Abstract
Cancer-associated cognitive deficits following chemotherapy have received increased attention in clinical research. Exercise has been shown to preserve cognitive function in cancer patients, though the overall effect is mixed. Here we present a scoping review of the published literature summarizing methods used to assess cognitive function in exercise oncology trials.
Keywords
Introduction
Cancer patients are at increased risk for developing cognitive impairments (CI) for a myriad of reasons. 1 The onset and progression of CI is multifocal. CI can be the result of tumor burden, treatment related side effects or age-related comorbidities. Patients and providers often refer to this as “chemobrain.” Impairments are often reported across multiple cognitive domains, including verbal memory, language, visual spatial skills, executive function, and psychomotor function.2,3 Attention, memory and executive functioning are the most frequently identified domains of CI impacted by cancer. 1 Multiple reviews have summarized the literature on the rates of CI in cancer patients.1,4-6 In a review by Janelsins et al 6 the authors reported that 30% of patients with cancer exhibited cognitive impairment prior to treatment, 75% may have CI during treatment and 35% will continue to exhibit cognitive impairment in the months to years following treatment. Several comorbid factors can impact cognitive functioning in cancer survivors. Depression, anxiety, and fatigue can adversely affect cognitive functioning by increasing cognitive load prior to, during and after the completion of treatment.7-9 Chemotherapy has been highly reported to be associated with cognitive impairment.4,5,10,11 The conceptual model of Chemotherapy-Related Change in Cognitive Function presented the physiological and psychological antecedents, moderators, mediators and consequences of chemotherapy for which future research can build upon. 12
Cognitive function can be assessed using objective and subjective. The current gold standard for measuring CI is objective neuropsychological assessment performed by a trained psychologist, which can target certain domains. 13 However, these can be time-consuming for both researchers and patients. 14 Alternatively, patient-reported outcomes (PRO) are widely used to understand the context of disease burden; however, reference values are instrument-specific and comparisons between outcomes using different PRO assessments are difficult. 15 Additionally, responses can be influenced by factors like fatigue or anxiety1,16 and fail to provide objective measurements of cognitive function. To harmonize the data elements collected to evaluate cognitive function in cancer trials, the International Cognition and Cancer Task Force (ICCTF) was convened to identify strategies to increase consistency in the design and endpoints used in oncology trials centered on cognitive function.17,18 The ICCTF provides recommendations for study design, measurement timepoints, tests to include to evaluate cognitive function, criteria to assess CI and criteria to assess changes in cognitive function. 18
To attenuate the cognitive function deficits that may occur in cancer patients, behavioral interventions such as exercise have been proposed. 3 Exercise has been associated with improved cognitive function in adults with and without cancer because of its ability to reduce inflammation and enhance neuroplasticity. 19 A number of reviews have demonstrated that various modalities of exercise can serve as a preventative and rehabilitative intervention for cognitive deficits in cancer patients, including aerobic exercise, resistance exercise and mind-body interventions like yoga and tai chi.19-21 While promising, the variability in the assessments and methodologies used to measure cognitive function in exercise oncology trials points to the need for more standardized measurements of cognitive function in rigorous clinical trials. In the absence of any curative intervention for cancer-associated CI, exercise is one strategy to help maintain cognition. However, in order for the medical and lay community to better understand the benefits of exercise as an integrative approach to the prevention and management of serious side effects associated with cancer and cancer therapy, a structured methodology of cognitive assessments need to be included in exercise oncology trials. The purpose of this scoping review is to summarize the strategies used to measure CI in exercise oncology trials in cancer patients that have completed chemotherapy. Additionally, we offer directions for future research in this field.
Methodology
Information Sources and Search Strategy
PRISMA Extension for Scoping Reviews (PRISMA-ScR) was used to guide this scoping review. 22 PubMed, CINAHL (EBSCO), and PsycINFO (OVID) databases were searched using the keywords “cognition,” “cancer” OR “neoplasm” OR “tumor,” “chemotherapy” and “exercise” OR “physical activity” on March 15, 2022. Keywords were entered into the search bar as a single entry with limits and filters applied based on the eligibility criteria described below.
Eligibility
After exclusion of duplicates, 2 experienced exercise oncology researchers (CM and DP) independently reviewed titles and abstracts of the identified articles. Studies eligible for inclusion in this review include those that were written in English, included a method of analyzing cognitive function in adult onset cancer patients, studies in which data were collected prospectively, studies in which an exercise modality or method of quantifying exercise habits was evident and published in peer-reviewed journals. Studies were excluded from this review if they were protocol papers, included a pediatric population, included patients that were not diagnosed with cancer, were systematic/narrative/scoping reviews or dissertation/theses. After final inclusion of abstracts, the respective full texts were reviewed independently by the same 2 reviewers to ensure appropriate inclusion in this review.
Data Extraction and Analysis
Results of the search strategy were compiled into Rayyan (rayyan.ai). Duplicates were removed and abstracts were reviewed using inclusion and exclusion criteria. Full text review was performed on articles meeting all criteria. Data collected from eligible studies included type of physical activity or exercise evaluated, oncological target population, methods used to measure cognitive function and whether cognitive function was measured as a primary or secondary endpoint. The Grades of Recommendations, Assessment, Development, and Evaluation (GRADE) ranking system was used to evaluate the quality of evidence for study outcomes. 23 Levels of quality (high, moderate, low, very low) were adjusted based on risk of bias related to sample size, cognitive function as a primary endpoint, objectivity of the instrumentation related to the primary endpoint, and specificity of the instrumentation. Classification of cognitive function endpoints was based in accordance with the authors’ intentions. Relevant data from the included full tests were extracted and organized into standardized data tables.
Results
A total of 332 articles were found using the keywords indicated. After 97 duplicates were removed, 235 records underwent abstract screening, of which 216 articles were excluded. A total of 29 studies met the inclusion criteria and were included in the review. Figure 1 displays the flow of studies through the review process.
PRISMA diagram.
Study Characteristics
Included studies were published between 2008 and 2022. The largest proportion of studies were conducted in the United States (n = 11, 35.5%) but the search also included studies conducted in Germany (n = 5, 16.1%), Canada (n = 4, 12.9%), China (n = 4, 12.9%), Sweden (n = 2, 6.5%), France (n = 2, 6.5%), Japan (n = 1, 3.2%), the United Kingdom (n = 1, 3.2%), and The Netherlands (n = 1, 3.2%).
Breast cancer comprised the largest population studied, with 15 studies (48.3% overall) focusing solely on breast cancer,24-38 while an additional 6 studies included breast cancer survivors in their study sample in addition to other cancer populations.39-44 Ovarian cancer was investigated in 2 studies45,46 as was colorectal cancer.47,48 Leukemia, 49 gastrointestinal, 50 and lung 51 cancer each had one study that focused only on that respective disease site. Seven studies collected data across several cancer populations including breast, colorectal, ovarian, Hodgkin’s lymphoma, leukemias, bone marrow, and prostate.39-44,52 Four studies investigated the role of adjuvant exercise while patients were going through therapy.45,48,49,52 A summary of these studies is presented in Tables 1 and 2.
Description of Studies Included in this Review.
Cognitive Function Endpoint: Indicates whether the reviewed study measured cognitive function as a primary or secondary endpoint.
Grade: The quality of evidence based on the Grades of Recommendation, Assessment, Development, and Evaluation (GRADE) ranking system. The GRADE system offers four levels of evidence: high (H), moderate (M), low (L), and very low (VL).
Summary of the Assessments Used to Measure Cognitive Function.
Physical Activity or Exercise Interventions Reported
This review includes a mix of interventional and observation studies. Interventional studies are defined for this review as prospective trials in which a structured exercise program was administered. Observational studies are defined for this review as studies that used a survey-based method or accelerometer to evaluate exercise habits in consenting participants. Nine studies tailored a combination of resistance and aerobic exercise to the needs, abilities, and progression of individual participants.26,27,31,33,39,44,46,49,52 Another 6 studies used a structured aerobic and resistance training24,35,38,41,48,51 program so that each patient performed the same frequency, intensity, time, and type of exercise performed for the intervention. Seven studies conducted observational studies looking at the amount of self-selected time and intensity (eg, light vs. moderate-to-vigorous vs. vigorous) of participants either through surveys like the Godin Leisure Time Exercise Questionnaire, physical tracking using actigraphy, or a combination of questionnaires and physical measures.25,28,30,34,37,40,42 Two studies focused on resistance training only24,38 while another 3 focused only on aerobic exercise.29,43,48 Three studies looked at the use of Qigong36,47,50, 1 incorporated yoga. 32
The duration of interventions also varied between studies. A majority of studies conducted prospective exercise trials ranging from 12 to 24 weeks.24,26,29,31,33,35,38,43,46-48,51 The shortest study investigated bimodal exercise in a 3-week intervention 41 while a yoga intervention was evaluated in a 4-week protocol. 32 The longest trial included a 12-month intervention of personalized bimodal exercise. 39
Levels of Evidence
Significant variability existed between the studies included in this review. Only 5 of the 15 RCTs were designed to investigate cognitive function as a primary outcomes,26,29,45-47 of which only 3 utilized objective measures.26,29,47 Of the remaining studies (14 non-RCTs), 11 were designed to measure cognitive function as a primary outcomes,24,25,28,30,34,37,40-44 of which 6 utilized objective measures.24,28,30,34,42,43 Based on the GRADE system, none of the studies met the benchmark for a high-ranking classification on the quality of evidence. Sixteen studies were ranked as moderate, of which 2 were aerobic exercise studies,26,29 2 were resistance exercise studies,24,38 8 were bimodal,27,35,39,45,46,49,51,52 1 was high-intensity interval training, 48 and 3 were Qigong.19,45,47 The remaining studies were ranking as low.25,28,30-34,37,40-44
Methods to Assess Cognitive Function
In total 29 different assessments were used to evaluate cognitive function. Studies utilized anywhere from a single assessment of cognitive function to 11 different assessments. To assess effects of PROs on cognitive function, 8 different assessments were used. Objective measures of cognitive function were more robust with 21 different assessments used. Furthermore, of the 29 studies included in this review, 16 studies included cognitive function as a primary outcome24-26,28-30,34,37,40-47 while the remaining 13 studies involved cognitive function measurement as secondary outcomes.27,31-33,35,36,38,39,48-52 Twelve studies utilized objective measures of cognitive function.24,26,28-30,34,36,38,39,42,43,47 The different methods of cognitive function measurements are summarized in Tables 3 and 4.
Summary of the Assessments Used to Measure Cognitive Function Based on Cancer Type.
Summary of the Assessments Used to Measure Cognitive Function Based on Exercise Study Type.
Domains Measured in Exercise Oncology Trials
The articles included in this study included a wide range of cognitive domains that were measured, both through subjective (patient reported outcomes) and objective (computerized or staff administered) assessments. Using the 8 overarching domains of cognition published by Harvey, 53 we evaluated the frequency of cognitive domains assessed in the exercise oncology studies included in this review. Of the 8 patient-reported surveys identified in the literature, memory and attention/concentration were the domains most frequently evaluated. Memory was assessed in the EORTC-QLQ-C30, FACT-Cog, PROMIS, Cognitive Failures Questionnaire, Frequency of Forgetting Scale, and the Modified Fatigue Impact Scale. Attention/concentration was assessed in the FACT-Cog, Piper Fatigue Scale, PROMIS survey, and the Modified Fatigue Impact Scale. We were unable to identify a specific domain for the Healthy Brain Questionnaire because it was used as a demographic tool rather than an assessment of cognitive function.
The objective assessments of cognitive function were much clearer in identifying specific cognitive domains and often included multiple domains within each assessment. Ten of the assessments found in the literature included memory and the subcomponents of memory (ie, working memory, episodic/declarative memory, procedural memory, etc.) and 9 assessments targeted attention and concentration. Executive function was included in 10 different assessments while processing speed was included in 6 studies. Table 5 presents each assessment and the respective cognitive domains targeted.
Cognitive Domains Evaluated in Exercise Oncology Trials in Patients After Chemotherapy.
Patient Reported Measures of Cognition
Twenty-two papers included patient-reported outcomes of which 17 publications relied solely on patient reported measures of cognitive function. Below we briefly summarize each of these instruments:
Cognitive Failures Questionnaire
The Cognitive Failures Questionnaire (CFQ) is a 25-item instrument using a 5-point Likert scale to measure cognitive function. 54 Three subscales are measured including perception slips, memory slips and slips in motor functioning. The CFQ can also be scored used as a summative score of absent-mindedness. This assessment was used as a stand-alone measure of cognitive function investigating multimodal exercise in breast cancer survivors. 32
EORTC-QLQ-C30
The EORTC-QLQ-C30 is a 30-item quality of life survey that measures 5 functional domains of quality of life (physical, emotional, social, role, cognitive), 8 symptoms (fatigue, pain, nausea/vomiting, constipation, diarrhea, insomnia, dyspnea and appetite loss), as well as global health/quality of life and financial impact. 55 Six studies in our review used the EORTC-QLQ-C30 to measure cognitive function in cohorts of patients with breast, lung, colorectal, gastrointestinal and a heterogenous sampling of hematological malignancies, respectively.27,35,48,50-52
FACT-Cog
The FACT-Cog is a 37-item survey of cancer patients related to cognitive problems using 5-point Likert-type scales. 56 This survey asks respondents to reflect on the past 7 days to answer questions that measure perceived cognitive impairments, impact of perceived cognitive impairments on quality of life, comments from others, and perceived cognitive abilities. Five studies, solely focused on exercise interventions in breast cancer patients, utilized the FACT-Cog to measure cognitive function.25,26,33,36,47
Frequency of Forgetting Scale
The Frequency of Forgetting Scale assesses feelings about aspects of memory with lower scores indicative of more impairment. 57 This is a 10-item survey which demonstrates comparable outcomes as compared to the original 33-item survey. The final score represents memory self-efficacy of the participant. Only one study in our review utilized this method. 30
Healthy Brain Questionnaire
The Healthy Brain Questionnaire is a 19-item survey that includes a mix of demographic information, diet, physical activity and 2 questions on cognitive function. 58 The questions on cognitive function ask the respondent to rate their memory and if they have received a clinical diagnosis of memory loss, Alzheimer’s disease, dementia, or senility. This assessment of cognitive function was utilized in studies investigating exercise on cognitive function in breast/colorectal cancer patients and was used in conjunction with the objectively measured Montreal Cognitive Assessment. 42
Modified Fatigue Impact Scale (MFIS)
The Modified Fatigue Impact Scale is a 21-item scale that assesses the effects of fatigue in terms of physical, cognitive, and psychosocial functioning. 59 The cognitive subscale consists of 10 questions with a score ranging from 0 to 40 points. The MFIS was used in 2 studies in a heterogenous sampling of hematological cancer patients.41,52
Piper Fatigue Scale
The Piper Fatigue Scale is a 22 item survey that measures 4 subscales: behavioral/severity, affective meaning, sensory, and cognitive/mood. 60 Each item is rated on a scale of 0 (none) to 10 (a great deal) and can be useful in assessing cancer patient’s concentration and nonverbal memory. The cognitive/mood subscale was used to measure the effects of fatigue on cognitive function in 5 studies which included breast cancer only, 2 studies in ovarian cancer patients and 2 studies engaging a heterogenous sampling of cancer patients.31,40,44-46
PROMIS survey
The PROMIS applied cognition is a short, self-reported assessment of cognitive abilities and concerns. It is a 16-item survey evaluating self-impressions of cognitive function “in the past 7 days” in areas such as mental acuity, concentration, and memory. 61 PROMIS was used as a standalone measure of cognitive function in 3 studies enrolling leukemia and breast cancer patients, respectively.36,37,49
Objective Measures of Cognitive Impairment
Thirteen studies utilized objective measured of cognitive function.24,26,28-30,34,36,38,39,42,43,47 Our review found 22 objective cognitive function assessments that were used, with only 6 assessments (Controlled Oral Word Association Test, Hopkins Verbal Learning Test, Stroop Test, Trail Making Test, Wechsler Adult Intelligence and Memory Scale, and Verbal Fluency Test) used in more than one study. Each of these objective measures of cognitive function are summarized below:
Animal Naming Test
The Animal Naming Test is an assessment of categorical fluency used by Campbell et al 26 In this assessment, participants are given 60 seconds to name as many animals as they can. If the participant gives no response for 15 seconds, the examiner prompts the participant. Gender distinctions (ie, mare, stallion gelding) and age distinctions (foal, horse) are usually awarded full credit.
Central Nervous System Vital Signs Battery
The Central Nervous System Vital Signs Battery is a computerized assessment of cognitive function that measures attention, executive functioning, psychomotor speed and memory. 62 This battery has 5 primary domain scores for memory, psychomotor speed, reaction time, complex attention, and cognitive flexibility. Six secondary domain scores are also calculated for verbal memory, visual memory, processing speed, executive function, simple attention, and motor speed. An overall summary score (Neurocognition Index) can also be calculated. This assessment takes approximately 30 minutes to administer. Crowgey et al 28 utilized this method in breast cancer patients.
Cognitive P300 Test
The highly technical Cognitive P300 Test measures the amplitude and latency of P300 using a 32-channel electroencephalography acquisition system. 63 The P300 wave is a measurable direct reaction of the brain to a certain sensory, cognitive, or mechanical stimulus. Latency is associated with information processing time while amplitude is related to attention level. Lin et al 47 used this objective measure in parallel with the FACT-Cog to evaluate cognitive function in colorectal cancer patients.
Controlled Oral Word Association Test
The Controlled Oral Word Association Test (COWAT) is an oral fluency test in which the patient is told to make verbal associations to different letters of the words which they can think of beginning with a given letter. 64 Three letters of progressively increasing difficulty are presented successively as a stimulus. Difficulty level of each letter is defined in terms of relative frequency of words beginning with that letter in each language, as found in the standard language dictionary. Campbell et al 26 and Peterson et al 43 used this assessment in a mixed cohort of cancer patients.
D2 Test of attention
The D2 Test assesses the patient’s capacity for sustained attention. 65 The test consists of rows of letters randomly interspersed with a designated target letter. The patient is asked to cross out all the target letters as an indication of attention and concentration. Baumann et al 24 utilized this method in conjunction with the Wilde and Memo tests.
Flanker task
The Flanker task test is a measure of executive functioning where reaction time is measured across incongruent trials. 66 The participant is presented with a central fixation cross for 500 ms, after which the patient is presented 5 arrows for 2000 ms. Randomly, the flanking arrows point in the same direction as the target arrow. The patient is instructed to focus on the central arrow and report as quickly and accurately as possible whether the arrow pointed to the right or left. The Flanker test was used in a single study of breast cancer patients as part of a battery of assessments. 30
Hopkins verbal learning test
This assessment is useful as a measure of verbal learning and memory. 67 The participant is read 12 words which they must recall immediately after the reading. Two additional learning trials are then administered, after which the participant is alerted that he/she might be asked to recall the list again later. The final trial is completed 20 to 25 minutes after the previous reading where the participant is asked to recall as many of the 12 words as possible (ie, delayed recall). Finally, the examiner reads a list of 24 words, including the 12 target words, and the participant is asked to indicate a “yes” or “no” to the word read out loud based on their recollection of whether the word was on the target list of not. Raw scores are derived for Total Recall, Delayed Recall, Retention (% retained), and Recognition Discrimination Index. Campbell et al 26 utilized this assessment as part of a battery of assessments in a cohort of breast cancer patients.
Letter/number sequencing/coding
The letter/number sequencing task is a subtest of the WAIS-III and WAIS-IV test battery that measures working memory capacity. 68 Traditionally, this task is administered orally, but also can be administered via electronic devices. Working memory capacity is the patient’s ability to concurrently process and store information. The letter/number sequencing tasks involved hearing a series of letters and numbers and then reporting back the stimuli with the letters in alphabetical order and the numbers recited in numerical order. Peterson et al 43 utilized this as part of a battery of assessments.
Logical memory I and II test
Logical memory subsets are core assessments in the Wechsler Memory Scale IV. 69 The logical memory I subtest assesses narrative memory under a free recall condition. Two short stories are presented orally to the patient. In the case of older adults, one story is presented twice. The patient is asked to retell each story from memory immediately after hearing it. Logical memory II assesses long-term narrative memory with free recall and recognition tasks. The patient is asked to retell both stories from the immediate recall and then asked yes/no questions about both stories, as conducted in Peterson et al. 43
Mazes Task
The Mazes Task is a simple objective measure of executive functioning where the patient is asked to use their finger to draws a line from the start of the maze to the finish. 70 Performance is measured as the total time spent drawing while completing the maze. The mazes task test was included in a battery of assessments in breast cancer patients. 30
Memo Test
The Memo Test is a 10-word learning test constructed according to Buschke and Fuld’s 71 selective reminding procedure. Five consecutive trials are performed. Delayed recall is tested after 15 minutes. Baumann et al 24 utilized this assessment as part of battery of tests to investigate the effects of resistance exercise in breast cancer patients.
Montreal cognitive assessment (MoCA)
The MoCA is a test useful to evaluate patients with memory loss or other symptoms of cognitive decline. 72 The assessment contains 30 questions and take less than 15 minutes to complete. The MoCA assesses different types of cognitive or thinking abilities, including orientation, short-term memory, executive functioning/visuospatial ability, language, abstraction, animal naming, attention, and clock-drawing test. Fitzpatrick et al 42 use the MoCA in conjunction with the Healthy Brain Questionnaire in a mixed cohort of breast and colorectal cancer survivors.
NeuroTrax Comprehensive Testing Suite
The NeuroTrax Comprehensive Testing suite is a 45-minutes computerized test designed to sample a range of cognitive domains using a variety of assessments. 73 Information processing speed is assessed by low- and medium-load stages of Staged Information Processing Speed test; memory is assessed using Verbal and Non-verbal Memory Tests; and executive function is assessed by the Stroop Interference test, Go-No-Go Response Inhibition test, and the Catch Game. The NeuroTrax suite of assessments was used to evaluate general physical activity levels and cognitive function in breast cancer patients. 34
Rey auditory verbal learning test
This test measures short term and long term memory following interpolated activity and allows for a comparison between retrieval efficiency and learning. 74 This test assesses the patient’s ability to encode, combine, store, and recover verbal information in different stages of memory. The effect of interference stimulus, delayed memory and recognition are evaluated with this assessment. This battery of assessments was used in breast cancer patients to evaluate the effect of Qigong on cognitive function. 36
Stroop test
The Stroop test assesses the patient’s ability to substitute an alternative response for more obvious reactions and is sensitive to impairments in executive functioning. 75 The test consists of 3 stimulus cards containing 100 words, 100 colored rectangles, and 100 color-words, respectively. Two studies, both in breast cancer, used the Stroop test for cognitive function.26,29
Sustained Attention to Response Task (SART)
The SART is a computerized go/no-go test that requires participants to withhold behavioral response to a single, infrequent target (ie, the number 3) presented against a background of frequent non-targets (ie, numbers 0-2, 4-9). 76 This specific style of measurement is indicative of working memory, sustained attention, and impulse/inhibitory control. Reaction time for all key presses is also collected. After completion, the patient is asked 2 questions regarding their attention using a 6-point Likert scale. Gokal et al 29 utilized SART as part of a battery of assessments in breast cancer survivors.
Task-switch task
The Task-switch assessment asks patients to decipher the color of the square (pink or blue) and whether a number fits within a range (1-4 or 6-9) or is odd or even. 77 Numbers are presented to the patient for 2500 ms. If the background color of the box is blue, patients are asked to respond with their left hand if the number was higher or lower than 5. If the background color of the box is pink, the patient is asked to decipher if the number was odd or even with the right hand. Patients complete 48 trials and performance is defined as the reaction time across the trials. The task-switch test was used in a single study of breast cancer patients as part of a battery of assessments. 30
Trail making tests A and B
The Trail making test is a measure of visual conceptual and visuomotor tracking given in 2 parts (Part A and Part B). 78 For Part A, the patient draws lines to connect consecutively numbered circles on a worksheet. For Part B, the patient connects the same number of consecutively numbered and lettered circles on another worksheet by alternating between the 2 sequences. Speed is a factor in this study as the patient is urged to connect the circles as fast as possible as a measurement of attention/concentration and mental flexibility. Five studies utilized the Trail making test, either as a stand-alone 38 or as part of a battery.26,30,36,43
Verbal fluency tests
The verbal fluency tests include participants completing 2 tasks: category fluency (or semantic fluency) and letter fluency (or phonemic fluency). 79 Participants are given 1 minute to produce as many unique words as possible within a semantic category (category fluency) or starting with a given letter (letter fluency). The participant’s score in each task is the number of unique correct words. Arrieta et al 39 utilized the verbal fluency test as a standalone measure of cognitive function while Myers et al 36 included this assessment as part of a battery of tests.
Wechsler Adult Intelligence Scale (WAIS)
The Wechsler Adult Intelligence Scale indexes 4 factors: verbal comprehension, perceptual reasoning, working memory and processing speed. 80 Of particular interest for cognitive functioning the digit span test of the WAIS involves forward and backwards repetitions of a series of numbers as a measure of concentration and speed (attention/concentration). The block design test of the WAIS is a measure of visual spatial ability, constructional praxis, motor skill and problem-solving ability. Two studies utilized a component of this scale.29,43
Wilde Intelligence subtest
The Wilde Intelligence subtest can be administered either electronically via a computer-based application or in a traditional paper and pencil version to test working memory in cancer patients. 81 The sole study utilizing these assessments was a study conducted in Germany with breast cancer patients by Baumann et al. 24
Discussion
This review examined the exercise oncology literature for the various methods by which cognitive function is measured in cancer survivors who had a history of chemotherapy treatment. A total of 29 prospective studies were reviewed that uncovered the inconsistency in cognitive function measurements. The main findings of this review suggest that: (1) patient reported outcome (PRO) measurements are most commonly used to assess cognitive function in cancer patients and (2) standardized methods for measuring cognitive function are needed.
Subjective PROs were used to measure cognitive function in 22 of the articles reviewed. The EORTC-QLQ-C30 (n = 592 patients) and the Piper Fatigue Scale (n = 705 patients) were the 2 PROs used most frequently to evaluate cognitive function in this review. While easy to administer and with low patient burden, the global scores for cognitive function resulting from these surveys limit the applicability of its use. The Piper Fatigue Scale measures a combination of cognition and mood through 3 items, asking patients to reflect on their ability to concentrate, remember and think clearly. 60 The EORTC-QLQ-C30 only evaluates cognitive functioning through 2 questions that generally ask the patient if they have difficulty in concentrating on things like reading a newspaper or watching TV and if they have difficulty remembering things. Though these assessments may be beneficial in evaluating cognitive function generally, this also points to the potential bias and limitations PRO-based subjective surveys may impose, particularly given that the Piper Fatigue Scale31,40,44-46 and the ERTOC-QLQ-C3027,35,48,50-52 were the sole measure of cognitive function used in a number of studies included in this review. The FACT-Cog was the third most frequently used survey (n = 454 patients). The FACT-Cog and the PROMIS Survey (n = 134 patients) are the only surveys that have been validated in the cancer population to measure CI.
The surveys utilized in the studies included in this paper also leave major gaps in the evaluation of cognitive function in cancer survivors. Of the 8 surveys found in the literature review, only the FACT-Cog, MFIS and PROMIS surveys demonstrated the ability to measure more than one domain of cognitive functioning (see Table 4). Memory, as a global construct, was evaluated most often in breast cancer survivors. This was expected, given that the preponderance of exercise studies published to date have focused on breast cancer patients. Seven studies, 5 of which were interventional,37,41,44-46 utilized only a single PRO to measure cognitive outcomes as a primary endpoint. A limitation to the use of subjective assessments in evaluating cognitive function is the singular domain that these surveys often focus on. This raises cause for concern on the reliability of the exercise effect on cognition given the subjective nature of PROs and the limited scope of evaluating the totality of domains associated with cognition.
A rationale for the use of PRO surveys, the EORTC-QLQ-C30 and the FACT-Cog in particular, is the reduced investigator and patient burden that comes with administering these surveys. These assessments can be done in minutes rather than hours, there is the flexibility of patients completing these surveys at home (reducing clinic time) and the useful ability of these surveys to provide immediate feedback to the investigator due to the speed of scoring. These 2 surveys are also validated in multiple languages, providing a depth of normative and comparative data for investigators to evaluate against.
Consistency was observed more frequently in the surveys that were utilized by studies in this review. While PROs can be simple to administer and cost-effective ways of gaging cognitive function, these assessments rely on participant evaluation of memory and cognition which pose challenges due to the subjective nature of these assessments. 82 PRO tools only measure a participant’s perceived level of cognitive functioning, rather than objective cognitive assessment tools that measure actual function across cognitive domains.
Objective measurements of cognitive function provide more accurate and reliable evaluation of the different cognitive domains. The Trail Making Test (n = 441 patients) and the Verbal Fluency Test (n = 351 patients) were the 2 most reported objective measures of cognitive function in this review. The Trail Making Test is a measure of visual conceptual and visuomotor tracking given in 2 parts. Generally, the Trail Making Test requires visual scanning and working memory to complete the assessment with time being the main criterion used to measure impairment. This assessment is useful for studies that require outcomes on executive function, concentration, or psychomotor ability. The verbal fluency test is a psychological test that asks patients to produce as many words as possible from a category in a set time (usually 60 seconds). The category can either be semantic (i.e., categories such as animals) or phonemic (ie, words that begin with the letter T). In addition to the total number of unique words the patient verbalizes, researchers can also stratify their analysis by evaluating the number of repetitions, length of words or use of words that do not fit either the semantic or phonemic categories. Though large sample sizes were subjected to these 2 tests, it should be noted that the Trail Making Test was in the highest number of studies (n = 5) included in this review,26,30,36,38,43 while the verbal fluency test was utilized in only 2.36,39 Similarly, the Flanker Task Test, Mazes Task, and Task-Switch Test were completed by 249 patients, though the 3 were utilized only in a single study. 30 This points to the need for more consistent, standardized battery of assessments and methods for measuring and quantifying cognitive function in exercise oncology trials.
Even if investigators combined well-validated PROs and more objective neuropsychological tests, there are limitations to generalizing results. Poor correlation has been observed between PRO and objective measures of cognitive function across multiple cancer populations.83,84 Horowitz et al 85 stated that objective neuropsychological approaches lack a consensus of what cognitive domains are measured by these tests. These authors add that the use of a dichotomous result (ie, cognitive impairment or no cognitive impairment) fails to capture a broader picture of cognitive function as most patients fall in what is considered the low range of normal on many tests. These explanations may help explain the large spectrum of findings regarding exercise and cognitive function in oncology patients in this review.
The diverse assessment techniques used in the studies included in this review point to the heterogeneity and complexity of standardizing cognitive function measurements. While this scoping review was not limited to reviewing studies in which patients were only on treatment, or post treatment, there is likely to be an effect with respect to the timing of the cognitive function evaluations performed. For example, chemotherapy treatment has been found to have a time-sensitive effect on cognitive function, though the further away from completion of chemotherapy, the better recovery of cognitive function exists.12,86,87 Conversely, chemotherapy related CI may not arise for many years after completing therapy, therefore, detecting change in a short period of time may not be possible. Therefore, considerations on the use of objective measures based on timing of therapy and time off therapy should be considered when identifying appropriate cognitive function measures.
To this same effect, the heterogeneity of exercise interventions may also have an impact on cognitive function in cancer patients. The studies included in this review used a variety of interventions or strategies to evaluate exercise and physical activity habits on cognitive function in cancer patients. Some studies implemented structured exercise programs, while others evaluated physical activity habits using surveys and patient reported outcomes. Further, variability in the exercise prescription also poses barriers to evaluating the effectiveness of exercise in promoting cognitive function in cancer patients. Though specificity of training goes beyond the scope of this review, it needs to be acknowledged that the type and duration of training likely contribute to the mixed results in this review.
Comparison to the ICCTF Recommendations
The recommendations made by the ICCTF provides guidance on 6 areas: study design, longitudinal versus cross-sectional studies, pretreatment cognitive performance, utilization of appropriate control groups with varying treatment modalities, utilization of appropriate patient populations, excluding patients with depression and patients taking selective serotonin-reuptake inhibitors because of its ability to impact cognition, and single site versus multisite studies to include appropriate sample sizes. The study design recommendation indicates that while the gold standard double-blind randomized controlled, prospective, longitudinal study design is ideal, it may not always be feasible. Therefore, the ICCTF provides an alternative recommendation for the use of appropriate control groups when observational studies are utilized. Because of this broad recommendation, every study that was included in this review met this recommendation. When determining the congruency of the included studies to the recommendation of longitudinal studies, 9 studies were interventional with cognition assessed at multiple timepoints.24,26,29,41,43-47 The overwhelming majority of studies that included cognitive outcomes as a primary aim did not include assessments prior to the initiation of therapy, with one exception. 29 Many of these studies were evaluating the effectiveness of exercise either during treatment, 46 after treatment,24,26,41,43-45,47 which also limits the ability to assess cognitive function prior to cancer treatment.
In their recommendations, the ICCTF recommended that tests which measure learning and memory, processing speed and executive function be included in cognitive function studies in cancer patients. While not discounting the role of self-reported measures, the group recommended including assessments that have adequate psychometric properties, and that are available in multiple languages or are not language dependent. The group recommends the Hopkins Verbal Learning Test-Revised (HVLT-R) for learning and memory, 67 Trail Making Test (TMT) for psychomotor speed and aspects of executive function,78,88 and the Controlled Oral Word Association of the Multilingual Aphasia Examination for executive function. 89 Only Campbell et al, 26 met this recommendation. Details about these assessments can be found in their publication.17,18
Strengths and Limitations
To the author’s knowledge, no previous study has critically reviewed the various measures of cognitive function used in exercise oncology trials in patients that have completed chemotherapy. The aim of this review was to identify the types of cognitive function measurement tools used among clinically diverse cancer samples to provide a resource to investigators for future research and consensus building. As identified throughout this review, the types of measurements utilized to assess cognitive function are very heterogenous across the exercise oncology paradigm. The strength of this scoping review allows us to highlight a range of studies and pinpoint key areas of future research to further refine how we measure cognitive function within exercise oncology, though we are limited in our interpretation of any new research that has been published since the review was initiated. The inclusion of observational and interventional studies allows for a complete listing and a clear focus on the assessments currently utilized in the field. Although the lack of a rating and recommendation of the assessments seems to be a weakness, this evaluation was deliberately avoided. Recommendations based on limited broad use of individual assessments appear inappropriate as the objective presentation of the assessments was the primary aim of this paper. Instead, the results enable researchers to select methods from this paper that correspond to their individual research questions while providing resources for investigators to develop multimodal assessment batteries for use in future studies. The current lack of consensus could also point to shortcomings and the need for better standardization of measurement.
Future Considerations
As evident in this review, the mixed approaches to evaluating cognitive function in cancer patients poses a major limitation to understanding the role of exercise in maintaining cognitive function and preventing impairment. While PROs are time efficient, they pose bias in reporting with the potential of under- or over-reporting of subjective domains. Furthermore, a number of PROs incorporate cognitive function assessments as a secondary effect of other side effects (ie, fatigue or depression). Therefore, future research would benefit from separating the primary cognitive assessment from the secondary endpoint. More objective measurements are oftentimes time consuming and may require individuals with specialty training to be on the research team. Ideally, objective validated assessments that can be administered with minimal training and completed in a fraction of the time would be ideal.
The Cambridge Neuropsychological Test Automated Battery (CANTAB) may be a potential solution to the problem of lack of harmonization. 90 CANTAB assessments are (a) well-validated and for which the constructs tested are well-defined; (b) feasible, portable, and easy to administer and to take; (c) able to specifically assess cognitive domains across multiple domains. CANTAB is language-independent, available with Spanish instructions, culturally neutral, and requires no technical knowledge or prior familiarity with computers. It is largely self-administered with basic supervision from trained study team members. Cognitive tests are available to measure (a) attention and psychomotor speed, (b) executive functioning, (c) memory, and (d) emotion and social cognition; domains that are not objectively measured in PRO surveys. A customized battery can be created based on the subdomains the investigative team is interested in evaluating. Subdomain assessments can range between 7 and 18 minutes, thus allowing the investigator to evaluate multiple domains in a short amount of time. Similarly the NIH Toolbox Cognition Battery includes a number of objective assessments that have been validated for patients up to 85 years of age and include many of the assessments included in the ICCTF recommendations. 91 Both of these options compare favorably to the more than 2-hour length of a traditional neuropsychological battery that assesses some of the same cognitive domains.
Finally, future research should focus on utilizing objective assessments to measure cognitive function in a manner that is not burdensome to the patient, covers multiple domains of cognition and can be interpreted across different clinical conditions and exercise interventions. This could help to compare measures from research and clinical work with oncology patients, identify impairments and react with appropriate interventions. Developing a battery of assessments based on the ICCTF’s recommendations will help in reaching this goal and reduce the heterogeneity currently exhibited in the field and will move exercise oncology research forward.
Footnotes
Author’s Note
Paola Gonzalo-Encabo is also affiliated to Departamento de Ciencias Biomédicas, Área de Educación Física y Deportiva, Facultad de Medicina y Ciencias de la Salud, Universidad de Alcalá, Madrid, España.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
