To uncover mechanisms underlying chemotherapy-induced cognitive impairment in breast cancer, we studied new biomarkers of neuroinflammation and neuronal survival. This cohort study included 74 women (47 ± 10 years) from 22 October 2017 until 20 August 2020. Nineteen chemotherapy-treated and 18 chemotherapy-naïve patients with breast cancer were assessed one month after the completion of surgery and/or chemotherapy, and 37 healthy controls were included. Assessments included neuropsychological testing, questionnaires, blood sampling for 17 inflammatory and two neuronal survival markers (neurofilament light-chain (NfL), and brain-derived neurotrophic factor (BDNF) and PET-MR neuroimaging. To investigate neuroinflammation, translocator protein (TSPO) [¹⁸F]DPA714-PET-MR was acquired for 15 participants per group, and evaluated by volume of distribution normalized to the cerebellum. Chemotherapy-treated patients showed higher TSPO expression, indicative for neuroinflammation, in the occipital and parietal lobe when compared to healthy controls or chemotherapy-naïve patients. After partial-volume correction, differences with healthy controls persisted (p_FWE < 0.05). Additionally, compared to healthy- or chemotherapy-naïve controls, cognitive impairment (17–22%) and altered levels in blood markers (F ≥ 3.7, p ≤ 0.031) were found in chemotherapy-treated patients. NfL, an axonal damage marker, was particularly sensitive in differentiating groups (F = 105, p = 4.2 × 10 ⁻²¹), with levels 20-fold higher in chemotherapy-treated patients. Lastly, in chemotherapy-treated patients alone, higher local TSPO expression was associated with worse cognitive performance, higher blood levels of BDNF/NfL, and decreased fiber cross-section in the corpus callosum (p_FWE < 0.05). These findings suggest that increased neuroinflammation is associated with chemotherapy-related cognitive impairment in breast cancer. Additionally, NfL could be a useful biomarker to assess neurotoxic effects of anticancer chemotherapies.

As survival rates increase, more emphasis has gone to possible cognitive sequelae in older cancer patients, which could be explained by accelerated brain aging. In this review, we provide a complete overview of studies investigating neuroimaging, neurocognitive, and neurodegenerative disorders in older cancer survivors (>65 years), based on three databases (Pubmed, Web of Science and Medline). Ninety-six studies were included. Evidence was found for functional and structural brain changes (frontal regions, basal ganglia, gray and white matter), compared to healthy controls. Cognitive decline was mainly found in memory functioning. Anti-hormonal treatments were repeatedly associated with cognitive decline (tamoxifen) and sometimes with an increased risk of Alzheimer’s disease (androgen deprivation therapy). Chemotherapy was inconsistently associated with later development of cognitive changes or dementia. Radiotherapy was not associated with cognition in patients with non-central nervous system cancer but can play a role in patients with central nervous system cancer, while neurosurgery seemed to improve their cognition in the short-term. Individual risk factors included cancer subtypes (e.g., brain cancer, hormone-related cancers), treatment (e.g., anti-hormonal therapy, chemotherapy, cranial radiation), genetic predisposition (e.g., APOE, COMT, BDNF), age, comorbidities (e.g., frailty, cognitive reserve), and psychological (e.g., depression, (post-traumatic) distress, sleep, fatigue) and social factors (e.g., loneliness, limited caregiver support, low SES). More research on accelerated aging is required to guide intervention studies.

• Differences observed in all blood marker categories, from on-therapy until years later. • Differences observed in metabolic and blood cell markers, even pre-chemotherapy. • Blood markers were associated with cognition mainly years post-chemotherapy. • Structural brain metrics associated with cognition shortly and years post-treatment.

Due to the rising use of chemotherapy treatment in cancer patients and growing survival rates, therapy-induced neurotoxic side effects are increasingly reported. Given the ambiguity about the prevalence and severity of leukoencephalopathy, one of such toxic side effects, in non-central nervous system (CNS) cancer patients, we performed a systematic literature search using the PubMed/Medline database to summarize existing literature regarding leukoencephalopathy epidemiology in non-CNS cancer patients and its potential cognitive sequelae. The search was based on the following terms: (‘MRI’ OR ‘T2-weighted MRI’ OR ‘FLAIR’) AND (‘cancer’ OR ‘tumour’ OR ‘leukaemia’ OR ‘neoplasms’) AND (‘chemotherapy’ OR ‘radiotherapy’) AND (‘posterior reversible encephalopathy’ OR ‘leukoencephalopathy’ OR ‘cerebral ischaemia’ OR ‘stroke’). Thirty-two studies discussing the occurrence of leukoencephalopathy in cancer patients were included, of which the majority investigated Acute Lymphoblastic Leukaemia (ALL) patients (n = 22). Regularly scanned ALL patients showed a prevalence of leukoencephalopathy between 17 − 87%, and 15 − 83% of patients presented with leukoencephalopathy when only scanned after a CNS event. When diagnosed with posterior reversible encephalopathy syndrome, 100% of patients showed leukoencephalopathy because its diagnosis is based in part on observable lesions. An increased prevalence was observed in ALL patients treated with higher doses of methotrexate (5 g/m² MTX, 42 − 87%) when compared to lower doses (< 5 g/m², 32 − 67%). By contrast, in breast cancer patients, white matter lesions were mainly detected in case of neurological symptoms, but not (yet) clearly associated with chemotherapy administration. However, chemotherapy treatment was associated with more infratentorial microbleeds in breast cancer patients . Up to 50% of other (neurologically asymptomatic) solid tumour patients presented white matter lesions, even years after treatment. When cognitive data were investigated, lesioned patients showed lower scores on neurocognitive tests in 50% of studies, years after ending therapy. In conclusion, leukoencephalopathy is well-documented for ALL patients (with a focus on methotrexate), but there is a lack of knowledge for other intravenous chemotherapeutics, other oncological populations, wider age ranges and possible risk factors (e.g. history of CNS event). Furthermore, the long-term neuropsychological impact and potential risk for neurodegenerative processes due to leukoencephalopathy remains inconclusive. Hence, large international databanks, epidemiological and prospective case-control studies are necessary to stratify risk groups for CNS-related side effects.

Purpose: To investigate the short-term cerebral metabolic effects of intravenous chemotherapy and their association with long-term fatigue/cognitive complaints.Experimental design: Using [¹⁸F]-FDG-PET/CT whole-body scans, we retrospectively quantified relative cerebral glucose metabolism before and after neoadjuvant chemotherapy in a cohort of patients treated for non-metastatic breast cancer (2009-2019). Self-report of cognitive complaints and fatigue were prospectively assessed 7 ± 3 years after therapy. Metabolic changes were estimated with i) robust mixed-effects modelling in regions-of-interest (frontal, parietal, temporal, occipital, and insular cortex) and ii) general-linear modelling of whole-brain voxel-wise outcomes. iii) The association between metabolic changes and self-reported outcomes was evaluated using linear regression-analysis.Results: Of the 667 screened patients, 263 underwent PET/CT before and after chemotherapy and 183 (48 ± 9 years) met the inclusion criteria. After chemotherapy, decreased frontal and increased parietal and insular metabolism were observed (|ß|>0.273, p_FDR0.244, p_FDR≤0.048). Voxel-based analysis (p_cluster-FWE<0.001) showed decreased metabolism in the paracingulate gyrus (-3.2 ± 3.9%) and putamen (3.1 ± 4.1%) and increased metabolism in the lateral cortex (L=2.9 ± 3.1%) and pericentral gyri (3.0 ± 4.4%). Except for the central sulcus, the same regions showed changes in EC, but not in FEC patients. Of the 97 self-reported responders, 23% and 27% experienced extreme fatigue and long-term cognitive complaints, respectively, which were not associated with metabolic changes.Conclusion: Both hyper- and hypometabolism were observed after chemotherapy for breast cancer. Combined with earlier findings, this study could support inflammatory mechanisms resulting in relative hypermetabolism, mainly in the parietal/occipital cortices. As early metabolic changes did not precede long-term complaints, further research is necessary to identify vulnerable patients.