Description of the problem

Cancer survivors first raised awareness that some patients experienced problems with their memory and/or concentration during and after chemotherapy. Also, complaints of difficulty with multitasking after undergoing chemotherapy are frequently reported by patients. For most patients the deficits appear subtle and often improve after ceasing treatment; however, for a subset of people the symptoms are sustained and have a significant impact on their quality of life and ability to function in their everyday activities.


It is only in the last decade that this area of research has begun to be systematically investigated. Most of the knowledge gained so far on the effects of chemotherapy on cognition, is based on studies investigating adjuvant regimens applied in the treatment of breast cancer. This is not because it is thought that specifically these cytototic regimens might be neurotoxic, but simply because these regimens are given to many patients, surviving for long periods of time, thereby providing a good model for study on the effects of chemotherapy on cognition.

Earlier studies reported cognitive impairment in 15-50% of adult solid tumor survivors who had received chemotherapy. Most of these studies were limited by sample size and were cross-sectional in design, with no evaluation of cognitive function prior to treatment and no longitudinal data. Four meta-analyses have concluded that there is evidence for cognitive changes associated with cancer and cancer treatments, but all reported that the effect size of the impairment was small to moderate.

A small number of pre-treatment and prospective, longitudinal studies have been published, and a number are ongoing. Studies evaluating patients prior to receiving chemotherapy have found higher than expected cognitive impairment prior to receiving adjuvant treatment, with up to one third of patients having impairment. Studies evaluating cognitive function longitudinally post-treatment have had varying results: some have reported a decline in cognitive function after chemotherapy while others show no significant change. This discrepancy may reflect the different chemotherapy regimens administered and/or methodological problems relating to the design and analysis of the studies (e.g., variations in normative data and reference groups, statistical cut-offs used to define cognitive impairment, use or not of corrections for practice effect and of alternate forms of neuropsychological tests to minimize practice effect, and control for test-retest variability in comparable people not undergoing chemotherapy).

There are also studies emerging that suggest that cognitive impairment may also occur in a subset of patients receiving hormonal anti-cancer treatment.

Nature of the cognitive difficulties

The earlier studies consistently reported that the cognitive sequelae were diffuse and relatively non-specific in nature, but most consistently involving the domains of attention and concentration, verbal and visual memory, and processing speed.  However, evidence from new studies suggests a frontal, subcortical toxicity profile, with cognitive dysfunction within domains of information and processing speed, attention, memory retrieval and executive function.

Duration of the impairment

The duration of the cognitive impairment after anti-cancer treatment remains unclear.  One study has found impairment in patients up to 10 years after systemic treatment, whilst others have found no difference at 4 years. Large longitudinal studies with longer term follow-up are required to answer this question; however, they will need to incorporate appropriate control groups so that practice effect can be accounted for.

Self-reported cognitive problems

A puzzling uniform observation from studies in cancer patients is the discrepancy between cognitive impairment as measured with standardized neuropsychological tests and cognitive complaints as reported by the patients.

Usually the percentage of patients that complain about cognitive problems after treatment, independent of the cytotoxic combination used, is considerably higher than the percentage of patients with detectable cognitive impairment. Also, the group of patients with detectable cognitive impairment does not per definition coincide with the patient group that expresses cognitive complaints.

In the literature, several explanations have been proposed for this discrepant finding, which is not unique for this cancer population and which is often observed in clinical neuropsychology; the subject of the relation between self-reported cognitive complaints and cognitive functioning as measured with objective cognitive tests is generally viewed as complex and difficult. In trying to understand the discrepancy found between objectively measured cognitive functioning and reported cognitive complaints, the following explanations can be considered.

Explanation for discrepant finding

Ecological relevance

The absence of a clear relation between patients’ self-reported complaints and objective test performance may be explained by the low ecological relevance of most neuropsychological tests. The domains assessed by neuropsychological tests of cognitive function show little overlap with the everyday experience on which patients base their self-report; traditional neuropsychological tests can be viewed as artificial in terms of everyday cognitive function. Thus, the absence of a relationship between laboratory and everyday tasks is not really surprising.


The possibility exists that self-perception could be a more sensitive method to assess subtle cognitive impairment than objective neuropsychological assessment. Particularly for highly educated persons, the tests may not be sensitive enough to measure cognitive impairment, because of ceiling effects. This impairment can nonetheless adversely affect quality of life and cause cognitive complaints.

Laboratory setting

The conditions under which the neuropsychological tests are assessed and the implications of these conditions for the performance measured should be taken into account. Laboratory tasks have been designed to control all except one single critical variable and to minimize the effects of individual differences. Unlike real-life demands, laboratory tasks are experienced under conditions of minimal distractions and stress. Patients are reassured and encouraged to do their best, leading to the full use of their potential cognitive abilities. In this respect, neuropsychological testing measures the capacity of a patient within the optimal circumstances of a test situation. The ability, however, to learn a list of words in a quiet atmosphere of a test room does not guarantee that a patient will remember the name of a person he/she was introduced to at a noisy social gathering, or what goods he/she had intended to purchase at a busy supermarket.


Furthermore, normal test results may be achieved only with an abnormal expenditure of effort. Patients may be able to perform well during a relatively short test period, but their daily functioning may be compromised and give rise to cognitive complaints. In other words, patients may be able to mobilize cognitive resources or to compensate for cognitive problems for a short time during a neuropsychological test session, but they are not able to sustain this for longer periods of time during everyday activities, leading to the report of cognitive problems. A technique that is used nowadays to address this issue, is functional MRI, in which brain activity patterns are studied during test performance. First results from studies using these techniques are supportive of this notion.

Psychological distress

Another possibility that has to be considered is that cognitive complaints expressed by cancer patients are only partly attributable to underlying cognitive dysfunction, and that psychological distress and fatigue play an important role in the occurrence of cognitive complaints. It is indeed often found that self-reported cognitive problems of patients are related to negative affect, whereas the objective test performance is not, indicating the possibility that cognitive complaints do not reflect actual cognitive deficits but a way in which a patient copes with stressful events.


It can also be argued that people may have only limited conscious access to their own cognitive processes. Especially subjects with cognitive deficits may have difficulty in objectively assessing their own daily cognitive failures. Whatever the degree of insight into one’s own cognitive abilities can be expected, people can only make relative comparisons, assessing their competence against their ability to cope with the particular demands which their lives make of them. This obviously reflects many more aspects besides cognitive functioning, for example self-respect or confidence.

Impairment, disability and handicap

Neuropsychological testing can be best viewed as a measure of the capacity of a patient. This implies that a patient can truly experience cognitive problems in daily life, leading to the report of cognitive complaints, without exhibiting impairment on neuropsychological testing, because full use of abilities is being employed at that moment. At the same time, neuropsychological tests may elicit deficits, without bringing about an effect in daily life functioning. In what way the cognitive capacity (as measured with neuropsychological tests) is expressed or experienced in daily life functioning by a patient is dependent on the circumstances (including personal and environmental factors) in which the patient lives. In other words, the injury occurring in brain functioning as a consequence of cytotoxic treatment (impairment) can be manifested in the cognitive capacity (disabilities) measured with the neuropsychological tests, which in turn can lead to a complaint (handicap) in daily life functioning. It is the extent to which a disability becomes a handicap that is determined by a number of factors.

Impairment, disability and handicap can all be expressions of a single underlying phenomenon, without clear-cut associations between these components. Which components should be accentuated and what outcomes will be most decisive is dependent on the questions asked.