The Breast Cancer to Bone (B2B) Metastases Research Program: a multi-disciplinary investigation of bone metastases from breast cancer

Study design

Population-based ascertainment

The population-based ascertainment of breast cancer patients for the B2B Research
Program was developed in partnership with the Alberta Cancer Research Biobank (ACRB)
and the Alberta Cancer Registry (ACR). The ACR is a province-wide cancer registry
that has been awarded a Gold Certification from the North American Association of
Central Cancer Registries since 1999, indicating the highest quality of completeness,
accuracy, and timeliness of cancer reporting. Prior to the establishment of the B2B
Research Program, the ACRB focused predominantly on the collection of fresh-frozen
tumour tissue from breast cancer patients in addition to a limited collection of largely
post-surgical blood samples. The need to recruit a population-based cohort and collect
pre-surgical blood samples, led to the development of the Comprehensive Biospecimen
Rapid Ascertainment (CoBRA) system (Fig. 2). The CoBRA system is responsible for the ascertainment of patients and their recruitment
into the ACRB. The ACRB is approved by both the institutional and provincial research
ethics boards (HREBA and CHREB, respectively).

Fig. 2. B2B Research Program timeline. Recruitment for the B2B Cohort began in 2010, and steadily
increased through successive operational enhancements, key partnerships, and implementation
of a centralized biospecimen ascertainment infrastructure

The CoBRA system was designed to identify all newly diagnosed breast cancer patients,
within the Calgary area, through seven multiply-redundant mechanisms. The primary
mechanism for patient identification is the pathology reports pertaining to their
diagnostic (pre-surgical) fine needle and core biopsy. All biopsy reports are submitted
to the ACR; a copy of the report is also submitted to dedicated ACRB personnel who
are designated as affiliates of the ACR and bound by the code of conduct and training
required by all ACR personnel. The additional six supplementary sources of patient
ascertainment are outlined in Table 1. Each new patient is recorded in the CoBRA database and all correspondence and patient
contacts, pertaining to their informed consent and biologic sample collection for
the ACRB, are managed within this system. Potential donors to the ACRB are contacted
only after their awareness of their diagnosis has been confirmed. Written informed
consent is sought from each patient to donate a pre-surgery (or pre-neoadjuvant therapy,
if applicable) blood sample and tumour tissue at surgery if sufficient tissue is available
without compromising pathologic assessment or future clinical care. The informed consent
process for the ACRB consists of a Consent Information Brochure and a separate consent
form on which patients select to participate or not and indicate their willingness
to be contacted regarding future research.

Table 1. Seven patient ascertainment and recruitment strategies to facilitate comprehensive
population-based biospecimen accrual and the potential for differential patient selection
associated with each individual approach

The comprehensive population-based ascertainment of breast cancer patients commenced
in February 2010. All potentially eligible participants for the B2B Cohort are selected
from patients who agreed to participate in ACRB. The CoBRA procedures will continue
to ascertain and recruit patients beyond the time frame of the B2B Research Program,
to support additional research projects and biospecimen requests.

Study population

Patients with incident primary breast cancer are eligible for recruitment into the
B2B Cohort if they meet the following criteria as determined through the CoBRA database:
(1) histologically-confirmed stage I to stage IIIc breast cancer diagnosed between
2010 and 2015; (2) residents of Calgary, Alberta and the surrounding areas; (3) females
?18 and ?80 years at initial diagnosis; (4) able to provide informed, written consent
and complete questionnaires and an in-person interview in English; and, (5) no previous
cancer diagnosis with the exception of cervical in-situ neoplasia (CIN) and non-melanoma
skin cancer. Cohort participants must have donated a pre-surgical blood sample to
the ACRB (Fig. 3) and indicated willingness to be contacted for future research. The contact details
of eligible patients are then exported to the B2B Research Program database for recruitment
into the B2B Cohort. The B2B Cohort is restricted to the Calgary area because of the
feasibility of processing blood samples collected from community laboratories within
24 h.

Fig. 3. Ascertainment recruitment, data and biospecimen collection and sharing scheme. Newly
diagnosed cancer patients are identified through the ACR, and are invited to donate
biospecimen samples by the ACRB utilizing the CoBRA infrastructure. Clinical data
and biospecimens are stored by the ACRB, and contact information from eligible and
consenting participants is sent to study coordinators of relevant research programs.
Subsequent blood samples and blood questionnaire data for routine study follow-up
are collected by the ACRB and act as a shared resource between the biorepository and
research study team

Recruitment

The recruitment of patients into the B2B Cohort was harmonized with recruitment for
the Alberta Moving beyond Breast cancER (AMBER) study 92] because the potential participants are drawn from the same population of breast cancer
patients (Fig. 4). If a patient agrees to be contacted regarding future research, contact details
for all eligible patients are imported into the AMBER B2B Recruitment Database. Patients are first invited to participate in the AMBER study so that their exercise
assessments can be completed prior to the delivery of systemic therapy 92].

Fig. 4. B2B/AMBER participant recruitment. Eligible patients are identified by the ACRB through
CoBRA processes, and the contact details of consenting biospecimen donors are sent
to a recruitment database shared by both the AMBER and B2B study coordinators. Patients
are invited to participate by each study, and if they agree, their information is
then imported into the specific study database. Some information sharing occurs between
the AMBER and B2B study database, such as whether or not shared questionnaires have
been completed

Patients are invited to participate in the B2B Cohort approximately 6–8 weeks post-surgery,
and after potential recruitment into the AMBER study. The B2B Research Program Study
Coordinator contacts eligible women by telephone to explain the research program.
If the potential participant verbally agrees to receive the recruitment package, their
information is imported into the B2B Tracking Database and they are mailed a letter of invitation, consent information brochure, consent
form and Pre Interview Worksheets.

Consenting participants complete baseline worksheets, an in-person interview and post-interview
questionnaires within six months of an initial breast cancer diagnosis, with follow-up
assessment occurring at 24, 48 and 72-month intervals post-diagnosis. Passive follow-up,
through chart abstraction of medical records, will occur at 10 years following the
completion of the active follow-up or for those who were lost to follow-up but did
not withdraw consent. We anticipate that the recruitment of the baseline B2B Cohort
will be completed by August 2015; recruitment of the baseline cohort of 600 participants
will have taken a total of 5.5 years. Analysis of data and biospecimens will commence
shortly afterwards.

Sample size

As a core infrastructure resource, the B2B Cohort sample size was not explicitly based
on the power to address a single hypothesis. However, to address our primary, outcome–based
hypotheses embedded within the core projects, we will follow all cohort members (~600
patients) for a median of five years. Less than 60 % of breast cancer recurrences
are apparent within three years of follow-up but ~80 % are evident after five years
of follow-up 3], 9], 10]. Within the 10-year time frame of the B2B Research Program, we expect 70-80 women
to present with clinically evident bone metastases 3], 9], 10]. Using serum vitamin D concentrations as an example a specific hypothesis to be tested,
broad inter-quintile ranges of serum [25-OHD] that are typical within North American
populations (Q1??14.9 ng/ml, Q5??35.3 ng/ml 93]). Therefore, we anticipate that, for the vitamin D and inflammation analyses, a 20
% difference in vitamin D exposures and inflammatory status between women with and
without bone metastases will provide 80 % power to detect a relative risk for bone
metastases of 1.8 even with this modest sample size. We anticipate that similar effect
magnitudes will be observed in the other core projects.

Data collection instruments

Pre-interview worksheets

Each participant is mailed pre-interview questionnaires including the Sun Exposure Worksheets and Past Year Dietary Worksheets as part of their recruitment package. Participants are asked to complete these questionnaires
and return them by mail. The Sun Exposure Worksheets collect information on residence, work history and vacation history for three time
periods: the 12 months prior to breast cancer diagnosis; the calendar year five years
prior to breast cancer diagnosis; and the calendar year 10 years prior to breast cancer
diagnosis. Although eligible participants must be free of detectable distant metastases
at diagnosis, breast cancer cells can be disseminated early in tumour development.
Capturing exposure data for an extended pre-diagnostic period, and during follow-up,
ensures that exposure can be estimated for the entire period that a patient was at
risk of disease dissemination.

The Past Year Dietary Worksheets collect data regarding food intake and frequency in the 12 months prior to breast
cancer diagnosis, and was developed to assess intake of certain foods and supplements
that have high vitamin D and calcium content, consumed at a reasonable frequency by
female study participants in Alberta 94]. The sun exposure, dietary, and supplement data will be used to estimate levels of
vitamin D during the relevant exposure period.

In-person interview

The completed pre-interview questionnaires are scanned using TELEform®, an optical
character recognition software program, then verified by study staff for completeness
before being exported into the Blaise® computer-assisted interviewing draw 1software
program to pre-populate corresponding responses in the B2B Baseline Interview. Furthermore, if the participant has completed the AMBER Baseline Health Questionnaire (BHQ), those verified responses are also used to pre-populate corresponding responses
in the B2B Baseline Interview. By pre-populating the interview with information provided by the participants in
the pre-interview questionnaires and the AMBER BHQ, we reduce participant interview
burden and expedite the in-person interview process.

One of the B2B Interviewers conducts the in-person interview at a time and place convenient
for the participant. The B2B Baseline Interview is typically an hour in length, and collects information regarding pregnancy and
menstruation, menopausal status, hormone replacement therapy, birth control and hormone
contraceptive use, personal health history/co-morbidity, medications (over-the-counter
and prescription), vitamins, minerals and herbal supplements, mobility and physical
activity, sun exposure, diet history, family history of cancer, smoking habits, alcohol
consumption history, and demographic information. Following the in-person interview,
the participant is provided with a Canadian Diet History Questionnaire II (DHQ II) and Past Year Total Physical Activity Questionnaire (PAQ) 95], which is to be completed and returned to the study office by mail. If a B2B participant
has already completed the DHQII and PAQ as part of the AMBER study, these responses are made available to the B2B Research
Program to further reduce participant burden.

Physical activity questionnaire

The PAQ is administered at four time points throughout the study: baseline (post-interview),
24, 48 and 72-month follow-up. The PAQ is a self-administered questionnaire in which participants report their occupational,
transportation, household and recreational/leisure physical activities over the previous
12 months. Participants report the number of hours spent in each activity per week,
allowing for analysis of each individual type of activity as well as a summation of
all four categories of activities to determine the participant’s total amount of physical
activity over the past year 95]. These measures are expressed as metabolic equivalents for each activity and are
reported in total MET-hours/week/year of activity 95].

Diet history questionnaire

The DHQ II96] is also administered at four time-points during the study at baseline (post-interview),
24, 48 and 72-month follow-up. It is a self-administered food frequency questionnaire
developed initially by the National Institute of Health and then adapted for use in
the Canadian population 96]. This FFQ is a comprehensive assessment of dietary intake in the previous 12 months
that has 164 questions about 134 food items and includes seasonal intake of a variety
of foods, the portion size and frequency of intake for each food item. Responses from
the DHQII provide comprehensive information on dietary habits, output of nutrients and the
amount foods and food groups consumed. Additionally, the dose and frequency of vitamin
and mineral supplementation over the past year is also obtained.

Biospecimen collection

Each participant’s baseline blood sample is collected as part of their ascertainment
and upstream recruitment into the ACRB according to the CoBRA procedures. Participants
receive a blood requisition form and are asked to donate a blood sample at any Calgary
Laboratory Services location. The baseline collection consists of a 60 ml of blood
sample collected in six 6 ml Red Top (clot activator) vacutainers and four 6 ml Lavender
Top (EDTA) vacutainers. The vacutainers are transported to a central processing laboratory
and fractionated by centrifugation to yield a total of 48 aliquots comprised of 26
serum, 14 plasma, 4 buffy coat and 4 red blood cells (400–500 ?l per aliquot) in 1
ml Matrix® 2D-barcoded tubes (Thermo Fisher Scientific Inc.). At the time of blood
collection, participants also complete a short Blood Questionnaire that records information regarding their fasting status, recent smoking, medication,
supplement use, family history of cancer and menstrual status.

Hematoxylin and eosin stained slides corresponding to formalin-fixed paraffin embedded
(FFPE) tissue blocks are retrieved for all participants for whom tissue is available.
Archived tissue blocks will be requested and retrieved according to the HE slide
pathology review; the pathologist will mark the area of the block from which triplicate
0.6 mm tissue cores should be collected for the construction of TMAs. In addition
to the collection of tissue cores, 10 ?m tissue scrolls will be collected for the
extraction of nucleic acids (DNA and RNA). The RNA will be used for the transcript
analysis in sub-project 2 and the DNA will be extracted at a later date for ancillary
projects potentially investigating mutational analyses. All blood and tissue samples
are stored within the ACRB.

Participant follow-up at 24, 48 or 72-months

Additional data and biospecimen collections occur at the specified follow-up intervals
of 24, 48 and 72-months from the participant’s primary breast cancer diagnosis. Each
month, the Study Coordinator queries the B2B Recruitment Database to generate a list of participants eligible for follow-up. The Study Coordinator
contacts each participant by telephone to confirm their address and willingness to
continue their participation in the B2B Research Program. If they agree, participants
are sent a follow-up package that includes a PAQ, DHQ II and the appropriate Follow-Up Questionnaire (24, 48 or 72-month); these are self-administered questionnaires to be completed
by the participants and returned by mail to the study office. The 24, 48 or 72-month Follow-Up Questionnaires request information on: personal health history, breast cancer progression (only
at 24 month follow up only), recurrence, contra-laterality and new primary diagnosis,
medications (prescription and over the counter), smoking habits, alcohol consumption,
sun exposure, dietary intake, mobility and physical activity and anthropometric measurements.

Follow-up blood samples are also collected at 24, 48 and 72 months. Each Follow-Up
Package contains a Research Blood Requisition form and participants are asked to donate a blood sample at any Calgary Lab Services
location. The follow-up blood collections consist of a 30 ml of blood sample collected
using three 6 ml Red Top (clot activator) vacutainers and two EDTA vacutainers. Again,
the vacutainers are transported to a central processing laboratory, fractionated by
centrifugation to yield serum, plasma red blood cells and buffy coat and stored in
32 aliquots of 400–500 ?l in 1 ml Matrix® 2D-barcoded tubes (Table 1).

Vital status check

The vital status of each participant is checked before each follow-up contact at 24,
48 and 72 months through a record linkage done by the Department of Cancer Surveillance
(Alberta Health Services). Vital Statistics Alberta (VSA) provides information on
all deaths that occurred in the province to the ACR, on request, with underlying cause
of death provided by Statistics Canada to VSA. There is an average three-month time
lag between the actual death occurrence and reporting to the ACR. Several mechanisms,
such as reciprocal agreements between other provinces and record linkages with the
Canadian Mortality Database, exist to capture the deaths of participants who left
the province of Alberta after their diagnosis. These agreements and processes ensure
that vital status can be determined for over 95 % of participants. Cause and date
of death will also be obtained from this source.

Medical record abstraction

Medical record abstraction will occur in the final year of the B2B Program operation
(commencing August 2018). Health Record Technicians from the ACR will use direct data
entry to a medical record abstraction form to collect data from the medical records
(both paper and electronic charts) for all participants in the B2B Cohort. The medical
record abstraction form was developed from standardized forms used in our past physical
activity and breast cancer cohort study conducted in Alberta 97], 98].

Baseline pathologic data, including clinical stage and pathologic stage (according
to American Joint Committee on Cancer criteria 99], tumor size, grade, histology, estrogen receptor status, progesterone receptor status
human epidermal growth factor receptor 2 status, type and results of computerized
tomography or positron emission tomography scans, status of margins (with breast conserving
surgery), and pathology of lymph nodes (if surgically sampled) are provided by the
ACRB/CoBRA database and verified during the medical record abstraction. Abstracted
variables will include the type of surgery, and all treatment and follow-up care including
data on chemotherapy, radiation therapy, and hormone therapy. Treatment completion
rates will be estimated for chemotherapy and hormone therapy but not for radiation
therapy since few patients fail to complete radiation therapy. For chemotherapy completion
rate, we will estimate the average relative dose intensity (RDI) received for the
originally planned regimen based on standard formulae as we have done in a previous
RCT 100]. For hormone therapy, the follow-up questionnaires ask participants to report if
they have stopped taking their prescribed hormone therapy at any time before its intended
completion and the reasons for stopping.

Disease endpoints are defined according to the Standardized Definitions for Efficacy
End Points in Adjuvant Breast Cancer Trials 101]. Our primary endpoint of interest is bone metastasis or a skeletal-related event
according to the definition in the NSABP 34 trial (http://clinicaltrials.gov/show/NCT00009945). We will also examine other composite disease endpoints as secondary endpoints including
overall survival, distant disease-free survival, distant relapse-free survival, and
distant recurrence-free interval. Finally, we will examine the single disease endpoints
of death from breast cancer and death from non-breast cancer. For participants who
have left the province and who are not known to be deceased, the date of leaving Alberta
will be used as the censoring time.