Examining the effects of adjuvant chemotherapy on cognition and the impact of any cognitive impairment on quality of life in colorectal cancer patients: study protocol

Design

This study will use a longitudinal design. Data will be collected using neuropsychological
assessments and QoL questionnaires at three time points: post-surgery but prior to
chemotherapy treatment (‘T1’), between 12 and 14 weeks after first scheduled chemotherapy
(‘T2’), and 3 months after last scheduled chemotherapy (‘T3’) (Please see Fig. 1). A total of 156 participants (50 % of whom will receive chemotherapy and 50 % will
be non-chemotherapy surgical patients) will be recruited from 5 NHS Healthcare Trusts
across London. For those patients who are not receiving chemotherapy data will be
collected at T1 and then in parallel with the chemotherapy group at T2 and T3.

Fig. 1. Study measurement time points

Participants

Adult CRC patients under the care of the Consultant Oncologists at five participating
NHS Trusts with London-based hospital sites who have had colorectal surgery will be
invited to take part in the study.

Patients who have had prior exposure to chemotherapy or significant psychiatric or
medical comorbidities, which might affect ability to participate in the study, will
be excluded. Patients who are not sufficiently literate in English will also be excluded
as a failure to understand English would make completion of the questionnaires impossible.
Only those patients over the age of 18 years who have had surgery following a diagnosis
of CRC will be eligible to participate, provided that they are then offered adjuvant
chemotherapy treatment and start it at least 3 weeks after surgery or no other cancer
related treatments at all and are fluent in written and spoken English.

During the post-surgery follow-up appointment nurses/trial co-ordinators at each location
will provide a consecutive series of patients (satisfying the inclusion criteria)
with information about the study. A research assistant will also be available at that
time to answer any questions that the patient may have about the study. Those patients
who provide the researcher with telephone numbers will be contacted after 48 h and
invited to participate in an interview, either at their chemotherapy clinic or at
home. This interview will take place prior to the commencement of chemotherapy treatment
for those in receipt of chemotherapy and at a parallel point in time for the surgery
only control group (T1). At the beginning of the interview the patient will be guided
through the information sheet again and the consent form by the researcher and written
informed consent will be obtained. In the event that a patient declines to provide
the researcher with a telephone number or refuses to take part he/she will not be
contacted again about the study.

The questionnaires and assessments will be completed by the patient in the hospital
at T1 and an appointment for the subsequent assessments (T2, T3) will be made. Patients’
participation in the research will take approximately 2 h and 15 min at T1 and 1 h
50 min at T2 and T3 and will take place at the time of the appropriate outpatient
appointment or at an equivalent point in time for the control group.

Measures

Pre-screening test

At T1, consented participants over the age of 65 will be asked to complete the Montreal
Cognitive Assessment (MoCA) version 3 23] as a pre-screening test in order to exclude those with mild cognitive impairment
(MCI) from taking part in the study. In the event that such a participant obtains
a raw score of less than 26 they will not progress into the study, as this is considered
to be the cut off point for MCI.

The following measures will be collected at T1, T2 and T3 unless otherwise specified:

Neuropsychological assessments

The following battery of assessments has been designed to measure a wide range of
cognitive domains and includes all of those recommended by the ICCTF 10]. All measures are standardised, validated and taken from published test batteries
with healthy population norms, which will provide the researchers with another important
comparison:

i) The Hopkins Verbal Learning Test-Revised (HVLT-R) 24] for verbal memory; this is a brief verbal learning and memory test that includes
delayed recall and recognition trials. Alternate forms will be used at each of T1,
T2 and T3.

ii) Trail Making Test (TMT) A and B 25] to measure psychomotor speed and aspects of executive function and spatial organisation,
visual pursuits, recall, and recognition.

iii) The Controlled Oral Word Association (COWA) of the Multilingual Aphasia Examination
26] that measures speeded lexical fluency requiring aspects of executive function.

The above-recommended measures will be supplemented with the following:

iv) The Digit Span subtest of the Wechsler Adult Intelligence Scales – Third Edition,
(WAIS – III Digit Span) 27] consisting of two mental activity tests involving auditory attention and short term
memory retention capacity.

v) The Symbol Digit Modalities Test (SDMT) 28], 29] assesses complex visual scanning and tracking 29]. It is a simple substitution task.

vi) Letter Cancellation of the Behavioural Inattention Test (BIT) 30], 31].

vii) Grooved Pegboard Test 32], 33], a manual dexterity test measuring visuo-motor coordination.

viii) The Benton Visual Retention Test (BVRT) 34] for visual perception, visual memory and visuo-constructive ability. There are three
near-equivalent forms (Forms C, D, and E) of the BVRT. Form C will be used at T1,
Form D at T2 and Form E at T3, which will allow for retesting while minimizing practice
effects. Administration A (of the 4 possible methods) will be used throughout.

Self-reported cognitive functioning

Functional Assessment of Cancer Therapy-Cognitive scale (FACT-Cog, Version 3) 35] is a validated self-report measure of cognitive function. It evaluates mental acuity,
attention and concentration, memory, verbal fluency, functional interference, deficits
observed by others but reported by the patient; change from previous functioning,
and impact on quality of life.

Mood

Anxiety and depression will be measured using the Hospital Anxiety and Depression
Scale (HADS) 36].

Fatigue

The Functional Assessment of Chronic Illness Therapy – Fatigue (FACIT-F, version 4)
37] a 13-item self-report subscale of the FACT-G (see below). The FACIT-F is a well-validated
quality of life instrument widely used for the assessment of cancer-related fatigue
in clinical trials 37]–40]. The items include physical and functional consequences of fatigue 37].

Quality of life

The Functional Assessment of Cancer Therapy – General (FACT-G, Version 4), will be
used to measure 4 quality of life domains 38]: physical, emotional, family/social and functional well being in the previous 7 days.
Participants will also complete the 9-item FACT-C subscale that evaluates symptoms
related specifically to CRC.

IQ

This will only be measured at T1 using the Wechsler Abbreviated Scale of Intelligence
– Second Edition (WASI –II) Vocabulary and Matrix Reasoning 41] to assess background level of intellectual ability.

Socio-demographic information

This information will be collected at T1 via a structured questionnaire and will include
age, sex, employment (i.e. full or part-time employment, retired and unemployed) and
marital status (married, cohabiting, single, separated, divorced, widowed). Specific
information relating to surgery date, planned treatments and comorbidities will also
be obtained.

Medical records and treatment plan

Participants’ disease and treatment-related factors will be recorded from medical
records (including the type of chemotherapy administered, any dose adjustments made,
the actual number of cycles completed, any neurotoxicity experienced and the anti-emetic
regimen) once the participant has consented.

Sample size

A recent meta-analysis of chemotherapy and cognitive function 42] estimated mean effect sizes in a range of cognitive domains. The effect sizes ranged
from d?=?0.11 to?0.51. A sample size calculation was performed using GPower 3.1.
Taking into consideration the resource constraints of the study, the sample size was
calculated with the aim of detecting a medium effect size. To detect an effect size
of?0.26 with 80 % power and a significance level of 0.05 at the final time point,
a minimum sample size of 120 participants was indicated. Based on medium effect sizes
in the meta-analysis, a sample size of 120 would allow effects to be detected in the
following domains: executive function, information processing speed, language, motor
function, verbal memory and visual memory. However, it is acknowledged small effects
may not be detected in the following domains: attention and visuospatial skills. Assuming
an overall attrition rate of 22 % (based on SCOT trial attrition rates¹
), a total sample size of 156 participants will be sought (78 per group).

Analysis

In order to detect subtle changes in cognitive function, pre- and post-chemotherapy
difference scores will be calculated. This approach has been successfully applied
in cardiac research exploring post-surgery cognitive decline 43]. A standardised score (z-score) will be computed for every patient on each neuropsychological
test by dividing the test score by the standard deviation of the pre-chemotherapy
test score of all study participants. A standardised score will be computed for each
test at all-time points using the pre-chemotherapy standard deviation. The post-chemotherapy
standardised score will then be subtracted from the pre-chemotherapy standardised
score to give a relative difference score for each patient. A total z-score can then
be computed for all neuropsychological tests.

ANCOVA will be used to compare mean difference z-scores between the chemotherapy and
surgery-only groups while controlling for the effects of gender, age, depression,
educational level and extent of disease. This method of analysis is preferable to
conventional deficit/no deficit analysis as it allows for detection of subtle changes
in cognition and accounts for pre-chemotherapy cognitive performance 44] and will increase the power of the analyses.

Multiple and logistic regression analyses will be used, as appropriate, to explore
the relationship between cognitive impairment (total z-score) and quality of life
(FACT G C), adjusting for age, gender, SES and anxiety and depression (HADS). Finally,
correlation and regression analyses will be also used to initially examine the relationship
between subjective (FACT-Cog) and objective (total z-score) cognitive impairment.

Ethics and acceptability feasibility

Ethical approval for the study was obtained from the NHS Health Research Authority
– NRES Committee South-West Cornwall Plymouth in August 2013.

All of the assessments are standardised and have been widely used across many patient
groups including cancer patients. At the beginning of each assessment participants
will be reminded that they have the right to withdraw at any time and can avoid answering
questions that are felt to be too personal or intrusive. Participants will be assured
that any future treatment will not be affected in any way should they choose to withdraw.
However, in the unlikely event that the assessments and content of the questionnaires
cause distress or any discomfort to any of the participants, the researcher will remind
the participant that he/she is entitled to refuse to answer any question that may
cause upset or distress and that he/she may stop and withdraw from the study at any
time. If they feel the need to have professional help they will be encouraged to raise
this with their consultant or the consultant will be informed by the researcher if
the patient would prefer.

Data management and data confidentiality

Confidentiality will be adhered to at all times. All questionnaires will be kept anonymous
by assigning codes to participants. All data will only be identified by that code,
not by the participant name or any other information that could identify them. All
questionnaires will be kept in locked cabinets and/or password protected computers.

Data will be collected, transferred and stored in compliance with the NHS data protection
requirements and be managed by a data manager. The data manager will also advise on
current regulatory framework regarding data protection and data management procedures
in compliance with the Data Protection Act 1998 and other regulations. The data manager
will design and set up a bespoke database in MS Access, which will have integrated
data validation checks and a full audit trail. Patient identifiable and pseudonymised
data will be stored separately. The data manager will advise on and set up data transfer
systems and encryption systems so that all patient identifiable data is encrypted.
The data manager will also advise on storage, back up and archiving of data to ensure
databases are regularly backed up to ensure data is safeguarded from accidental loss.
The study master file and all study documentation will be archived for 10 years.