Prospective Study of Clostridium Difficile in Children Undergoing Colonoscopy

Clostridium difficile (CD) is a gram positive, spore forming anaerobic bacillus that is not
a normal commensal of the gastrointestinal tract. It was initially named "difficult
clostridium" because it was difficult to isolate and grow on culture media. CD can exist in
spore and vegetative forms. Outside the colon, it survives in a spore form that is resistant
to heat, acid and antibiotics. In the colon, CD is found in the vegetative, toxin-producing
form and is susceptible to antibiotics. CD is widely distributed and infection is acquired
through the fecal-oral route from person to person contact, although the organism can be
cultured from toilets, bedpans, floors, telephones, shoes of hospital personnel and other
items. CD is most often identified as a disease acquired from medical facilities, but
community-acquired CD (CACD) is also described. CD infection is positively associated with
the use of antibiotics and gastric acid suppressing agents. CD infection is described in
patients who have inflammatory bowel disease and is significantly associated with their need
for hospitalization. Those who are hospitalized for inflammatory bowel disease and CD have
higher mortality and longer hospital stays than patients with inflammatory bowel disease but
no CD infection.

CD is a serious problem among adults. Less is known about CD in children. Although CD was
first identified in newborns, it is not considered a pathogen in children less than one year
of age. In older children CD causes a spectrum of disease similar to that seen in adults but
at a lower rate and with less severity. The epidemiology of CD in children has not been
studied extensively or in a systematic manner. However, intestinal colonization occurs in 5%
of all children older than 2 years of age. Recently, the investigators found colonic
aspirates positive for CD toxin from children at the time of their colonoscopies and in whom
CD was not suspected.

Our retrospective chart review of children undergoing colonoscopy from 9/1/2006 through
8/30/2008 identified 41 (17%) CD positive colonic aspirates. There was no difference between
CD positive and negative patients except abdominal pain and weight loss together occurred
significantly more often in the CD positive group. Significantly more CD positive patients
(p = 0.01) used an H2RA within one month of colonoscopy. Too few patient charts contained
information on antibiotic use to determine if a relationship existed with CD. Of the CD
positive patients 5 were lost to follow up, 36 were treated and all except 1 patient had
resolution of symptoms. The investigators concluded that CD infection, likely acquired in
the community, is common in children undergoing a colonoscopy and symptoms cannot be used to
discriminate which children will be positive for CD.

The investigators also cultured patient care sites at Women and Children's Hospital and
found no CD contamination of colonoscopes, medical areas, or supplies was identified.
Colonic aspirates were obtained from 73% of colonoscopies.

This retrospective chart review had several weaknesses. The most important weakness is that
the study design did not allow us to discriminate between intestinal colonization that
occurs in 5% of all children (15) older than 2 years of age or infection. The study did not
permit an assessment of whether treatment had a significant effect on the symptoms for which
the children presented because there was no control group. The investigators address these
weaknesses in this prospective study.

d. Experimental or Study Design: i. Brief: Double blind crossover study with a control arm.
ii. Detailed: Patients undergoing colonoscopy will be screened at time of initial
recruitment for exclusion criteria. At time of endoscopy a stool aspirate culture (routine)
will be collected. At time of endoscopy patients will be subjected to a second round of
exclusion criteria based on visual findings. Patients with CD (+) stool by aspirate will be
randomized into one of two treatment arms (early, delay). In addition, a third arm of CD (-)
age match controls will also be followed.

e. Sample Size i. A power calculation performed by Chang-Xing Ma PhD. based on an
improvement in symptoms of 1 point between groups on pain scale was performed. Recruiting 25
patients to each sample group and assuming a 30% drop out rate, final group sample sizes of
17 and 17 achieve 80.7% power to detect a difference of 1.0 between two group means with
estimated group standard deviations of 1.0 and 1.0 and with a significance level (alpha) of
0.05000 using a two-sided two-sample t-test.

ii. Based on the data from the retrospective review the investigators know that there is a
large amount of variability in this population and expect 1-2 patients enrolled per month.

f. Variables of Interest i. Subjective: nausea, bloating, ii. Objective: pain, stool
frequency, stool consistency, vomiting, weight loss, antibiotic use, use of acid suppression
iii. Biochemical: stool glutamate dehydrogenase, lactoferrin, CD culture, CD toxin

g. Completion i. Recruitment period: 18 months ii. Completion of therapy after recruitment:
1 month iii. Data analysis: 1 month iv. Total anticipated duration: 2 years

B. Characteristics of the Research Population

1. Gender: male/female

2. Age: age greater than 7 less than 21

3. Racial and Ethnic Origen: Any

4. Inclusion Criteria:

i. Any gender ii. Any race iii. Age between 7-21 iv. Greater then 21 kg v. Able to swallow
pills e. Exclusion Criteria (at time of screening): i. marked abdominal tenderness and
distention with minimal diarrhea ii. history of inflammatory bowel disease iii. fever iv.
systemic toxicity v. allergy to metronidazole vi. short bowel syndrome vii. history of bowel
resection viii. history of abdominal surgery (excluding appendectomy or cholecystectomy) ix.
history of hirschsprungs disease x. active pregnancy xi. inability to participate in phone
surveys xii. inability to maintain scheduled clinic follow up xiii. bloody stools at time of
screening f. Exclusion Criteria (at time of colonoscopy) i. Visual evidence of
pseudomembranes ii. Visible ulcerations iii. Mucosal friability iv. Mayo score greater than
4 v. Fissures/fistulas

g. Vulnerable Subjects- The impact of chronic abdominal pain is noted to have a high
prevalence, strong health care use, and considerable restrictions in the daily life of
children and Adolescents (16, 17). As a result of this disease burden it is prudent to study
this vulnerable population to determine if a biological trigger is present, and if present
how can this trigger be monitored and treated. To protect this population in this
investigation all patients participating under the age of 18 will be required to have both a
parental consent and a participant assent. Patients older than 18 years of age will be
required to have a documented consent. Participation in the project is voluntary and
permission to participate may be withdrawn at any time without risk of changing accepted
standard of care. All tests and procedures are non-invasive, and monetary reward for
completion of study is sufficient to compensate for time and travel costs so not to incur a
financial burden on the family, in addition the monies is not so great as to encourage
participation for the sake of monetary gain.

C. Methods and Procedures

a. Recruitment and enrollment methods i. Patients will be recruited from the Department of
Pediatrics, Digestive Disease and Nutrition Center. The patient population of interest does
not necessitate recruitment of patients outside of the normal scope of practice ii.
Enrollment will be voluntary. Eligible patients will be screened at time of clinic visit and
offered the potential opportunity to participate in the study pending the results of routine
screening.

iii. Patients who fulfill inclusion/exclusion criteria at the end of initial screening and
have self identified themselves as interested in participation will be given the opportunity
to participate in the study.

b. Privacy Protection i. See Section C part e and f.

c. Research Setting i. Women and Children's Hospital of Buffalo and allied clinics

d. Summary of Research Design i. See Section A part d.

e. Randomization i. Group assignment will be assigned at time of colonoscopy ii. Group
assignment will be in place for those patients assigned to a therapeutic intervention iii.
Random group assignment list by patient enrollment number was generated by Chang-Xang Ma,
PhD

f. HIPAA i. The patients "Informed Consent Form" (please see section 6 of application) will
be accompanied by a separate document incorporating Health Insurance Portability and
Accountability Act (HIPAA) compliant wording by which patients/parents/legal guardian
authorize the use and disclosure of their Protected Health information by the investigator
and by those persons who need that information for the purposes of the study.

The Written Informed Consent will explain that for data verification purposes, study
personnel, a regulatory authority, or an IRB may require direct access to parts of the
hospital or practice records relevant to the study, including patients' medical history.

g. Data Storage and Confidentiality i. All study data will be securely housed at the
Digestive Disease and Nutrition Center located at 219 Bryant Street ii. All electronic data
is securely housed in Cerner Powerchart monitored by Kaleida Health

h. Payments/compensation i. Patients will be compensated $60.00 at the completion of the
study

i. Data Analysis Plan i. Statistical analysis will be performed by a professional
statistician, Chang-Xang Ma, PhD, at the University at Buffalo, Buffalo, NY. Comparisons
treatment 1 and treatment 2 groups will be performed by SAS statistical software using chi
square, Fisher's exact tests, t-test and multivariate analysis as indicated.

D. Risks and Potential Benefits of Participation in the Research a. Subject Risk i. The risk
to subjects for the initial colonoscopy is no greater than the standard risk. The potential
risk attributable to home stool collection includes contamination of food, and spread of
fecal oral organisms if appropriate hand washing is not performed.

b. Subject potential benefit i. The potential benefit to patients is resolution of abdominal
pain symptoms

E. Informed Consent Process (Consent/Permission/Assent)

a. The principal investigator will ensure that the patient's parent/guardian is given full
and adequate oral and written information about the nature, purpose, possible risk and
benefit of the study. As all patients will be greater than assent age, assent will be
obtained for all children enrolled in the study. Patients and the patient's parent/guardian
will be given the opportunity to ask questions and allowed time to consider the information
provided.

The patient/guardian's signed and dated informed consent as well as the patient's assent
will be obtained before conducting any procedure or test specifically for the study.

Inclusion Criteria:

- Aspirate at colonoscopy positive for C. diff.

- Any gender

- Any race

- Age between 7-21

- Greater than 21 kg

- Able to swallow pills

Exclusion Criteria:

- Abdominal tenderness and distention with minimal diarrhea

- History of inflammatory bowel disease

- Fever

- Systemic toxicity or allergy to metronidazole

- Short bowel syndrome

- History of bowel resection vii

- History of abdominal surgery (excluding appendectomy or cholecystectomy)

- History of hirschsprungs disease x

- Pregnant

- Inability to participate in phone surveys

- Inability to maintain scheduled clinic follow up

At time of Colonoscopy:

- Visual evidence of pseudomembranes

- Visible ulcerations

- Mucosal friability iv.

- Fissures/fistulas