Cognitive impairment after cancer treatment: mechanisms, clinical characterization, and management
BMJ 2023; 380 doi: https://doi.org/10.1136/bmj-2022-071726 (Published 15 March 2023) Cite this as: BMJ 2023;380:e071726
- Ben Fleming, research assistant1,
- Paul Edison, clinical reader in neuroscience and professor, honorary consultant1 2,
- Laura Kenny, clinical senior lecturer, consultant medical oncologist3
- 1Department of Brain Sciences, Faculty of Medicine, Imperial College London, London, UK
- 2College of Biomedical and Life Sciences, Cardiff University, Cardiff, UK
- 3Department of Surgery and Cancer, Faculty of Medicine, Imperial College London, London, UK
- Correspondence to: L Kenny l.kenny{at}imperial.ac.uk
Abstract
Cognitive impairment is a debilitating side effect experienced by patients with cancer treated with systemically administered anticancer therapies. With around 19.3 million new cases of cancer worldwide in 2020 and the five year survival rate growing from 50% in 1970 to 67% in 2013, an urgent need exists to understand enduring side effects with severe implications for quality of life. Whereas cognitive impairment associated with chemotherapy is recognized in patients with breast cancer, researchers have started to identify cognitive impairment associated with other treatments such as immune, endocrine, and targeted therapies only recently. The underlying mechanisms are diverse and therapy specific, so further evaluation is needed to develop effective therapeutic interventions. Drug and non-drug management strategies are emerging that target mechanistic pathways or the cognitive deficits themselves, but they need to be rigorously evaluated. Clinically, consistent use of objective diagnostic tools is necessary for accurate diagnosis and clinical characterization of cognitive impairment in patients treated with anticancer therapies. This should be supplemented with clinical guidelines that could be implemented in daily practice. This review summarizes the recent advances in the mechanisms, clinical characterization, and novel management strategies of cognitive impairment associated with treatment of non-central nervous system cancers.
Introduction
The global cancer burden is estimated to have increased by 19.3 million new cases in 2020, and cancer is a leading cause of death and reduced life expectancy globally.1 Chemotherapy has improved year-on-year survival rates for patients with cancer, with the 50% survival rate extending from one to 10 years over a 40 year period.2 However, a multitude of detrimental side effects after treatment, such as nausea, fatigue, hair loss, and cognitive impairment, erode the quality of life of survivors.3 Multi-organ toxicity across the body often underlies these side effects; in the brain, chemotherapeutic agents and chimeric antigen receptor (CAR) T cell immunotherapy can cause cytokine release syndrome,4 in which elevated cytokine concentrations dysregulate blood-brain barrier permeability, giving rise to neurotoxic CD8 positive T cell infiltration and cognitive impairment.5
Cognitive impairment associated with cancer treatment, colloquially termed “chemobrain” or “chemofog,” is characterized by a decline in performance in cognitive function related to learning, attention, executive functions, memory, multitasking, and processing speed,6 and is associated with chemotherapy, hormone therapy, immunotherapy, and targeted therapies.7 Prevalence of clinically significant cognitive impairment varies between 17% and 78% with self-reported measures,8 and is approximately 33% using objective neurocognitive testing in post-chemotherapy patients with breast cancer.9 Although cognitive impairment is a key factor in preventing patients from regaining their previous quality of life, no management strategies or clinical guidelines are available. Here, we review the epidemiology, recent evidence, and mechanistic underpinnings of cognitive impairment associated with chemotherapy, hormone therapy, immunotherapy, and some targeted therapies in non-central nervous system (CNS) cancers. Additionally, we discuss the risk factors, biomarkers for diagnosis, and guidelines for the evaluation of these patients, outlined in figure 1.
Cognitive effects, genetic risk factors, mechanisms, and potential biomarkers implicated in cognitive impairment associated with cancer treatment. IL=interleukin; TNF=tumor necrosis factor
Sources and selection criteria
We searched PubMed and Embase for articles published from 1980 up to 8 April 2022 by using the following search terms in the title or abstract: “chemobrain” or “chemofog” or “chemotherapy” or “immunotherapy” or “hormone therapy” or “targeted therapy” and “cognitive impairment” or “cognitive dysfunction” or “cognitive processes” or “memory problems”. This provided 1270 articles, which we filtered to exclude books, documents, case reports, commentaries, and other irrelevant articles and then screened for relevance. This left 502 articles for subsequent review and consideration. We applied limited exclusion criteria, excluding articles concerned with cancer related cognitive impairments only or regarding patients with cognitive impairment and CNS cancers. Larger sample studies using validated neuropsychological testing are highlighted in this review. We included within participant longitudinal designs that measure cognition throughout and after cancer treatment, as well as cross sectional investigations. Searches of clinicaltrials.gov found 37 ongoing clinical trials testing drug and non-drug interventions for the cognitive sequelae of systemic anticancer therapy, which we screened for relevance. Where clinical trial data were not available, we gained insights from animal studies. Moreover, we included review articles that provide more specific overviews of mechanistic subsections. We did searches for cognitive deficits associated with chemotherapy, immunotherapy, hormone therapy, and targeted therapy. In total, we included 200 references.
Epidemiology
Differences in study designs, including measures and definitions of cognitive impairment and lack of pre-treatment measurement of cognitive function, make ascertaining the incidence of cognitive impairment after cancer treatment difficult.10 Prevalence rates of self-reported cognitive complaints associated with chemotherapy are variable, ranging between 21% and 34% in patients with breast cancer.9 Despite most clinical studies investigating cognitive impairment having been conducted in chemotherapy treated breast cancer populations,11 evidence in support of chemotherapy induced cognitive impairment (CICI) spans a variety of prevalent cancers, including lung, colorectal, and gynecological cancers.121314 However, these studies are few, and a clear need exists to establish more robust evidence for CICI in cancers other than the most studied breast cancer group, especially neuroimaging studies in which sex differences can affect results.
Comparison of chemotherapy treated survivors of breast cancer 10 years post-chemotherapy treatment and healthy matched controls has shown cognitive deficits in aspects of working memory, psychomotor speed, and executive function.15 Additionally, objective neuropsychological changes in executive function, verbal memory, and processing speed and white matter microstructural alterations can be observed up to 10-20 years after chemotherapy treatment.1617 A recent review of longitudinal cohort studies in patients with breast cancer suggests that 15% of these patients improve within one month of treatment and 31% improve after one year of follow-up,18 which leaves a large proportion of patients having persistent symptoms.
Other factors that influence CICI include dosage, age, and genetic risk factors. Dosage of chemotherapy treatment is thought to influence incidence rates, with one study reporting 32% of patients treated with high doses experiencing cognitive deficits compared with 17% of those treated with the standard dose.19 Moreover, postmenopausal patients older than 60 with breast cancer showed a greater decline in processing speed compared with premenopausal patients under 52,20 suggesting that age may play a role in the severity CICI. However, memory complaints that increase alongside hormonal changes during the menopausal transition might affect cognition, which is a possible confounder.21
Clinical characterization
Neurocognitive testing
Neuropsychological tests are the de facto method of objectively identifying cognitive impairment in recipients of cancer treatment.11 Methods for recognizing cognitive impairment vary considerably between studies, which has made comparison difficult.22 However, calls to harmonize neurocognitive testing by the International Cognition and Cancer Task Force (ICCTF)23 have increased the number of studies using standardized tests. The ICCTF recommended tests include the Hopkins Verbal Learning Test-Revised (HVLT-R),24 the Controlled Oral Word Association Test,25 and the Trail Making Test.2 Simple cognitive batteries such as the Montreal Cognitive Assessment are not adequately sensitive.26 Despite efforts to standardize, a disparity remains between self-reported cognitive complaints and objective neurocognitive testing, with a weak association between them.27 This suggests that current testing is inadequate to detect early and subtle changes in cognitive function.
Objective neurocognitive data support the notion that cognitive deficits are most common in the domains of executive functioning, working memory, attentional resource mobilization, and processing speed.28 Research shows a decline in cognitive performance throughout the progression of chemotherapy for non-CNS cancers, with most taking the form of longitudinal studies.11
Longitudinal studies lasting up to 10 years suggest that in some cases cognitive impairment can persist after cancer treatment. One longitudinal self-report study comparing 356 survivors of breast cancer with 256 healthy matched controls eight years after chemotherapy showed that chemotherapy treated patients reported significant cognitive deficits compared with controls.29 Furthermore, research investigating cognitive performance in survivors of breast cancer (over age 65) 10 years after treatment found lower scores in measures of working memory, psychomotor speed, and executive function compared with controls.30 The risk factors for persistent cognitive impairment, along with age and other comorbidities, need further evaluation.
Neuroimaging
Structural imaging
Comparison of brain volumes using voxel-wise analysis of magnetic resonance imaging (MRI) images showed that cancer survivors treated with chemotherapy had reduced gray matter volumes in the prefrontal and anterior cingulate cortex, compared with survivors treated without chemotherapy.31 Furthermore, brain network alterations include impaired white matter microstructural integrity.32 Voxel based morphometry has shown lower gray matter densities associated with lower cognitive scores in chemotherapy treated patients with breast cancer.33
Diffusion tensor imaging (DTI) provides information about tissue microstructures such as the orientation of white matter fibers and tracts. Alterations to the corpus callosum have been identified, including lower fractional anisotropy scores for patients experiencing cognitive impairment with chemotherapy treatment, with slower processing speeds than controls.34 Furthermore, longitudinal DTI analysis supports the idea that cerebral white matter organizations shift over time in patients experiencing CICI, in tandem with deteriorating performance on cognitive tests.35
Functional imaging
Functional MRI data have shown specific, abnormal patterns of hypo-activation in frontal-parietal areas, consistent with the reported deficits in executive function.36 This is corroborated by longitudinal research tracking the long term effects of chemotherapy in patients with testicular cancer, suggesting that functional hyperconnectivity in the prefrontal cortex compensates for pathophysiological disruption during treatment.37
Positron emission tomography (PET) has enabled the measurement of acute alterations to cerebral blood flow and metabolic functioning. Comparison of women treated for breast cancer five to 10 years previously with and without chemotherapy suggested that modulation of cerebral blood flow in the cerebellum and frontal cortex was impaired in chemotherapy treated participants but not in untreated participants. Additionally, lower resting metabolism in participants in the chemotherapy group correlated with lower scores on short term memory tasks.38 Other PET studies have shown that patients treated with chemotherapy have frontal hypometabolism.39 More recently, 18F-DPA714-PET was used to quantify expression of translocator protein, a biomarker of microglial activation and neuroinflammation, in chemotherapy, non-chemotherapy, and healthy control groups. Patients treated with epirubicin, cyclophosphamide, and paclitaxel had higher translocator protein expression than both non-chemotherapy treated and healthy control groups, indicative of elevated neuroinflammation.40 Thus far, the number of PET studies with human participants is limited. Uniquely, PET can target specific metabolic functions, offering significant value in understanding the mechanisms of cancer treatment related cognitive impairment in the future.
Mechanisms underlying cognitive impairment after cancer treatment
Studies investigating the mechanisms by which cognitive impairment arises after cancer treatment converge on several key pathways. Each anticancer treatment and drug class has its own mechanism of action, shown in table 1, but some share common potential contributory pathways to neurotoxicity. Blood-brain barrier (BBB) dysfunction, oxidative stress, neuroinflammation, impaired neurogenesis, and disruption to neurotransmission are all hypothesized mechanisms. Figure 2 provides an overview of the mechanisms linked to chemotherapy induced cognitive impairment.
Summary of commonly studied systemic anticancer treatments and possible mechanisms underlying cognitive impairment
Putative mechanisms of chemotherapy induced cognitive impairment. Blood-brain barrier (BBB) breakdown: chemotherapeutic agents elevate peripheral pro-inflammatory cytokine and reactive oxygen species (ROS) concentrations, which disrupt tight junctions of the BBB, leading to influx of neurotoxic substances that damage brain tissues. Neuroinflammation: peripherally generated pro-inflammatory cytokines cross the BBB causing microglial activation and a self-perpetuating neuroinflammatory response leading to neuronal cell death. Oxidative stress: systemic excess of ROS can cause DNA damage, which subsequently triggers apoptotic pathways and cell death. Impaired neurogenesis: dysfunction of neuron formation in the hippocampus could arise from cytokine related neuroinflammation, ROS damaging neural precursor cells, or cytotoxic reactions from the chemotherapeutic agent. Decreased neurotransmission: single nucleotide polymorphisms in catechol methyltransferase, an enzyme that degrades catecholamines, plays a role in risk of chemotherapy induced cognitive impairment, implicating neurotransmitter concentrations. COMT=catechol-O-methyltransferase; IL=interleukin; TNF=tumor necrosis factor
Mechanisms underlying chemotherapy induced cognitive impairment
Blood-brain barrier
The BBB protects the brain by preventing the influx of harmful solutes. The BBB comprises astrocytes, pericytes, and endothelial cells forming a selective diffusion barrier with transmembrane efflux pumps, such as P-glycoprotein, disposing of unwanted compounds.57 Several contributory factors may lead to the increased permeability of the BBB during chemotherapy, which could give rise to neurotoxicity. Systemic administration of chemotherapeutic agents such as anthracyclines—for example, doxorubicin58—induce oxidative stress, which can damage BBB structures. The integrity of the BBB has been suggested to be compromised by pro-inflammatory cytokines such as tumor necrosis factor-α (TNF-α) generated by peripheral inflammation caused by chemotherapeutic agents acting on non-CNS tissues.59 Subsequent disruption of tight junction protein complexes connecting endothelial cells through damage to pericytes and astrocytes compromises the BBB, increasing permeability.60 This leads to the entry of pro-inflammatory cytokines (TNF-α, interleukin-6) and reactive oxygen species (ROS), causing microglial and astrocyte activation in a neuroinflammatory response. Subsequent DNA damage and neuronal death contribute to cognitive impairment.
Oxidative stress
Oxidative stress is characterized by an excess of ROS that are involved in apoptosis, gene expression, and cell signaling. Mitochondria and peroxisomes are essential in the regulation of the reduction-oxidation balance, and disruption caused by ROS can heighten the level of oxidative stress in the brain.61 Excessive concentrations of ROS cause subsequent degradation of subcellular structures and DNA damage and promote cell death.62 With its relatively high oxygen consumption, poor antioxidant ability, and density of polyunsaturated fatty acids, the brain is particularly vulnerable to oxidative stress.63
Several antineoplastic agents induce cellular ROS production in non-tumor tissues, including anthracyclines and platinum based agents.64 In the case of doxorubicin, the quinone component of its molecular structure can be reduced to semiquinone, generating ROS.58 Intravenous administration of cyclophosphamide and doxorubicin to rodents for three weeks caused oxidative damage to nucleic acids and RNA in the hippocampus, indicated by the presence of oxidized 8-hydroxyguanosine, a key oxidative damage marker.65 Furthermore, rats given doxorubicin showed an 80% decrease in manganese superoxide dismutase concentrations in the hippocampal homogenate, a key antioxidant cell component.66 Elevated ROS cause neuronal damage by disrupting antioxidant enzymes such as manganese superoxide dismutase, leading to mitochondrial dysfunction, inducing apoptosis via exposure to the pro-apoptotic proteins bcl-2-like protein 4 and p53,67 triggered by elevated ROS and related DNA damage. The subsequent secretion of cytochrome c triggers caspase signaling, notably apoptosis initiator caspase 9 and then caspase 3, causing apoptosis,68 with deleterious effects on cognition.
Neuroinflammation
Neuroinflammation is associated with pro-inflammatory cytokines in the CNS and activation of microglia and astrocytes when peripheral toxins cross the BBB.69 Major pro-inflammatory cytokines associated with CICI include TNFα, a potent pro-inflammatory cytokine, interleukin-6 and interleukin-1β.70 Patients with breast cancer receiving anthracycline treatment have been shown to have higher concentrations of tumor necrosis factor receptors (sTNFRI and sTNFRII) and were associated with lower cognitive scores.70 High concentrations of TNF-α and interkeukin-6 in the whole brain, hippocampus, and prefrontal cortex have been found in mice receiving a combination of docetaxel, adriamycin, and cyclophosphamide.71 Pro-inflammatory mediators generated in the periphery can enter the brain via diffusion, influx transport, or ROS induced damage to the BBB.72
Chemotherapy related neuroinflammation may arise via several possible mechanisms.11 Elevated cerebral ROS induced by antineoplastic action in the periphery may oxidize apolipoprotein-1, inhibiting its ability to mediate the transcription 3 signaling cascade, triggering macrophage produced pro-inflammatory cytokines.73 Next, peripherally available pro-inflammatory cytokines—namely, interleukin-1β and TNF-α—pass through the BBB via simple diffusion. Once they are in the brain, subsequent microglial and astrocyte activation triggers a neuroinflammatory response and generates a further excess of pro-inflammatory cytokines.73 Cytokines disrupt the cellular processes involved in proliferation and neuronal differentiation, to the detriment of neuroplasticity.74 Together, the self-perpetuating neuroinflammatory response to chemotherapeutic agents, originating in the periphery, could contribute to CICI.
Neurotransmission
Neurotransmitter dysfunction has been associated with CICI. Catechol-O-methyltransferase (COMT) plays an important role in the metabolism of catecholamines including dopamine, epinephrine, and norepinephrine.75 Dopamine neurotransmission is involved in episodic memory encoding and retrieval via nerve cell signaling, and people homozygous for the COMT Val allele have lower concentrations of cortical dopamine.76 The valine allele of COMT has been identified as a risk factor for the later development of CICI.77 Comparison of COMT (rs165599, rs737865) polymorphisms in a longitudinal cohort study of patients with breast cancer showed that patients with the COMT (rs737865) A/G and G/G genotype had higher Mini-Mental State Examination and time based prospective memory scores than did patients with the COMT A/A (rs165599) genotype.78
Neurogenesis
Chemotherapeutic agents of many types have been associated with a decline in hippocampal cell proliferation in rodents, including anthracyclines such as doxorubicin and antimetabolites such as methotrexate.7980 Furthermore, pro-inflammatory cytokines such as TNF-α, elevated during chemotherapy, are also associated with a decrease in hippocampal cell proliferation indicated by a reduction in neuro-specific nuclear antigen bromodeoxyuridine labeled cells in rodent models.81 Moreover, even at low doses, cisplatin reduced hippocampal dendritic spine density and induced neuronal apoptosis, then impairing neurogenesis in rodents.82 Adult neurogenesis in the hippocampus has been linked to learning and memory, as immature neurons are thought to possess higher synaptic and structural plasticity, which can facilitate hippocampal function.83
Putative mechanisms of endocrine therapy induced cognitive impairment
Breast cancer
Estrogen receptor positive breast cancers respond to endocrine therapies via disruption of hormone dependent tumor growth pathways.84 Endocrine therapies have been shown to increase both subjective cognitive complaints and objective cognitive testing.85 Current evidence shows that the deleterious effects of endocrine therapies may predominantly affect verbal memory and processing speed, rather than global cognitive deficits.86 Several studies suggest that verbal memory is mostly affected in endocrine therapies.87 However, one study evaluating longitudinal measurement of cognition after endocrine therapy suggested that the adverse cognitive effects may not last, with no significant cognitive deficits seen six years after treatment.88 Furthermore, menopause has been linked to an increase in cognitive complaints,89 which could be confounding the cognitive effects of endocrine therapies in some cases. Limited conclusions can be drawn regarding the effect of endocrine therapies on cognition in patients with breast cancer owing to the lack of robust studies.90 Methodological errors including small sample sizes, non-consideration of menopause as a potential confounder, and differences in cognitive testing must be corrected in future investigation.
Mechanism
Aromatase inhibitors and selective estrogen receptor modulators/agonists can readily cross the BBB,91 and estrogen positively influences biological systems beneficial to cognition.92 Estrogen receptors mediate the effects of estradiol and are associated with many signaling mechanisms that offer neuroprotection in the brain.93 For example, estrogen offers antioxidant effects that can reduce ROS induced mitochondrial damage.94 Additionally, estradiol has been understood to stimulate dendritic spine production and enhance neuroplasticity, possibly via enhancement of brain derived neurotrophic factor (BDNF) related transcription, protein synthesis, and dendritic spine plasticity, particularly in the prefrontal cortex.95 Selective estrogen receptor modulators have been suggested to impair cognition to a greater extent than non-steroidal aromatase inhibitors, as observed with tamoxifen.96 This may be explained by the differences in estrogen receptor type (ERα or ERβ) and expression in brain regions.97 Recurrent reports of primarily verbal memory deficits in patients with breast cancer receiving endocrine therapies corroborate this, as associated areas of the brain show increased expression of estrogen receptors and may be more vulnerable to disruption by endocrine therapies.87 Aromatase inhibitors directly reduce estrogen synthesis, which could reduce the associated neuroprotective effects, and selective estrogen receptor modulators may disrupt cognitively beneficial estrogen related intracellular functionality triggered via estrogen receptor binding. Figure 3 provides an overview of these possible mechanisms.
Possible mechanisms underlying cognitive impairment associated with hormone, immune, and targeted therapy. Hormone therapy: decreased estradiol concentrations may reduce N-methyl-D-aspartate receptor concentrations, down-regulate brain derived neurotrophic factor expression, and reduce the antioxidant ability of cells in the brain to counteract elevated reactive oxygen species concentrations. Reduction in testosterone may cause similar effects to antioxidants and brain derived neurotrophic factors, with impairments to neprilysin that may impair the clearance of amyloid β. Immunotherapy: Chimeric antigen receptor (CAR)-19 T cell binding with cell surface antigens on target molecules can lead to cytokine release syndrome, in which concentrations of pro-inflammatory cytokines such as interleukin-6 (IL-6), tumor necrosis factor- α (TNF-α), and interferon‐γ (IFN-γ) are elevated. These cytokines can then cross the blood-brain barrier, activating microglia and causing a neuroinflammatory response leading to neuronal cell death. Targeted therapy: anti-angiogenic tyrosine kinase inhibitors disrupt vascular endothelial growth factors, which may reduce cerebral blood flow and impair neurogenesis. Proteasome inhibitors may damage mitochondria and elevate pro-inflammatory cytokine concentrations, giving rise to neuroinflammation and leading to cell death. VEGF=vascular endothelial growth factor; VEGFR=vascular endothelial growth factor receptor
Prostate cancer
Androgen deprivation therapy aims to lower testosterone concentrations from the testes and adrenal glands or reduce the effectiveness of androgen to limit tumor growth in prostate cancer.98 Lowering testosterone production can be achieved via localized administration of luteinizing hormone releasing hormone agonists or antagonists.99 Alternatively, blocking enzymes involved in the biosynthesis of testosterone, such as CYP17 (17α-hydroxylase/17,20-lyase) can also diminish cancer cell growth mediated androgen receptor binding.100
Recent systematic reviews of studies evaluating cognitive impairment in men treated with androgen deprivation therapy for prostate cancer report mixed results.101102 Over the duration of three months, patients treated with enzalutamide showed a significant decline in Functional Assessment of Cancer Therapy-Cognitive Function (FACT-Cog) self-report scores compared with those receiving abiraterone acetate plus prednisone.103 However, one prospective study found no cognitive deficits across any domains between patients treated with androgen deprivation therapy compared with the no treatment group.104 Systematic review has shown that many randomized trials investigating the relation between androgen deprivation therapy and investigator assessed or self-reported cognition often fail to use adequate cognitive evaluation tools,105 which must be corrected to accurately quantify incidence rates.
Mechanism
The absence of male androgens can have a deleterious effect on cognition through several mechanisms.106 The accumulation of amyloid β is central to Alzheimer’s disease pathology; this amyloid β accumulation is associated with neuroinflammation via microglial activation and neuronal loss via phagocytosis.107 Testosterone is understood to play a role in the production of the amyloid β catabolizing enzyme neprilysin, which aids in the clearance of amyloid β.108 Animal studies have shown that dihydrotestosterone increased neprilysin expression, which causes reduced amyloid β aggregation in hippocampal neurons, but these effects were not observed in rats without androgen receptor expression.109 Thus, impaired amyloid β clearance could underlie the observed cognitive impairment. Testosterone may be associated with the transcription of antioxidant genes, increasing superoxide dismutase which can counteract excessive ROS and prevent mitochondrial damage and subsequent cell death.110 Testosterone has also been suggested to influence BDNF concentrations, as observed in rodent models in which testosterone reduction as a result of gonadectomy reduced hippocampal BDNF.111 Conversely, testosterone is an estradiol precursor, and a reduction in systemic concentrations of testosterone could decrease estradiol synthesis in the male brain. Hence, the deleterious effects on cognition could be attributed to the mechanisms of the reduction in the neuroprotective effects associated with estradiol, as outlined in figure 3.
Putative mechanisms of immunotherapy induced cognitive impairment
Adverse events can occur during immunotherapy treatment, including neurologic complications such as encephalitis.112 Several main classes of immunotherapies exist, including immune checkpoint inhibitors and CAR T cell therapies.113 The incidence rate of neurologic toxicities associated with immunotherapy has been reported to be 12% in patients receiving both anti-PD-1 and anti-CTLA4 drugs, and several adverse neurologic events have been reported during immunotherapy treatment, including polyneuropathy and demyelination.114 Most research into the association between immunotherapy and cognition concerns immune checkpoint inhibitors and CAR T cell therapies, although conclusions are limited.49
Immune checkpoint inhibitors
Immune checkpoint inhibitors influence the binding of checkpoint proteins such as programmed cell death protein 1 (PD-1) and programmed death ligand 1 (PD-L1), which mediate tumor induced cell death, and blockage of these proteins enables T cells to kill target tumor cells via the release of effecter cytokines.115 Neurotoxicity is a well established side effect of immune checkpoint inhibitor treatment affecting cognition. A cross sectional study investigated patients with immune checkpoint inhibitor induced neurotoxicities and identified a wide variety of complications including demyelinating disorder (28%) and gray matter autoimmune encephalitis (17%), with the onset of neurotoxicity averaging at five weeks.116 Recent results from a single center longitudinal study observing 25 patients treated with pembrolizumab concluded that 32% of patients had neurocognitive impairment related to treatment.117 Similar cross sectional research observed functional deficits in working memory and verbal memory in up to 41% of patients treated with ipilimumab.118 Again, studies are limited by small sample sizes.
Mechanism
T cell overactivation arising from CNS-borne autoantigen reactions has been hypothesized to underly cognitive impairment after immune checkpoint inhibitor treatment.119 T cells play a role in anti-N-methyl D-aspartate receptor encephalitis, especially under conditions of elevated pro-inflammatory cytokines.120 Amplification of an ongoing autoimmune response is also posited as a mechanism, as patients with multiple sclerosis treated with checkpoint inhibitors can have severe adverse neurologic outcomes.121 Furthermore, cells with the often targeted PD-L1 receptor exist in the CNS, such as neurons, leading to immune reactions with off-target cells and subsequent CNS tissue damage.122
CAR T cell therapy
CAR T cell therapy is an emerging therapy used primarily for the treatment of hematologic malignancies. In CAR T cell therapies, T cells can be genetically altered to express chimeric antigen receptors that are engineered to target antigens on tumorous cells—for example, B lymphocyte antigen CD19 in the case of lymphoma.123
Evidence
Studies evaluating cognitive changes associated with side effects of CAR T cell immunotherapy are few, but some indicatory evidence exists. One multicenter cohort study showed that 43% of 84 patients treated for lymphoma with CD19 targeted CAR T cell therapy had neurotoxicity, with the most frequent neurologic sign being severe cognitive impairment (36%).124 Longitudinal study of 40 recipients of CAR T cell treatment observing cognitive changes over three years suggested that 37% of patients reported cognitive deficits, especially in domains such as memory (35%), attention (22.5%), and executive functioning (12.5%).125 Recent one year longitudinal observation of cognition in patients being treated with CAR T cell therapy for non-Hodgkin’s lymphoma showed that cognitive scores had declined at day 90 of treatment but recovered from day 90 to 360.126 Future investigation into how cognitive impairment unfolds over time and the primary aspects of cognition affected will further inform potential treatment strategies.
Mechanism
CD19 T cell activation from reactions with target cells can increase concentrations of pro-inflammatory cytokines, including interferon-γ, TNF-α, and interleukin-6, in the periphery, known as cytokine release syndrome.127 These cytokines may cross the BBB, activating microglia in a neuroinflammatory response and triggering apoptotic pathways that may ultimately cause neuronal cell death. Moreover, increased BBB permeability owing to endothelial activation by pro-inflammatory cytokines could lead to more severe neurotoxicity.128Figure 3 gives a depiction of this possible mechanism.
Putative mechanisms of targeted therapy induced cognitive impairment
Targeted therapy aims to prevent the spread and growth of cancer cells via alterations to molecularly specific tumor associated targets.129 Targets can include the tumor cells themselves, through the triggering of extrinsic and intrinsic apoptotic regulatory pathways, such as the inhibition of Bcl-2 proteins, which can trigger apoptosis in chronic lymphocytic leukemia.130 Alternatively, molecular modulation of the tumor microenvironment, such as activation of anti-tumor immune responses and hypoxia, can also restrict cancer cell proliferation and migration.131
Antiangiogenic therapies
Antiangiogenic tyrosine kinase inhibitors (TKIs) have been subject to some scrutiny in terms of their effect on cognition. One prospective study found that 31% of patients with metastatic renal cancer treated with antiangiogenic vascular endothelial growth factor (VEGF) inhibitors showed a cognitive decline between the three month and six month time points, specifically in domains of information processing speed and working memory, independent of fatigue.53 Furthermore, a clinical cross sectional study of sunitinib and sorafenib treatment showed that learning, memory, and executive functioning were specific areas of impairment.54 TKIs have also been associated with cognitive complaints when treating chronic myeloid leukemia in the chronic phase through a single center cohort study.132
Mechanism
Antiangiogenic TKIs can target VEGF, implicated in the formation of new blood vessels, facilitating tumor growth. VEGF has been associated with neuroprotective effects such as neuronal protection, enhanced neurogenesis, and cerebral blood flow.133 VEGF is also implicated in CNS health to ensure adequate oxygen supply, promoting neurogenesis and synaptic plasticity, but the relation between VEGF and cognition needs further elucidation.134 As many VEGF inhibitors cannot cross the BBB owing to their pharmacokinetic profile or susceptibility to efflux pumps, TKIs targeting VEGF are thought to cause cognitive impairment through inhibition of peripheral VEGF.134 Animal models investigating the cognitive effects of the TKI sunitinib on cognition suggest that TKIs may induce cognitive impairment via the dysregulation and suppression of VEGF, VEGF receptor, and nuclear factor kappa B (NFκB) signaling involved in apoptotic pathways.135 Ultimately, disruption of the maintenance of normal cerebral microvasculature could disrupt cognition, as shown in figure 3.
Proteasome inhibitors
In the limited research into the effects of proteasome inhibitors on cognition, few have been associated with impairment. Exploration of the effects of bortezomib on cognition found that it was not associated with BBB penetration or cognitive deficits at therapeutic doses in mouse and human models.136 However, cognitive dysfunction was listed as a dose limiting side effect as part of the phase 1 clinical trial for the investigation of marizomib but has not since been subject to rigorous investigation.137 Evaluation of effects on quality of life, including cognition, of proteasome inhibitors in patients with multiple myeloma suggests that they are not strongly associated with worsening cognitive impairment in the case of ixazomib.138
Mechanism
Proteasome inhibitors have been suggested to damage mitochondria via disruption of the cellular microenvironment, increasing oxidative stress, which can lead to cell death.139 Furthermore, proteasome inhibitors have been associated with inhibition of NFκB activity,140 which could cause a decrease in BDNF via connected pathways, stunting neurogenesis and limiting support of existing neurons. Lastly, the aggresomal formation also increases with proteasome inhibition in healthy cells—for example, neurons in the case of proteasomal inhibitor lactacystin,141 which can lead to stimulation of pro-apoptotic signaling pathways and neuronal cell death, as shown in figure 3.
Emerging biomarkers
Plasma and cerebrospinal fluid
Plasma and cerebrospinal fluid based markers of neuroinflammation provide the most practical utility for determining risk of developing CICI. Pro-inflammatory cytokine concentrations, notably interkeukin-6, TNF-α, and interleukin-1β, are temporally linked to onset of cognitive impairment.142 One multicenter prospective cohort study suggested that increased interleukin-6 and interleukin-1β concentrations correlated with severity of cognitive impairment in patients treated with chemotherapy.143 Promising findings have shown that plasma concentrations of neurofilament light chain and translocator protein were significantly elevated in patients with breast cancer who received chemotherapy compared with chemotherapy-naive patients and controls, showing greater axonal damage and glial cell activation respectively.40 These results suggest that biomarkers of neuroinflammation may serve as useful measures indicative of the cognitive adverse effects of anticancer treatment.
Neuroimaging
Functional and structural changes to the brain have been seen in neuroimaging studies investigating cognitive impairment after cancer treatment, which could be used to characterize CICI. Glucose hypometabolism has been shown in the cerebral cortex in patients with CICI.144 In terms of structural changes, decreases in gray matter volume have been identified as recurrent markers of chemotherapy related cognitive impairment.16 Meta-analysis of functional MRI studies has suggested that hypofrontal-parietal activity in different cognitive domains can be used to differentiate patients treated with chemotherapy from controls.36 These patterns of abnormalities could be further refined and validated to establish objective neuroimaging thresholds for diagnosis.
Risk factors
Genetic risk factors for CICI have been explored on the basis of their role in other neurodegenerative diseases. Apolipoprotein E4 is the strongest and most well established genetic risk factor for sporadic Alzheimer’s disease connected with CICI.145 The limited clinical studies investigating the role of the apolipoprotein E4 genotype in CICI converge on the notion that carriers of apolipoprotein E4 have greater cognitive deficits, particularly in working memory and processing speeds,146 compared with patients negative for the apolipoprotein E4 allele. Mechanisms underlying this could include compromised BBB integrity and poor ability to detoxify ROS, both associated with apolipoprotein E4 carriers and connected with CICI mechanisms.147
COMT plays a role in the enzymatic breakdown of catecholamines such as dopamine, epinephrine, and norepinephrine.148 Expression of the Val158Met COMT polymorphism is understood to decrease the availability of dopamine and detrimentally affect neurotransmission and, consequently, aspects of cognition such as executive functioning.75 In the context of CICI, the COMT-Val allele has been associated with poorer scores in attention, verbal fluency, and motor speed in patients with cancer,149 and is therefore a possible risk factor.
BDNF plays a role in the formation of dendric spines and the growth of new neurons and synapses.150 The Val/Val BDNF genotype has been associated with a higher likelihood of experiencing cognitive deficits compared with the Met/Met homozygote.151 Furthermore, longitudinal examination of plasma BDNF concentrations in patients with early stage breast cancer showed a significant change after chemotherapy treatment, except for Met/Met homozygous carriers of the BDNF genotype.152 Together, investigations suggest that BDNF Val/Val homozygous carriers may be more susceptible to CICI.
Management strategies
Drug treatment
The putative pathological mechanisms underlying CICI provide a diverse group of potential pharmacologic targets. Current research primarily focuses on CICI with repurposed CNS stimulants and drugs for dementia and memory impairment being the most studied.153 As randomized controlled trials (RCTs) in humans are sparse in the field, insights are gained from animal data when human trials do not exist.
Oxidative stress
Antioxidative compounds administered exogenously can protect against oxidative stress by accepting an electron donation or inhibiting the activity of the free radical generating enzymes nicotinamide adenine dinucleotide phosphate oxidase or myeloperoxidase.154 Predominantly animal studies exist exploring the effects of antioxidant interventions.
Antioxidants may be able to improve the cognitive effects of chemotherapy without compromising the efficacy of cancer treatment. Zinc sulfate prevented short term memory deficits in rats treated with the cytotoxic agent carmustine compared with controls.155 Interestingly, a study reported that mulmina mango ameliorated cognitive impairments and aided the maintenance of regular cytokine concentrations in rodents, with tumor size considerably reduced during cyclophosphamide, methotrexate, and 5-fluorouracil treatment.156 Phytochemical and antioxidant alkaloid piperlongumine prevented measurable social memory deficits in mice treated with a doxorubicin, cyclophosphamide, and docetaxel cancer regimen.157 However, an RCT (n=166) showed that ginkgo biloba did not prevent chemotherapy related cognitive impairment in patients with breast cancer treated with anthracycline and taxane based chemotherapy compared with placebo, as measured by the High Sensitivity Cognitive Screen (HSCS) (mean change 2.7; P=0.84).158 However, consideration of the paradoxical relation between pro-tumor and anti-tumor ROS activities is important so as not to undermine the efficacy of cancer treatment.
Neuroinflammation
Modulating microglia mediated pro-inflammatory cytokine release could be a potential treatment strategy for the prevention of cognitive impairment arising from cancer treatment. Rodent models constructed to explore well documented anti-inflammatory drugs suggest ineffectiveness in improving chemotherapy related memory deficits in the case of both aspirin and naproxen.159160 Naturally derived compounds have been associated with suppressed pro-inflammatory cytokines TNF-α and interleukin-6 and elevated anti-inflammatory cytokines interleukin-4 and interleukin-10 in several brain tissues in cases of CICI in rodent models, including ginsenoside rg1,161 c-phycocyanin,162 and reservatrol.163 Mechanistically, these compounds have been hypothesized to modulate the cytokine regulating NFκB pathway, known to rapidly regulate harmful stimuli including ROS, TNF-α, and interleukin-1β, ameliorating neuroinflammation.164 As for other emergent treatments for CICI, translational randomized clinical trials are needed.
Neurogenesis
Disruption of the formation of hippocampal neurons from neural stem and progenitor cells has been suggested to contribute to chemotherapy induced cognitive impairment.165 Fluoxetine is a commonly prescribed antidepressant, elevating serotonin via blockage of the serotonin reuptake pump located at the neural membrane.166 Cellular responses to fluoxetine also include the stimulation of hippocampal neurogenesis, increase in BDNF, and improvements in neuronal survival rates, sparking interest in its ability to treat CICI.167 In vitro research has shown that fluoxetine may increase neurogenesis via the up-regulation of the phosphorylation of Ser9 of glycogen synthase kinase-3-β and increasing β-catenin in the nucleus, increasing proliferation of neural progenitor cells.168 Evidence from rodent models supports the notion that concurrent or previous administration of fluoxetine in chemotherapy prevents spatial memory deficits associated with 5-fluorouracil treatment.169 Despite promising findings, clinical trials are needed to show efficacy of fluoxetine in humans.
Other potential neurogenesis targeted treatments include metformin, an anti-hyperglycemic first line therapy for type 2 diabetes mellitus.170 Metformin may promote neurogenesis by activating a protein kinase CPB transcriptional coactivator pathway involved with the typical genesis of neurons from neural progenitor cells.171 Rodent models investigating metformin as a treatment for cancer treatment related cognitive impairment have yielded conflicting results. Metformin effectively ameliorated cognitive deficits associated with doxorubicin in some rodent models,172 but in other cases it did not.173 However, differences could be attributed to differences in dosage between studies. Conversely, metformin administered in combination with standard cyclophosphamide, methotrexate, and 5-fluorouracil cancer regimens was shown to potentially induce an inflammatory response marked by an increase in pro-inflammatory cytokines—namely, interkeukin-1α.174 A review suggests that metformin may be effective only in ameliorating cognitive impairments associated with cisplatin and cyclophosphamide and not doxorubicin and may even negatively affect survival rates overall.170 Despite possible enhancements to neurogenesis, more clarity is needed on the relation between metformin, its influence on chemotherapy efficacy, and neuroinflammation, which may counteract benefits altogether.
Neurotransmitter
Donepezil is among the few treatments for Alzheimer’s disease related cognitive impairment and acts via reversibly inactivating cholinesterases, decreasing acetylcholine hydrolysis, which increases the availability of acetylcholine at cholinergic synapses.175 Rodent studies support a role for donepezil in the amelioration of chemotherapy related learning and memory deficits when co-administered with doxorubicin, achieved by the attenuation of oxidative stress and neuroinflammation and the enhancement of neurogenesis.176 Moreover, a small randomized controlled trial (n=47) showed that cognitive deficits after systemic chemotherapy for non-CNS cancer were partially restored by donepezil, with the treatment group showing significantly enhanced performance compared with controls on the HVLT-R Total Recall, with a mean difference 2.78 (95% confidence interval 0.23 to 5.34) and HVLT-R Discrimination, with a mean difference 0.89 (0.06 to 1.71) even one to five years after chemotherapy.177 Donepezil preserves cholinergic neurons in the basal forebrain in patients with Alzheimer’s disease,178 and it could be improving cognitive impairment symptoms in this way.
Methylphenidate is a CNS stimulant that blocks the reuptake of norepinephrine and dopamine, enhancing dopaminergic and noradrenergic activity in the prefrontal cortex that is used for treating attention deficit/hyperactivity disorder.179 An RCT (n=57) found no significant difference in cognitive scores as measured by HSCS or HVLT-R between the methylphenidate treated group and controls.180 However, one other small sample (n=33) RCT showed a slight improvement only in attention.181 The catecholaminergic drug modafinil is also a dopamine reuptake inhibitor that can improve cognition via enhancement of dopaminergic transmission.182 One RCT with 82 patients with breast cancer receiving chemotherapy showed enhancement in episodic memory (P=0.02), memory speed (P=0.03), and attention (P=0.01) after modafinil treatment.183 Furthermore, the antioxidant drug 2-mercaptoethanol sulfonate sodium prevented cognitive dysfunction and doxorubicin associated decreases in choline-containing compounds in a mouse model via preservation of phospholipase activity, important for choline for acetylcholine generation through the enzymatic breakdown of phosphatidylcholine.184 Future elucidation of the role of neurotransmitters in CICI will support development of neurotransmitter directed treatment.
Non-drug treatment
Non-drug therapeutic approaches to management of CICI are mostly centered around either cognitive training or physical activity.185 One RCT (n=167) tested the efficacy of a computer assisted cognitive rehabilitation intervention for cancer patients expressing cognitive complaints. Patients receiving rehabilitation had a significant improvement in working memory as measured by the digit span backward Wechsler Adult Intelligence Scale, fourth edition (P=0.03) compared with a control group.186 Another RCT of 50 survivors of childhood cancer assessed a physical exercise intervention designed to increase moderate to vigorous physical activity. It found improved neurocognitive outcomes as measured by Wechsler Abbreviated Scale of Intelligence full scale IQ z-score (mean change 0.13; P=0.04) and Delis-Kaplan Executive Function System, Color-Word Interference Test (mean change 0.23; P=0.007) in the intervention group compared with controls.187 Although both cognitive training and exercise show promise as accessible, low cost methods of mitigating symptoms of CICI, more evidence of efficacy is needed. Moreover, limitations of current research include methodological heterogeneity, differences in intervention programs, and variability in adherence rates.
Emerging treatments
Many drug interventions under investigation have been selected for their ability to target potential mechanistic pathways of CICI. Antioxidant N-acetylcysteine is undergoing testing for safety and tolerability in patients with ovarian cancer receiving platinum based chemotherapeutic agents (clinicaltrials.gov, NCT04520139). Additionally, ginkgo biloba, which has shown potential clinical efficacy for Alzheimer’s disease, is also being tested for antioxidative and anti-inflammatory properties (NCT00046891). Furthermore, mesna may reduce plasma protein oxidation that occurs in the blood of doxorubicin and cyclophosphamide treated patients causing release of TNF-α, which has shown promise in rodent models and is now being tested in humans (NCT01205503). Other treatments under investigation target neuronal protection, differentiation, and neurogenesis, such as ganglioside-monosialic acid (NCT05239663), present on the outer layer of plasma membranes and involved with injury repair, and transdermal nicotine (NCT02312934), which could interact with nerve cells and improve important learning and memory functions.
Brain training through the practice of cognitive tasks, physical brain stimulation using techniques such as transcranial magnetic stimulation, and exercise are among the non-drug therapeutic avenues under investigation. Brain training aims to improve cognition by augmenting neuroplasticity. Training programs being tested to treat CICI in RCTs include computerized cognitive remediation (NCT05283629, NCT04230863), cognitive behavioral therapy based memory and attention adaptation training (NCT04586530), and other brain fitness programs that aim to improve the processing speed of auditory information, attention (NCT02515487), memory, and cognition (NCT00387062). In terms of brain stimulation, transcranial magnetic stimulation and neuromodulation in the form of intermittent theta-burst transcranial stimulation to frontal regions of interest including the prefrontal cortex are also being tested to see whether they can induce measurable changes in patients with CICI (NCT04966520). Six months of regular exercise (NCT00495703) and intermittent bouts of high intensity exercise (NCT04789187) are being tested to see whether this will improve cognition scores, reduce detrimental changes in MRI volumetry, and decrease inflammatory biomarkers such as TNF-α, interleukin-6, interleukin-1α, and interleukin-1β in patients undergoing chemotherapy treatment.
Guidelines
Virtually no clinical guidelines exist for the identification and management of CICI. The need for formalized care standards and guidelines for practice has been raised on several occasions but has not yet come to fruition.188 This could be attributed to the lack of effective interventions and imprecise diagnostic criteria, but data exist that could be used to create guidelines. The applicability of the National Institute on Aging/Alzheimer’s Association criteria for mild cognitive impairment has been successfully shown in the context of chemotherapy treated patients with breast cancer.189 This provides a good basis to formulate fundamental guidelines, which could then be refined over time. The ICCTF was established in 2006 to encourage international collaboration and promote research into cancer and cognition.190 Members have called for standardization of neuropsychological evaluation of patients treated for cancer, along with neuroimaging methods.191 Adherence to these standards has increased the number of studies with standardized neuropsychological testing, accelerating clinical characterization and diagnostic thresholds. However, the onus falls on patients to self-report problems before a specialist referral is made. The body of research to date is enough to establish preliminary guidelines for clinicians, to improve support for patients, simplify diagnosis, and encourage dedicated service provision.
To close this gap, we suggest drawing from guidelines proposed by the American Academy of Neurology to identify and manage mild cognitive impairment.192 Firstly, as part of the process of beginning cancer treatment, administration of a battery of validated and objective cognitive tests, such as Addenbrooke's Cognitive Examination-III,193 HVLT-R,24 Trail Making Test,194 and Controlled Oral Word Association,25 should be conducted before treatment as a baseline value. Furthermore, quality of life assessments should be made, such as the 36-Item Short-Form Survey,195 with depression and anxiety screened using the Hospital Anxiety and Depression Scale.196 In support of accessibility, convenience, and accuracy, and to overcome practical constraints such as the lack of specialist neuropsychologists, cognitive tests could be administered online. Subsequently, clinicians should administer the same tests at regular intervals to monitor cognition and test against baseline values to determine cognitive impairment and make a diagnosis based on National Institute on Aging/Alzheimer’s Association criteria for mild cognitive impairment, where possible.189 The ICCTF suggests a diagnostic threshold of two tests at or below −1.5 standard deviations or a single test score below −2.0 standard deviations from the normative mean for the recommended neurocognitive tests.23 From here, further investigation can take place along with support offerings, management strategies, and ongoing monitoring as outlined in box 1.
Sequential overview of suggested guidelines for management of cognitive impairment associated with cancer treatment
Initial evaluation
Before administration of cancer treatment, patient should undergo objective cognitive evaluation using validated structured cognitive instruments (examples include Addenbrooke's Cognitive Examination-III, Hopkins Verbal Learning Test-Revised, Trail Making Test, Controlled Oral Word Association)
This will establish a baseline performance of the patient, which will aid in quantifying the extent of cognitive impairment
Patients should also undergo quality of life assessment (example: 36-Item Short-Form Survey) and depression and anxiety screening (example: Hospital Anxiety and Depression Scale)
Patients with cognitive complaints can be directed to local resources (UK examples: Cancer Research UK197 or Macmillan198), which summarize typical presentations and coping strategies for cognitive changes in relation to cancer treatment
Follow-up
Patients should have check-up appointments for cognitive evaluation in accordance with chemotherapy treatment program (3, 6, 9, 12 month intervals) and scores compared with baseline values
Diagnostic threshold recommended as −1.5 standard deviations from normative mean (as per ICCTF guidance) with aid from mild cognitive impairment criteria—ie, evidence of cognitive complaints, objective cognitive impairment from neurocognitive evaluation, and some deficits in daily tasks
If no differences are found in cognitive testing, continue monitoring patient at regular intervals
Investigation
Patients with objective cognitive deficits should be referred for comprehensive neurocognitive assessment by a specialist
In more severe cases, a case may exist for specialist referral and use of more in-depth investigative tools such as:
Neuroimaging
Cerebrospinal fluid/plasma biomarkers
Planning future care
The clinical team should assess patient’s quality of life assessment to clarify areas of life that are most affected by impaired cognition and use this to:
Inform family/friends/carer about areas in which patient may need support to carry out normal aspects of life
Give advice and information on self-management
Management
No treatments are approved for cognitive impairment associated with cancer therapy
Assess whether exercise or brain training programs may be useful for patient
Follow-up and monitoring
Open discussions should be held between patients, primary carers, and clinicians about how best to monitor cognition and whether this can be integrated into cancer treatment program
Neurocognitive testing: Addenbrooke’s Cognitive Examination-III,193 Hopkins Verbal Learning Test-Revised,24 Trail Making Test,194 Controlled Oral Word Association.25 Quality of life: 36-Item Short-Form Survey.195 Depression and anxiety screening: Hospital Anxiety and Depression Scale.196 Diagnostic threshold: International Cognition and Cancer Task Force (ICCTF),23 mild cognitive impairment criteria: International Working Group on Mild Cognitive Impairment199
Conclusion
As the cancer landscape changes to place more emphasis on survival and quality of life, cognitive dysfunction associated with cancer treatment is arguably the most debilitating side effect of all. The epidemiologic picture is becoming clearer, with high self-reported rates being validated with objective neuropsychological testing, but data are needed for different types of cancer and therapies. This emphasizes the need to shift objective neurocognitive assessment from research to the clinic. Moreover, the disparity and the weak association between self-reported rates and objective neurocognitive testing rates raise the question of whether the tests used are of adequate sensitivity. However, tools to investigate mild cognitive impairment, neuroimaging techniques such as PET, and biomarkers such as neurofilament light chain and translocator protein imaging could help to detect cognitive dysfunction earlier and with higher sensitivity and objectivity. The mechanisms by which each anticancer therapy could be causing cognitive impairment are diverse and heterogeneous, but some overlaps between therapies exist that could act as targets for biomarker development or drug interventions. The next five years hold promise for the development of new and repurposed drug interventions that may have a beneficial effect on these patients. The role of exercise and cognitive training is yet to be determined. Looking to the future, clinicians should routinely integrate objective neurocognitive testing before and during cancer treatments. Translational research should aim to clarify how treatments other than chemotherapy affect cognition, and clinical trials should focus on novel therapeutic strategies.
Glossary of abbreviations
BBB—blood-brain barrier
BDNF—brain derived neurotrophic factor
CAR—chimeric antigen receptor
CICI—chemotherapy induced cognitive impairment
CNS—central nervous system
COMT—catechol-O-methyltransferase
DTI—diffusion tensor imaging
HSCS—High Sensitivity Cognitive Screen
HVLT-R—Hopkins Verbal Learning Test-Revised
ICCTF—International Cognition and Cancer Task Force
MRI—magnetic resonance imaging
NFκB—nuclear factor kappa B
PD-1—programmed cell death protein 1
PD-L1—programmed death ligand 1
PET—positron emission tomography
RCT—randomized controlled trial
ROS—reactive oxygen species
TKI—tyrosine kinase inhibitor
TNF-α—tumor necrosis factor-α
VEGF—vascular endothelial growth factor
Questions for future research
What are the prevalence rates for cognitive impairment in types of cancer other than breast cancer and therapies other than chemotherapy?
Does a characteristic pattern of cognitive domains and brain regions affected exist? What can this tell us about the mechanistic underpinnings?
Of the emergent biomarkers, which best predicts the onset of cognitive impairment during cancer treatment?
Can drug interventions used in other neurologic conditions affecting cognition be repurposed to treat cognitive impairment from cancer therapy?
What is the best way to integrate neurocognitive testing into clinics, and how can this be incorporated into overall cancer treatment plans?
How patients were involved in the creation of this article
We contacted people with lived experience of cancer treatment related side effects to review initial drafts of the article by using The BMJ’s patient reviewer questions. Key points raised included the need for more detail on guidelines with suggestions, including patient support as part of care, and more emphasis on improving quality of life during treatment. These points informed changes to the guidelines section
Footnotes
Series explanation: State of the Art Reviews are commissioned on the basis of their relevance to academics and specialists in the US and internationally. For this reason they are written predominantly by US authors
Contributors: BF did the initial literature search. BF was the first author of the manuscript and created the first draft and revised it. PE and LK conceptualized and designed the work. LK and PE reviewed and revised the manuscript. PE and LK are the guarantors.
Competing interests: We have read and understood the BMJ policy on declaration of interests and declare the following interests: PE was funded by the Medical Research Council and is now funded by Higher Education Funding Council for England (HEFCE); he has also received grants from Alzheimer’s Research UK, Alzheimer’s Drug Discovery Foundation, Alzheimer’s Society UK, Medical Research Council, Alzheimer’s Association US, Van-Geest foundation, and the European Union; he is a consultant to Roche, Pfizer, and Novo Nordisk and has received educational and research grants from GE Healthcare, Novo Nordisk, Piramal Life Science/Life Molecular Imaging, Avid Radiopharmaceuticals, and Eli Lilly; he was a member of the Scientific Advisory Board at Novo Nordisk and is on the scientific advisory board of CytoDyn; LK was a member of the scientific advisory board of General Electric Healthcare and is a consultant to General Electric Healthcare.
Provenance and peer review: Commissioned; externally peer reviewed.