1. Queensland Institute of Medical Research
2. Wesley Medical Centre
3. Wesley Research Institute
Outcomes research is an important focus of public and research policy in Australia. Recent reports of cognitive impairment after chemotherapy for breast cancer have highlighted the emerging importance of cognition as a clinically relevant outcome. To date, most studies have utilised cross-sectional study designs to investigate impairment after treatment. Future research requires a within-group study design that utilises a longitudinal repeated measures approach to identify the nature and magnitude of cognitive change after adjuvant chemotherapy and to evaluate factors that mediate cognitive functioning. The results of within-group studies will also provide the foundations for individual patient assessment, with its focus on the development of tailored behavioural interventions for significantly impaired patients. As cognitive research matures, the neuropsychologist will become an important addition to the multi-disciplinary cancer care team.
The neurotoxicity of chemotherapy is well known, but cognitive impairment in the absence of a demonstrable neurological disability has been recognised only recently as an important clinical problem. To date, most of the research literature has reported cognitive functioning in women with early breast cancer who have been treated with adjuvant chemotherapy. This review briefly describes the conceptualisation of cognition and its measurement and summarises the results of studies that report cognitive impairment in women after adjuvant chemotherapy for breast cancer. It also evaluates the research agenda required to better understand the nature and extent of cognitive change after chemotherapy and to develop tailored interventions for women diagnosed with cognitive impairment.
Cognition is a knowledge based process that recognises, stores and retrieves information. In contrast to the cognitive states of other animal species, human cognition is characterised by a far more richly diverse interplay of perception, memory and thinking. Furthermore, human cognition displays enormous plasticity throughout life in response to brain development and fluctuations in physical, emotional and social health.
The main components of cognition incorporate attention, memory, language, conceptualisation and visuospatial abilities.1 These components in turn comprise a multi-faceted array of interlocking domains, each representing a specific cognitive attribute. An additional construct, executive functioning, describes higher order control and coordination of cognitive operations that are required for normal daily living activities such as planning, organisational capacity, strategic thinking and problem solving.2 The discipline of neuropsychology is devoted to the study of brain – mind relationships and numerous tests have been devised by research neuropsychologists over the past 60 years to investigate cognitive functioning.
Cognitive tests can be classified as global, component or construct specific and disorder specific. Ideal tests are characterised by appropriate conceptualisation, robust psychometric properties, the capacity to detect changes over time and validated alternate forms that neutralise the potentially confounding effect of test recollection after repeated administration. The selection of an appropriate testing protocol for a particular clinical setting should take into account the cognitive phenotype of the disease by profiling the cognitive symptoms described by patients. The terms ‘chemobrain’ and ‘chemofog’ are often used by women with breast cancer to describe the effects of adjuvant chemotherapy on memory, concentration and mental agility. In order to fully understand the nature and severity of these symptoms, validated tests capable of reliably detecting change over time should be selected from cognitive domains that target this clinical phenotype. Table 1 summarises examples of symptoms described by women, the domains of cognitive functioning into which these symptoms fit and examples of cognitive tasks that can be selected to test these domains.
Since the first report a decade ago, numerous studies utilising a cross-sectional design have investigated cognitive functioning in women with breast cancer after adjuvant chemotherapy.3-7 In these studies, cognitive functioning was assessed during chemotherapy, at six months after treatment and one to ten years after treatment. The findings were compared with various control groups. The neuropsychologic tests varied widely between studies and included a self-report scale of cognition, a self-administered instrument of cognitive functioning and a battery of administered neuropsychological tests designed to assess a variety of cognitive domains. Self-report measures of quality of life, depression, anxiety and fatigue were also evaluated in various studies to examine the relationships between these constructs and cognitive functioning. Findings from each of the cross-sectional studies indicated that treatment groups performed more poorly than control groups in at least some areas of cognitive functioning. While between-group analyses failed to identify significant differences in many cognitive tasks, women in treatment groups were more likely to be impaired on memory, attention, processing speed and motor tasks than control groups.4-6
Three studies have utilised a longitudinal design, two with pre-treatment assessments and one with the first assessment after commencement of adjuvant chemotherapy. In one study, cognitive functioning was assessed in 18 patients before commencement of chemotherapy, three to four weeks after chemotherapy and 12 months after chemotherapy.8 A large battery of neuropsychological tests, taking several hours to complete, assessed attention, processing speed, learning, memory, executive functioning, visuospatial processing and motor skills. Participants also completed self-report scales of anxiety and depression and quality of life. This study reported that 61% of patients experienced a decline in cognitive functioning relative to baseline in one or more domains immediately after chemotherapy, with 45% of those patients showing improvement at 12 months post-chemotherapy. Importantly, 35% of patients exhibited cognitive impairment before commencement of chemotherapy. The second study with baseline measures reported preliminary results in 50 women undergoing adjuvant chemotherapy for early breast cancer.9 A comprehensive battery of neuropsychological tests evaluated verbal and visual memory, working memory, processing speed and executive functioning. In comparison with healthy controls, chemotherapy treated patients had a measurable cognitive decline, especially in tasks of verbal and working memory. In the third study, cognitive functioning was assessed using a global screening tool in 100 women undergoing chemotherapy and in a patient nominated, age matched control group.10 The results indicated that patients were more likely to show cognitive impairment at the time of completion of chemotherapy than the control group, but there was a subsequent trend towards recovery.
The recently recognised effect of chemotherapy on cognitive functioning has stimulated numerous reviews, a meta-analysis and a workshop to summarise results to date and recommend future directions for research. A meta-analysis of 29 studies identified three basic research methods that compared post-treatment performances to normative data, control groups, or baseline assessments.11 In comparison with normative data and control groups, chemotherapy treated groups exhibited consistent reductions of task performance in the domains of verbal memory, motor function and executive functioning. The findings from the workshop recognised that even small changes in some areas of cognition, such as attention and executive functioning, could be associated with reductions in the ability to function effectively in work, home and family/social environments.12 Recommendations from the meta-analysis, workshop and other reviews placed emphasis on the development of longitudinal studies with larger samples and pre-treatment assessments to better understand the nature, degree and duration of cognitive changes after chemotherapy.
Although the results of published studies are proof of principle that administration of chemotherapy is associated with altered cognitive functioning, significant conceptual and methodological gaps separate the notion of a measurable change and the management of a clinically relevant impairment. For example, what does the term ‘cognitive dysfunction’ mean and how is it diagnosed and managed? What is the appropriate structure of future clinical studies to investigate causality and which cognitive tests should be used? The results of published studies to date provide only limited insights into these issues. Partnership with the mature and expanding discipline of neuropsychology is required to understand gaps in knowledge and to establish an appropriate assessment/treatment model.
The selection of appropriate cognitive tasks is a crucial requirement to assess and treat clinically relevant deficits. Global cognitive measures have the advantage of ease of administration, but are less likely to detect changes over time in a domain that might be specifically relevant to a particular disease or group of patients. However, there are also limitations of tests that tap specific domains. To date, no domain specific test has been devised that addresses all aspects of that domain. At the same time, tasks intended to be domain specific frequently cross into other domains. Considerable expertise is therefore required in the interpretation of tests that apply artificial cognitive scenarios distant from daily living. While self-administered tests have been the hallmark of research in psycho-oncology, investigation of specific domains of cognitive functioning requires administration and interpretation of tests by board registered neuropsychologists.
Advances in study design are also required in order to provide more detailed information about factors that mediate cognitive functioning or confound the interpretation of results. While inter-group comparisons provide a snapshot of a clinical problem, intra-group comparisons are required to identify factors that both influence cognitive performance and confound the interpretation of results. For example, baseline and follow-up assessments not only provide additional information about the nature and magnitude of change, but also assess factors that influence task performance (eg. psychotropic medication, change of ovarian function after chemotherapy, use of endocrine treatments). Furthermore, within-subject studies provide intra-individual data that provide the basis for diagnosing and managing cognitive impairment. Thus the term “cognitive dysfunction” described in published cross-sectional studies, assumes that a statistically significant difference of scores between chemotherapy and control groups describes a clinically significant impairment, rather than merely a change. Since some of these statistically significant differences occur when mean scores in both groups are still in the normal range, within-subject studies are necessary to distinguish impairment and change.
By way of example, consider the following three scenarios of individual women treated with adjuvant chemotherapy whose baseline cognitive scores are known and who participate in a cross-sectional study that reports impaired post-treatment scores in comparison with those of a control group. In the first scenario, a woman with a baseline score just below the mean of the post-treatment group does not change as a result of chemotherapy. Her result contributes to the overall poorer score of the post-treatment group, but this woman does not have a chemotherapy induced impairment. In the second scenario, a woman with a baseline score two standard deviations above the mean of the control group drops to the mean of the control group after chemotherapy. She is a computer programmer who now has difficulty in performing complex tasks at work. She reports that she is concerned about losing her job and is now anxious about her declining work performance. Her score after chemotherapy does not contribute to the overall poorer result of the post-treatment group, but clearly she has an impairment. In the third scenario, a woman has a baseline score equal to the mean of the control group and drops to the mean of the post-treatment group after chemotherapy. She is a housewife whose daily activities have not changed but she perceives that she is different now and is distressed by this awareness. This woman’s score contributes to the worse result of the chemotherapy group and her perception of change and objective change match. She is anxious about the change, but her daily living cognitive performance as a housewife remains normal.
These scenarios illustrate the need for longitudinal studies with baseline measures in order to provide a better understanding of the nature of cognitive change after adjuvant chemotherapy and its relationship to other patient reported outcomes. Although published cross-sectional studies have employed self-report scales of quality of life, anxiety and depression to ‘correct’ for differences in cognitive scores between groups, the nature of this relationship is unknown. Changes in quality of life, anxiety and depression are well documented in women with early breast cancer after chemotherapy, but scenario two above suggests that impaired cognitive functioning may mediate anxiety as a result of significant impairment of daily living. In scenario three, the perception of cognitive impairment, not matched by impairment of daily living activities, could also plausibly mediate anxiety.
These examples highlight the emerging importance of the cognitive domain in psycho-oncology research. Unlike self-report scales of quality of life however, tests of cognitive functioning are applied and interpreted by trained neuropsychologists. The addition of the neuropsychologist to the multi-disciplinary cancer team can therefore add a new dimension both to psycho-oncology research and cancer care by integrating the clinical skills of history taking and the application and interpretation of cognitive tests to assess impairment and to develop individualised behavioural interventions.
4. Van Dam FSAM, et al. Impairment of cognitive function in women receiving adjuvant treatment for high-risk breast cancer: High-dose versus standard-dose chemotherapy. J Natl Cancer Inst. 1998; 90(3): 210-218.
9. Shilling V, Jenkins V, Morris R, Deutsch G, Bloomfield D. The effect of adjuvant chemotherapy of cognition in women with early breast cancer – preliminary results of an observational longitudinal study. The Breast. 2005; 14: 142-150.