Abstract
Background:
Chemotherapy is essential for treating acute myeloid leukemia (AML). Prior studies concluded that survivors of cancer who were treated with chemotherapy experience cognitive impairment. Therefore, it is important to understand cognitive function in survivors of AML.
Objective:
To explore distributions and correlates of cognitive function; and prediction of cognitive function on other outcomes in adults with AML who were treated with chemotherapy.
Methods:
A health science librarian systematically searched PubMed, CINAHL, PsycINFO, and Embase databases. Two reviewers independently conducted the title, abstract, and full-text screening. Data were extracted and synthesized based on the aims of the review.
Results:
A total of 10 articles were included. Findings indicate that up to 62.2% of adults with AML experienced impaired cognitive function after starting chemotherapy. Three studies found cognitive function remained stable over time. Education and cytokines were potential correlates of cognitive function. Worse cognitive function may predict lower physical performance and higher mortality, although the results were inconsistent across studies.
Conclusion:
Impaired cognitive function was observed in adults with AML who were treated with chemotherapy. However, no study used a validated subjective cognitive function-specific patient-reported questionnaire and previous studies focusing on cognitive function included relatively young samples. Hence, further research on cognitive function in older adults with AML is needed.
Implication for practice:
Due to the high prevalence of cognitive impairment identified, it is important to screen cognitive function in adults with AML who are planning to receive chemotherapy to intervene and provide support earlier.
Keywords: Cognitive Function, Cognitive Impairment, Acute Myeloid Leukemia, Chemotherapy, Systematic Review
Introduction
Cognitive impairment has been reported in survivors of cancer and involves deficits in memory, attention, executive function, and processing speed.1, 2 With these deficits, survivors of cancer may experience social isolation,3 decreased working capacity,3 lower quality of life (QOL),4 and increased mortality risk.5 According to Lange and colleagues,2 cognitive impairment may result from various factors, such as aging, cancer itself, genes and biomarkers, and toxicities of specific cancer treatments. Regardless of causality, decrements in cognitive function in survivors of cancer is known as cancer-related cognitive impairment (CRCI).6
CRCI—which has been colloquially labeled as “chemobrain” or “chemo fog”—has been reported in survivors of cancer undergoing anticancer therapy, during or after treatment completion,7, 8 and during long-term survivorship.9 Existing research focuses primarily on survivors of cancer with a solid tumor diagnosis, such as breast cancer.10 Prior studies of hematological cancers focused on pediatric populations with acute lymphoblastic leukemia or primary central nervous system lymphoma.11
For survivors with a diagnosis of acute myeloid leukemia (AML), chemotherapy is a standard treatment.12 In the United States, AML is the most common acute leukemia, with over 20 050 new cases anticipated in 2022.13 AML is especially common in older adults as the median age at diagnosis of 68 years.14 Because a large percentage of adults with AML receive chemotherapy and are older, it is important to understand the CRCI in this population. Furthermore, because radiation and surgery are rarely employed in adults with AML, cognitive impairment in this population after therapy may be due either to chemotherapy, the existential trauma of the diagnosis, biologic consequences of the leukemia, and aging.
The objectives of this systematic review were to 1) describe cognitive function at pre-, during, or post-chemotherapy; 2) identify potential correlates of cognitive function; and 3) explore how cognitive function predicts other outcomes in adults with AML.
Methods
This review was guided by the Preferred Reporting Items for Systematic reviews and Meta-Analyses.15 The review protocol is registered in PROSPERO (protocol # CRD 42020170338).
Search Strategy
The PubMed, CINAHL Plus with Full Text (EBSCOhost), APA PsycInfo (EBSCOhost), and Embase databases were searched by a health science librarian (JLC) on November 17, 2021. The search included a combination of key words and subject headings for AML, chemotherapy, and CRCI (Appendix I). No date restrictions and limitations were set except for the Embase database. During the Embase database search, we limited the search by including only articles as the publication type. All articles were imported into EndNote; after removing duplicates, the articles were then imported into Covidence for screening.
Data Extraction and Synthesis
Three authors (YNC, SB, and ALB) independently screened the title/abstract and full-text articles to identify studies for inclusion. Discrepancies were resolved by a third author. The inclusion criteria were: 1) full-text articles published in English, 2) included participants ≥18 years old, 3) reported cognitive function in adults with AML or a mixed sample of adults with AML or myelodysplastic syndromes (MDS) who were treated with chemotherapy, and 4) research studies with observational or experimental designs. We considered the studies eligible if they included participants with MDS because those with high-risk MDS have an increased risk of progressing to AML and may be considered for intensive chemotherapy or stem cell transplant.16 In addition, we included all chemotherapy regimens and dosage if they were identified as AML chemotherapy by the studies. For cognitive function outcome, we included overall cognitive function and cognitive domains results from the studies assessed using patient-reported questionnaires, objective neuropsychological assessments, or neuroimaging on adults with AML. Studies including survivors who received chemotherapy and those who did not receive chemotherapy but did not separately report cognitive function in participants who received chemotherapy were excluded.
Data were extracted using an Excel template created by the authors (YNC and SB) and included purpose, study design, sample size and characteristics, measure tool and time points, main findings, strengths and weaknesses, and implications. Findings were synthesized based on the aims of this review.
Quality Assessment
Instead of using the Mixed Methods Appraisal Tool (MMAT) registered in the PROSPERO, one author (YNC) assessed the quality of the included articles using the Joanna Briggs Institute (JBI) Systematic Reviews Critical Appraisal Tools.17 Specifically, for articles using longitudinal or cohort study design, we used the Checklist for Cohort Studies. For articles using cross-sectional design, we used the Checklist for Analytical Cross-Sectional Studies. Each item was assessed using “Yes”, “No”, “Unclear”, or “Not Applicable.” We did not exclude any articles based on the quality assessment results.
Results
The initial search identified 1,114 articles. After excluding duplicates and screening title/abstract, 44 articles were identified. After full-text screening, 10 articles were retained in the final synthesis (see Figure).
Figure.
PRISMA Flowchart of Study Selection Process
Study Characteristics and Study Quality
The 10 articles, from eight independent studies, were published between 2009 and 2021 across six countries (Table 1). Four were prospective longitudinal,18–21 four were prospective cohort,22–25 and two were cross-sectional.26, 27 For studies used repeated assessments over time, and intervals ranged from a month to a couple cycles of chemotherapy, with total study duration up to over 12 months. In the five studies for which cognitive function was not the main outcome, cognitive function was measured exclusively at baseline.23–27 The majority of studies focused on geriatric assessment.19, 22–24, 26 Three studies had cognitive function as the main outcome.20, 21, 27
Table 1.
Characteristics of the Included Studies (n=10)
First Author (Year) | Design | n | Treatment | Male (%) | Age in Years, Mean(SD)/Median | Length of Study |
---|---|---|---|---|---|---|
Alibhai (2009)18 | L | 20 | CT | 55 | 70 (NS) | 12m |
Jouzier (2021)22 | C | 397 | Intensive CT | 57 | 69 (NS) | >12m |
Klepin (2011)26a | CS | 54 | Intensive CT | 59 | 71 (6) | N/A |
Klepin (2013)23a | C | 74 | Intensive CT | 54 | 70 (6) or 69 | ? |
Klepin (2016)19a | L | 49 | Intensive CT | 57 | 69 | ≈3–4m |
Kotb (2019)27 | CS | 45b | CT | NS | NS | N/A |
Meyers (2005)20 | L | 54c | CT | 56 | 60 (NS) | 1m |
Modzelewski (2011)21 | L | 20 | Intensive CT | 45 | 38 (12) | 6m |
Molga (2020)24 | C | 98c | CT & non-CT | 63 | 77 | ? |
Oliva (2011)25 | C | 113 | CT & non-CT | 51 | 72 (6) | 12m |
The articles were from the same study.
The sample included other malignant hematological disease and only presented the number of adults with AML.
The sample included adults with myelodysplastic syndromes.
Abbreviations: ≈, approximately; ?, not able to determine; C, cohort; CS, cross-sectional; CT, chemotherapy; L, longitudinal; m, months; N/A, not applicable; non-CT, non-chemotherapy; NS, non-specified; SD, standard deviation.
All articles focused on recruiting a single group of participants with exposure of AML or MDS treatment (ie. chemotherapy or supportive care). The majority of the articles (70%) pre-identified and managed confounding factors by using statistical methods or publishing normative data. The quality of the included articles, appraised using the JBI Systematic Reviews Critical Appraisal Tools, can be found in Table 2.
Table 2.
Quality Assessment of the Included Studies (n=10)
First Author (Year) | Alibhai (2009)18 | Jouzier (2021)22 | Klepin (2011)26 | Klepin (2013)23 | Klepin (2016)19 | Kotb (2019)27 | Meyers (2005)20 | Modzelewski (2011)21 | Molga, (2020)24 | Oliva (2011)25 |
---|---|---|---|---|---|---|---|---|---|---|
Were the two groups similar and recruited from the same population? | Y | Y | - | Y | Y | - | Y | Y | Y | Y |
Were the exposures measured similarly to assign people to both exposed and unexposed groups? | N/A | N/A | - | N/A | N/A | - | N/A | N/A | N/A | N/A |
Was the exposure measured in a valid and reliable way? | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A |
Were confounding factors identified? | N | Y | Y | Y | Y | N | Y | Y | N | Y |
Were strategies to deal with confounding factors stated? | N | Y | Y | Y | Y | N | Y | Y | N | Y |
Were the groups/participants free of the outcome at the start of the study (or at the moment of exposure)? | N/A | N/A | - | N/A | N/A | - | N/A | N/A | N/A | N/A |
Were the outcomes measured in a valid and reliable way? | Y | Y | Y | Y | Y | Y | Y | Y | Y | Y |
Was the follow up time reported and sufficient to be long enough for outcomes to occur? | Y | Y | - | U | Y | - | Y | Y | U | Y |
Was follow up complete, and if not, were the reasons to loss to follow up described and explored? | Y | Y | - | Y | Y | - | Y | Y | Y | Y |
Were strategies to address incomplete follow up utilized? | N | Y | - | Y | N | - | N | N | Y | Y |
Was appropriate statistical analysis used? | Y | Y | Y | Y | Y | Y | Y | Y | Y | Y |
Were the criteria for inclusion in the sample clearly defined? | - | - | Y | - | - | Y | - | - | - | - |
Were the study subjects and the setting described in detail? | - | - | Y | - | - | Y | - | - | - | - |
Were objective, standard criteria used for measurement of the condition? | - | - | Y | - | - | Y | - | - | - | - |
Abbreviations: Y, yes; N, no; N/A, not applicable; U, unclear.
Sample Characteristics
Study sample sizes ranged between 20 and 397. Two studies included participants with a diagnosis of MDS20, 24 and one recruited participants with other types of malignant hematological disease.27 Most studies included participants with intensive chemotherapy only.19, 21–23, 26 Two studies also included participants who were treated with supportive care or immunomodulatory agents (for MDS), which did not include chemotherapy.24, 25 Across studies, similar numbers of men and women were included; participant mean age ranged from 38–72 years, and median age ranged from 69–77 years. For studies reporting race, White was predominant (95.9%–96.3%).19, 23, 26 For studies reporting education level, college degree or above was predominant (51.4%–57.5%).19, 23, 26
Main Findings
Change in cognitive function across chemotherapy continuum
Various approaches (patient-reported questionnaires, neuropsychological assessments, and neuroimaging) were used across studies to measure cognitive function (Table 3). Two studies reported subjective or perceived cognitive function while eight studies reported neuropsychological assessment and one study used additional neuroimaging to capture objective data. Of studies with subjective measures, all assessed cognitive function using the EORTC QLQ-C30 two-items cognitive function subscale with an aim to explore QOL among an AML population and reported mean or median cognitive function scores.18, 25 Specifically, one longitudinal study further examined the change of cognitive function over time and found a stable cognitive function across the treatment continuum up to 12 months.18
Table 3.
Main Cognition Outcome of the Included Studies (n=10)
First Author (Year) | Measures | Assessment Timepoints | Reported Format / Time Point | Change Overtime | Correlates Tested |
---|---|---|---|---|---|
Cognition assessed more than once | |||||
Alibhai (2009)18 | 2 items | pre-CT, 1m, 4m, 6m, 9m, 12m | mean score / all time points | Stable (no p-value) | remission status at 6m |
Jouzier (2021)22 | MMSE (<26 CI) | at dx, 3rd cycle re-ICT, 6th cycle re-ICT, end-CCT | mean score / at dx P of CI at dx: 16% |
NS | gender, performance status, age, grade 3–4 toxicity during induction, time of induction, lomustine during induction and postinduction therapy |
Klepin (2016)19 | 3MS (<77 CI) | pre-ICT, post-ICT | mean score / all time points P of CI: ≈> 20%→≈< 20% |
NS | - |
Meyers (2005)20 | Digit Span, Digit Symbol, HVLT, COWA-verbal fluency, TMT, GPT | pre-CT, 1m | mean score / pre-CT P of impaired domains- attention: 7%→8% psychomotor speed: 8%→13% total recall: 44%→58% immediate recognition: 7%→25% delayed recall: 41%→58% verbal fluency: 17%→25% visual scanning: 28%→38% executive function: 29%→46% dexterity: 37%→54% |
P of impaired dexterity: S↑ P of other impaired domains: NS |
age, education, hemoglobin, fatigue, cytokines |
Modzelewski (2011)21 | SPECT, ICARS, BI, TMT, 9HPT, DRS, Digit Span, GBVLT | at dx, pre-CCT, 2nd cycle CCT, during CCT, post-CCT, 6m | mean score / all time points P of normal test score: 100% |
SPECT at dx vs. pre-CCT vs. 2nd cycle CCT: NS R’t hand dexterity: S↑ during CCT–post-CCT L’t hand dexterity: S↓ 2nd cycle CCT–during CCT & S↑ post-CCT–6m attention: S↑ post-CCT–6m working memory: S↑ post-CCT–6m & S↑ at dx–pre-CCT verbal memory: S↑ at dx–6m global efficiency: S↑ at dx–6m |
- |
Cognition assessed once | |||||
Klepin (2011)26 | 3MS (<80 CI) | within 5d of initial hospitalization | mean score P of CI: 31.5% |
N/A | cytogenetics |
Klepin (2013)23 | 3MS (<77 CI) | within 5d of initial hospitalization | median score P of CI: 28.8% |
N/A | - |
Kotb (2019)27 | MOCA (<26 CI) | 6m–2y post-CT | mean score of each domain P of CI: 62.2% |
N/A | - |
Molga (2020)24 | MMSE (<24 CI) | pre-treatment | N/A | N/A | - |
Oliva (2011)25 | 2 items | at dx | IC group median score |
N/A | treatment |
Note: 2 items= cognitive function domain from the European Organization for the Research and Treatment of Cancer Quality of Life Questionnaire (EORTC QLQ-C30), ≈ approximately, ↑ increased, ↓ decreased, S significant, L’t=left, R’t=right
Abbreviations: 3MS, Modified Mini-Mental State Examination; 9HPT, Nine Hole Peg Test; BI, Barthel Index; CCT, consolidation chemotherapy; CF, cognitive function; CI, cognitive impairment; COWA, Controlled Oral Word Association; CT, chemotherapy; d, days; DRS, Mattis Dementia Rating Scale; dx, diagnosis; GBVLT, Grober and Buschke Verbal Learning test; GPT, Grooved Pegboard Test; HVLT, Hopkins Verbal Learning Test; IC, intensive chemotherapy; ICARS, International Cooperative Ataxia Rating Scale; ICT, induction chemotherapy; m, months; MMSE, Mini-Mental State Exam; MOCA, Montreal Cognitive Assessment; N/A, not applicable; NS, non-specified; p, prevalence; SPECT, single-photon emission computerized tomography; TMT, Trial Making Test; y, years.
Among the eight studies using objective assessment tools, two used the Mini-Mental State Exam (MMSE), three used the Modified Mini-Mental State Examination (3MS), one used the Montreal Cognitive Assessment (MOCA), and two used a battery of neuropsychological assessments. For the six studies using the MMSE, 3MS, and MOCA to assess overall cognitive function, two longitudinal studies assessed the change of cognitive function using the MMSE and 3MS, respectively, from diagnosis of AML up to the end of consolidation chemotherapy and found that cognitive function remained stable over time (p=.55; p=.72).19, 22 In addition to reporting mean or median score, a specific cut-off score for cognitive impairment was defined and the prevalence of cognitive impairment was further identified. Three articles from the same study included the 3MS as part of the geriatric assessment and found the prevalence of cognitive impairment ranged from above 20% to 31.5% prior to induction chemotherapy/within 5 days of initial hospitalization for AML19, 23, 26 and less than 20% after induction chemotherapy.19 One study conducted the assessment using the MMSE at AML diagnosis and reported 16% of participants had cognitive impairment.22 A cross-sectional study assessed cognitive function at six months to two years after chemotherapy using the MOCA and observed 62.2% prevalence of cognitive impairment.27 These studies concluded that although cognitive function remained stable over time, the prevalence of cognitive function was 16%–31.5% at the initiation of chemotherapy and up to 62.2% after starting chemotherapy.
CRCI shows impairment in different cognitive domains; therefore, researchers in two studies used batteries of neuropsychological assessments to better understand how each domain is impacted.20, 21 Specifically, in one study, mixed results were identified in severity of impairment related to hand dexterity, attention, working memory, verbal memory, and global efficiency between time points across the treatment continuum.21 However, the other study identified a non-significant increase in prevalence of impairment related to attention, psychomotor speed, total recall, immediate recognition, delayed recall, verbal fluency, visual scanning, and executive function between pre- and post-induction chemotherapy20 and a significant increase in prevalence of impaired dexterity.20 Unfortunately, with only two studies using batteries of neuropsychological assessment with a follow-up time frame between one and six months, we had limited evidence for drawing a powerful conclusion on how chemotherapy impacts domains of cognitive function in the AML population. These findings highlight the necessity for using batteries of neuropsychological assessments in future CRCI research.
Neuroimaging is an evolving method to understand cognitive function. In addition to the neuropsychological assessments, one study further used single-photon emission computerized tomography (SPECT) to measure brain perfusion and indicated non-significant differences between pre-chemotherapy and post-chemotherapy.21 However, with this single study, we are hesitant to draw conclusions.
Potential correlates of cognitive function
The following potential correlates were tested across studies:
Age.
The relationship between age and cognitive function was tested in two studies and both reported non-significant findings.20, 22 Specifically, one study found that age did not correlate with the change of cognitive function over time.22 Another study found a non-significant relationship between age and cognitive function.20
Gender.
One study observed no significant difference in the change of cognitive function over time between gender.22
Education.
One study found a significant positive association between years of education and cognitive function.20
Biomarkers.
One study examined the relationship between biomarkers and cognitive function and reported mixed findings.20 For cytogenetics, the study found a significant negative association between IL-6 and executive function, and a positive association between IL-8 and memory at pre-chemotherapy.20 Yet, no significant relationship was found between hemoglobin and cognitive function at pre-chemotherapy.20
Disease characteristics.
Relationships between disease characteristics and cognitive function was assessed in two studies that included: cytogenetics (favorable to intermediate vs. unfavorable),26 and remission status at 6 months (p=.88);18 however, no significant relationships were identified.
Chemotherapy regimens.
Comparisons in cognitive function across different AML chemotherapy regimens were made in two studies; both generated no significant findings.22, 25 One cross-sectional study reported no difference in cognitive function at AML diagnosis between patients receiving intensive and palliative treatment (p=.52).25 Similarly, the other longitudinal study reported that with or without lomustine during induction and post-induction chemotherapy did not associate with the change in cognitive function over time (p=.61).22
Time of induction.
Jouzier and colleagues22 further tested the relationship with time of induction and also discovered non-significant findings.
Fatigue.
In one study, no significant relationship between fatigue and cognitive function was found prior to starting chemotherapy.20
Functional status.
Jouzier and colleagues22 reported a 16% and 34% of cognitive impairment in participants with the Eastern Cooperative Oncology Group (ECOG) performance status 0–1 and those with the ECOG performance status 2, respectively (no p-value tested). This same study also found a non-significant relationship between the ECOG performance status and the change of cognitive function over time.22
To summarize, only education and cytokines were identified to significantly correlate with cognitive function. However, these findings were only reported in one study. In contrast, age, gender, hemoglobin, cytogenetics, remission status, treatment regimens, hemoglobin, fatigue, and functional status were not significantly associated with cognitive function.
Prediction of cognitive function on other outcomes
Cohort studies with a focus on geriatric assessments examined the prediction of cognitive function on disease status, treatment plan, symptoms, and physical performance among adults with AML. One study found that cognitive function at initiation of chemotherapy had non-significant effect on remission (3MS>77 vs. 3MS<77= 67% vs. 57%, p=.41).23 Another study reported that participants with cognitive impairment at pre-chemotherapy completed significant fewer cycles of azacitidine chemotherapy than those without cognitive impairment at pre-chemotherapy (MMSE<24 vs. MMSE≥24: 3.5 ± 2.1 vs. 10.9 ± 7.9; p=.03).24 Using six cycles of azacytidine chemotherapy as a cut-off point, this same study also found a significant higher percentage of participants with cognitive impairment at pre-chemotherapy were not able to complete the six cycles than those without cognitive impairment at pre-chemotherapy (MMSE<24 vs. MMSE≥24: 75% vs. 24%, p=.05).24
Another cohort study identified that cognitive function at diagnosis did not predict the occurrence of grade 3–4 toxicities during induction chemotherapy (MMSE≥26 vs. MMSE<26: no toxicities number= 83.3% vs. 16.7% & toxicities number= 78.7% vs. 21.4%, p=.31).22 In terms of the prediction on physical performance, one longitudinal study identified a significant positive relationship between baseline cognitive function and the improvement of physical performance from pre- to post-induction chemotherapy, after adjusting for cytogenetic risk, depression, balance, and performance status (p=.05).19 Also, this same study reported that participants with cognitive impairment (3MS<77) at both pre- and post-induction chemotherapy had more decrease in physical performance (Short Physical Performance Battery: 8.0±1.7 vs. 3.6±1.7, p=.07) compared to those without cognitive impairment (Short Physical Performance Battery: 8.4±0.7 vs. 6.9±0.7, p=0.12).19 However, all these findings were tested in only one study. We are therefore unable to surmise a definitive conclusion about the prediction of cognitive function on disease, treatment plan, symptoms, and physical performance.
Findings about the prediction of cognitive function on mortality were tested in one study.23 Klepin and colleagues23 found a non-significant higher percentage of 30-day mortality (within 30 days since starting induction chemotherapy) in participants with cognitive impairment than those who did not have cognitive impairment (3MS<77: 23.8%, [95%CI=8.2–47.2] vs. 3MS≥77: 9.6%, [95%CI=3.0–21.0], p=.14). However, this same study also discovered a significant positive relationship between cognitive impairment and risk of death after adjusting for covariates (HR 2.5, [95%CI=1.2–5.5]).23 The prediction of cognitive function on survival was mixed across three studies.22–24 Klepin and colleagues23 discovered that participants with cognitive impairment within 5 days of initial hospitalization had a significantly lower median overall survival than those without cognitive impairment (3MS<77: 5.2 months vs. 3MS≥77: 15.6 months, p=.002). However, Jouzier and colleagues22 found that there was non-significant prediction of cognitive function at diagnosis on both overall survival (MMSE≥26 vs. MMSE<26: median overall survival= 26 months [95%CI=22.4–31.2] vs. 21 months [95%CI=13.1–36.8], p=.67) and event free survival (MMSE≥26 vs. MMSE<26: median event free survival=16 months [95%CI=12.5–19] vs. 13 months [95%CI=9.7–22.5], p=.81). Similarly, another study also identified that there was no significant difference in survival between participants who received azacytidine chemotherapy with and without cognitive impairment (12 months vs. 19 months, p=1).24 Although these longitudinal or cohort studies with a focus on geriatric assessments emphasize on a comprehensive domains of aging, such as functional status, mobility, nutrition, and cognition; these exploratory findings reinforce the importance of understanding cognition in this patient population because of its potential prediction on multiple aspects of adults with AML.
Discussion
We found in our systematic review that, in extant AML research, cognitive function was studied as one of the components of QOL and as an important factor of geriatric assessment. Our review concluded that up to 62.2% of adults with AML experienced cognitive impairment; however, stable cognitive function after initiating chemotherapy was identified when measured either using patient-reported questionnaires or objective assessments.
Although chemotherapy is a standard treatment for adults with AML, we identified limited research focusing on the impact of AML chemotherapy on CRCI. Of the two studies including patient-reported questionnaire, both used the EORTC QLQ-C30 cognitive function subscale.18, 25 However, considering the EORTC QLQ-C30 is a QOL measure, and the cognitive function subscale only measures two cognitive domains,28 it might not be sensitive enough to capture changes in cognitive function following chemotherapy. Additionally, EORTC QLQ-C30 provides different information about cognitive function compared to neuropsychological assessments. Consequently, we identified a gap in researchers’ understanding of perceived cognitive function as reported via cognitive function-specific self-reported questionnaires. Perceived cognitive function is critically important29 to facilitate early recognition of declines in cognitive function30 and increase clinicians’ ability to detect CRCI.31 Hence, these findings highlight a necessity for researchers to understand CRCI in the AML population and use cognitive function-specific patient-reported measures.
Objective assessments were widely used in CRCI studies. The MMSE and 3MS (which was expanded from the MMSE)32 were the most used objective assessment in our review, specifically in studies focusing on geriatric assessment.19,22–24,26 However, the MMSE may not be sensitive enough to detect cognitive changes in cancer populations31 because it is a screening measure of cognitive function and thus provides limited information. Although batteries of neuropsychological assessments and neuroimaging are recommended by the International Cancer and Cognition Task Force,1,33 only two studies used them.20,21 With this small number of studies, we were unable to conclude how chemotherapy impacts cognition domains in adults with AML. To address this gap, we highlight a need for researchers to use batteries of neuropsychological assessment to understand CRCI in AML.
Understanding the correlates of cognitive function will enable early identification and intervention. Lower education correlated with worse cognitive function,20 which was consistent with findings of a prior study of lymphoma survivors.34 Although aging may contribute to a lower cognitive function in some domains, our review found no relationship between age and cognitive function.20,22 One study included in our review found no relationship between fatigue and cognitive function,20 which was different from the results of a prior study reporting fatigue as a predictor of decreased cognitive function in survivors of cancer.35 This inconsistency might be due to the measures used: Meyers, Albitar, and Estey20 used a battery of neuropsychological assessments, while Oh35 used the MMSE and a patient-reported questionnaire. The differences between patient-reported cognitive function and neuropsychological assessments have been identified in prior research. Specifically, O’Farrell, Smith, and Collins36 found that patient-reported cognitive function was negatively associated with anxiety and fatigue, while objective cognitive function measured by a battery of neuropsychological assessments was not. In addition, survivors of cancer with a low hemoglobin level may experience symptoms (i.e. fatigue, dizziness, headache) which might contribute to cognitive decline.37 However, no significant relationship was found between hemoglobin level and cognitive function in our review.20 Researchers proposed that inflammation is a possible etiology of CRCI.2 A prior study also identified a significant negative relationship between cytokines (IL-1β, TNF-α and IL-4) and cognitive function.38 However, the identified cytokines were different from the ones identified in our review.20 Finally, no psychological correlates were tested. A recent systematic review of 19 studies focusing on survivors of breast cancer concluded that survivors with higher levels of psychological distress indicated worse cognitive function.39 Therefore, the relationship between emotional distress and cognitive function should be further studied in AML survivors. To summarize, the small number of studies testing each correlate and their inconsistent use of cognitive function measures limits the power of our findings. The lack of strong evidence reinforces the importance of using cognitive function-specific measures and further testing the correlates in future studies.
Limited studies included in our review found that worse cognitive function contributes to poorer outcomes in adults with AML. In our review, one study found worse cognitive function causes an early cessation of azacytidine therapy.24 Azacytidine, one of the hypomethylating agents, was prescribed in combination with Venetoclax to treat AML survivors who are precluded from intensive chemotherapy.40 This treatment plan includes daily oral medication, Venetoclax, which requires AML survivors to adhere to their treatment protocol. Unfortunately, prior research found that poor cognitive function may lead to a nonadherence to oral cancer treatment over time.41 Another study reported that adults with AML experiencing worse cognitive function have poorer chances of survival.23 This finding aligned with prior studies, which indicated a significant positive relationship between decreased cognitive function and mortality in community-dwelling older adults.42 Studies focusing on other cancer diagnosis found that survivors of cancer with CRCI may experience social isolation,3 poor working capacity,3 and decreased QOL.4 Considering differences in patients’ characteristics, how CRCI impacts survivors of cancer may vary. For example, Boykoff, Moieni, and Subramanian43 found that the symptoms of CRCI made survivors of breast cancer easily distractible and decreased work efficiency, forcing them to switch to jobs with lighter workloads. However, survivors of breast cancer have a median age of diagnosis of 62 years,44 while AML is 68 years, which is over standard retirement age.14 Hence, work capacity might not be the focus in the AML population. Instead, AML survivors may be more focused on other tasks, so researchers cannot assume that CRCI has the same impact in AML survivors as it does in different cancer populations. A research gap exists, and qualitative studies are needed to understand experiences of older adults with AML with CRCI.
Our review included ten articles reporting cognitive function; because we did not restrict to CRCI studies, we obtained a broader understanding of current cognitive function study in AML population. However, the sample sizes were varied across studies and some major findings were only tested in a single study, which made it hard to draw conclusions on the cognitive function in this population. In addition, the three studies on CRCI were not representative of the general AML population because the age of their samples (mean age: 38–60 years) were younger than the median age of diagnosis (68 years)20, 21 and did not focus exclusively on AML survivors.20, 27 Due to these weaknesses in the sample ages and diagnosis, we know little about CRCI in older adults with AML. Thus, older adults should be recruited in future CRCI research in the AML population.
This review is not without limitation. First, no gray literature was searched. Therefore, unpublished reports and conference proceedings related to CRCI in AML survivors were missed during the searching process. Second, due to the limited research about CRCI in hematological malignancies, studies recruited samples with a MDS, or AML diagnosis were included in the review. The results we captured might not be the experiences of AML survivors exclusively.
Conclusion
Our review provided a thorough understanding of current research on cognitive function in adults with AML who were treated with chemotherapy. Cognitive function may impact survivors in different aspects. However, stable cognitive function was identified after initiating chemotherapy. Furthermore, some potential correlates of cognitive function—education and cytokines—were identified.
Our review identified several CRCI research gaps in the AML population. First, a lack of using cognitive function-specific measures limited the power of findings related to understanding experiences and correlates of CRCI. Second, there is no research exploring how CRCI impacts adults with AML. Lastly, CRCI studies included a variety of diagnoses and relatively younger samples. Therefore, future researchers should use patient-reported questionnaires and objective assessments (such as a battery of neuropsychological assessments) to explore CRCI and correlates in adults with AML—in particular, older adults. Furthermore, given the emergence of newer treatments for AML in older adults, there will be more choices in therapies and the risks of CRCI with such therapies (both new and old treatments) need to be consistently defined in order to empower patients to choose according to their goals. Finally, considering differences in patients’ characteristics (ie. Age), how CRCI impacts survivors of cancer may vary. It is also crucial to use a qualitative approach to explore how CRCI impacts this patient population. Once a preliminary understanding is obtained, quantitative studies with fully powered sample sizes and intervention development can be further conducted. By doing so, we will have an in-depth and comprehensive understanding of CRCI in adults with AML.
Supplementary Material
Funding Support:
Ya-Ning Chan received Carol Ann Beerstecher Graduate Nursing Scholarship (2019–2020), Class of 1967 Forever Fund Scholarship (2020–2021), Helen Watkins Umphlet Graduate Scholarship (2020–2021), and Elizabeth Scott Carrington Nursing Scholarship (2021–2022) from the School of Nursing at University of North Carolina at Chapel Hill and Doctoral Degree Scholarship in Cancer Nursing (DSCN-20-076-01-SCN) from the American Cancer Society; Stephanie Betancur received funding support from the Hillman Scholars Program in Nursing Innovation and T32 Training grant from the National Institutes of Health.
Footnotes
Conflicts of interests: The authors have no conflicts of interests to disclose.
References
- 1.Deprez S, Kesler SR, Saykin AJ, Silverman DH, De Ruiter MB, McDonald BC. International cognition and cancer task force recommendations for neuroimaging methods in the study of cognitive impairment in non-CNS cancer patients. JNCI. 2018;110(3):223–231. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Lange M, Joly F, Vardy J, et al. Cancer-related cognitive impairment: an update on state of the art, detection, and management strategies in cancer survivors. Ann Oncol. 2019;30(12):1925–1940. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Selamat MH, Loh SY, Mackenzie L, Vardy J. Chemobrain experienced by breast cancer survivors: a meta-ethnography study investigating research and care implications. PLoS One. 2014;9(9):e108002–e108002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Lycke M, Lefebvre T, Pottel L, et al. Subjective, but not objective, cognitive complaints impact long-term quality of life in cancer patients. J Psychosoc Oncol. 2019;37(4):427–440. [DOI] [PubMed] [Google Scholar]
- 5.Robb C, Boulware D, Overcash J, Extermann M. Patterns of care and survival in cancer patients with cognitive impairment. Crit Rev Oncol Hematol. 2010;74(3):218–224. [DOI] [PubMed] [Google Scholar]
- 6.Lindner OC, Phillips B, McCabe MG, et al. A meta-analysis of cognitive impairment following adult cancer chemotherapy. Neuropsychology. 2014;28(5):726–740. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Hess LM, Huang HQ, Hanlon AL, et al. Cognitive function during and six months following chemotherapy for front-line treatment of ovarian, primary peritoneal or fallopian tube cancer: an NRG oncology/gynecologic oncology group study. Gynecol. Oncol 2015;139(3):541–545. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Moore K, Stutzman S, Priddy L, Olson D. Chemobrain: A Pilot Study Exploring the Severity and Onset of Chemotherapy-Related Cognitive Impairment. Clin J Oncol Nurs. 2019;23(4):411–416. [DOI] [PubMed] [Google Scholar]
- 9.Stouten-Kemperman MM, de Ruiter MB, Caan MW, et al. Lower cognitive performance and white matter changes in testicular cancer survivors 10 years after chemotherapy. Hum Brain Mapp. 2015;36(11):4638–4647. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Bray VJ, Dhillon HM, Vardy JL. Systematic review of self-reported cognitive function in cancer patients following chemotherapy treatment. J Cancer Surviv. 2018;12(4):537–559. [DOI] [PubMed] [Google Scholar]
- 11.Williams AM, Zent CS, Janelsins MC. What is known and unknown about chemotherapy-related cognitive impairment in patients with haematological malignancies and areas of needed research. British journal of haematology. 2016;174(6):835–846. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.National Comprehensive Cancer Network. NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines): Acute Myeloid Leukemia Version 1.2018 Available at: https://oncolife.com.ua/doc/nccn/Acute_Myeloid_Leukemia.pdf. Accessed August 16, 2019.
- 13.Siegel RL, Miller KD, Fuchs HE, Jemal A. Cancer statistics, 2022. Cancer. 2022;72(1):7–33. [DOI] [PubMed] [Google Scholar]
- 14.National Cancer Institute. SEER Cancer Stat Facts: Colorectal Cancer. Available at: https://seer.cancer.gov/statfacts/html/colorect.html. Accessed September 30, 2019.
- 15.Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Ann Intern Med. 2009;151(4):264–269. [DOI] [PubMed] [Google Scholar]
- 16.Bewersdorf JP, Carraway H, Prebet T. Emerging treatment options for patients with high-risk myelodysplastic syndrome. Ther Adv Hematol. 2020;11:1–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Moola SMZ, Tufanaru C, Aromataris E, et al. JBI Manual for Evidence Synthesis (M. Z. Aromataris E, Ed.). Available at: https://synthesismanual.jbi.global Accessed December 28, 2021. [Google Scholar]
- 18.Alibhai SM, Leach M, Gupta V, et al. Quality of life beyond 6 months after diagnosis in older adults with acute myeloid leukemia. Crit Rev Oncol Hematol. 2009;69(2):168–174. [DOI] [PubMed] [Google Scholar]
- 19.Klepin HD, Tooze JA, Pardee TS, et al. Effect of intensive chemotherapy on physical, cognitive, and emotional health of older adults with acute myeloid leukemia. J Am Geriatr Soc. 2016;64(10):1988–1995. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Meyers CA, Albitar M, Estey E. Cognitive impairment, fatigue, and cytokine levels in patients with acute myelogenous leukemia or myelodysplastic syndrome. Cancer. 2005;104(4):788–793. [DOI] [PubMed] [Google Scholar]
- 21.Modzelewski R, Lepretre S, Martinaud O, et al. Brain perfusion in adult patients with acute myeloblastic leukemia before and after cytosine arabinoside. Mol Imaging Biol. 2011;13(4):747–753. [DOI] [PubMed] [Google Scholar]
- 22.Jouzier C, Hamel J-F, Dumas P-Y, et al. Conventional chemotherapy for acute myeloid leukemia in older adults: Impact on nutritional, cognitive, and functional status. Eur J Haematol. 2021;106(6):859–867. [DOI] [PubMed] [Google Scholar]
- 23.Klepin HD, Geiger AM, Tooze JA, et al. Geriatric assessment predicts survival for older adults receiving induction chemotherapy for acute myelogenous leukemia. Am J Hematol. 2013;121(21):4287–4294. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Molga A, Wall M, Chhetri R, et al. Comprehensive geriatric assessment predicts azacitidine treatment duration and survival in older patients with myelodysplastic syndromes. J Geriatr Oncol. 2020;11(1):114–120. [DOI] [PubMed] [Google Scholar]
- 25.Oliva EN, Nobile F, Alimena G, et al. Quality of life in elderly patients with acute myeloid leukemia: patients may be more accurate than physicians. Haematologica. 2011;96(5):696–702. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Klepin HD, Geiger AM, Tooze JA, et al. The feasibility of inpatient geriatric assessment for older adults receiving induction chemotherapy for acute myelogenous leukemia. J Am Geriatr Soc. 2011;59(10):1837–1846. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Kotb MG, Ibrahim RI, Said RMM, El Din MMW. Chemotherapy-induced cognitive impairment in hematological malignancies. Egypt J Neurol Psychiatr Neurosurg. 2019;55(1):56. [Google Scholar]
- 28.European Organisation for Research and Treatment of Cancer. Quality of life of cancer patients. Available at: http://qol.eortc.org/questionnaire/eortc-qlq-c30/. Accessed November 11, 2018.
- 29.Savard J, Ganz PA. Subjective or Objective Measures of Cognitive Functioning—What’s More Important?Subjective vs Objective Measures of Cognitive FunctioningSubjective vs Objective Measures of Cognitive Functioning. JAMA Oncology. 2016;2(10):1263–1264. [DOI] [PubMed] [Google Scholar]
- 30.Lai J-S, Butt Z, Wagner L, et al. Evaluating the Dimensionality of Perceived Cognitive Function. Journal of Pain and Symptom Management. 2009/06/01/ 2009;37(6):982–995. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Isenberg-Grzeda E, Huband H, Lam H. A review of cognitive screening tools in cancer. Curr Opin Support Palliat Care. 2017;11(1):24–31. [DOI] [PubMed] [Google Scholar]
- 32.Teng EL, Chui HC. The Modified Mini-Mental State (3MS) examination. J Clin Psychiatry. 1987;48(8):314–318. [PubMed] [Google Scholar]
- 33.Wefel JS, Vardy J, Ahles T, Schagen SB. International Cognition and Cancer Task Force recommendations to harmonise studies of cognitive function in patients with cancer. Lancet Oncol. 2011;12(7):703–708. [DOI] [PubMed] [Google Scholar]
- 34.Wouters H, Baars JW, Schagen SB. Neurocognitive function of lymphoma patients after treatment with chemotherapy. Acta Oncologica. 2016;55(9–10):1121–1125. [DOI] [PubMed] [Google Scholar]
- 35.Oh P-J. Predictors of cognitive decline in people with cancer undergoing chemotherapy. Eur J Oncol Nurs. 2017;27:53–59. [DOI] [PubMed] [Google Scholar]
- 36.O’Farrell E, Smith A, Collins B. Objective–subjective disparity in cancer-related cognitive impairment: does the use of change measures help reconcile the difference? Psycho-oncology. 2017;26(10):1667–1674. [DOI] [PubMed] [Google Scholar]
- 37.Cunningham RS. Anemia in the oncology patient: cognitive function and cancer. Cancer Nurs. 2003;26:38S–42S. [DOI] [PubMed] [Google Scholar]
- 38.Zhao J, Zuo H, Ding K, Zhang X, Bi Z, Cheng H. Changes in plasma IL-1β, TNF-α and IL-4 levels are involved in chemotherapy-related cognitive impairment in early-stage breast cancer patients. Am J Transl Res. 2020;12(6):3046–3056. [PMC free article] [PubMed] [Google Scholar]
- 39.Yang Y, Hendrix CC. Cancer-Related Cognitive Impairment in Breast Cancer Patients: Influences of Psychological Variables. Asia-Pac J Oncol Nurs. 2018;5(3):296–306. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Food and Drug Administration. FDA approves venetoclax in combination for AML in adults. Available at: https://www.fda.gov/drugs/fda-approves-venetoclax-combination-aml-adults.
- 41.Bender CM, Gentry AL, Brufsky AM, et al. Influence of patient and treatment factors on adherence to adjuvant endocrine therapy in breast cancer. Oncol Nurs Forum. 2014;41(3):274–285. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Lv X, Li W, Ma Y, et al. Cognitive decline and mortality among community-dwelling Chinese older people. BMC Medicine. 2019/03/15 2019;17(1):63. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Boykoff N, Moieni M, Subramanian SK. Confronting chemobrain: an in-depth look at survivors’ reports of impact on work, social networks, and health care response. J Cancer Surviv. 2009;3(4):223. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.National Cancer Institute. SEER Cancer Stat Facts: Breast Cancer. Available at: https://seer.cancer.gov/statfacts/html/breast.html. Accessed September 30, 2019.
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.