Stratifying Risk in Barrett*s Esophagus: A Pilot Study for Biomarker-Based Patient Management

Barrett*s esophagus (BE) is the precursor lesion of esophageal adenocarcinoma
(EAC), a cancer with a >500% increase in incidence in the U.S. since the
1970*s. Effective prevention of EAC is hindered by poor risk stratification.
Our only commonly used prognostic factor, dysplasia, is both poorly
reproducible and inaccurate. Recently developed biomarker panels may allow for
accurate risk stratification, such that subjects at higher risk may be
identified early for endoscopic intervention.

EAC is increasing rapidly in incidence, and is associated with an abysmal
5-year survival rate of less than 15%. Early diagnosis has been shown to
improve outcomes and should be feasible in view of the presence of the
premalignant condition BE. However, because of the low conversion rate of BE to
cancer, as well as inherent difficulties with sampling bias and histological
assessment of specimens, the benefits of surveillance are controversial. On the
basis of cost-effectiveness analyses, it has become apparent that only patients
with a high risk of developing cancer should be targeted for further
intervention, such as RFA.1,2 Our inability to accurately stratify BE patients
on the basis of histology has precluded this more elegant approach to patient
management and as such, all BE patients, including those with low risk of
progression, remain under surveillance. Current clinical practice for patients
diagnosed with BE involves regular endoscopy and multiple random biopsies for
the assessment of dysplasia.3 Unless there is an obvious abnormality (ulcer or
nodule), then the biopsies may not sample the dysplastic area, leading to
under-staging of the degree of dysplasia. Furthermore, dysplasia is a highly
subjective pathological diagnosis which may miscategorize patients.4,5 The
management algorithm hinges on the diagnosis of dysplasia, such that
individuals with no dysplasia often undergo continued surveillance whereas
those with HGD or intramucosal carcinoma (IMC) are offered a therapeutic
intervention with either endoscopic therapy or surgery depending on local
expertise, patient fitness, and depth of invasion. Subjects with LGD may
sometimes undergo endoscopic therapy, or may be retained in an intensified
surveillance program, depending on patient and physician consultation. The
results of this algorithm are that many patients at very low risk are subjected
to repeated endoscopy with associated risks and psychological stress, as well
as financial costs. On the other hand, patients at high risk may be missed due
to sampling bias, misclassification of dysplasia, or lack of surveillance due
to poor compliance or inavailability of services. Our current system is thus
fraught with errors and excess expense.

Determining which patients with BE have the greatest risk for progression to
EAC has the potential to improve outcomes and to deliver more cost-effective
care. During a retrospective validation study, Investigators have identified
and validated a biomarker panel that will be used for this study to identify
subjects at high risk of developing EAC. This is a pilot study to gather
information on the feasibility of a prospective, international validation
study, which would randomize subjects with BE and no dysplasia or low-grade
dysplasia who are recognized to be at high risk for progression to either an
active intervention (RFA) or to endoscopic surveillance. Subjects deemed at low
risk by the biomarker panel would be retained in surveillance. Comparisons
would include progression outcomes in the high risk surveillance arm to the low
risk surveillance arm (to assess the accuracy of the panel at predicting risk)
and progression outcomes of the high risk intervention arm (RFA) to the high
risk surveillance arm (to assess if active intervention in appropriately
selected individuals attenuates the risk of progression to EAC). Such an
approach heralds a new paradigm, in which patients are accurately managed
according to their future risk of cancer on the basis of objective, easily
assayed markers. If successful, we will be able to focus resources on those at
greatest risk for cancer, thus sparing low risk patients unnecessary invasive
and expensive interventions, while reducing the progression to cancer in those
at high risk. Technological advances in biomarker research, as well as safe and
effective endoscopic therapeutic modalities (endoscopic mucosal resection and
RFA) make the time ripe for such a study.

Primary objective:
To gather information on the feasibility of a prospective, international
multicenter study in which, on the basis of biomarker-based risk
stratification, patients at high risk for progression are randomized to
radiofrequency ablation (RFA) vs. a surveillance arm (yearly EGD), while
patients in the low risk group have an attenuated EGD surveillance regimen per
current standard of care.

Secondary objective:
To develop the protocols, operating procedures, and other materials necessary
for application to the National Cancer Institutes for support of the
multi-center trial of ablative therapy in BE.

The Investigators have previously completed a validation study in which they
have identified a panel of biomarkers that can predict progression of subjects
with BE and no dysplasia or low-grade dysplasia to esophageal adenocarcinoma
(EAC). The biomarker panel, performed by the Fitzgerald lab, included ploidy
(by image cytometry), AOL (histochemistry (IHC)), p53 (IHC), cyclin A (IHC),
and dysplasia. These markers have been validated and demonstrated to be highly
predictive of both progression to EAC, as well as the presence of occult
malignancy elsewhere in the specimen (field effect). The final panel of
validated biomarkers will be used in this study to identify patients at high
risk of developing EAC.

The study will recruit 100 patients across 5 sites (University of North
Carolina, UHCMC, Cambridge, Medisch Centrum Alkmaar and Amsterdam). The
specific aims of this pilot study are to:
1) Demonstrate that the international, multicenter team can work together,
2) Define the logistics of assaying biomarkers in real time such that in the
future interventional trial, results could influence clinical decision-making,
and,
3) Provide further data to inform a power calculation for the full trial.

Subjects enrolled in the study will complete a questionnaire gathering
hypothetical willingness to be randomized to receive endoscopic treatment
intervention (RFA) or surveillance endoscopy. Subjects enrolled in the Dutch
centers will also complete a quality of life questionnaire, focusing on general
health related quality of life, as well as health perception and fear of
progression, including:
- The SF-36: measures generic quality of life. Its 36 items form eight
subscales: physical functioning, social functioning, physical role-functioning,
emotional role-functioning, vitality, bodily pain, mental health and general
health.
- The EORTC-QLQ-C30: is a cancer specific questionnaire with 30 items,
evaluating five functional scales (physical functioning, emotional functioning,
cognitive functioning, social functioning and role functioning); three general
cancer symptom scales (fatigue, pain, nausea and vomiting) and five separate
cancer symptoms (shortness of breath, insomnia, appetite loss, constipation,
and diarrhea).
- The EORTC-QLQ-OES18: measures esophageal cancer-specific quality of life.
This 18 item questionnaire measures 10 esophageal cancer specific symptoms:
dysphagia, reflux, pain, eating problems, trouble speaking, coughing, trouble
swallowing, choking, dry mouth and taste.18
- The Hospital Anxiety and Depression Scale: measures general anxiety and
depression in 14 items.
- The Fear of Progression questionnaire Short Form: measures fear of cancer
progression in 12 items.
- The Brief Illness Perception Questionnaire (Brief IPQ), a nine-item scale
designed to rapidly assess the cognitive and emotional representations of
illness.

Biopsy samples will be obtained from all subjects and tested for all biomarkers
in the panel. Results of the biomarker panel will not be communicated to
sites. Subjects with low grade dysplasia will be offered the option of
receiving radiofrequency ablation (RFA) as part of routine care. Subjects with
low grade dysplasia who agree to RFA will receive RFA as part of routine care.
Subjects with low grade dysplasia who do not want to receive RFA will receive a
surveillance endoscopy in 1 year as part of routine care. Subjects with no
dysplasia will undergo regular surveillance endoscopy in 3-5 years as per
routine clinical care.

The goals of this pilot study are to ascertain the proportion of subjects in
the high risk arm, to demonstrate the plausibility of performing the biomarker
analysis efficiently in a sizable group of patients, to demonstrate the
feasibility of delivering the endoscopic intervention (RFA), to obtain 1 year
pilot data regarding progression in the high risk arm for use in sample size
calculations, and to document collaboration among the centers.

Patients with Barrett*s esophagus undergo regular endoscopic surveillance with
biopsies to assess the presence of dysplasia. This surveillance is performed as
per routine clinical care, recommended by current international guidelines.
Every patient that undergoes a surveillance endoscopy with biopsies stands a
minute risk of bleeding as a result of these biopsies. For this study,
participating patients will only undergo esophagogastroduodenoscopies as per
routine clinical care. During endoscopies performed for this study, however,
additional biopsies will be obtained from the Barrett*s segment (for every 2
centimeters of the Barrett*s segment 2 additional biopsies will be obtained).
However, these additional biopsies will not significantly increase the bleeding
risk for participating patients, as compared to the standard amount of
surveillance biopsies that would have been obtained.
The endoscopic procedure might be 10-15 minutes longer in duration compared
with a regular surveillance endoscopy.
Patients will be administered a small questionnaire, consisting of a single
question. This questionnaire will address the hypothetical question if patients
would be willing to participate in a study that randomizes patients to
prophylactic treatment or regular endoscopic surveillance. Additionally,
patients will be administered a quality of life questionnaire, measuring
general health related quality of life, as well as health perception and fear
of malignant progression. Completing this quality of life questionnaire will
take no more than 15 minutes and this questionnaire does not include questions
that will threaten the personal integrity of the patient.

Taking into account the minor burden and (additional) risks for patients
associated with participation in this study, we find that execution of this
research study is justified.