ASSOCIATION FOR ACADEMIC SURGERY
Neutropenic Colitis in Children
Frankie B. Fike, M.D., Vincent Mortellaro, M.D., David Juang, M.D., Shawn D. St. Peter, M.D.,
Walter S. Andrews, M.D., and Charles L. Snyder, M.D.1
Department of Surgery, Children’s Mercy Hospital, Kansas City, Missouri
Submitted for publication December 21, 2010
Background. Neutropenic colitis (NC) or typhlitis
has an incidence of approximately 5% in patients re-
ceiving chemotherapy for malignancy. The precise
cause is unknown, but almost all patients are neutro-
with neutropenic colitis over the past 10 y to assess the
incidence, management, and outcome.
Materials and Methods. Hospital records were re-
viewed after obtaining IRB approval (# 10-10-184E).
There were 1224 children treated for cancer at our in-
stitution over the study interval. Neutropenic colitis
was strictly defined as the presence of both clinical
and radiographic findings consistent with the diagno-
tients with confirmed appendicitis were excluded.
Medical management consisted of bowel rest and/or
decompression, broad-spectrum IV antibiotics and
anti-fungal coverage, and serial clinical exams and ra-
diographic studies. Clinical deterioration and free air
were the primary indications for surgery. Demo-
graphic factors, signs, symptoms, clinical presenta-
tion, underlying disease process, white blood cell
count (WBC), and absolute neutrophil count (ANC)
prior to the onset of the disease, treatment, and out-
come were analyzed.
Results. Neutropenic colitis was confirmed in 17
children (1.4% incidence). Three patients had more
than one episode. Leukemia was the most common un-
inance, and the mean age at diagnosis was 8.4 y. Mean
WBC and ANC at onset were 670 and 164, respectively.
A sharp decline in the ANC usually immediately pre-
ceded the onset of NC; 2/17 (12%) underwent operation,
butboth patientshadonlypneumatosis atexploration.
Overall survival was 9/17 (53%), but no deaths were di-
rectly attributable to the colitis.
Conclusion. Neutropenic colitis is an uncommon oc-
currence in children with neoplasia (1.4% in the cur-
rent study). Leukemia is the most common cause. A
precipitous decline in ANC usually occurs prior to
the onset of NC. Mostpatients do not require operation
and the overall mortality directly attributable to NC is
? 2011 Elsevier Inc. All rights reserved.
Key Words: typhlitis; neutropenia; colitis; immuno-
Neutropenic colitis or typhlitis has an overall inci-
dence ranging from less than 1% to as high as 26%, al-
chemotherapy for malignancy . The criteria for the
diagnosis are unclear, and most studies are retrospec-
tive case reviews, which include a heterogenous group
of adult and pediatric patients with a variety of diagno-
ses (although most cases are secondary to leukemia).
Some studies include infectious colitis (C. difficle and
others), while some reviews exclude these patients.
The precise cause is unknown, but almost all patients
are profoundly neutropenic. The symptoms are vague
(fever, nausea and vomiting, abdominal discomfort/dis-
tention) and are often present in unaffected patients in
the clinician to establish the diagnosis and select the
ill and expire after NC is identified, yet it is not the di-
rect cause of death. We reviewed our experience with
neutropenic colitis in children at a single institution
over the past 10 y to assess the incidence, presentation,
management, and outcome.
After obtaining IRB approval (no. 10-10-184E) we reviewed the
charts and medical records of all children with a confirmed (clinical
1To whom correspondence and reprint requests should be ad-
dressed at Department of Surgery, Children’s Mercy Hospital, 2401
Gillham Road, Kansas City, MO 64108. E-mail: firstname.lastname@example.org.
? 2011 Elsevier Inc. All rights reserved.
Journal of Surgical Research 170, 73–76 (2011)
and radiologic) diagnosis of NC over a 10-y interval (January 2000 to
March 2010). Only children with both symptoms and radiologic find-
ings consistent with the diagnosis were included. Patients with ap-
pendicitis were excluded.
Demographic parameters, symptoms, and signs of typhlitis, under-
lying disease process, age at onset, radiographic results, trends in the
white blood cell, and absolute neutrophil count (ANC) in the 2 mo
prior to the diagnosis and 1 mo afterwards, management, and out-
come were reviewed. Temporal changes in ANCs were calculated
There were 17 patients with the definitive diagnosis
of NC during the study interval. There were 1224 pa-
tients treated for malignancies, for an incidence of
1.4%. During the study interval, 465 patients were
treated for leukemia (2% developed NC) and 137 for
Three patients had two episodes at different times,
for a total of 20 episodes of NC. There were eight males
and nine females. The mean age at diagnosis was 8.4 y,
with a range of 1.8 to 18.3 y. The three patients who de-
veloped recurrent NC did so at 19,4.4, and 1 mo after
the primary episode.
The underlying malignancies were: AML (acute mye-
logenous leukemia) in five children, acute lymphocytic
leukemia (ALL) in four, neuroblastoma in three pa-
tients, and isolated cases of Burkitt’s lymphoma, atyp-
ical teratoid rhabdoid tumor of the brain, osteogenic
sarcoma, retroperitoneal desmoplastic round cell tu-
mor, and hepatocellular carcinoma. Two of the cases
motherapeutic protocols had been used: no clear pat-
tern could be discerned in this small series.
The most common presenting symptoms were ab-
dominal pain (88%, 15/17), fever (41%, 7/17), diarrhea
and vomiting (18%, 3/17), and GI bleeding (6%, 1/17).
Mucositis was documented two patients (12%). Mean
white blood cell count (WBC) at diagnosis was 670,
and mean ANC (absolute neutrophil count) was 164.
Only 12% (2/17) patients were not neutropenic at the
time of diagnosis. All patients had bowel wall thicken-
ing in the cecum or ascending colon or pneumatosis
identified on CT or US.
Stool cultures and C. difficle (toxin/cultures) were ob-
tained in almost all patients and were uniformly nega-
tive. A single patient did have C. difficle enterocolitis,
diologic studies confirming the diagnosis.
The ANC counts were evaluated for the 2 mo prior to
the onset of NC, and for 1 mo afterwards. The ANC
counts for each patient were plotted over the 3-mo in-
terval surrounding their diagnosis of NC, and are
shown in Figure 1. Limited values were available for
two patients. All patients except two were receiving/
had received granulocyte colony-stimulating factor
Nine of the 17 patients were long-term survivors, but
no deaths were clearly causally related to the NC.
Twelve percent (2/17) underwent operation, both for
free air with negative findings at exploratory laparot-
omy (except pneumatosis).
The true incidence of neutropenic enterocolitis is un-
known; a review of 145 pooled adult studies found that
(Table 1); a busy pediatric cancer center reported 24
cases over a 30-year period . Another report identi-
fied 12 children over a 10 year interval . The autopsy
incidence of NC in children in with hematologic malig-
nancies is high - 24% of children with acute leukemia
had postmortem evidence of typhlitis in one series ;
this may be due to the obvious selection bias in autopsy
The cause of NC in patients with malignancies re-
mains unclear. Direct cytotoxicity to the intestinal mu-
cosal cells, overgrowth of specific bacterial flora,
generalized immunodeficiency with decreased resis-
tance even to normal intestinal bacteria, production of
bacterial endotoxins, impaired microvascular blood
supply to the gut lining (with the larger cecum being
the most susceptible region), and many other possible
causative agents have been postulated . Most cases
involve the cecum, but the ileum may also be involved,
as well as the ascending colon . The postulated se-
quence of events is inflammation -> edema -> ulcera-
tion -> necrosis -> perforation. The most common
underlying disease in children is usually leukemia .
In one series of 42 children, 76% had underlying hema-
tologic malignancies: Burkitt’s lymphoma (15%) and
acute myeloblastic leukemia (12%) predominated .
One multivariate analysis found factors significantly
associated with the development of typhlitis were mu-
cositis (OR 30.7), stem cell transplantation (OR 58.9),
and chemotherapy within the prior 2 wk (OR 12.9) .
Age greater than 16 y at the time of cancer diagnosis
was a risk factor for developing NC in another large se-
Profound neutropenia is almost always present.
However, 12% of 83 patients in one pediatric series
were not neutropenic. We found that almost all of our
patients (88%) were neutropenic. Some have suggested
opment of NC; in our series, it appeared that the rate of
change of the ANC was significant. Figure 1 demon-
strates a rapid decline in ANC prior to the onset of
JOURNAL OF SURGICAL RESEARCH: VOL. 170, NO. 1, SEPTEMBER 2011
agnosis of NC, and 1 mo afterwards. Sharp declines in the ANC were not universally seen, but did occur in most children just prior to the
diagnosis of NC. (Color version of figure is available online.)
Graphs of absolute neutrophil count over time for 15 patients with NC are shown. Data shown are for the 2 mo prior to the di-
FIKE ET AL.: NEUTROPENIC COLITIS IN CHILDREN
NCinmostofourpatients.Obviously,arapid declinein Download full-text
ANC may occur without the development of NC, but in
patients with NC, most have experienced a recent pre-
cipitous drop in ANC.
The most common presenting symptoms are usually
abdominal pain (91%), fever (84%), abdominal tender-
ness (82%), and diarrhea (72%) . However, the onset
is often subtle, and CT scan is generally the gold stan-
dard for the diagnosis. A high index of suspicion must
be maintained in order to avoid missing the diagnosis.
Clinical examination and plain films were comparable
to CT scan for diagnostic accuracy in one report .
The degree of bowel wall thickness on US and the dura-
tion of neutropenia, fever, and abdominal discomfort
wereassociated withalongerduration ofNC symptoms
in one study of 83 children .
Appendicitis can be difficult to distinguish clinically
from NC, although CT scans can be very helpful in
this regard. Five of seven children with hematologic
malignancy had an atypical appendicitis presentation
; only fever and diarrhea distinguished the typhlitis
patients in one review.
decompression, broad-spectrum IV antibiotics (includ-
ing anti-fungal agents), and sometimes G-CSF. Pa-
tients with perforation, obstruction, peritonitis, active
GI bleeding, or clinical deterioration may require oper-
ation; usually colectomy and diversion. Most children
do not require operation (Table 1) [6, 9, 3, 10]. This
was certainly the case in our series, where only 2/17 un-
derwent operation. Both patients were found to have
pneumatosis but viable bowel, and neither diversion
nor resection was performed.
Mortality rates currently range from 30% to 50% in
adults . Mortality is markedly lower in children -
approximately 5% in compiled series (Table 1) [6, 9,
3, 10] In our series, 2 children died shortly after the de-
velopment of NC, but in both cases, the death was not
directly attributable to the NC.
Neutropenic colitis is relatively uncommon in chil-
dren with hematologic malignancies. Laparotomy is
rarely necessary, and mortality directly from the colitis
is low. A precipitous drop in the ANC should raise the
level of clinical vigilance/suspicion, particularly if other
signs and symptoms are present.
1. Gorschl€ uter M, Mey U, Strehl J, et al. Neutropenic enterocolitis
in adults:Systematic analysis of evidencequality.EurJ Haema-
Sloas MM, Flynn PM, Kaste SC, et al. Typhlitis in children with
cancer: A 30-year experience. Clin Infect Dis 1993;17:484.
Schlatter M, Snyder K, Freyer D. Successful nonoperative man-
agement of typhlitis in pediatric oncology patients. J Pediatr
perience and postmortem review. Cancer 1990;65:1041.
Davila ML. Neutropenic enterocolitis. Curr Opin Gastroenterol
Moran H, Yaniv I, Ashkenazi S, et al. Risk factors for typhlitis in
pediatric patients with cancer. J Pediatr Hematol Oncol 2009;
McCarville MB, Adelman CS, Li C, et al. Typhlitis in childhood
cancer. Cancer 2005;104:380.
Hobson MJ, Carney DE, Molik KA, et al. Appendicitis in child-
hood hematologic malignancies: Analysis and comparison with
typhlitis. J Pediatr Surg 2005;40:214. discussion 219.
Otaibi AA, Barker C, Anderson R, et al. Neutropenic enterocoli-
tis (typhlitis) after pediatric bone marrow transplant. J Pediatr
and outcomes of suspected typhlitis in oncology patients—
experience in a tertiary pediatric surgical center in the United
Kingdom. J Pediatr Surg 2009;44:381.
colitis. Surg Infect (Larchmt) 2009;10:307.
Neutropenic Colitis in Pediatric Series
Author YearNo. casesIncidence (%)Operation (%)Mortality (%)
JOURNAL OF SURGICAL RESEARCH: VOL. 170, NO. 1, SEPTEMBER 2011