|Home | About | Journals | Submit | Contact Us | Français|
Chemotherapy-induced cognitive changes have been an increasing concern among cancer survivors. Using adjuvant treatment for breast cancer as the prototype, this manuscript reviews research from neuropsychological, imaging, genetic, and animal model studies that have examined the clinical presentation and potential mechanisms for cognitive changes associated with exposure to chemotherapy. An impressive body of research supports the hypothesis that a subgroup of patients is vulnerable to post-treatment cognitive changes, although not exclusively related to chemotherapy. Further, imaging and animal model studies are providing accumulating evidence for putative mechanisms for chemotherapy-induced cognitive change. Models of aging are also reviewed in support of the proposal that cognitive changes associated with cancer and cancer treatments can be viewed in the context of factors that affect the trajectory of normal aging.
References to cognitive changes associated with chemotherapy can be found dating back to the 1980s (1); however, serious scientific attention was not paid to the topic until the mid-90s (2–4). Post-treatment cognitive changes frequently include problems in attention, concentration, working memory and executive function. This review will use cognitive changes associated with adjuvant treatment for breast cancer as an illustrative example since the bulk of the research has been conducted in this area. Further, the relevance of viewing this research within the context of models of aging will be explored.
To date, 21 longitudinal studies of breast cancer patients (5–25) that include pre-and post-treatment assessments have been reported and the majority of studies (16) have found evidence for post-treatment cognitive change in a subgroup of individuals, while 5 studies reported negative findings (10, 14, 16, 20–21). Estimates of the prevalence of post-treatment cognitive change vary among studies, likely due to differences in patient populations, assessment instruments used, criteria for defining change, and other aspects of study methods. Many investigators site the incidence of post-treatment cognitive problems as ranging from 15–25% (26), although percentages as high as 61% have been reported (5).
Inclusion of pretreatment assessments revealed an unanticipated result in that studies have found that 20–30% of breast cancer patients have lower than expected cognitive performance based on age and education prior to receiving adjuvant treatment (e.g., 27–28). Interestingly, lower than expected level of performance does not appear to be related to psychological factors (depression or anxiety), fatigue, or surgical factors (e.g., type and length of general anesthesia) (27). Two, non-mutually exclusive hypotheses have been proposed to explain this finding (29): 1) The biology of cancer (e.g., an inflammatory response triggering neurotoxic cytokines) may contribute to lower than expected cognitive performance and/or 2) Common risk factors for the development of both breast cancer and mild cognitive changes over years may exists (e.g., poor DNA repair mechanisms have been linked to risk of cancer and neurodegenerative disorders).
Further, the assumption that cognitive changes were due to chemotherapy exposure alone has also been questioned as evidence emerged suggesting that the combination of chemotherapy and endocrine therapy or endocrine therapy alone may cause cognitive change (30). Initial examination of this issue produced mixed results; however, most studies were not powered to adequately examine the independent effects of endocrine therapy. Schilder et al (31) conducted neuropsychological assessments in the context of a longitudinal study examining patients not treated with chemotherapy, who were randomized to treatment with tamoxifen or exemestane. They demonstrated that patients treated with tamoxifen, but not exemestane experienced cognitive change compared to healthy controls. Early investigators assumed that they were studying the effects of chemotherapy; however, most breast cancer patients receive multi-modality treatment; surgery with exposure to general anesthesia, radiation therapy, and endocrine therapy in addition to chemotherapy. This in combination with the evidence for pretreatment cognitive problems led Hurria and colleagues to propose the phrase “cancer and cancer treatment associated cognitive change” more accurately describes this phenomenon (32).
The finding that only a subgroup of patients experience persistent post-treatment cognitive decline, leads logically to the examination of risk factors for cognitive change. Age is a well established risk factor for cognitive decline, and researchers have speculated that older adults may be more vulnerable to cognitive side effects of cancer treatments. Cognitive reserve, which represents innate and developed cognitive capacity (influenced by education, occupational attainment, and lifestyle) has also been associated resiliency (high) or vulnerability (low) to cognitive decline following various brain insults. Support for an interaction of age, cognitive reserve and exposure to chemotherapy as risk factors for cognitive decline has been reported (19); older patients with lower levels of pretreatment cognitive reserve exposed to chemotherapy demonstrated significantly reduced performance on post-treatment processing speed. Exploratory analyses conducted by Schilder et al (31) also revealed that, in older breast cancer patients (>65), tamoxifen had a larger effect on more cognitive domains suggesting an age-dependency of the impact of tamoxifen on cognitive functioning.
Genetic factors have also been examined as potential risk factors for cognitive decline. Apolipoprotein E (ApoE) is a complex glycolipoprotein that facilitates the uptake, transport, and distribution of lipids and plays an important role in neuronal repair and plasticity after injury. The human E4 allele has been associated with a variety of disorders with prominent cognitive dysfunction including healthy individuals with memory difficulties, Alzheimer’s disease, and poor outcomes in stroke and traumatic brain injury (33). Ahles et al. (34) evaluated the relationship of the ApoE genotype to neuropsychological performance in long-term cancer survivors treated with standard dose chemotherapy. The results demonstrated that survivors with at least one E4 allele scored significantly lower in the visual memory and spatial ability domains, with a trend to score lower in the psychomotor domain, as compared to survivors who did not carry an E4 allele.
Small et al. (35) studied catechol-o-methyl transferase (COMT) which influences the metabolic breakdown of catecholamines through the methylation of dopamine (DA). The valine version (val allele) is approximately four times as active as the methionine version of the gene (met allele). Individuals homozygous for the val allele presumably metabolize DA much more rapidly (i.e., have lower levels of DA) than those with the met allele. Therefore, COMT is a major modulator of dopaminergic tone in the frontal cortex. These researchers found that breast cancer patients who had the COMT-Val allele and were treated with chemotherapy performed more poorly on tests of attention, verbal fluency and motor speed as compared to COMT-Met homozygotes.
Other genetic factors that have been suggested as potential candidates for increasing risk for chemotherapy-induced cognitive change include genes that regulate DNA repair (e.g., X-ray repair cross complementing protein 1, XRCC1; Meiotic recombination 11 homolog A, MRE11A), cytokine regulation (e.g., Interleukin 1, IL1; IL6; tumor necrosis factor alpha TNF-alpha), neurotransmitter activity (e.g., BDNF), and blood brain barrier efficiency (e.g., multidrug resistance 1, MDR1; organic anion transporting polypeptide, OATP). However, no studies to date have directly examined 7 the relationship between these genes and chemotherapy-induced cognitive dysfunction (29).
Several studies have not found evidence for cognitive changes associated with chemotherapy or other treatments. This inconsistent pattern of results may be related to variability in study design and choice of comparison groups. Two of the studies compared patients treated with chemotherapy to patients treated with endocrine therapy, but did not include a healthy control group (14, 21). However, both chemotherapy and endocrine treated patients may experience cognitive change, although through different mechanisms, which may explain the lack of group differences. Further, the pattern of post-treatment cognitive deficits may be influenced by sample characteristics like age and cognitive reserve. Therefore, if a study population consists of young, highly educated (one proxy for cognitive reserve) patients, then one might expect less evidence of post-treatment cognitive deficits as compared to a study that includes older, less educated individuals. Two of the studies with negative findings included cancer patients with a mean age in the 40s (16, 20). Further, in modest size studies, the mix of patients with the vulnerable alleles of genes like APOE and COMT can vary significantly.
Several cross-sectional, post-treatment studies utilizing magnetic resonance imaging (MRI) have documented reductions in gray matter, primarily in frontal structures and hippocampus, and white matter integrity in cancer survivors treated with chemotherapy (36–40), although negative results have been reported (41). Longitudinal studies have reported similar results: 1) decreased gray matter density in bilateral frontal, temporal (including hippocampus), and cerebellar regions and right thalamus at one month post-chemotherapy with only partial recovery at one year post-chemotherapy in several structures, contrasted with no significant changes in gray matter over time in the no chemotherapy cancer group or the healthy controls (42) and 2) Decreased frontal, parietal, and occipital white matter integrity in chemotherapy exposed patients with no changes in either no chemotherapy or healthy controls at post-treatment (43).
Cross-sectional studies of cancer survivors utilizing functional imaging techniques including functional MRI (fMRI) (44–47) and functional positron emission tomography (fPET) (48) have demonstrated areas of decreased activation during performance of a cognitive task in survivors exposed to chemotherapy as compared to controls in areas similar to the structural differences described above. McDonald et al (49) conducted a longitudinal study utilizing fMRI and found frontal lobe hyperactivation to support a working memory task prior to treatment, decreased activation one month post-chemotherapy, and a return to pretreatment hyperactivation at one year post-treatment. A similar pattern was seen in patients treated with endocrine therapy. Interestingly, two other studies have reported over-activation during a memory task prior to treatment in cancer patients compared to healthy controls, consistent with the reports of neuropsychological deficits at pretreatment (50–51). One interpretation is that pretreatment over activation represents an attempt to compensate for pretreatment decreases in brain resources; however, over years, patients lose the ability for compensatory activation as a result of exposure to cancer treatments and/or age-associated changes in the brain (see the section on Cancer, Cognition and Aging below).
Seigers and Fardell (52) recently reviewed the animal studies of chemotherapy-induced cognitive impairment. Studies utilizing common chemotherapeutic agents demonstrated changes in memory and learning which parallel the deficits seen in cancer survivors. Further, animal studies have demonstrated evidence for a variety of potential mechanisms for the effect of chemotherapy on the brain including: 1) inhibition of hippocampal neurogenesis; 2) oxidative damage; 3) white matter damage, including progressive change associated with 5-FU; 4) decreased hypothalamic-pituitary-adrenal axis activity; and 5) reduced brain vascularization and blood flow. Additionally, concentrations of chemotherapy agents which are ineffective in killing tumor cells have been shown to increase cell death and decrease cell division in brain regions including hippocampus suggesting that small amounts of chemotherapy crossing the blood brain barrier can have toxic effects (53).
Emerging evidence from animal studies supports the efficacy of antioxidants in blocking the behavioral and physiological effects in the brain when co-administered with chemotherapy (52). Although this is an interesting proof of principal, antioxidants may not have immediate clinical utility because of concerns that they may decrease the efficacy of chemotherapy. Fluoxetine has been shown to prevent deficits in behavior and hippocampal function when administered before and during administration of 5-FU and may represent a more promising preventative approach (54–55).
Data from imaging and animal studies support the hypothesis that chemotherapy affects brain structure and function and begin to provide evidence for candidate mechanisms of chemotherapy-induced cognitive change. Similar studies examining other aspects of cancer treatments such as endocrine therapy for breast cancer and hormone ablation therapy for prostate cancer are clearly needed.
Given the data described above suggesting an association between age and post-treatment cognitive decline, a potentially useful perspective may be viewing cognitive change within the context of factors that influence the trajectory of normal aging. Cancer and aging are linked, although the molecular mechanisms responsible for the increasing risk of cancer with increasing age are not completely understood. Aging is associated with a variety of biological changes including increased cell senescence, DNA damage, oxidative stress, inflammation, and decreased telomere length (telomerase activity) (56–57). Chemotherapy has been associated with increased DNA damage, oxidative stress, inflammation and shortened telomeres (29, 58). Further, research has suggested that the targets for certain cancer treatments negatively impact biological markers of aging, e.g., increases in tumor suppressor mechanisms through the P53 pathway are associated with increased cell senescence systemically (59). Tamoxifen has also been shown to be genotoxic and other endocrine therapies may be associated with increased DNA damage because of decreased antioxidant capacity (60). Finally, all of the above processes have been implicated in cognitive decline and the development of neurodegenerative diseases (29, 58). This research suggests that biological processes underlying cancer, the impact of cancer treatments, aging, neurodegeneration and cognitive decline are linked, leading to the hypothesis that cancer treatments may accelerate the aging process (58).
In addition to examining specific pathways associated with aging, theoreticians have proposed systems theories of aging which provide interesting insights and hypotheses regarding cognition and cancer treatment. The reliability theory of aging is an example of a model of aging that is not specific to a particular biological process, but is consistent with a systems biology perspective (64). Reliability theory proposes that complex biological systems have developed a high level of redundancy to support survival. In a highly redundant system, failure of one or more components will not result in system failure if other components are available to support a specific pathway. Therefore, aging is determined by the failure rate of systems (loss of redundancy). Loss of redundancy is influenced by the initial extent of system redundancy (primarily genetically determined), the systems repair potential, and factors that increase failure rate such as poor healthcare, lifestyle risk factors, and/or exposure to environmental toxins. Someone with a low failure rate and/or high repair potential will show fewer signs of biological aging as they age chronologically, whereas someone with a high failure rate and/or low repair potential will age more rapidly as evidenced by the development of a disease associated with a specific set of system failures or frailty with a patchwork of failures across multiple systems; hence the difference between chronological and biological aging.
One implication of reliability theory is that vulnerability to post-treatment cognitive change does not necessarily depend upon a given treatment affecting a specific biological pathway. Rather, different patterns of failure rate (redundancy loss) across various biological systems may confer more or less vulnerability to specific treatments for each individual. Therefore, one patient may be vulnerable to the DNA damaging effects of a chemotherapy regimen, whereas another patient may be vulnerable to the impact on the hormonal milieu of endocrine treatments. This vulnerability may be strongly influenced by the pattern of systems failure prior to cancer diagnosis.
Further, investigators have assumed that long-term cognitive problems result from the lack of recovery from the acute effects of treatment, but remain stable after initial recovery (26). However, viewed within the context of models of aging, two additional hypotheses emerge. First, the initial effect of cancer treatment may produce a cascade of biological events which causes continued cognitive decline with aging. Second, a given treatment may not be sufficient to cause enough redundancy loss to immediately effect cognitive function but may produce a delayed effect as aging continues. Support for each of these patterns was reported by Wefel et al (22) who studied patients treated with regimens that included 5-fluorouracil (5-FU): 1) stable cognitive functioning over time after an acute post-treatment decline; 2) Continued cognitive decline over one year; and 3) no acute cognitive decline with new evidence of cognitive decline at one year post-treatment.
Other investigators have proposed that normal aging is a curvilinear process with relatively little change in young adulthood to middle age and increasing levels of decline in older adults (62). The slope of change in older adults is influenced by a variety of factors including cognitive reserve, diet and exercise, comorbid conditions, and genetic factors (e.g., APOE and COMT). Therefore, even if a given cancer treatment like chemotherapy has the same impact on brain resources across the lifespan, the effect on cognitive performance may differ depending on the age of the individual and the slope of cognitive aging. This point is illustrated in Figure 1 where the same change in brain resources can have a minimal effect on cognitive performance in a young adult (a), a moderate effect in an older adult with high cognitive reserve (b) and a greater effect on an older adult with low cognitive reserve (c). This model may partially explain the emerging pattern of results from imaging studies which seem to indicate group effects of chemotherapy on brain structure and function (suggesting a more consistent effect across individuals) whereas the neuropsychological studies suggest that only a subgroup of patients demonstrate a decline in cognitive performance.
These considerations suggest the need for studying the short and long-term effects of cancer treatments in older cancer patients. Despite the fact that the majority of breast cancer patients are diagnosed at age 65 and older and that the number of older breast cancer survivors is growing dramatically, nearly all of the published research has focused on younger breast cancer patients (mean age <60). Longitudinal studies (9) suggest that older breast cancer patients experience objective cognitive declines shortly after treatment; however, larger scale prospective studies are needed. Additionally, a cross-sectional study of older (>65) long-term breast survivors found lower performance on measures of executive function, working memory, and divided attention as compared to healthy controls (63).
Although the recent focus of research has been on longitudinal studies with pretreatment assessments, data suggesting the possibility of continued or delayed cognitive decline demonstrates the critical need for studies examining the impact of cancer and cancer treatments on the trajectory of age-associated cognitive change, particularly in older long-term survivors. Cross-sectional studies suggest that older, long-term cancer survivors will have lower performance in various areas of neurocognitive functioning as compared to matched older adults without a cancer history (63–65). However, longitudinal assessments are important to define whether age-associated declines parallel those of older adults with no cancer history (Phase Shift Hypothesis) or a steeper slope of decline (Accelerated Aging Hypothesis) (Figure 2). These are not mutually exclusive hypotheses in that one group of survivors may demonstrate the Phase Shift pattern whereas another, vulnerable population may demonstrate the Accelerated Aging pattern. Further, it is critical to define whether the impact on the trajectory of cognitive aging is the same for someone treated as a younger versus older adult.
Although longitudinal studies that include pretreatment assessments continue to be important to answer certain questions, there is an urgent need to study the growing number of older women who are long-term survivors of breast cancer treatment in order to evaluate the trajectory of change of cognitive function with aging. Beginning a large-scale study with pretreatment assessment with the goal of studying the impact of cognitive functioning on long-term survivors is logistically difficult and would delay answering critical questions for 10 or more years. Further, difficulties inherent in the recruitment of patients at diagnosis makes selection on critical factors (e.g., cognitive reserve, age, smoking history) extremely difficult; leaving investigators to conduct post-hoc analyses with samples that may or may not be large enough to allow for sufficient power to test specific hypotheses. An advantage of examining a large survivor cohort is that groups with specific characteristics like age, level of cognitive reserve, smoking history etc can be recruited so that specific hypotheses can be tested.
If cancer treatments accelerate the aging process, some overlap in brain structures affected by cancer treatments and aging would be expected. Imaging studies have demonstrated that total gray matter volume reliably decreases with advancing age (beginning in the mid-40s), with regional changes exhibited mainly in frontal cortex and in regions around the central sulcus (66). Global white matter decreases with advancing age with a trend for anterior white matter integrity decreasing earlier than posterior sites have been found (66–67). As described above, similar areas of the brain are affected by chemotherapy. Therefore, change in brain structure and function may be an interaction between the effects of cancer treatments and changes associated with aging.
If cancer treatments accelerate aging, then it is reasonable to assess whether these treatments increase the risk of dementia. Three studies utilized SEER data linked to Medicare claims data to evaluate the increased risk of dementia in breast cancer survivors who were or were not exposed to chemotherapy (68–70). One study (68) found a significantly increased risk for dementia for survivors who had been exposed to chemotherapy (hazard ration-1.20, 95% confidence interval-1.08–1.33), whereas the other two found no association between chemotherapy exposure and risk of dementia (69–70). An earlier twin study reported that the twin who was a cancer survivor was twice as likely to develop dementia as their co-twin, although this difference was not statistically significant (71). The authors of the first three studies recognized the limitations of Medicare claims data for the diagnosis of dementia and recommended further studies with formal neuropsychological assessments. Further, none of these studies evaluated whether exposure to chemotherapy increased risk for dementia in people who had other risk factors for dementia (e.g., APOE4 positive). Therefore, exposure to cancer treatments may not increase risk of dementia generally, but only in those people with existing risk factors for dementia.
Few studies designed to evaluate interventions to treat cognitive changes have been reported, although a number of treatment trials are in process. Two studies have found support for the efficacy of modafinil, a psychostimulant, in improving memory and attention and reducing fatigue (72–73). Cognitive rehabilitation approaches are also being developed with initial reports of positive results (74). A recent review of factors associated with prevention of cognitive decline with aging reported evidence for cognitive training, physical exercise and possibly diet as efficacious interventions (75). These data suggest the value of testing exercise and dietary interventions to preserve cognitive function in cancer survivors.
An impressive body of research supports the hypothesis that a subgroup of breast cancer patients is vulnerable to post-treatment cognitive changes, although not exclusively related to chemotherapy. Further, imaging and animal model studies are providing accumulating evidence for putative mechanisms for chemotherapy-induced cognitive change. Additional research with other cancer groups and treatment modalities (e.g., androgen ablation for prostate cancer) is growing; however, significantly more research is required in order to determine if the findings in breast cancer are generalizable to other cancer types and treatments. Finally, models of aging are reviewed to suggest that a useful perspective is to view cognitive changes associated with cancer and cancer treatments in the context of factors that affect the trajectory of normal aging.
Supported by grants (R01 CA87845, R01 CA101318, R01 CA129769, and U54 CA132378) from the National Cancer Institute, Bethesda, MD and from the Starr Foundation.
Based on the presentation given as the Wiley-Blackwell invited speaker for the IPOS World Congress 2011
I have no conflict of interest to report with regard to the preparation of this manuscript.