Abstract
BACKGROUND
Both fatigue and cognitive impairment are highly prevalent in individuals with glioma and have been associated with diminished health-related quality of life. Although the two are thought to be related, evidence remains sparse, in part, due to the limited availability of data. Therefore, we aimed to investigate the relationship between fatigue and cognition in adult glioma patients.
MATERIAL AND METHODS
We analyzed cognitive data from adult patients with diffuse glioma at the University of California San Francisco (UCSF, n=100) and Amsterdam UMC (n=127). Comparable cognitive tests were combined by domain: attention, information processing, mental flexibility, semantic fluency, verbal learning, verbal memory, and visual search. Z-scores of ≤ -1.5 SD were considered impaired. Fatigue was assessed with the Fatigue Symptom Inventory (UCSF) and the Checklist Individual Strength (Amsterdam UMC). Linear mixed models were used to assess associations between cognitive domains and fatigue adjusted for site, grade, histology, sex, radiotherapy, and chemotherapy. Latent profile analysis was used to identify distinct fatigue-cognitive phenotypes.
RESULTS
UCSF patients were older (50.3 vs. 44.9 years, p=0.002), more likely to have grade 4 tumors (32% vs. 18%, p=0.025), and more years since diagnosis (2.72 vs 1.27, p<0.001). Interestingly, though UCSF patients were more impaired across nearly all cognitive domains, there was no difference in fatigue between institutions (-0.0457 vs. -0.178, p=0.60). Attention and impaired semantic fluency were associated with fatigue severity in univariate models. Using backward selection, the best-fit model included only impaired semantic fluency (estimated effect = 0.35, p= 0.014, r2=0.03). We identified four distinct cognitive phenotypes in the pooled dataset, which validated across sites, though fatigue severity did not contribute to these phenotypes.
CONCLUSION
Although fatigue and cognitive symptoms can overlap, the association between cognitive impairment and fatigue severity is weak. Interestingly, this important finding was validated across institutions and glioma subtypes. Further study is needed to address questions regarding the longitudinal change of symptoms. Given the limited relational overlap, fatigue and cognitive impairments should both be explicitly assessed and treated with targeted therapies.
SUPPORT/DISCLOSURE
Stichting Anita Veldman Foundation CCA-2019-2-21; Amsterdam University Fund 4056-2022-7-8; CCA travel grant 2022-10-10; Prins Bernhard Cultuurfondsbeurs 40042576/NOD; R.A. Laan fonds 2022-10-24; The loglio Collective; Sheri Sobrato Brisson Brain Cancer Survivorship Program.
Both fatigue and cognitive impairment are highly prevalent in individuals with glioma and have been associated with diminished health-related quality of life. Although the two are thought to be related, evidence remains sparse, in part, due to the limited availability of data. Therefore, we aimed to investigate the relationship between fatigue and cognition in adult glioma patients.
MATERIAL AND METHODS
We analyzed cognitive data from adult patients with diffuse glioma at the University of California San Francisco (UCSF, n=100) and Amsterdam UMC (n=127). Comparable cognitive tests were combined by domain: attention, information processing, mental flexibility, semantic fluency, verbal learning, verbal memory, and visual search. Z-scores of ≤ -1.5 SD were considered impaired. Fatigue was assessed with the Fatigue Symptom Inventory (UCSF) and the Checklist Individual Strength (Amsterdam UMC). Linear mixed models were used to assess associations between cognitive domains and fatigue adjusted for site, grade, histology, sex, radiotherapy, and chemotherapy. Latent profile analysis was used to identify distinct fatigue-cognitive phenotypes.
RESULTS
UCSF patients were older (50.3 vs. 44.9 years, p=0.002), more likely to have grade 4 tumors (32% vs. 18%, p=0.025), and more years since diagnosis (2.72 vs 1.27, p<0.001). Interestingly, though UCSF patients were more impaired across nearly all cognitive domains, there was no difference in fatigue between institutions (-0.0457 vs. -0.178, p=0.60). Attention and impaired semantic fluency were associated with fatigue severity in univariate models. Using backward selection, the best-fit model included only impaired semantic fluency (estimated effect = 0.35, p= 0.014, r2=0.03). We identified four distinct cognitive phenotypes in the pooled dataset, which validated across sites, though fatigue severity did not contribute to these phenotypes.
CONCLUSION
Although fatigue and cognitive symptoms can overlap, the association between cognitive impairment and fatigue severity is weak. Interestingly, this important finding was validated across institutions and glioma subtypes. Further study is needed to address questions regarding the longitudinal change of symptoms. Given the limited relational overlap, fatigue and cognitive impairments should both be explicitly assessed and treated with targeted therapies.
SUPPORT/DISCLOSURE
Stichting Anita Veldman Foundation CCA-2019-2-21; Amsterdam University Fund 4056-2022-7-8; CCA travel grant 2022-10-10; Prins Bernhard Cultuurfondsbeurs 40042576/NOD; R.A. Laan fonds 2022-10-24; The loglio Collective; Sheri Sobrato Brisson Brain Cancer Survivorship Program.
Original language | English |
---|---|
Journal | Neuro-oncology |
DOIs | |
Publication status | Published - 2023 |