Abstract
Context
Researchers have described a constellation of cognitive deficits (e.g., impairments in executive functions, working memory, attention, and information-processing speed) associated with cancer treatment, and specifically chemotherapy, for non-central nervous system tumors. However, findings have been inconsistent, largely because of measurement and study design issues.
Objectives
To propose ways for researchers to more clearly delineate and characterize the mild cognitive deficits and related outcomes that appear to affect a subset of cancer patients and suggest methods to make more effective use of the existing data to understand risk factors for impaired neuropsychological functioning.
Methods
We examined the literature on the relationship between chemotherapy and cognitive impairment, as well as related literature on neuropsychological measurement, structural and functional neuroimaging, alternative measures of health outcomes, and integrative data analysis.
Results
A more comprehensive picture of cognitive functioning might be obtained by incorporating nontraditional ecological measures, self-reports, computational modeling, new neuroimaging methods, and markers of occupational functioning. Case-control and integrative data analytic techniques potentially could leverage existing data to identify risk factors for cognitive dysfunction and test hypotheses about the etiology of these effects.
Conclusion
There is a need to apply new research approaches to understand the real-world functional implications of the cognitive side effects of chemotherapy to develop and implement strategies to minimize and remediate these effects before, during, and after cancer treatment.
Key Words
Is it permanent? I had chemo 11 years ago and I'm still in a fog.
-- Cancer survivor posting on Facebook
Introduction
When the neurotoxic effects of antineoplastic drugs were first recognized nearly 40 years ago, it was assumed that neurotoxicity was secondary to central nervous system metastasis as it was generally believed that most cytotoxic drugs did not cross the blood-brain barrier.
1
, 2
However, in 1980, doctors at Dartmouth Medical School found that cancer chemotherapy patients who did not have metastases to the central nervous system were experiencing cognitive impairment that was independent of affective disorders or other psychiatric conditions.3
Since then, the neurotoxicity of many commonly used chemotherapeutic agents has been widely recognized,4
although the underlying mechanism has not been elucidated. The survivor community, however, has been acutely aware of these cognitive effects for some time, and it is widely believed that they coined the term “chemo brain” to describe the diffuse mental slowing and fogginess that often accompanies cancer treatment, and specifically chemotherapy, often for months or years.Numerous researchers have described a constellation of cognitive deficits associated with chemotherapy for non-central nervous system tumors. These include decrements in executive functions, working memory, information-processing speed, attention, concentration, reaction time, psychomotor speed, and visuospatial ability
2
(Table 1). Although impairment in these domains can range from subtle to profound, particularly for patients receiving high-dose chemotherapy,2
, 5
most studies, including five meta-analytic reviews,6
, 7
, 8
, 9
, 10
report deficits in the mild-to-moderate range. However, severe neurotoxicity has been documented in case reports of toxic leukoencephalopathy and progressive multifocal leukoencephalopathy occurring in patients receiving oral and intravenous chemotherapy.11
, 12
Table 1Cognitive Deficits Associated With Chemotherapy
| Executive function: Cognitive abilities required for goal-directed purposive behavior and adaptation to novel situations. Includes the ability to plan and organize tasks, problem solve, self-monitor, and initiate appropriate behavior. |
| Working memory: The ability to temporarily store and manipulate information needed for complex tasks, such as learning and reasoning. |
| Information-processing speed: The ability to perform cognitive tasks fluently and automatically, particularly under the demands of focused attention and concentration. |
| Attention: The ability to focus on a particular stimulus while ignoring competing stimuli. Involves the ability to sustain attention (concentration), divide attention between competing stimuli, and shift attention appropriately by determining where attentional resources should be directed. |
| Reaction time: The amount of time that elapses between experiencing a stimulus and initiating a response. |
| Psychomotor speed: Refers to the length of time between a decision and initiation of a response. Usually measured by reaction time. |
| Visuospatial ability: The ability to understand visual representations and spatial relationships in two and three dimensions. |
Findings from clinical studies have been inconsistent, largely because of wide variations in study design, methodology, and definitions of cognitive impairment.
13
Because researchers have operationalized cognitive impairment in different ways, it is often difficult to make cross-study comparisons. To illustrate the extent of this problem, Shilling et al.14
analyzed their breast cancer patient data using seven different methods of calculating cognitive impairment reported in the literature: impairment ranged from 12% to 68%, depending on the analytical method used. Measurement of cognitive impairment across studies is further hampered by differences in inclusion criteria, comparison groups (e.g., published normative data, healthy matched controls, non-chemotherapy-exposed cancer patients, and medical patients without cancer), clinical regimens, neuropsychological test protocols and cutoff scores, timing of cognitive assessments, and self-report instruments. Within patient samples, patients often differ with respect to stage of disease, concomitant treatments (e.g., pain medication and hormone therapy), and concomitant illnesses.Although most of the literature supports an association between chemotherapy and cognitive decline for a subset of cancer patients,
15
, 16
, 17
a few studies involving breast cancer and testicular cancer patients have failed to find one.18
, 19
, 20
Some have suggested that changes in cognition subsequent to chemotherapy are overdetermined and may represent normal variation in cognitive functioning, low levels of pretreatment cognitive functioning, effects of treatment-induced menopause, classification error, or artifacts of study design.19
, 20
, 21
Genetic variation in blood-brain transporters, DNA-repair genes, and neural repair genes (e.g., apolipoprotein E) also may be contributory factors.2
Others have proposed that chemotherapy actually may be associated with improvement in cognitive functioning for certain patients.21
, 22
It is important to remember that when considering improvement in cognitive performance, improvement should be compared with normal functioning (e.g., if a patient tests 2 SDs below the mean for processing speed one month after chemotherapy and improves to 1 SD below the mean by six months, the patient has improved but is still impaired relative to normal functioning). More recently, several studies have documented compromised cognitive function in a subset of patients before chemotherapy,22
, 23
, 24
raising the possibility that cognitive impairment previously ascribed to chemotherapy may be related, at least in part, to the disease process itself (e.g., cytokine dysregulation and inflammation), other sequelae of cancer treatment (e.g., fatigue and anemia), or the stress of having a cancer diagnosis (e.g., depression and anxiety). Whether these effects are short lived or long lasting is an ongoing question. Some studies have reported that cognitive changes experienced during and shortly after chemotherapy resolve within a year,25
whereas others have documented long-term (10 years) and delayed-onset impairment after chemotherapy.15
, 17
The combination of mixed results and generally mild deficits has prompted some to be skeptical about so-called chemo brain or to consider these deficits a minor side effect of treatment. Recognizing the complex nature of cognitive impairment and the complicated state of the literature, the purpose of this article was to propose new research approaches to study chemotherapy-associated cognitive changes. We describe how ecological measures of cognitive functioning, computational modeling, novel applications of neuroimaging techniques and biomarkers, and leveraging existing clinical data can advance our understanding of this phenomenon. These methods are not intended to replace neuropsychological testing but rather to complement conventional testing in order to identify potential risk factors for cognitive dysfunction and test hypotheses about the etiology of these effects.
Delineating the Nature of Cognitive Impairment
An ongoing source of confusion regarding the nature of cognitive impairment stems from the discrepancy between subjective self-reports of cognitive deficits and objective measures of neuropsychological functioning observed in numerous studies of patients who received chemotherapy.
16
, 19
, 21
, 26
, 27
, 28
Patients frequently perform within normal limits on neuropsychological tests—although often scoring lower than controls—yet report feeling that they are functioning well below their usual level. The reason for this discrepancy is open to interpretation, and there is likely no simple answer. This discordance is not unusual because poor correlations between objective measures and self-reports are not uncommon in neuropsychology.29
In fact, self-reports and neuropsychological test findings may each be a valid reflection of different dimensions of cognitive function;30
depending on the research question, one method of assessment may be more appropriate than the other. Emotional factors may significantly influence subjective reports of well-being, and a number of studies have found subjective cognitive complaints related to depression, anxiety, fatigue, and psychological distress.28
, 31
Use of neuropsychological tests that are not sensitive enough to detect mild cognitive deficits30
also may contribute to the lack of correspondence between objective and subjective assessments. To this end, the 2003 Workshop on Cognitive Impairment Associated with Chemotherapy recommended that future research focus on identifying neuropsychological tests that are relevant to everyday life and sensitive to subtle cognitive changes and appropriate self-report instruments.32
Conventional Assessment of Cognitive Functioning
Neuropsychological tests are considered the gold standard for measuring cognitive function, and a select battery of tests for assessing learning, memory, processing speed, and executive function in cancer patients has been recommended by the International Cognition and Cancer Task Force.
33
However, by using nontraditional measures, researchers may gain insight into how a patient may be able to compensate for a cognitive deficit by drawing on other strengths outside the examination room. Conversely, a patient may test well but encounter difficulty functioning in a busy office environment.The original purpose of neuropsychological testing was to localize and diagnose severe brain pathology, such as traumatic head injury, stroke, and dementia, which presents a distinctly different clinical picture from the mild cognitive impairment commonly seen in patients who have received chemotherapy. Whereas signs of head injury, stroke, and dementia are often obvious to observers, the typically subtle cognitive deficits associated with chemotherapy are frequently not visible and often not discernible to anyone other than the patient. As one breast cancer survivor lamented: “… it's not like I'm suddenly a dull normal. I'm still able to function. It's just the fine degree of memory or the speed at which I'd be able to recall information.”
34
(p. 226)Conventional neuropsychological testing may not be sensitive enough to detect the subtle changes in cognitive function, and especially executive function, typically experienced by chemotherapy patients.
35
As Lezak et al.36
(p. 612) noted: “A limited number of established examination techniques give the subject sufficient leeway to think of and choose alternatives as needed to demonstrate the main components of executive behavior.” Another concern is that certain neuropsychological tests may not assess the upper bounds of cognitive ability.37
Because many studies of cancer patients include highly educated individuals, ceiling effects of certain tests may obscure changes in cognitive function, leading one to underestimate the extent of impairment. This is critical because even small deficits in cognitive function for someone in an intellectually demanding or highly skilled occupation may have profound professional consequences.Complementary Approaches to Understanding Cognitive Functioning
Simply because a patient tests “within normal limits” does not mean that that he or she is able to function effectively outside the structured test situation. Researchers could glean additional insight into a patient's cognitive functioning by supplementing neuropsychological testing with ecologically valid assessment measures, computational modeling approaches, and novel applications of neuroimaging techniques and biomarkers.
Ecologically Valid Assessment
From a quality-of-life perspective, some of the most distressing problems reported by chemotherapy patients are related to declines in executive function. Problems with short-term memory, attention, planning, cognitive flexibility, and multitasking seem to afflict a subset of survivors.
2
Qualitative studies have documented the profound impact that these deficits have on social functioning, occupational performance, and general well-being.38
Dissatisfaction with certain traditional measures of executive functioning, such as those that assess the ability to plan and problem solve,
39
and ones that appear to have little relevance to everyday life, have led some to advocate using more ecologically valid measures.40
, 41
Conventional neuropsychological tests have “… obstacles that limit the degree to which they achieve [real world validity]…. For example, testing in a quiet environment may not reveal the problems that patients have with concentration or memory in their natural work or home environment with their numerous distractions.”- Campbell Z.
- Zakzanis K.K.
- Jovanovski D.
- Joordens S.
Utilizing virtual reality to improve the ecological validity of clinical neuropsychology: an fMRI case study elucidating the neural basis of planning by comparing the Tower of London with a three-dimensional navigation task.
Appl Neuropsychol. 2009; 16: 295-306
42
(p. 126)Ecological validity has been defined as “… the functional and predictive relationship between the patient's performance on a set of neuropsychological tests and the patient's behavior in a variety of real-world settings.”
43
(p. 16) Such tests may reveal deficits in memory, organization, motivation, and other executive functions that conventional tests may not detect. Just as tests have been developed to enhance the ecological validity of neuropsychological assessment for special patient groups, such as adolescents with traumatic brain injury,44
tests targeted to the real-world challenges faced by chemotherapy patients could be developed.The more well-known Rivermead Behavioural Memory Test,
45
Test of Everyday Attention,46
and Multiple Errands Test47
, 48
were developed to assess “real-world” functioning. For example, the Rivermead Behavioural Memory Test includes analogues of everyday situations that test immediate and delayed visual, visuospatial, and verbal memory. Because it was developed as a screening test for severe memory impairment, the original test was not sensitive enough to identify mild memory deficits.49
However, the extended version of the Rivermead Behavioural Memory Test, which doubled the amount of material to be remembered and increased the level of difficulty of certain items, can detect mild memory impairment–such as that due to medication or a stressor–in normal adults.50
, 51
The Test of Everyday Attention is a reliable and valid measure of selective attention, sustained attention, and attentional switching. It is sensitive enough to show age effects in normal adults and can be used to detect subtle differences in attention in neurologically intact individuals.49
Unlike the Rivermead Behavioural Memory Test and Test of Everyday Attention, the Multiple Errands Test is an in vivo test that “… captures the non-routine, problem-solving, planning, organization and initiative required for everyday functioning”48
(p. 213) by assessing the ability to conduct several ordinary tasks simultaneously. Patients are taken to a shopping mall and assigned a variety of routine tasks, such as purchasing a loaf of bread and recording the price of tomatoes. A simplified version of the Multiple Errands Test52
and a hospital version of the test53
also are available.Virtual reality simulation technology also may be adopted to enhance the ecological validity of neuropsychological assessment and rehabilitation in cancer survivors.
41
, - Campbell Z.
- Zakzanis K.K.
- Jovanovski D.
- Joordens S.
Utilizing virtual reality to improve the ecological validity of clinical neuropsychology: an fMRI case study elucidating the neural basis of planning by comparing the Tower of London with a three-dimensional navigation task.
Appl Neuropsychol. 2009; 16: 295-306
54
, 55
Using virtual reality technology, patients can be safely immersed in context-relevant environments, such as an office, classroom, or busy city street, and their neurobehavioral responses recorded. Because virtual environments “… allow for precise presentation and control of dynamic perceptual stimuli (visual, auditory, olfactory, gustatory, ambulatory, and haptic conditions), they can provide ecologically valid assessments that combine the veridical control and rigor of laboratory measures with a verisimilitude that reflects real life situations.”56
(p. 280) Because examiners can modulate task complexity and intensity, the degree of challenge in a particular task can be adjusted according to the response characteristics of the individual.55
An ecologically valid virtual environment should include tasks that reflect real-world activities, are representative of the target population, and are relevant to problems that a patient is facing (e.g., returning to work).56
This technology is particularly well suited for examining the cognitive domains of attention, memory, executive function, and visuospatial ability, which appear to be most affected by adjuvant chemotherapy.10
, 28
, 57
For example, virtual reality spatial navigation paradigms have been used to study how different levels of structure affect mental representations of space,58
sex differences in navigational strategies,59
the effects of age and dementia on route learning and memory,60
and neurodevelopmental processes in adolescents and adults.61
Researchers and clinicians should consider using this kind of technology with cancer patients who have undergone chemotherapy because it can simulate everyday challenges under highly controlled conditions.Of course, there are potential limitations to such ecologically valid tests. For one, some lack extensive normative and psychometric data, and there are only limited data regarding their sensitivity for detecting chemotherapy-associated cognitive change. Furthermore, because some measures may be too simple for cancer patients, new tests may need to be developed, or existing tests modified, for this population. There is also the question of whether ecologically valid tests yield substantively different conclusions than conventional cognitive neuropsychological tests. Although some cognitive aging studies have found that conventional tests adequately predict employment status
62
and Alzheimer's dementia patients' driving performance,63
other studies have found that ecological tests are better predictors of everyday functioning.40
, 64
In the absence of studies that have examined how ecological tests perform with cancer patients, we can only speculate that given the multiple burdens of everyday life occasioned by a major illness, cancer survivors may exhibit deficits that are considerably more severe than those observed in the neuropsychological testing environment. Despite these potential limitations, ecological measures can provide a window into the cognitive difficulties encountered by cancer patients.Computational Modeling
Traditional neuropsychological tests yield scores that quantify functioning in various cognitive domains, such as attention, memory, speed of information processing, and language, but they do not provide information on the multiple component processes that underlie task performance. For example, in the Rey Auditory Verbal Learning Test, 15 nouns are read aloud by the examiner for five consecutive trials, each trial followed by a free-recall test.
65
The summary score, which is the number of words correctly recalled, does not just reflect retrieval, but multiple processes, including attention, encoding, and decision making. Recognizing that cognitive performance comprises subcomponent processes, researchers developed computational modeling techniques to decompose tasks into their constituent processes.66
“The ability of these models to identify sources of cognitive performance variability has been a major advance because they allow noise in processing to be separated into sources that occur at different points in the stream of processing, from encoding to decision.”67
(p. 247) For example, computational modeling has been successfully applied to the Digit Symbol Substitution task that is commonly used in conventional neuropsychological testing.68
, 69
Although these approaches have been used to study cognitive processing in clinical disorders (e.g., anxiety and hypoglycemia),Roring RW. Age-related changes in information processing on tasks of perceptual speed. 2005. Electronic Theses, Treatises and Dissertations. Paper 1774. Available from http://diginole.lib.fsu.edu/etd/1774. Accessed November 5, 2012.
70
, 71
they have not yet been applied to cancer patients. Using computational modeling, one could more finely characterize some types of cognitive deficits experienced by chemotherapy patients and develop tailored cognitive retraining plans.Whereas computational modeling has been most commonly used for binary choice tasks, it also has been applied to cognitive tasks involving multiple-choice alternatives, which is more common in neuropsychological tests.
67
Although some cognitive domains assessed by conventional neuropsychological tests may not be amenable to computational modeling because of incompatible response formats, the potential of this approach to elucidate the specific cognitive processes impaired by cancer treatment and cytotoxic drugs should at least be considered. Used in conjunction with brain imaging (see below), these techniques might identify specific brain regions and cognitive processes affected by chemotherapy. This knowledge could potentially inform treatment decisions and patient retraining efforts.Neuroimaging Techniques and Biomarkers
Advances in neuroimaging and biomarker assessment have allowed researchers to more precisely assess changes in brain integrity. Although these techniques have been applied to study structural and functional changes in cognitive aging and dementia, fewer than 20 studies published to date have used these methods to study chemotherapy patients. Alterations in brain morphology and function after cancer treatment constitute some of the strongest evidence that chemotherapy impairs cognition and may provide clues for understanding the duration and reversibility of these changes. Before describing the potential of newer techniques to examine brain structure and function, we provide a brief synopsis of relevant studies.
Nine empirical studies have assessed reduction in brain volume in breast cancer patients using structural imaging with magnetic resonance imaging (MRI). These cross-sectional studies vary with respect to time of assessment, from a few months to 20 years after treatment.
72
, 73
, 74
, 75
, 76
, 77
, 78
, 79
, 80
Most of these studies found that chemotherapy patients had reduced brain volume compared with breast cancer patients who did not receive chemotherapy or with healthy controls. In some studies, both gray and white matter changes were observed, whereas in others, only gray matter volume was reduced. There also have been null findings.72
Interestingly, one prospective MRI study81
documented decreased gray matter density one month after completion of chemotherapy with partial recovery by 12 months, which underscores the importance of the timing of cognitive testing (see also 79
, 82
).Although these studies documented reductions in brain volume, they could not determine whether these changes represented dehydration, edema, or neural degeneration, all of which have implications for whether cognitive recovery is possible.
81
However, a newer MRI approach, diffusion tensor imaging (DTI), is able to characterize water diffusion and microstructure in biological tissues. This allows investigators to identify degradation of neural structures and determine whether axonal death and/or deterioration of the myelin sheath are involved.83
Four studies74
, 78
, 82
, 84
used DTI to study brain structure in breast cancer patients three months to 10 years after chemotherapy: All found evidence of neural damage, and two studies78
, 82
reported specific evidence of axonal and myelin degeneration. However, what we still do not know, and what DTI may be able to provide, is information about the onset of neural degeneration and how these changes vary with the type and dose of chemotherapy.DTI is less precise in areas of the brain in which neural fibers cross.
83
Fortunately, newer DTI techniques (e.g., q-space imaging and hybrid diffusion imaging), which involve more complex sampling and model fitting, can correct for fiber crossing and provide a clearer picture of neural degeneration.85
However, these newer techniques have yet to be used to study chemotherapy patients.Proton magnetic resonance spectroscopy (PMRS), which is used in conjunction with DTI, is an imaging technique that can ascertain whether white matter changes represent inflammation or axonal death by detecting changes in brain metabolites, specifically N-acetylaspartate (NAA). Although the initial PMRS studies of chemotherapy patients did not find changes in NAA
75
, 79
, 86
—likely because of small sample sizes and low-field scanning—a recent study using state-of-the-art scanning found reductions in NAA consistent with axonal degeneration.78
Because axonal death is irreversible, this raises questions about whether cognitive function can eventually be restored. Although these findings are provocative, they need to be replicated because they are based on a single study. Collecting neuropsychological test data in conjunction with PMRS would provide strong convergent evidence linking neural integrity and cognitive impairment in chemotherapy-treated patients.Functional imaging, such as positron emission tomography, single-photon emission computed tomography, and functional magnetic resonance imaging (fMRI), also has received limited attention from researchers studying the effects of chemotherapy on cognition. fMRI differs from the other functional measures in that it does not require a radioactive tracer or prolonged scan times, making it useful for tracking brain activity as patients perform cognitive tasks in the scanner. Only five studies have used functional measures to assess how task activity correlates with brain activity in chemotherapy patients.
77
, 87
, 88
, 89
, 90
In three of these studies, when chemotherapy patients performed as well as non-chemotherapy patients, their brain scans showed multiple areas of hyperresponsiveness; when they performed more poorly than non-chemotherapy patients, their brain scans indicated hyporesponsiveness.77
, 87
, 88
These findings suggest that hyperresponsiveness represents overcompensation whereby more brain regions are recruited to perform at a normal level. However, a fourth study found evidence of hyporesponsiveness only during executive function tasks,90
and the final study showed hyporesponsiveness during encoding but hyperresponsiveness during recall.89
The authors proposed that the hyporesponsiveness might indicate insufficient encoding, whereas the hyperresponsiveness might represent increased effort to recall material that was suboptimally encoded. These differences in brain activity for encoding and recall could be explored further by using computational modeling in conjunction with fMRI. In that way, patterns of regional brain activity could be correlated with specific cognitive processes involved in particular cognitive tasks. However, because there are so few functional imaging studies, the above findings need to be replicated.Functional connectivity MRI, also referred to as intrinsic functional connectivity,
91
uses fMRI data to compare levels of activity among brain regions and determine which areas communicate with one another. Certain brain networks are active even when an individual is not deliberatively engaged in a cognitive task.92
Brain networks that exhibit correlated fluctuations are referred to as “resting state networks.” The structures in the visual system are a good example of such a resting state network. Although these structures are located in different parts of the brain, they show changes in activity that are correlated.93
The strength of correlations between functionally coupled regions potentially can be used as an index of brain system integrity. Coherent fluctuations are consistently found in persons with normal cognitive function. In contrast, resting state is disrupted in patients with Alzheimer's disease and persons with mild cognitive impairment at high risk of dementia.94
However, no empirical studies of resting state have been conducted in chemotherapy patients. Such research might potentially lead to identifying a neurobiomarker for cancer patients at risk for treatment-related cognitive impairment. Even more intriguing, hybrid approaches that combine DTI, functional connectivity MRI, and conventional fMRI could potentially provide correlative evidence of neural degeneration, functional deficits in particular brain regions, and disruption in more global brain networks.In addition to imaging, biomarkers potentially could be used to study brain alterations associated with cancer treatment. For example, NfH-SM135, a protein released when the cell membrane disintegrates, indicates axonal death and can be used as a serum protein biomarker for neurodegeneration. High levels of NfH have been found in patients who received chemotherapy before bone marrow transplant and sustained cognitive deficits.
95
The potential for NfH to serve as a biomarker of cognitive impairment merits consideration.Other Sources of Data on Cognitive Function
Cancer treatment studies often include self-reports that assess health-related quality of life in physical, emotional, psychological, social, and cognitive domains. A nonexhaustive review of the literature on cognitive impairment associated with cancer treatment and chemotherapy identified more than 25 such self-report instruments; however, few of these instruments specifically address multiple aspects of cognition. The importance of perceived cognitive function
96
in understanding the effects of cancer treatment is underscored by a case study in which monozygotic twins discordant for cancer chemotherapy performed within normal limits on neuropsychological tests but had substantial differences in self-perceptions of cognitive functioning and on structural and fMRI.77
Other studies involving different study populations have found subjective complaints of cognitive dysfunction in the absence of objective signs of cognitive impairment to be associated with changes in brain structure or function.96
, 97
, 98
These studies suggest that self-reports of cognitive function may capture subtle nuances in functioning, as well as predict underlying structural or functional changes.96
Because neuropsychological testing is not always available, feasible, or affordable, an easily administered comprehensive measure of perceived cognitive function, such as the Functional Assessment of Cancer Therapy-Cognitive Function (FACT-Cog),
99
could be used to identify individuals who might benefit from more extensive neuropsychological or neurological evaluation. The FACT-Cog, a 37-item patient-reported outcome measure of cognitive functioning and quality of life, addresses perceived cognitive abilities (e.g., “I am able to shift back and forth between two activities that require thinking”), perceived cognitive impairments (e.g., “I have trouble finding my way to a familiar place”), perceptions of others (e.g., “Other people have told me I seem confused”), and impact on quality of life (e.g., “These problems have interfered with my ability to work”). Several clinical trials have been collecting neuropsychological test data and neuroimaging data in addition to the FACT-Cog, which would allow one to assess whether perceived cognitive changes are correlated with or predictive of structural or functional changes in the brain.An advantage of self-reports is that they summarize performance across time and situations, unlike neuropsychological tests that assess performance at a single point in time.
32
Rather than consider them as an alternative measure of cognitive function, it may be useful to regard self-reports and neuropsychological test data as independent sources of data on neuropsychological functioning.Returning to work after cancer treatment is a milestone for many patients and represents an important step toward recovery. Aside from the obvious economic benefits, work provides a means of social reintegration and may offer the patient a semblance of normalcy. Employment also may confer a sense of control and boost to one's self-esteem.
100
However, in contrast to the more protected and controlled home environment, the demands of the workplace, such as meeting deadlines and performance standards, interacting with fellow workers, and organizing daily tasks, may expose an individual's cognitive vulnerabilities. Work performance may be heavily influenced by physical changes associated with cancer treatment (e.g., peripheral neuropathy and limited mobility), which may make it difficult to isolate cognitive causes of changes in work status.Although a substantial proportion of cancer patients return to work during and after treatment,
101
many patients experience an unwanted change in work status as a result of their diagnosis or treatment.102
A meta-analysis that assessed the risk of unemployment for adult cancer survivors compared with healthy controls found that survivors were 1.37 times more likely to be unemployed.103
Not surprisingly, receipt of chemotherapy was associated with longer work absence, impaired ability to work, and changes in the work situation.104
, 105
Although one might expect that cognitive dysfunction would account for many of these adverse consequences, most studies cite impaired physical functioning secondary to cancer as the major reason for change in work status. Only rarely are cognitive complaints mentioned as a proximal cause of change in work status. However, qualitative studies describing the experience of cancer patients' return to work have noted significant difficulties resulting from problems with memory, attention, concentration, decision making, multitasking, and a general loss of confidence in cognitive abilities.
38
, 106
Fatigue is often cited as a major factor leading to changes in work status, yet the nature of fatigue (i.e., physical or mental) is rarely specified. It is conceivable that cognitive slowing and diminished executive function may be experienced and reported as fatigue. A more fine-grained approach is needed to understand these nuances and the effects of mild cognitive impairment on occupational functioning.Although many patients are able to continue working, a subset of patients lose their jobs, voluntarily or involuntarily retire, go on disability, reduce their work hours, or change jobs altogether for treatment-related reasons.
107
, 108
, 109
Furthermore, the cognitive, affective, and physical sequelae of cancer treatment may result in more limited job opportunities and chances for promotion. A more in-depth exploration of the relationship between cognitive dysfunction and changes in occupational status and functioning might illuminate the kind of cognitive retraining or type of cognitive aids that would help cancer survivors function most effectively when they return to the workplace.Leveraging Existing Data
Recognizing the inconsistencies in study design and methodology, the International Cognition and Cancer Task Force recommended that researchers adopt a more standardized approach to facilitate combining studies, harmonizing data, and making between-study comparisons.
33
They have called for larger studies to “make it more feasible to accurately identify specific treatments (modalities, regimens, dosimetry, and timing) and patient characteristics (age, cognitive reserve, genetic risk factors, comorbid conditions, and other cancer-related symptoms) that constitute risk factors for cognitive decline.”33
(p. 1) In the absence of larger studies, there may be ways to leverage existing data to answer some of these questions. Case-control studies and integrative data analysis may provide insight into risk factors and mechanisms of chemotherapy-associated cognitive impairment.Case-control studies offer a relatively efficient way to identify multiple potential risk factors for cognitive impairment.
110
These studies could be nested within existing observational studies of cancer patients that collected data on neuropsychological functioning (e.g., neuropsychological test data36
or proxy measures111
). For example, one could take a cohort of chemotherapy-exposed patients and identify two subsets of patients: those who experienced significant cognitive impairment (i.e., cases) and those who did not (i.e., controls). To define these groups, one could statistically extract them by applying multivariate finite mixture models., 113
Similar to cluster analysis, this technique defines mutually exclusive groups based on shared features, such as a particular cognitive test profile. It has an advantage over conventional techniques, such as cluster analysis, in that it adjusts for the probability or certainty that classifications are accurate. Cases and controls could then be compared with respect to potential medical, psychological, and behavioral risk factors for cognitive impairment. To the best of our knowledge, this study design has not been used to identify risk factors for cognitive impairment.Another technique that makes use of existing data is integrative data analysis, which is “the statistical analysis of a single data set that consists of two or more separate samples that have been pooled into one.”
114
(p. 82) Unlike meta-analysis, which aggregates summary-level data from each study, integrative data analysis combines individual raw data from independent studies, allowing models to be fitted to the original data. Although the merits of pooling data to create a more cumulative science have been touted by statisticians and behavioral scientists for some time, this technique has only recently gained currency as a result of the availability of newer statistical approaches, sophisticated software, and a greater emphasis on data sharing.114
, 115
Integrative data analysis hinges on whether measures from different test batteries are commensurable, that is, comparable by a common metric. Statisticians have developed modeling strategies to determine whether measures can be pooled. If pooling is feasible, studies can be combined to test important questions that previously had been precluded because of small sample size.Integrative data analysis has been used to study risk factors for a variety of cancers.
116
, 117
, 118
For example, van den Brandt et al.119
pooled data from seven cohort studies to study the relationship between height, weight, body mass index, and breast cancer risk. By pooling data, researchers could potentially increase the statistical power needed to test hypotheses that cannot be tested with small samples. One hypothesis that potentially could be tested is whether high-intensity chemotherapy causes greater cognitive impairment than conventional low-dose chemotherapy. Because certain chemotherapeutic agents are believed to cross the blood-brain barrier2
or produce systemic inflammation that leads to central nervous system inflammation,120
one would predict that higher doses of these drugs would be associated with greater cognitive impairment.5
However, because manipulating intensity of chemotherapy is neither ethical nor feasible, an alternative would be to test for a dose-response relationship using a pooled data set. Another question that pooling data might help to address is the timing of cognitive side effects. Some researchers have reported that cognitive deficits tend to be short lived, whereas others have found that deficits can persist 10 years after chemotherapy.15
, 25
Pooling patient data from studies that have assessed cognitive functioning at various time points (e.g., at one month, six months, one year, or 10 years after chemotherapy) might help to clarify the onset and duration of chemotherapy-related cognitive impairment.Conclusion
Ever since chemo brain entered the cancer lexicon, researchers have attempted to describe and quantify the cognitive side effects experienced by countless cancer patients. Although the methods that have been used to study this problem have yielded valuable insight into the nature of these deficits and possible candidate mechanisms,
2
these efforts could be broadened by incorporating a more multidisciplinary perspective and taking new and creative approaches to measurement. In addition to traditional neuropsychological testing, researchers should consider novel neuroimaging techniques, potential biomarkers, alternative measures of health outcomes, and computational modeling as valuable sources of data. There also may be much we can learn by systematically mining existing data to refine our understanding of which patients are at risk for experiencing cognitive impairment; the onset, duration, and reversibility of these side effects; and the specific cognitive processes and brain regions that are affected.More work is needed to understand the real-world functional implications of these deficits so that strategies to minimize and remediate these side effects can be developed and implemented before, during, and after cancer treatment. Moreover, understanding the everyday functional consequences of these cognitive side effects is vitally important from an ethical informed consent perspective. Future research conducted along the novel lines that we have suggested could augment the knowledge base to adequately inform cancer patients and health care providers about the potential cognitive effects of treatment.
Disclosures and Acknowledgments
This work had no specific funding source, and the authors have no financial interests to disclose.
The authors gratefully acknowledge and thank Roger Ratcliff, PhD, and Lynne Padgett, PhD, for their valuable advice.
References
- Neurotoxicity of commonly used antineoplastic agents.N Engl J Med. 1974; 291: 75-133
- Candidate mechanisms for chemotherapy-induced cognitive changes.Nat Rev Cancer. 2007; 7: 192-201
- Psychosocial aspects of neoplastic disease: II. Affective and cognitive effects of chemotherapy in cancer patients.Am J Psychiatry. 1980; 137: 597-601
- Neurotoxic complications of chemotherapy in patients with cancer.Drugs. 2003; 63: 1549-1563
- 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: 210-218
- Neuropsychological effects of treatments for adults with cancer: a meta-analysis and review of the literature.J Int Neuropsychol Soc. 2003; 9: 967-982
- The nature and severity of cognitive impairment associated with adjuvant chemotherapy in women with breast cancer: a meta-analysis of the current literature.Brain Cogn. 2005; 59: 60-70
- A metaanalysis of studies of the effects of cancer chemotherapy on various domains of cognitive function.Cancer. 2005; 104: 2222-2233
- A meta-analysis of the neuropsychological effects of adjuvant chemotherapy treatment in women treated for breast cancer.Clin Neuropsychol. 2006; 20: 76-89
- Meta-analysis of cognitive functioning in breast cancer survivors previously treated with standard-dose chemotherapy.J Clin Oncol. 2012; 30: 3578-3587
- Toxic leukoencephalopathy.N Engl J Med. 2001; 345: 425-432
- Progressive multifocal leukoencephalopathy in chronic lymphocytic leukemia: a report of three cases and review of the literature.Clin Lymphoma Myeloma Leuk. 2010; 10: E1-E9
- Evaluation of cognitive function associated with chemotherapy: a review of published studies and recommendations for future research.J Clin Oncol. 2007; 25: 2455-2463
- The (mis)classification of chemo-fog—methodological inconsistencies in the investigation of cognitive impairment after chemotherapy.Breast Cancer Res Treat. 2006; 95: 125-129
- Neuropsychologic impact of standard-dose systemic chemotherapy in long-term survivors of breast cancer and lymphoma.J Clin Oncol. 2002; 20: 485-493
- Cognitive deficits after postoperative adjuvant chemotherapy for breast carcinoma.Cancer. 1999; 85: 640-650
- Acute and late onset cognitive dysfunction associated with chemotherapy in women with breast cancer.Cancer. 2010; 116: 3348-3356
- Long-term cognitive function following chemotherapy in patients with testicular cancer.J Int Neuropsychol Soc. 2009; 15: 296-301
- Cognitive functioning after adjuvant chemotherapy and/or radiotherapy for early-stage breast carcinoma.Cancer. 2005; 104: 2499-2507
- No indications of cognitive side-effects in a prospective study of breast cancer patients receiving adjuvant chemotherapy.Psychooncology. 2009; 18: 248-257
- A 3-year prospective study of the effects of adjuvant treatments on cognition in women with early stage breast cancer.Br J Cancer. 2006; 94: 828-834
- Cognitive function during neoadjuvant chemotherapy for breast cancer.Cancer. 2007; 109: 1905-1913
- The cognitive sequelae of standard-dose adjuvant chemotherapy in women with breast cancer.Cancer. 2004; 100: 2292-2299
- Prechemotherapy alterations in brain function in women with breast cancer.J Clin Exp Neuropsychol. 2010; 32: 324-331
- Cognitive effects of chemotherapy in post-menopausal breast cancer patients 1 year after treatment.Psychooncology. 2009; 18: 134-143
- Preliminary results of a longitudinal study of changes in cognitive function in breast cancer patients undergoing chemotherapy with doxorubicin and cyclophosphamide.Psychooncology. 2008; 17: 1189-1195
- Cognitive impairment associated with adjuvant therapy in breast cancer.Psychooncology. 2006; 15: 422-430
- Neurocognitive performance in breast cancer survivors exposed to adjuvant chemotherapy and tamoxifen.J Clin Exp Neuropsychol. 2004; 26: 955-969
- Assessing cognitive deficits in bipolar disorder: are self-reports valid?.Psychiatry Res. 2005; 136: 43-50
- Neuropsychological functioning following systemic treatment in women treated for breast cancer: a review.Eur J Cancer. 2003; 39: 2288-2297
- Pre-treatment factors related to cognitive functioning in women newly diagnosed with breast cancer.Psychooncology. 2005; 14: 70-78
- Cognitive impairment associated with chemotherapy for cancer: report of a workshop.J Clin Oncol. 2004; 22: 2233-2239
- International Cognition and Cancer Task Force recommendations to harmonise studies of cognitive function in patients with cancer.Lancet Oncol. 2011; 12: 703-708
- Confronting chemobrain: an in-depth look at survivors' reports of impact on work, social networks, and health care response.J Cancer Surviv. 2009; 3: 223-232
- A meta-analysis of the sensitivity of various neuropsychological tests used to detect chemotherapy-induced cognitive impairment in patients with breast cancer.Oncol Nurs Forum. 2007; 34: 997-1005
- Neuropsychological assessment.4th ed. Oxford University Press, New York2004
- Cognitive adverse effects of chemotherapy in breast cancer patients.Curr Opin Support Palliat Care. 2007; 1: 57-62
- Women's perceptions of chemotherapy-induced cognitive side affects on work ability: a focus group study.J Clin Nurs. 2010; 19: 1362-1370
- On testing the face validity of planning/problem-solving tasks in a normal population.J Int Neuropsychol Soc. 1997; 3: 108-119
- The ecological validity of neuropsychological tests: a review of the literature on everyday cognitive skills.Neuropsychol Rev. 2003; 13: 181-197
- Utilizing virtual reality to improve the ecological validity of clinical neuropsychology: an fMRI case study elucidating the neural basis of planning by comparing the Tower of London with a three-dimensional navigation task.Appl Neuropsychol. 2009; 16: 295-306
- Neuropsychological assessment.5th ed. Oxford University Press, New York2012
- Ecological validity: some critical issues for the neuropsychologist.in: Sbordone R.J. Long C.J. Ecological validity of neuropsychological testing. GR Press/St. Lucie Press, Delray Beach, FL1996: 15-41
- The party planning task: a useful tool in the functional assessment of planning skills in adolescents with TBI.Brain Inj. 2011; 25: 1080-1090
- The development and validation of a test battery for detecting and monitoring everyday memory problems.J Clin Exp Neuropsychol. 1989; 11: 855-870
- The structure of normal human attention: the Test of Everyday Attention.J Int Neuropsychol Soc. 1996; 2: 525-534
- Deficits in strategy application following frontal lobe damage in man.Brain. 1991; 114: 727-741
- Ecological validity of neuropsychological assessment: do neuropsychological indexes predict performance in everyday activities?.Appl Prev Psychol. 1993; 2: 209-215
- Ecological validity in neuropsychological assessment: a case for greater consideration in research with neurologically intact populations.Arch Clin Neuropsychol. 2006; 2: 327-337
- The Extended Rivermead Behavioural Memory Test: a measure of everyday memory performance in normal adults.Memory. 1994; 2: 149-166
- HRT and everyday memory at menopause: a comparison of two samples of mid-aged women.Women Health. 2006; 43: 37-57
- Ecological validity of a simplified version of the multiple errands shopping test.J Int Neuropsychol Soc. 2003; 9: 31-44
- Development of a simplified version of the multiple errands test for use in hospital settings.Neuropsychol Rehabil. 2002; 12: 231-255
- Virtual reality based tools for the rehabilitation of cognitive and executive functions: the V-STORE.PsychNology J. 2003; 1: 310-325
- Analysis of assets for virtual reality applications in neuropsychology.Neuropsychol Rehabil. 2004; 14: 207-239
- Neuropsychological assessment using virtual environments: enhanced assessment technology for improved ecological validity.in: Brahnam S. Jain L.C. Advanced computational intelligence paradigms in healthcare 6. Springer-Verlag, Berlin2011: 271-289
- Neuropsychological assessment of attentional processing using virtual reality.Annu Rev CyberTherapy Telemed. 2008; 6: 23-28
- Mental representation and the spatial structure of virtual environments.Environ Behav. 2000; 32: 427-442
- Quantitative and qualitative sex differences in spatial navigation.Scand J Psychol. 2007; 48: 353-358
- Age and dementia-related differences in spatial navigation within an immersive virtual environment.Med Sci Monit. 2009; 15: CR140-CR150
- Neurodevelopmental aspects of spatial navigation: a virtual reality fMRI study.Neuroimage. 2002; 15: 396-406
- Neurocognitive functioning is associated with employment status: a quantitative review.J Clin Exp Neuropsychol. 2003; 25: 1186-1191
- The relationship between neuropsychological functioning and driving ability in dementia: a meta-analysis.Neuropsychology. 2004; 18: 85-93
- Prediction of functional status in older adults: the ecological validity of four Delis-Kaplan Executive Function System tests.J Clin Exp Neuropsychol. 2008; 30: 683-690
- A compendium of neuropsychological tests.Oxford University Press, Oxford1998
- A diffusion decision model: theory and data for two-choice decision tasks.Neural Comput. 2008; 20: 873-922
- Modeling reaction time and accuracy of multiple-alternative decisions.Atten Percept Psychophys. 2010; 72: 246-273
- Taking a computational approach to aging: the SPAN theory of working memory.Psychol Aging. 1998; 13: 309-322
Roring RW. Age-related changes in information processing on tasks of perceptual speed. 2005. Electronic Theses, Treatises and Dissertations. Paper 1774. Available from http://diginole.lib.fsu.edu/etd/1774. Accessed November 5, 2012.
- Using diffusion models to understand clinical disorders.J Math Psychol. 2010; 54: 39-52
- Modeling the effects of hypoglycemia on a two-choice task in adult humans.Neuropsychology. 2010; 24: 652-660
- No adverse effects of adjuvant chemotherapy on hippocampal volume in Japanese breast cancer survivors.Breast Cancer Res Treat. 2005; 92: 81-84
- Smaller regional volumes of brain gray and white matter demonstrated in breast cancer survivors exposed to adjuvant chemotherapy.Cancer. 2007; 109: 146-156
- Adjuvant chemotherapy for breast cancer: effects on cerebral white matter seen in diffusion tensor imaging.Clin Breast Cancer. 2008; 8: 88-91
- MR and proton spectroscopy of white matter disease induced by high-dose chemotherapy with bone marrow transplant in advanced breast carcinoma.Am J Neuroradiol. 1995; 16: 2013-2020
- 5-Fluorouracil-induced leukoencephalopathy in patients with breast cancer.J Korean Med Sci. 2001; 16: 328-334
- Brain structure and function differences in monozygotic twins: possible effects of breast cancer chemotherapy.J Clin Oncol. 2007; 25: 3866-3870
- Late effects of high-dose adjuvant chemotherapy on white and gray matter in breast cancer survivors: converging results from multimodal magnetic resonance imaging.Hum Brain Mapp. 2012; 30: 3675-3686
- White matter disease induced by high-dose chemotherapy: longitudinal study with MR imaging and proton spectroscopy.Am J Neuroradiol. 1998; 19: 217-221
- Global and focal brain volume in long-term breast cancer survivors exposed to adjuvant chemotherapy.Breast Cancer Res Treat. 2012; 132: 1099-1106
- Gray matter reduction associated with systemic chemotherapy for breast cancer: a prospective MRI study.Breast Cancer Res Treat. 2010; 123: 819-828
- Chemotherapy-induced structural changes in cerebral white matter and its correlation with impaired cognitive functioning in breast cancer patients.Hum Brain Mapp. 2011; 32: 480-493
- Diffusion tensor imaging in the brain.Neurotherapeutics. 2007; 4: 316-329
- Longitudinal assessment of chemotherapy-induced structural changes in cerebral white matter and its correlation with impaired cognitive functioning.J Clin Oncol. 2012; 30: 274-281
- Age- and gender-related changes in the normal human brain using hybrid diffusion imaging (HYDI).Neuroimage. 2011; 54: 1840-1853
- White matter changes in patients with breast cancer treated with high-dose chemotherapy and autologous bone marrow support.AJNR Am J Neuroradiol. 1994; 15: 1267-1273
- Altered frontocortical, cerebellar, and basal ganglia activity in adjuvant-treated breast cancer survivors 5-10 years after chemotherapy.Breast Cancer Res Treat. 2007; 103: 303-311
- Cerebral hyporesponsiveness and cognitive impairment 10 years after chemotherapy for breast cancer.Hum Brain Mapp. 2011; 32: 1208-1219
- Regional brain activation during verbal declarative memory in metastatic breast cancer.Clin Cancer Res. 2009; 15: 6665-6673
- Prefrontal cortex and executive function impairments in primary breast cancer.Arch Neurol. 2011; 68: 1447-1453
- Characterizing variation in the functional connectome: promise and pitfalls.Trends Cogn Sci. 2012; 16: 181-188
- Functional connectivity in the motor cortex of resting human brain using echo-planar MRI.Magn Reson Med. 1995; 34: 537-541
- Intrinsic functional connectivity as a tool for human connectomics: theory, properties and optimization.J Neurophysiol. 2010; 103: 297-321
- Abnormal connectivity in the posterior cingulated and hippocampus in early Alzheimer’s disease and mild cognitive impairment.Alzheimers Dement. 2008; 4: 265-270
- Evidence for acute neurotoxicity after chemotherapy.Ann Neurol. 2010; 68: 806-815
- Evaluating the dimensionality of perceived cognitive function.J Pain Symptom Manage. 2009; 37: 982-995
- Older adults with cognitive complaints show brain atrophy similar to that of amnestic MCI.Neurology. 2006; 67: 834-842
- Cerebral white matter lesions and subjective cognitive dysfunction: the Rotterdam Scan Study.Neurology. 2001; 56: 1539-1545
- Measuring patient self-reported cognitive function: development of the Functional Assessment of Cancer Therapy-Cognitive Function instrument.J Support Oncol. 2009; 7: W32-W39
- Cancer and the meaning of work.Gen Hosp Psychiatry. 2000; 22: 200-205
- Factors reported to influence the return to work of cancer survivors: a literature review.Psychooncology. 2002; 11: 124-131
- Employment and work-related issues in cancer survivors.Crit Rev Oncol Hematol. 2011; 77: 109-130
- Cancer survivors and unemployment: a meta-analysis and meta-regression.JAMA. 2009; 301: 753-762
- Return to work after early-stage breast cancer: a cohort study into the effects of treatment and cancer-related symptoms.J Occup Rehabil. 2008; 18: 267-272
- Long-term cancer survivors experience work changes after diagnosis: results of a population-based study.Psychooncology. 2009; 18: 1252-1260
- A longitudinal qualitative study of the experience of working following treatment for gynaecological cancer.Psychooncology. 2012; 21: 82-89
- Cancer and the risk for taking early retirement pension: a Danish cohort study.Scand J Public Health. 2008; 36: 117-125
- Job loss and reemployment after a cancer diagnosis in Koreans—a prospective cohort study.Psychooncology. 2007; 16: 205-213
- Factors influencing changes in employment among women with newly diagnosed breast cancer.Cancer. 2009; 115: 2775-2782
- Case-control studies: Design, conduct, analysis.Oxford University Press, New York1982
- Telephone assessment of cognitive function in adulthood: the Brief Test of Adult Cognition by telephone.Age Ageing. 2006; 35: 629-632
- Finite mixture models.Wiley, New York2000
- Major habitual dietary patterns are associated with acute myocardial infarction and cardiovascular risk markers in a southern European population.J Am Diet Assoc. 2011; 111: 241-250
- Integrative data analysis: the simultaneous analysis of multiple data sets.Psychol Methods. 2009; 14: 81-100
- The relative benefits of meta-analysis conducted with individual participant data versus aggregated data.Psychol Methods. 2009; 14: 165-176
- Association between body mass index and the colorectal cancer risk in Japan: pooled analysis of population-based cohort studies in Japan.Ann Oncol. 2012; 23: 479-490
- Height at diagnosis and birth-weight as risk factors for osteosarcoma.Cancer Causes Control. 2011; 6: 899-908
- Persistence of disseminated tumor cells in the bone marrow of breast cancer patients predicts increased risk for relapse—a European pooled analysis.Clin Cancer Res. 2011; 17: 2967-2976
- Pooled analysis of prospective cohort studies on height, weight, and breast cancer risk.Am J Epidemiol. 2000; 152: 514-527
- Alterations in brain antioxidant enzymes and redox proteomic identification of oxidized brain proteins induced by the anti-cancer drug adriamycin: implications for oxidative stress-mediated chemobrain.Neuroscience. 2010; 166: 796-807
Article info
Publication history
Published online: March 25, 2013
Accepted:
December 7,
2012
Identification
Copyright
© 2013 U.S. Cancer Pain Relief Committee. Published by Elsevier Inc.
