Cognitive function after cancer and cancer treatment

INTRODUCTION — There are over 18 million cancer survivors in the United States and over 40 million survivors worldwide [1,2]. This number is expected to grow due to improvements in cancer screening, increases in life expectancy following definitive cancer treatment, and the aging of the population. Despite these improvements, the late effects of treatment, in particular those related to chemotherapy, have become more concerning to patients. These late effects may include what patients have termed "chemobrain" or "chemofog," which can persist in some patients long after treatment has ended. It is also referred to as cancer-related cognitive impairment (CRCI) and chemotherapy-induced cognitive impairment (CICI) in the literature.

This topic will discuss cognitive function changes after cancer treatment, particularly as they relate to chemotherapy. Immunotherapy is now also a treatment modality for various cancers and can have adverse neurologic effects and fatigue, but there are no clinical studies as of yet that focus on cognitive functioning.

Cognitive function following craniospinal irradiation, as well as other issues in cancer survivorship, are discussed separately. (See "Delayed complications of cranial irradiation" and "Management of late complications of head and neck cancer and its treatment" and "Overview of cancer survivorship care for primary care and oncology providers".)

IMPACT OF CANCER AND CANCER TREATMENT ON COGNITIVE FUNCTION

Neurobiologic basis — The mechanisms by which cancer and cancer treatment impact cognitive function have not been elucidated. The following mechanisms have been proposed:

Effect of cancer therapy on trajectory of cognitive aging — Some investigators have proposed that cancer treatment can either shift the trajectory of normal cognitive aging or accelerate the aging process [3]. Evidence that chemotherapy-related cognitive dysfunction is accompanied by structural changes in the central nervous system comes from cross-sectional magnetic resonance imaging (MRI) studies, most of which included breast cancer patients who were treated with chemotherapy months to years earlier; the control groups consisted of healthy individuals and breast cancer patients not treated with chemotherapy [4-6]. Patients who received chemotherapy were more likely than controls to manifest a long-term persisting diffuse decrease in gray matter and white matter volumes post-treatment, and these changes correlated with neurocognitive deficits.

Other research suggests that chemotherapy-related cognitive impairment may be related not only to altered neuronal integrity, but also to disruption of brain structural networks that process and integrate information across the brain. In a prospective study of 64 males with newly diagnosed testicular cancer who underwent orchiectomy, patients underwent baseline and six-month follow-up neurocognitive testing and MRI scanning, which provided data to construct theoretical brain structure networks [7]. Twenty-two males received cisplatin and 42 had surveillance only. At six-month follow-up, compared with the surveillance group, males who had chemotherapy exhibited altered neuronal integrity and disruption of structural networks, as well as poorer cognitive performance. Patients with lower cognitive reserve appeared to be more vulnerable to these changes and to postchemotherapy cognitive decline.

It is likely that many mechanisms contribute to cognitive decline in cancer patients, including inflammation [8], direct neurotoxic effects of treatments, and damage to progenitor cells, among others [9]. In addition, it is also likely that the observed cognitive dysfunction in some patients reflects an interaction between multiple factors. As an example, in one review of the literature, the multifactorial nature of cognitive deficits in breast cancer patients was discussed [10]. Factors other than chemotherapy were deemed to be important beyond the impact from surgery or medical therapy. These included psychological states (eg, depression, anxiety), the presence of fatigue or menopause, and other comorbid conditions. These data highlighted the need for larger studies to better understand both the magnitude and mechanism of cognitive dysfunction in this population. Another potential interacting factor is lower pretreatment cognitive reserve, which may make patients more vulnerable to postchemotherapy cognitive decline [7,11,12].

It is important to note that the incidence of cognitive changes (due to chemotherapy predominantly) has been primarily conducted in patients treated with either hematological or breast malignancies. Earlier studies were retrospective and primarily within a year of undergoing treatment. Later studies have included prospective designs and longer follow-ups, with attempts to identify risk factors for developing chemotherapy-related cognitive changes as well as identify the underlying mechanisms.

Soil, seed, and pesticide model — Others have conceptualized the interaction between cancer therapy and cognitive function as a "soil, seed, and pesticides" model. In this model [13]:

•The "soil" includes host-related factors (ie, what the individuals themselves provide in the development and treatment reactivity to cancer) such as genetics, immune reactivity, nutrition, and cognitive reserve.

•The "seed" refers to disease-related factors, including the cancer itself. It also includes tumor genetics, mutations, cytokines, and paraneoplastic disorders.

•The "pesticides" include treatments for the cancer. Although chemotherapy is used in the model, one could also extend it to include hormonal treatments and radiation therapy (RT). The mechanisms by which chemotherapy can impact cognition and brain function are not entirely understood. Many factors have been suggested, but none are mutually exclusive to the risk.

Specific cognitive complaints — Individuals who have been treated with chemotherapy may bring out issues related to cognition in a variety of ways. As examples, patient complaints taken from our practice include:

●"I have complete memory blocks"

●"My mind feels like jello…"

●"Some days I feel like a ditz"

●"I cannot multitask"

●"I lose my place and attention when trying to read"

Studies have reported a fairly consistent pattern regarding the experience of cognitive impairment, such that declines tend to be in the realms of memory, processing speed, and complex attention [14]. However, there has been a lack of consistency between studies assessing cognitive outcomes in cancer patients with respect to design and defining cognitive impairment. For example, one comprehensive review of the literature highlighted the discrepancies in the assessment of cognitive functioning, particularly among patients who have undergone chemotherapy [14,15]. These include:

●A lack of a set standard to define cognitive impairment

●Variability in which neuropsychological tests were utilized

●The range of statistical analyses employed across studies

Indications for referral — While cognitive complaints may be subtle, they should be taken seriously by the patient's clinicians and medical providers. If the complaints persist months after the cessation of treatment and are impacting a patient's quality of life, including performance at work or home, a referral for neuropsychological evaluation is indicated in order to help with the differential with respect to the etiology of the cognitive issues and to help guide subsequent management.

EVALUATING COGNITIVE FUNCTION — The routine assessment of cognitive function (either before or after treatment) is not a standard practice, and evaluation is typically reserved for patients in whom cognitive impairment is negatively impacting quality of life or functional performance.

In general, assessments are mostly done as part of clinical research rather than as part of routine clinical practice. Most investigators include measures of neuropsychological functioning that characterize both cognitive domains and intellectual function at baseline, during treatment, and into follow-up [16]. As with most clinical research, control for bias is important, which is done through the use of prospective designs, incorporating control groups, and assessing for any cognitive decline compared with baseline using validated measures.

There are a number of methods to assess changes in cognitive function. One example is called the reliable change index (RCI). The RCI is the difference between two obtained scores for an individual divided by the standard error of the difference of the test utilized. It is a quantification of how much an individual score has changed and whether or not that change is reliable and clinically significant. The RCI has since been utilized by a variety of researchers. However, other investigators have made adjustments to the index [17].

Other methods to evaluate cognitive function in patients with cancer are available, such as the Mini Mental State Examination and the Montreal Cognitive Assessment. Further details on these assessments are discussed separately. (See "Patients with cancer: Clinical features and diagnosis of cognitive impairment and delirium", section on 'Cognitive impairment' and "Evaluation of cognitive impairment and dementia", section on 'Cognitive testing'.)

SPECIFIC CANCERS — Cognitive dysfunction can be apparent in various cancers at the time of diagnosis and can range from as high as 90 percent in patients with a history of a brain tumor, 20 to 30 percent in adult patients with leukemia, to between 17 and 75 percent in patients following breast cancer [18]. The data associated with cognitive impairment and these tumors are discussed below.

Brain tumors — There are few studies assessing the cognitive impact from the treatment of brain tumors, as it is difficult to parcel out the roles of the tumor itself and cranial irradiation effects with respect to any neuropsychological changes. (See "Delayed complications of cranial irradiation", section on 'Neurocognitive effects'.)

However, as the disease-free survival rate improves for these patients with the use of adjunctive chemotherapy and immunotherapy, the cognitive outcomes will likely become more important to assess. The impact of treatment response on cognition was shown in a phase II study in patients with recurrent glioblastoma treated with bevacizumab, all of whom underwent neuropsychological testing at baseline and every six weeks posttreatment to assess cognition [19]. Those who showed an objective response and survival over six months showed cognitive impairments at baseline relative to the general population, but had improved scores at the 24-week time point. Those who had progressive disease showed a decline.

Hematologic malignancies — Patients undergoing hematopoietic cell transplantation (HCT) may develop evidence of post-transplant cognitive decline, although in many cases, baseline cognitive dysfunction is present prior to HCT, possibly because they have previously received chemotherapy and/or radiation [20-22]. As an example, in one prospective study where a fairly liberal definition of cognitive decline was utilized (performance one standard deviation below the mean compared with normative data), 58 percent of patients were classified as having a baseline cognitive impairment prior to HCT [20].

In the setting of HCT, at least some data suggest that the intensity of the conditioning regimen influences the risk and timing of post-treatment cognitive impairment. In a longitudinal prospective study of 477 HCT recipients (236 autologous, 128 reduced-intensity allogeneic, 113 standard myeloablative allogeneic), those undergoing a full-intensity allogeneic HCT were at greatest risk of cognitive decline post-treatment, which was present at six months and persisted beyond three years [12]. Individuals undergoing reduced-intensity allogeneic HCT had evidence of delayed decline in cognitive functioning (primarily in visual memory and fine motor dexterity) that did not emerge until three years post-transplant, and cognitive functioning was largely spared in most patients after autologous HCT. At three years post-HCT, global cognitive impairment was present in 19 percent of autologous HCT recipients compared with 36 percent of allogeneic HCT recipients. Older age, male sex, lower education, income, and pretreatment cognitive reserve were all associated with post-HCT cognitive impairment.

Others have noted the following risk factors for post-treatment cognitive decline: cognitive reserve [11], educational level [23], disease and health status [11,20,23], use of total body irradiation (TBI) as a component of the conditioning regimen [24], chronic graft-versus-host disease, and fatigue [23]. Affective state has not necessarily emerged as a contributor to cognitive changes/impairments in group studies [11,22], though in individual patients, this can certainly be a factor.

Follow-up studies have shown that cognitive dysfunction may improve over time, but this has mainly been observed in individuals undergoing autologous HCT [11,23,25-30]. Studies that focused on allogenic HCT mostly show a decline in cognitive impairment over time with limited recovery [24,31,32]. Special care must be given to the psychological health of the transplant survivor and that of their spouse/partner, and involvement of mental health providers in the long-term care of the transplant survivor is essential. Recommendations for post-transplant evaluation and management are discussed in detail elsewhere. (See "Long-term care of the adult hematopoietic cell transplantation survivor", section on 'Mental health'.)

Breast cancer — The cognitive late effects of chemotherapy have been studied extensively in patients treated for breast cancer for both chemotherapy and/or endocrine therapy.

●Chemotherapy – Patients with breast cancer who receive chemotherapy report higher impacts on cognitive function compared with those without cancer. As an example, in one case-control study involving 581 patients with breast cancer and 364 noncancer controls, 37 percent of patients reported a decline in cognitive function at six months follow-up after chemotherapy relative to pretreatment, versus 14 percent of controls over the same time period [33].

Studies have also identified specific clinical factors associated with a higher risk of cognitive decline in patients receiving chemotherapy, such as baseline fatigue, anxiety, depression, decreased cognitive reserve, and functional well-being [33,34]. In one study, older individuals with lower cognitive reserve exposed to chemotherapy had lower scores on tests of processing speed compared with controls and a group that did not have chemotherapy [34].

In addition, other factors may predict for an increased predisposition to cognitive difficulties, such as changes in brain white matter microstructure [35] and specific molecular alterations [36]. As an example, one study which assessed both cognitive functioning and functional magnetic resonance imaging (MRI) documented slower processing speed and poorer reported executive functioning in breast cancer patients treated with chemotherapy compared with normal controls; those with an apolipoprotein E4 (APOE4) allele had significantly slower processing speed [36]. Other candidate genes (COMT, MDR1, BDNF and GST) did not influence cognitive performance. However, impaired functional connectivity was related to genetic polymorphisms among patients with breast cancer [36].

Cognitive impairment was also noted primarily in the domains of verbal and visuospatial abilities in one meta-analysis of 17 studies focusing on survivors treated with standard-dose chemotherapy for longer than six months [37]. Relevant modifiers (age, education, time since treatment, adjunct hormonal treatment) did not emerge as significant modifiers.

●Endocrine therapy – Beyond chemotherapy, data suggest that endocrine therapy (eg, selective estrogen receptor modulators and aromatase inhibitors) can negatively impact cognition [38-40]. However, the data have not been consistently reported and require further evaluation [41,42]. Imaging studies in this patient population have also been conducted and provide additional evidence of the potential impact of cytotoxic chemotherapy on the brain [43-48].

In a randomized trial that included 552 patients with breast cancer assigned to adjuvant endocrine therapy with or without chemotherapy, all patients had a decline in self-reported cognitive function, but it was greater in those who additionally received chemotherapy at the three- and six-month follow-up assessments. However, these findings were not sustained at one year or beyond [49].

Ovarian cancer — Evidence suggests that some patients may experience cognitive decline with chemotherapy treatments for ovarian cancer, though further study is needed to identify those patients at highest risk of impairment.

A study of 230 patients with newly diagnosed ovarian cancer utilized web-based cognitive screening assessments and self-reporting to examine processing speed, attention, and motor reaction times before, during, and after chemotherapy [50]. In this study, approximately 25 percent of patients experienced a decline in at least one of these areas during treatment, though the proportion fell to 18 percent six months following therapy. Typically, impairments were limited to just one area of function. Exploratory analyses suggested that cognitive decline with chemotherapy was associated with patient’s age and education level.

In another study of 18 patients with ovarian cancer, participants underwent structural and functional brain MRI and neuropsychological testing within four months of completing chemotherapy [51]. Compared with an age-, sex-, and education-matched control group, patients with ovarian cancer had apparent structural and functional alterations in the frontal and parietal brain regions, but there were no differences in neuropsychological measures between the two groups.

Colon cancer — Patients with colorectal cancer may experience cognitive decline relative to healthy controls irrespective of chemotherapy treatment and stage of disease.

A study of almost 300 patients with localized colorectal cancer and approximately 70 healthy controls assessed the effect of colorectal cancer on cognition using neuropsychological assessments at baseline (prior to chemotherapy, if given) and at six-month intervals [52]. Cognitive impairment was more frequent at baseline among the colorectal cancer patients than the healthy controls (43 versus 15 percent), and the difference persisted at the later assessments (6 and 12 months). Attention, memory, verbal learning, and processing speed were the most affected parameters. Among patients with colorectal cancer, there were no differences in cognitive assessments at any time point between those who received chemotherapy and those who did not. Additionally, 73 patients with metastatic cancer were compared with the patients with localized cancer and found to have similar levels of cognitive impairment. There were no associations between other modifiers (cognitive symptoms, fatigue, cytokine levels, and APOE4 allele) and cognitive impairment.

Prostate cancer — The effects of androgen deprivation therapy on cognitive function in prostate cancer survivors are discussed separately. (See "Overview of approach to prostate cancer survivors".)

Childhood cancers — In general, individuals who were treated at a young age with higher doses of cranial irradiation may incur greater deficits [53]. Fortunately, due to changes in treatment protocols, the late cognitive effects in individuals treated for leukemia tend to be lower today than the original reports that dated back to the early 1980s. Multiple factors are associated with the reported incidence of cognitive decline, including the age of the child during treatment and type of treatment received. In addition, a preexisting history of a learning disability may make someone more vulnerable to developing late cognitive effects. These issues are particularly relevant given other data suggesting that childhood cancer survivors may have an increasing incidence of frailty, which may represent a more rapid aging process compared with their peers [54].

In adolescent and adult survivors of acute lymphoblastic leukemia (ALL), neuropsychological late effects may be apparent and seem to be dependent on the factors as above [55]. In studies assessing the late effects of chemotherapy (which is used as a central nervous system [CNS] prophylaxis), primarily intrathecal methotrexate (IT-MTX), the results are variable. There are those studies that have found cognitive deficits or lower scores compared with normative data in these patients [56,57], particularly among those with a greater number of intrathecal injections [58], and others where no major deficits were observed [59,60]. One study has implicated folate depletion and homocysteine elevation as possible mediators of cognitive impairment in the ALL survivors [61]. Genetic polymorphisms in methionine synthase as well as in genes related to oxidative stress have also been associated with cognitive deficits in survivors of childhood ALL [62,63].

INTERVENTIONS — Since the etiology of potential cognitive inefficiencies postchemotherapy is not well established, it is difficult to identify possible modifiers. In the meantime, interventions are focused on behavioral strategies and psychopharmacological approaches.

In cancer survivors who report persistent cognitive concerns after completing therapy, a combination of both behavioral and pharmacological treatments may result in improved cognitive functioning and perceived improvement in quality of life. Optimizing modifiable factors (ie, improved sleep, regular exercise, good nutrition, reducing stress) can lead to subjective improvements. Pharmacological interventions target specific symptoms and are patient-specific. However, a review of both pharmacological and non-pharmacologic interventions favored the latter in breast cancer patients [64-67]. In addition, psychoeducation of patients and partners has been shown to help improve affective state and develop coping strategies [68,69].

Behavioral interventions — Interventions aimed at developing compensatory strategies for cognitive issues prior to or during treatment for cancer have been studied in breast cancer populations [64,66,70-72] and in children treated for leukemia [73-75]. However, there are few randomized clinical trials assessing many of these interventions in cancer patients or survivors. Some cancer centers are now offering cognitive rehabilitation either in group or individual settings during and/or after treatment in order to help patients devise compensatory strategies proactively, which may not only help cope with cognitive impairment, but also improve distress.

Examples of cognitive and behavioral remediation include:

●Relaxation – Relaxation and other mindfulness training is an emerging area of interest. Mindfulness-based relaxation techniques teach individuals to focus their attention and awareness in their everyday life as a way to manage their emotions and aspects of cognition. This is sometimes recommended given the usefulness of mindfulness treatment for attentional disorders [76]. While there are studies assessing its usefulness in improving the quality of life in cancer patients [77], there are no high quality data for whether these techniques improve cognition.

●Physical exercise programs – Physical exercise is a reasonable recommendation for patients with cancer who have cognitive concerns [78]. Although physical exercise is associated with improved cognition in a variety of studies (notably in children and older adults), randomized studies in patients who are treated for cancer are limited [79]. However, studies in breast cancer patients found that exercise interventions are associated with improved cognitive function [80-84]. Further details on role of physical activity in cancer survivors are discussed separately. (See "The roles of diet, physical activity, and body weight in cancer survivors".)

●Occupational therapy – This is usually referred to as cognitive rehabilitation, which is typically done by some speech therapists, occupational therapists, and neuropsychologists. (See "Physical rehabilitation for cancer survivors", section on 'Cognitive dysfunction'.)

●Brain-training programs – Brain-training computer-based programs (sometimes referred to as neurobics) use repetitive mental exercises to increase the brain's performance. These programs are commercially available but may be quite expensive.

Several randomized trials suggest that these computer-based programs may improve and maintain cognitive function. In one randomized trial of mostly breast cancer patients treated with chemotherapy, computer-assisted cognitive rehabilitation in the presence of a neuropsychologist improved cognition and working memory compared with other interventions, such as home cognitive exercises or phone follow-up [85]. Similar results were seen in another randomized trial also conducted in breast cancer patients, in which memory and speed of processing computer training was compared with a wait list control [64]. While both showed some efficacy, speed of processing improvements resulted in transfer to other domains. In a more diverse group of cancer survivors treated with chemotherapy who reported persistent cognitive symptoms, a home/web-based cognitive rehabilitation program improved cognitive symptoms compared with standard care (no intervention) [86]. A separate randomized clinical trial in childhood survivors of acute lymphoblastic leukemia showed that computerized training benefits on cognition were maintained six months following training [87].

●Electroencephalography (EEG) biofeedback – This technique, also termed neurofeedback, was assessed in a small group of breast cancer patients with self-reported cognitive impairments [67]. The treatment resulted in improvement in cognitive measures, fatigue, psychological scales, and sleep. These improvements were maintained at four weeks posttreatment.

●Group cognitive rehabilitation intervention – A randomized clinical study of 48 cancer survivors demonstrated that small-group intervention, including psychoeducation and cognitive exercises, led to improvements in patients’ self-reported cognitive complaints as well as memory on neurocognitive testing [88].

●Psychoeducation – This entails educating patients regarding the possible cognitive late effects of treatment and is important as they transition from cancer patients to cancer survivors.

Pharmacologic interventions — Pharmacological treatment may be an appropriate intervention in order to improve cognition in patients previously treated for cancer. However, the data to support its use are limited, and no randomized clinical trials have been performed [89-92].

●Central nervous system (CNS) stimulants – Modafinil and methylphenidate are two CNS stimulants that have been evaluated in patients with cognitive declines after cancer treatment [89,91,92]. The limited data suggest they may improve cognitive function, although formal randomized trials specifically evaluating their impact on cognitive function have not been performed.

●Donepezil – Donepezil, a reversible acetylcholinesterase inhibitor, improved multiple areas of cognition (including attention/concentration, verbal and figural memory) at 24 weeks compared with baseline in one phase II trial of 35 patients with a brain tumor [90]. The use of this agent and other reversible acetylcholinesterase inhibitors for mild cognitive impairment in the general population are discussed separately. (See "Mild cognitive impairment: Prognosis and treatment", section on 'Acetylcholinesterase inhibitors'.)

Further management options for mild cognitive impairment in the general population are discussed separately. (See "Mild cognitive impairment: Prognosis and treatment".)

SUMMARY AND RECOMMENDATIONS

●For patients with cancer, the late effects of treatment, in particular those related to chemotherapy, may include cognitive problems, which have been termed "chemobrain" or "chemofog." These issues can persist in some patients long after treatment has ended. (See 'Introduction' above.)

●The mechanism by which cancer and cancer treatment impacts cognition has not been entirely elucidated. Some propose that cancer and cancer treatment may have an adverse effect on age-related cognitive changes. Others have conceptualized the interaction as a "soil, seed, and pesticides" model. In this model (see 'Impact of cancer and cancer treatment on cognitive function' above):

•The "soil" includes host-related factors (ie, what the individuals themselves provide in the development and treatment reactivity to cancer) such as genetics, immune reactivity, nutrition, and cognitive reserve.

•The "seed" refers to disease-related factors, including the cancer itself. It also includes tumor genetics, mutations, cytokines, and paraneoplastic disorders.

•The "pesticides" include treatments for the cancer. Although chemotherapy is used in the model, one could also extend it to include hormonal treatments and radiation therapy (RT). The mechanisms by which chemotherapy can impact cognition and brain function are not entirely understood. Many factors have been suggested, but none are mutually exclusive to the risk.

●The routine assessment of cognitive function (either before or after treatment) is not standard clinical practice, and evaluation is reserved for patients in whom cognitive impairment is negatively impacting quality of life or functional performance. (See 'Evaluating cognitive function' above.)

●Cognitive dysfunction can be apparent in various cancers at the time of diagnosis and can range from as high as 90 percent in patients with a brain tumor, 20 to 30 percent in adult patients with leukemia, to between 17 and 75 percent in patients with breast cancer. (See 'Specific cancers' above.)

●In cancer survivors who report persistent cognitive concerns after completing therapy, a combination of both behavioral and pharmacological treatments is likely to result in improved cognitive functioning and perceived improvement in quality of life. Optimizing modifiable factors (ie, improved sleep, regular exercise, good nutrition, reducing stress) can lead to subjective improvements. Pharmacological interventions target specific symptoms and are patient-specific. In addition, psychoeducation of patients and partners has been shown to help improve affective state and develop coping strategies. (See 'Interventions' above.)