil count [ANC],
and CRP) biomarkers in relation to the duration of
abdominal pain in children with suspected appendicitis.
By using duration of pain as a surrogate for the degree
of inammation, we sought to identify which biomar-
kers might be most useful for distinguishing appendici-
tis from other causes of acute abdominal pain over the
course of illness.
METHODS
Study Design
We conducted a prospective, observational pilot study.
We obtained written informed consent from all parents
and assent from children 7 years of age and older. The
study was approved by the local institutional review
board.
Study Setting and Population
The study site was an urban, tertiary care, pediatric
emergency department (ED), with approximately 50,000
visits annually. From August 2008 to November 2009,
children 3 to 18 years of age who presented to the ED
with acute abdominal pain of less than 96 hours dura-
tion and who were being evaluated for possible
appendicitis were enrolled. We dened possible
appendicitis as when the treating physician obtained
blood tests, radiologic studies (CT and or ultrasound),
or a surgical consultation for the purpose of diagnosing
appendicitis. In our ED, it is standard practice to obtain
a WBC count with differential for all patients with sus-
pected appendicitis. Radiologic studies or surgical con-
sults are obtained at the discretion of the treating
physician. We excluded patients with any of the follow-
ing conditions: pregnancy, prior abdominal surgery
(e.g., gastrostomy tube, abdominal hernia repair),
chronic illness that potentially affected the gastrointesti-
nal system (e.g., cystic brosis, inammatory bowel dis-
ease, sickle cell anemia, chronic pancreatitis, diabetes,
immunosuppression), or a medical condition affecting
the providers ability to obtain an accurate history (e.g.,
substantial language or developmental delay). We also
excluded patients who had radiologic studies (CT or
ultrasound) of the abdomen performed prior to ED
arrival or a history of abdominal trauma within 7 days
of the ED evaluation.
Study Protocol
Pediatric emergency physicians completed standardized
histories and physical examinations prior to knowledge
of any radiologic studies, if obtained. As part of the
patient history, clinicians categorized the duration of
pain in 12-hour intervals. To assess the interobserver
reliability of the duration of pain, a second physician
assessor was asked to complete an additional data col-
lection tool on a subset of patients within 30 minutes of
the rst assessor.
The patients medical record was abstracted to obtain
data from laboratory, radiology, pathology, and opera-
tive reports. A single research assistant (YC) abstracted
the medical record and entered it into Microsoft Access
(Microsoft Corp., Redmond, WA); all data were double-
checked for accuracy by one author (ABK), including a
review of data interpretation and data entry. We
reviewed the daily ED admission log and electronic
tracking system to identify potentially eligible patients
who were not enrolled (i.e., missed).
Serum Collection and Analysis
We obtained serum WBC count and automated differ-
entials per standard hospital procedure. We obtained
an additional aliquot of patient serum (35 mL) and
placed it into a serum separator tube to measure IL-6,
IL-8, and CRP levels. These additional samples were
sent to the hospital clinical laboratory within 1 hour of
collection where they were spun at a speed of 1300 rel-
ative centrifugal force for 10 minutes. During the hours
of 0900 to 1600, trained laboratory technicians immedi-
ately divided the sample into two aliquots and froze the
blood at )80C. After 1600 on weekdays and at all times
on weekends, spun samples were stored at 4C. The fol-
lowing business day, the laboratory technicians divided
the samples into two aliquots and froze the blood at
)80 C. We recorded the time the serum was drawn,
delivered to the laboratory, and frozen.
The study laboratory technicians analyzed the serum
for IL-6 and IL-8 (R&D Systems, Minneapolis, MN) lev-
els by enzyme-linked immunosorbent assay (ELISA).
CRP (Siemens Healthcare Diagnostics, Deereld, IL)
was measured by particle enhanced immunonephelom-
etry. If the initial IL-6 level was at the lower detection
limit of 3.12 pg mL on the primary ELISA assay, an
additional hypersensitive IL-6 assay (R&D Systems) was
performed to assess the full spectrum of IL-6 levels in
patients with acute abdominal pain. The technicians
were blinded to the patients nal diagnosis.
Measures
The primary outcome was the presence or absence of
appendicitis, determined by reviewing the written
attending pathologist histopathology report for patients
who had an appendectomy. A perforated appendix was
determined by the attending surgeons written postop-
erative note. For patients who did not have surgery, we
ACADEMIC EMERGENCY MEDICINE
June 2011, Vol. 18, No. 6

www.aemj.org 569
determined the outcome by a follow-up telephone call
14 to 21 days following the index ED visit. If the family
could not be reached, a review of the hospital elec-
tronic record system was conducted to assess for oper-
ations (i.e., appendectomy), hospitalizations or ED visits
during the follow-up period. Those assessing the out-
come were blinded to the biomarker levels.
Data Analysis
For each biomarker, we conducted descriptive analyses,
exploring ranges, means with standard deviations
(SDs), medians with interquartile ranges (IQRs), and the
respective 95% condence intervals (CIs). We assessed
the association between each biomarker and the pres-
ence or absence of appendicitis with the unpaired t-test
or Mann-Whitney U-test for normally and asymmetri-
cally distributed data, respectively. To assess biomarker
trends and accuracy among groups with similar illness
severity, we then stratied our analyses by appendicitis
status (e.g., perforated appendicitis, nonperforated
appendicitis, or no appendicitis). Finally, we con-
structed receiver operating characteristic (ROC) curves
to determine the area under the curve (AUC) for the
various serum biomarkers based on the duration of
abdominal pain in three clinically relevant time inter-
vals: <24, 24 to 48, and >48 hours of pain. Within each
time category, we calculated the cut-point from the
ROC curve that maximized sensitivity and specicity
(based on the Youden index).
18
The test performance
characteristics (including likelihood ratio [LR]) for this
cut-point were then calculated. We determined the inte-
robserver reliability between physicians for their
assessment of duration of abdominal pain using Co-
hens unweighted kappa statistic.
1921
All statistical
analyses were performed using SPSS (Version 18.0,
SPSS Inc., Chicago, IL.
RESULTS
Study Population
Over the 15-month study period, 380 patients 3 to
18 years of age presented to the ED with acute abdomi-
nal pain and were considered for appendicitis. A total
of 291 were potentially eligible for participation, 280
were enrolled (96%), and 259 had complete biomarker
data available for analysis of trends and discriminative
characteristics (Figure 1).
Patient Characteristics
Of the 280 enrolled patients, the median age was
11.3 years (IQR = 8.6, 14.8 years), and 160 (57%) were
male. Comparisons between patients with appendicitis
and those without revealed no statistical difference by
sex, age, duration of abdominal pain, history of right
lower quadrant pain, mean temperature in the ED, ten-
derness on exam, or percentage who underwent imag-
ing (p > 0.05 for all comparisons). Pairs of physicians
(n = 41) showed good interobserver agreement on
their assessments of the duration of abdominal pain
(j = 0.63).
Ninety-four patients were diagnosed with appendici-
tis (33%), of whom 22 (23%) had a perforated appendix.
We completed telephone follow-up on 177 (99%) of the
179 patients who did not undergo an operation; none
had an appendectomy during the follow-up period.
Medical records for the two patients lost to telephone
follow-up revealed no further ED visits, operations, or
hospitalizations within 2 months of enrollment. Com-
pared to enrolled patients, eligible patients who were
not enrolled were slightly older (median age =
13.1 years) and more likely to be female (60%), but had
similar likelihoods of undergoing CT or ultrasound
(73%) and having appendicitis (36%).
Levels of Biomarkers in Patients With and Without
Appendicitis
The biomarker levels in patients without appendicitis,
with nonperforated appendicitis, and with perforated
appendicitis are shown in Table 1. Each serum biomar-
ker level except IL-8 was statistically higher for patients
with nonperforated appendicitis compared to those
without appendicitis (p < 0.001). The serum levels of all
biomarkers differed signicantly for children with per-
forated appendicitis compared to those with nonperfo-
rated appendicitis (p < 0.001).
Relationship of Duration of Abdominal Pain
to Biomarker Levels
In Figures 2A2D, the biomarker levels for IL-6, IL-8,
WBCs (ANC results closely approximate the WBC
count and are not shown), and CRP are shown for
each time interval as stratied by appendicitis status.
As seen in each gure, biomarker levels in patients
without appendicitis remain low and steadily decline
across the time intervals. In comparison, in cases of
either nonperforated or perforated appendicitis, bio-
marker levels rise and reach maximum values at differ-
ent rates. In nonperforated appendicitis, median IL-6
and WBC levels are maximal by 24 hours of pain, and
CRP by 24 to 48 hours, and median IL-8 levels were
similar across time categories. In perforated appendici-
tis, median IL-8 level is highest by 24 hours, WBC
count and IL-6 by 24 to 48 hours, and CRP after
48 hours of pain.
Time-dependent Accuracy of Biomarkers
Table 2 displays the discriminative ability of the biomar-
kers studied to distinguish patients with appendicitis
(perforated and nonperforated) from those without
appendicitis, stratied by duration of abdominal pain.
380 Screened
89 Excluded
48 Transferred
with imaging
35 Had chronic
illness
6 Met other
exclusion
criteria
11 Missed 280 Enrolled
21 Incomplete
sample
collecon
259 Complete
serum samples
Figure 1. Study patient ow.
570 Kharbanda et al.
TEMPORAL CHANGES IN BIOMARKERS FOR APPENDICITIS
Table 1
Biomarker Levels Stratied by Appendicitis Status
Biomarker
No Appendicitis
(n = 186)
Nonperforated
Appendicitis (n = 72)
Perforated
Appendicitis (n = 22)
WBC count (10
9
L)* (n = 276) 10.4 (4.8)
[9.711.1]
15.8 (4.5)
[14.816.9]
17.3 (4.4)
[15.319.2]
ANC (10
9
L)* (n = 262) 7.5 (4.7)
[6.88.2]
12.7 (4.6)
[11.613.8]
14.7 (4.5)
[12.616.7]
IL-6 (pg mL)|| (n = 259) 4.4 (1.316.1)
[2.45.7]
21.0 (9.837.0)
[15.725.1]
122.3 (42.2388.6)
[43.8250.6]
IL-8 (pg ml)|| (n = 259) 16.3 (11.024.2)
[13.917.5]
13.5 (10.6- 21.2)
[11.916.2]
25.2 (19.055.6)
[19.138.2]
CRP (mg L)|| (n = 260) 5.3 (0.4917.8)
[2.77.4]
14.1 (4.934.4)
[7.322.9]
115.1 (38.3179.0)
[38.5155]
ANC = absolute neutrophil count; CRP = C-reactive protein; IL-6 = interleukin-6; IL-8 = interleukin-8; IQR = interquartile range;
WBC = white blood cell.
*Mean (SD). Data for WBC count, ANC, and neutrophil count were normally distributed.
p < 0.001 for comparisons between no appendicitis and nonperforated appendicitis groups.
p < 0.001 for comparisons between nonperforated appendicitis and perforated appendicitis groups.
Values in brackets are 95% CIs.
||Median (IQR). Data for IL-6, IL-8, and CRP levels were skewed and thus not normally distributed.
(a) (b)
(c) (d)
Figure 2. (A) WBC levels based on duration of pain and appendicitis status. (B) IL-6 levels based on duration of pain and appendi-
citis status. (C) IL-8 levels based on duration of pain and appendicitis status. (D) CRP levels based on duration of pain and appendi-
citis status. CRP = C-reactive protein; IL-6 = interleukin-6; IL-8 = interleukin-8; WBC = white blood cell.
ACADEMIC EMERGENCY MEDICINE
June 2011, Vol. 18, No. 6

www.aemj.org 571
The WBC count, ANC, and IL-6 had relatively similar
AUCs for children with less than 24 hours of pain,
whereas CRP had a trend toward greater discriminative
accuracy (by AUC) in the 24- to 48-hour time period,
and the WBC count appeared to be more accurate than
other biomarkers in those with longer than 48 hours of
abdominal pain. Despite these trends, the 95% CIs for
the AUCs of these biomarkers did overlap.
Within each time category, we also provide cut-
points and test performance characteristics for each bi-
omarker, as calculated from their respective ROC
curves. Each biomarker exhibited uctuations in test
performance as a function of duration of pain. For
example, the test performance of the WBC count was
maximal (LR+ of 3.8) in the >48-hour time period, while
for IL-6 and CRP, optimal performance was seen in the
24- to 48-hour time period (LR+ of 3.3 and 6.6, respec-
tively). Based on the LR+, the WBC count appeared to
be the most useful marker to predict appendicitis in
those with <24 or >48 hours of pain, while CRP was the
most useful in those with 24 to 48 hours of pain.
DISCUSSION
In this pilot study, we have demonstrated the uctua-
tions in serum levels and accuracy of novel and tradi-
tional biomarkers for pediatric appendicitis as a function
of duration of pain. In patients with nonperforated
appendicitis, median serum IL-6, the WBC count, and
ANC levels were highest in the 24 hours after symptoms
of pain began, while median CRP levels trended higher
in those with 24 to 48 hours of pain. Furthermore, we
found that the novel biomarker IL-6 had substantial
discriminative ability, whereas IL-8 was seemingly less
useful for diagnosing nonperforated appendicitis.
We believe that an understanding of the uctuations
in biomarker levels in relation to the duration of
abdominal pain may be clinically useful. IL-6 increased
early in relation to symptoms of pain, consistent with
its role in stimulating the immune system to contain
infection, whereas CRP, which is stimulated by IL-6,
required 24 to 36 hours for up-regulation. In compari-
son, the WBC count and ANC showed little uctua-
tions in patients with nonperforated appendicitis. We
found it interesting that the biomarkers studied in this
pilot did not necessarily exhibit their best test perfor-
mance during time periods when their levels would be
expected to peak. For example, although overall CRP
levels peaked in those with >48 hours of pain, CRP
exhibited its best test performance in those with 24 to
48 hours of pain. Similarly, the WBC count revealed
improving test performance as duration of pain
increased. If further research conrms our ndings,
clinicians could tailor which laboratory test(s) they
obtained in relation to a patients history of symptoms.
In addition, the thresholds (as described in Table 2)
that would trigger further action would also change
based on an understanding of these uctuations.
Based on our results, the WBC count appears to be
the most useful biomarker in those patients with <24
or >48 hours of pain, while the CRP may be the most
useful in those with 24 to 48 hours of pain. In addi-
tion, if IL-6 is to be utilized in the diagnosis of appen-
dicitis, it may be most useful in those with symptoms
for 24 to 48 hours or it may be used to distinguish
which patients with appendicitis have a perforated
appendix.
For each of the biomarkers analyzed in this study,
serum levels remained low or declined in patients with-
out appendicitis, even as duration of symptoms
increased. We believe that the improved overall test
performance of the WBC count, as duration of pain
increased, can be attributed to the steady decline in the
WBC in patients without appendicitis (who likely have
Table 2
Biomarker Performance Based on Duration of Abdominal Pain
Biomarker
Duration of
Pain (Hours) AUC (95% CI)
Cut-point
Maximizing
Overall Accuracy
Sensitivity at
Cut-point,
% (95% CI)
Specicity at
Cut-point,
% (95% CI) LR+
WBC count
(10
9
L)
<24 0.78 (0.700.85) 14.6 67.8 (5479) 80.0 (7087) 3.4
2448 0.87 (0.780.95) 10.0 95.6 (76100.0) 71.4 (5584) 3.3
>48 0.92 (0.840.99) 10.0 100 (70100.0) 73.3 (5885) 3.8
ANC (10
9
L) <24 0.78 (0.700.85) 11.0 69.1 (5581) 74.7 (6483) 2.7
2448 0.85 (0.760.95) 8.4 90.5 (6898) 80.0 (6490) 4.5
>48 0.89 (0.800.98) 6.0 91.7 (60100.0) 66.7 (5080) 2.8
IL-6 (pg mL) <24 0.78 (0.710.86) 11.31 82.1 (6991) 68.5 (5878) 2.6
2448 0.87 (0.790.95) 5.7 95.6 (76100.0) 71.1 (5484) 3.3
>48 0.76 (0.600.93) 2.5 91.7 (60100.0) 58.5 (4273) 1.3
IL-8 (pg mL) 24 0.49 (0.400.59) 64.8 5.3 (1.814.6) 81.8 (7388) 0.3
2448 0.59 (0.440.74) 15.9 65.2 (4383) 63.2 (4678) 1.8
>48 0.57 (0.360.78) 24.1 41.2 (1771) 86.8 (7195) 3.2
CRP (mg L) <24 0.64 (0.550.73) 0.6 92.9 (8298) 28.1 (1939) 1.3
2448 0.89 (0.810.97) 20.8 87.0 (6597) 86.8 (7195) 6.6
>48 0.85 (0.730.96) 7.0 100 (7099) 56.1 (4071) 2.3
ANC = absolute neutrophil count; AUC = area under the receiver operating characteristic curve; CRP = C-reactive protein;
IL-6 = interleukin-6; IL-8 = interleukin-8; LR+ = likelihood ratio positive; WBC = white blood cell.
572 Kharbanda et al.
TEMPORAL CHANGES IN BIOMARKERS FOR APPENDICITIS
conditions that do not illicit as aggressive of an immune
response). These ndings add to prior literature, in
which low levels of the WBC and ANC have provided
reassurance for the absence of appendicitis.
1,22,23
Fur-
thermore, two recent studies have provided results that
would indicate that a normal WBC count combined
with a normal CRP value would make the probability of
appendicitis low.
24,25
Further research is needed to
determine whether a combination of the biomarkers
analyzed in this study provide marginal benet over
individual biomarkers for the purpose of diagnosing
appendicitis when correlated to a patients symptoms
of abdominal pain.
Similar to our ndings, several prior publications
have described improved test performance characteris-
tics of CRP in those with several days of symptoms.
24,26
In addition, prior researchers have noted the utility of
serial CRP measurements with acute abdominal pain,
describing that an increasing CRP level over time was
concerning for appendicitis.
15,17
Although previous
authors have noted that the dynamic nature of CRP lev-
els may make serial measurement useful, it may be clin-
ically more feasible to associate a biomarker with
duration of symptoms, as serial blood draws can be
problematic in the ED setting and in children.
17,26
We
agree with previous authors that an elevation in CRP
(either serial measurements or as correlated to dura-
tions of symptoms) in children with acute abdominal
pain may aid in the diagnosis of appendicitis.
Our IL-6 ndings differ from prior studies in which
serum IL-6 levels were only elevated in children with
perforated appendicitis.
911,13
Paajanen et al.
10
mea-
sured cytokine levels in 80 patients (27 of them younger
than 20 years) undergoing surgery for appendectomy
and found that IL-6 levels correlated with increasing
severity of inammation seen on histopathology, but
that no signicant differences were noted in IL-6 levels
between patients with nonperforated appendicitis and
those without appendicitis. Similarly, Sack et al.
9
recently reported on 211 children with suspected
appendicitis (189 underwent appendectomies) and
found that IL-6 levels in patients without appendicitis
and with focal, nonperforated appendicitis by histopa-
thology were both low (median IL-6 levels 2.5 and 5.9
pg mL, respectively). Our results may have differed due
to the inclusion of a population of patients who mostly
did not have appendicitis (as opposed to higher rates of
appendicitis in previous studies), which provided us a
more accurate estimate of IL-6 levels in those without
appendicitis. Additionally, we were unable to determine
the timing of the biomarker assessments in the previous
IL-6 studies. IL-6 is up-regulated by mRNA transcrip-
tion (which takes 4 to 6 hours), and protein levels peak
by 24 hours following the development of a localized
inammatory process.
10,11,13
Therefore, diagnostic per-
formance may depend on when in the disease course
serum is collected.
Although IL-8 levels were elevated in children with
perforated appendicitis, serum levels do not distinguish
patients with nonperforated appendicitis from those
without appendicitis. We studied IL-8 as recent
research demonstrated that IL-8 gene expression was
highly elevated in appendicitis and that expression
correlated with the degree of inammation on pathol-
ogy.
6
Although our results were consistent with this
nding for those with perforation, it is not clear why
serum IL-8 levels were not elevated in patients with non-
perforated appendicitis. Similar to our ndings, Yoon
et al.
11
found that serum IL-8 levels were signicantly
elevated only in patients with perforated appendicitis.
LIMITATIONS
The current study has several limitations. First,
although a relatively large number of patients were
prospectively enrolled, there were a limited number of
patients with a longer duration of abdominal pain and
nonperforated appendicitis; this resulted in wide 95%
CIs for estimates of diagnostic accuracy. Second, the
estimates of biomarker accuracy were based on the
premise that physicians could reliably determine, and
parents and children could reliably report, duration of
pain. Although this is a potential source of variability,
there was good interobserver reliability between physi-
cians. Additionally, sequential serum samples were not
collected from patients; this mimics the routine ED set-
ting where only a single sample would typically be
obtained. It should be noted that although we had a
very high follow-up rate, two patients were lost to tele-
phone follow-up. We cannot exclude that these patients
sought care at an alternative medical facility. Finally,
the human IL-6 and IL-8 assay kits are presently
intended for laboratory investigation, take many hours
to complete, and are not FDA-approved for clinical use.
Further work must be done to conrm biomarker utility
and to develop assays to provide biomarker results in a
clinically useful time frame. Nevertheless, this study
provides important baseline information to guide future
research on novel biomarkers.
CONCLUSIONS
Although interleuken-6, the white blood cell count, the
absolute neutrophil count, and C-reactive protein are
all increased in patients with appendicitis, levels of
these markers uctuate over the course of illness.
Serum interleuken-6 is a potentially useful novel bio-
marker for patients with suspected appendicitis. Dura-
tion of symptoms may be an important variable to
consider when interpreting laboratory values in
patients with acute abdominal pain.
References
1. Bundy DG, Byerley JS, Liles EA, Perrin EM, Katz-
nelson J, Rice HE. Does this child have appendici-
tis? JAMA. 2007; 298:43851.
2. Flum DR, Morris A, Koepsell T, Dellinger EP. Has
misdiagnosis of appendicitis decreased over time?
A population-based analysis. JAMA. 2001; 286:
174853.
3. Brenner D, Elliston C, Hall E, Berdon W. Esti-
mated risks of radiation-induced fatal cancer from
pediatric CT. AJR Am J Roentgenol. 2001; 176:289
96.
ACADEMIC EMERGENCY MEDICINE
June 2011, Vol. 18, No. 6

www.aemj.org 573
4. Smith-Bindman R, Lipson J, Marcus R, et al. Radia-
tion dose associated with common computed
tomography examinations and the associated life-
time attributable risk of cancer. Arch Intern Med.
2009; 169:207886.
5. Kentsis A, Lin YY, Kurek K, et al. Discovery and val-
idation of urine markers of acute pediatric appendi-
citis using high-accuracy mass spectrometry. Ann
Emerg Med. 2010; 55:6270.
6. Murphy CG, Glickman JN, Tomczak K, et al. Acute
appendicitis is characterized by a uniform and
highly selective pattern of inammatory gene
expression. Mucosal Immunol. 2008; 1:297308.
7. Groselj-Grenc M, Repse S, Dolenc-Strazar Z, Hojker
S, Derganc M. Interleukin-6 and lipopolysaccharide-
binding protein in acute appendicitis in children.
Scand J Clin Lab Invest. 2007; 67:197206.
8. Rivera-Chavez FA, Peters-Hybki DL, Barber RC,
et al. Innate immunity genes inuence the severity
of acute appendicitis. Ann Surg. 2004; 240:269
277.
9. Sack U, Biereder B, Elouahidi T, Bauer K, Keller T,
Trobs RB. Diagnostic value of blood inammatory
markers for detection of acute appendicitis in chil-
dren. BMC Surg. 2006; 6:e15.
10. Paajanen H, Mansikka A, Laato M, Ristamaki R, Pul-
kki K, Kostiainen S. Novel serum inammatory
markers in acute appendicitis. Scand J Clin Lab
Invest. 2002; 62:57984.
11. Yoon DY, Chu J, Chandler C, Hiyama S, Thompson
JE, Hines OJ. Human cytokine levels in nonperfo-
rated versus perforated appendicitis: molecular
serum markers for extent of disease? Am Surg.
2002; 68:10337.
12. Gronroos JM, Gronroos P. Leucocyte count and C-
reactive protein in the diagnosis of acute appendici-
tis. Br J Surg. 1999; 86:5014.
13. Gurleyik G, Gurleyik E, Cetinkaya F, Unalmiser S.
Serum interleukin-6 measurement in the diagnosis
of acute appendicitis. ANZ J Surg. 2002; 72:6657.
14. Goodwin AT, Swift RI, Bartlett MJ, Fernando BS,
Chadwick SJ. Can serum interleukin-6 levels predict
the outcome of patients with right iliac fossa pain?
Ann R Coll Surg Engl. 1997; 79:1303.
15. Andersson RE, Hugander A, Ravn H, et al.
Repeated clinical and laboratory examinations in
patients with an equivocal diagnosis of appendicitis.
World J Surg. 2000; 24:47985.
16. Turkyilmaz Z, Sonmez K, Karabulut R, et al.
Sequential cytokine levels in the diagnosis of appen-
dicitis. Scand J Clin Lab Invest. 2006; 66:72331.
17. Wu HP, Huang CY, Chang YJ, Chou CC, Lin CY.
Use of changes over time in serum inammatory
parameters in patients with equivocal appendicitis.
Surgery. 2006; 139:78996.
18. Fluss R, Faraggi D, Reiser B. Estimation of the You-
den Index and its associated cutoff point. Biom J.
2005; 47:45872.
19. Bland JM, Altman DG. Statistical methods for
assessing agreement between two methods of clini-
cal measurement. Lancet. 1986; 1:30710.
20. Landis JR, Koch GG. The measurement of observer
agreement for categorical data. Biometrics. 1977;
33:15974.
21. Cicchetti D, Bronen R, Spencer S, et al. Rating
scales, scales of measurement, issues of reliability:
resolving some critical issues for clinicians and
researchers. J Nerv Ment Dis. 2006; 194:55764.
22. Kharbanda AB, Taylor GA, Fishman SJ, Bachur RG.
A clinical decision rule to identify children at low
risk for appendicitis. Pediatrics. 2005; 116:70916.
23. Andersson RE. Meta-analysis of the clinical and lab-
oratory diagnosis of appendicitis. Br J Surg. 2004;
91:2837.
24. Ortega-Deballon P, Ruiz de Adana-Belbel JC, Her-
nandez-Matias A, Garcia-Septiem J, Moreno-Azco-
ita M. Usefulness of laboratory data in the
management of right iliac fossa pain in adults. Dis
Colon Rectum. 2008; 51:10939.
25. Sengupta A, Bax G, Paterson-Brown S. White cell
count and C-reactive protein measurement in
patients with possible appendicitis. Ann R Coll Surg
Engl. 2009; 91:1135.
26. Wu HP, Fu YC. Application with repeated serum
biomarkers in pediatric appendicitis in clinical
surgery. Pediatr Surg Int. 2010; 26:1616.
574 Kharbanda et al.
TEMPORAL CHANGES IN BIOMARKERS FOR APPENDICITIS