The PRISMA flow diagram (Figure X) shows the number of manuscripts generated by the initial search along with a description of exclusion criteria that were applied to achieve the resulting 41 articles ultimately included in the review. Originally 16,757 publications were uploaded to RAYYAN and cross-referenced with the PubMed database. Then, 4,278 duplicates were identified and removed. Using the defined exclusion criteria, the remaining 12,479 publications were reviewed by the authors, leaving 295 publications. From these remaining 295 publications, an additional 255 were excluded due to lack of reportable data—such as an abstract without the full study, fewer than 10 participants, or undefined CICI assessment tools. The final number of included papers for this systematic literature review is 40 publications. Table One included in this paper gives an in-depth summary of the type of non-modifiable factor studied, type of CICI and tool assessments used, the sample size, type of study, type of cancer and results of the study.
The findings indicate that most papers included in the review had a low risk of bias except when random sequence generation was not part of the study design. Selection bias is present due to the non-modifiable characteristics of the populations, such as type of cancer or demographics. Furthermore, the allocation concealment (selection bias) was marked as “unsure” following the yes and no answer scheme provided to the RoB2 tool.
Table 1
Table indicates the differences between the included studies and why meta-analysis could not be conducted
| NFS | CICI Assessment | Tool Assessment | Type of Cancer /Type of treatment | Number of Participants | Study Design | Results |
Apple et al.; 2017 | Brain Structure | Neuro-quality of life for self-reported CICI and National Institute of Health (NIH) Toolbox | 3T MRI | Pre-menopausal breast cancer survivors (BCS); 18 months post chemotherapy | 16 BCSs and 18 healthy controls | Cross-sectional study | Concluded that hippocampal deformity is seen with CICI in BCSs. |
Apple et al.; 2018 | Brain Structure | NIH Toolbox Cognition Battery Test | MRI; 3T TIM Trio Scanner with a 32-channel head coil | Pre-menopausal BCSs; 18 months post chemotherapy | 16 BCSs and 18 healthy controls | Cross-sectional study | Concluded that higher connectivity of the precuneus with hippocampus was significantly associated with worse cognitive concern in survivors. |
Arsdale et al.; 2016 | Type of cancer- Gynecologic | Montreal Cognitive Assessment (MoCA) | Patient history | Gynecological | 165 cancer participants | Cross- sectional study | Conclude that there is a high prevalence of CICI but more research needs to be done. |
Blommaert et al.; 2019 | Brain Structure | Cognitive Failure Questionnaire and Dutch Adult Reading Test. | MRI T1 | BCSs treated with chemotherapy | 72 BCSs treated w/chemotherapy, 49 BCSs w/out chemotherapy, and 59 healthy controls | Longitudinal Study | Concluded that white matter expansion after chemotherapy is more pronounced in premenopausal patients and they showed less cognitive decline. This suggests that CICI may also have an age component. |
Carreira et al.; 2020 | Demographic | Quality of Life in Adult Cancer Survivors scale | Electronic health record, Index of Multiple Deprivation | BCSs who were 8 years post diagnosis | 356 BCSs and 252 healthy controls | Cross-sectional study | Concluded that socio-demographic variables data were tested and collected but the results were not included because of lack of statistical significance. |
Carroll et al.;2019 | Inflammation | Functional Assessment of Cancer Therapy-Cognitive (FACT-Cog) v3 | Telomere repeat amplification protocol | Early-stage breast cancer | 94 cancer participants | Cross-sectional study | Concluded that higher rates of DNA damage and lower telomerase activity were statistically significant to increases in CICI. |
Chan et al.; 2019 | Genetics | Fact-Cog v3 | 10 mL blood into an EDTA tube and QIAamp DNA Blood Mini Kit (Qiagen) | Early-stage breast cancer patients (BCP) on chemotherapy | 351 participants | Prospective Longitudinal Study | Concluded that DNMT1 rs2162560 polymorphism in cancer patients is indicative of CICI. |
Chen et al.;2018 | Brain Structure | NIH Toolbox for Cognition | MRI | Breast cancer receiving adjuvant chemotherapy | 16 breast cancer and 14 age/sex matched healthy controls | Pilot Longitudinal study | Concluded that there was difference in brain volumes between both groups, especially in the temporal lobe. Could be contributing to CICI. |
Cheng et al.;2013 | Type of Cancer-Breast | Mini-Mental State Examination (MMSE), Event Based Prospective Memory (EBPM) and Time-Based Prospective Memory (TBPM) | Patient history | Breast cancer | 40 cancer participants and 40 healthy matched controls | Cross sectional study | Conclude that overall lower scores for BCPs, with significant differences in scores for the mini-mental state examination, verbal fluency test, and EBPM. |
Dos Santos et al.;2021 | Demographic | MMSE | Recorded at inclusion stage of study when screening participants. | > 70 years older who have colon, pancreas, stomach, ovary, bladder, prostate, lung, non-Hodgkin’s lymphoma | 310 cancer participants | Prospective study | Concluded that older cancer patients (≥ 70 yrs) and concluded that they were also at higher risk for CICI complaints through predictive factors. |
Durán-Gómez et al.;2022 | Type of Cancer-Breast | FACT-Cog and Perceived cognitive impairment | Patient History | BCPs | 180 cancer participants | Observational cross-sectional study | Concluded that 41.7% of the BCP reported that they had perceived cognitive impairment . |
Ganz et al.;2013 | Inflammation | Neuropsychological assessments | EDTA tubes and ELISA assay | BCSs | 93 participants | Longitudinal study | Concluded that longitudinal decline in sTNF-RII was significantly associated with the reporting of fewer memory complaints over 12 month. Tumor necrosis factor (TNF) also can play a role in CICI. |
Henneghan et al.; 2019 | Inflammation | Hopkins Verbal Learning Test, Revised, Trail Making Test A & B and Controlled Oral Word Association Test (COWA), | Multiplex assays | BCSs | 65 cancer participants | Cross Sectional Study | Concluded that relationship that concentration of ABeta42, ABeta-40, tau, and cytokines indicated the possibility of CICI complaints in BCSs. |
Henneghan et al.;2020 | Inflammation | Perceived Cognitive Impairments and FACT-Cog v3 | Multiplex assays from EMD Millipore | BCSs | 66 cancer participants | Cross Sectional Study | Concluded that there was a partial correlation of interleukin-2 with fatigue, loneliness, stress, and perceived cognitive impairment. |
Henneghan et al.;2021 | Brain Structure | Rey Auditory Verbal Learning Test (RAVLT, Comprehensive Trail Making Test (CTMT), eControlled Oral Word Association test (COWA) | GE Discovery MR650 3.0 Tesla Whole Body Scanner- MRI | BCPs stages I-IIIa who received any type of treatment | 43 BCPs vs 60 healthy controls | Longitudinal study | Concluded that increased cortical brain aging in middle-age BCPs with chemotherapy treatment is associated with CICI (decreased verbal memory). |
Janelsins et al.; 2017 | Type of Cancer-Breast | FACT-Cog | Patient history | Breast cancer | 581 cancer participants | Multicenter prospective longitudinal study | Concluded that BCS reported significantly greater cognitive difficulties on the FACT-Cog. |
Kesler et al.; 2013 | Inflammation | Multifactorial Memory Questionnaire Ability Scale, and e Hopkins Verbal Learning Test Revised | MRI and high sensitivity multiplexed sandwich immunoassay | Chemotherapy treated BCPs | 42 cancer patients and 35 healthy controls | Cross sectional study | Concluded that the effects of an elevated Interleukin-6 (IL) and TNF-α factor in the BCS group and compared them to the control group. |
Kim et al.;2021 | Type of Cancer-Colorectal | Patient history/ ICD-10 code | Patient history | Colorectal cancer | 95,303 cancer participants | Cohort study | Concluded that chemotherapy did not increase the risk of cognitive impairment in colorectal cancer patients. |
Lange et al,; 2016 | Demographic | FACT-Cog | Sociodemographic measures included age and education level. | Early-stage breast cancer | 119 participants | Prospective Study | Concluded that 49% of patients showed cognitive decline after adjuvant treatment and 65% out of the 49% later went on to develop CICI after treatment was finished. |
Lepage et al.; 2014 | Brain Structure | Pencil-and-paper neuropsychological test battery focused on processing speed, working memory, verbal memory and visual memory. | MRI; 1.5 Tesla Siemens Magnetom Symphony MR scanner | Chemotherapy treated BCPs | 19 BCPs and 19 healthy controls | Prospective study | After one month of treatment, patients show distributed grey matter volume reductions. |
Li et al.; 2020 | Genetics | Mini-mental state examination, Verbal Fluency Test, and Digit Span Test | QIAGEN kit | BCPs | 232 participants | Cross-sectional Study | Concluded that the COMT polymorphism (rs737865) was identified as a potential genetic marker for cognitive impairment activated by chemotherapy. |
Li et al.; 2022 | Genetics | MMSE | 3–5 mL SCA blood tubes and DNA Qiagen Kit | HER2-/+ Breast cancer | 203 HER2- and 118 HER2+ | Not defined | Concluded that patients with the ER-/PR- gene had higher complaints of CICI after chemotherapy. |
Lund et al.;2022 | Demographic | Mini-Cog | G8 score | Pancreatic, colorectal, or prostate cancer | 188 participants | Recap Study | Concluded that patients of older age had a higher risk of CICI. |
Lyon et al.;2016 | Inflammation | Central Nervous System Vital Signs | Assay | Early-stage breast cancer | 75 cancer participants | Longitudinal study | Concluded that as IL-1β and IL-12 levels increased with increasing CICI. |
Miao et al.;2016 | Brain Structure | MoCA used to assess general cognitive function. Stroop Interference Test for executive function. | fMRI | BCSs treated with 36 months post chemotherapy | 23 BCSs and 26 healthyhy controls | Cross-Sectional Study | Changes in connectivity from chemotherapy are pathological and can be related to CICI. |
Ng et al.; 2017 | Inflammation | FACT-Cog | ELISA kit | Early-stage breast cancer | 51 cancer participants | Longitudinal Study | Concluded that BDNF levels decreased over the duration of the study. Furthermore, the decreasing levels of BDNF were associated with an increase in CICI through the assessment of FACT Cog Domains in mental acuity, concentration, memory, verbal functional and multi-tasking. |
Park et al.;2021 | Brain Structure | Wechsler Adult Intelligence Scale Digit span, Wechsler Adult Intelligence Scale Digit symbol, Raven’s Standard progressive Matrices, Rey-Kim Auditory verbal learning test | rs-fMRI using a 3.0T MRI Scanner with a 32-channel head coil. | BCSs with post-6 months | 20 BCSs and 17 healthy counterparts | Cross-Sectional Design | Changes in the cerebellar-cortical connectivity may be a pathophysiological basis for CICI. |
Schroyen et al.;2021 | Inflammation | 9 neuropsychological tests | s [18F] DPA714-PET-MR neuroimaging on a GESIGNA scanner | Breast cancer | 19 chemotherapy treated participants, 18 chemotherapies naive, and 37 controls | Prospective cohort study | Concluded that around 17–22% participants showed cognitive impairment and altered levels of blood markers in chemotherapy treated patients. |
Shaffer et al.;2021 | Type of cancer- breast and colorectal | Telephone Interview for cognitive status assessment | Patient History | Breast and colorectal cancer | 141 breast cancer and 224 colorectal cancers | Prospective Cohort Study | Concluded that there were no differences in the rates of cognitive decline before and after diagnosis for patients who received chemotherapy in adjusted models. |
Tao et al.; 2020 | Brain Structure | MoCA for global normal cognitive function. Digit Span was attention. Stroop Color Word Test and Trail Making Test for executive function. Auditory Verbal Learning Test for memory. | MRI T1 | Premenopausal BCPs who received chemotherapy treatment only. | 35 BCPs and 32 age and sex matched healthy controls | Cross-sectional study | Connectivity was decreased in the cingulated posterior, insular and postcentral regions. |
Torre et al.;2021 | Inflammation | Past medical history from patient charts. | Immunohistochemistry was performed on 5-micron formalin-fixed paraffin-embedded sections | Various types of cancers | 15 chemotherapy participants, 10 cancer patients w/out chemotherapy, and 10 healthy controls | Retrograde study | Concluded that cancer patients treated with chemo had increased staining for oxidative stress and DNA damage. |
Wang et al.; 2016 | Bran structure | FACT-Cog v1 | Siemens 3T Trio MRI scanner and fMRI, T1 image | Solid, non-central nervous system tumor or hematologic malignancy of any stage, with competition of 3 or more cycles within 6 months | 15 cancer survivors and 14 healthy controls | Cross-sectional study | Concluded that decreased recruitment/activation of prefrontal cortex were associated with CICI. |
Van et al.; 2021 | Genetics | FACT-Cog | TaqMan assay and Fluidigm Genotyping | Apolipoprotein E (APOE) e2 genotype in older BCSs | 427 BCSs and 407 nonmatched controls | Multisite Prospective Study | Concluded that the chemotherapy group with the e2 allele had higher scores at baseline and maintained these score differentials compared to those without the e2 allele. |
Van der Willik et al.;2018 | Inflammation | MMSE, Letter-Digit Substitution Test, Word Fluency Test, Stroop Test, Purdue Pegboard Test, and 15-Word Learning Test | COULTER® Ac·T diff2™ Hematology Analyzer | BCSs | 166 cancer participants and 1,344 healthy controls | Cohort Study | Concluded that there was a statistically significant association between higher levels of inflammatory markers and lower general cognitive factor in cancer survivors. |
Vardy et al.; 2014 | Type of Cancer- Colorectal | FACT-Cog | Patient History | Colorectal cancer | 291 cancer participants | Longitudinal study | Concluded that cognitive impairment was noted to be correlated with complex processing speed, attention/working memory, and verbal learning efficiency. |
Yao et al.; 2019 | Genetics | FACT-Cog | Illumina 450k methylation array | Early-stage BCPs | 93 cancer participants and 48 non-cancer controls | Prospective Study | Concluded that chemotherapy alters composition and DNA methylation leading to a higher chance of developing CICI. |
Zeng et al.; 2017 | Type of Cancer- Gynecological | FACT-Cog | Patient history | Gynecological cancer | 150 cancer participants and 130 healthy controls | Cross sectional study | Concluded that the cancer group reported impairments in subjective cognitive functioning, perceived cognitive impairment, and perceived cognitive ability. |
Zhang et al.;2020 | Type of Cancer- Lung | MMSE, MoCA | Patient history | Lung Cancer | 21 post chemotherapy, 27 non chemotherapy lung cancer and 30 healthy controls | Cross sectional study | Concluded that decreased FC from the posterior cingulate cortex to bilateral anterior cingulate cortex in post-chemotherapy patients was positively associated with reduced MoCA scores. |
Zhang et al.;2022 | Type of Cancer- Lung | MMSE, MoCA | Patient history | Lung Cancer | 21 non-small cell lung cancer w/chemo, 25 non-small cell lung cacer w/out chemo | Cross sectional study | Concluded that the chemotherapy group showed higher cerebral blood flow in the left insula, right caudate, right superior occipital gyrus, left superior temporal gyrus, and right middle frontal gyrus. Furthermore, increased CBF in the right MFG was negatively correlated with MoCA scores and memory scores |
Zhao et al.; 2020 | Inflammation | MMSE | 2 mL blood sample into an EDTA tube and ELISAs | Early-stage BCPs | 190 cancer participants | Cross-sectional study | Concluded that an increase in inflammatory markers led to an increase in CICI complaints in BCPs. |
I. Demographics
Six studies demonstrate the relationship between demographic variables and CICI. Demographic categories impact both mental and physical health. The studies in this demographic section involve exploring patient age as a factor for predicting CICI.
Lund et al. used the Mini-Cog assessment to see if sociodemographic information had any effect on the CICI complaints of 188 newly diagnosed pancreatic, colorectal, and prostate cancer patients. They concluded that patients of older age (aOR 0.94 [95% CI, 0.88–0.98]) had higher risk of developing CICI (aOR 0.56 [95% CI, 0.33–0.91]).4
Dos Santos et al. used the Mini-Mental State Examination (MMSE) assessment in older cancer patients (≥ 70 yrs) and concluded they were also at higher risk of CICI complaints through predictive factors (OR = 2.58, p = 0.022). The predictive factors mentioned were old age, malnutrition, and pain.5
Lange et al. followed 119 early-stage breast cancer patients (BCP) who were older than 65 yrs and concluded that 49% of patients showed cognitive decline after adjuvant treatment and 65% out of the 49% later developed CICI after concluding treatment. Furthermore, they observed that older patients (> 70 years old) were more likely to have the most severe decline in cognitive impairment (p = 0.05).6
Kesler et al. looked at 42 chemotherapy-treated breast cancer survivors (BCS) and compared them to 25 healthy female controls by discovering MRI changes to bilateral hippocampal volume. They concluded that the breast cancer group’s higher interleukin-6 (IL) was associated with older age (β = 0.673, p = 0.004; adjusted R2 = 0.414, F = 11.62, p = 0.004), and linked to increased risk of CICI in BCS. More effects of inflammation on CICI are discussed in the Inflammation section.7
Arsale et al. tested the cognitive function of 165 gynecological cancer patients and found that lower Montreal Cognitive Assessment (MoCA) scores (< 22) were associated with older age, non-white race, lower education level, uterine and vulvar cancers, and pain scores ≥ 5 (p < 0.05).8
In contrast to the forementioned studies, Carreira et al. explored the UK Clinical Practice Research Datalink Gold primary care database and created a group of BCS (n = 356) and healthy controls (n = 252) to ascertain CICI complaints. They concluded that socio-demographic variables were present, but the results were not included due to lack of statistical significance.9
Overall, the findings from these included studies suggest that those who are of older age, non-white race, have lower educational levels, and report higher pain scores have a higher likelihood of developing CICI compared to their counterparts.
II. Genetics
Using the established inclusion and exclusion criteria, the authors found six studies exploring genetics and CICI.
Yao et al. concluded that chemotherapy led to an increase in DNA methylation and leukocyte circulation in BCPs, possibly providing an epigenetic link to cognitive impairment in BCPs. This was seen in 4.2% of the CpG sites that underwent genetic changes after chemotherapy treatment (p < 1x10− 7). It was documented that CpG cg 16936953 (p < 1.710− 16) was associated with the greatest cognitive decline in BCPs.10
Chan et al. studied the effects of the DNMTI rs2162560 A allele in a cohort of BCPs. They conclude that the DNMT1 rs2162560 A allele is related to both complaints of self-reported CICI (OR = 0.45, 95% CI: 0.25–0.82, P = 0.01) and functional interference (OR = 0.48, 95% CI: 0.24–0.95, P = 0.03). Interestingly, there was no significant association between DNMT1 rs2162560 and objective CICI.11
Li et al., studied 232 BCPs that were ER-/PR- (n = 113) or ER+/PR+ (n = 119) after receiving chemotherapy, and recorded their cognitive impairment. They also focused on six single-nucleotide polymorphisms (SNPs), including brain-derived neurotrophic factor (BDNF) (rs6265), Apolipoprotein E (APOE) (rs429358, rs7412), and COMT (rs165599, rs4680, rs737865). They concluded that patients with the ER-/PR- gene had more complaints of CICI after chemotherapy (z=-7.831, p < 0.01). Furthermore, the COMT polymorphism rs737865 was identified as a potential genetic marker for cognitive impairment activated by chemotherapy.12
Van Dyk et al., studied the protective effects of the APOE e2 genotype in 427 older BCS compared to 407 healthy controls. They discovered the chemotherapy group with the e2 allele had higher scores at baseline and maintained these scores compared to those without the e2 allele. This protective effect was not seen in the healthy control group. This led to the conclusion that this allele, the APOE e2 genotype, provided a protective benefit on cognition (β = 0.32, 95% CI = 0.00 to 0.65).13
Carroll et al. conducted a cross-sectional sample of 94 women for early-stage breast cancer. They concluded that higher rates of DNA damage and lower telomerase activity was statistically significant to increased CICI (Standardized Coefficients [B] = -0.23 & 0.30, P’s < 0.05). They established that lower telomerase activity led to worse attention and motor speed (B’s = 0.30 & 0.24, P < 0.05).14
Li et al. studied the effect of COMT (rs165599, rs737865, and rs4680) in HER- (n = 103) and Her+ (n = 118) BCPs using peripheral blood samples. Cognitive function was assessed using Event Based Prospective Memory (EBPM) and Time-Based Prospective Memory (TBPM) assessments. The HER2- group survivors’ cognition assessments were lower compared to the HER2 + group (HER2-: z = -7.181, p < 0.01; HER2+: z = -2.205 p < 0.05). The COMT (rs165599) A/A genotype carriers showed decreased risk of developing CICI (dominant model: adjusted, OR = 2.21, CI (95%) = 1.156–4.225, p = 0.016) and scored higher on the TBPM test. Lastly, patients with COMT (rs737865) A/G and G/G genotypes demonstrated mild protective function against CICI.15
Overall, the findings from the included studies suggest that increased leukocyte circulation, the DNMT1 rs2162560 A allele, 11 ER-/PR- genotype, HER- genotype, and increased DNA methylation—especially at the CpG regions—leads to increased complaints of CICI in BCPs and survivors. 12 Furthermore, having the APOE e2 genotype serves as a protective factor against CICI.
III. Inflammation
There are increased levels of inflammatory markers in cancer patients; this persists in survivors' years after treatment. The 10 included studies show the relationship between increased inflammatory markers and worsened cognitive impairment.
Van Der Wilk et al. compared inflammatory markers in 166 BCS to 1,344 cancer-free women as controls. They found lower general cognitive factors among the BCS compared to the healthy controls (mean difference = -0.21; 95% confidence interval (CI) -0.35 to -0.06). Additionally, inflammatory markers were higher among the BCS compared to the controls. They concluded there was a statistically significant association between higher levels of inflammatory markers and lower general cognitive factor in cancer survivors. 16
Henneghan et al. used network analysis to establish the greatest degree of closeness with IL-2, IL-13, and granulocyte macrophage colony stimulating factor (GM-CSF) with perceived cognitive impairment. They found a partial correlation of IL-2 with fatigue, loneliness, stress, and perceived cognitive impairment (rs=-0.27 to 0.37, ps < 0.05). Furthermore, there was significant negative correlation between perceived cognitive impairment and GM-CSF ( p < 0.01) suggesting a protective benefit.17
Lyon et al. conducted a longitudinal study that followed 75 women with early-stage breast cancer before initiating treatment, through 24 months post-treatment initiation. They associated increased IL-1β ( p < 0.01) and IL-12 ( p < 0.01) with impairment of psychomotor speed and verbal memory, indicating CICI development. They observed positive associations with inflammatory markers, such as IL-5 and IL-7, and psychomotor speed, while IL-13 was positively associated with verbal memory.18
As discussed in demographics, Kesler et al. additionally explored the effects of elevated IL-6 and Tumor necrosis factor-α (TNF) in BCS compared against the control group. Through serum levels and MRI imaging, they concluded that BCS had smaller left hippocampal volume (p = 0.04), slightly reduced right hippocampal volume (p = 0.07), and increased IL-6 and TNF-α (β=-2.46, p = 0.006), which linked to significant deficits in memory performance compared to the control group.19
Zhao et al. divided 190 BC patients into a before-chemotherapy group and after-chemotherapy group. They found a correlation between cognition and quality of life in the before-chemotherapy patients (r =-0.790, 0.852, 0.847, and 0.937). They observed another correlation between cognition and cytokine levels (IL-1β, TNF-α, and IL-4) in the before-chemotherapy patients (r = -0.681, -0.572, and − 0.626), which correlates increased inflammatory markers to increased CICI development in BCPs.20
Schroyen et al. studied 74 BCPs for three years and compared them to 37 healthy controls. They concluded 17–22% participants show cognitive impairment and altered levels of blood markers ( p ≤ 0.031) when treated with chemotherapy. Furthermore, neurofilament light-chains increased 20-fold in chemotherapy-treated patients.21
Torre et al. explored oxidative stress and DNA damage in cortical neurons by conducting frontal lobe autopsies in cancer patients who were treated with chemotherapy. They compared the brain tissue between patients treated with chemotherapy (n = 15), cancer patients not treated with chemotherapy (n = 10), and patients without a history of cancer (n = 10) for oxidative stress markers. They observe increased staining for oxidative stress and DNA damage (p < 0.05) in patients treated with chemotherapy.22
Henneghan et al. positive correlated αβ42, αβ40, Tau, cytokines, and CICI complaints to BCS (R2 = 0.71, F = 9.01, p < 0.0001).23
Ganz et al. observed how decreased sTNF-RII was significantly associated with fewer memory complaints over a 12-month period ( p = 0.04) in BCS. This indicated that TNF has a role in CICI developing in BCS.24
Ng et al. conducted a longitudinal study focused on the effects of BDNF on self-perceived CICI for 51 early-stage BCPs. BDNF plays a role in regulating synaptic plasticity, neuronal growth, and survival indicating its importance in cognition. They concluded that BDNF levels decreased over the duration of the study (T1:5423.0 pg/ml; T2:5313.6 pg/ml; T3:4050.3 pg/ml; p < 0.01). Furthermore, the decreasing BDNF was associated with increased CICI via FACT Cog domains: mental acuity (SE=-0.00004,p = 0.669), concentration ( p = 0.032), memory (SE=-0.00003, p = 0.732), verbal fluency (SE = 0.00005, p = 0.511), functional capacity (SE=-0.00006, p = 0.617), and multi-tasking (SE = 0.00009, p = 0.138).25
Overall, the findings from the included studies suggest that inflammation plays a role in determining the presence of CICI in cancer patients and survivors. These studies conclude that increased inflammatory markers, such as IL-2, IL-4, IL-6, IL-12, IL-1β, and TNF-α, and decreased BDNF is associated with increased severity of CICI. Furthermore, decreased sTNF-RII is associated with milder CICI complaints in BCS.
IV. Brain Structure
Changes in brain structure leads to changes in cognitive ability. Some cancer treatments directly treat the brain, while others still cross the blood brain barrier, which physically damages and changes brain structure. The authors accepted nine papers exploring this relationship.
Chen et al. compared volume changes in 16 BCPs and 14 healthy age and sex-matched controls. For the 16 cancer patients, seven were on docetaxel and cyclophosphamide, while nine were on other regimens. They concluded the cancer group has greater volume reduction in the temporal lobe (change=-0.26). The cancer group also had decreased oral reading recognition scores (change=-6.94).26
Park et al. explored resting-state functional connectivity and observed disruptions to functional connectivity in cancer patients with cognitive complaints. They concluded that BCPs with cognitive complaints (n = 36) were negatively correlated with executive function in the functional connectivity of the dorsolateral prefrontal cortex lobule VII (r=-0.38, P = 0.026).27
A study by Lepage et al. observed decreased gray matter volume one month after chemotherapy treatment. Furthermore, participants had changes to the frontal and temporal lobes one year following treatment. This indicated that brain structure changes led to cognitive impairment and patient complaints were valid (p < 0.001).28
Blommaert et al. concluded that white matter volume changes were linked to cognitive impairment in cancer patients, especially those that were older in age. For example, they found older subgroups receiving chemotherapy had greater volume reduction in the right corona radiata and decreased verbal memory (p = 0.49) .29
Henneghan et al. concluded that BCPs showed decreased cortical thickness (p < 0.001) compared to the control group (p = 0.08). Furthermore, greater decreases in the cortical thickness were associated with worse performance on verbal learning tasks (r=-0.48, p < 0.01).30
Tao et al., used the Stroop Interference Test (x̄ =36.04 ± 7.87, r=-0.238, p = 0.088 ), Stroop Word Test (x̄ =21.05 ± 4.70, r=-0.042, p = 0.407), and Trail Making B Test (x̄ =106.71 ± 11.99, r=-0.361, p = 0.018) and correlated CICI and cingulate posterior volume changes in BCPs compared to healthy controls. They concluded that as the volume of the cingulate posterior decreased the more severe CICI was in 35 pre-menopausal women with breast cancer.31
Miao et al. established a relationship between connectivity changes and CICI complaints via exploring how the anterior cingulate cortex was primarily involved in response-related processes and how its function can be evaluated using the Stroop Test. They concluded that significant negative correlation exists between response times in the Stroop Interference Test and fMRI functional connectivity strength in the anterior cingulate cortex in low signal areas, including the left superior frontal gyrus (p = 0.0309) and the left medial frontal gyrus (p = 0.0465) in the BCG compared to the healthy control group.32
Studies that focused on the hippocampus established correlations between hippocampal volume and CICI. Apple et al. observed how the cancer group (n = 15) showed higher hippocampal connectivity in the left cuneus, left lingual, left precuneus, and right middle frontal gyrus compared to healthy controls (n = 18). Multiple linear regression revealed how precuneus connectivity (p = 0.004) in the hippocampus was significantly associated with subjective cognitive concerns for the four identified clusters (p = 0.022, ).33
Apple et al. also discovered that aberrant hippocampal changes in size and activity were seen in other studies exploring CICI. They found that survivors scored significantly lower on the episodic memory subtest of the age-adjusted NIH Toolbox Cognitive Battery (Picture Sequence Memory test) (p = 0.041). Furthermore, the survivor group showed significantly more inward deformation across both hemispheres compared to control (survivors’ x̄ =-0.068mm, ± 0.13; controls’ x̄ =0.026mm ± 0.12; F(1, 31) = 5.76, p = 0.023) and smaller total average hippocampal volumes ( x̄ =2102mm3 ± 192) compared to control ( x̄ =2247mm3 ± 173). This indicates that changes to hippocampal volume in cancer patients plays a role in CICI.34
Overall, the findings of the included studies suggest any change to brain structure and any decrease in brain volume leads to increased severity of CICI in cancer patients. The brain structures reviewed include changes to volume and structure of the temporal lobe, volume of grey and white matter, cortical thickness, anterior and posterior cingulate, and hippocampus; decreases in any of these volumes indicates a higher likelihood of developing CICI.
V. Type of Cancer
Differing cancers are linked to differing severity of cognitive impairment. The authors accepted 21 publications that discuss the relationship between cancer type to impairment severity. This systematic review accepted three breast cancer papers, two lung cancer papers, three colorectal cancer papers, and two gynecological papers.
A. Breast Cancer
Cheng et al measured cognitive impairment using MMSE, EBPM, and TBPM for 40 BCPs treated with chemotherapy against education-matched healthy women for controls. They found overall lower scores for BCPs with significant differences for MMSE ( p < 0.01), verbal fluency ( p < 0.01), and EBPM ( p < 0.01); there was no significant difference on the TBPM ( p > 0.05). This indicated that BCPs do have CICI.35
Duran-Gomez et al. studied the presence of CICI in 180 BCPs that were divided equally into chemotherapy treated groups (CT+) and no treatment groups (CT-). They conclude that 41.7% of the breast cancer group report they had perceived cognitive impairment. The CT + group scored worse on perceived cognitive impairment (X = 50.60 ± 15.64 vs. X = 55.01 ± 12.10; p = 0.005), phonological fluency tasks (X = 41.99 ± 9.52 vs. -=47.03 ± 9.31; p < 0.001) and semantic verbal fluency tasks (X = 33.43 ± 11.0 vs. X = 36.14 ± 10.68; p < 0.001) compared to the CT- group.36
Janelsins et al. discussed the presence of CICI in 583 BCS who received chemotherapy treatment compared to 364 healthy age-matched controls. They find that BCS report significantly greater cognitive difficulties on FACT-Cog (p < 0.001).37
Overall, the findings from these three studies suggest that breast cancer leads to CICI complaints in both BCPs and BCSs.
B. Lung Cancer
Zhang et al. investigated the functional connectivity pattern within the default mode network and its relation to CICI in 21 post-chemotherapy lung cancer patients using fMRI. Results were compared to 27 non-chemotherapy lung cancer patients and 30 healthy controls. They concluded that decreased functional connectivity from the posterior cingulate cortex to the bilateral anterior cingulate cortex in post-chemotherapy patients is positively correlated with reduced MoCA scores (left: p = 0.029; right: p = 0.015).38
Zhang et al. also explored the effect of cerebral blood flow on connectivity deficits in lung cancer patients after chemotherapy. They concluded that the chemotherapy group showed higher cerebral blood flow in the left insula, right caudate, right superior occipital gyrus, left superior temporal gyrus, and right middle frontal gyrus. Furthermore, increased CBF in the right middle frontal gyrus negatively correlated with MoCA and memory scores (p = 0.023).39
Overall, the findings from these two included studies suggest lungs cancer leads to CICI complaints in cancer patients.
C. Colorectal Cancer
Vardy et al. evaluated 291 patients with early-stage colorectal cancer, 72 patients with metastatic disease, and 72 healthy controls. Using the Geriatric Depression Scale, they found 45% of participants with early-stage colorectal cancer have cognitive impairment versus 15% of healthy controls (OR = 4.51, 95% CI 2.28–8.93; P < 0.001). Cognitive impairment was correlated with complex processing speed, attention, working memory, and verbal learning efficiency.40
In contrast to the study discussed above, Kim et al. concluded that chemotherapy did not increase the risk of cognitive impairment in colorectal cancer patients (colon cancer: HR, 0.92; 95% CI 0.83–1.03; rectal cancer: HR, 0.88; 95% CI, 0.75–1.04), while radiotherapy negatively correlated with cognitive impairment in rectal cancer patients (HR, 0.01; 95% CI, 0.84 to 0.99).41
Shaffer et al. also found no difference in the rates of cognitive decline before and after diagnosis for patients who receive chemotherapy in adjusted models (p = 0.86, one sided 95% PI lower bound: 0.09 worse after chemotherapy).42
Overall, the findings from these three studies suggest colorectal cancer may lead to CICI complaints in cancer patients, and that treatment type may affect the risk of developing CICI.
D. Gynecological Cancer
Zeng et al conducted a cross-sectional study that followed 158 participants with gynecological cancer and 130 age-matched non-cancer controls. They concluded that the cancer group reported impairments in subjective cognitive functioning, perceived cognitive impairment, and perceived cognitive ability (all p values < 0.001).43
Similarly, Arsdale et al., tested the cognitive function of 165 gynecological cancer patients and concluded that the mean MoCA score for the entire cohort was 24.1 (range 13–30). There was a trend towards lower cognition scores for women who were treated with both chemotherapy and radiation (p = 0.10).8
Overall, the findings from the two included studies suggest gynecological cancer patients suffer from CICI complaints.